Circadian Regulation: Methods and Protocols (Methods in Molecular Biology, 2482) 107162248X, 9781071622483

Citation preview

Methods in Molecular Biology 2482

Guiomar Solanas · Patrick-Simon Welz Editors

Circadian Regulation Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

Circadian Regulation Methods and Protocols

Edited by

Guiomar Solanas Institute for Research in Biomedicine, Barcelona Institute for Science and Tech, Barcelona, Spain

Patrick-Simon Welz Hospital del Mar Medical Research Institute (IMIM), Barcelona, Spain

Editors Guiomar Solanas Institute for Research in Biomedicine Barcelona Institute for Science and Tech Barcelona, Spain

Patrick-Simon Welz Hospital del Mar Medical Research Institute (IMIM) Barcelona, Spain

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-2248-3 ISBN 978-1-0716-2249-0 (eBook) https://doi.org/10.1007/978-1-0716-2249-0 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022 Chapter 3, 4, 8, and 17 are licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). For further details see license information in the chapter. This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface This book compiles a series of methods on several aspects of circadian research. These methods are thoroughly explained for newcomers to the field and contain some of the latest techniques for expert scientists to be updated. The reader will find protocols covering the wide variety of daily rhythmic processes, using diverse model organisms, for the analysis of circadian rhythms in the SCN and in peripheral organs, describing in vitro systems and in silico methods. Barcelona, Spain

Guiomar Solanas Patrick-Simon Welz

v

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v ix

1 Sleep Under Preindustrial Conditions: What We Can Learn from It . . . . . . . . . . . Leandro Casiraghi and Horacio O. de la Iglesia 2 The Structure-Based Molecular-Docking Screen Against Core Clock Proteins to Identify Small Molecules to Modulate the Circadian Clock. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Seref Gul and Ibrahim Halil Kavakli 3 Analysis of Complex Circadian Time Series Data Using Wavelets . . . . . . . . . . . . . Christoph Schmal, Gregor Mo¨nke, and Adria´n E. Granada 4 Mathematical Modeling in Circadian Rhythmicity . . . . . . . . . . . . . . . . . . . . . . . . . . Marta del Olmo, Saskia Grabe, and Hanspeter Herzel 5 Bioinformatics and Systems Biology of Circadian Rhythms: BIO_CYCLE and CircadiOmics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Muntaha Samad, Forest Agostinelli, and Pierre Baldi 6 Cell-Based Phenotypic Screens to Discover Circadian Clock-Modulating Compounds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Megumi Hatori and Tsuyoshi Hirota 7 Methods for Assessing Circadian Rhythms and Cell Cycle in Intestinal Enteroids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Miri Park, Yuhui Cao, and Christian I. Hong 8 Using ALLIGATORs to Capture Circadian Bioluminescence. . . . . . . . . . . . . . . . . Aiwei Zeng and John S. O’Neill 9 Studying Circadian Clock Entrainment by Hormonal Signals. . . . . . . . . . . . . . . . . Violetta Pilorz, Iwona Olejniczak, and Henrik Oster 10 In Vitro Assays for Measuring Intercellular Coupling Among Peripheral Circadian Oscillators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anna-Marie Finger 11 Circadian Control of Transcriptional and Metabolic Rhythms in Primary Hepatocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sung Kook Chun and Selma Masri 12 Electrophysiology of the Suprachiasmatic Nucleus: Single-Unit Recording . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martha U. Gillette and Jennifer W. Mitchell 13 Anatomical Methods to Study the Suprachiasmatic Nucleus. . . . . . . . . . . . . . . . . . Eric L. Bittman 14 Circadian Analysis of Rodent Locomotor Activity in Home Cages . . . . . . . . . . . . Paul Petrus and Paolo Sassone-Corsi

1

vii

15 35 55

81

95

105 125 137

153

169

181 191 211

viii

15

16

17

18

19

20

21

22

23 24

25

Contents

Recording of Diurnal Gene Expression in Peripheral Organs of Mice Using the RT-Biolumicorder. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Georgia Katsioudi, Alejandro Osorio-Forero, Flore Sinturel, Claudia Hagedorn, Florian Kreppel, Ueli Schibler, and David Gatfield Isolation and Sorting of Epidermal Interfollicular Stem Cells for the Study of Circadian Rhythms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Paloma Sola´ and Valentina M. Zinna Detecting Circadian Rhythms in Human Red Blood Cells by Dielectrophoresis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Andrew D. Beale, Fatima H. Labeed, Stephen J. Kitcatt, and John S. O’Neill Measuring Circadian Neutrophil Infiltration in Tissues by Paired Whole-Mount Tissue Clearing and Flow Cytometry . . . . . . . . . . . . . . . . . . Tommaso Vicanolo, Andres Hidalgo, and Jose M. Adrover In Vivo Imaging of Circadian NET Formation During Lung Injury by Four-Dimensional Intravital Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . Alejandra Aroca-Creville´n, Andres Hidalgo, and Jose M. Adrover Real-Time Measurement of Energy Metabolism Over Circadian Time Using Indirect Calorimetry-Enabled Metabolic Cages . . . . . . . . . . . . . . . . . Kevin B. Koronowski and Paolo Sassone-Corsi Untargeted and Targeted Circadian Metabolomics Using Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) and Flow Injection-Electrospray Ionization-Tandem Mass Spectrometry (FIA-ESI-MS/MS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anna Artati, Cornelia Prehn, Dominik Lutter, and Kenneth Allen Dyar Time-Restricted Feeding and Caloric Restriction: Two Feeding Regimens at the Crossroad of Metabolic and Circadian Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amandine Chaix Chromatin Immunoprecipitation and Circadian Rhythms. . . . . . . . . . . . . . . . . . . . Kenichiro Kinouchi, Kazutoshi Miyashita, and Hiroshi Itoh Fluorescent Reporters for Studying Circadian Rhythms in Drosophila melanogaster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kathyani Parasram, Daniela Bachetti, Vania Carmona-Alcocer, and Phillip Karpowicz Visualization of Mutant Aggregates from Clock Neurons by Agarose Gel Electrophoresis (AGERA) in Drosophila melanogaster . . . . . . . . . Laura Delfino, Susanna Campesan, Giorgio Fedele, Edward W. Green, Flaviano Giorgini, Charalambos P. Kyriacou, and Ezio Rosato

217

243

255

265

285

301

311

329 341

353

373

Contents

26

27

ix

Methods for Delivery of dsRNAi Against Canonical Clock Genes and Immunocytodetection of Clock Proteins in Crustacea . . . . . . . . . . . . . 385 David C. Wilcockson, Lin Zhang, and Charalambos P. Kyriacou In Vivo Bioluminescence Analyses of Circadian Rhythms in Arabidopsis thaliana Using a Microplate Luminometer . . . . . . . . . . . . . . . . . . . . . . 395 Masaaki Okada and Paloma Mas

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

407

Contributors JOSE M. ADROVER • Area of Cell and Developmental Biology, Fundaci
on Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain; Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA FOREST AGOSTINELLI • Institute for Genomics and Bioinformatics, University of California, Irvine, CA, USA; Department of Computer Science and Engineering, University of South Carolina, Columbia, SC, USA ALEJANDRA AROCA-CREVILLE´N • Area of Cell and Developmental Biology, Fundaci
on Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain ANNA ARTATI • Metabolomics and Proteomics Core Facility, Helmholtz Zentrum Mu¨nchen, German Research Center for Environmental Health (GmbH), Neuherberg, Germany DANIELA BACHETTI • Department of Biomedical Sciences, University of Windsor, Windsor, ON, Canada PIERRE BALDI • Department of Computer Science, University of California Irvine, Irvine, CA, USA; Institute for Genomics and Bioinformatics, University of California, Irvine, CA, USA ANDREW D. BEALE • MRC Laboratory of Molecular Biology, Cambridge, UK ERIC L. BITTMAN • Department of Biology and Program in Neuroscience & Behavior, University of Massachusetts, Amherst, MA, USA SUSANNA CAMPESAN • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK YUHUI CAO • Department of Pharmacology and Systems Physiology, University of Cincinnati College of Medicine, Cincinnati, OH, USA VANIA CARMONA-ALCOCER • Department of Biomedical Sciences, University of Windsor, Windsor, ON, Canada LEANDRO CASIRAGHI • Department of Biology, University of Washington, Seattle, WA, USA AMANDINE CHAIX • Department of Nutrition and Integrative Physiology, University of Utah, Salt Lake City, UT, USA SUNG KOOK CHUN • Department of Biological Chemistry, Center for Epigenetics and Metabolism, Chao Family Comprehensive Cancer Center, University of California, Irvine (UCI), Irvine, CA, USA LAURA DELFINO • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK KENNETH ALLEN DYAR • German Center for Diabetes Research (DZD), Neuherberg, Germany; Metabolic Physiology, Institute for Diabetes and Cancer (IDC), Helmholtz Diabetes Center, Helmholtz Zentrum Mu¨nchen, Neuherberg, Germany GIORGIO FEDELE • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK ANNA-MARIE FINGER • Charite´ Universit€ a tsmedizin Berlin, Berlin, Germany; Freie Universit€ at Berlin, Berlin, Germany; Humboldt-Universit€ a t zu Berlin, Berlin, Germany; Laboratory of Chronobiology, Berlin Institute of Health, Berlin, Germany DAVID GATFIELD • Center for Integrative Genomics, University of Lausanne, Lausanne, Switzerland

xi

xii

Contributors

MARTHA U. GILLETTE • Department of Cell and Developmental Biology and Neuroscience Program, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, USA FLAVIANO GIORGINI • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK SASKIA GRABE • Institute for Theoretical Biology, Charite and Humboldt Universit€ a t zu Berlin, Berlin, Germany ADRIA´N E. GRANADA • Charite´ Comprehensive Cancer Center, Charite´ Universit€ atsmedizin Berlin, Berlin, Germany; German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ), Heidelberg, Germany EDWARD W. GREEN • German Cancer Research Center (DKFZ), Heidelberg, Germany SEREF GUL • Chemical and Biological Engineering, Koc¸ University, Istanbul, Turkey CLAUDIA HAGEDORN • Biochemistry and Molecular Medicine, Center for Biomedical Education and Research (ZBAF), Faculty of Health, Witten/Herdecke University, Witten, Germany MEGUMI HATORI • Institute of Transformative Bio-Molecules, Nagoya University, Nagoya, Japan HANSPETER HERZEL • Institute for Theoretical Biology, Charite and Humboldt Universit€ at zu Berlin, Berlin, Germany ANDRES HIDALGO • Area of Cell and Developmental Biology, Fundaci
on Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain TSUYOSHI HIROTA • Institute of Transformative Bio-Molecules, Nagoya University, Nagoya, Japan CHRISTIAN I. HONG • Department of Pharmacology and Systems Physiology, University of Cincinnati College of Medicine, Cincinnati, OH, USA HORACIO O. DE LA IGLESIA • Department of Biology, University of Washington, Seattle, WA, USA HIROSHI ITOH • Department of Endocrinology, Metabolism, and Nephrology, Keio University School of Medicine, Tokyo, Japan PHILLIP KARPOWICZ • Department of Biomedical Sciences, University of Windsor, Windsor, ON, Canada GEORGIA KATSIOUDI • Center for Integrative Genomics, University of Lausanne, Lausanne, Switzerland IBRAHIM HALIL KAVAKLI • Chemical and Biological Engineering, Koc¸ University, Istanbul, Turkey; Molecular Biology and Genetics, Koc¸ University, Istanbul, Turkey KENICHIRO KINOUCHI • Department of Endocrinology, Metabolism, and Nephrology, Keio University School of Medicine, Tokyo, Japan STEPHEN J. KITCATT • Department of Mechanical Engineering Sciences, University of Surrey, Surrey, UK KEVIN B. KORONOWSKI • Center for Epigenetics and Metabolism, U1233 INSERM, Department of Biological Chemistry, University of California, Irvine, CA, USA FLORIAN KREPPEL • Biochemistry and Molecular Medicine, Center for Biomedical Education and Research (ZBAF), Faculty of Health, Witten/Herdecke University, Witten, Germany CHARALAMBOS P. KYRIACOU • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK FATIMA H. LABEED • Department of Mechanical Engineering Sciences, University of Surrey, Surrey, UK

Contributors

xiii

DOMINIK LUTTER • Computational Discovery Research, Institute for Diabetes and Obesity (IDO), Helmholtz Diabetes Center, Helmholtz Zentrum Mu¨nchen, Neuherberg, Germany; German Center for Diabetes Research (DZD), Neuherberg, Germany PALOMA MAS • Centre for Research in Agricultural Genomics (CRAG), CSIC-IRTA-UABUB, Barcelona, Spain; Consejo Superior de Investigaciones Cientı´ficas (CSIC), Barcelona, Spain SELMA MASRI • Department of Biological Chemistry, Center for Epigenetics and Metabolism, Chao Family Comprehensive Cancer Center, University of California, Irvine (UCI), Irvine, CA, USA JENNIFER W. MITCHELL • Department of Cell and Developmental Biology and Neuroscience Program, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, USA KAZUTOSHI MIYASHITA • Department of Endocrinology, Metabolism, and Nephrology, Keio University School of Medicine, Tokyo, Japan GREGOR MO¨NKE • European Molecular Biology Laboratory, Heidelberg, Germany JOHN S. O’NEILL • MRC Laboratory of Molecular Biology, Cambridge, UK MASAAKI OKADA • Centre for Research in Agricultural Genomics (CRAG), CSIC-IRTAUAB-UB, Barcelona, Spain IWONA OLEJNICZAK • Institute of Neurobiology, Center of Brain, Behavior & Metabolism, University of Lu¨beck, Lu¨beck, Germany MARTA DEL OLMO • Institute for Theoretical Biology, Charite and Humboldt Universit€ at zu Berlin, Berlin, Germany ALEJANDRO OSORIO-FORERO • Department of Fundamental Neurosciences, University of Lausanne, Lausanne, Switzerland HENRIK OSTER • Institute of Neurobiology, Center of Brain, Behavior & Metabolism, University of Lu¨beck, Lu¨beck, Germany KATHYANI PARASRAM • Department of Biomedical Sciences, University of Windsor, Windsor, ON, Canada MIRI PARK • Department of Pharmacology and Systems Physiology, University of Cincinnati College of Medicine, Cincinnati, OH, USA PAUL PETRUS • Center for Epigenetics and Metabolism, U1233 INSERM, Department of Biological Chemistry, University of California, Irvine, CA, USA VIOLETTA PILORZ • Institute of Neurobiology, Center of Brain, Behavior & Metabolism, University of Lu¨beck, Lu¨beck, Germany CORNELIA PREHN • Metabolomics and Proteomics Core Facility, Helmholtz Zentrum Mu¨nchen, German Research Center for Environmental Health (GmbH), Neuherberg, Germany EZIO ROSATO • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK MUNTAHA SAMAD • Department of Computer Science, University of California Irvine, Irvine, CA, USA; Institute for Genomics and Bioinformatics, University of California, Irvine, CA, USA UELI SCHIBLER • Department of Molecular Biology, Faculty of Sciences, University of Geneva, Geneva, Switzerland CHRISTOPH SCHMAL • Institute for Theoretical Biology, Humboldt Universit€ a t zu Berlin, Berlin, Germany FLORE SINTUREL • Division of Endocrinology, Diabetes, Nutrition and Patient Education, Department of Medicine, University Hospital of Geneva, Geneva, Switzerland;

xiv

Contributors

Department of Cell Physiology and Metabolism, Faculty of Medicine, University of Geneva, Geneva, Switzerland PALOMA SOLA´ • Institute for Research in Biomedicine (IRB Barcelona), Barcelona Institute of Science and Technology, Barcelona, Spain TOMMASO VICANOLO • Area of Cell and Developmental Biology, Fundaci
on Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain DAVID C. WILCOCKSON • Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Aberystwyth, UK AIWEI ZENG • MRC Laboratory of Molecular Biology, Cambridge, UK LIN ZHANG • Department of Genetics and Genome Biology, University of Leicester, Leicester, UK VALENTINA M. ZINNA • Institute for Research in Biomedicine (IRB Barcelona), Barcelona Institute of Science and Technology, Barcelona, Spain

Chapter 1 Sleep Under Preindustrial Conditions: What We Can Learn from It Leandro Casiraghi and Horacio O. de la Iglesia Abstract Human sleep is regulated by light in two fundamental ways: The light-dark (LD) cycle entrains a circadian clock that in turn regulates sleep timing, and light per se can acutely inhibit sleep. Throughout evolution, these sleep regulatory systems became highly sensitive to the effects of light and they can be affected by the relatively low light intensities that are used indoors. Thus, postindustrial living conditions have created built environments that have isolated humans from the natural LD cycle and exposed them to an artificial one that can affect daily sleep timing. Studying indigenous communities that have differential access to electricity, as well as communities living in highly urbanized areas, we and others have shown that human access to artificial light has delayed the daily onset of sleep but has had a smaller effect on its offset, leading to an overall reduction in sleep duration that is pervasive in modern societies. In this chapter we discuss these studies, highlight their main findings, and point to their limitations. Key words Sleep, Artificial light, Indigenous communities, Preindustrial societies, Modern societies

1

The Conquest of the Night Life on earth evolved under the inescapable selective pressure of the alternation between day and night. Human evolution was no exception and collection of food by hunter-gatherer communities was limited to the daytime while the night was a time to seek refuge in a safe sleeping place. The transition to agricultural communities created not only more sustainable food resources but also built environments in which humans could shield themselves from predators, extreme weather, and other dangers associated with a nomadic life. These built environments demanded the use of artificial light, which became a key resource particularly during the long winter nights of higher latitudes. However, it was not until the industrial revolution, and particularly until the use of electric light became more widespread, that artificial light eliminated natural daylight as a condition for human

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_1, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

1

2

Leandro Casiraghi and Horacio O. de la Iglesia

activity. The second half of the twentieth century saw an exponential increase in electricity production and use, and a concomitant exponential decrease in its cost, brightly lighting the indoor environment irrespective of the solar day [1]. Labor, social life, and the acquisition and transmission of knowledge could suddenly continue beyond the daytime, transforming nighttime from a time for rest and sleep to a time during which activity could continue at will. Natural dusk, which was the inescapable call to the end of daily activity, was replaced by electric dusk, the timing of which could be controlled freely. This manipulation of light and the associated conquest of the night came at a cost, and a major consequence has been the progressive change in the daily timing of sleep as humans transitioned from agricultural, to industrial, to post-industrial and highly urbanized communities. Whether and how human sleep duration changed in the last century remains a matter of debate; however, before the night was artificially lit, it would have been hard to imagine the levels of activity and social interaction that are now typical of any urban community after dusk. A quick visual representation of the sleep and light exposure patterns in a rural community without access to electricity and a highly urbanized community clearly reveals the impact of the access to electric light (Fig. 1).

Fig. 1 The access to electric light delays sleep timing. Clock plots displaying light exposure and actimetrybased sleep timing. The plot on the left is from participants in a Toba/Qom community that has no access to electricity. The plot on the right is from participants living in a highly urbanized community (Seattle, WA, USA) and sampled during the weekend. Green bars represent the interquartile range for the first and last times that individuals are exposed to a light intensity of 50 lux. Purple and blue bars represent the times of sleep onset and offset, respectively. Note that in both cases sleep starts after the last exposure to 50 lux, but this occurs much later relative to solar time in the highly urbanized community, in which the sleep onset occurs after solar midnight. Gray and white in the pie represent the natural night and day, respectively. Time represents the local clock time

Sleep Under Preindustrial Conditions: What We Can Learn from It

2

3

Artificial Light and Its Impact on Sleep The 24-h cycle of day and night represented an evolutionary selective pressure for the temporal distribution of physiological and behavioral outputs. For some species it became adaptive to forage and feed during the night whereas for others this was the time to sleep. Although biological systems could have simply responded to 24-h cyclic events in the environment with specific behavioral and physiological outcomes, it was likely more adaptive to be able to predict these events before their arrival. As a consequence, circadian clocks, namely biological mechanisms that oscillate autonomously with a ~24-h period, were selected and are now a defining feature of living systems. The main advantage of an endogenous clock is its ability to predict cyclic events in the environment; for instance, animals could seek insulation from the cold night before they experienced low nocturnal temperatures. Importantly, circadian clocks have periods that are not exactly 24 h and therefore need to be entrained by environmental cycles. In any given habitat, cycles that more reliably relay solar time are more likely to be selected as proximate environmental factors to entrain the clock, and for virtually all species the light-dark (LD) is the most prominent entraining cycle or zeitgeber. A consequence of evolutionary selection of the LD cycle as the principal zeitgeber is that circadian clocks and the biological rhythms they sustain are extremely sensitive to the influence of light. The human circadian system is no exception to this, and its master circadian clock can be entrained by relatively low-intensity light [2]. Humans, however, are different from other species in their unique ability to control their own LD cycle, and this artificial LD cycle can impact circadian rhythms, including the sleep-wake cycle, in many ways. First, light can directly affect the phase of circadian rhythms through direct retinal projections to the suprachiasmatic nucleus (SCN), which houses the master circadian clock of mammals [3]. Light has an effect on the clock’s phase that depends on the time at which the light stimulus reaches the clock [4]. Light during the end of the night and early morning advances the phase of the clock whereas light during the end of the evening and beginning of the night delays the clock. Typically, humans use artificial light to extend the end of the daytime instead of advancing its beginning. In other words, most humans with access to electricity who can choose their timing of activity will keep the lights on many hours after dusk (up to 8 h in high latitudes during the winter) but will rarely turn the lights on many hours before dawn. This asymmetric use of artificial light leads to a net delay—relative to the natural LD cycle—in the circadian clock and the rhythms it sustains, leading to a later daily timing of sleep.

4

Leandro Casiraghi and Horacio O. de la Iglesia

Among the circadian rhythms that are consistently delayed is the nocturnal release of melatonin, a hallmark of the biological night. Melatonin is released by the pineal gland, and the circadian timing of its release is regulated by the SCN. Under normal conditions, the onset of melatonin occurs a few hours before bedtime, and it represents an important cue to prepare the brain for the onset of sleep. However, the signal is delayed by exposure to evening artificial light. Second, light can have acute effects on behavioral and physiological rhythms. Brightly lit environments are particularly stimulating to humans. This is the consequence of the inhibitory effect that light can have on sleep but also of the fact that light enables humans to engage in different activities—TV watching, reading, social interactions—that further inhibit sleep. Interestingly, light also inhibits the nocturnal release of melatonin. The pineal releases melatonin in a circadian fashion, but light also has the ability to acutely reduce melatonin synthesis and release. Thus, light in the late evening and early night will not only delay the clock but will also abolish the release of the hormone that marks the beginning of the night. The acute effects of light on sleep and melatonin release push sleep onset to a later time in the night. Of note, these effects of light have been exacerbated by the use of screens that have become more widespread in the last two decades. A great part of human evening activity is spent watching TV screens, monitors, or portable devices, which are not only highly stimulating but themselves emit light that is sufficient to elicit physiological effects [5]. A third and typically more ignored effect of artificial light is that it allows humans to spend more time indoors. Most humans in urbanized communities spend their working or school days within artificially lit environments. Although these environments are appropriately lit for indoor activities, they expose humans to light levels that are typically an order of magnitude lower than natural daylight. This shielding from natural daylight reduces the ability of natural light to entrain circadian rhythms, which are therefore more amenable to be entrained by mistimed artificial light. All these effects of artificial light act together to delay human sleep onset relative to what the sleep onset would be under natural light conditions. This has been shown both by field and intervention studies that have compared human sleep in the absence and presence of artificial sleep. Notably, humans in highly urbanized communities still wake up around sunrise to go to work or school, and it is reasonable to predict that a trend to later sleep start without delaying its end should shorten daily sleep. However, whether daily sleep duration has historically diminished is still a matter of controversy, one we may never fully elucidate. After all, objective measures of sleep have only been available many decades after electric light became popular. Nevertheless, it is clear that a large proportion of humans living in modern cities sleep less than

Sleep Under Preindustrial Conditions: What We Can Learn from It

5

what is considered necessary to sustain good health, and this shorter than needed sleep would be much harder to achieve without access to electric light.

3

Methods to Study the Effects of Electric Light on Human Sleep The study of human circadian rhythms, including the sleep-wake cycle, has been approached through a wide range of methods. Isolation studies to demonstrate the endogenous nature of circadian rhythms were first done in natural caves and later in artificial isolation labs, which allowed for the removal of light cues, social interactions, and other environmental variables that could provide information about external time [6, 7]. Initially, these studies suggested that the human circadian system was more sensitive to social cues than environmental light cues, but later studies clearly established that light could act as a strong zeitgeber in humans [8, 9]. Furthermore, studies under carefully controlled laboratory conditions were critical in demonstrating that electric light was effective in resetting circadian rhythms even at relatively low intensities used in indoor environments [2]. These studies, however, could not assess the direct impact of electric light on human circadian rhythms under field conditions. The fact that a single light pulse of a given intensity can lead to a reliable phase shift of a given rhythm in the lab does not mean the same pulse will elicit the same shift in subjects living their normal life and exposed to other sources of light throughout the day. Field experiments are more appropriate to determine the real impact of artificial light on real-life circadian physiology. However, they obviously present limitations in terms of the physiological variables that can be measured. These variables not only provide a more holistic view of circadian outputs but in some cases, they are critical to elucidate the mechanisms behind the changes in the timing of behavioral rhythms. As described above, the same delay on the onset of sleep could result from an acute effect of evening light on behavior or from an effect of evening light on the phase of the circadian clock. In either case, the behavioral expression of the effect of light is the same, i.e., a later sleep start, but the mechanism behind each case is radically different. To distinguish between them, a more reliable marker of circadian phase such as the dim-light melatonin onset (DLMO) is needed [10]. The DLMO can be determined through serial blood or saliva samples. However, a limitation is that the samples need to be taken under dim-light, which is sometimes difficult or not possible under field conditions. The ideal field experiment to assess the impact of electric light on human circadian rhythms would be a longitudinal experiment in which a group of subjects living in the same place gain or lose access to electricity and their rhythms are monitored before and after the

6

Leandro Casiraghi and Horacio O. de la Iglesia

introduction of electric light. Such a scenario is difficult to come by, but a similar longitudinal approach was used by Wright and colleagues [11]. The team monitored their own actimetry-based sleep and DLMO in their typical urban setting during the summer, and later monitored the same rhythms during a camping trip in which they used no artificial sources of light. They found that in the presence of electricity sleep timing was delayed, and this delay was consistent with a later daily melatonin rhythm release, clearly indicating a direct effect of electric light on the phase of the circadian system. Although this study did not show a shortening of daily sleep duration, the team performed a similar study during the winter and found that sleep onset was advanced during winter camping without a change in sleep offset, leading to a lengthening of daily sleep duration [12]. In most field scenarios longitudinal studies are not feasible, and investigators have gravitated toward communities with limited or no access to electricity to assess sleep timing. Actimetry represents the gold standard to determine sleep timing in field experiments and in some groups has been shown to be much more accurate than diary-based self-assessment of sleep [13]. Although wrist actimetry has been validated with polysomnographic recordings as a reliable estimate of sleep timing, it does not allow for the determination of sleep stages [14–16]. Thus, actimetry alone provides little information on the physiological basis of changes in sleep parameters, but it does provide an accurate reflection of how the timing of sleep changes under different environmental conditions. A major limitation of field studies of communities living under preindustrial conditions is the lack of appropriate control groups. It is not hard to find communities still living under rural conditions and even without access to electricity, but it is usually much harder to find communities that can serve as controls. Ideally, communities that are ethnically and culturally similar but do have access to electricity or live in more urbanized communities should be used as comparison groups.

4

Lessons from Sleep Under Preindustrial Conditions Until recently, studies assessing sleep timing in rural communities and, more specifically, communities without access to electricity were based on diaries on which participants self-assessed their times of sleep onset and offsets. Although sleep diaries can provide an accurate reflection of sleep timing under well-controlled conditions, they are less reliable in long-term studies where investigators cannot assure daily compliance, and in communities with low literacy levels. Furthermore, they may be particularly less informative in communities in which activities follow no strict clock time.

Sleep Under Preindustrial Conditions: What We Can Learn from It

7

Nevertheless, early as well as more recent studies based on data from diaries and interviews have provided valuable information about sleep in different populations around the world (Table 1). Worthman and Brown [17] studied Egyptian villagers and found that sleeping arrangements, and particularly cosleeping, which was highly prevalent in this population, affected the timing and quality of sleep. Beckinschtein and collaborators [18] followed a native Mapuche community in the Argentinian Patagonia and described a clear seasonal difference in the timing of sleep which was also tied to a seasonal migratory pattern. Louzada and Menna-Barreto [19] compared the sleep habits of adolescents in a rural area in Southern Brazil and found that those with access to electricity had later slept times, and that the presence of a TV in their house correlated with later bedtimes. Even in studies that use actimetry, diaries and interviews provide invaluable contextual information. For instance, Yetish et al. [20] report that neither the San people in Namibia nor the Tsimane people in Bolivia have a word for “insomnia”—a pretty telling observation. Most importantly, these studies as a whole underscore that we cannot simply frame sleep studies on whether people live under preindustrial conditions and have no access to electric light; sleep can be affected vastly by factors including geography and weather, cultural traits, familial composition, and socioeconomic status. The amenability of wrist activity monitors provided a noninvasive, objective, and quantitative method to determine the time of sleep onset and offset with high accuracy. These devices can also measure light exposure as well as other physiological variables and have become an invaluable tool for field sleep studies in remote areas. In the last few years, multiple research teams have used wrist activity as a powerful tool to objectively measure sleep timing, and in many cases, to demystify assumptions about sleep under preindustrial conditions. Indeed, studies have nothing but proved that there is no normative sleep timing. A first pioneering study using this technology showed sleep patterns in members of a small community living without electricity in a small island near mainland New Guinea [21]. Although the study lacked a control group (urban, with electricity), the authors made truly interesting findings. For instance, whereas wake-up time was highly synchronous among the study participants, their sleep onset was much more variable, both within and between individuals. This result contrasts with the very low variability in both sleep onset and offset that we found in the Toba/Qom people living without electricity (Fig. 1; [22]). Similarly, whereas the access to electric light is typically associated with a delayed sleep onset, it was shown to shorten sleep in some communities but not others (Table 1). Furthermore, although fragmented sleep and chronotype variability seem to be the norm and may represent an adaptive trait in Hadza hunter-

Actimetry, diaries, DLMO

Diaries, actimetry, DLMO

Controlled, access to electric Amazonian Amazonian light vs. no access rubber tappers Extractive Reserve, Brazil

Formosa, Argentina

Toba/Qom

Controlled, different levels of access to electric light and urbanity

Observational, population with Actimetry no access to electricity only

Fondwa, Haiti

Haitian rural villagers

Diaries

Observational

Egyptian villagers El Cairo, Egypt

Questionnaires, actimetry, DLMO

Paraná State, Brazil

Rural village adolescents

Controlled, access to electric light vs. no access; different school times

Observational, population with Sleep diaries no access to electricity only

Neuque´n, Argentina

Mapuche

Methodology

Observational, population with Actimetry no access to electricity only

Type of study

Tauwema Village, Papu´a Nueva Guinea

Region

Trobriand islanders

Population/ community

[18]

[21]

References

[36]

[17, 35]

Access to electricity and urbanization correlated with delayed and shorter sleep.

[26, 37, 38]

[22, 24, 25] Access to electricity correlated with delayed sleep onset and reduced sleep duration. Larger differences between electricity levels in the winter. Electric light associated with delayed onset of melatonin release during the winter.

Sleep fragmentation decreased with age. Longer times of staying-in-bed than in industrialized societies, but similar effective sleep duration.

Cosleeping correlated with less sleep but of better quality. High occurrence of daytime napping.

[19, 33, 34] Adolescents with no access to electric light slept more and showed no sleep deprivation. School times correlated with sleep duration and onset. Access to electric light correlated with delayed phase of sleep.

Wake-up times are delayed in the winter.

Young infants displayed ultradian sleep patterns and slept more. Wives in couples slept more than their husbands.

Main findings

Table 1 Summary of studies classified by the population studied and the research team or teams. When more than one work is referenced for the same population and team, methods and main findings are summarized indistinctly for the sake of simplicity. Rows are ordered chronologically

8 Leandro Casiraghi and Horacio O. de la Iglesia

a

Observational, population with Actimetry, PSG no access to electric light only Controlled, urban vs. rural conditions

Mandena, Madagascar

Zambe´zia, Mozambique

Southern Brazil

Tanna Island, Vanuatu

Milange and Tengua communities

Quilombolas

Indigenous Melanesians

Yetish et al. [20] compares these three preindustrial populations

Controlled, access to electric light vs. no access

Controlled, different levels of access to electric light and urbanity

Short sleep duration compared to industrial societies. Communal “sentinel” pattern of sleep.

Shorter and later sleep in the summer. Low incidence of insomnia.

Shorter and later sleep in the summer. Low incidence of insomnia.

[27]

[28]

[42]

[40, 41]

[20, 23, 39]

[20]

[20]

Access to electricity associated with sleep onset [29] and reduces sleep duration and efficiency.

Access to electricity and urbanization correlated with delayed and shorter sleep.

Actimetry, MCT Questionnaire Actimetry

Urbanity was associated with delayed sleep onset. No differences in sleep duration.

Actimetry

High prevalence of “segmented” night sleep and daytime napping. Short sleep duration compared to industrial societies.

Actimetry, DLMO, One of the earliest chronotypes in literature. questionnaires Women slept more than men.

Malagasy rural villagers

Observational

Baependi, Brazil

Rural villagers

Actimetry, interviews

Observational, mixeda

Northern Tanzania

Hadza people

Actimetry, interviews

Mixeda, communities with no access to electric light

Beni Department, Bolivia

Tsimane people

Actimetry, interviews

Mixeda, communities with no access to electric light

Northeastern Namibia

San people

Sleep Under Preindustrial Conditions: What We Can Learn from It 9

10

Leandro Casiraghi and Horacio O. de la Iglesia

gatherers, >8 h consolidated sleep with no chronotype variation and no naps appears to be the norm for the Toba/Qom people [22, 23]. Beyond the temporal distribution of sleep, total daily sleep duration in preindustrial communities—under the assumption that it could represent some indicator of a minimal need for daily sleep—has received particular attention. Although there is nothing normative about preindustrial sleep (see Subheading 5) it is important to highlight that total daily sleep is equally variable among communities living in rural areas without electricity. Several studies confirmed the long-held intuition that access to electricity delays sleep onset, even when involving communities that differ ethnically, geographically, and culturally (Table 1): Toba/Qom communities in the north of Argentina [22, 24, 25], Amazonian workers [26], Quilombola natives in Brazil [27], Milange and Tengua communities in Mozambique [28], or indigenous Melanesians in Vanuatu [29]. Importantly, the delay in sleep onset associated with access to electric light shortened sleep duration in some of these communities but not in others, suggesting that changes in sleep duration may be associated with different trade-offs in different communities. This is further underscored by the wide variability in sleep duration under preindustrial conditions, which can range from 6.25 h (e.g., the Tsimane people in Bolivia) to 8.3 h (the Toba/ Qom people in Argentina) ([20, 25, 30]; Fig. 1). Interestingly, even when the total sleep times vary widely among these groups, they all show seasonal differences, with a strikingly similar lengthening of sleep during the winter, although the mechanisms behind this longer winter sleep may differ. The access to electricity is not only associated with the amenability of electric light but also with the access to other light-emitting devices like TVs and screens from computers, tablets, and cell phones. Furthermore, electric light allows for evening social activity, which together with these devices promotes a later bedtime. This has raised the question of whether the delayed sleep associated with access to electricity can directly be attributed to light or depends on other factors associated with urban settings. By assessing the DLMO in communities with and without electricity, we and others have shown that, at least in part, the delayed sleep timing is a consequence of a delayed circadian clock [24, 26].

5

Limitations of Preindustrial Sleep Studies It is tempting to interpret sleep patterns found in communities living in preindustrial conditions as a proxy of human ancestral sleep. However, in the same manner in which no extant species can be seen as basal or ancestral, sleep patterns we observe today, regardless of the environmental conditions in which they occur, do

Sleep Under Preindustrial Conditions: What We Can Learn from It

11

not necessarily reflect how humans slept historically or prehistorically. It is important to keep in mind that over 300,000 years of evolution separate modern humans from the first humans that appear in Africa [31]. Several authors have also suggested sleep patterns under non-industrial conditions are normative for sleep in humans living in highly urban environments. The rationale is that sleep under more “natural” conditions should reflect how much and what kind of sleep we need under modern living conditions. This premise is flawed for several reasons. First, there is really not such a thing as “natural sleeping habitat.” The multidimensional nature of sleep function determines that each environment demands a specific timing and quality of sleep. In the same way that different preindustrial environments demand different sleep patterns, highly urbanized environments likely demand drastically different sleep behavior. In other words, the natural sleeping environment for a person living in New York City is New York City and not some remote place isolated from urban development. Second, sleep timing and the manifestation of specific sleep stages are not simply a consequence of the sleep needs that a specific environment demands but also of the sleep that environment permits. In other words, members of a community living under preindustrial conditions could also be subjected to chronic or acute sleep deprivation because of environmental constraints. Third, the possibility exists that original ethnic communities from different continents differ in sleep genetic traits [32]. It is conceivable that as Homo sapiens migrated from Africa to Australia, Europe, Asia, and eventually the Americas, specific environments selected for specific genetically determined sleep traits. For instance, the Americas are believed to have been colonized by migrations that reached the new continent through the Bering Strait about 15,000 years ago, when the temperatures were considerably lower than now. This environment likely demanded sleep patterns that differed from those on the African continent. Furthermore, human colonization of new continents could have created evolutionary bottlenecks for genetic sleep traits, which could reduce the adaptability of sleep patterns to new environments.

6

Conclusions The last two decades have seen a marked increase in studies that have capitalized on the access to communities living in a range of preindustrial conditions—from hunting-gathering to agricultural peoples—to gain insight into sleep patterns in the absence of urbanization. These studies have provided unequivocal evidence that urbanization, and its associated increased exposure to electric light, delays daily sleep timing and, in many cases, shortens its

12

Leandro Casiraghi and Horacio O. de la Iglesia

duration. These effects on sleep parameters are, in some cases, mediated in part by a delay on the circadian system. Collectively, these studies have confirmed that sleep under preindustrial conditions can vary tremendously both in duration and temporal distribution, corroborating the plasticity of sleep as a behavioral trait and underscoring that sleep under low levels of urbanization should not be taken as a normative sleep for other human communities.

Acknowledgments We are thankful to Isabelle Hua and Gideon Dunster for their help with Fig. 1, and to Dr. Ray Sanchez for comments on the original manuscript. Supported by NSF RAPID award #1743364 and Leakey Foundation grant #1266 to HOD. References 1. Tillman DA (2018) Coal-fired electricity and emissions control: efficiency and effectiveness. Elsevier Science & Technology, Saint Louis 2. Boivin DB, Duffy JF, Kronauer RE, Czeisler CA (1996) Dose-response relationships for resetting of human circadian clock by light. Nature 379(6565):540–542. https://doi. org/10.1038/379540a0 3. Hughes S, Jagannath A, Hankins MW, Foster RG, Peirson SN (2015) Photic regulation of clock systems. Methods Enzymol 552:125– 143. https://doi.org/10.1016/bs.mie.2014. 10.018 4. Johnson CH, Elliott JA, Foster R (2003) Entrainment of circadian programs. Chronobiol Int 20(5):741–774 5. Chang AM, Aeschbach D, Duffy JF, Czeisler CA (2015) Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness. Proc Natl Acad Sci U S A 112(4):1232–1237. https://doi.org/ 10.1073/pnas.1418490112 6. Kleitman N (1939) Sleep and wakefulness as alternating phases in the cycle of existence. The University of Chicago Press, Chicago 7. Aschoff J (1965) Circadian rhythms in man. Science 148:1427–1432 8. Aschoff J, Fatranska M, Giedke H, Doerr P, Stamm D, Wisser H (1971) Human circadian rhythms in continuous darkness: entrainment by social cues. Science 171(967):213–215 9. Czeisler CA, Kronauer RE, Allan JS, Duffy JF, Jewett ME, Brown EN, Ronda JM (1989) Bright light induction of strong (type 0)

resetting of the human circadian pacemaker. Science 244(4910):1328–1333 10. Klerman EB, Gershengorn HB, Duffy JF, Kronauer RE (2002) Comparisons of the variability of three markers of the human circadian pacemaker. J Biol Rhythm 17(2):181–193 11. Wright KP Jr, McHill AW, Birks BR, Griffin BR, Rusterholz T, Chinoy ED (2013) Entrainment of the human circadian clock to the natural light-dark cycle. Curr Biol 23(16): 1554–1558. https://doi.org/10.1016/j.cub. 2013.06.039 12. Stothard ER, McHill AW, Depner CM, Birks BR, Moehlman TM, Ritchie HK, Guzzetti JR, Chinoy ED, LeBourgeois MK, Axelsson J, Wright KP Jr (2017) Circadian entrainment to the natural light-dark cycle across seasons and the weekend. Curr Biol 27(4):508–513. https://doi.org/10.1016/j.cub.2016.12.041 13. Dunster GP, de la Iglesia L, Ben-Hamo M, Nave C, Fleischer JG, Panda S, de la Iglesia HO (2018) Sleepmore in Seattle: later school start times are associated with more sleep and better performance in high school students. Sci Adv 4(12):eaau6200. https://doi.org/10. 1126/sciadv.aau6200 14. van de Wouw E, Evenhuis HM, Echteld MA (2013) Comparison of two types of Actiwatch with polysomnography in older adults with intellectual disability: a pilot study. J Intellect Develop Disabil 38(3):265–273. https://doi. org/10.3109/13668250.2013.816274 15. Rupp TL, Balkin TJ (2011) Comparison of Motionlogger Watch and Actiwatch actigraphs

Sleep Under Preindustrial Conditions: What We Can Learn from It to polysomnography for sleep/wake estimation in healthy young adults. Behav Res Methods 43(4):1152–1160. https://doi.org/10. 3758/s13428-011-0098-4 16. Kushida CA, Chang A, Gadkary C, Guilleminault C, Carrillo O, Dement WC (2001) Comparison of actigraphic, polysomnographic, and subjective assessment of sleep parameters in sleep-disordered patients. Sleep Med 2(5):389–396. https://doi.org/10. 1016/s1389-9457(00)00098-8 17. Worthman CM, Brown RA (2007) Companionable sleep: social regulation of sleep and cosleeping in Egyptian families. J Fam Psychol 21(1):124–135. https://doi.org/10.1037/ 0893-3200.21.1.124 18. Bekinschtein TA, Negro A, Goldı´n A, Fernández MP, Rosenbaum S, Golombek DA (2004) Seasonality in a Mapuche native population. Biol Rhythm Res 35(1–2):145–152 19. Louzada F, Menna-Barreto L (2004) Sleep– wake cycle in rural populations. Biol Rhythm Res 35(1–2):153–157. https://doi.org/10. 1080/09291010412331313304 20. Yetish G, Kaplan H, Gurven M, Wood B, Pontzer H, Manger PR, Wilson C, McGregor R, Siegel JM (2015) Natural sleep and its seasonal variations in three pre-industrial societies. Curr Biol 25(21): 2862–2868. https://doi.org/10.1016/j.cub. 2015.09.046 21. Siegmund R, Tittel M, Schiefenho¨vel W (1998) Activity monitoring of the inhabitants in Tauwema, a traditional Melanesian Village: rest/activity behaviour of Trobriand Islanders (Papua New Guinea). Biol Rhythm Res 29(1): 49–59. https://doi.org/10.1076/brhm.29.1. 49.3045 22. de la Iglesia HO, Fernandez-Duque E, Golombek DA, Lanza N, Duffy JF, Czeisler CA, Valeggia CR (2015) Access to electric light is associated with shorter sleep duration in a traditionally hunter-gatherer community. J Biol Rhythm 30(4):342–350. https://doi.org/10. 1177/0748730415590702 23. Samson DR, Crittenden AN, Mabulla IA, Mabulla AZP, Nunn CL (2017) Chronotype variation drives night-time sentinel-like behaviour in hunter-gatherers. Proc Biol Sci 284(1858). https://doi.org/10.1098/rspb. 2017.0967 24. Casiraghi LP, Plano SA, Fernandez-Duque E, Valeggia C, Golombek DA, de la Iglesia HO (2020) Access to electric light is associated with delays of the dim-light melatonin onset in a traditionally hunter-gatherer Toba/Qom community. J Pineal Res 2020:e12689. https:// doi.org/10.1111/jpi.12689

13

25. Casiraghi L, Spiousas I, Dunster GP, McGlothlen K, Fernández-Duque E, Valeggia C, de la Iglesia HO (2021) Moonstruck sleep: synchronization of human sleep with the moon cycle under field conditions. Sci Adv 7(5):eabe0465. https://doi.org/10. 1126/sciadv.abe0465 26. Moreno CR, Vasconcelos S, Marqueze EC, Lowden A, Middleton B, Fischer FM, Louzada FM, Skene DJ (2015) Sleep patterns in Amazon rubber tappers with and without electric light at home. Sci Rep 5:14074. https://doi. org/10.1038/srep14074 27. Pilz LK, Levandovski R, Oliveira MAB, Hidalgo MP, Roenneberg T (2018) Sleep and light exposure across different levels of urbanisation in Brazilian communities. Sci Rep 8(1): 11389. https://doi.org/10.1038/s41598018-29494-4 28. Beale AD, Pedrazzoli M, Goncalves B, Beijamini F, Duarte NE, Egan KJ, Knutson KL, Schantz MV, Roden LC (2017) Comparison between an African town and a neighbouring village shows delayed, but not decreased, sleep during the early stages of urbanisation. Sci Rep 7(1):5697. https://doi.org/10. 1038/s41598-017-05712-3 29. Smit AN, Broesch T, Siegel JM, Mistlberger RE (2019) Sleep timing and duration in indigenous villages with and without electric lighting on Tanna Island, Vanuatu. Sci Rep 9(1):17278. https://doi.org/10.1038/ s41598-019-53635-y 30. Yetish G, McGregor R (2019) Chapter 21 Hunter-gatherer sleep and novel human sleep adaptations. In: Dringenberg HC (ed) Handbook of behavioral neuroscience, vol 30. Elsevier, Amsterdam, pp 317–331. https://d oi.org/10.1016 /B97 8-0-12813743-7.00021-9 31. Hublin JJ, Ben-Ncer A, Bailey SE, Freidline SE, Neubauer S, Skinner MM, Bergmann I, Le Cabec A, Benazzi S, Harvati K, Gunz P (2017) New fossils from Jebel Irhoud, Morocco and the pan-African origin of Homo sapiens. Nature 546(7657):289–292. https:// doi.org/10.1038/nature22336 32. Nadkarni NA, Weale ME, von Schantz M, Thomas MG (2005) Evolution of a length polymorphism in the human PER3 gene, a component of the circadian system. J Biol Rhythm 20(6):490–499 33. Peixoto CA, da Silva AG, Carskadon MA, Louzada FM (2009) Adolescents living in homes without electric lighting have earlier sleep times. Behav Sleep Med 7(2):73–80. https:// doi.org/10.1080/15402000902762311

14

Leandro Casiraghi and Horacio O. de la Iglesia

34. Pereira E´F, Louzada FM, Moreno CR (2010) Not all adolescents are sleep deprived: a study of rural populations. Sleep Biol Rhythms 8(4): 267–273. https://doi.org/10.1111/j. 1479-8425.2010.00458.x 35. Worthman CM, Brown RA (2012) Sleep budgets in a globalizing world: biocultural interactions influence sleep sufficiency among Egyptian families. Soc Sci Med. https://doi. org/10.1016/j.socscimed.2012.03.048 36. Knutson KL (2014) Sleep duration, quality, and timing and their associations with age in a community without electricity in Haiti. Am J Hum Biol 26(1):80–86. https://doi.org/10. 1002/ajhb.22481 37. Martins AJ, Vasconcelos SP, Skene DJ, Lowden A, de Castro Moreno CR (2016) Effects of physical activity at work and lifestyle on sleep in workers from an Amazonian Extractivist Reserve. Sleep Sci 9(4):289–294. https://doi.org/10.1016/j.slsci.2016.10.001 38. Martins AJ, Isherwood CM, Vasconcelos SP, Lowden A, Skene DJ, Moreno CRC (2020) The effect of urbanization on sleep, sleep/ wake routine, and metabolic health of residents in the Amazon region of Brazil. Chronobiol Int 37(9–10):1335–1343. https://doi.org/10. 1080/07420528.2020.1802287

39. Crittenden AN, Samson DR, Herlosky KN, Mabulla IA, Mabulla AZP, McKenna JJ (2018) Infant co-sleeping patterns and maternal sleep quality among Hadza huntergatherers. Sleep Health 4(6):527–534. https://doi.org/10.1016/j.sleh.2018.10.005 40. Beijamini F, Knutson KL, Lorenzi-Filho G, Egan KJ, Taporoski TP, De Paula LK, Negrao AB, Horimoto AR, Duarte NE, Vallada H, Krieger JE, Pedrazzoli M, Pereira AC, von Schantz M (2016) Timing and quality of sleep in a rural Brazilian family-based cohort, the Baependi Heart Study. Sci Rep 6:39283. https://doi.org/10.1038/srep39283 41. Ruiz FS, Beijamini F, Beale AD, Goncalves B, Vartanian D, Taporoski TP, Middleton B, Krieger JE, Vallada H, Arendt J, Pereira AC, Knutson KL, Pedrazzoli M, von Schantz M (2020) Early chronotype with advanced activity rhythms and dim light melatonin onset in a rural population. J Pineal Res 69(3):e12675. https://doi.org/10.1111/jpi.12675 42. Samson DR, Manus MB, Krystal AD, Fakir E, Yu JJ, Nunn CL (2017) Segmented sleep in a nonelectric, small-scale agricultural society in Madagascar. Am J Hum Biol 29(4):e22979. https://doi.org/10.1002/ajhb.22979

Chapter 2 The Structure-Based Molecular-Docking Screen Against Core Clock Proteins to Identify Small Molecules to Modulate the Circadian Clock Seref Gul and Ibrahim Halil Kavakli Abstract Circadian rhythms are part of the body’s clock, which regulates several physiological and biochemical variables according to the 24-h cycle. Ample evidence indicated disturbance of the circadian clock leads to an increased susceptibility to several diseases. Therefore, a great effort has been made to find small molecules that regulate circadian rhythm by high-throughput methods. Having crystal structures of core clock proteins, makes them amenable to structure-based drug design studies. Here, we describe virtual screening methods that can be utilized for the identification of small molecules regulating the activity of core clock protein Cryptochrome 1. Key words Circadian rhythm, Virtual screening, Cryptochrome, Docking, Drug discovery

1

Introduction The circadian rhythms are endogenous autonomous oscillators, which enable organisms to adapt their physiological variables to 24-h day/night cycles [1]. They are generated by a mechanism known as the circadian clock, which modulates memory, blood pressure, immune responses, etc. by a periodic transcriptional regulation [2–4]. The cellular clockwork is generated by positive and negative transcriptional and translational feedback loops [5, 6]. In mammals, the positive loop consists of two core clock proteins, BMAL1 and CLOCK. These proteins form heterodimers and bind to E-BOX elements (CACGTG) in the Period (Per), Cryptochrome (Cry) and clock control genes. PER and CRY along with casein kinase Iε (CKIε) form trimeric structure and translocate to the nucleus where it represses BMAL1/CLOCK-driven transcription. Upon phosphorylation, CRYs are ubiquitinated by E3 ubiquitin

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_2, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

15

16

Seref Gul and Ibrahim Halil Kavakli

ligases, e.g., FBXL3 and FBXL21, and directed to the proteasome for degradation. FBXL3 and FBXL21 bind to the primary (FAD binding) pocket of CRYs and act antagonistically and differentially on their stability in the cytosol and the nucleus [7, 8]. This core clock machinery drives rhythmic output called clock-controlled genes in almost every tissue. Indeed, 50% of detected metabolites are under circadian control in mouse liver [9] and nearly 43% of all protein coding genes are under the control of the circadian clock in at least one tissue [10]. A recent study highlights the importance of a post-translational switch that modulates the activity of the cystathionine ß-synthase by CRY1 to regulate amino acid metabolism [11]. Epidemiological and genetic studies have indicated that circadian disturbances are associated with increased risks of different types of diseases. A defect in PER2 produces familial advanced sleep phase syndrome [12, 13]; an analogous mutation causes the same phenotype in mice [14]. People with a causal mutation in CSNK1D and an associated variant in CSNK1E display advanced sleep phase syndrome and delayed sleep phase syndrome, respectively [15, 16]. A human CLOCK variant is associated with diurnal sleep preference [17]. Circadian core clock genes are also associated with a host of neurological disorders including schizophrenia [18– 21], unipolar major depression [22], and bipolar disorder [23, 24]. Another study leads to an identification of an SNP in DEC2 in short sleeper patients [25]. A human gain-of-function CRY1 variant (exon 11 skipping mutation in C-tail of CRY1), and another CRY1 mutant (exon 6 skipping mutation in PHR of CRY1) were found in people suffering from familial delayed sleep phase disorder and attention deficit/hyperactivity disorder [26, 27]. The importance of a robust circadian clock for health is increasingly recognized, and therefore the identification of molecules that modulate the circadian clock became a hot topic. The high-throughput screening was instrumental for identifying the molecules affecting the circadian clock [28–31]. However, that approach is expensive and slow, to find specific molecules for core clock proteins. To identify a molecule library that regulates the core clock proteins (PER, BMAL1, CLOCK, and CRY) is currently a scientific challenge. With the recent reports of resolved crystal structures of core clock proteins and their interacting partners (CRY-FBXL3 (pdbID: 4K0R), BMAL1-CLOCK (pdbID: 4F3L), and CRY-PER (pdbID: 4U8H)), now it is possible to perform in silico analysis against them. A recent study with in silico methods was able to identify a molecule (CLK8) to target the CLOCK protein [32]. Further in vivo and in vitro studies revealed that CLK8 regulates CLOCK/BMAL1 entry into the nucleus and regulates the circadian rhythm.

Structure-Based Drug Discovery for Clock Proteins

2

17

Materials All of the programs listed below can be executed in Linux-based computers. Researchers who want to use these procedures should have at least a beginner level of Linux computer usage.

2.1 Homology Modeling Server

If the protein does not have the crystal structure, homology modeling tools should be used to construct 3D structure. Similarly, missing atoms of the crystal structures can be added with the same approach. Freely available and user-friendly homology modeling servers can be used: RaptorX (http://raptorx.uchicago.edu/ StructPredV2/predict/), Swiss-Model (https://swissmodel. expasy.org/interactive), and Phyre (http://www.sbg.bio.ic.ac.uk/ ~phyre2/html/page.cgi?id¼index).

2.2 Programs to Visualize Protein Structures

Pymol, Chimera, and VMD programs can be used to visualize and manipulate protein structures. Manual and tutorial of these programs are available online to perform various operations such as visualization, mutation, rotamer selection, and structure preparation for molecular dynamics (MD) simulations. Chimera and VMD are free of charge for academic users with registration requirements through their websites (https://www.cgl.ucsf.edu/chimera/down load.html, https://www.ks.uiuc.edu/Development/Download/ download.cgi?PackageName¼VMD). Pymol offers an education version for free and an alternatively open-source version can be installed (https://pymol.org/dokuwiki/?id¼installation). The usage of these programs will be mentioned in relevant sections.

2.3 Retrieving Protein Structures from Protein Data Bank (PDB)

Crystal structures of CRY1 (PDB ID: 4K0R, 5T5X, or 6KX4) can be downloaded from the protein data bank [33–35]. 4K0R is the first resolved photolyase homology region (PHR) of mouse CRY1 structure with 2.65 Å in 2013 [33]. Higher-resolution PHR domain of CRY1 structures with 1.84 and 2.00 Å were reported recently [34, 35].

2.4 NAMD Program and CHARMM Force Field

NAMD software is required to run MD simulations of protein. Depending on the computer power, various versions of NAMD have been developed and can be freely downloaded by academic users (https://www.ks.uiuc.edu/Development/Download/down load.cgi?PackageName¼NAMD). To simulate biological macromolecules, we need to calculate forces between atoms within a molecule and between molecules. Thus, to describe these forces, set of parameters that are generated after extensive benchmarking studies, named force fields, are used. Ample high-quality force fields are available in the literature and can be readily used with NAMD. Since we used the CHARMM force field in our studies we

18

Seref Gul and Ibrahim Halil Kavakli

suggest using the latest version of CHARMM (http://www. charmm-gui.org/?doc¼toppar, https://www.charmm.org/ charmm/resources/charmm-force-fields/#charmm). 2.5

Small Molecules

2.6 Autodock Vina and Autodock Tools

3 3.1

Commercially available small molecule (SM) libraries are available in sdf format with 2D structural information. Since 3D structural information and partial charges of atoms in molecules are required for docking simulations, sdf format should be converted to pdbqt format. Autodock Tools (http://autodock.scripps.edu/resources/adt) and Autodock Vina (http://vina.scripps.edu/download.html) are used to prepare SMs and dock them to the receptor, respectively. Openbabel can convert sdf format to pdb format, then Autodock Tools is used to get pdbqt format of molecules from pdb files. In addition to SM conversion, Autodock Tools is also used to prepare the pdbqt format of the receptor (protein) and to determine the coordinates of the docking site on the receptor. Vina is used for docking simulations of SMs to the target pocket of receptors. At the end of the docking simulations, Vina generates the binding mode of the molecule in the pocket and calculates Vina binding energy between molecule and receptor. The Autodock Tools program is used to visualize and analyze docking results.

Methods MD Simulation

3.1.1 Homology Modeling of CRY1

Two approaches are utilized for docking simulations in the literature: (a) docking molecules to the crystal structure of the receptor and (b) docking molecules to the structure of the receptor equilibrated under physiological conditions. We are not going to compare the advantages and disadvantages of these approaches in this chapter, but simply we prefer to use equilibrated receptor structures in our docking simulations. Before starting, generate a new folder named “structurebased,” download and save the below-mentioned files into this folder. Since all crystal structures of CRY1 have missing atoms, homology modeling should be performed. Swiss-Model server can be used to get the full-length CRY1-PHR structure. The protein sequence of mouse CRY1 can be retrieved from NCBI data bank. 1. Go to the website of NCBI (https://www.ncbi.nlm.nih.gov/), select “Protein” from the drop-down menu and search for “AAD39548.1”. 2. Click “FASTA” to get the fasta format of the CRY1 amino acid sequence.

Structure-Based Drug Discovery for Clock Proteins

19

3. Copy and paste sequence of CRY1 to SWISS-MODEL server (https://swissmodel.expasy.org/interactive) (Fig. 1a) (see Note 1). 4. Clicking “Search For Templates” will search for the crystal structures having the highest identity and similarity to the query sequence. 5. Click 5t5x that has the highest resolution among available structures (Fig. 1b) (see Note 2). 6. Let’s choose 5t5x that has the highest coverage, identity, and resolution among available structures. Then click “Build Models.” Download the PDB file by clicking Model > PDB Format > save as CRY1.pdb (Fig. 1c) (see Note 3). 3.1.2 Preparing CRY1 for MD Simulation

1. To assign protonation states of residues by using PROPKA server (http://server.poissonboltzmann.org/pdb2pqr) (see Note 4). (a) Click “Upload a PDB file” to upload CRY1.PDB. (b) Select “Use PROPKA to assign protonation states at provided pH.” (c) Forcefield to use: CHARMM. (d) Output naming scheme to use: CHARMM. (e) In “Additional Options” check the following: “Ensure that new atoms are not rebuilt too close to existing atoms,” “Optimize the hydrogen bonding network,” and “Add/keep chain IDs in the PQR file.” (f) Download output “jobid.pqr”. Rename file as CRY1_p. pqr. (g) Open CRY1_p.pqr with PyMol, first open Pymol, then File > Open > (browse molecule) > Open File menu > Save molecule > check “one file” and “global” and click OK > CRY1_p.pdb (into the “structurebased” folder) 2. To solvate protein and add an appropriate concentration of salt we are going to use VMD. (a) Download top_all36_prot.rtf into the “structurebased” folder from http://www.charmm-gui.org/?doc¼toppar (see Note 5). (b) First, we need to generate psf file. Generate a text file “psfgen.in” and type commands given in Subheading 4. Open VMD in the same directory with CRY1_p.pdb and the topology file (in “structurebased” folder). In the “VMD Main” panel

20

Seref Gul and Ibrahim Halil Kavakli

Fig. 1 Generating homology model of CRY1 using SWISS-MODEL server. (a) Query page, (b) template selection page, (c) model page

Structure-Based Drug Discovery for Clock Proteins

21

Fig. 2 Snapshot from VMD session generated after CRY1 solvation

Extension menu > Tk console: source psfgen.in (see Note 6). (c) Protein will be solvated and NaCl will be added. Similar to the previous step generate a text file “solvation.in” and type the following to Tk console: resetpsf mol delete all source solvation.in

At the end of this step ionized.pdb and ionized.psf files will be produced (see Note 7) (Fig. 2). 3. Next, the dimension of system should be measured and generate pdb file with restrained Cα atoms that is required for running the MD simulations. (a) Type the following on the Tk console (without deleting anything from the previous step): set everyatom [atomselect top all] measure minmax $everyatom (see Note 8)

(b) In MD simulation the system should be gradually heated. Thus, Cα atoms of all residues are restrained and let side chains of amino acids to move by sequential heating. At

22

Seref Gul and Ibrahim Halil Kavakli

this step, we are going to set up a pdb file (CA_rest.pdb) containing restrained Cα atoms. Start a new VMD session: File > New Molecule > Browse > ionized.psf. Right click on ionized.psf (on VMD Main) > Load Data onto Molecule > Browse > ionized. pdb > OK > Load Type the following to Tk console: source restrain.in (see Note 9). 3.1.3 Running MD Simulation

Before running the simulation, minimization of the system should be performed. Then gradually heat and equilibrate the system. Finally, we are going to use coordinates, velocity, and extended system configuration files of minimized, heated, and equilibrated system to run the MD simulation. 1. Following parameter files should be downloaded to working directory (into the “structurebased” folder): par_all36_prot. prm, par_all36_lipid.prm, par_all36_carb.prm, par_all36_cgenff.prm from (http://www.charmm-gui.org/? doc¼toppar) and toppar_water_ions_namd.str from (http:// mackerell.umar yland.edu/~kenno/cgenf f/program. php#namd). The last file is the edited version of toppar_water_ions.str to be able to use it in NAMD. 2. Make four folders and name them as Minimized, Heat, Equilibrium, and Production. Next move parameter files given in supplementary materials minimization.in, heating.in, equilibrium.in, production.in to corresponding folders. 3. Go to “Minimize” folder and run minimize.in via NAMD by typing the following: ~/namd/namd2 +p8 minimize.in > CRY1_minimize.log (see Note 10)

4. After successful completion of the minimization, go to “Heating” folder and run the following: ~/namd/namd2 +p8 heating.in > CRY1_heating.log (see Note 11)

Upon completion of the heating, go to “Equilibrium” folder and run the following: ~/namd/namd2 +p8 equilibrium.in > CRY1_equilibrium.log

Upon completion of the equilibration, go to “Production” folder and run the following: ~/namd/namd2 +p8 production.in > CRY1_production.log

Structure-Based Drug Discovery for Clock Proteins

23

5. The equilibriated frame of the production run can be used for docking simulations. To extract the last frame, go to “Production” file and run VMD then: File > New Molecule > Browse > ionized. psf > Load > Browse > CRY1_production.dcd > Load Then select (click on) ionized.psf on the VMD Main panel and File > Save Coordinates From the Save Trajectory panel type protein into Selected atoms section, Type the last step in the “First” and “Last” sections > Save > CRY1.pdb 6. Now we have CRY1 structure equilibrated under physiological conditions and ready to be used for the docking. Docking

Determining the target pocket, preparing molecules and receptor (s) in pdbqt format are crucial steps in docking studies. Below we are going to describe these steps for CRY1. Before going into details generate a folder named “docking.” Move CRY1.pdb file to the docking folder.

3.2.1 Preparation of CRY1 for Docking Simulations

1. Open Autodock Tools program from the terminal, e.g., ~/ MGLTools-1.5.6/MGLToolsPckgs/AutodockTools then File > Rad Molecule > CRY1.pdb Grid > Macromolecule > Choose > CRY1 > Select molecule New panel should appear: type CRY1.pdbqt > Save (see Note 12).

3.2

2. A grid center point and box size should be determined for docking simulations. Since the primary pocket of CRY1 is targeted, coordinates of the side chain terminal carbon atom of Arg358 (CZ in pdb file) can be used as grid center. Then, fine tune the center by using the “Grid Options” panel. To visualize the target pocket, the grid box panel should be used as follows: Grid > Grid Box > number of points in x-dimension: 18; y-dimension: 20; z-dimension: 20; x center: 11.7; y center: 35.7; z center: 16.4 (Fig. 3) (see Note 13). 3. Target pocket size and center information determined in step 2 will be provided in the configuration file that includes the following keywords: receptor, out, center_x, center_y, center_z, size_x, size_y, size_z. Although these parameters are self-explanatory, options that can be specified in AutodockVina are explained in notes (see Note 14). 4. A sample configuration file named as “cry1_conf.txt” is provided below:

24

Seref Gul and Ibrahim Halil Kavakli

Fig. 3 CRY1.pdbqt is visualized using AutodockTools. Grid box is colored in red-green-blue in x, y, and z directions, respectively receptor = CRY1.pdbqt center_x = -11.7 center_y = 35.7 center_z = -16.4 size_x = 18.0 size_y = 20.0 size_z = 20.0

3.2.2 Preparation of Small Molecules for Docking Simulations

Small molecule libraries from different companies are freely available and can be downloaded from their websites. One of the commonly used libraries is ZINC library that has many tranches classified according to properties of molecules such as molecular weight and LogP. In addition to these physical properties, molecules are under predefined categories, e.g., fragments, lead-like, and drug-like. 1. Go to website: https://zinc.docking.org/tranches/home/

Structure-Based Drug Discovery for Clock Proteins

25

Then click a tranch to download. On the new page click # symbol on the top of the page and select the format of molecules. We suggest downloading the sdf format. 2. Rename the downloaded file to CRY1_sm.sdf. CRY1_sm.sdf includes all molecules in one file. To extract molecules individually we are going to use the openbabel program. Type the following to the terminal: obabel CRY1_sm.sdf -O sm.sdf -m (see Note 15)

3. Molecules are in two-dimensional form. Openbabel (http:// openbabel.org/wiki/Main_Page) will generate threedimensional molecules with only polar hydrogen atoms in the pdb format. For this type the following to terminal: obabel sm1.sdf -O sm1.pdb --gen3D -d –AddPolarH

4. Next, to convert the pdb file to pdbqt format which includes the charge information we are going to use the prepare_ligand4 module of AutoDockTools. Type the following to terminal ~/MGLTools-1.5.6./MGLToolsPckgs/AutoDockTools/Utilities24/ prepare_ligand4.py -l sm1.pdb

This command is going to produce the pdbqt file with the same name of the input pdb (see Note 16). 3.2.3 Running and Analyzing AutoDock Vina Simulations

Our receptor (CRY1.pdbqt) and ligands (sm1.pdbqt and others) are ready for docking simulations. We are going to use AutodockVina software to run simulations. This program does not need to be installed. Once the program is downloaded, giving a path to the “vina” executable will be sufficient to run the program. “cry1_conf. txt” was given in step 4 of Subheading 3.2.1. If all SMs and CRY1. pdbqt is in the same folder, 1. Type following to the terminal: ~/autodock_vina_1_1_2_linux_x86/bin/vina --config cry1_conf. txt

--ligand

sm1.pdbqt

--out

cry1_sm1_out.pdbqt

--log

cry1_sm1_log.txt

2. Vina Binding Energy is printed on the screen. Also, output and log file with the name given after --out and --log options, respectively, are generated (see Note 17).

26

Seref Gul and Ibrahim Halil Kavakli

3. Docking simulations are completed now. Vina binding energies can be sorted by using a python script available in the vina manual page (http://vina.scripps.edu/manual.html) (vina_screen_get_top.py). On the terminal, script can be run as the following to get the top ten molecules: ./vina_screen_get_top.py 10 (see Note 18)

4. We know our best molecules now. We are going to use AutoDock Tools to analyze the results. Open AutodockTools File > Read Molecule > CRY1.pdbqt > Open Analyze > Docking > Open AutoDock vina result > (select the best molecule) > Open Select “Single molecule with multiple conformations” on the “Load MODEL as” panel > OK. To visualize the interacting residues: Analyze > Macromolecules > Choose > CRY1 > Select Macromolecule Analyze > Docking > Show interactions (Fig. 4) (see Note 19)

Fig. 4 AutoDock Tools snapshot from analysis of docking results

Structure-Based Drug Discovery for Clock Proteins

4

27

Notes 1. Providing “Project Title” will be beneficial for future analysis since this query will be saved in the server with this entry. Some searches for finding template(s) may take a couple of hours or even days. When search is over, results will be notified through “Email” provided. 2. Since novel molecules will be designed to PHR of CRY1, crystal structures without ligand should be chosen for modeling. 4k0r, 5t5x, and 6kx4 PDB entries having the highest coverage, 100% identity, and no ligand bound crystals can be used to model full-length PHR. The 5t5x structure includes residues between 1 and 496 found in the PHR of CRY1; however, some residues in this domain are still missing. SWISS-MODEL adds coordinates of missing atoms. The C-tail of CRY1 is unstructured and is not resolved in 5t5x. Thus, C-tail is also missing in our model. Overall, the model includes only PHR of CRY1 that will be used for structurebased small-molecule screening. 3. For each model SWISS-MODEL server provides quality scores such as GMQE and QMEAN. GMQE is a score between 0 and 1 that is calculated based on alignment of target-template and structure of the template. QMEAN is a Z-score that predicts the nativeness of the structural features of the model. Basically, it compares QMEAN scores of model and similar size crystal structures. Scores below 4 is the indication of a poor model. “Thumbs-up” or “Thumbs-down” symbol near to QMEAN scores is the indication of “good” or “bad” model, respectively. Some other quality controls and information of the model can be found by clicking “\/” symbols on the right-hand side of the panel. 4. Protonation states of ionizable amino acid residues such as histidine, aspartic acid, and glutamic acid depend on the polarity of the milieu and are quite substantial to design molecules to the target pocket. 5. CHARMM forcefield is composed of topology and parameter files. top_all36_prot.rtf file contains topology information for proteins that is needed to generate protein structure file (PSF). Topology file includes name, type, bond, and partial charges of atoms in all amino acid residues. Parameter file includes force constants for all bond-dependent terms and van der Waals and nonbonded interaction terms. 6. PSF includes data for all molecules that enables the application of a forcefield to the system. In this step a psf file for protein is generated which is going to be used in the next steps. “psfgen. in” has the following commands:

28

Seref Gul and Ibrahim Halil Kavakli package require psfgen topology top_all36_prot.rtf segment U {pdb CRY1_p.pdb} coordpdb CRY1_p.pdb U guesscoord writepdb CRY1_p_vmd.pdb writepsf CRY1_p_vmd.psf

7. To run MD simulations, we need to solvate our protein. CRY1 should be placed in a box with 15 Å water layer in every direction from the outermost atom in that direction. Then add NaCl salt 150 mM as final concentration to mimic the physiological environment. ionized.pdb and ionized.psf files now include CRY1, water molecules and salt (NaCl) molecules, ready for the MD simulation. “solvation.in” has the following commands: package require solvate solvate CRY1_p_vmd.psf CRY1_p_vmd.pdb -t 15 -o CRY1_wb

package require autoionize autoionize -psf CRY1_wb.psf -pdb CRY1_wb.pdb -sc 0.15

8. This command will measure the min and max coordinates in x, y, and z directions. Our measurement gave the following {52.63 10.58 79.04} {32.88 81.79 25.79}. These values are going to be used to calculate the origin of the system and vector length is needed in MD simulations. 9. restrain.in has the following commands: set everyone [atomselect top all] set fix [atomselect top “protein and name CA”] $everyone set beta 0 $fix set beta 1 $everyone writepdb CA_rest.pdb

10. minimize.in can be found in supplementary materials. We minimized the system by using conjugate gradient method for 20,000 steps. Meaning of parameters used in MD scripts can be found in this link: http://www.ks.uiuc.edu/Research/ namd/cvs/ug.pdf. Among many parameters, cellOrigin, cellBasisVector1, cellBasisVector2, and cellBasisVector3 are calculated from measurements given in Note 8. The arithmetic average of x, y, and z coordinates is provided as the cellOrigin. Extend of x, y, and z coordinates are calculated and provided as (x 0 0), (0 y 0), (0 0 z) format for cellBasisVector1, 2, and

Structure-Based Drug Discovery for Clock Proteins

29

3, respectively. The same cellBasisVector and cellOrigin coordinates will be used in heating.in, equilibrium.in, and production.in. After each MD step, log files should be checked for error messages. 11. Configuration files heating.in, equilibrium.in, and production. in can be found in the supplementary materials. In minimization and heating steps canonical NVT (N, particles; V, volume; T, temperature conserved) ensemble was used. In equilibration and production simulations isothermal-isobaric NPT (N, particles; P, pressure; T, temperature conserved) ensemble was used. 12. By following the above steps non-polar hydrogens were merged with atoms they bound and Gasteiger charges were added to each atom. 13. Since the Vina program is using “Spacing (angstrom)” 1, set this value 1. x-center, y-center, and z-center values will be different for each user because of the randomness in MD simulations. We choose coordinates of Arg358 since mutation of corresponding amino acid in CRY1 diminished interaction with FBXL3 [36]. Pocket size is determined to cover the whole primary pocket. PDBs are readable files, so coordinates of the CZ atom in Arg358 can be read from a pdb file using a text editor. 14. Following information is given in AutodockVina help menu: Input: --receptor arg rigid part of the receptor (PDBQT) --flex arg flexible side chains, if any (PDBQT) --ligand arg ligand (PDBQT)

Search space (required): --center_x arg X coordinate of the center --center_y arg Y coordinate of the center --center_z arg Z coordinate of the center --size_x arg size in the X dimension (Angstroms) --size_y arg size in the Y dimension (Angstroms) --size_z arg size in the Z dimension (Angstroms)

Output (optional): --out arg output models (PDBQT), the default is chosen based on the ligand file name --log arg optionally, write log file

30

Seref Gul and Ibrahim Halil Kavakli

Misc (optional): --cpu arg the number of CPUs to use (the default is to try to detect the number of CPUs or, failing that, use 1) --seed arg explicit random seed --exhaustiveness arg (=8) exhaustiveness of the global search (roughly proportional to time): 1+ --num_modes arg (=9) maximum number of binding modes to generate --energy_range arg (=3) maximum energy difference between the best binding mode and the worst one displayed (kcal/mol)

Configuration file (optional): --config arg the above options can be put here

Information (optional): --help display usage summary --help_advanced display usage summary with advanced options --version display program version

15. We assume that the openbabel program is downloaded and installed properly. Using the command provided, each molecule will be written to a single file, starting from the prefix given after the -O option. File names will increase one by one, e.g., sm1.sdf, sm2.sdf, sm3.sdf, etc. Openbabel program has many options that can be seen by typing the following: obabel -H

16. Full path to prepare_ligand4.py depends on where users downloaded the MGLTools. “~” stands for the path of home directory. After -l option, input is provided (in pdb format) and pdbqt format is generated automatically with the same name. All available options for this module can be found by typing the following: ~/MGLTools-1.5.6./MGLToolsPckgs/AutoDockTools/Utilities24/ prepare_ligand4.py

To convert all pdb files to pdbqt for loop script can be used. 17. To dock all molecules, a for loop script can be coded. Specifying the number of CPUs will help to run docking simulations

Structure-Based Drug Discovery for Clock Proteins

31

efficiently. For example, --cpu 8 option will allow programs to use only 8 CPUs and the rest of the processors can be used for other jobs. Unless --cpu is specified, Vina will try to use all available processors. Another option that can be helpful to mention is the exhaustiveness. While the default value of the exhaustiveness is 8, increasing this value linearly increases the computational time but exponentially decreases the probability of not finding the minimum binding energy. Increasing the exhaustiveness value, e.g., --exhaustiveness 12, will help to improve your results especially if the target pocket is large. However, computational cost should be taken into account when screening a large library. 18. A few factors should be regarded to run this script. Script was written for Python 2.7 or earlier versions; however, recent Python versions (3 and newer versions) handle some variables differently. Thus, to run this script without any editing default python version should be considered. If multiple python versions were installed version can be specified to run script on the terminal such as: python2.7 vina_screen_get_top.py 10

Another point to mention is that the script is written such that results were generated under a new folder. A quick change in the script will allow us to analyze results generated using the options given above. Change: “file_names = glob.glob(’*/*.pdbqt’)” to file_names = glob. glob(’*_out.pdbqt’)

19. Various options are available to visualize the molecule and interacting residues. First one is “to update display for each conformation.” Once it is clicked, which comes as the default, as conformation of the molecule is changed by using arrow keys, interacting residues will be updated automatically. Second, to show molecules in surface representation that is the default. Third, “close contact” atoms can be represented as solid spheres or wireframes. Close contact residues can be changed by playing with the “VDW Scaling Factor” option. Another option is to show ribbon for “near residues” or “for all residues.” Labels on the contact amino acids can be shown or hidden by clicking on “display labels on residues.” Pi-pi and pi-cation interactions can be visualized, as well. Finally, the generated favorite representation can be saved as a png file by clicking on “Save Image.”

32

Seref Gul and Ibrahim Halil Kavakli

References 1. Xie YL, Tang QM, Chen GJ, Xie MR, Yu SL, Zhao JJ, Chen LL (2019) New insights into the circadian rhythm and its related diseases. Front Physiol 10:ARTN 682. https://doi.org/ 10.3389/fphys.2019.00682 2. Gachon F, Nagoshi E, Brown SA, Ripperger J, Schibler U (2004) The mammalian circadian timing system: from gene expression to physiology. Chromosoma 113(3):103–112. https://doi.org/10.1007/s00412-0040296-2 3. Kavakli IH, Sancar A (2002) Circadian photoreception in humans and mice. Mol Interv 2(8):484–492. https://doi.org/10.1124/mi. 2.8.484 4. Kyriacou CP, Hastings MH (2010) Circadian clocks: genes, sleep, and cognition. Trends Cogn Sci 14(6):259–267. https://doi.org/ 10.1016/j.tics.2010.03.007 5. Kavakli IH, Baris I, Tardu M, Gul S, Oner H, Cal S, Bulut S, Yarparvar D, Berkel C, Ustaoglu P, Aydin C (2017) The photolyase/ cryptochrome family of proteins as DNA repair enzymes and transcriptional repressors. Photochem Photobiol 93(1):93–103. https://doi. org/10.1111/php.12669 6. Takahashi JS (2017) Transcriptional architecture of the mammalian circadian clock. Nat Rev Genet 18(3):164–179. https://doi.org/10. 1038/nrg.2016.150 7. Hirano A, Yumimoto K, Tsunematsu R, Matsumoto M, Oyama M, Kozuka-Hata H, Nakagawa T, Lanjakornsiripan D, Nakayama Keiichi I, Fukada Y (2013) FBXL21 regulates oscillation of the circadian clock through ubiquitination and stabilization of cryptochromes. Cell 152(5):1106–1118. https://doi.org/10. 1016/j.cell.2013.01.054 8. Yoo SH, Mohawk JA, Siepka SM, Shan Y, Huh SK, Hong HK, Kornblum I, Kumar V, Koike N, Xu M, Nussbaum J, Liu X, Chen Z, Chen ZJ, Green CB, Takahashi JS (2013) Competing E3 ubiquitin ligases govern circadian periodicity by degradation of CRY in nucleus and cytoplasm. Cell 152(5): 1091–1105. https://doi.org/10.1016/j.cell. 2013.01.055 9. Krishnaiah SY, Wu G, Altman BJ, Growe J, Rhoades SD, Coldren F, Venkataraman A, Olarerin-George AO, Francey LJ, Mukherjee S, Girish S, Selby CP, Cal S, Ubeydullah ER, Sianati B, Sengupta A, Anafi RC, Kavakli IH, Sancar A, Baur JA, Dang CV, Hogenesch JB, Weljie AM (2017) Clock regulation of metabolites reveals coupling between

transcription and metabolism. Cell Metab 25(4):961. https://doi.org/10.1016/j.cmet. 2017.03.019 10. Zhang R, Lahens NF, Ballance HI, Hughes ME, Hogenesch JB (2014) A circadian gene expression atlas in mammals: implications for biology and medicine. Proc Natl Acad Sci U S A 111(45):16219–16224. https://doi.org/ 10.1073/pnas.1408886111 11. Cal-Kayitmazbatir S, Kulkoyluoglu-Cotul E, Growe J, Selby CP, Rhoades SD, Malik D, Oner H, Asimgil H, Francey LJ, Sancar A, Kruger WD, Hogenesch JB, Weljie A, Anafi RC, Kavakli IH (2020) CRY1-CBS binding regulates circadian clock function and metabolism. FEBS J. https://doi.org/10.1111/febs.15360 12. Toh KL, Jones CR, He Y, Eide EJ, Hinz WA, Virshup DM, Ptacek LJ, Fu YH (2001) An hPer2 phosphorylation site mutation in familial advanced sleep phase syndrome. Science 291(5506):1040–1043 13. Vanselow K, Vanselow JT, Westermark PO, Reischl S, Maier B, Korte T, Herrmann A, Herzel H, Schlosser A, Kramer A (2006) Differential effects of PER2 phosphorylation: molecular basis for the human familial advanced sleep phase syndrome (FASPS). Genes Dev 20(19):2660–2672. https://doi. org/10.1101/gad.397006 14. Xu Y, Toh KL, Jones CR, Shin JY, Fu YH, Ptacek LJ (2007) Modeling of a human circadian mutation yields insights into clock regulation by PER2. Cell 128(1):59–70. https://doi. org/10.1016/j.cell.2006.11.043 15. Takano A, Uchiyama M, Kajimura N, Mishima K, Inoue Y, Kamei Y, Kitajima T, Shibui K, Katoh M, Watanabe T, Hashimotodani Y, Nakajima T, Ozeki Y, Hori T, Yamada N, Toyoshima R, Ozaki N, Okawa M, Nagai K, Takahashi K, Isojima Y, Yamauchi T, Ebisawa T (2004) A missense variation in human casein kinase I epsilon gene that induces functional alteration and shows an inverse association with circadian rhythm sleep disorders. Neuropsychopharmacology 29(10):1901–1909. https://doi.org/10. 1038/sj.npp.1300503 16. Weitzman ED, Czeisler CA, Coleman RM, Spielman AJ, Zimmerman JC, Dement W, Richardson G, Pollak CP (1981) Delayed sleep phase syndrome. A chronobiological disorder with sleep-onset insomnia. Arch Gen Psychiatry 38(7):737–746 17. Katzenberg D, Young T, Finn L, Lin L, King DP, Takahashi JS, Mignot E (1998) A CLOCK

Structure-Based Drug Discovery for Clock Proteins polymorphism associated with human diurnal preference. Sleep 21(6):569–576 18. Lamont EW, Coutu DL, Cermakian N, Boivin DB (2010) Circadian rhythms and clock genes in psychotic disorders. Isr J Psychiatry Relat Sci 47(1):27–35 19. Lamont EW, Legault-Coutu D, Cermakian N, Boivin DB (2007) The role of circadian clock genes in mental disorders. Dialogues Clin Neurosci 9(3):333–342 20. Mansour HA, Wood J, Logue T, Chowdari KV, Dayal M, Kupfer DJ, Monk TH, Devlin B, Nimgaonkar VL (2006) Association study of eight circadian genes with bipolar I disorder, schizoaffective disorder and schizophrenia. Genes Brain Behav 5(2):150–157. https:// doi.org/10.1111/j.1601-183X.2005. 00147.x 21. Moons T, Claes S, Martens GJ, Peuskens J, Van Loo KM, Van Schijndel JE, De Hert M, van Winkel R (2011) Clock genes and body composition in patients with schizophrenia under treatment with antipsychotic drugs. Schizophr Res 125(2–3):187–193. https://doi.org/10. 1016/j.schres.2010.10.008 22. Soria V, Martinez-Amoros E, Escaramis G, Valero J, Perez-Egea R, Garcia C, GutierrezZotes A, Puigdemont D, Bayes M, Crespo JM, Martorell L, Vilella E, Labad A, Vallejo J, Perez V, Menchon JM, Estivill X, Gratacos M, Urretavizcaya M (2010) Differential association of circadian genes with mood disorders: CRY1 and NPAS2 are associated with unipolar major depression and CLOCK and VIP with bipolar disorder. Neuropsychopharmacology 35(6):1279–1289. https://doi.org/10.1038/ npp.2009.230 23. Benedetti F, Dallaspezia S, Colombo C, Pirovano A, Marino E, Smeraldi E (2008) A length polymorphism in the circadian clock gene Per3 influences age at onset of bipolar disorder. Neurosci Lett 445(2):184–187. https://doi.org/10.1016/j.neulet.2008. 09.002 24. McCarthy MJ, Welsh DK (2012) Cellular circadian clocks in mood disorders. J Biol Rhythm 27(5):339–352. https://doi.org/10.1177/ 0748730412456367 25. Pellegrino R, Kavakli IH, Goel N, Cardinale CJ, Dinges DF, Kuna ST, Maislin G, Van Dongen HP, Tufik S, Hogenesch JB, Hakonarson H, Pack AI (2014) A novel BHLHE41 variant is associated with short sleep and resistance to sleep deprivation in humans. Sleep 37(8):1327–1336. https://doi. org/10.5665/sleep.3924 26. Onat OE, Kars ME, Gul S, Bilguvar K, Wu Y, Ozhan A, Aydin C, Basak AN, Trusso MA,

33

Goracci A, Fallerini C, Renieri A, Casanova JL, Itan Y, Atbasoglu CE, Saka MC, Kavakli IH, Ozcelik T (2020) Human CRY1 variants associate with attention deficit/hyperactivity disorder. J Clin Invest 130(7):3885–3900. https://doi.org/10.1172/JCI135500 27. Patke A, Murphy PJ, Onat OE, Krieger AC, Ozcelik T, Campbell SS, Young MW (2017) Mutation of the human circadian clock gene CRY1 in familial delayed sleep phase disorder. Cell 169(2):203–215. https://doi.org/10. 1016/j.cell.2017.03.027 28. Chen Z, Yoo SH, Park YS, Kim KH, Wei S, Buhr E, Ye ZY, Pan HL, Takahashi JS (2012) Identification of diverse modulators of central and peripheral circadian clocks by highthroughput chemical screening. Proc Natl Acad Sci U S A 109(1):101–106. https://doi. org/10.1073/pnas.1118034108 29. Hirota T, Kay SA (2015) Identification of small-molecule modulators of the circadian clock. Methods Enzymol 551:267–282. https://doi.org/10.1016/bs.mie.2014. 10.015 30. Hirota T, Lewis WG, Liu AC, Lee JW, Schultz PG, Kay SA (2008) A chemical biology approach reveals period shortening of the mammalian circadian clock by specific inhibition of GSK-3beta. Proc Natl Acad Sci U S A 105(52):20746–20751. https://doi.org/10. 1073/pnas.0811410106 31. Isojima Y, Nakajima M, Ukai H, Fujishima H, Yamada RG, Masumoto KH, Kiuchi R, Ishida M, Ukai-Tadenuma M, Minami Y, Kito R, Nakao K, Kishimoto W, Yoo SH, Shimomura K, Takao T, Takano A, Kojima T, Nagai K, Sakaki Y, Takahashi JS, Ueda HR (2009) CKI epsilon/delta-dependent phosphorylation is a temperature-insensitive, period-determining process in the mammalian circadian clock. Proc Natl Acad Sci U S A 106(37):15744–15749. https://doi.org/10. 1073/pnas.0908733106 32. Doruk YU, Yarparvar D, Akyel YK, Gul S, Taskin AC, Yilmaz F, Baris I, Ozturk N, Turkay M, Ozturk N, Okyar A, Kavakli IH (2020) A CLOCK-binding small molecule disrupts the interaction between CLOCK and BMAL1 and enhances circadian rhythm amplitude. J Biol Chem 295(11):3518–3531. https://doi.org/10.1074/jbc.RA119.011332 33. Czarna A, Berndt A, Singh HR, Grudziecki A, Ladurner AG, Timinszky G, Kramer A, Wolf E (2013) Structures of Drosophila cryptochrome and mouse Cryptochrome1 provide insight into circadian function. Cell 153(6): 1394–1405. https://doi.org/10.1016/j.cell. 2013.05.011

34

Seref Gul and Ibrahim Halil Kavakli

34. Michael AK, Fribourgh JL, Chelliah Y, Sandate CR, Hura GL, Schneidman-Duhovny D, Tripathi SM, Takahashi JS, Partch CL (2017) Formation of a repressive complex in the mammalian circadian clock is mediated by the secondary pocket of CRY1. Proc Natl Acad Sci U S A 114(7):1560–1565. https://doi.org/10. 1073/pnas.1615310114 35. Miller S, Son YL, Aikawa Y, Makino E, Nagai Y, Srivastava A, Oshima T, Sugiyama A, Hara A, Abe K, Hirata K, Oishi S, Hagihara S,

Sato A, Tama F, Itami K, Kay SA, Hatori M, Hirota T (2020) Isoform-selective regulation of mammalian cryptochromes. Nat Chem Biol 16(6):676. https://doi.org/10.1038/ s41589-020-0505-1 36. Xing WM, Busino L, Hinds TR, Marionni ST, Saifee NH, Bush MF, Pagano M, Zheng N (2013) SCFFBXL3 ubiquitin ligase targets cryptochromes at their cofactor pocket. Nature 496(7443):64. https://doi.org/10.1038/ nature11964

Chapter 3 Analysis of Complex Circadian Time Series Data Using Wavelets Christoph Schmal, Gregor Mo¨nke, and Adria´n E. Granada Abstract Experiments that compare rhythmic properties across different genetic alterations and entrainment conditions underlie some of the most important breakthroughs in circadian biology. A robust estimation of the rhythmic properties of the circadian signals goes hand in hand with these discoveries. Widely applied traditional signal analysis methods such as fitting cosine functions or Fourier transformations rely on the assumption that oscillation periods do not change over time. However, novel high-resolution recording techniques have shown that, most commonly, circadian signals exhibit time-dependent changes of periods and amplitudes which cannot be captured with the traditional approaches. In this chapter we introduce a method to determine time-dependent properties of oscillatory signals, using the novel open-source Python-based Biological Oscillations Analysis Toolkit (pyBOAT). We show with examples how to detect rhythms, compute and interpret high-resolution time-dependent spectral results, analyze the main oscillatory component, and to subsequently determine these main components’ time-dependent instantaneous period, amplitude, and phase. We introduce step-by-step how such an analysis can be done by means of the easy-to-use point-and-click graphical user interface (GUI) provided by pyBOAT or executed within a Python programming environment. Concepts are explained using simulated signals as well as experimentally obtained time series. Key words Circadian clocks, Data analysis, Oscillations, Time series analysis, Wavelets, Nonstationary signals, Spectral analysis, Synchronization

1

Introduction Circadian oscillations are present at all scales of an organism, from the cellular up to the behavioral level. Recent improvements in the experimental techniques have allowed unprecedented long-term high-resolution recordings in cultures of individual cells, ex vivo tissues and even in vivo from freely moving animals [1–4]. In some cases, these rhythms show robust stable oscillations with steady period and amplitude, but most frequently they show timedependent fluctuations in period, amplitude, and trends. Nevertheless, when it comes to quantifying these rhythms, the vast majority

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_3, © The Author(s) 2022

35

36

Christoph Schmal et al.

of the circadian community still relies on software tools that, at their core, rely on methods designed under the premise that oscillations have static time-independent components, also known as the stationarity assumption. Among others, stationarity-based methods implemented ubiquitously in analytical software tools include the well-known Fourier transformations, Lomb Scargle periodograms, and cosinor analysis [5]. For specific cases, the biological data analysis community has developed data-analysis tools tailored to the characterization of nonstationary oscillatory components [6, 7]. These new set of robust software solutions are in practice incorporated as an additional step within larger data analysis pipelines that typically include preprocessing denoising, detrending, and normalization. We have recently shown that such multistep pipelines that combine preprocessing steps with a subsequent analysis of oscillatory properties can lead to significant spectral artifacts that often remain undetected [8]. In this chapter, we describe through examples how to use a recently published multistep open-source software tool, pyBOAT, that integrates all required steps for the analysis of raw circadian data. PyBOAT implements wavelet analysis and was specifically designed for noisy nonstationary datasets that by-design overcomes potential spectral artifacts of the most frequent preprocessing steps in time series data analysis. In Subheading 2 of this chapter we provide a set of online sources to download and install pyBOAT. In Subheading 3.1, we describe the graphical user interface to carry out a spectral analysis and generate figures of the results. Finally, in Subheading 3.2, we introduce a flexible scripting-based implementation of pyBOAT.

2

Materials Software: 1. pyBOAT is a freely available open-access software that runs on multiple mainstream operating systems such as Linux, MacOS, and Windows. It requires a Python 3.x version to be installed on the system. 2. A convenient approach to install Python together with pyBOAT is by means of Anaconda, an open-source Python and R programming language distribution that aims at simplifying package management for scientific computing. Anaconda has a graphical user interface (GUI), the Anaconda Navigator, and thus requires no use of the command-line. An installation manual for Anaconda can be found on: https://docs.anaconda. com/anaconda/install/

Wavelet Analysis

37

3. A detailed guide how to install pyBOAT using Anaconda can be found at https://github.com/tensionhead/pyBOAT or by following the steps in the pyBOAT’s video installation tutorial http://granadalab.org/media/ 4. In case Anaconda or its package manager conda is already installed on the machine, pyBOAT can be installed via the command line by typing conda config --add channels conda-forge conda install pyboat

5. pyBOAT can be installed without conda using the packagemanagement system pip by typing pip install pyboat

into the command-line.

3

Methods

3.1 Graphical User Interface

pyBOAT contains an easy-to-use graphical user interface (GUI) that requires no programming experience. In the following paragraphs we will introduce step by step how to perform a wavelet analysis of circadian time series data using the GUI of pyBOAT. In order to illustrate the strength of our wavelet analysis approach, we will first investigate a simulated oscillatory time series with wellknown properties. Experimental data is then subsequently analyzed in Subheadings 3.1.8 and 3.2.6.

3.1.1 Example Data

Commonly applied time series analysis methods such as Fourier analysis, Lomb-Scargle periodograms, or fitting of harmonic functions reliably estimate the period of main oscillatory components as long as oscillatory properties remain stable or vary little over time. In contrast, time-frequency methods such as wavelet analysis are well suited to uncover time-dependent (i.e., instantaneous) periods and amplitudes. In order to illustrate the strength of wavelet analysis, we analyze a simulated nonstationary, noisy oscillatory time series whose period changes linearly from 22 to 26 h within a week. This synthetic signal additionally exhibits a nonlinear baseline expression trend as well as a decaying amplitude, signal properties often encountered in practice.

3.1.2 Download Example Data

In Subheading 3.2.1 we describe how to simulate and save the example time series data that we are going to analyze in the next paragraphs, using Python commands. Alternatively, the data can be downloaded from the following link: https://github.com/ cschmal/chapter-wavelets

38

Christoph Schmal et al.

Fig. 1 Data import and parameter setup for the analytic wavelet transform. (a) Main window of pyBOAT. (b) Data Import Options window. (c) The DataViewer shows the imported data, plots the time series and trend of interest, and allows to define parameters for sinc filter detrending and the subsequent wavelet analysis 3.1.3 Launch pyBOAT

The graphical user interface of pyBOAT can be started either by launching it from the Anaconda Navigator or by typing pyboat into the command line.

3.1.4 Import Data

The main window of pyBOAT (Fig. 1a) contains three elements. The Start Generator button on the right column launches a signal generator to analyze synthetic (simulated) signals (see Note 1). This can be useful for teaching purposes or to accommodate with the

Wavelet Analysis

39

properties and features of pyBOAT and its underlying wavelet analysis. The buttons within the left column allow to import external data: 1. pyBOAT expects tabular data in one of the supported formats .xls, .xlsx, .csv, .tsv, or .txt. Each column should contain a time series signal, sampled at equidistant time points (see Note 2). 2. Click on the Open button and select the file that contains the data to be imported. Column names are automatically inferred from the first row of the data file. We can use this option to import the data set from Subheading 3.1.2. 3. Click on the Import button for more importing options as shown in Fig. 1b. For nonstandard data formats, uncheck Separator from extension and specify a custom Column separator in the corresponding box. In case the first row of the data file does not contain column names, check the No header row present box. In addition, you can choose to check the Interpolate missing values option for a linear interpolation of gaps in your data set (see Note 3). 4. After importing the data, the DataViewer window opens as shown in Fig. 1c. Within the DataViewer you can see the head of the imported data and set various options for the subsequent analysis. 5. First, choose a data column for further analysis. We will analyze the second column since the first column contains the time points of data sampling in our data set. This column can be chosen either by clicking on the respective column or by choosing the column name within the Select Signal box. 6. Second, we need to specify the sampling interval and corresponding time unit. In our example we analyze (simulated) data that has been sampled at a 15-min interval. Thus, we choose “0.25” within the Sampling Interval box and “hours” within the Time Unit box, see upper part of Fig. 1c. 7. Click the Refresh Plot button. One should now see the raw time series signal for the chosen time axis units in the Signal and Trend plotting window (Fig. 1c bottom left). 3.1.5 Detrending

Circadian time series often exhibit long-term changes in their magnitude of oscillation. While changes in this magnitude can be informative by themselves, it is often useful to remove this baseline trend for a better representation of oscillatory components and further analysis. In Mo¨nke et al. [8] we argue that the sinc filter is a good choice for removing nonlinear trends (low frequency components) of oscillating time series, while minimizing common detrending artifacts such as spurious oscillations. The sinc filter works as a step function in the frequency domain. It removes signal

40

Christoph Schmal et al.

components that are larger than a predefined cutoff period and neither attenuates nor amplifies components below this cutoff period [8]. 1. Choose a cutoff period for the sinc filter by typing a numerical value into the Cut-off Period box of the Sinc Detrending panel within the DataViewer. Here, we chose a cutoff period of 48 h, i.e., roughly twice as long as the expected period of the circadian signal (Fig. 1c, see Note 4). 2. Check the box Trend within the Plotting Options of the DataViewer to plot the trend, determined by the sinc filter. The nonlinear parabola-shaped trend of this synthetic time series is nicely captured (Fig. 1c purple line). 3. One can plot the detrended time series by additionally checking the Detrended Signal box. 4. The raw data, trend, and detrended time series data can be saved into a three-column data file via the Save Filter Results button. Supported output formats are *.txt, *.csv, or *.xlsx. 3.1.6 Analysis and Detection of Periodic Signals Using Wavelets

1. For wavelet-based time frequency analysis, a mother wavelet (see Note 5) probes the signal of interest along the time axis for a range of predefined frequencies or periods. This range of periods of interest has to be supplied by the user within the Analysis panel of the DataViewer. One can specify the periods to be analyzed between a Lowest period and a Highest period at equidistant steps for a given total Number of periods by typing numerical values into the corresponding boxes. Periods outside of this user-defined range will not be included in the subsequent analysis, so especially for an explorative analysis it is recommended to initially choose a wide interval of potential periods. 2. Here, we chose a lowest period of 0.5 h, tantamount to the Nyquist period given by two-times the sampling interval. The Nyquist period is also the default value used by pyBOAT (see Fig. 1c). 3. In order to perform the wavelet analysis on the sinc-detrended time series one has to check the Use Detrended Signal box. 4. Click on the Analyze Signal button to perform the wavelet analysis. 5. After the computation is done, the Wavelet Spectrum window opens, see Fig. 2a. The upper part of the window shows the analyzed signal, i.e., in our case the detrended time series. The middle part of the window shows the wavelet spectrogram, the main result of our time frequency analysis. Such spectrogram gives a detailed time-resolved picture that is able to unveil timedependent oscillatory properties (see Fig. 2a) as well as multiple

Wavelet Analysis

41

Fig. 2 Wavelet analysis and ridge readout. (a) Wavelet spectrum window. Upper window depicts the analyzed signal, i.e., in our case the sinc-detrended time series. Bottom window depicts the wavelet spectrogram together with the detected maximum power ridge (bold red line) and the cones of influence (gray dasheddotted lines). (b) Wavelet Results window, obtained by clicking on the Plot Ridge Readout button in panel a. Depicted are the time-dependent (instantaneous) period (upper left), phase (upper right), amplitude (bottom left), and maximum power values (bottom right) as evaluated from the maximum power ridge in panel a. (c) Comma-separated value (csv) readout, obtained by clicking on the Save Results button in panel b

oscillatory components such as ultradian rhythms subordinated to circadian oscillations [7]. 6. The region of maximum power shows a clear trend toward longer periods for later times t and thus successfully captures the linear evolution from a period of 22 h up to 26 h within 1 week as described in Subheading 3.1.1. The decaying power for later times t reflects the decaying amplitude of the signal. 3.1.7 Ridge Analysis Reveals the Main Rhythmic Component

Although the wavelet spectrogram gives a complete picture of the time-resolved oscillatory properties, potentially including multiple dominant periods, one is often interested in identifying a main oscillatory component and its properties. Such main oscillatory component can be deduced from a wavelet ridge (see Note 6). pyBOAT connects the set of maximal power values in the spectrogram along successive time points to determine the ridge: 1. Click on the Detect Maximum Ridge button within the Ridge Detection panel of the Wavelet Spectrum window to compute the maximum power ridge. Subsequently, a bold red line depicts the ridge within the wavelet spectrogram (see Fig. 2a).

42

Christoph Schmal et al.

2. In order to avoid evaluation of the ridge within a low spectral power (noise) regime, one can set a minimum wavelet power threshold (see Note 7) by typing a numerical value into the Ridge Threshold box of the Ridge Detection panel. 3. To work around sudden jumps within the ridge, e.g., due to poor spectral resolution of the transform, the ridge can also be smoothed by choosing a Savitzky-Golay window size. 4. Click on the Plot Ridge Readout button to evaluate the timedependent (instantaneous) period, amplitude, power, and phase of the main oscillatory component. A new window termed Wavelet Results will subsequently open, see Fig. 2b. 5. The Wavelet Results window shows the time-dependent oscillation period (upper left), phase (upper right), and amplitude (bottom left) as well as the spectrogram power along the maximum ridge (bottom right). Values within or outside the cone of influence are depicted by dashed or bold lines, respectively (see Note 8). 6. To save these results click on Save Results at the bottom of the window. Supported formats are *.txt, *.csv, and *.xlsx. See Fig. 2c for an example readout. 3.1.8 Ensemble Analysis

The GUI of pyBOAT provides a convenient way to analyze large ensembles of time series data (see Note 9) and gives various summary statistics such as the ensembles period and amplitude distribution or phase coherence. We demonstrate this functionality using a PER2::LUC bioluminescence recording within coronal slices of the mammalian central pacemaker—the suprachiasmatic nucleus (SCN)—as previously published in Abel et al. [1]. Within this data set, SCN slices have been treated with tetrodotoxin (TTX) to suspend spike-associated couplings, 4 days after starting the in vitro recordings. After another 6 days, TTX has been washed out from the medium. Circadian oscillatory signals of individual SCN neurons could be identified and tracked throughout the time lapse recordings, see Abel et al. for further details. Within the next steps we study the effect of TTX on dynamical properties of the SCN neurons circadian PER2::LUC oscillations using the batch analysis function of pyBOAT: 1. PER2::LUC oscillatory time series and the corresponding locations of SCN neurons within the time lapse recordings of ex vivo SCN slices as published in Abel et al. can be downloaded from: https://github.com/JohnAbel/scnresynchronization-data-2016 2. Here, we chose the data set “scn2_full_data.csv.” It contains 264 times series recorded from the tracked individual SCN neurons. Since the file lacks descriptive headers in the first row, we import the data via the Import button of pyBOATs main screen and check the No header row present box.

Wavelet Analysis

43

Fig. 3 Analysis of large data sets. (a) The DataViewer window of pyBOAT. A large data set has been imported that contains 264 bioluminescence time series signals obtained by individual tracking of SCN neurons within a cultured SCN slice at a sampling rate of 1 h. (b) Wavelet analysis of a bioluminescence recording from an individual SCN neuron stored in column one as shown in panel a. (c) The Batch Processing window allows to specify options for the analysis of large ensemble data sets. (d) Summary statistics of the corresponding ensemble analysis of all 264 recordings. Bold lines denote averages while shaded areas denote standard deviations of oscillation properties evaluated along the maximum power ridges of all signals

3. After importing the data, we set the Sampling Interval to 1 and the Time Unit as “h” as used in the experimental protocol, see Fig. 3a. 4. Amplitudes of PER2::LUC oscillations drastically decrease upon TTX treatment. To reduce edge effects at this transition for sinc filter detrending, we choose a relatively low filter cutoff of 36 h (see Note 10), compare Fig. 3a. 5. Once suitable parameters for sinc detrending and the subsequent wavelet analysis have been found (compare Fig. 3a, b), one can run the batch or ensemble analysis by clicking on the Analyze All button. 6. Choose the Summary Statistics of interest, Ridge Detection parameters as described above and required export options within the Batch Processing window. 7. Check the Ensemble Dynamics box in the Summary Statistics panel and click on the Run for 264 Signals button to execute the batch analysis.

44

Christoph Schmal et al.

8. The Ensemble Dynamics window depicts the mean and standard deviation of the time-dependent (instantaneous) period, amplitude and wavelet spectrum power as well as the phase coherence for the whole ensemble of signals. It can be seen that application of TTX reversibly broadens the period distribution (i.e., standard deviation increases), decreases the phase coherence, and leads to a decrease in amplitude which can be indicative for a reduced relative coupling strength as described previously [9–11]. 3.2 Implementing pyBOAT Within a Python Script

Even though pyBOAT has an easy-to-use graphical user interface (GUI), it can be more convenient in some cases to run the analysis routines of pyBOAT within a Python script. This applies, for example, to cases where the wavelet analysis provided by pyBOAT is only part of a larger analysis pipeline or to cases of extremely long time series data where, due to the long computation time of wavelet spectra, the practitioner could want to run the analysis on an external computing cluster. In the following sections we demonstrate how to analyze time series data with pyBOAT using the Python programming language. The code can be run either by executing each line in an interactive command shell such as IPython or a Jupyter Notebook, or by copying the code into a file—e.g., “example.py”—and running it in the shell using the command python example.py.

3.2.1 Generating Simluated Complex Oscillatory Time Series Data

The following steps show how to create and save the example time series data as analyzed in Subheadings 3.1.4–3.1.7. 1. The NumPy scientific computing library for Python provides convenient ways to simulate rhythmic time series. The library can be imported via import numpy as np

2. We define an equidistant set of time points, using a sampling interval of 15 min for an overall length of 1 week: dt = 0.25 # Sampling interval in hours tvec = np.arange(0, 7 * 24 + dt, dt) # Time array

Please note that the Python interpreter ignores everything that follows a “#” within a given line. 3. We next define a time-dependent (instantaneous) period of the simulated time series that linearly lengthens from 22 to 26 h within 1 week: T = 22. + 4. / (7. * 24.) * tvec

Wavelet Analysis

45

4. An oscillatory signal based on this instantaneous period T is constructed via the NumPy cosine function: signal = np.cos(2 * np.pi / T * tvec)

5. In many cases, such as bioluminescence recordings in cell culture or tissue slices, the corresponding circadian time series exhibit a decay in amplitude. We model this by introducing an exponential decay at a half-life of 48 h: signal = signal * np.e**(-np.log(2) / 48. * tvec)

6. In addition, many experimental circadian time series show a nonlinear trend in baseline expression or magnitude. Here, we add a simple nonlinear baseline trend to our signal, given by a mirrored parabola signal += -tvec * (tvec – 8 * 24) * 0.0002

7. Noise is omnipresent in biological signals and can be due to uncertainties in the measurement process or due to the intrinsic probabilisitic nature of biological processes. Noise can confound otherwise precise deterministic observables of interest but can, on the other hand, also be a critical function to drive biological processes [12, 13]. Here, we add for illustrative purposes uncorrelated (Gaussian) white noise to our simulated time series by adding an array of normally distributed random numbers of standard deviation 0.15 and of mean 0, using the “normal” function of NumPy’s “random” package: signal += np.random.normal(loc=0, scale=0.15, size=len(tvec))

The impact of different kinds of noise as well as noise strength are discussed in more detail in Mo¨nke et al. [8]. 3.2.2 Import and Initialization of the Wavelet Analyzer

1. Import the wavelet analyzer of pyBOAT by typing from pyboat import WAnalyzer

2. Analogously to Subheading 3.1.6, step 1, we first define a set of periods to be analyzed by the wavelet based time-frequency analysis. For this sake, we use the linspace function of the Numpy package: periods = np.linspace(start=2*dt, stop=48, num=200)

Above function generates an array of elements with values between the Nyquist period 2*dt and 48 h in 200 equidistant steps.

46

Christoph Schmal et al.

3. We choose the time unit (here “hours”) and initialize the wavelet analyzer via wAn = WAnalyzer(periods, dt, time_unit_label=’hours’)

4. pyBOAT follows up on the pythonic idea of “introspection,” e.g., typing help(wAn)

shows a comprehensive documentation of the WAnalyzer and its methods. 5. In order to show pyBOAT’s results interactively we import the Python Matplotlib library via import matplotlib.pyplot as plt

6. Typing plt.ion()

turns the interactive plotting mode on. 3.2.3 Detrending

Analogously to Subheading 3.1.5, we first detrend the raw time series using sinc filter smoothing as implemented within the pyBOAT package: 1. Define a sinc filter cutoff period in hours T_c = 48

2. The sinc_smooth function of the pyBOAT package requires two arguments, the raw signal that we aim to detrend as well as the cutoff period from step 1 of this section: trend = wAn.sinc_smooth(signal, T_c=T_c)

3. We subsequently obtain the detrended time series via subtracting the trend from the original signal detrended_signal = signal – trend

4. pyBOAT offers functions to plot the time series and corresponding trend: wAn.plot_signal(signal, label=’Raw signal’, color=’red’, alpha=0.5) wAn.plot_trend(trend, label=’Trend’)

Wavelet Analysis

47

Note that any arguments (e.g., “label,” “color,” or “alpha”) that are accepted by the plot function of the Matplotlib package can be provided to above functions. 3.2.4 Computing the Wavelet Spectrum

1. We defined the parameters that are required for the wavelet analysis already in Subheading 3.2.2, step 3. We now perform the wavelet analysis on the detrended time series by typing modulus, transform = wAn.compute_spectrum(detrended_signal)

This function also spawns a plot, showing the signal timealigned with the wavelet power spectrum as obtained from the GUI (Fig. 2a). 2. Here, “modulus” is a two-dimensional real valued array containing the Wavelet power spectrum while “transform” is a two-dimensional array representing the direct output of the complex convolutions with the Morlet mother wavelet. The number of rows equals the number of periods and the number of columns equals the length of the signal for both arrays. 3. Both objects modulus and transform are Numpy arrays and can be saved using, e.g., the Numpy function np.savetxt or the Python pickle module. 3.2.5 Detect and Evaluate the Wavelet Ridge

Analogously to Subheading 3.1.7, we compute and evaluate the maximal power ridge of the wavelet spectrum: 1.

ridge_results = wAn.get_maxRidge()

determines the maximum power ridge of the wavelet spectrum computed in Subheading 3.2.4, step 1 (see Note 11). 2. To plot the maximum ridge as a red line in the wavelet spectrogram, type wAn.draw_Ridge()

3. The object “ridge_results” defined in step 1 contains the instantaneous period, phase, amplitude, and maximum power values along the maximum power ridge of the wavelet spectrum. It is in a pandas DataFrame format such that the convenient I/O functions of the pandas software library can be used to save the results. For example, ridge_results.to_csv("save_ridge.csv")

saves the results as a comma-separated values (csv) file, named “save_ridge.csv”. Excel users might want to save their data into an Excel sheet using ridge_results.to_excel("save_ridge.xlsx")

48

Christoph Schmal et al.

4. These instantaneous properties can be plotted by typing wAn.plot_readout(draw_coi=True)

which reproduces Fig. 2b. 3.2.6 Import and Analysis of Experimental Datasets

In this paragraph we study a previously published experimental data set. We show how such data can be imported within a Python scripting environment and subsequently analyze it. The data set: Circadian rhythms are generated at the intracellular level through multiple interlocked regulatory feedback loops [14]. In mammals the negative feedback loop is composed of the Period (Per1, Per2, Per3), Cryptochrome (Cry1, Cry2), and Bmal1 clock genes, see Fig. 4a. This core loop is intertwined with multiple other loops such as the Bmal1 and Reverb (Reva and Revb) negative feedback loop [15]. Recently, it has been shown that perturbations of the system given by jet-lag, light pulses, SCN slice preparations, or culture medium exchange induce differential dynamical changes among different clock genes [16–20] and that these differential perturbation responses could be explained by the topology of the intracellular regulatory network [21]. Such differential responses translate into (at least transiently) different instantaneous amplitudes and periods among different clock genes and should thus be analyzed by a time-frequency analysis that can account for these complex and time-varying dynamical properties. In the next paragraphs we analyze Bmal1-ELuc and Per1-luc reporter expression within SCN slices of double-transgenic mice that express both reporters simultaneously as previously described in Ono et al. [19]. 1. The data has been stored in a text-file “bioluminescence_raw. txt” containing three different columns of numerical values, i.e., the time instances of measurements as well as bioluminescence intensities of the Bmal1-ELuc and Per1-luc reporter constructs, respectively. The first row contains the data description, see Fig. 4b. 2. There are multiple ways to load such data within a Python environment. One of the easiest is to use the loadtxt function of Numpy: import numpy as np np.loadtxt("./bioluminescence_raw.txt", skiprows=1) t, Bmal1, Per1 = Data.T

Here, the file “bioluminescence_raw.txt” has to be in the same folder as the Python script. An alternative way is to use the convenient read_csv function of the Pandas library which is especially well suited for large data sets: import pandas as pd Data = pd.read_csv("./bioluminescence_raw.txt", sep="\t")

Wavelet Analysis

49

Fig. 4 Detecting differential phase responses of Bmal1-ELuc and Per1-Luc clock gene reporter expression. (a) Sketch of the mammalian circadian core clock regulatory network. (b) Illustration of the data set. (c) Raw non-detrended Bmal1-ELuc (green) and Per1-luc (blue) bioluminescence given in relative light units (RLU; 1RLU ¼ 1000 counts per 15 s), measured via a photomultiplier tube (PMT) as previously described [19]. Nonlinear baseline expression trends (magnitudes), determined by a sinc-filter, are depicted by a black dotted or dashed line in case of Bmal1-ELuc or Per1-luc signals, respectively. (d) Corresponding detrended bioluminescence signals, calculated by subtracting the trend from the original raw time series data from panel (c). (e) Time-dependent instantaneous periods, evaluated from the maximum power ridge of the corresponding wavelet spectrograms (data not shown). (f) Difference of the time-dependent instantaneous Bmal1-ELuc and Per1-luc oscillation phases t = Data["Time"].values Bmal1 = Data["Bmal1"].values Per1 = Data["Per1"].values

Figure 4c depicts the raw bioluminescence time series data as measured experimentally. Note that the Bmal1-ELuc reporter has a much brighter overall light intensity but a smaller relative amplitude in comparison to the Per1-luc signal. 3. pyBOAT allows to detrend the raw time series signal and compute the wavelet spectrum in a single step which we will showcase in the next points.

50

Christoph Schmal et al.

4. Define the experimentally used sampling interval of the data, choose the corresponding time unit, set up the periods of interest, and initialize the wavelet analyzer via: dt = 10./60 # Experimental sampling interval in hours periods = np.linspace(2*dt, 48, 600) wAn = WAnalyzer(periods, dt, time_unit_label=’hours’)

5. Define a cutoff period for sinc detrending, compute the wavelet spectra, and determine the maximum power ridge for the Bmal1-ELuc T_cutoff = 96 # Define cutoff period in h B1_modulus,

B1_transform

=

wAn.compute_spectrum(Bmal1,

T_c=T_cutoff) B1_ridge = wAn.get_maxRidge()

and Per1-luc data P1_modulus,

P1_transform

=

wAn.compute_spectrum(Per1,

T_c=T_cutoff) P1_ridge = wAn.get_maxRidge()

compare with Subheadings 3.2.3 and 3.2.4. Here, providing a parameter T_c within the compute_spectrum function of the wavelet analyzer allows to sinc-detrend and compute the spectrum within a single command. The detrended bioluminescence intensities of Bmal1-ELuc and Per1-luc reporters clearly show differential dynamics of oscillatory phases, i.e., an internal dynamical dissociation, see Fig. 4d. 6. To further quantify this effect, we depict the instantaneous time-dependent periods from the Bmal1-ELuc and Per1-luc reporter as accessed via Bmal1_period = B1_ridge["periods"] Per1_period = P1_ridge["periods"]

in Fig. 4e. It can be seen that the Bmal1-ELuc oscillation period is slower than the Per1-luc period but then speeds up to values smaller than those of Per1-luc. Toward the end of the experiment, both periods approach each other again, indicating a rather transient dynamical dissociation followed by a subsequent resynchronization. 7. The differential period responses naturally translate into differential phase dynamics. We calculate the phase difference between the Bmal1-ELuc and Per1-luc oscillations,

Wavelet Analysis

51

from numpy import arctan2, sin, cos Phasediff

=

arctan2(

sin(P1_ridge["phase"]

-

B1_ridge

["phase"]), cos(P1_ridge["phase"] - B1_ridge["phase"]) )

using a distance metric that accounts for the cyclic nature of phase variables as previously described [22]. It can be seen that the initial large phase gap between Per1-luc and Bmal1-ELuc oscillations evolves toward an antiphasic relationship at around t ¼ 10d and ultimately saturates at a smaller phase difference of about 115 , see Fig. 4f. 8. In conclusion, our wavelet-based time-frequency analysis helps to identify complex differential dynamical features in clock gene expression after SCN slice preparation and in vitro culturing.

4

Notes 1. The Synthetic Signal Generator allows to simulate oscillatory time series signals composed of the superposition of two nonstationary oscillations (so-called “chirps”) with different period behavior, an exponential decay as well as AR1 noise. Setting the AR1 parameter to zero corresponds to Gaussian white noise. 2. The wavelet analysis routine of pyBOAT expects equidistant time series sampling. Gaps in the recording can be interpolated, as described in the next point. 3. If there is missing data (gaps) in the time series, pyBOAT offers a simple linear interpolation in between existing data points. See the GUI tooltip for the “Set missing values entry” for the set of default characters encoding missing data (e.g., “NaN”) or define your own. Note that stretches of missing data at the beginning or end of a signal can only be interpolated to constant values. 4. The sinc detrending filter as implemented in pyBOAT acts like a step function in the period domain, i.e., period components of a signal that are below a certain threshold or cutoff-period are neither attenuated nor amplified while period components above the threshold can be related with the trend of the signal. Since the sinc filter has a nonzero roll-off in practical implementations for finite time series length, one has to carefully choose the threshold. Here, we have chosen a cutoff period of 48 h since it is significantly above the expected oscillatory time scale of ~24 h and thus detrending does not perturb the oscillatory properties of the signal, while it keeps the filter “flexible” enough to reliably remove the nonlinear baseline trend.

52

Christoph Schmal et al.

5. The choice of the mother wavelet function has a strong impact on the outcome of the analysis regarding the absolute power values. pyBOAT uses Morlet mother wavelets as a default which is one of the most widely used mother wavelet and known to fit sinusoidal-like signals well. 6. The main oscillatory component extracted via the wavelet ridge is strongly linked to the results obtained by the Hilbert transform, which is another commonly used nonstationary signal analysis approach. The Hilbert transform however is very noise vulnerable and generally requires a pre-smoothing of data obtained experimentally. In addition, the phase extracted via the Hilbert transform is different from results obtained by a wavelet analysis as it is waveform dependent [23]. 7. In order to decide on a meaningful power threshold that divides the background noise from the signal components of interest, one needs a good null model for the background noise spectrum (see also Note 12). In case of a white noise null model, the (period or frequency independent) threshold at a 95% confidence level is three [8], i.e., parts of the signal with a wavelet power larger than this can be assumed as statistically significant oscillations. However, the background spectrum of biological signals can significantly deviate from the white noise model due to correlations present. In case a reasonable null model is missing, one can estimate an empirical background spectrum from the experimental data itself (see also Note 12). For the sake of consistency and reproducibility, it is important to keep and report the same threshold for the whole analysis, using a given experimental setup. 8. pyBOAT’s underlying mathematical analysis is based on convolutions, which inherently display edge effects visualized by the cone of influence (COI). For very short signals it is possible that the entire ridge is inside the COI. As shown in detail in Mo¨nke et al. [8], the phase, power, and amplitude estimates are unreliable in these cases. However, period estimates show only very minor deviations. As a rule of thumb, the signal should have a length of at least three oscillations. 9. For repetitive analysis of similar data sets one can fix the default analysis parameters (e.g., the sampling interval or cutoff period) via the “Settings” menu entry of the main window. 10. PER2::LUC oscillations show a strong decline in amplitude after TTX application. We have selected a relatively low sinc filter cutoff value of 36 h that is close to the expected dominant period of ~24 h. Thus, we have chosen a compromise between potential mild perturbations of the main oscillatory component (see also Note 4) and the ability to reliably detrend the signal around the sudden jumps in oscillatory amplitude due to TTX treatment (compare Fig. 3).

Wavelet Analysis

53

11. Analogously to the GUI of pyBOAT one can provide two parameters for the determination of the maximum power ridge, i.e., wAn.get_maxRidge(power_thresh=10, smoothing_wsize=20)

While power_thresh gives a minimum wavelet spectrum power for which the ridge is determined, smooting_wsize provides the window size of the Savityky-Golay filter for ridge smoothing. 12. If it is possible to record known non-oscillatory signals within the same experimental setting (e.g., a nuclear fluorescent marker), pyBOAT can show the time-averaged wavelet power distribution for the whole ensemble (Batch Processing -> Fourier Spectra Distribution). The powers of this empirical background spectrum allow for a good estimate of a sensitive ridge power threshold to robustly detect oscillations. As shown in Mo¨nke et al. [8], the minimal power required to statistically classify as “oscillation” is three times the background power.

Acknowledgments We gratefully acknowledge Daisuke Ono, Sato Honma, Ken-ichi Honma, John Abel, and Erik Herzog for sharing experimental data as well as Carolin Ector for comments on our manuscript. Funding: Christoph Schmal’s research was supported by the Deutsche Forschungsgemeinschaft (DFG) through grant SCHM3362/2-1. Gregor Mo¨nke’s research was supported by the EMBL Interdisciplinary Postdoc Programme (EIPOD) under Marie Slodowska-Curie Actions COFUND grant number 664726. Adria´n E. Granada’s research was supported by the German Federal Ministry for Education and Research (BMBF) through the Junior Network in Systems Medicine, under the auspices of the e:Med Programme (grant 01ZX1917C). The funders had no role in study design, data collection and analyses, decision to publish, or preparation of the manuscript.

Conflicts of Interest The authors declare no conflict of interest. References 1. Abel JH et al (2016) Functional network inference of the suprachiasmatic nucleus. Proc Natl Acad Sci U S A 113(16):4512–4517 2. Saini C et al (2013) Real-time recording of circadian liver gene expression in freely moving mice

reveals the phase-setting behavior of hepatocyte clocks. Genes Dev 27(13):1526–1536 3. Gabriel C et al (2021) Live-cell imaging of circadian clock protein dynamics in CRISPRgenerated knock-in cells. Nat Commun 12:3796

54

Christoph Schmal et al.

4. Evans JA, Leise TL, Castanon-Cervantes O, Davidson AJ (2013) Dynamic interactions mediated by nonredundant signaling mechanisms couple circadian clock neurons. Neuron 80(4):973–983 5. Moore A, Zielinski T, Millar AJ (2014) Online period estimation and determination of rhythmicity in circadian data, using the BioDare data infrastructure. Methods Mol Biol 1158:13–44 6. Price TS, Baggs JE, Curtis AM, FitzGerald GA, Hogenesch JB (2008) WAVECLOCK: wavelet analysis of circadian oscillation. Bioinformatics 24(23):2794–2795 7. Leise TL (2013) Wavelet analysis of circadian and ultradian behavioral rhythms. J Circadian Rhythms 11(1):5 8. Mo¨nke G, Sorgenfrei F, Schmal C, Granada A (2020) Optimal time frequency analysis for biological data pyBOAT. bioRxiv 179(4568):985–986. https://doi.org/10. 1101/2020.04.29.067744 9. Schmal C, Herzog ED, Herzel H (2018) Measuring relative coupling strength in circadian systems. J Biol Rhythm 33(1):84–98 10. Abraham U, Granada AE, Westermark PO, Heine M, Kramer A, Herzel H (2010) Coupling governs entrainment range of circadian clocks. Mol Syst Biol 6(1):438 11. Pikovsky A, Rosenblum M, Kurths J (2001) Synchronization. Cambridge University Press, Cambridge 12. Eldar A, Elowitz MB (2010) Functional roles for noise in genetic circuits. Nature 467(7312): 167–173 13. H€anggi P (2002) Stochastic resonance in biology: how noise can enhance detection of weak signals and help improve biological information processing. ChemPhysChem 3(3):285–290 14. Ko CH, Takahashi JS (2006) Molecular components of the mammalian circadian clock. Hum Mol Genet 15(Suppl. 2):271–277

15. Preitner N et al (2002) The orphan nuclear receptor REV-ERBα controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 110(2):251–260 16. Reddy AB, Field MD, Maywood ES, Hastings MH (2002) Differential resynchronisation of circadian clock gene expression within the suprachiasmatic nuclei of mice subjected to experimental jet lag. J Neurosci 22(17): 7326–7330 17. Kiessling S, Eichele G, Oster H (2010) Adrenal glucocorticoids have a key role in circadian resynchronization in a mouse model of jet lag. J Clin Invest 120(7):2600–2609 18. Myung J, Hong S, Hatanaka F, Nakajima Y, De Schutter E, Takumi T (2012) Period coding of Bmal1 oscillators in the suprachiasmatic nucleus. J Neurosci 32(26):8900–8918 19. Ono D, Honma S, Nakajima Y, Kuroda S, Enoki R, Honma K (2017) Dissociation of Per1 and Bmal1 circadian rhythms in the suprachiasmatic nucleus in parallel with behavioral outputs. Proc Natl Acad Sci U S A 114(18): E3699–E3708 20. Nishide S, Honma S, Honma K-I (2018) Two coupled circadian oscillations regulate Bmal1ELuc and Per2-SLR2 expression in the mouse suprachiasmatic nucleus. Sci Rep 8(1):1–12 21. Schmal C et al (2019) Weak coupling between intracellular feedback loops explains dissociation of clock gene dynamics. PLOS Comp Biol 15(9):e1007330 22. Schmal C, Myung J, Herzel H, Bordyugov G (2017) Moran’s I quantifies spatio-temporal pattern formation in neural imaging data. Bioinformatics 33(19):3072–3079 23. Kralemann B, Cimponeriu L, Rosenblum M, Pikovsky A, Mrowka R (2008) Phase dynamics of coupled oscillators reconstructed from data. Phys Rev E Stat Nonlin Soft Matter Phys 77(6 Pt 2):066205

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), 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 chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Chapter 4 Mathematical Modeling in Circadian Rhythmicity Marta del Olmo, Saskia Grabe, and Hanspeter Herzel Abstract Circadian clocks are autonomous systems able to oscillate in a self-sustained manner in the absence of external cues, although such Zeitgebers are typically present. At the cellular level, the molecular clockwork consists of a complex network of interlocked feedback loops. This chapter discusses self-sustained circadian oscillators in the context of nonlinear dynamics theory. We suggest basic steps that can help in constructing a mathematical model and introduce how self-sustained generations can be modeled using ordinary differential equations. Moreover, we discuss how coupled oscillators synchronize among themselves or entrain to periodic signals. The development of mathematical models over the last years has helped to understand such complex network systems and to highlight the basic building blocks in which oscillating systems are built upon. We argue that, through theoretical predictions, the use of simple models can guide experimental research and is thus suitable to model biological systems qualitatively. Key words Oscillations, Clocks, Modeling, Ordinary differential equations, Limit cycles, Bifurcations, Nonlinearities, Feedback loops, Coupled oscillators, Synchronization, Entrainment

1

Introduction Physiological rhythms are central to life. Some rhythms appear during certain phases in an individual’s life, like the somite clock during embryonic development, and some others, like circadian clocks, are maintained throughout life. Understanding the mechanisms of physiological rhythms requires an approach that integrates mathematics and physiology. Of particular relevance is a branch of mathematics called nonlinear dynamics [1, 2]. Dynamics is the subject that deals with change, with systems that evolve throughout time. Whether the system in question settles down to an equilibrium, keeps repeating in cycles or does something more

Supplementary Information The online version contains supplementary material available at (https://doi. org/10.1007/978-1-0716-2249-0_4). Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_4, © The Author(s) 2022

55

56

Marta del Olmo et al.

complicated, it is the theory of nonlinear dynamics what we use to analyze the behavior. The roots of nonlinear dynamics were set by Henri Poincare´ at the end of the nineteenth century, but have seen remarkable development over the past 50 years, especially in the application to biological systems. The development of theoretical models in biology is not a recent field. Conceptual models have been investigated, for example in population biology, long before the first genes were discovered. Nevertheless, with the molecular biology revolution of the 1980s, many new examples of gene regulatory or proteininteraction networks came to light. With so many feedback and feed-forward loops underlying the complex dynamics of cellular processes, many questions became impossible to understand with intuitive reasoning; and thus, mathematical models gained popularity. Especially in the context of oscillations and clocks, which are the type of processes in which mathematical models give a good chance not only to describe them, but also to understand them. Through numerical simulations, models can highlight the role of key parameters in oscillations and can be used to predict the system’s behavior in conditions that have not yet been experimentally tested. Mathematical models can also help in grasping the dynamic properties of molecular mechanisms that are responsible for the generation of robust oscillations, both at the cellular and intercellular level. They even provide tools to artificially construct biological networks that can aid in understanding the design principles of biochemical oscillating systems. Elowitz and Leibler were pioneers on this and designed an oscillating network termed “repressilator” in Escherichia coli [3]. In this chapter we address how circadian clock models can be developed and what insights they provide. We start by introducing some important terms in Subheading 2 and exemplify the concepts with a simple generic oscillator model, the Goodwin model. We then apply the logic of the Goodwin model to understand how more complex models have been developed in the context of circadian clocks, and what mathematical biologists have learned from such models (Subheading 3). But circadian clocks do not exist in isolation; they are subjected to a number of inputs (light, feeding cues, etc.) and they also govern physiological output responses. Thus, in Subheading 4 we overview the interaction of clocks with their environment. We end by summarizing the main points and reviewing modeling limitations. Throughout the chapter there are 9 boxes with practical examples. Scripts for the analyses are provided as Supplementary Material. More extensive introductions to mathematical modeling, however, can be found in the excellent textbooks of Glass and Mackey [1], Kaplan and Glass [4], Segel [5], Murray [6], Goldbeter [7], Ingalls [8] and Jackson [9].

Mathematical Modeling in Circadian Rhythmicity

57

2 Clock Modeling Fundamentals: Mathematical Preliminaries, Notations and Basic Concepts Designing a model requires some understanding of the system of interest. It is necessary to gather and summarize information on what are the components and key interactions. Because biological systems are typically of great complexity, it is important to differentiate between essential and superfluous variables. In this sense, drawing a scheme of the system of interest is often helpful, even before formulating equations. Putting all the known information on a single picture helps us clarifying the nature of the interactions, and to order the different molecular processes. Contrary to many expectations, this first step in the process of building a model is often the most time-consuming. Although the mathematical and computer tools that are used to simulate models are standard, there is no consensus on how to construct a model. This always requires some considerations and assumptions, and a number of questions naturally arise when we have to write the equations. What are the key variables? How many equations should be considered? What kind of equations? Are all kinetic constants (model parameters) known? If not, how can we set them? Modelers have to make choices that depend, first of all, on the biological question to be answered, but also on personal tastes and experiences. Simple generic models are useful to study general properties of circadian rhythms, such as coupling of large oscillator ensembles [10, 11], entrainment of clocks to external Zeitgebers [12–14] or the role of positive feedback loops in the generation of oscillations [15]. On the other hand, if the focus is to understand the molecular details, more complex models with a larger number of variables are generated. A number of detailed models are now available for the circadian clock in mammals [16– 19], Neurospora [20–22] or Drosophila [23–25], among other organisms. The model, be it relatively simple, with just a few variables, or in contrast very complex, will be a precise representation of what we believe to be true. The modeler’s task, as the mathematical biologist Tyson says, is “to determine whether it is a good or useful representation of [that] truth” [26]. 2.1 Ordinary Differential Equations (ODEs)

Most circadian clock models are described with ordinary differential equations (ODEs), which take the form dx ¼ f ðx, y, zÞ: dt

ð1Þ

ODEs are often used to describe how a dynamical system changes over time [1, 2, 8, 26]. The function f(x, y, z) describes the rate of change of a variable x(t) as a function of (potentially) all the

58

Marta del Olmo et al.

variables x, y, z. . . that define the dynamical system. For example, x(t) might stand for the concentration of a given protein, whose evolution depends on a number of time-dependent variables (amount of mRNA) and on time-independent parameters that have a physical interpretation (rate of synthesis, degradation, modification, complex formation, transport, etc.).

Box A Formulating a simple ODE The concentration of a certain protein (let us call it variable y) is controlled by synthesis and degradation. Even though protein production processes are highly complex and might be modulated by ribosome and tRNA availability, a simple way to model production is with a first order reaction, in which the absolute production rate of protein y is proportional to the mRNA abundance (let us call it x). In the same lines, degradation processes can be assumed to follow first order kinetics (absolute degradation rate proportional to the protein abundance), although there might be additional regulatory mechanisms. The mathematical translation of the protein concentration ( y) differential equation is:

dy ¼ px  dy dt

ð2Þ

a linear first-order ODE with two parameters: production rate p and degradation rate d. We can read this equation as follows: the change of protein y over time (dy dt ) is equal to its production (proportional to the mRNA abundance x and the protein synthesis rate p) minus its degradation (proportional to the protein abundance y and the degradation rate). (Note the negative sign in front of the degradation term, since it contributes to the removal of protein y).

Using standard mass action and enzymatic kinetics, we can convert the network diagram that we have drawn from the biological system into a set of ODEs. In such equations, concentrations of variables (e.g., the reactant species) are associated with rates of biochemical reactions (transcription, translation, degradation, phosphorylation, etc.). The equations can then be solved numerically, this is, by letting the computer work out the implications of the complex feedback and feed-forward loops in the network, without having to solve the system analytically (i.e., by hand). Many programming languages have built-in strategies that allow numerical solving of ODEs, such as odeint (as part of the scipy library) in Python, ode (as part of the deSolve package) in R, or ode45 in Matlab, among others.

Mathematical Modeling in Circadian Rhythmicity

59

Box B Goodwin model for circadian clocks, part I—Scheme and ODEs One of the simplest and most famous ODE-based oscillator models is the one imagined by Goodwin [27]. In 1965, when Goodwin developed his model, the molecular mechanisms of circadian clocks were not yet known. He proposed the model as a prototypical biomolecular oscillator. The Goodwin model is based on a delayed negative feedback loop, where the final product of a 3-step chain of reactions inhibits the production of the first component (Fig. 1a, b). In the context of circadian rhythms, the model is interpreted as follows: a clock gene mRNA (x) produces a clock protein ( y) that, in turn, activates a transcriptional inhibitor (z) that represses the synthesis of the x mRNA, closing the negative feedback loop. This generic model can be seen as the minimal backbone for circadian oscillations, as it accounts for the negative feedback exerted by PER and CRY proteins in their own genes. The Goodwin model, later refined by Gonze [10], is still used today to describe fundamental properties of the core circadian oscillator [15, 28, 29] or the synchronization of an ensemble of coupled circadian oscillators [10, 30, 31].

Fig. 1 Goodwin model for self-sustained circadian oscillations. (a) Scheme of the model. (b) Ordinary differential equations of the model: three variables are considered and account for a clock mRNA (x) that produces a clock protein ( y) which activates a transcriptional inhibitor (z). Production reactions are modeled with mass action kinetics; degradation reactions are modeled assuming Michaelis Menten kinetics; repression is modeled with a Hill equation

2.2

Limit Cycles

Most circadian clock models generate stable limit cycle oscillations. Limit cycles are isolated trajectories characterized by a given period and amplitude [1, 2, 32]. Isolated means that neighboring trajectories are not closed, and they spiral either towards or out of the limit cycle [2]. If all neighboring trajectories approach the limit cycle, we say that the limit cycle is stable or attracting (Fig. 2a). This way, a small perturbation that pushes the system out of a stable limit cycle will eventually dampen out, and the trajectory of the perturbed variable will be “attracted” back to the stable limit cycle (red and blue curves in Fig. 2a). In the terms of nonlinear

60

Marta del Olmo et al.

Fig. 2 Limit cycle oscillations (a) and conservative oscillations (b) in phase space. Limit cycles are isolated trajectories and thus any perturbation that pushes the system out of the cycle (red or blue curves) will dampen out, and the system will asymptotically return to the limit cycle (thick black line). Conservative oscillations, on the other hand, are not isolated. As a consequence, there is no damping and, after perturbation, the amplitude of the oscillation changes and does not recover its initial value

dynamics, stable limit cycles represent a type of attractor, since any perturbation will asymptotically return back to the limit cycle with time. Note that not all oscillations are limit cycles: some, like those idealized by the pendulum, represent another type of oscillators (conservative oscillators), in which neighboring trajectories are closed. In these types of oscillations, unlike in limit cycles, amplitude depends on initial conditions (Fig. 2b) [2]. Limit cycle oscillations, in this sense, ensure robustness to small perturbations in the environment. Limit cycles exhibit self-sustained oscillations, this is, they intrinsically oscillate, even in the absence of external periodic cues. This self-sustained oscillating nature is often observed among biological systems [2, 7, 33]. Phenomena such as heart beats, circadian clocks or neuronal activity are just some biological limit cycle oscillations among countless examples [7]. In each case, if the system is slightly perturbed, it always returns to standard cycle. 2.2.1 Cooking Recipe for Oscillations

For limit cycle oscillations to occur, the biological system of interest must fulfill a series of requirements that have been reviewed by Ferrell, Gonze and Tyson in [32, 34–36], among other theoreticians. 1. First of all, a negative feedback is necessary to carry the reaction network back to the point where the oscillation started [34, 35, 37].

Mathematical Modeling in Circadian Rhythmicity

61

2. Second, the negative feedback signal must be sufficiently delayed in time so that reactions do not settle on a stable steady state. This time delay can be achieved by explicitly introducing time delays in the equations (the so-called “delay-differential equations”), or by a long chain of reaction intermediates. The more variables in the loop, the longer the time delay is, and the “easier” it is to generate oscillations [34, 38]. 3. In addition, nonlinear kinetic processes must present in the system to destabilize the steady state. Such nonlinear processes are also commonly referred to as switches [22, 39] or ultrasensitive processes [40–42], and they help to keep the system away from the stable steady state. In Eq. 2, both production and degradation terms of protein concentration were linear; in Eqs. 3, however, degradation of all variables was assumed to follow nonlinear Michaelis Menten kinetics. Phosphorylation, active transport, cooperative binding, sequestration, or other enzymatic events are commonly described by nonlinear terms, such as Michaelis Menten- or Hill-like kinetics. These terms often provide the necessary source of nonlinearity [43–45]. 4. Lastly, the system must be open; this is, it must be equipped with dissipative mechanisms (e.g., degradation processes) and sources of energy (e.g., mRNA or protein synthesis), so that oscillations that grow too large are dampened out, and oscillations that become too small are pumped up [38]. The Goodwin model presented in Box B includes these four “ingredients” and thus, self-sustained oscillations can be expected for proper choice of parameter values.

Box C Goodwin model for circadian clocks, part II—Limit cycle oscillations The original Goodwin model [27] only contains one nonlinear term, which is given by the Hill equation used to model the z-mediated repression of x. Griffith demonstrated that the Hill coefficient had to be sufficiently large (n ¼ 8) for the model to generate self-sustained oscillations [46]. But, since such high Hill coefficients are not biologically meaningful, the original model was modified by Gonze and others by including additional nonlinearities. This reduced the need of such a high Hill coefficient [10, 15, 45]. In the modified Goodwin model represented in Fig. 1, there are two sources of nonlinearities: the repressive Hill term and the Michaelis Menten-like kinetics assumed for degradation processes [10]. Numerical integration of the Goodwin model equations (Fig. 1b) over time, with an appropriate parameter choice, can result in self-sustained limit cycle oscillations (Fig. 3). We can plot the solution as time series, in which we illustrate how the concentration of the different species change over time (Fig. 3a), or in phase space, which is the illustration of the space of all possible states (Fig. 3b).

(continued)

62

Marta del Olmo et al.

Fig. 3 Limit cycle circadian oscillations of the Goodwin model. Limit cycle oscillations, plotted as time series (a) or in phase space (b). Results were obtained by numerical integration of the equations in Fig. 1b for the following parameter values: ν1 ¼ 0.70 nM/h, ν2 ¼ 0.45 nM/h, ν3 ¼ 0.70 h1, ν4 ¼ 0.35 nM/h, ν5 ¼ 0.70 h1, ν6 ¼ 0.35 nM/h, K1 ¼ 1 nM, K2 ¼ 1 nM, K4 ¼ 1 nM, K6 ¼ 1 nM, n ¼ 7. Oscillations were normalized to their mean

2.3 Bifurcation Diagrams

Another important term that frequently appears in the field of theoretical chronobiology is that of bifurcation diagrams. To understand this concept, we have to be aware of how the dynamics of a system greatly depend on parameter values. Unfortunately, kinetic rates and equilibrium constants are not measured experimentally in many cases, so this poses a challenge for constructing the model. Although a model with several parameters leads to a combinatorial number of possible parameter values, there are many constraints that fortunately allow us to narrow down the range of suitable values. In the context of circadian clocks, examples of such constraints might be given by the oscillation period (which needs to be circadian), by phase relationships of variables (if known), by effects of some mutations (assuming that we know which parameters are affected) or from some biochemical constants known from in vitro experiments. The unknown parameters that have to be “guessed” should be chosen within realistic physiological ranges. Usually, when a parameter value changes, the characteristics of the limit cycle (i.e., period, amplitude, phase relationship between variables. . .) also change. These changes can be illustrated in bifurcation diagrams. Novak and Tyson make the analogy of bifurcation diagrams being for modelers what signal-response curves are for experimentalists [26]. In a physiology experiment, biologists measure how some behavior of the cell (e.g., oscillation amplitude or period) depends on the value of an experimentally controlled signal

Mathematical Modeling in Circadian Rhythmicity

63

(e.g., the concentration of a certain synchronizing factor in culture medium). The signal is held at a constant value until the response settles on a definitive value. Then, the signal is changed to a new value and the new response is recorded. A one-parameter bifurcation diagram illustrates the same concept. It shows how the final states of a mathematical model (e.g., period or amplitude of oscillations, or most commonly the maximum and minimum of a variable, plotted in the y axis) depend on a control parameter of the model (plotted in the x axis). Variation in parameter values can cause qualitative changes in the long-term behavior of the system. For example, the number of steady states or their stability properties can vary. These qualitative changes in the dynamics are called bifurcations, and the parameter values at which they occur are called bifurcation points [2]. In the context of oscillations, Hopf bifurcations are the most important type of bifurcation point. They occur in dynamic systems when a periodic solution (limit cycle) arises from a stable steady state that loses its stability (Fig. 4). By manually exploring the parameter space (i.e., by analyzing how solutions change as parameter values are varied), we can make predictions of how the model might behave under different conditions, and its sensitivity towards parameter changes.

Box D Goodwin model for circadian clocks, part III—Bifurcation diagrams The bifurcation diagram of a given control parameter can be numerically constructed in the same way as Fig. 3 was built, but iterating this process over a range of parameter values. Thus, for each value of the control parameter, we let the computer solve the set of ODEs (i.e., we simulate the system) and retain the maximum and minimum values (or any other oscillation parameter) reached by a given variable once the system has converged to its stable regime. If the system converges to a steady state, then the minimum and maximum will be indistinguishable and a single point will be plotted on the bifurcation diagram. If, on the contrary, the system oscillates for the simulated parameter value, then two points will be plotted. Such maximum-minimum bifurcation plot is shown in Fig. 4a. We can also build the period bifurcation diagram (Fig. 4b), keeping in mind that the period can only be estimated when oscillations appear. The bifurcation plots show, firstly, that oscillations disappear when the degradation rate is > 0.93 nM/h (this is the Hopf bifurcation point). Moreover, it illustrates that both amplitude and period of x decrease as its degradation rate increases. The model thus predicts that decreasing the degradation rate should lead to oscillations with longer period. In this way, modeling and bifurcation analyses can be used to predict the behavior of the system in conditions that might have not yet been tested experimentally. Of note is that there are other methods, such as those implemented in XPP-AUTO, that allow the calculation of bifurcations without intensive simulation for each parameter value.

(continued)

64

Marta del Olmo et al.

Fig. 4 Bifurcation diagrams of the Goodwin model as a function of the x degradation rate ν2. Effect that changes in ν2 have on maxima and minima (a) or on the period (b) of x dynamics. The diagrams were built numerically by simulating the ODEs for each value of the control parameter ν2 (the rest of the parameters take their default value, given in the caption of Fig. 3) and retaining either the maximum and minimum values of the x oscillations, or the period, once the system converged to its stable regime. Maximum and minimum values or periods are plotted against the control parameter values. When the system converges to a stable steady state, maximum and minimum values are indistinguishable: a single point is plotted in the maximum-minimum bifurcation diagram, and no point is plotted in the period bifurcation plot. Oscillations vanish at ν2 ¼ 0.93 nM/h (Hopf bifurcation point). Red points indicate the default parameter value that produces 24 h circadian oscillations, ν2 ¼ 0.45 nM/h

3

Learnings from Modeling Interlocked Feedback Loops The Goodwin oscillator exemplified in Box B demonstrates that, with a minimum number of “ingredients”, it is possible to generate stable limit cycle oscillations (Fig. 3). But in biological systems, however, the picture is more complex. Over the last years, the discovery of additional clock genes and regulations has led to the realization that the circadian timing system involves multiple sources of nonlinearity and interlocked feedback loops. Consequently, these findings have motivated the development of more detailed molecular models [16–19, 47–49] that contain additional positive and negative feedback loops, which have also been shown to contribute to the generation of stable and robust oscillations. Nevertheless, their structure most of the times relies on a Goodwin-like negative feedback loop.

Mathematical Modeling in Circadian Rhythmicity

65

Box E Positive feedback loops promote oscillations in the Goodwin model There have been many useful refinements of the Goodwin oscillator. It has been shown, for example, that complementary positive feedback loops can enhance the capabilities of rhythm generation [15, 50–52]. However, it is important to note that not all feedback loops (positive or negative) are always explicitly visible. Sequestration (formation of inactive protein complexes), for instance, can also form implicit feedback loops. A prominent example is the sequestration of KaiA molecules in the cyanobacterial clock [53]. Although the majority of models use Hill functions to describe transcriptional repression, some have also introduced protein sequestration-based repression [54]. Ananthasubramaniam et al. showed very elegantly in 2014 that addition of positive feedback loops to the Goodwin model promotes oscillations at lower Hill coefficients. Figure 5 summarizes their findings: a Goodwin-like motif with a Hill coefficient n ¼ 4 (lower n than in the simulations from Fig. 3, where n ¼ 7) cannot oscillate;

A

B

x y z

1.50 1.25 1.00 0.75 0.50 0

120

240

time [h]

360

480

modeling with positive feedback loop

concentration [a.u.]

concentration [a.u.]

modeling without positive feedback loop

x y z

1.50 1.25 1.00 0.75 0.50 0

24

48

72

96

120

time [h]

Fig. 5 A positive feedback loop promotes oscillations in a Goodwin-like motif. (a) Simulation of a Goodwin-like motif with a Hill coefficient n ¼ 4 does not produce self-sustained limit cycle oscillations. Instead, the system approaches a stable steady state. (b) When a self-activating positive feedback loop is included on x, limit cycle oscillations emerge for the same parameter values. Results were obtained by numerical integration of the equations in Fig. 1b for the following parameter values: ν1 ¼ 0.70 nM/h, ν2 ¼ 0.45 nM/h, ν3 ¼ 0.70 h1, ν4 ¼ 0.35 nM/h, ν5 ¼ 0.70 h1, ν6 ¼ 0.35 nM/h, K1 ¼ 1 nM, K2 ¼ 1 nM, K4 ¼ 1 nM, K6 ¼ 1 nM, n ¼ 4. Oscillations were normalized to their mean

(continued)

66

Marta del Olmo et al.

nevertheless, the same Goodwin-like motif with an explicit positive feedback loop on x (which can be interpreted as the BMAL1-Ror loop) generates limit cycle oscillations. In the study, the authors highlighted additional mechanisms that may facilitate the emergence of oscillations, such as cross-activation (explicit feedback loop) or Michaelis Menten degradation (implicit feedback loop) of variables [15].

In 2011, Relo´gio et al. developed an extensive 19 variable ODE-based model. The Relo´gio system contains clock transcripts, cytoplasmatic and nuclear proteins, either alone or in complex with other clock proteins [19]. It was built from available data on phases and amplitudes of clock components to understand the mechanisms that govern circadian rhythm generation in mammalian cells [19]. It allowed to independently study the roles of the Ror-BMAL1-RevErb and Per2:Cry1 loops, as well as the role of Per2 degradation rate in the dynamics of the system. The authors provided in silico evidence, for the first time, that the Ror-BMAL1RevErb loop could act as an oscillator independently of the Per2: Cry1 loop and they showed that in silico overexpression of RevErbα and RevErbβ resulted in the loss of oscillations [19]. This theoretical prediction was experimentally validated one year after in mouse embryonic fibroblasts [55]. Taken together, the computational findings from the Relo´gio model challenged the role of the Ror-BMAL1-RevErb loop as a merely auxiliary loop and illustrate how models can be used to make predictions. In the same lines, a more recent study from Pett et al. showed that a repressilator motif containing regulated expression of Cry1, Per2 and RevErbα is sufficient to generate 24 h rhythmicity, thus constituting a core loop of the mammalian oscillator [48]. In a later bioinformatic study, the authors proposed that the most essential feedback loops which result in rhythm generation can differ among tissues [56]. It has been suggested that the primary rhythmgenerating loop in adrenal gland and heart is the BMAL1-RevErb loop, whereas self-inhibitions of Per and Cry genes are more characteristic for models of suprachiasmatic nucleus clocks [56]. Of note, though, is that the authors did not use ODEs in their study, but instead delay differential equations, in which time delays are introduced explicitly in the equations. A new simple ODE model of the mammalian circadian clockwork was published by Almeida in the early 2020. It was tailored to identify the essential interactions that are needed to generate phase opposition between the activating CLOCK:BMAL1 and the repressing Per2:Cry1 complexes [49]. Van Soest and colleagues performed extensive bifurcation analyses on the Almeida model and found that changes in degradation rates of clock proteins could generate arrhythmic dynamics. These findings suggested (and predicted) that not only knockout [57, 58] or overexpression

Mathematical Modeling in Circadian Rhythmicity

67

[55] of core clock components can lead to arrhythmicities, but also changes in degradation or transcription activation rates. The in vitro or in vivo significance of such observations, however, remains to be validated. The take-home message from these computational studies is that modeling can help to study alternative network architectures and can in this way guide experimental research. Mathematical models are usually made to answer a specific question, but a big advantage of modeling is that we can then make predictions and ask ourselves additional questions, that can be later tested (and hopefully validated) experimentally.

4

Interaction of Clocks with the Environment The major role of circadian rhythms is to coordinate physiological and behavioral processes with the natural daily variation. To do this, molecular circadian clocks need to integrate signals from the external world (Zeitgebers) and to transmit such signals to the whole organism. Figure 6 illustrates the key paradigm in biological clock research: although clocks are able to tick by themselves, they respond to inputs (Zeitgebers) and perform output responses (coordination of physiology or behavior with external time). In mathematical terms, a free-running clock is an autonomous system. An autonomous system is a system of ordinary differential equations which does not explicitly depend on the independent variable. When the variable is time, like in our case, they are also called time-invariant systems. When mathematical biologists want to model how external signals confer timekeeping information to an autonomous clock, they typically add a time-dependent term on the right-hand side of an ODE. This is commonly referred to as “adding a forcing” or “driving term” to the system, and thus the resulting system is said to be forced, driven or non-autonomous (since now the system depends on time explicitly) (Box F).

Fig. 6 Simplified scheme of circadian clock systems. Circadian clock systems consist of a network of input pathways that integrate external Zeitgeber timing cues, the central oscillator (pacemaker) and output pathways. Central oscillators generate the endogenous rhythm and must be able to synchronize to environmental Zeitgebers (e.g., light, food, temperature) via input pathways. Consequently, pacemakers drive output pathways (e.g., physiology, behavior) and clock-controlled activities by synchronizing downstream oscillators

Marta del Olmo et al.

Box F Amplitude-phase models Amplitude-phase oscillators are one of the most abstract yet intuitive class of models that can relate to any observed rhythm. They describe phenomenologically the dynamics of a system, independently of any molecular details [11, 14, 32]. Such models have been used in clock research to study generic properties of phase response curves [59], entrainment [12, 14], the behavior of ensembles of coupled oscillators [11, 38, 60– 63], or to interpret experimental results [64]. They are described with only two variables, namely radius r and phase ϕ of the oscillation, and thus they do not necessarily account for the levels of a given protein or transcript. The equations, in polar coordinates, read:

dr dt dϕ dt

¼ λrðA 0  rÞ,

ð3Þ

2π ¼ τ

where r and ϕ represent the variables (radius and phase, respectively), and λ, A0 and τ, the parameters (amplitude relaxation rate, oscillation amplitude and period, respectively). The amplitude relaxation rate is a relatively abstract concept that describes how fast a perturbation relaxes back to the limit cycle [11, 12, 38]. Basic calculus allows the transformation pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi of any point in the polar plane into Cartesian coordinates, since r ¼ x 2 þ y 2 and ϕ ¼ arctan ðxy Þ (Fig. 7). Thus, the amplitude-phase model can be converted into Cartesian coordinates, reading:

Y y

r

φ

r cos φ

P (r, φ) Polar P (x, y) Cartesian r sin φ

68

x

X

Fig. 7 Converting between polar and Cartesian coordinates. A point in polar coordinates is characterized by the variables r and ϕ.pPolar coordinates can be converted to the Cartesian ffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi coordinates x and y with r > 0 by r ¼ x 2 þ y 2 (as in the Pythagorean theorem) and ϕ ¼ arctan ðyx Þ

(continued)

Mathematical Modeling in Circadian Rhythmicity

dx dt dy dt

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2π x 2 þ y 2 Þ  y, τ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2π ¼ λyðA 0  x 2 þ y 2 Þ þ x: τ

¼ λxðA 0 

69

ð4Þ

Period and amplitude can easily be calculated from the oscillatory time series. Amplitude relaxation rate can be determined by estimating the rate at which a perturbation decays back to the limit cycle. The reader is encouraged to calculate this parameter for any variable of the Goodwin oscillator described in Fig. 3. So far, the right-hand sides of the amplitude-phase model ODEs do not contain time t. Thus, these equations describe an autonomous system that oscillates by itself, i.e., in the absence of external timing cues. But clocks usually respond to external timekeeping cues, and thus they can be driven (forced) by these signals. Assuming that the forcing is done by a sinusoidal Zeitgeber Z(t) with period T and amplitude F, 2π Z ðtÞ ¼ F cos ð t þ ϕÞ, T

ð5Þ

and that the Zeitgeber Z(t) drives oscillations of x, we can now incorporate the forcing term (Eq. 5) in the right-hand side of the x ODE (Eq. 4) as follows: pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dx 2π ¼ λxðA 0  x 2 þ y 2 Þ  y þ Z ðtÞ, τ dt ð6Þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dy 2π 2 2 ¼ λyðA 0  x þ y Þ þ x: τ dt The system has become forced or non-autonomous, since time t is now explicitly included in the ODEs inside the Z(t) term.

4.1 Coupled Oscillators, Synchronization and Entrainment

Both theoretical and experimental scientists have long been puzzled by the existence of spontaneous order (and thus synchronization) that exists in the universe. The “science of synchronization” centers on the study of coupled oscillators, which are widespread throughout biological systems: groups of fireflies, pacemaker cells or circadian clocks are collection of oscillators in which one can find some “underlying order”. Understanding the basic rules of coupled oscillator theory can help us to gain insights into how coupling results in synchronization or entrainment. Two or more oscillators are said to be coupled if some physical or chemical process allows them to influence one another [2]. Fireflies communicate with light, heart cells exchange electrical currents. . . The result of this mutual influence is often synchrony. When synchrony occurs, oscillators acquire a rational mn ratio, meaning that one oscillator will undergo m cycles in the time in which the second one undergoes n cycles. Some examples can be seen by the 1:4 frequency locking between respiratory and cardiac rhythms in some individuals (i.e., 1 inhalation and exhalation occur for every 4 heart beats) [65, 66], or by the 1:1 synchronization of circadian

70

Marta del Olmo et al.

clocks to the day-night cycle. We can distinguish between intrinsic or multidirectional and extrinsic or unidirectional coupling in biological systems: l

Intrinsic or multidirectional coupling can be seen as the coupling that exists among oscillators without explicit external time information. Here, all oscillators exchange information with one another. For example, coupling among cardiac cells exists to produce a coherent heart beat as output response. In the context of circadian clocks, it is often said that feedback loops that comprise the core clock network are coupled: in principle, different feedback loops are able to oscillate independently [56], but all loops are coupled such that they all oscillate with a 24 h periodicity. Another example of intrinsic coupling occurs clearly in neurons from the suprachiasmatic nucleus (SCN). Although circadian rhythms can be observed on the single cell level [67, 68], synaptic connections, gap junctions and neurotransmitters are believed to couple (and thus synchronize) SCN neurons in a robust manner [68].

l

Extrinsic or unidirectional coupling, on the other hand, requires an explicit periodic signal (Zeitgeber) present in the surrounding of an oscillator, such as the alternation of night and day, feeding-fasting rhythms or tidal rhythms. This external rhythm affects the intrinsic clock, but not the other way around (thus unidirectional coupling). When the intrinsic clock adapts (synchronizes) to the external timing signal, entrainment results.

We have defined some terminology related to coupling between oscillators. In Boxes G and H we illustrate some of the key behaviors that are seen in coupled oscillator systems, namely spontaneous synchronization and entrainment.

Box G Coupled circadian oscillators synchronize spontaneously A remarkable property of circadian rhythms in the SCN is their robust nature. Although the free-running periods of isolated neurons are broadly distributed [68], the SCN as an ensemble oscillates very robustly with a clear periodicity. This indicates that a coupling mechanism must operate between the neurons, which is known to be achieved by periodic neurotransmitter release and synaptic connections [68]. Based on this, it is a reasonable hypothesis to assume global coupling among all oscillators in the SCN, achieved through a mean-field M. The mean-field can be defined as the average concentration of neurotransmitter xi as follows:

M ¼

1 N

XN

i¼1

xi

ð7Þ

The network dynamics of an ensemble of N amplitude-phase models, in Cartesian coordinates, that describes the oscillatory dynamics of N neurotransmitter xi in the presence of mean-field coupling can then be given by (continued)

Mathematical Modeling in Circadian Rhythmicity

dx i dt dy i dt

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2π x 2i þ y 2i Þ  y i þ K coup M , τi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2π ¼ λy i ðA 0  x 2i þ y 2i Þ þ x i : τi

71

¼ λx i ðA 0 

ð8Þ

where the parameter Kcoup denotes the strength of the coupling between the mean field and the single oscillatory units. We assume, in the lines of previous studies [11, 13], that the mean-field M additively couples only to the x coordinate. Note that this is, strictly speaking, not a forced system, since there is no explicit time dependence in the right-hand side of the equations. Figure 8 shows how an ensemble of N ¼ 50 heterogeneous oscillators with periods chosen from a normal distribution with mean μ ¼ 24 h and a standard deviation σ ¼ 1.5 h can spontaneously synchronize when they are coupled. In the absence of

Fig. 8 Spontaneous synchronization of coupled circadian oscillators. 50 heterogeneous amplitudephase oscillators run at their own pace in the absence of coupling (a), but they spontaneously synchronize when coupled through a mean-field M (b). Grey thin lines represent individual oscillators; thick lines (blue, red) represent the signal of the average population (bulk). (c) Distribution of the individual periods in the uncoupled (Kcoup ¼ 0, blue) and coupled (Kcoup ¼ 0.1, red) systems. (d) Coupling leads to higher bulk amplitudes due to resonance. Results were obtained by numerical integration of Eqs. 8, for 100 days and the following parameter values: A0 ¼ 1, λ ¼ 0.03 h1, individual periods τi taken from a normal distribution with mean μ ¼ 24 h and standard deviation σ ¼ 1.5 h, and varying Kcoup values. Bulk amplitudes were calculated as the mean peak-to-trough distance of the average signal (thick lines in panels (a) and (b)) during the last 5 days of simulations

(continued)

72

Marta del Olmo et al.

inter-oscillator coupling (Fig. 8a), each oscillator runs with its free-running period τi, and the average bulk signal (in blue thick line) does not display robust rhythms. When the individual oscillators are coupled through their mean-field, on the other hand, order emerges: oscillators start running at the same pace and locked to the mean-field (Fig. 8b). Consequently, the period distribution of the individual oscillators becomes narrower (Fig. 8c). It is well known from the theory of coupled oscillators that if a periodic stimulus (in this case the mean-field) is of the same or nearly the same frequency as the natural vibrating frequency of a system, the amplitude of the system will increase; a phenomenon called resonance [69]. For a network of oscillators, like in this case, resonance can be interpreted as amplification of the amplitude of individual oscillators. Figure 8d shows precisely this phenomenon: as the coupling strength increases, so does the amplitude of the bulk signal due to resonance effects (to values that are even bigger than that of the individual oscillators, A0 ¼ 1 in the simulations). It is important to mention that the emergent properties of the coupled ensemble depend not only on the characteristics of coupling, but also on the properties of individual oscillators. For example, amplitude relaxation rate λ of individual oscillators is inversely correlated with amplitude resonance: as the oscillator relaxation rate increases, amplitude expansions decrease [11, 12]. The reader is encouraged to analyze through simulations how the curve from Fig. 8d changes in systems with varying λ.

4.1.1 Entrainment and Arnold Tongues

A characteristic property of circadian rhythms is their ability to be synchronized, or entrained, by external Zeitgebers. Thus, although circadian rhythms can persist in the absence of external timing cues, normally such cues are present and rhythms are aligned to them. This alignment is called entrainment, and it occurs when the strength of the Zeitgeber (the “coupling strength”) is capable of overcoming the period mismatch between its period T and the clock’s intrinsic period τ. If this happens, the Zeitgeber will enforce its natural periodicity T on the clock. The range of period mismatches τ–T for which entrainment occurs is called the range of entrainment and it depends on the Zeitgeber strength as well as on the clock’s properties (Box H). It is in fact the difference of both periods τ and T, rather than the single periods per se, what determines whether a clock can be entrained or not for a given Zeitgeber strength [1, 13, 14]. When entrainment occurs, the system adopts a specific phase relationship with the phase of the Zeitgeber, and this difference is known as phase of entrainment Ψ. In more general terms, entrainment is the process in which oscillators synchronize to an external signal at a fixed m to n ratio, and it is common to all systems of coupled oscillators. In the field of circadian clocks, 1:1 entrainment is the common scenario (the period τ of the circadian system adjusts such that it equals the period T of the Zeitgeber); nevertheless, other entrainment ratios might exist under some circumstances. The regions of mn

Mathematical Modeling in Circadian Rhythmicity

73

synchronization can be plotted as Arnold tongues (Box H), named after the mathematician Arnold, who described them in the 1960s. The dynamics are more or less simple at low coupling: tori (limit cycles of two frequencies that are not locked) and some zones of synchronization (and thus period- and phase-locking) dominate the parameter space. Higher coupling increases the regions in which the tongues (and thus synchronization) exist, but can also lead to more complex behavior, including chaos. Arnold tongues are generic to coupled oscillators and there is a vast amount of literature on their theory and the results of mathematical modeling [1, 13, 14, 70–72].

Box H Entrainment and Arnold tongue of a circadian amplitude-phase model We now compute the Arnold tongue of a circadian amplitude-phase oscillator driven by a sinusoidal Zeitgeber (Eqs. 5 and 6) and explore which combinations of Zeitgeber strength F and Zeitgeber period T lead to entrainment. From the theory of coupled oscillators we know that sufficiently strong Zeitgebers (with high “unidirectional” coupling strength F) can entrain oscillators even if the period difference between the intrinsic oscillator and the Zeitgeber (i.e., the period mismatch τ–T) is large [1, 2, 12– 14, 69, 70, 72]. The tongue indeed shows that the range of entrainment increases with Zeitgeber strength F (Fig. 9a). Nevertheless, not only the coupling strength F, but also the intrinsic oscillator properties can affect the entrainment range. Stronger oscillators with high amplitude relaxation rates λ display narrow ranges of entrainment, whereas weaker oscillators (lower λ) have wider entrainment ranges. The reader is encouraged to compute the Arnold tongues of an amplitude-phase oscillator with changing values of λ. These principles explain experimental findings, namely that peripheral clocks in the lung entrain to extreme Zeitgeber cycles, while SCN clocks do not [12]. The simulations reproduce the observations that the phase of entrainment Ψ increases (i.e., the intrinsic clock becomes later) with period mismatch τ–T (Fig. 9a) [14]. These theoretical observations can be translated into biological words and associated to the spread of chronotypes. Under natural conditions of T ¼ 24 h, variations of intrinsic period τ lead to different phases of entrainment: short endogenous periods τ often lead to early phases or entrainment (“morning larks”), whereas longer periods τ correspond to later phases (“night owls”) [13, 14, 73, 74]. The strength of a Zeitgeber has also been suggested to modulate the phase of entrainment [75]. If we move vertically along the τ > T region of the tongue (red vertical line in Fig. 9a), we see how an increase in the Zeitgeber strength results in earlier entrainment phases. On the other hand, stronger Zeitgebers can lead to later phases Ψ within the τ < T region (blue vertical line in Fig. 9a). Thus, increasing Zeitgeber strength can lead to both decrease or increase of Ψ depending on the mismatch τ–T (Fig. 9b). Experimental predictions can relate light intensity and chronotypes in the base of these observations: we expect that more light leads to earlier entrainment phases for night owls with τ > T, but later values of Ψ for morning larks with τ < T. Taken together, these theoretical results predict that strong Zeitgebers should lead to narrower distributions of chronotypes [13, 14, 76].

(continued)

74

Marta del Olmo et al.

Fig. 9 Entrainment of an amplitude-phase oscillator to an external sinusoidal Zeitgeber. (a) Arnold tongue for a circadian amplitude-phase oscillator driven by a Zeitgeber, where phases of entrainment Ψ are color-coded. (b) Phase of entrainment in dependence on the Zeitgeber strength F along the vertical blue and red lines in (a). Results were obtained by numerical integration of Eq. 6 for the following parameter values: A0 ¼ 1, τ ¼ 24 h, λ ¼ 0.01 h1, Zeitgeber strengths F ranging from 0 to 0.1 and Zeitgeber periods T ranging from 19 h to 29 h

4.2 Output Regulation

So far we have focused on the ability of clocks to free run (exhibit self-sustained limit cycle oscillations, Subheading 2) and on how external signals impinge on the oscillator (Subheading 4.1). Nevertheless, clocks also regulate a myriad of output signals. In fact, transcriptomic studies have shown that the expression of almost 10% of all genes in peripheral tissues is regulated rhythmically (despite small overlap between tissues) [56, 77–79]. Modeling can also help in studying properties of the expression of clockcontrolled genes, as described in [80].

Box I Modeling driven expression of clock-controlled genes In many cases, transcription factors like BMAL1 activate the expression of the so-called clock-controlled genes, and this can lead to the following model generalization:

dx 2π ¼ pð1 þ F sin ð tÞÞ  dx, τ dt

ð9Þ

where the production rate p of an mRNA x is periodically driven by a core clock element with an amplitude F and a period τ of about 24 h. This (non-autonomous) equation can be solved numerically (letting the computer run) or analytically, and the (continued)

Mathematical Modeling in Circadian Rhythmicity

75

A

mRNA abundance

solution will oscillate periodically around its mean (dp , unless normalized like in Fig. 10a). Some interesting insights emerge from these simulations: first, amplitude and, second, phase of the clock-controlled transcript depend strongly on its own halflife (Fig. 10b, c). Short-lived transcripts display large amplitudes and are almost in phase with the transcriptional modulator. Long lived genes have larger delays (approaching 6 h) but smaller amplitudes as lifetimes increase. Indeed, such dependencies have been found for many clock-controlled genes [81].

clock element

1.2

CCG1, d=1.0h

1

CCG2, d=0.15h

1.0

1

0.8 0

24

48

72

96

time [h]

C

0.4

phase delay [h]

CCG amplitude

B

0.3 0.2 0.1 0

5

Half life [h]

10

4

2

0 0

5

10

Half life [h]

Fig. 10 Modeling the expression of clock-controlled genes. (a) Oscillations in mRNA abundance of two clock-controlled genes with different half-lives. Amplitude (b) and phase delay (c) of the driven clock-controlled gene depend on its half-life. Results were obtained by numerical integration of Eq. 9 for the following parameter values: p ¼ 1 (units of concentration/h), F ¼ 0.20 and τ ¼ 24 h. Oscillations were normalized to their mean. The amplitude values depicted in panel (c) were calculated as peak-totrough distances

76

5

Marta del Olmo et al.

Concluding Remarks and Modeling Limitations We have seen how two simple and generic models, namely the Goodwin and an amplitude-phase model, properly reproduce core features of the circadian clock, notably its self-sustained nature, its response to parameter variations, to coupling and to Zeitgeber entrainment. These “toy models” give hints, ideas and speculative explanations, but they are also subjected to several caveats. The most common criticism towards models is that the type of equations and the model parameters are (mostly) arbitrary. Whereas molecular models are empirically based on well-established genetic regulations, the quantitative details of the molecular mechanisms are usually unknown. For instance, the choice of Michaelis Menten or Hill-like functions are realistic representations of enzymatic processes (and they account for the necessary degree of nonlinearity that models need to oscillate) but the hypotheses underlying these approximations are not always satisfied. Thus, theoretical models like the ones presented in this chapter should be regarded as semiquantitative and phenomenological models. Simple models usually do not allow the investigation of quantitative details of physiological processes, but they allow to study qualitatively the dynamic properties of oscillating systems. Second, most circadian clock models are based on ODEs. These models, as well as their stochastic versions, only account for the regulation of physiological responses in time, and neglect the aspects in space. They assume that the underlying molecular mechanisms occur in well-stirred reaction vessels, and that the variables move freely around the cell. But eukaryotic cells are far from being well-stirred reaction vessels. Cells are very crowded spaces and cellular processes are not only organized in time but also in space. They are divided into compartments, which might need to be modeled individually to take into account space and diffusion, two variables that likely play critical roles in the dynamics of cellular systems. We must be aware that none of the models, as detailed as they may be, bring really definitive answers. Rather, they provide elements for reflection. For example, the role of positive feedback loops in the molecular mechanism of circadian clocks is not fully elucidated yet. But modeling provides us with clues to possible functions of these additional loops: increasing robustness to parameter variations, allowing period tunability, etc. But that is in fact the beauty of simple models: they provide us with additional perspectives of a system and allow the constant self-formulation of new questions. In the words of the great Albert Einstein, and invoking Occam’s razor “everything should be made as simple as possible, but not simpler”.

Mathematical Modeling in Circadian Rhythmicity

77

References 1. Glass L, Mackey MC (1988) From clocks to chaos. Princeton University Press, Princeton 2. Strogatz SH (2000) Nonlinear dynamics and chaos: with applications to physics, biology, chemistry and engineering. Westview Press, Cambridge 3. Elowitz MB, Leibler S (2000) A synthetic oscillatory network of transcriptional regulators. Nature 403:335–338 4. Kaplan D, Glass L (1995) Understanding nonlinear dynamics. Springer, New York 5. Segel LA (1984) Modeling dynamic phenomena in molecular and cellular biology. Cambridge University Press, Cambridge 6. Murray JD (2002) Mathematical biology I: an introduction, vol 17. Interdisciplinary applied mathematics. Springer, New York 7. Goldbeter A (1996) Biochemical oscillations and cellular rhythms. The molecular bases of periodic and chaotic behavior. Cambridge University Press, Cambridge 8. Ingalls BP (2013) Mathematical modeling in systems biology: an introduction. MIT Press, Cambridge 9. Jackson E (1989) Perspectives of nonlinear dynamics 1. Cambridge University Press, Cambridge 10. Gonze D, Bernard S, Waltermann C, Kramer A, Herzel H (2005) Spontaneous synchronization of coupled circadian oscillations. Biophys J 89:120–129 11. Schmal C, Herzog ED, Herzel H (2018) Measuring relative coupling strength in circadian systems. J Biol Rhythms 33:84–98 12. Abraham U, Granada AE, Westermark PO, Heine M, Kramer A, Herzel H (2010) Coupling governs entrainment range of circadian clocks. Mol Syst Biol 6:438 13. Bordyugov G, Abraham U, Granada AE, Rose P, Imkeller K, Kramer A, Herzel H (2015) Tuning the phase of circadian entrainment. J R Soc Interface 12:20150282 14. Granada AE, Bordyugov G, Kramer A, Herzel H (2013) Human chronotypes from a theoretical perspective. PLoS One 8:e59464 15. Ananthasubramaniam B, Herzel H (2014) Positive feedback promotes oscillations in negative feedback loops. PLoS One 9:e104761 16. Becker-Weimann S, Wolf J, Herzel H, Kramer A (2004) Modeling feedback loops of the mammalian circadian oscillator. Biophys J 87: 3023–3034 17. Forger DB, Peskin CS (2003) A detailed predictive model of the mammalian circadian

clock. Proc Natl Acad Sci USA 100: 14806–14811 18. Leloup JC, Goldbeter A (2003) Towards a detailed computational model for the mammalian circadian clock. Proc Natl Acad Sci USA 100:7051–7056 19. Relo´gio A, Westermark PO, Wallach T, Schellenberg K, Kramer A, Herzel H (2011) Tuning the mammalian circadian clock: robust synergy of two loops. PLoS Comput Biol 7: e1002309 20. Francois P (2005) A model for the Neurospora circadian clock. Biophys J 88:2369–2383 21. Hong CI, Jolma IW, Loros JJ, Dunlap JC, Ruoff P (2008) Simulating dark expressions and interactions of frq and wc-1 in the Neurospora circadian clock. Biophys J 94:1221–1232 22. Upadhyay A, Brunner M, Herzel H (2019) An inactivation switch enables rhythms in a Neurospora clock model. Int J Mol Sci 20:2985 23. Fathallah-Shaykh HM, Bona JL, Kadener S (2009) Mathematical model of the Drosophila circadian clock: loop regulation and transcriptional integration. Biophys J 97:2399–2408 24. Leloup JC, Goldbeter A (1998) A model for circadian rhythms in Drosophila incorporating the formation of a complex between the PER and TIM proteins. J Biol Rhythm 13:70–87 25. Ueda HR, Hagiwara M, Kitano H (2001) Robust oscillations within the interlocked feedback model of Drosophila circadian rhythm. J Theor Biol 210:401–406 26. Tyson JJ, Novak B (2020) A dynamical paradigm for molecular cell biology. Trends Cell Biol 30:504–515 27. Goodwin BC (1965) Oscillatory behavior in enzymatic control processes. Adv Enzyme Regul 3:425–438 28. Ruoff P, Vinsjevik M, Monnerjahn C, Rensing L (2001) The goodwin model: simulating the effect of light pulses on the circadian sporulation rhythm of Neurospora crassa. J Theor Biol 209:29–42 29. Ruoff P, Loros JJ, Dunlap JC (2005) The relationship between FRQ-protein stability and temperature compensation in the Neurospora circadian clock. Proc Natl Acad Sci USA 102: 17681–17686 30. Locke JC, Westermark PO, Kramer A, Herzel H (2008) Global parameter search reveals design principles of the mammalian circadian clock. BMC Syst Biol 2:22 31. Komin N, Murza AC, Hernandez-Garcia E, Toral R (2011) Synchronization and

78

Marta del Olmo et al.

entrainment of coupled circadian oscillators. Interface Focus 1:167–76 32. Gonze D (2011) Modeling circadian clocks: from equations to oscillations. Centr Eur J Biol 6:699–711 2011. 33. Forger DB (2017) Biological clocks, rhythms, and oscillations: the theory of biological timekeeping. University Press, Cambridge 34. Ferrell Jr JE, Tsai TY, Yang Q (2011) Modeling the cell cycle: why do certain circuits oscillate? Cell 144:874–885 35. Nova´k B, Tyson JJ (2008) Design principles of biochemical oscillators. Nat Rev Mol Cell Biol 9:981–991 36. Gonze D, Ruoff P (2020) The Goodwin oscillator and its legacy. Acta Biotheoretica 37. Tyson JJ, Chen KC, Novak B (2003) Sniffers, buzzers, toggles and blinkers: dynamics of regulatory and signaling pathways in the cell. Curr Opin Cell Biol 15:221–231 38. Bordyugov G, Westermark PO, Korencˇicˇ A, Bernard S, Herzel H (2013) Mathematical modeling in chronobiology. In: Kramer A, Merrow M (ed) Circadian clocks. Springr, Berlin, pp 45–66 39. Del Olmo M, Kramer A, Herzel H (2019) A robust model for circadian redox oscillations. Int J Mol Sci 20:2368 40. Ferrell Jr JE, Ha SH (2014) Ultrasensitivity part I: Michaelian responses and zero-order ultrasensitivity. Trends Biochem Sci 39: 496–503 41. Ferrell Jr JE, Ha SH (2014) Ultrasensitivity part II: multisite phosphorylation, stoichiometric inhibitors, and positive feedback. Trends Biochem Sci 39:556–69 42. Ferrell Jr JE, Ha SH (2014) Ultrasensitivity part III: cascades, bistable switches, and oscillators. Trends Biochem Sci 39:612–618 43. Cornish-Bowden A (1995) Fundamentals of enzyme kinetics. Portland Press, London 44. Alon U (2007) An introduction to systems biology – design principles of biological circuits. CRC Press, Boca Raton 45. Gonze D, Abou-Jaoude´ W (2013) The Goodwin model: behind the hill function. PLoS One 8:e69573 46. Griffith JS (1968) Mathematics of cellular control processes. I. Negative feedback to one gene. J Theor Biol 20:202–208 47. Korencˇicˇ A, Kosˇir R, Bordyugov G, Lehmann R, Rozman D, Herzel H (2014) Timing of circadian genes in mammalian tissues. Sci Rep 4:5782

48. Pett JP, Korencˇicˇ A, Wesener F, Kramer A, Herzel H (2016) Feedback loops of the mammalian circadian clock constitute repressilator. PLoS Comput Biol 12:e1005266 49. Almeida S, Chaves M, Delaunay F (2020) Transcription-based circadian mechanism controls the duration of molecular clock states in response to signaling inputs. J Theor Biol 484: 110015 50. Meinhardt H (1982) Models of biological pattern formation. Academic Press, London 51. Tsai TY, Choi YS, Ma W, Pomerening JR, Tang C, Ferrell Jr JE (2008) Robust, tunable biological oscillations from interlinked positive and negative feedback loops. Science 321: 126–129 52. Stricker J, Cookson S, Bennett MR, Mather WH, Tsimring LS, Hasty J (2008) A fast, robust and tunable synthetic gene oscillator. Nature 456:516–519 53. Clodong S, Du¨hring U, Kronk L, Wilde A, Axmann I, Herzel H, Kollmann M (2007) Functioning and robustness of a bacterial circadian clock. Mol Syst Biol 3:90 54. Kim JK (2016) Protein sequestration versus Hill-type repression in circadian clock models. IET Syst Biol 10:125–135 55. Bugge A, Feng D, Everett LJ, Briggs ER, Mullican SE, Wang F, Jager J, Lazar MA (2012) Rev-erbα and Rev-erbβ coordinately protect the circadian clock and normal metabolic function. Genes Dev 26:657–667 56. Pett JP, Kondoff M, Bordyugov G, Kramer A, Herzel H (2018) Co-existing feedback loops generate tissue-specific circadian rhythms. Life Sci Alliance 1:e201800078 57. Edwards MD, Brancaccio M, Chesham JE, Maywood ES, Hastings MH (2016) Rhythmic expression of cryptochrome induces the circadian clock of arrhythmic suprachiasmatic nuclei through arginine vasopressin signaling. Proc Natl Acad Sci USA 113(10):2732–2737 58. Maywood ES, Elliott TS, Patton AP, Krogager TP, Chesham JE, Ernst RJ, Bera´nek V, Brancaccio M, Chin JW, Hastings MH (2018) Translational switching of Cry1 protein expression confers reversible control of circadian behavior in arrhythmic Cry-deficient mice. Proc Natl Acad Sci USA 115(52): E12388–12397 59. Granada A, Hennig RM, Ronacher B, Kramer A, Herzel H (2009) Phase response curves: elucidating the dynamics of coupled oscillators. In: Johnson ML, Brand L (eds)

Mathematical Modeling in Circadian Rhythmicity Methods in enzymology, chapter 1. Academic Press, Cambridge 60. Amdaoud M, Vallade M, Weiss-Schaber C, Mihalcescu I (2007) Cyanobacterial clock, a stable phase oscillator with negligible intercellular coupling. Proc Natl Acad Sci U S A 104: 7051–7056 61. Rougemont J, Naef F (2006) Collective synchronization in populations of globally coupled phase oscillators with drifting frequencies. Phys Rev E Stat Nonlin Soft Matter Phys 73:011104 62. Rougemont J, Naef F (2008) Stochastic phase oscillator models for circadian clocks. Adv Exp Med Biol 641:141–149 63. Winfree AT (1967) Biological rhythms and the behavior of populations of coupled oscillators. J Theor Biol 16:15–42 64. Yang Q, Pando BF, Dong G, Golden SS, Oudenaarden A (2010) Circadian gating of the cell cycle revealed in single cyanobacterial cells. Science 327:1522–1526 65. Ahn S, Solfest J, Rubchinsky LL (2014) Fine temporal structure of cardiorespiratory synchronization. Am J Physiol Heart Circ Physiol 306:H755–H763 66. Seidel H, Herzel H (1998) Analyzing entrainment of heartbeat and respiration with surrogates. IEEE Eng Med Biol Mag 17:54–57 67. Honma S, Shirakawa T, Katsuno Y, Namihira M, Honma K (1998) Circadian periods of single suprachiasmatic neurons in rats. Neurosci Lett 250:157–160 68. Yamaguchi S, Isejima H, Matsuo T, Okura R, Yagita K, Kobayashi M, Okamura H (2003) Synchronization of cellular clocks in the suprachiasmatic nucleus. Science 302:1408–1412 69. Pikovsky A, Rosenblum M, Kurths J (2001) Synchronization: a universal concept in nonlinear sciences. Cambridge University Press, Cambridge 70. Arnold V (1988) Geometrical methods in the theory of ordinary differential equations. Springer, New York 71. Glass L, Be´lair J (1986) Continuation of arnold tongues in mathematical models of periodically forced biological oscillators. In: Othmer HG (ed) Nonlinear oscillations in biology and chemistry. Springer, New York

79

72. Schmal C, Myung J, Herzel H, Bordyugov G (2015) A theoretical study on seasonality. Front Neurol 6:94 73. Duffy JF, Czeisler CA (2002) Age-related change in the relationship between circadian period, circadian phase, and diurnal preference in humans. Neurosci Lett 318:117–120 74. Roenneberg T, Kuehnle T, Pramstaller PP, Ricken J, Havel M, Guth A, Merrow M (2004) A marker for the end of adolescence. Curr Biol 14:R1038–R1039 75. Johnson CH, Elliott JA, Foster R (2003) Entrainment of circadian programs. Chronobiol Int 20:741–774 76. Roenneberg T, Daan S, Merrow M (2003) The art of entrainment. J Biol Rhythms 18: 183–194 77. Dibner C, Schibler U, Albrecht U (2010) The mammalian circadian timing system: organization and coordination of central and peripheral clocks. Annu Rev Physiol 72:517–549 78. Miller BH, McDearmon EL, Panda S, Hayes KR, Zhang J, Andrews JL, Antoch MP, Walker JR, Esser KA, Hogenesch JB, Takahashi JS (2007) Circadian and CLOCK-controlled regulation of the mouse transcriptome and cell proliferation. Proc Natl Acad Sci U S A 104: 3342–3347 79. Storch K, Lipan O, Leykin I, Viswanathan N, Davis FC, Wong WH, Weitz CJ (2002) Extensive and divergent circadian gene expression in liver and heart. Nature 417:78–83 80. Pett JP, Westermark PO, Herzel H (2021) Simple kinetic models in molecular chronobiology. In: Brown SA (ed) Circadian clocks. Humana, New York 81. Le Martelot G, Canella D, Symul L, Migliavacca E, Gilardi F, Liechti R, Martin O, Harshman K, Delorenzi M, Desvergne B, Herr W, Deplancke B, Schibler U, Rougemont J, Guex N, Hernandez N, Naef F, and the CycliX consortium (2012) Genome-wide RNA polymerase II profiles and RNA accumulation reveal kinetics of transcription and associated epigenetic changes during diurnal cycles. PLoS Biol 10:1–16

80

Marta del Olmo et al.

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), 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 chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Chapter 5 Bioinformatics and Systems Biology of Circadian Rhythms: BIO_CYCLE and CircadiOmics Muntaha Samad, Forest Agostinelli, and Pierre Baldi Abstract Circadian rhythms are fundamental to biology and medicine and today these can be studied at the molecular level in high-throughput fashion using various omic technologies. We briefly present two resources for the study of circadian omic (e.g. transcriptomic, metabolomic, proteomic) time series. First, BIO_CYCLE is a deep-learning-based program and web server that can analyze omic time series and statistically assess their periodic nature and, when periodic, accurately infer the corresponding period, amplitude, and phase. Second, CircadiOmics is the larges annotated repository of circadian omic time series, containing over 260 experiments and 90 million individual measurements, across multiple organs and tissues, and across 9 different species. In combination, these tools enable powerful bioinformatics and systems biology analyses. The are currently being deployed in a host of different projects where they are enabling significant discoveries: both tools are publicly available over the web at: http://circadiomics.ics.uci.edu/. Key words Circadian, Rhythms, Omic, Transcriptomic, Period, Amplitude, Phase, Bioinformatics

1

Introduction Circadian rhythms are a ubiquitous phenomenon in biology—they are found in cyanobacteria, fungi, plants, and animals—and play a fundamental role in biological organization [1–8] Circadian oscillations of molecular species are responsible for regulating a variety of physiological and metabolic processes including: sleep/wake cycle, hormone secretion, diet related metabolism, and neural function [9–12]. Due to the ubiquitous and vital nature of circadian rhythms, any disruption to them can result in serious health problems such as cancer, diabetes, obesity and premature aging [2, 13–25]. Circadian rhythms may also play an important role in a host of therapeutic interventions, for instance in optimizing the times of administration of drugs and other treatments in precision medicine [26].

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_5, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

81

82

Muntaha Samad et al.

The advance of modern high-throughput technologies has made it possible to investigate circadian rhythms on the molecular level. Measuring the concentrations of molecular species across time has shown that circadian oscillations do indeed occur in the cell. It is estimated that 10% of all transcripts and metabolites are periodic with a period of 24 h [27–33]. In addition, differences in oscillations occur based on tissue type and condition [31, 34, 35]. Therefore, gaining an in-depth understanding of circadian oscillations continues to be the focus of intense research. Such research depends crucially on the availability of effective computational tools for analyzing omic-time series and for comparing and integrating such time series across different experiments. We present two tools for addressing these needs: BIO_CYCLE and CircadiOmics.

2

BIO_CYCLE Circadian omic experiments measure the concentration of molecular species over a period of time, typically 20–48 h, with multiple biological replicates obtained at each time-point. The fundamental computational problems are: (1) to determine which species appear to exhibit rhythmic concentrations patterns and quantify the corresponding degree of confidence (e.g. in terms of p-values and q-values); and (2) for the oscillating species, determine their period, amplitude, and phase with possibly the corresponding error bars. While the problem of detecting periodic patterns in time series is not new, circadian experiments present unique challenges: (1) the sparsity of the data (due to experimental costs, measurements are typically only carried out every 4 h); (2) noise from multiple sources including biological variability, often accentuated by the need to sacrifice animals after each measurement, and measurement noise; (3) the small number of replicates (typically three or fewer per timepoint); (4) missing data; and (5) uneven time-point sampling [36]. BIO_CYCLE is a program and software package that addresses these issues and has been shown to have superior detection of periodic species as well as superior estimation of periods and phases [37]. BIO_CYCLE is based on a deep learning approach, i.e. it uses deep neural networks to perform these tasks. Most remarkably, BIO_CYCLE can be trained using primarily synthetic data. Since its original publication, a new version of BIO_CYCLE has been developed and is currently being released, with several improvements including: (1) Modeling real-world replicated experimental data to produce more realistic p-values; (2) Implementation in Python to take advantage of state-of-the-art deep learning software packages (TensorFlow, PyTorch,Keras) and hypeparameter optimisation frameworks (Sherpa [38]); (3) Handling of missing timepoints; (4) Improved amplitude estimation; (5) Addition of offset

BIO_CYCLE and CircadiOmics

83

estimation; and (6) Hosting on a public web server. While the BIO_CYCLE code is freely available online for download, running BIO_CYCLE is only one step to analyzing a circadian experiment. To assist with the analysis of circadian experiments, we have created the BIO_CYCLE web server:http://circadiomics.igb.uci.edu/ biocycle. The web server runs BIO_CYCLE on user uploaded circadian datasets and provides the user with easy-to-use analysis tools, which include: • Histograms of periods, phases, amplitudes, and offsets • Querying of molecular species based on a user-selected P-value or q-value cutoff • Visualization of molecular concentrations across time • Analysis at 24, 12, and 8 h periods. We use synthetic data to train the deep neural networks of BIO_CYCLE because these allow one to control the type of signal along with its parameters, giving one definitive ground truth values for periodic/aperiodic classification as well as for estimation of the period and other parameters of the oscillations. Much of the synthetic data comes from [39]. There are 7 different types of periodic trajectories: cosine, cosine peaked, triangle wave, square wave, cosine plus a linear component, cosine plus an exponential component, and cosine multiplied by an exponential component. There are 5 different types of aperiodic trajectories: whitenoise, linear, flat, an auto-regressive process of order 1, and data generated from a Gaussian process [40]. Figure 1 shows an example of each different type of trajectory from the synthetic dataset. Of course, to calibrate and validate the performance of BIO_CYCLE, we also use biological data, taken from CircadiOmics (see next section). For instance, the behavior of the core clock genes can be used to further assess the performance of BIO_CYCLE. Additional technical details BIO_CYCLE and the underlying deep learning and statistical methods are provided in the corresponding publications. Figure 2 shows how BIO_CYCLE (BC) outperforms other programs that have been developed to assess periodicity in time series, including ARSER (ARS) [41], Lomb-Scargle (LS) [42], and JTK_CYCLE (JTK) [43]. An independent assessment of these programs, confirming our conclusions, can be found in [44]. 2.1 The BIO_CYCLE Web Server

2.2

Input Format

The web server accepts a user-uploaded file containing the measurements related to the concentrations of molecular species across time-points (e.g. transcript levels measured every 4 h). Each row

84

Muntaha Samad et al.

Fig. 1 Samples of synthetic data. Periodic trajectories are in green and aperiodic trajectories are in red

must contain the ID of the molecular species, followed by the concentration measurement at each time-point. Each column corresponds to a different time-point or replicate. A visualization of the interface is shown in Fig. 3. 2.2.1 Running BIO_CYCLE

After the file is uploaded, the server will run BIO_CYCLE on the uploaded file for three separate possible period ranges: 20 through 28 h, 10 through 14 h, and 7 through 9 h. A separate deep neural network (DNN) is trained for each set of timep-points and for each period range. If the DNNs are already trained, then the results are ready within about 1 min. If the DNNs are not trained, BIO_CYCLE will automatically train them and the results will be ready within about 2 min.

2.2.2 Period Selection

Since cycling transcripts have been found with periods of 24 h, 12 h, and 8 h [45], the BIO_CYCLE server will have a drop-down menu that allows the user to select which harmonic to investigate as well as which p-value or q-value cutoff to use to distinguish between periodic and aperiodic molecular species.

2.2.3 Histograms

Given the selected period and the p-value or q-value cutoff, the webserver will produce histograms of periods, lags, amplitudes, and offsets.

BIO_CYCLE and CircadiOmics

85

Fig. 2 Mean absolute error for prediction of the period and phase on a labeled biological dataset for several programs. ARS ¼ ARSER, LS ¼ Lomb-Scargle, JTK ¼ JTK_CYCLE, and BC ¼ BIO_CYCLE

86

Muntaha Samad et al.

Fig. 3 The BIO_CYCLE web server interface

Fig. 4 The web server shows the IDs of the trajectories that match the user’s search 2.2.4 Trajectory Visualization

Users can see the IDs of the trajectories that match their search (Fig. 4). If the user wishes to view the trajectories of specific molecular species they may do so in the trajectory visualization tab (Fig. 5).

BIO_CYCLE and CircadiOmics

87

Fig. 5 The web server shows visualizations of the user’s data

3

CircadiOmics In order to develop BIO_CYCLE and, more importantly, to enable comparative analyses and aggregated inferences about circadian rhythms at the molecular level, we have developed CircadiOmics. CircadiOmics is the largest database and web-server of circadian omic time series. CircadiOmics(circadiomics.ics.uci.edu) was developed to be a comprehensive web server for accessing and mining circadian omic datasets. The latest version contains 263 highthroughput omic datasets corresponding to over 90 million individual. Via the CircadiOmics web server, researchers can visualize and compare the oscillations of molecular species across tissues, species and conditions. Additionally statistics regarding periodicity including: period, amplitude, phase, p-value, and q-value are displayed and can be easily downloaded in several different formats. These periodicity statistics are obtained from BIO_CYCLE and, in the current version, from JTK_CYCLE as well. In Fig. 6 we see the output of a sample query to the web server.

3.1 Dataset Statistics

CircadiOmics is currently the largest single repository of circadian data available, with 263 datasets and over 90 million data points. There are also 24 different types of tissues (grouped into 14 categories) and 9 different species represented—this is significantly more than any other circadian data repository. Figure 7 shows a breakdown of the number and types of datasets available in several of the most prominent circadian data repositories [46–49]. The majority of datasets in CircadiOmics are collected from the species Mus musculus (mouse) and Papio anibus (babboon) and

88

Muntaha Samad et al.

Fig. 6 Visualization of queries for Arntl in two different experimental datasets

Fig. 7 Comparison of CircadiOmics with other circadian repositories

from liver and brain tissue. In addition to a wide variety of species and tissues, CircadiOmics also has a diverse set of experimental conditions represented. Some of the experimental conditions represented include: knock-downs, knock-outs, diet changes, exercise, and drug treatments. In addition, CircadiOmics uniquely contains data from different omic experiments, including transcriptome, metabolome, proteome, and acetylome experiments. Figure 8 summarizes the number of available datasets by detailed categories. 3.2 Dataset Collection

The datasets accessible through the CircadiOmics web portal are primarily collected via collaborations with other laboratories, and automated discovery. Due to exponential growth of circadian omic data, we use several automated approaches to identify newly available circadian datasets including a web crawler developed in-house and the publicly available web service PubCrawler [50]. The web crawler developed in-house uses the Python packages scholarly and

BIO_CYCLE and CircadiOmics

89

Fig. 8 Breakdown of datasets by species, tissues, experimental conditions, and omic categories

geotools to search the literature to discover new circadian omic studies and their affiliated datasets. To find new datasets, the crawler performs keyword searches on published abstracts, filters the selected abstracts based on publishing journal, author and provided supplementary materials, and then uses logistic regression on the extracted features to classify whether or not a dataset is a good candidate for inclusion in CircadiOmics. The datasets discovered by the crawler are then manually inspected for quality and manually processed to be included into CircadiOmics. Using this crawler in tandem with PubCrawler, which sends a daily email containing a list of possible publications of interest, we are able to keep CircadiOmics current by continuously adding the latest cutting-edge research in circadian rhythms to the web portal. 3.3 CircadiOmics Features 3.3.1 Search

The main functionality of CircadiOmics is the search, which allows users to compare and visualize the oscillation trends of molecular species. The user can select a single dataset, or multiple datasets, from within the repository and search any molecular species. CircadiOmics allows for the overlay of multiple searches together to enable comparative studies. When datasets do not have the same time course, results are displayed from the minimum to the maximum time point over all selected datasets. The absolute mean expression values for oscillating species can often have large differences in intensity, therefore to ensure easy visual comparisons the mean expression can be dynamically scaled with the minimum and maximum values normalized to zero and one, respectively. For each query, a table of periodicity statistics including: period, amplitude,

90

Muntaha Samad et al.

phase, p-value, and q-value is displayed. Because CircadiOmics is a large repository and is constantly expanding, filtering candidate datasets based on species, tissue and experimental condition is necessary to limit the scope of the datasets displayed for selection. In addition to what is displayed on the search page, additional details for each dataset can be found in tabular form under the dataset tab. These details include a brief description of the experimental protocol, the citation, the GEO accession number, and other summary information. 3.3.2 Metabolic Atlas

Another feature provided by CircadiOmics is the The Metabolic Atlas web portal. The Metabolic Atlas web portal (http:// circadiomics.ics.uci.edu/metabolicatlas), allows researchers to generate and visualize interactive metabolic networks. These networks are derived in part from the KEGG database and can be filtered using BIO_CYCLE statistics [51]. To create a metabolic network, users start by selecting a dataset and a particular metabolite. From there, the user can select options to create a network. For example, one option is to display a network of all metabolites that are oscillating in-phase with the selected metabolite. Another possible option is to display a network of all metabolites that are involved in the same pathways at the selected metabolite. There are six possible options for the user to select from for network creation. Once the network is displayed, the user can choose to filter out edges based on BIO_CYCLE statistics. Figure 9 shows The Metabolic Atlas web portal interface with adenine as the selected metabolite. To create a network, the user would select one of the buttons under the “Add Edge” heading and then filter the results by using the “Edge Parameter” options.

3.3.3 BIO_CYCLE Web Portal within CircadiOmics

All of the periodicity statistics displayed via CircadiOmics are obtained by processing each dataset with the software packages BIO_CYCLE and JTK_CYCLE [52, 53]. To identify which molecular species are oscillating in a circadian manner, researchers can filter the displayed data using the outputted p-value and q-value at a selected threshold. The literature has shown that BIO_CYCLE is consistently more accurate at determining periodicity over alternative methods [52]. Therefore, to facilitate the usage of this software we have included a BIO_CYCLE portal within CircadiOmics (http://circadiomics.ics.uci.edu/biocycle) that allows users to upload a dataset of their choosing to be processed with BIO_CYCLE. The processed file is automatically downloaded and contains the p-value, q-value, period, amplitude and phase for each molecular species. Additionally, summary figures displaying the distribution of each periodicity statistic in the user provided dataset are also downloaded. To help make the upload process as intuitive as possible, an example dataset is provided to give the user a template for the desired data format. Figure 10 shows the interface for the BIO_CYCLE portal.

BIO_CYCLE and CircadiOmics

Fig. 9 Screenshot of the Metabolic Atlas Web Portal with adenine selected

Fig. 10 Screenshot of the BIO_CYCLE Portal

91

92

4

Muntaha Samad et al.

Biological Discovery In addition to the projects in the references already mentioned, BIO_CYCLE and CircadiOmics are being used in a host of other projects, for instance to probe circadian rhythms in the liver and in various brain areas in relation to diet and nutritional challenges, exercise, metabolic activity, epilepsy, cocaine administration and many other genetic, epigenetic, and environmental perturbations (e.g. [54–63]). The tools are freely available to the academic research community and, as we continuously update them, we welcome any feedback and suggestions for improving them further.

References 1. Eckel-Mahan K, Sassone-Corsi P (2009) Metabolism control by the circadian clock and vice versa. Nat Struct Mol Biol 16(5):462–467 2. Froy O (2011) Circadian rhythms, aging, and life span in mammals. Physiology (Bethesda) 26(4):225–235 3. Takahashi JS et al (2008) The genetics of mammalian circadian order and disorder: implications for physiology and disease. Nat Rev Genet 9(10):764–775 4. Yoo SH et al (2004). PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101(15):5339–5346 5. Gerstner JR et al (2009a) Cycling behavior and memory formation. J Neurosci 29(41): 12824–12830 6. Eckel-Mahan KL et al (2008) Circadian oscillation of hippocampal MAPK activity and cAMP: implications for memory persistence. Nat Neurosci 11(9):1074–1082 7. Panda S, Hogenesch JB, Kay SA (2002) Circadian rhythms from flies to human. Nature 417(6886):329–335. https://doi.org/10.103 8/417329a 8. Patel VR et al (2015) The pervasiveness and plasticity of circadian oscillations: the coupled circadian-oscillators framework. Bioinformatics 31(19):3181–3188. https://doi.org/10. 1093/bioinformatics/btv353 9. Bass J (2012) Circadian topology of metabolism. Nature 491(7424):348–356. https:// doi.org/10.1038/nature11704 10. Dibner C, Schibler U, Albrecht U (2010) The mammalian circadian timing system: organization and coordination of central and peripheral clocks. Annu Rev Phys Chem 72(1):517–549.

https://doi.org/10.1146/annurev-physiol021909-135821 11. Gerstner JR et al (2009) Cycling behavior and memory formation. J Neurosci 29(41): 12824–12830. https://doi.org/10.1523/ jneurosci.3353-09.2009 12. Menet JS, Rosbash M (2011) When brain clocks lose track of time: cause or consequence of neuropsychiatric disorders. Curr Opin Neur o b i o l 2 1 ( 6 ) : 8 4 9 – 8 5 7 . h t t p s : // d o i . org/10.1016/j.conb.2011.06.008 13. Karlsson B, Knutsson A, Lindahl B (2001) Is there an association between shift work and having a metabolic syndrome? Results from a population based study of 27,485 people. Occup Environ Med 58(11):747–752 14. Antunes LC et al (2010) Obesity and shift work: chronobiological aspects. Nutr Res Rev 23(1):155–168 15. Froy O (2010) Metabolism and circadian rhythms–implications for obesity. Endocr Rev 31(1):1–24 16. Kohsaka A et al (2007) High-fat diet disrupts behavioral and molecular circadian rhythms in mice. Cell Metab 6(5):414–421 17. Sharifian A et al (2005) Shift work as an oxidative stressor. J Circadian Rhythms 3:15 18. Turek FW et al (2005) Obesity and metabolic syndrome in circadian Clock mutant mice. Science 308(5724):1043–1045 19. Shi S-q et al (2013) Circadian disruption leads to insulin resistance and obesity. Curr Biol 23(5):372–381 20. Kondratov RV et al (2006) Early aging and age-related pathologies in mice deficient in BMAL1, the core componentof the circadian clock. Genes Develop 20(14):1868–1873

BIO_CYCLE and CircadiOmics 21. Adser H et al (2011) Interleukin-6 modifies mRNA expression in mouse skeletal muscle. Acta Physiol. 202(2):165–173. https://doi. org/10.1111/j.1748-1716.2011.02269.x 22. Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabolism, and the circadian clock. Cell 161(1):84–92. https://doi.org/10.1016/j. cell.2015.03.015 23. Fu L, Lee CC (2003) The circadian clock: pacemaker and tumour suppressor. Nat Rev Cancer 3(5):350–361. https://doi.org/10. 1038/nrc1072 24. Partch CL, Green CB, Takahashi JS (2014) Molecular architecture of the mammalian circadian clock. Trends Cell Biol 24(2):90–99. https://doi.org/10.1016/j.tcb.2013.07.002 25. Roenneberg T, Merrow M (2016) The circadian clock and human health. Curr Biol 26(10). https://doi.org/10.1016/j.cub.201 6.04.011 26. Zhang R et al (2014) A circadian gene expression atlas in mammals: implications for biology and medicine. Proc Natl Acad Sci 111(45): 16219–16224 27. Hughes ME et al (2009) Harmonics of circadian gene transcription in mammals. PLoS Genet 5(4):e1000442 28. Miller BH et al (2007) Circadian and CLOCKcontrolled regulation of the mouse transcriptome and cell proliferation. Proc Natl Acad Sci 104(9):3342–3347 29. Eckel-Mahan KL et al (2012) Coordination of the transcriptome and metabolome by the circadian clock. Proc Natl Acad Sci U S A 109(14):5541–5546 30. Eckel-Mahan KL et al (2013) Reprogramming of the circadian clock by nutritional challenge. Cell 155(7):1464–1478. https://doi. org/10.1016/j.cell.2013.11.034. http:// www.cell.com/cell/abstract/S0092-8674(13) 01485-2 31. Panda S et al (2002) Coordinated transcription of key pathways in the mouse by the circadian clock. Cell 109(3):307–320 32. Andrews JL et al (2010) CLOCK and BMAL1 regulate MyoD and are necessary for maintenance of skeletal muscle phenotype and function. Proc Natl Acad Sci U S A 107(44): 19090–19095 33. Masri S et al (2014) SIRT6 defines circadian transcription leading to control of lipid metabolism. Cell. In press 34. Storch KF et al (2002) Extensive and divergent circadian gene expression in liver and heart. Nature 417(6884):78–83

93

35. Yan J et al (2008) Analysis of gene regulatory networks in the mammalian circadian rhythm. PLoS Comput Biol 4(10):e1000193 36. Hughes ME et al (2017) Guidelines for genome-scale analysis of biological rhythms. J Biol Rhythm 32(5):380–393 37. Agostinelli F et al (2016) What time is it? Deep learning approaches for circadian rhythms. Bioinformatics 32(12):i8–i17 38. Hertel L et al (2020) Sherpa: robust hyperparameter optimization for machine learning. SoftwareX. Also arXiv:2005.04048. Software available at: https://github.com/sherpa-ai/ sherpa 39. Deckard A et al (2013) Design and analysis of large-scale biological rhythm studies: a comparison of algorithms for detecting periodic signals in biological data. Bioinformatics 29(24):3174–3180 40. Rasmussen CE (2006) Gaussian processes for machine learning. MIT Press, Cambridge 41. Yang R, Su Z (2010) Analyzing circadian expression data by harmonic regression based on autoregressive spectral estimation. Bioinformatics 26(12):i168–i174 42. Glynn EF, Chen J, Mushegian AR (2006) Detecting periodic patterns in un-evenly spaced gene expression time series using Lomb–Scargle periodograms. Bioinformatics 22(3):310–316 43. Hughes ME, Hogenesch JB, Kornacker K (2010) JTK_CYCLE: an efficient nonparametric algorithm for detecting rhythmic components in genome-scale data sets. J Biol Rhythm 25(5):372–380 44. Mei W et al (2020) Genome-wide circadian rhythm detection methods: systematic evaluations and practical guidelines. Brief Bioinform 45. Hughes ME et al (2009) Harmonics of circadian gene transcription in mammals. PLoS Genet 5(4):e1000442 46. Ceglia N et al (2018) CircadiOmics: circadian omic web portal. Nucleic Acids Res 46 (W1) https://doi.org/10.1093/nar/gky441 47. Pizarro A et al (2012) CircaDB: a database of mammalian circadian gene expression profiles. Nucleic Acids Res 41(D1). https://doi.org/ 10.1093/nar/gks1161 48. Li D et al (2010) BioClock: a web server and database aimed for interpreting circadian rhythm 49. Li X et al (2017) CirGRDB: a database for the genome-wide deciphering circadian genes and regulators. Nucleic Acids Res 46(D1). https:// doi.org/10.1093/nar/gkx944

94

Muntaha Samad et al.

50. Hokamp K, Wolfe KH (2004) PubCrawler: keeping up comfortably with PubMed and Gen-Bank. Nucleic Acids Res 32. Web Server. https://doi.org/10.1093/nar/gkh453 51. Kanehisa M (2000) KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28(1):27–30. https://doi.org/10.1093/nar/ 28.1.27 52. Agostinelli F et al (2016) What time is it? Deep learning approaches for circadian rhythms. Bioinformatics 32(19):3051–3051. https://doi. org/10.1093/bioinformatics/btw504 53. Hughes ME, Hogenesch JB, Kacker K (2010) JTK_CYCLE: an efficient nonparametric algorithm for detecting rhythmic components in genome-scale data sets. J Biol Rhythm 25(5): 372–380. https://doi.org/10.1177/074 8730410379711 54. Greco CM et al (2020) S-adenosyl-l-homocysteine hydrolase links methionine metabolism to the circadian clock and chromatin remodeling. Sci Adv 6(51):eabc5629 55. Tognini P et al (2020) Reshaping circadian metabolism in the suprachiasmatic nucleus and prefrontal cortex by nutritional challenge. Proc Natl Acad Sci 117(47):29904–29913 56. Debski KJ et al (2020) The circadian dynamics of the hippocampal transcriptome and proteome is altered in experimental temporal lobe

epilepsy. Sci Adv 6(41). Also bioRxiv, eaat5979. https://doi.org/10.1101/199372 57. Brami-Cherrier K et al (2020) Cocainemediated circadian reprogramming in the striatum through dopamine D2R and PPARγactivation. Nat Commun 11(1):1–14 58. Greco CM et al (2020) A non-pharmacological therapeutic approach in the gut triggers distal metabolic rewiring capable of ameliorating diet-induced dysfunctions encompassed by metabolic syndrome. Sci Rep 10(1):1–13 59. Murakami M et al (2016) Gut microbiota directs PPARγ-driven reprogramming of the liver circadian clock by nutritional challenge. EMBO Rep 17(9):1292–1303 60. Sato S et al (2019) Time of exercise specifies the impact on muscle metabolic pathways and systemic energy homeostasis. Cell Metab 30(1):92–110 61. Koronowski KB et al (2019) Defining the independence of the liver circadian clock. Cell 177(6):1448–1462 62. Kwapis JL et al (2018) Epigenetic regulation of the circadian gene Per1 contributes to age-related changes in hippocampal memory. Nat Commun 9(1):1–14 63. Tognini P et al (2017) Distinct circadian signatures in liver and gut clocks revealed by ketogenic diet. Cell Metab 26(3):523–538

Chapter 6 Cell-Based Phenotypic Screens to Discover Circadian Clock-Modulating Compounds Megumi Hatori and Tsuyoshi Hirota Abstract There is increasing demand to control circadian clock functions in a conditional manner for deeper understanding of the circadian system as well as for potential treatment of clock-related diseases. Smallmolecule compounds provide powerful tools to reveal novel functions of target proteins in the circadian clock mechanism, and can be great therapeutic candidates. Here we describe the detailed methods of measuring cellular circadian rhythms in a high-throughput manner for chemical screening to identify compounds that affect circadian rhythms by targeting clock-related proteins. Key words Circadian clock, Period, Luciferase reporter, Cell-based phenotypic screen, Highthroughput screening, Small-molecule compounds

1

Introduction The mammalian circadian rhythms are generated by transcriptiontranslation based feedback loops that are composed of clock proteins, PER, CRY, CLOCK, and BMAL1, in a cell autonomous manner [1]. Small-molecule compounds facilitate a better understanding of the circadian system, and the treatment of clock-related diseases, by enabling control of target protein functions in dosedependent and conditional manners across species [2, 3]. To search for circadian clock-modulating compounds in an unbiased manner, cell-based phenotypic screens are required. The oscillatory mechanism of the circadian clock resides not only in the suprachiasmatic nucleus (SCN) in the hypothalamus but also peripheral tissues and cells. In addition to clock gene expression, the SCN neurons show robust circadian variations in electrophysiological activities that can be measured by multielectrode recordings [4, 5]. Fluorescent proteins and fluorescent proteinbased probes [6] provide powerful tools to monitor Ca2+ (such as cameleon, GCaMP, and GECO), membrane voltage (such as

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_6, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

95

96

Megumi Hatori and Tsuyoshi Hirota

ArcLight), and gene expression (such as destabilized forms of GFP) at the single-cell level, and circadian rhythms of these parameters have been successfully visualized in the SCN [7–9]. Achilles, a fastmaturing YFP variant, was recently developed, enabling visualization of the segmentation clock oscillations with a period of 2–3 h at the single-cell level [10]. To visualize circadian gene expression, however, luminescence-based technologies that do not require excitation light have been broadly used at the cell population level as well as the single-cell level [11]. The most commonly used reporter is firefly luciferase and its degradable form (dLuc; halflife ~0.5 h). Triggered by the discovery by Schibler and colleagues that cultured fibroblasts show circadian gene expression [12], a clock gene promoter-driven luciferase reporter system has been introduced to monitor circadian rhythms in cultured cells [13–16]. By applying cell-based circadian assays, we and others have established high-throughput phenotypic screening systems and identified circadian clock-modulating compounds such as GSK-3 inhibitors [17], CKI inhibitors [18–21], a CK2 inhibitor [22], CRY activators [23–26], and a steroid hormone DHEA [27] for period regulation; a natural flavonoid nobiletin for amplitude regulation [28]; and an adenosine analog cordycepin for phase regulation [29]. In addition to phenotypic screening, target-based approaches have identified CKI inhibitors [30–32], REV-ERB agonists [33], a CRY inhibitor [34], and a CLOCK inhibitor [35], as circadian clock-modulating compounds. In this protocol, we describe the detailed methods of cell-based phenotypic screening to identify circadian clock-modulating compounds.

2 2.1

Materials Cell Lines

2.2 Cell Culture Reagents and Supplies

1. Stable U2OS cells harboring a Bmal1-dLuc or Per2-dLuc reporter [17, 36] (see Note 1). 1. Cell culture medium: DMEM (11995-073, Gibco) supplemented with 10% fetal bovine serum, 0.29 mg/mL L-glutamine, 100 units/mL penicillin, and 100 μg/mL streptomycin (see Note 2). 2. Explant medium: DMEM powder (12800-017, Gibco) dissolved in cell culture grade water and supplemented with 2% B27 (Gibco), 10 mM HEPES, 0.38 mg/mL sodium bicarbonate, 0.29 mg/mL L-glutamine, 100 units/mL penicillin, 100 μg/mL streptomycin, and 0.2 mM luciferin; pH 7.2 (see Note 2). 3. DPBS: Dulbecco’s phosphate-buffered saline. 4. Trypsin-EDTA: EDTA
4Na.

0.05%

(w/v)

trypsin

and

0.53

mM

Circadian Chemical Screening

97

5. 15-cm Dish. 6. Pipette. 7. 50 mL tube. 8. 500 mL Glass bottle and magnetic stirrer bar: Sterilize using an autoclave. 9. White solid-bottom 384-well plate (catalog number 781073, Greiner) (see Note 3). 10. Optically clear film. 2.3

Compounds

1. 1 mM Solution in DMSO (see Note 4), aliquoted to each well of 384-well compound plates. Each compound plate requires DMSO control wells. For targeted screens, libraries of wellcharacterized compounds, such as LOPAC1280 (Library of Pharmacologically Active Compounds, Sigma-Aldrich), can be used. LOPAC covers wide variety of signaling pathways and drug target classes, such as GPCRs, kinases, ion channels, and transporters, providing a starting point for compound screens. In contrast, a large-scale library of uncharacterized compounds with diverse chemical scaffolds has the potential to probe many classes of targets, compared with limited numbers of well-characterized compounds. Screening facilities generally provide their own collection of structurally diverse compounds from commercial sources and academic labs. In this protocol, final compound concentration was determined as 7 μM (applying 500 nL of 1 mM compound to 70 μL medium) by a pilot screening of LOPAC1280 [17]. 2. 1 mM or 10 mM stock solution of the hit compounds.

2.4

Equipment

1. Microplate dispenser, MultiFlo FX (Biotek) (see Note 5). 2. Liquid handling workstation, Sciclone (Perkin Elmer) (see Note 5). 3. Plate reader, SynergyH1 (BioTek) with stacker, BioStack4 (BioTek). 4. Electronic Scientific).

multichannel

pipette,

E1-ClipTip

(Thermo

5. Magnetic stirrer. 6. CO2 incubator. 2.5

Software

1. Curve fitting software, MultiCycle or CellulaRhythm (see Note 6). Recording sixteen 384-well plates every 2 h for 5 days generates hundreds of thousands of data points. To obtain circadian parameters such as period, phase, and amplitude from the large amount of data, a specialized algorithm is required.

98

3

Megumi Hatori and Tsuyoshi Hirota

Methods

3.1 Cell Plating (For Sixteen 384-Well Plates)

1. Prepare Bmal1-dLuc reporter cells (see Note 1) on six 15-cm dishes at 60–70% confluency in culture medium. 2. Wash cells with 10 mL of DPBS (per dish) and treat with 3 mL of trypsin-EDTA at 37  C to fully detach (see Note 7). 3. Add 7 mL of cell culture medium and mix well with a pipette to stop trypsinization. 4. Transfer to 50 mL tube by combining three 15-cm dishes (total two tubes) and take 10 μL for cell number counting. 5. Centrifuge at 100  g for 5 min to pellet down the cells. 6. Discard supernatant and resuspend cells thoroughly (see Note 7) with an appropriate amount of cell culture medium for 2  106 cells/mL (combine two tubes). 7. Re-count cell number and dilute with cell culture medium to 1  105 cells/mL in a sterilized 500 mL glass bottle with a magnetic stirrer bar (prepare more than 220 mL; 12 mL for each plate, and more than 25 mL for priming etc. of a microplate dispenser). 8. Dispense 30 μL (3000 cells) per well onto white solid-bottom 384-well plates using a microplate dispenser. During this process, keep gently mixing the cell suspension in a glass bottle with a magnetic stirrer (see Note 8). 9. Grow cells at 37  C in a CO2 incubator for 2 days to reach confluence (see Note 9).

3.2 Compound Application

1. Add 40 μL of explant medium per well using a microplate dispenser. 2. Apply 500 nL of compound (dissolved in DMSO) to each well from a compound plate using a liquid handling workstation (see Note 10). 3. Cover each plate with optically clear film (see Note 11).

3.3 Luminescence Measurement

1. Set eight plates to a plate stacker (use two sets of a plate reader with stacker) placed in a 32  C room (see Note 12). 2. Measure luminescence (2 s exposure per well) every 120 min for 5 days. Each plate takes ~15 min for measurement including restacking time (see Note 13).

3.4

Data Analysis

1. Export raw luminescence data to text files (one plate per file). 2. Open with a curve fitting software and determine circadian parameters (see Note 14). Exclude the first day data from analysis, because of transient changes in luminescence upon explant medium application.

Circadian Chemical Screening

99

3. Choose primary hit compounds that cause period lengthening or shortening by more than 2 h compared with DMSO control (generally corresponding to 5 standard deviation (SD)). Period change cannot be evaluated for extremely low amplitude (arrhythmic) compounds that are potentially potent and/or toxic. Such compounds would be reevaluated at lower concentration(s) (for example, 2.4 μM). 3.5 Hit Compound Validation

1. Obtain 1 mM or 10 mM stock of the hit compounds (see Note 15) and test against two reporter cells (Bmal1-dLuc and Per2dLuc) by making 8 points of a threefold dilution series (total ~6500-fold dilution). When the number of compounds is relatively small, all liquid handling can be done manually by using electronic multichannel pipettes, instead of a microplate dispenser and liquid handling workstation. 2. Choose compounds that cause dose-dependent period changes in both reporter cells for mechanistic analyses (see Note 16).

4

Notes 1. It is important to use cells that show robust circadian rhythms for successful screening. Human U2OS osteosarcoma cells, mouse NIH-3T3 fibroblasts, and mouse primary fibroblasts have been used in circadian high-throughput chemical screening [17–29]. These cells are suitable for long-term recordings, because they survive under confluent conditions. A luciferase reporter driven by a clock gene promoter (Bmal1, Per2, etc.) can be introduced to the cells by virus infection or plasmid transfection to establish stable lines [37]. Because there is a heterogeneity of integration locus and copy number, it is better to select a clonal line that shows high amplitude, low dampening rate, and high intensity rhythms. Alternatively, immortalized cells can be prepared from mice harboring a Per2∷Luc knock-in reporter in which luciferase is fused to the C-terminus of PER2 protein at the endogenous Per2 locus [15, 38]. Note that their luminescence intensity can be lower because of the nature of the knock-in reporter (2 copies per cell) and difficult to detect with conventional plate readers, compared to the cells with a multi-copy reporter. Therefore, an additional SV40 poly (A) signal was introduced to enhance the signal [21]. 2. Phenol red-containing medium enables visual inspection of pH after the screening (contaminated wells become yellow), while slightly reduces luminescence signal. In explant medium, HEPES buffer maintains the pH in the absence of 5% CO2, and B-27 supplement synchronizes the cellular clock. Because the luciferase substrate luciferin slightly affects circadian rhythmicity [39], its concentration needs to be optimized.

100

Megumi Hatori and Tsuyoshi Hirota

3. White solid-bottom plates give a higher luminescence signal than other plates due to light reflection. When luminescence signals are low, it is better to choose plates with lower background counts to obtain a higher signal-to-noise ratio. 4. Concentration of compounds in the screening needs to be optimized, because higher concentration results in a larger number of hit compounds, including less potent ones, making follow-up studies laborious. 5. To obtain consistent and high-quality data from thousands of wells in the screening, it is important to use robotics that enable highly accurate liquid handling. Screening facilities generally provide such machines. Compound and cell plate information can be managed with barcodes by using a barcode reader. 6. CellulaRhythm is an algorithm for data analysis and visualization [17] based on the R-project computing environment (www.r-project.org). It fits raw luminescence data to a damped cosine curve by using nonlinear least squares to obtain circadian parameters. It also creates luminescence traces for each well and heat maps for the entire plate, which help visual inspection of the validity of the curve fit. MultiCycle (Actimetrics) is a commercially available sophisticated software that is detailed in Note 14. 7. Formation of cell clumps causes variability among wells. Therefore, it is important to dissociate the cells well with trypsinization and resuspension. 8. To avoid sedimentation of the cells, the cell suspension needs to be continuously mixed gently. 9. The cell number and medium volume need to be optimized for each cell type. Cell growth can be checked on a clear-bottom plate prepared in parallel. Evaporation, especially at the edge wells of the plate, is not trivial because of the small volume of the medium. Frequent opening of the incubator door results in higher evaporation and therefore needs to be minimized. 10. Height (i.e., distance from the cells) and speed of compound application need to be optimized to minimize cellular damage. Although 0.7% DMSO (500 nL of DMSO in 70 μL of medium) is tolerated by U2OS cells, some cells are more sensitive to DMSO, and its concentration needs to be optimized. The entire process should be kept as simple as possible to minimize liquid handling steps. 11. To avoid evaporation, it is necessary to stick the film well to the plate, especially at the edges. Protruding part of the film may cause the plates to be stuck in the machines (especially a plate stacker), and therefore needs to be cut off with a razor blade.

Circadian Chemical Screening

101

12. To maintain the temperature of plates in the stacker, the entire system (plate reader and stacker) was set in a warm room. Some machines may not work properly at high temperatures such as 37  C. 13. Throughput of measurement depends on the detection method of the plate reader. A photomultiplier tube (PMT)based reader is less expensive but takes a longer time for measurement by scanning the wells one by one (for example, 2 s per well takes ~15 min per plate). A CCD camera-based reader can measure all wells at once and therefore quickly (about 30 s per plate) but is very expensive. Exposure time needs to be optimized to reduce noise, while maintaining throughput. To minimize cross talk between neighboring wells, the distance between the detector and the plate needs to be adjusted. 14. MultiCycle software detrends raw luminescence data with a polynomial curve (usually first-order) or a 24-h running average, and then fits to a sine curve to obtain circadian parameters. It also visualizes raw or detrended luminescence profiles from multiple wells together with obtained parameters, enabling efficient analytical optimization and visual inspection. For the curve fitting, both the start and the end time points affect parameter estimation. They need to be optimized so that the obtained parameters properly reflect the raw data. Manual inspection of the profiles is required to filter out poorly fitted data that usually come from low amplitude. While the period length is consistent irrespective of the well position, edge wells tend to have shifted phase, making phase analysis complicated. 15. Hit compounds need to be evaluated by using fresh stocks to check the possibility of cross-contamination or decomposition in the compound plates used for the screening. 16. It is likely that the compounds with dose-dependent effects target key regulators (i.e., rheostats) of the clock. Different mechanisms of action of the compounds could result in reporter-specific effects. For example, CRY-stabilizing compounds cause signal reduction and dampening of Per2 reporter rhythms at higher concentrations compared with Bmal1, in addition to period lengthening, while CKI inhibitors similarly affect both reporters [23, 40]. The hit compounds need to be prioritized based on their effect, potency, structure, and property, because subsequent mechanistic studies, especially target identification, require a lot of effort and time.

102

Megumi Hatori and Tsuyoshi Hirota

Acknowledgments We thank Dr. Ayato Sato (ITbM Chemical Library Center, Nagoya University) for collaboration, and Dr. Simon Miller for proofreading. This work was supported in part by JSPS grants 18H02402, 20K21269, and 21H04766 to T.H., and AMED grant JP19gm6110026 to M.H. References 1. Takahashi JS (2017) Transcriptional architecture of the mammalian circadian clock. Nat Rev Genet 18:164–179. https://doi.org/10. 1038/nrg.2016.150 2. Chen Z, Yoo SH, Takahashi JS (2018) Development and therapeutic potential of smallmolecule modulators of circadian systems. Annu Rev Pharmacol Toxicol 58:231–252. https://doi.org/10.1146/annurev-pharmtox010617-052645 3. Miller S, Hirota T (2020) Pharmacological interventions to circadian clocks and their molecular bases. J Mol Biol 432:3498–3514. https://doi.org/10.1016/j.jmb.2020.01.003 4. Hatori M, Gill S, Mure LS, Goulding M, O’Leary DD, Panda S (2014) Lhx1 maintains synchrony among circadian oscillator neurons of the SCN. Elife 3:e03357. https://doi.org/ 10.7554/eLife.03357 5. Paul S, Hanna L, Harding C, Hayter EA, Walmsley L, Bechtold DA, Brown TM (2020) Output from VIP cells of the mammalian central clock regulates daily physiological rhythms. Nat Commun 11:1453. https://doi.org/10. 1038/s41467-020-15277-x 6. Miyawaki A, Niino Y (2015) Molecular spies for bioimaging—fluorescent protein-based probes. Mol Cell 58:632–643. https://doi. org/10.1016/j.molcel.2015.03.002 7. Ikeda M, Sugiyama T, Wallace CS, Gompf HS, Yoshioka T, Miyawaki A, Allen CN (2003) Circadian dynamics of cytosolic and nuclear Ca2+ in single suprachiasmatic nucleus neurons. Neuron 38:253–263. https://doi.org/10. 1016/s0896-6273(03)00164-8 8. Enoki R, Oda Y, Mieda M, Ono D, Honma S, Honma KI (2017) Synchronous circadian voltage rhythms with asynchronous calcium rhythms in the suprachiasmatic nucleus. Proc Natl Acad Sci U S A 114:E2476–E2485. https://doi.org/10.1073/pnas.1616815114 9. Ohta H, Yamazaki S, McMahon DG (2005) Constant light desynchronizes mammalian clock neurons. Nat Neurosci 8:267–269. https://doi.org/10.1038/nn1395

10. Yoshioka-Kobayashi K, Matsumiya M, Niino Y, Isomura A, Kori H, Miyawaki A, Kageyama R (2020) Coupling delay controls synchronized oscillation in the segmentation clock. Nature 580:119–123. https://doi.org/10.1038/ s41586-019-1882-z 11. Welsh DK, Imaizumi T, Kay SA (2005) Realtime reporting of circadian-regulated gene expression by luciferase imaging in plants and mammalian cells. Methods Enzymol 393:269– 288. https://doi.org/10.1016/S0076-6879 (05)93011-5 12. Balsalobre A, Damiola F, Schibler U (1998) A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93:929– 937. https://doi.org/10.1016/s0092-8674 (00)81199-x 13. Ueda HR, Chen W, Adachi A, Wakamatsu H, Hayashi S, Takasugi T, Nagano M, Nakahama K, Suzuki Y, Sugano S, Iino M, Shigeyoshi Y, Hashimoto S (2002) A transcription factor response element for gene expression during circadian night. Nature 418:534– 539. https://doi.org/10.1038/nature00906 14. Nagoshi E, Saini C, Bauer C, Laroche T, Naef F, Schibler U (2004) Circadian gene expression in individual fibroblasts: cellautonomous and self-sustained oscillators pass time to daughter cells. Cell 119:693–705. https://doi.org/10.1016/j.cell.2004.11.015 15. Welsh DK, Yoo SH, Liu AC, Takahashi JS, Kay SA (2004) Bioluminescence imaging of individual fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression. Curr Biol 14:2289–2295. https:// doi.org/10.1016/j.cub.2004.11.057 16. Liu AC, Tran HG, Zhang EE, Priest AA, Welsh DK, Kay SA (2008) Redundant function of REV-ERBalpha and beta and non-essential role for Bmal1 cycling in transcriptional regulation of intracellular circadian rhythms. PLoS Genet 4:e1000023. https://doi.org/10. 1371/journal.pgen.1000023 17. Hirota T, Lewis WG, Liu AC, Lee JW, Schultz PG, Kay SA (2008) A chemical biology

Circadian Chemical Screening approach reveals period shortening of the mammalian circadian clock by specific inhibition of GSK-3beta. Proc Natl Acad Sci U S A 105:20746–20751. https://doi.org/10. 1073/pnas.0811410106 18. Isojima Y, Nakajima M, Ukai H, Fujishima H, Yamada RG, Masumoto KH, Kiuchi R, Ishida M, Ukai-Tadenuma M, Minami Y, Kito R, Nakao K, Kishimoto W, Yoo SH, Shimomura K, Takao T, Takano A, Kojima T, Nagai K, Sakaki Y, Takahashi JS, Ueda HR (2009) CKIepsilon/delta-dependent phosphorylation is a temperature-insensitive, period-determining process in the mammalian circadian clock. Proc Natl Acad Sci U S A 106: 15744–15749. https://doi.org/10.1073/ pnas.0908733106 19. Hirota T, Lee JW, Lewis WG, Zhang EE, Breton G, Liu X, Garcia M, Peters EC, Etchegaray JP, Traver D, Schultz PG, Kay SA (2010) High-throughput chemical screen identifies a novel potent modulator of cellular circadian rhythms and reveals CKIalpha as a clock regulatory kinase. PLoS Biol 8: e1000559. https://doi.org/10.1371/journal. pbio.1000559 20. Lee JW, Hirota T, Peters EC, Garcia M, Gonzalez R, Cho CY, Wu X, Schultz PG, Kay SA (2011) A small molecule modulates circadian rhythms through phosphorylation of the period protein. Angew Chem Int Ed Engl 50: 10608–10611. https://doi.org/10.1002/ anie.201103915 21. Chen Z, Yoo SH, Park YS, Kim KH, Wei S, Buhr E, Ye ZY, Pan HL, Takahashi JS (2012) Identification of diverse modulators of central and peripheral circadian clocks by highthroughput chemical screening. Proc Natl Acad Sci U S A 109:101–106. https://doi. org/10.1073/pnas.1118034108 22. Oshima T, Niwa Y, Kuwata K, Srivastava A, Hyoda T, Tsuchiya Y, Kumagai M, Tsuyuguchi M, Tamaru T, Sugiyama A, Ono N, Zolboot N, Aikawa Y, Oishi S, Nonami A, Arai F, Hagihara S, Yamaguchi J, Tama F, Kunisaki Y, Yagita K, Ikeda M, Kinoshita T, Kay SA, Itami K, Hirota T (2019) Cell-based screen identifies a new potent and highly selective CK2 inhibitor for modulation of circadian rhythms and cancer cell growth. Sci Adv 5:eaau9060. https://doi. org/10.1126/sciadv.aau9060 23. Hirota T, Lee JW, St John PC, Sawa M, Iwaisako K, Noguchi T, Pongsawakul PY, Sonntag T, Welsh DK, Brenner DA, Doyle FJ III, Schultz PG, Kay SA (2012) Identification of small molecule activators of cryptochrome.

103

Science 337:1094–1097. https://doi.org/10. 1126/science.1223710 24. Miller S, Son YL, Aikawa Y, Makino E, Nagai Y, Srivastava A, Oshima T, Sugiyama A, Hara A, Abe K, Hirata K, Oishi S, Hagihara S, Sato A, Tama F, Itami K, Kay SA, Hatori M, Hirota T (2020) Isoform-selective regulation of mammalian cryptochromes. Nat Chem Biol 16:676–685. https://doi.org/10.1038/ s41589-020-0505-1 25. Miller S, Aikawa Y, Sugiyama A, Nagai Y, Hara A, Oshima T, Amaike K, Kay SA, Itami K, Hirota T (2020) An isoform-selective modulator of cryptochrome 1 regulates circadian rhythms in mammals. Cell Chem Biol 27: 1192–1198. https://doi.org/10.1016/j. chembiol.2020.05.008 26. Kolarski D, Miller S, Oshima T, Nagai Y, Aoki Y, Kobauri P, Srivastava A, Sugiyama A, Amaike K, Sato A, Tama F, Szymanski W, Feringa B, Itami K, Hirota T (2021) Photopharmacological manipulation of mammalian CRY1 for regulation of the circadian clock. J Am Chem Soc 143:2078-2087. https://doi. org/10.1021/jacs.0c12280 27. Tamai TK, Nakane Y, Ota W, Kobayashi A, Ishiguro M, Kadofusa N, Ikegami K, Yagita K, Shigeyoshi Y, Sudo M, NishiwakiOhkawa T, Sato A, Yoshimura T (2018) Identification of circadian clock modulators from existing drugs. EMBO Mol Med 10:e8724. h t t p s : // d o i . o r g / 1 0 . 1 5 2 5 2 / e m m m . 201708724 28. He B, Nohara K, Park N, Park YS, Guillory B, Zhao Z, Garcia JM, Koike N, Lee CC, Takahashi JS, Yoo SH, Chen Z (2016) The small molecule nobiletin targets the molecular oscillator to enhance circadian rhythms and protect against metabolic syndrome. Cell Metab 23: 610–621. https://doi.org/10.1016/j.cmet. 2016.03.007 29. Ju D, Zhang W, Yan J, Zhao H, Li W, Wang J, Liao M, Xu Z, Wang Z, Zhou G, Mei L, Hou N, Ying S, Cai T, Chen S, Xie X, Lai L, Tang C, Park N, Takahashi JS, Huang N, Qi X, Zhang EE (2020) Chemical perturbations reveal that RUVBL2 regulates the circadian phase in mammals. Sci Transl Med 12: eaba0 769. https://do i.org/10.1 126/ scitranslmed.aba0769 30. Badura L, Swanson T, Adamowicz W, Adams J, Cianfrogna J, Fisher K, Holland J, Kleiman R, Nelson F, Reynolds L, St Germain K, Schaeffer E, Tate B, Sprouse J (2007) An inhibitor of casein kinase I epsilon induces phase delays in circadian rhythms under free-running and entrained conditions. J Pharmacol Exp

104

Megumi Hatori and Tsuyoshi Hirota

Ther 322:730–738. https://doi.org/10. 1124/jpet.107.122846 31. Walton KM, Fisher K, Rubitski D, Marconi M, Meng QJ, Sladek M, Adams J, Bass M, Chandrasekaran R, Butler T, Griffor M, Rajamohan F, Serpa M, Chen Y, Claffey M, Hastings M, Loudon A, Maywood E, Ohren J, Doran A, Wager TT (2009) Selective inhibition of casein kinase 1 epsilon minimally alters circadian clock period. J Pharmacol Exp Ther 330:430–439. https://doi.org/10. 1124/jpet.109.151415 32. Meng QJ, Maywood ES, Bechtold DA, Lu WQ, Li J, Gibbs JE, Dupre SM, Chesham JE, Rajamohan F, Knafels J, Sneed B, Zawadzke LE, Ohren JF, Walton KM, Wager TT, Hastings MH, Loudon AS (2010) Entrainment of disrupted circadian behavior through inhibition of casein kinase 1 (CK1) enzymes. Proc Natl Acad Sci U S A 107:15240–15245. https://doi.org/10.1073/pnas.1005101107 33. Solt LA, Wang Y, Banerjee S, Hughes T, Kojetin DJ, Lundasen T, Shin Y, Liu J, Cameron MD, Noel R, Yoo SH, Takahashi JS, Butler AA, Kamenecka TM, Burris TP (2012) Regulation of circadian behaviour and metabolism by synthetic REV-ERB agonists. Nature 485:62–68. https://doi.org/10.1038/nature11030 34. Chun SK, Jang J, Chung S, Yun H, Kim NJ, Jung JW, Son GH, Suh YG, Kim K (2014) Identification and validation of cryptochrome inhibitors that modulate the molecular circadian clock. ACS Chem Biol 9:703–710. https://doi.org/10.1021/cb400752k 35. Doruk YU, Yarparvar D, Akyel YK, Gul S, Taskin AC, Yilmaz F, Baris I, Ozturk N, Turkay M, Ozturk N, Okyar A, Kavakli IH (2020) A CLOCK-binding small molecule

disrupts the interaction between CLOCK and BMAL1 and enhances circadian rhythm amplitude. J Biol Chem 295:3518–3531. https:// doi.org/10.1074/jbc.RA119.011332 36. Zhang EE, Liu AC, Hirota T, Miraglia LJ, Welch G, Pongsawakul PY, Liu X, Atwood A, Huss JW III, Janes J, Su AI, Hogenesch JB, Kay SA (2009) A genome-wide RNAi screen for modifiers of the circadian clock in human cells. Cell 139:199–210. https://doi.org/10. 1016/j.cell.2009.08.031 37. Ramanathan C, Khan SK, Kathale ND, Xu H, Liu AC (2012) Monitoring cell-autonomous circadian clock rhythms of gene expression using luciferase bioluminescence reporters. J Vis Exp 2012:e4234. https://doi.org/10. 3791/4234 38. Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, Siepka SM, Hong HK, Oh WJ, Yoo OJ, Menaker M, Takahashi JS (2004) PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101:5339–5346. https://doi.org/10. 1073/pnas.0308709101 39. Feeney KA, Putker M, Brancaccio M, O’Neill JS (2016) In-depth characterization of firefly luciferase as a reporter of circadian gene expression in mammalian cells. J Biol Rhythm 31: 5 4 0 – 5 5 0 . h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 0748730416668898 40. St John PC, Hirota T, Kay SA, Doyle FJ III (2014) Spatiotemporal separation of PER and CRY posttranslational regulation in the mammalian circadian clock. Proc Natl Acad Sci U S A 111:2040–2045. https://doi.org/10. 1073/pnas.1323618111

Chapter 7 Methods for Assessing Circadian Rhythms and Cell Cycle in Intestinal Enteroids Miri Park, Yuhui Cao, and Christian I. Hong Abstract Endogenous circadian clocks play a key role in regulating a vast array of biological processes from cell cycle to metabolism, and disruption of circadian rhythms exacerbates a range of human ailments including cardiovascular, metabolic, and gastrointestinal diseases. Determining the state of a patient’s circadian rhythms and clock-controlled signaling pathways has important implications for precision and personalized medicine, from improving the diagnosis of circadian-related disorders to optimizing the timing of drug delivery. Patient-derived 3-dimensional enteroids or in vitro “mini gut” is an attractive model uncovering human- and patient-specific circadian target genes that may be critical for personalized medicine. Here, we introduce several procedures to assess circadian rhythms and cell cycle dynamics in enteroids through time course sample collection methods and assay techniques including immunofluorescence, live cell confocal microscopy, and bioluminescence. These methods can be applied to evaluate the state of circadian rhythms and circadian clock-gated cell division cycles using mouse and human intestinal enteroids. Key words Circadian rhythm, Cell cycle, Intestinal enteroids, Time course sampling, Immunofluorescence, Bioluminescence assay

1

Introduction The circadian clock regulates numerous repertoires of periodic biochemical, physiological, and behavioral processes to anticipate and align physiological functions according to daily environmental changes. At the cellular level, circadian rhythms are controlled by a complex network of transcriptional-translational feedback loops (TTFL) generating autonomous circadian oscillations even in single cells [1, 2]. In mammals, the heterodimeric circadian transcription factors (positive elements), CLOCK and BMAL1, activate core

Miri Park and Yuhui Cao contributed equally with all other contributors. Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_7, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

105

106

Miri Park et al.

circadian clock genes (negative elements), Period (Per1,2,3) and Cryptochrome (Cry1,2). Subsequently, PER and CRY complexes inhibit CLOCK/BMAL1, which establishes a time-delayed negative feedback loop generating autonomous circadian oscillations [3–5]. Recent findings uncovered tissue-specific functions of circadian rhythms regulating numerous clock-controlled signaling pathways including stem cell regeneration and tissue homeostasis [6– 9]. Importantly, disruption of circadian rhythms has been associated with increased severity of human diseases including cancer, cardiovascular, and gastrointestinal diseases. This suggests that assessments of clock-controlled genes and signaling pathways in human disease tissues could provide critical information for disease risk prediction, diagnostic assay, and chronotherapeutic treatment regimens. Most of the circadian clock-related studies, however, relied heavily on genetically engineered mice or immortalized human cell lines that may not appropriately reflect patients’ disease states with respect to their circadian-related conditions. In contrast, patient-derived organoids can uncover human- and disease-specific mechanisms that will be critical to design novel chronotherapeutic regimens. 3D intestinal organoids (enteroids) are multicellular in vitro model systems that mimic in vivo tissue structure and physiological properties, which cannot be obtained in homogeneous cell lines. Hans Clevers’ group first demonstrated that isolated Lgr5+ (leucine-rich repeat-containing G-protein coupled) adult intestinal stem cells from crypt domains spontaneously grow into 3D structure in the Matrigel™ with appropriate medium conditions [10, 11]. Since this discovery, enteroids and other organotypic organoids have emerged as powerful tools to investigate human diseases using patient tissue-derived organoids. In our previous work, we utilized both mouse and human intestinal enteroids to characterize robustness of circadian rhythms, clock-controlled genes (CCGs), circadian clock-gated cell division cycles, and circadian time-dependent necrotic cell death responses [12– 15]. Intriguingly, we observe anti-phasic necrotic cell death responses in mouse vs. human intestinal enteroids. Our data indicate that integrated usages of both mouse and human intestinal enteroids can elucidate both conserved and species-specific mechanisms between mouse and human intestine. Here, we provide experimental methods to investigate circadian rhythms, CCGs, and circadian clock-gated cell divisions. These methods can be used to investigate functions of the peripheral circadian clocks in mouse and human enteroids, which may uncover potential circadian target genes and chronotherapeutic regimens for translational applications.

Assessing Circadian Rhythms and Cell Cycle in Enteroids

2 2.1

107

Materials Animals

1. C57BL/6J (Stock No. 000644, Jackson Laboratory) (see Note 1). 2. PER2::LUCIFERASE (Stock No. 0066852, Jackson Laboratory; [16]). 3. FUCCI2 (Riken Acc. No. CDB0203T; [17]).

2.2 Growth and Expansion of Enteroids

1. CO2 incubator. 2. Tissue forceps. 3. Surgical scissors. 4. 50 ml Conical centrifuge tubes (Corning, 14-432-22). 5. 15 ml Conical centrifuge tubes (Corning, 14-959-53A). 6. 4-Well cell culture plates (Nunc, 144444). 7. 24-Well cell culture plates (Nunc, 142475). 8. 48-Well cell culture plates (Nunc, 150687). 9. 35
10 mm BD Falcon tissue culture dish (Fisher, 08-772A). 10. BD 1-ml conventional insulin syringes for mouse enteroids (BD, 14-820-28). 11. BD Micro-Fine™ IV insulin syringes for human enteroids (BD, 14-829-1B). 12. Cryogenic vials (Corning, 13-700-502). 13. Growth factor reduced Matrigel™ (Corning, CB-40230C). 14. 70 μm Cell strainer (Fisherbrand, 22363548). 15. Minigut medium: Advanced DMEM/F12 (Life Technologies, 12634-028) supplemented with 2 mM GlutaMAX™ (Gibco, 35050061), 10 mM HEPES (Gibco, 15630080), 100 U/ml penicillin/100 μg/ml streptomycin (Corning, 30002CI), 1
N2 (Gibco, 17502-048), 1
B27 supplements (Gibco, 17504-044), 10% R-spondin1 conditioned medium, 10% Noggin conditioned medium, and 50 ng/ml EGF (PeproTech, 315-09). 16. Chelation buffer: Freshly prepare 30 ml of crypt chelation buffer from stocks of 0.5 M Ethylenediaminetetraacetic acid, pH 8 (EDTA) and PBS. The final concentration of 2 mM EDTA solution will be used. 17. Cryopreservation medium: Recovery Cell Culture Freezing Medium (Gibco, 12648-010). 18. R-Spondin-producing cell line (available from ATCC). 19. Noggin-producing cell line (available from ATCC).

108

Miri Park et al.

20. Cell line growing medium: DMEM (Life technologies 10569010) supplemented with 1% streptomycin/penicillin and 10% FBS. 21. R-Spondin primary culture medium: Opti-MEM reducedserum medium (Life technologies 51985034) supplemented with 1% streptomycin/penicillin. 22. Noggin primary culture medium: Advanced DMEM/F12 (Life Technologies, 12634-028) supplemented with 1% streptomycin/penicillin, 1% 1 M Hepes and 1% GlutaMAX™. 23. Zeocin (Life Technologies R250-01). 24. Geneticine G418 (Life Technologies 10131-035). 25. Trypsin, 2.5% (wt/vol) (Invitrogen 15090046). 26. Cell culture flasks, 150 cm2 (Corning 430825). 2.3 Media for Human Intestinal Enteroids (HIEs)

1. Expansion medium: Intesticult Organoid Growth Medium (StemCell, 06010).

2.4 Lentiviral Transduction of Mouse and Human Enteroids

1. Bmal1-luciferase (Addgene, pABpuro-BluF #46824).

2. Differentiation medium: Intesticult OGM Human Basal Medium (StemCell, 100-0190) mix at a 1:1 ratio with Advanced DMEM/F12 supplemented with 15 mM HEPES.

2. Viral stock with a titer of 106. 3. Glycogen synthase kinase 3 (GSK-3) inhibitor (CHIR99021; Cayman Chemical, 13122). 4. Rho-associated protein kinase (ROCK) inhibitor (Y-27632). 5. 0.05% Trypsin/0.5 mM EDTA solution (Gibco, 25200056). 6. Polybrene (Millipore-Sigma, TR-1003-G). 7. Puromycin (Invivogen, ant-pr-1).

2.5 RNA Isolation and Processing

1. RNeasy mini-kit column (Qiagen, 74106).

2.6 Whole-Mount Staining [18]

1. Primary antibody (Abcam Rabbit monoclonal to Ki67, Cat# ab16667).

2. Superscript III 18080085).

Reverse

transcriptase

kit

(Invitrogen,

2. Secondary antibody (Donkey anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor 647, Cat # A31573, Thermo Fisher Scientific). 3. 1 μg/ml DAPI (ThermoFisher Scientific Cat# 62247). 4. 0.01% Triton X-100 (Sigma-Aldrich Cat# 9002-93-1). 5. Serum (Normal Donkey Serum, 017-000-121, Jackson ImmunoResearch Laboratories, INC.).

Assessing Circadian Rhythms and Cell Cycle in Enteroids

3

109

Methods

3.1 Preparation of RSpondin and Noggin Conditioned Media [19]

1. Prewarm 25 ml of cell line growing medium to 37  C in a 50 ml conical Falcon tube. 2. Recover a cryotube of R-spondin-producing cells and/or a cryotube of Noggin-producing cells in a 37  C water bath. 3. Transfer the recovered cells to the cell line growing medium in the conical Falcon tube. 4. Transfer the homogenized medium from the Falcon tube into a 150 cm2 cell culture flask and incubate them for 1 day. 5. Add Zeocin (600 μg/ml) and Geneticine G418 (500 μg/ml) to R-Spondin and Noggin cells, respectively, for selection. 6. Grow the cells until they become confluent (2–3 days). 7. Expand culture into five dishes. (a) Wash the cells with 20 ml of PBS and aspirate. (b) Add 1 ml of trypsin to the flask. Tap the flask several times to coat the plate surface. (c) Incubate the flask for 5 min at 37  C. (d) Suspend the cells in 12 ml of cell line growing medium. (e) Add 120 ml of cell line growing medium into the flask. Divide the cell suspension into five 150 cm2 cell culture flasks (25 ml per flask). (f) Incubate the flasks in a cell culture incubator until the cells reach to 100% confluency (3–4 days). (g) Wash the cells with 10 ml of R-Spondin or Noggin primary culture medium, respectively, and aspirate. (h) Add 25 ml of R-Spondin or Noggin primary culture medium, correspondingly. (i) Incubate the flasks in a cell culture incubator for 24 h. 8. Collection of conditioned media. (a) Collect the conditioned media into 50-ml centrifuge tubes and add a new corresponding primary culture medium to the flasks. Return the flasks to the cell culture incubator. (b) Centrifuge the tubes at 2000
g for 15 min at 4  C and decant the supernatant carefully in a 1-l bottle. Store the bottle at 4  C (first batch of conditioned medium). (c) Every 24 h, collect the second, third, and fourth conditioned media as described above. 9. Discard the cell lines after the fourth collection.

110

Miri Park et al.

3.2 Isolation, Expansion, and Maintenance of Mouse Enteroids [20, 21]

Briefly: 1. Dissect approximately 3–4 in. of jejunum from a 3–6-monthold mouse. 2. Place it on a petri dish and flush with ice-cold PBS through the intestine to clean and remove waste. 3. Cut it into three pieces and splay open with scissors, villi facing upward. 4. Drag glass slides over the top to remove the villi. 5. Place the cleaned intestinal tissue in a conical tube filled with 25–30 ml ice-cold PBS. 6. Invert the tube a few times for washing and repeat the process by placing it in a new tube with PBS. 7. Put them in an ice-cold chelation buffer. 8. Place the suspension on a rocking table at 4  C for 30 min (see Note 2). 9. After the crypts are separated, remove the tissue using forceps, pellet the intestinal crypt suspension at 800 rpm for 5 min, and wash with 15 ml Advanced DMEM. 10. Spin down at 600 rpm for 2 min, resuspend with 5 ml Advanced DMEM, and then filter it through a 70 μm cell strainer into a new 50 ml conical tube. 11. Spin down intestinal crypts using a centrifuge at 1000 rpm for 5 min and resuspend them in an appropriate amount of Matrigel (40 μl Matrigel/well). Then, plate the Matrigel suspension as a single 3D Matrigel dome onto a 24-well tissue culture plate. 12. Polymerize the Matrigel suspension at 37  C for 20–30 min and add 400 μl minigut medium. 13. Replace medium every 2–3 days. 14. Maintain enteroids in culture for 6–7 days before being propagated into additional wells or used for experimental procedures. 15. Passage enteroids using a syringe.

3.3 Preparation of HIE Culture [22]

1. Plate HIEs in three separate 10 μl Matrigel bubbles/well on a 24-well plate (see Note 3). 2. Add 350 μl of expansion medium to each well and incubate at 37  C for 2 days (see Note 4). 3. Replace expansion medium with differentiation medium and maintain for 4 days before the experiments are performed (see Note 5).

Assessing Circadian Rhythms and Cell Cycle in Enteroids

3.4 Forty-Eight-Hour Time Course Sample Collection with 2h Resolution for RNA and/or Protein Isolation

111

1. Prepare 24 dishes of mouse or human enteroids using 35-mm dishes. 2. For mouse enteroids: (a) Prepare 48 wells of mouse enteroids in two 24-well plates with each well containing a 40 μl Matrigel bubble and 400 μl of minigut medium. Each well contains about 100–200 mouse enteroids (see Note 6). (b) Passage enteroids to 24 35-mm dishes with 7 12 μl-Matrigel bubbles/dish, and add 2 ml of minigut medium postMatrigel solidification. 3. For human intestinal enteroids: (a) Expand HIEs to 48 wells using 24-well plates, each well containing three 10 μl Matrigel bubbles and 350 μl of expansion medium. (b) Induce differentiation: After 2 days, change expansion medium to 350 μl differentiation medium. (c) Transfer HIEs from 24-well plates to 24 35-mm dishes: Collect HIEs from 48 wells, spin them down, and remove supernatant as much as possible to eliminate Matrigel improving pellet density. (d) Re-pellet HIEs in 2 ml PBS to check the density. If the density of HIEs is 100–200/40 μl, then remove PBS and resuspend HIEs in 2 ml Matrigel. (e) Plate HIEs to 24 35-mm dishes with 7 12 μl-Matrigel bubbles/dish, and add 2 ml of differentiation medium/ dish (see Note 7). 4. After incubation at 37  C for 23 h, add 100 nM dexamethasone (Dex) for 1 h to synchronize circadian rhythms (see Note 8). 5. Aspirate the supernatant and add 3 ml of fresh medium (minigut medium for mouse enteroids and differentiation medium to HIEs) to each dish, and then place dishes at 37  C. 6. Timepoint collection: Harvest enteroids every 2 h over 48 h from 24 to 70 h post Dex reset [23]: (a) Spin down. (b) Remove supernatant. (c) Resuspend enteroids pellet in 1 ml per dish of ice-cold Trizol. (d) Vortex for 30 s to 1 min until the remaining Matrigel is completely dissolved in Trizol. (e) Immediately snap-freeze tube in liquid N2 and store at 80  C until the RNA/protein is isolated (see Note 9). 7. Isolate RNA: After completion of the time course, isolate RNA via a RNeasy mini-kit column and generate cDNA using the

112

Miri Park et al.

Fig. 1 Rhythmic expression of representative circadian clock genes in mouse and human enteroids. (a) qRT-PCR data of mRNA expressions of Bmal1 (black) and Rev-erbα (red) in mouse enteroids. (b) qRT-PCR data of mRNA expressions of Bmal1 (black) and Rev-erbα (red) in HIEs

Superscript III Reverse transcriptase kit according to manufacturer’s instructions. Perform quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) (see Note 10) (Fig. 1). 3.5 Bioluminescent Assay

For real-time monitoring of circadian oscillations, prepare enteroids possessing circadian bioluminescent reporters. We either use PER2::LUC mice [16] to derive PER2::LUC mouse enteroids or lentiviral transduction for stable expression of Bmal1-luc (Addgene #46824 [24]) in organoids of interest. Bmal1-luc and PER2::LUC are transcriptional and translational reporters reflecting Bmal1 promoter activity and PER2 protein, respectively. Lentiviral transduction of mouse and human enteroids is described in previous publications [25]. We use a photomultiplier tube (PMT)-based ATTO Kronos Dio incubating luminometer which also maintains 37  C and 5% CO2. We suggest the users optimize their protocol if they are using other equipment such as LumiCycle by Actimetrics. 1. Plate mouse or human enteroids with one 30 μl Matrigel bubble per dish by passage for mouse enteroids and replating for HIEs (see Subheading 3.1, step 1). Add 2 ml of medium and incubate at 37  C (see Note 11). 2. After 23 h, add 100 nM Dex and incubate at 37  C for 1 h. 3. Replace the medium with 2 ml of fresh minigut (mouse enteroids) or differentiation (HIEs) medium with 200 μM Beetle Luciferin Potassium salt (Promega, E1601). 4. Place dishes in a Kronos Dio incubating luminometer (ATTO, AB-2550) and monitor signals up to 3–5 days (see Note 12). 5. Analyze oscillation data using the Kronos software (Fig. 2) (see Note 13).

Assessing Circadian Rhythms and Cell Cycle in Enteroids

113

Fig. 2 Population-level analysis of circadian rhythms in mouse and human enteroids using bioluminescent reporters. Representative recordings of the bioluminescence from PER2::LUC in mouse enteroids (a) and Bmal1-luc in HIEs (b). Raw data, noise filtered data with moving average, and detrended data are shown 3.6 Time Course Experiments to Measure Cell Proliferation Index Using Immunofluorescence Imaging

1. For a 24-h time course with 4-h resolution, prepare 48 wells of mouse enteroids using six 8-well chamber slides. (a) Prepare 24 wells of mouse enteroids in one 24-well plate with each well containing one 40-μl of Matrigel bubble and 400 μl minigut medium. Each well contains about 100–200 mouse enteroids (see Note 14). (b) Passage enteroids to six 8-well chamber slides with one 20-μl Matrigel bubbles/well, and add 250 μl minigut medium post-Matrigel solidification. 2. After incubation at 37  C for 23 h, add 100 nM dexamethasone (Dex) for 1 h to synchronize circadian rhythms. 3. Aspirate the supernatant and add 250 μl minigut medium to each well, and then place chamber slides at 37  C. 4. Fixation: Collect samples from different time points and fix enteroids using 3.7% formaldehyde (see Note 15). (a) Remove mini gut medium and add 250 μl of PBS to wash mini gut medium (see Note 16). (b) Remove PBS immediately and add 3.7% formaldehyde in PBS (250 μl) for 15 min to fix samples. (c) Wash formaldehyde: Remove 3.7% formaldehyde, add 250 μl of PBS, and wait for 5 min (see Note 17). 5. Whole-mount staining of intestinal enteroids.

114

Miri Park et al.

(a) Permeabilize the cell membrane: Add 250 μl of 0.5% Triton X-100 in PBS for 20 min at RT. (b) Remove the 250 μl of 0.5% Triton X-100 and add 250 μl of 0.01% Triton X-100 in PBS to wash for 5 min at RT. (c) Blocking: Remove the 250 μl of 0.01% Triton X-100 and add 250 μl of 2% donkey serum diluted in 0.01% Triton X-100 for 1 h at RT. (d) Primary antibody treatment: Remove blocking buffer and incubate enteroids with 250 μl of primary antibody diluted 1:200 in 0.01% Triton X-100 in PBS at 4  C overnight. (e) Transfer enteroids from 4  C to RT and let them equilibrate for 5–10 min to prevent Matrigel dissolution. (f) Wash enteroids twice with 250 μl of 0.01% Triton X-100 in PBS for 5 min. (g) Secondary antibody treatment: Remove the 250 μl of 0.01% Triton X-100 and add 250 μl of secondary antibody at 4  C overnight (1:100 dilution of 647 donkey anti-rabbit antibody) (see Note 18). (h) Transfer them to RT and equilibrate for 10 min, and wash them with 250 μl of 0.01% Triton X-100 in PBS. (i) For imaging of the nuclei, add 250 μl of 1 μg/ml DAPI in PBS for 2–10 min at RT. (j) Remove DAPI and wash with 250 μl of 0.01% Triton X-100 in PBS. (k) To store samples for a longer period, add 250 μl of PBS and cover chamber slides with aluminum foil. Samples can be stored at 4  C up to 1 week. (l) Image acquisition: Use a confocal microscope to detect Ki67 and DAPI with 633 nm and 405 nm wavelength excitation lasers, respectively (Fig. 3) (see Note 19). (m) Data analysis can be done using a microscopy software and measure cell proliferation index by quantifying the ratio between Ki67 and DAPI positive cells (see Note 20). 3.7 Time Course Live Cell Confocal Microscopy for Cell Cycle Analysis

One can use fluorescent cell cycle reporters to quantify cell cycle dynamics in mouse and human enteroids. We used Fluorescent Ubiquitination-based Cell Cycle Indicator (FUCCI2) system, which labels G1 and S/G2/M phases with mCherry-hCDT1 and mVenus-hGeminin, respectively [26], and Zeiss LSM 710 confocal microscope.

Assessing Circadian Rhythms and Cell Cycle in Enteroids

115

Fig. 3 Representative images of fixed and stained mouse enteroids. Proliferative cells are stained with Ki67 (magenta) antibody and nuclei are labeled with DAPI (blue). Scale bar ¼ 40 μm

1. For a 48- to 70-h time course with 30-min resolution confocal microscopy, prepare 8 wells of mouse enteroids using one 8-well chamber slide. (a) Prepare 4 wells of mouse enteroids in one 4-well plate with each well containing one 40-μl Matrigel bubble and 400 μl minigut medium. Each well contains about 100–200 mouse enteroids. (b) Passage enteroids to one 8-well chamber slides with one 20-μl Matrigel bubble/well, and add 250 μl of minigut medium post-Matrigel solidification. Each well contains about 20–50 mouse enteroids. 2. After incubation at 37  C for 2 days, image enteroids by confocal microscope. (a) Replace minigut medium before initiating confocal microscopy imaging. (b) 560 nm and 514 nm wavelength excitation lasers were used to detect mCherry-hCDT1 and mVenus-hGeminin signals, respectively.

116

Miri Park et al.

(c) The distance interval between two consecutive Z-stacks was 4 μm. (d) Images are acquired for 60–70 h with a 30-min time interval (see Note 21). 3. Image processing can be done by different microscopy software. Below we describe our protocol using ZEN, NIS-element, and Imaris. (a) Use ZEN software to export all the raw images: After the time course, export all the images from ZEN in Tagged Image File (tif) format by Series and in RGB color. Make sure to save all images in one folder. (b) Use NIS-element software to improve image quality. Reopen all the images by time and Z-stack using NIS-element. Change Intensity of each channel by using visualization controls_LUTS function from VIEW. Apply image processing by using Advanced Denoising function from IMAGE. This is optional. Please apply this only when necessary because this might cause loss of signals. Apply image processing by using the Smooth function from IMAGE. This is optional. Please apply this only when necessary because this might cause loss of signals. (c) Use Imaris software to change scales in Z direction and use this file for subsequent analysis. 4. Manual tracking of cell cycle times (CCTs) in mouse enteroids using Imaris software. (a) Use SPOT module to identify green tracks (S/G2/M cells). SPOT module can identify spots at any frame and position. Identify cells in consecutive frames manually and then link them together (see Note 22). (b) Match each green track (S/G2/M cell) with a red track (G1 cell) by manually navigating through previous frames (Fig. 4) (see Note 23). (c) Export information of each matched red and green tracks, and calculate CCTs of single cells by subtracting the first time point of the red signal from the last time point of the corresponding green signal. 5. Automatic tracking of the number of cells in S/G2/M phases over time using Imaris software (Fig. 5). (a) Add new spots. (b) Check track spots and click the forward arrow. (c) Input estimated XY diameter by measuring dimensions via 2D image version (called Slice), click forward arrow.

Assessing Circadian Rhythms and Cell Cycle in Enteroids

117

Fig. 4 Representative cell track of FUCCI2 enteroids transitioning from G1 to S/G2/M phase. Cells in G1 and S/G2/M phases are indicated by mCherry-hCDT1 and mVenus-hGeminin, respectively. A representative single cell is tracked to measure cell cycle time from G1 to the end of Mitosis (yellow circle). Yellow arrow indicates transition from G1 to S/G2/M phase. Time interval ¼ 30 min. Scale bar ¼ 10 μm

Fig. 5 Measurements of the number of cells in the S/G2/M phase over time in a single mouse enteroid. (a) Representative images showing the spatial distribution of mVenus-hGeminin (green, S/G2/M) and mCherryhCdt1 (red G0/G1) at 12 (left) and 30 (right) h. Scale bar, 50 μm. (b) Representative traces of the number of mVenus-hGeminin-positive (green) cells in a single FUCCI2 enteroid

118

Miri Park et al.

(d) Select a filter type suitable for the spot to be measured and adjust the threshold levels, click forward arrow. The default filter is suitable for spot tracking in most cases. (e) Check the time-lapse images and correct any spots that need to be added or deleted, click forward arrow. (f) Select an appropriate algorithm according to the movement of the spot, modify the parameters for optimized tracking, and click forward arrow. (g) Check the tracked image and click the double-green arrow to finish if there is no problem. (h) Go to the Statistics and click Save to check the number of spots per time point in an Excel file (see Note 24).

4

Notes 1. We used mouse enteroids derived from C57BL/6J mice for RNA-Seq so that our data can be compared with other RNA-Seq data with the same genotype, but users may choose another control based on their experimental design. 2. Gently shake the conical tube by hand for 1 min every 10 min to dissociate epithelium from the basement membrane. 3. Invert plates while solidifying Matrigel to minimize the distance between enteroids and medium by positioning enteroids to the surface of Matrigel. 4. Change medium every 2-days for the HIEs because they shrink rapidly if nutrients are depleted. 5. This step is necessary to increase the number of terminally differentiated cells representing different cell types in the small intestine. Passage/expansion is not possible using differentiation media. Prepare an appropriate amount of enteroids before switching to differentiation media for experiments. 6. Determine the optimal enteroid seeding density before adding Matrigel. 7. Seed HIEs at the desired seeding density with Matrigel for optimal HIEs growth. Low seeding density may significantly decrease cell survival. Split enteroids-Matrigel suspension to 2–3 separate Eppendorf’s tubes before plating to prevent Matrigel solidification while plating. We use 24 separate dishes for sample collections at 2-h intervals over 48 h to minimize potential variabilities. 8. Dex induces the expression of a core clock gene, Per1, and works as a strong resetting cue [27]. 9. Pipet 2–3 times before spinning down enteroids to dissociate enteroids from the Matrigel. This process helps separating enteroid from the Matrigel during centrifugation.

Assessing Circadian Rhythms and Cell Cycle in Enteroids

119

Table 1 A representative timepoint schedule for a continuous time course Day0

Day1

Day2

Day3

Day4

Midnight

Midnight

Midnight

Midnight

Midnight

1:00

1:00

1:00

1:00

CT40 1:00

2:00

2:00

2:00

2:00

2:00

3:00

3:00

3:00

3:00

CT42 3:00

4:00

4:00

4:00

4:00

4:00

5:00

5:00

5:00

5:00

CT44 5:00

6:00

6:00

6:00

6:00

6:00

7:00

7:00

7:00

7:00

CT46 7:00

8:00

8:00

Dex 8:00 shock

8:00

8:00

9:00

Change Media

CT24 9:00

CT48 9:00

9:00

Passage enteroids

9:00

10:00

10:00

10:00

10:00

10:00

11:00

11:00

11:00

CT26 11:00

CT50 11:00

Noon

Noon

Noon

Noon

Noon

1:00

1:00

1:00

CT28 1:00

CT52 1:00

2:00

2:00

2:00

2:00

2:00

3:00

3:00

3:00

CT30 3:00

CT54 3:00

4:00

4:00

4:00

4:00

4:00

5:00

5:00

5:00

CT32 5:00

CT56 5:00

6:00

6:00

6:00

6:00

6:00

7:00

7:00

7:00

CT34 7:00

CT58 7:00

8:00

8:00

8:00

8:00

8:00

9:00

9:00

9:00

CT36 9:00

CT60 9:00

10:00

10:00

10:00

10:00

10:00

11:00

11:00

11:00

CT38 11:00

CT62 11:00

CT64

CT66

CT68

CT70

10. We use TATA-Box Binding protein (Tbp) expression as a housekeeping gene (Fig. 1). For RNA-Sequencing, we use a continuous time course protocol as described in Table 1 where two to three people take shifts to minimize potential variabilities from a staggering method. Alternatively, one can use a staggering method, which uses different timing of dexamethasone treatments to consolidate the timing of sample collections as described in Table 2.

Noon

1:00

2:00

3:00

4:00

1:00

2:00

3:00

4:00

8:00

8:00

Noon

7:00

7:00

11:00

6:00

6:00

11:00

5:00

5:00

10:00

4:00

4:00

10:00

3:00

3:00

9:00

2:00

2:00

9:00

1:00

1:00

Passage enteroids

Midnight

Batch #1, 2, 3 Time

Time

Midnight

Day1

Day0

Batch #2

6:00

5:00

4:00

3:00

2:00

1:00

Midnight

Time

4:00

3:00

2:00

1:00

Noon

11:00

10:00

9:00

Change 8:00 Media

Dex 7:00 shock

Batch #3

Day2

Table 2 A representative timepoint schedule for a staggering method

Change Media

Dex shock

Batch #1

Batch #2

Batch #3

4:00

3:00

2:00

1:00

Noon

11:00

10:00

9:00

8:00

7:00

6:00

5:00

4:00

3:00

2:00

1:00

Midnight

Time

Day3

CT32

CT30

CT28

CT26

CT24

Batch #1

CT44

CT42

CT40

CT38

CT36

Batch #2

CT56

CT54

CT52

CT50

CT48

Batch #3

4:00

3:00

2:00

1:00

Noon

11:00

10:00

9:00

8:00

7:00

6:00

5:00

4:00

3:00

2:00

1:00

Midnight

Time

Day4

CT68

CT66

CT64

CT62

CT60

Batch #2

120 Miri Park et al.

5:00

6:00

7:00

8:00

9:00

10:00

11:00

5:00

6:00

7:00

8:00

9:00

10:00

11:00

Change Media

Dex shock

11:00

10:00

9:00

8:00

7:00

6:00

5:00

11:00

10:00

9:00

8:00

7:00

6:00

5:00 CT34

CT46

CT58

11:00

10:00

9:00

8:00

7:00

6:00

5:00 CT70

Assessing Circadian Rhythms and Cell Cycle in Enteroids 121

122

Miri Park et al.

11. ATTO Kronos Dio can measure up to eight 35-mm dishes per batch in real time. To reduce variabilities (e.g., intensity of bioluminescence, amplitude of oscillations), one can control the density of enteroids by manually picking the crypt or HIEs with a microscope and seed a fixed number of enteroids/plate. 12. For 8 dishes, we measure bioluminescence with 9-min intervals. 13. The Kronos software provides data analysis functions such as moving average and detrending. 14. Determine the optimal enteroid seeding density before adding Matrigel. 15. As described above, users can adapt continuous or staggered time course schedule modifying Table 1 or 2, respectively. 16. Any solution applied to the chamber slide must be at room temperature (RT) to avoid Matrigel dissolution. 17. To store the sample for a longer period of time (up to 1 week at 4  C), replace the PBS and store at 4  C. 18. From this step, chamber slides should be covered with aluminum foil until image acquisition. 19. For each genotype and time point, we recommend to image 20–25 enteroids for quantification. 20. We use Imaris Microscopy Software by Bitplane. 21. 37  C and 5% CO2 were maintained in the incubation system for Zeiss 710. 22. There is a significantly lower number of cells expressing mVenus-hGeminin in mouse enteroids, because mCherryhCDT1 is also expressed in terminally differentiated cell types. Therefore, it is more efficient to identify the cells expressing mVenus-hGeminin and measure cell cycle times. 23. We observe that some of the green tracks do not have corresponding red tracks. Potential reasons may be from the following: The red track was not captured by a confocal microscope because they were not included in the Z-stack range. In some cells, the migration distance between two consecutive time points could be larger, and this leads to the signal drift so that we lose those tracks of cells. 24. We suggest users to read the following information: Introduction to Time Lapse in Imaris (https://imaris.oxinst.com/ learning/view/article/introduction-to-time-lapse-in-imaris), Become a Master of Tracking in Imaris (https://imaris.oxinst. com/learning/view/article/mastering-tracking-in-imaris-3d4d-visualization).

Assessing Circadian Rhythms and Cell Cycle in Enteroids

123

Acknowledgements This work was supported by NIH grants R01 DK117005 (CIH), U19 AI116491 (CIH), R21 CA227379 (CIH), and National Research Foundation of Korea 2020R1A6A3A03038405 (MP). References 1. Nagoshi E, Saini C, Bauer C, Laroche T, Naef F, Schibler U (2004) Circadian gene expression in individual fibroblasts: cellautonomous and self-sustained oscillators pass time to daughter cells. Cell 119(5):693–705. https://doi.org/10.1016/j.cell.2004.11.015 2. Leise TL, Wang CW, Gitis PJ, Welsh DK (2012) Persistent cell-autonomous circadian oscillations in fibroblasts revealed by six-week single-cell imaging of PER2::LUC bioluminescence. PLoS One 7(3):e33334. https://doi. org/10.1371/journal.pone.0033334 3. Dunlap JC (1999) Molecular bases for circadian clocks. Cell 96(2):271–290. https://doi. org/10.1016/s0092-8674(00)80566-8 4. Cox KH, Takahashi JS (2019) Circadian clock genes and the transcriptional architecture of the clock mechanism. J Mol Endocrinol 63(4):R93–r102. https://doi.org/10.1530/ jme-19-0153 5. Hurley JM, Loros JJ, Dunlap JC (2016) Circadian oscillators: around the transcriptiontranslation feedback loop and on to output. Trends Biochem Sci 41(10):834–846. https://doi.org/10.1016/j.tibs.2016.07.009 6. Parasram K, Karpowicz P (2020) Time after time: circadian clock regulation of intestinal stem cells. Cell Mol Life Sci 77(7): 1267–1288. https://doi.org/10.1007/ s00018-019-03323-x 7. Dierickx P, Van Laake LW, Geijsen N (2018) Circadian clocks: from stem cells to tissue homeostasis and regeneration. EMBO Rep 19(1):18–28. https://doi.org/10.15252/ embr.201745130 8. Welz PS, Zinna VM, Symeonidi A, Koronowski KB, Kinouchi K, Smith JG, Guille´n IM, Castellanos A, Furrow S, Arago´n F, Crainiciuc G, Prats N, Caballero JM, Hidalgo A, Sassone-Corsi P, Benitah SA (2019) BMAL1-driven tissue clocks respond independently to light to maintain homeostasis. Cell 177(6):1436–1447.e1412. https:// doi.org/10.1016/j.cell.2019.05.009 9. Ruben MD, Wu G, Smith DF, Schmidt RE, Francey LJ, Lee YY, Anafi RC, Hogenesch JB (2018) A database of tissue-specific

rhythmically expressed human genes has potential applications in circadian medicine. Sci Transl Med 10(458). https://doi.org/10. 1126/scitranslmed.aat8806 10. Sato T, Vries RG, Snippert HJ, van de Wetering M, Barker N, Stange DE, van Es JH, Abo A, Kujala P, Peters PJ, Clevers H (2009) Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459(7244):262–265. https:// doi.org/10.1038/nature07935 11. Sato T, van Es JH, Snippert HJ, Stange DE, Vries RG, van den Born M, Barker N, Shroyer NF, van de Wetering M, Clevers H (2011) Paneth cells constitute the niche for Lgr5 stem cells in intestinal crypts. Nature 469(7330):415–418. https://doi.org/10. 1038/nature09637 12. Matsu-Ura T, Moore SR, Hong CI (2018) WNT takes two to tango: molecular links between the circadian clock and the cell cycle in adult stem cells. J Biol Rhythm 33(1):5–14. h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 0748730417745913 13. Matsu-Ura T, Dovzhenok A, Aihara E, Rood J, Le H, Ren Y, Rosselot AE, Zhang T, Lee C, Obrietan K, Montrose MH, Lim S, Moore SR, Hong CI (2016) Intercellular coupling of the cell cycle and circadian clock in adult stem cell culture. Mol Cell 64(5):900–912. https://doi. org/10.1016/j.molcel.2016.10.015 14. Moore SR, Pruszka J, Vallance J, Aihara E, Matsuura T, Montrose MH, Shroyer NF, Hong CI (2014) Robust circadian rhythms in organoid cultures from PERIOD2::LUCIFERASE mouse small intestine. Dis Model Mech 7(9):1123–1130. https://doi.org/10. 1242/dmm.014399 15. Rosselot AE, Park M, Matsu-Ura T, Wu G, Flores DE, Subramanian KR, Lee S, Sundaram N, Broda TR, McCauley HA et al (2022) Ontogeny and function of the circadian clock in intestinal organoids. EMBO J 41(2): e106973 16. Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, Siepka SM, Hong HK, Oh WJ, Yoo OJ, Menaker M,

124

Miri Park et al.

Takahashi JS (2004) PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101(15):5339–5346. https://doi.org/ 10.1073/pnas.0308709101 17. Abe T, Sakaue-Sawano A, Kiyonari H, Shioi G, Inoue K, Horiuchi T, Nakao K, Miyawaki A, Aizawa S, Fujimori T (2013) Visualization of cell cycle in mouse embryos with Fucci2 reporter directed by Rosa26 promoter. Development 140(1):237–246. https://doi.org/10. 1242/dev.084111 18. Bertaux-Skeirik N, Wunderlich M, Teal E, Chakrabarti J, Biesiada J, Mahe M, Sundaram N, Gabre J, Hawkins J, Jian G, Engevik AC, Yang L, Wang J, Goldenring JR, Qualls JE, Medvedovic M, Helmrath MA, Diwan T, Mulloy JC, Zavros Y (2017) CD44 variant isoform 9 emerges in response to injury and contributes to the regeneration of the gastric epithelium. J Pathol 242(4):463–475. https://doi.org/10.1002/path.4918 19. Miyoshi H, Stappenbeck TS (2013) In vitro expansion and genetic modification of gastrointestinal stem cells in spheroid culture. Nat Protoc 8(12):2471–2482. https://doi.org/ 10.1038/nprot.2013.153 20. Mahe MM, Aihara E, Schumacher MA, Zavros Y, Montrose MH, Helmrath MA, Sato T, Shroyer NF (2013) Establishment of gastrointestinal epithelial organoids. Curr Protoc Mouse Biol 3(4):217–240. https://doi. org/10.1002/9780470942390.mo130179 21. Tsai YH, Czerwinski M, Wu A, Dame MK, Attili D, Hill E, Colacino JA, Nowacki LM, Shroyer NF, Higgins PDR, Kao JY, Spence JR (2018) A method for cryogenic preservation of human biopsy specimens and subsequent organoid culture. Cell Mol Gastroenterol Hepatol 6(2):218–222.e217. https://doi.org/10. 1016/j.jcmgh.2018.04.008 22. Mahe MM, Sundaram N, Watson CL, Shroyer NF, Helmrath MA (2015) Establishment of human epithelial enteroids and colonoids from whole tissue and biopsy. J Vis Exp 97. https://doi.org/10.3791/52483 23. Hughes ME, Abruzzi KC, Allada R, Anafi R, Arpat AB, Asher G, Baldi P, de Bekker C, BellPedersen D, Blau J, Brown S, Ceriani MF, Chen Z, Chiu JC, Cox J, Crowell AM,

DeBruyne JP, Dijk DJ, DiTacchio L, Doyle FJ, Duffield GE, Dunlap JC, Eckel-Mahan K, Esser KA, FitzGerald GA, Forger DB, Francey LJ, Fu YH, Gachon F, Gatfield D, de Goede P, Golden SS, Green C, Harer J, Harmer S, Haspel J, Hastings MH, Herzel H, Herzog ED, Hoffmann C, Hong C, Hughey JJ, Hurley JM, de la Iglesia HO, Johnson C, Kay SA, Koike N, Kornacker K, Kramer A, Lamia K, Leise T, Lewis SA, Li J, Li X, Liu AC, Loros JJ, Martino TA, Menet JS, Merrow M, Millar AJ, Mockler T, Naef F, Nagoshi E, Nitabach MN, Olmedo M, Nusinow DA, Ptácˇek LJ, Rand D, Reddy AB, Robles MS, Roenneberg T, Rosbash M, Ruben MD, Rund SSC, Sancar A, Sassone-Corsi P, Sehgal A, Sherrill-Mix S, Skene DJ, Storch KF, Takahashi JS, Ueda HR, Wang H, Weitz C, Westermark PO, Wijnen H, Xu Y, Wu G, Yoo SH, Young M, Zhang EE, Zielinski T, Hogenesch JB (2017) Guidelines for genome-scale analysis of biological rhythms. J Biol Rhythm 32(5):380–393. h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 0748730417728663 24. Brown SA, Fleury-Olela F, Nagoshi E, Hauser C, Juge C, Meier CA, Chicheportiche R, Dayer JM, Albrecht U, Schibler U (2005) The period length of fibroblast circadian gene expression varies widely among human individuals. PLoS Biol 3(10): e338. https://doi.org/10.1371/journal.pbio. 0030338 25. Lidth V, de Jeude JF, Vermeulen JL, Montenegro-Miranda PS, Van den Brink GR, Heijmans J (2015) A protocol for lentiviral transduction and downstream analysis of intestinal organoids. J Vis Exp 98. https://doi.org/ 10.3791/52531 26. Sakaue-Sawano A, Kurokawa H, Morimura T, Hanyu A, Hama H, Osawa H, Kashiwagi S, Fukami K, Miyata T, Miyoshi H, Imamura T, Ogawa M, Masai H, Miyawaki A (2008) Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell 132(3):487–498. https://doi.org/10.1016/j.cell.2007.12.033 27. Izumo M, Sato TR, Straume M, Johnson CH (2006) Quantitative analyses of circadian gene expression in mammalian cell cultures. PLoS Comput Biol 2(10):e136. https://doi.org/ 10.1371/journal.pcbi.0020136

Chapter 8 Using ALLIGATORs to Capture Circadian Bioluminescence Aiwei Zeng and John S. O’Neill Abstract Luciferases are a popular tool in circadian biology research as longitudinal reporters of gene expression. Here, we describe a short updated protocol for the use of an Automated Longitudinal Luciferase Imaging Gas and Temperature-Optimized Recorder (ALLIGATOR) to record cellular bioluminescence over many days. The ALLIGATOR has superior capacity and flexibility compared with traditional luminometers that employ photomultiplier tubes (PMTs), with high-throughput capability and spatial resolution. It can be readily adapted to a wide variety of applications, such as different sample types and plate sizes, under a wide range of physiologically relevant conditions. Key words Circadian rhythm, Bioluminescence, Luciferase, Cellular clock biology, ALLIGATOR, PERIOD2

1

Introduction Luciferases are commonly used in circadian biology to report circadian gene expression in cells and tissues, across a wide range of organisms, from mammals to fungi [1, 2]. Luciferases can be used as a reporter of transcriptional activity, such as PER2:LUC [3], or as a fusion protein to report translational activity, such as PER2:: LUC [4]. Although the light emitted from luciferases is dim relative to fluorescently tagged reporters such as GFP, they have several advantages which make them suitable for the long-term nature of circadian experiments. Unlike fluorescently tagged reporters, there is no need for exogenous illumination of luciferases, meaning that samples do not experience phototoxicity or photobleaching. In addition, samples typically have low background, whereas autofluorescence and low signal:noise ratio are common problems for fluorescence measurement over several days. Of all the natural luciferases, Firefly luciferase (Fluc) is particularly well suited for real-time recording of circadian regulation of gene expression, and has been extensively characterized as a reporter of circadian rhythms. Although Fluc protein is a fairly

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_8, © The Author(s) 2022

125

126

Aiwei Zeng and John S. O’Neill

stable protein (t½ > 2 days), the half-life of Fluc activity in the presence of substrate is relatively short, and dependent on substrate concentration and enzymatic turnover [5]. Expressed in mammalian cells Fluc activity has a half-life around 2–5 h under saturating concentrations of luciferin (between 0.1 and 1 mM extracellularly, depending on cell/tissue type). When used in longitudinal recordings as a transcriptional or translational reporter, Fluc bioluminescence correlates linearly with transcriptional activity of the gene of interest or recent protein synthesis, respectively, and thus reliably reports changes in gene expression or protein synthesis [5, 6]. There are several methods for measuring bioluminescence in cultured tissues and cell populations, which typically involve the use of photomultiplier tubes (PMTs). While PMTs are highly sensitive to photons and can therefore resolve dim samples, their use in measuring bioluminescence has several disadvantages. First, bioluminescent recordings must be performed under non-humidified tissue culture conditions, with samples held in air-tight chambers, to protect PMTs from water damage. Consequently, and to prevent evaporation, biological samples must be maintained in media buffered by 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) or 3-(N-morpholino)propanesulfonic acid (MOPS), rather than more physiological bicarbonate buffering systems. In addition, PMTs lack the flexibility to measure spatial differences in bioluminescence across samples. Moreover, by virtue of their housing, PMTs are typically limited to measurement from specific dish or plate sizes, using either one PMT per sample, or one PMT for many samples but with corresponding loss of sensitivity and temporal resolution. A superior method for long-term detection of bioluminescence is presented by the Automated Longitudinal Luciferase Imaging Gas and Temperature-Optimized Recorder (ALLIGATOR) (Fig. 1). The ALLIGATOR comprises a standard humidified CO2/N2 170 L tissue culture incubator and a water-chilled electron-multiplying charge-coupled device (EM CCD) camera with anti-mist optics that prevent condensation. The ALLIGATOR affords programmable control of temperature, %CO2, and %O2 allowing long-term bioluminescence recordings under physiologically relevant conditions. In addition, the ALLIGATOR can image up to six 96- or 384-well plates simultaneously, and supports recording from a wide variety of cell types, tissues, and organotypic slices, as well as nonstandard formats, such as perfused tissue cultures [7, 8]. Here, we describe an updated protocol for using of an ALLIGATOR to record quantitative bioluminescence time series from adherent cells. In this case, we use the ALLIGATOR to test the effect that the addition of cell culture media growth supplements, such as fetal bovine serum (FBS) and B27, has on cellular circadian rhythms in human U2OS cells expressing Fluc under the control of the PERIOD2 gene promoter (PER2:LUC). In Figs. 2 and 3, we

Using ALLIGATORs to Capture Circadian Bioluminescence

127

Fig. 1 Illustration of an ALLIGATOR. The ALLIGATOR consists of a standard tissue culture incubator with a top-mounted EM-CCD camera, and can image up to 6
96-well plates under physiological conditions. The shelf height can be adjusted according to the sample, while the black template surround size should be chosen according to the number of plates used. When recording, the cloth surround should be secured in place to prevent any environmental light from entering the ALLIGATOR incubator

demonstrate that compared with serum-free media, the addition of 10% serum or 1
B27 significantly decreases the period and increases the amplitude of circadian bioluminescence oscillations, as well as inducing phase shifts. These observations suggest that, when perturbing cells with pharmacological, genetic, or environmental manipulations, careful control and consideration of cell culture medium composition is required before any changes in circadian phase, period, or amplitude can be understood since such changes may result from an interaction between the perturbation and the response to serum factors in the media, rather than being solely attributable to the perturbation [9].

2

Materials

2.1 Cell Culture and Reagents

1. U2OS cells stably transfected with pGL4.20 constructs (Promega) to express Fluc under the control of the PER2 promoter (as described in [10]). 2. Culture medium: Dulbecco’s Modified Eagle Medium (DMEM) Glutamax, with high glucose, L-glutamine and phenol red (Gibco), supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin.

128

Aiwei Zeng and John S. O’Neill

A. Draw ROIs around each well, measure integrated density

Raw grouped data 250000

0% serum

Bioluminescence

B. Subtract background data

10% serum

200000

150000

100000 0

24

48

72

96

120

144

168

Time (h)

Detrended data with curve fit

C. Detrend each replicate and fit to a damped cosine wave

Bioluminescence

15000

10% serum 5000

0% serum

-5000

-15000 0

24

48

72

96

120

144

168

Time (h)

Calculated parameters 30000

25.5

Amplitude

D. Calculate parameters e.g. period, amplitude, phase, damping rate

Period (h)

25.0 24.5

20000

10000

24.0

0

23.5 0%

0%

10%

Serum

*

ns

0.027

2

0.026

1

0.025

Phase (h)

Damping rate

10%

Serum

0.024 0.023

0 -1 -2

0.022 0.021 0%

10%

Serum

-3 0%

10%

Serum

Fig. 2 Workflow for the analysis of bioluminescence of U2OS PER2:LUC cells supplemented with no serum or 10% serum. (a) Each image taken is an integration over 30 min of exposure time. After de-noising and removing cosmic ray interference, regions of interest (ROIs) should be drawn around each well, and integrated density measured across the time series. (b) Individual replicate traces plotted after removing background

Using ALLIGATORs to Capture Circadian Bioluminescence

129

3. Fetal bovine serum (GE Healthcare). 4. (Optional) 50
B27 supplement (Gibco) (see Note 1). 5. Phosphate-buffered saline (PBS). 6. 0.25% Trypsin, 0.04% EDTA solution, diluted 1:4 in PBS. 7. 100 mM D-Luciferin potassium salt (Biosynth), dissolved in Milli-Q water (see Note 2). 2.2

Other Materials

1. 96-Well plates, black, transparent-bottom tissue culture treated (Costar) (see Note 3). 2. Gas-permeable plate seals (4titude) (see Note 4). 3. White vinyl plate stickers (VWR). 4. 8-Channel multichannel pipette (Eppendorf). 5. 37  C Isothermal pad (see Note 5).

3

Methods

3.1 Seeding and Entrainment

1. Lift cells from cell culture flasks as per standard methods, e.g., wash cells with PBS and incubate with trypsin/EDTA at 37  C for 5 min. 2. Seed cells in culture medium supplemented with 10% fetal bovine serum (FBS) and 1 mM luciferin at a relatively high density. In Fig. 2, U2OS cells stably transfected with PER2: LUC were seeded into black 96-well plates at a density of approximately 2
104 cells/cm2. The presence of luciferin prior to bioluminescence imaging from transcriptional reporters is required to inactivate luciferase that is expressed prior to recording. 3. Grow cells to confluency in a temperature cycling incubator (12 h at 32  C, 12 h at 37  C, controlled by a computer connected to the incubator through its serial RS-232 port or similar), until the monolayer is 100% confluent (approximately 4–7 days) (see Notes 6 and 7).

3.2

Recording

1. Before recording begins, warm up recording medium (culture medium containing any compounds to be tested) containing 1 mM luciferin in a 37  C water bath (see Note 8).


Fig. 2 (continued) bioluminescence and excluding wells containing dead cells. (c) Detrending with a 24-h moving average and fitting the curve to a damped cosine wave, excluding the first 24 h of recording. (d) Additional parameters calculated, such as period, amplitude, damping rate, and phase (unpaired two-tailed ttest, mean  SEM, n ¼ 4 throughout)

Raw data

A

Detrended data with curve fit

240000

20000

Bioluminescence

Bioluminescence

0% serum

200000

160000

120000

10000 0 -10000 -20000

240000

20000

Bioluminescence

Bioluminescence

5% serum

200000

160000

120000

10000 0 -10000 -20000

240000

20000

Bioluminescence

Bioluminescence

10% serum

200000

160000

120000

10000 0 -10000 -20000

240000

20000

Bioluminescence

Bioluminescence

0.5X B27

200000

160000

120000

10000 0 -10000 -20000

240000

20000

Bioluminescence

Bioluminescence

1X B27

200000

160000

120000

10000 0 -10000 -20000

48

72

B

96

120

144

168

0

24

48

Time (h)

40000

10000

144

168

4

Phase (h)

2

0

-2

5% m se 10 rum % se ru m 0. 5X B 27 1X B 27

ru

se ru m 0. 5X B 2 1X 7 B 27

m

m

ru se

ru

%

10

27 B

1X

m

27 B

ru se

5X

%

0.

ru

ru se

5%

se

m

m

0.020

10

27 B

1X

m

B 27

ru se

5X 0.

m

m

ru

ru

%

10

se

se

0.024 0.022

0

5%

0.026

se

23.0

120

0.028

5%

23.5

0%

Damping rate

Amplitude

24.0

20000

0%

Period (h)

30000

24.5

96

ns

0.030

25.0

0%

25.5

72

Time (h)

se

24

0%

0

Fig. 3 The effect of differing medium supplements on U2OS PER2:LUC rhythms. (a) Left: raw background subtracted data of U2OS PER2:LUC cells recorded in 0%, 5%, or 10% serum, or 0.5
or 1
B27. Right: data detrended with a 24-h moving average subtraction, overlaid with a fitted damped cosine wave (solid lines). (b) Analysis of period, amplitude, damping rate, and phase (ordinary one-way ANOVA with Dunnett’s multiple comparisons test, mean  SEM, n ¼ 4 throughout)

Using ALLIGATORs to Capture Circadian Bioluminescence

131

2. Set up the ALLIGATOR using Optomorph software. For detailed instructions of Andor Solis, an alternative software package, see Crosby et al. [8]. (a) Turn on the camera, incubator, water bath, demister, and computer. (b) Allow camera to cool to its lowest temperature before using (usually 70 to 100  C for modern cameras). (c) Focus the camera: l

In Optomorph, select “Apps” ¼> “Multi Dimensional Acquisition” (MDA) ¼> “Wavelength.” Set “Digitizer” to “2.75 MHz (Standard)” and “Exposure” to “50 ms.” Click “Live.”

l

Unplug the demister ring and gently unscrew (take care not to touch the glass).

l

Use the focus knob to set the correct focal plane.

l

Screw the demister back on and plug in.

l

Place an object at the distance of the sample. Double check that it is in focus.

l

Click “Live” to turn off live mode.

(d) Fill incubator water tray with deionized water. (e) Set O2, CO2, and temperature to the desired levels on the control panel on the front of the incubator, or connected computer, as desired. 3. Remove cells from tissue culture incubator onto an isothermal pad at 37  C, quickly aspirate cell media and replace with pre-warmed recording medium (e.g. see Notes 8 and 9). Seal plate with a gas-permeable plate seal. Attach white vinyl plate stickers to the plate bottom to increase light reflected to the camera. 4. Transfer plates/dishes from isothermal pad into the ALLIGATOR, taking care to minimize any change in temperature experienced by the cells/tissues. 5. Choose recording parameters for the experiment. (a) In “Apps” ¼> “MDA” ¼> “Wavelength.” Standard conditions used for adherent cells: “Digitizer” ¼ 690 KHz (EM Gain), “Gain” ¼ 1
(default), “Illumination” ¼ “None.” Exposure time and EM Gain should be determined empirically according to cell type; exposure time is typically between 30 and 60 min but pixel saturation must be avoided (see Note 10). (b) In “MDA” ¼> “Timelapse,” set “Time Interval” accordingly, e.g., for an exposure time of 15 min every 30 min, set “Time Interval” to 30 min. Set “Duration” to the total length of the experiment.

132

Aiwei Zeng and John S. O’Neill

(c) In “MDA” ¼> “Saving,” choose a base name and tick “Increment base name,” and select an appropriate file saving location. 6. To begin recording, click “Acquire” in the MDA window. 7. If any manipulations are required during recording, press “Pause” and wait until the last exposure is taken before opening the incubator door. 3.3

Analysis

1. Open the image stack using the “.nd” file in Fiji/ImageJ. Adjust brightness/contrast. 2. Remove noise using “Process” ¼> “Noise” ¼> “Despeckle.” Remove cosmic ray interference using the “Cosmic Noise Reducer” plugin. 3. Take an average intensity z-projection and create region of interest (ROI) templates by drawing an “Ellipse” around each well, with an additional ROI for the background (Fig. 2). Save the ROIs to the ROI Manager (“Analyze” ¼> “Tools” ¼> “ROI Manager”). Apply the ROI template to the time series. 4. Measure integrated density per well and background across the time series using the “Multi Measure” function (“Analyse” ¼> “Set Measurements” ¼> “Integrated density”; “ROI Manager” ¼> “More” ¼> “Multi Measure”). 5. Export raw data to Excel. Subtract background bioluminescence, and detrend by subtracting a 24 h moving average. Exclude wells where cells have died (e.g., due to evaporation or contamination). 6. The first 24 h of acquisition should be excluded from analysis, as cellular bioluminescence may exhibit non-circadian changes during this time, due to the acute response following a media change. 7. Further analysis can be performed in other data analysis software programs, such as GraphPad Prism. For example, nonlinear regression and fitting to a damped cosine wave can be performed using the following equation: y ¼ amplitude  eðkX Þ  cos ð2  π  ðX  phaseÞ=periodÞ where y ¼ detrended and smoothed bioluminescence (signal in relative luminescent units; RLU), x ¼ time (h), k ¼ rate constant of signal decay, amplitude ¼ the highest value of y during the first cycle, period ¼ the time taken for a complete cycle to occur in units of x, phase is the value of x at which the highest value of y occurs during the first cycle.

Using ALLIGATORs to Capture Circadian Bioluminescence

4

133

Notes 1. Depending on the type of experiment, serum and B27 may be considered optional supplements, and the concentrations (and serum type) should be optimized for each cell type and experimental objective. In general, serum and B27 promote cell survival, proliferation, and attachment, and thus may be required for viability of some cell types. The presence of serum and/or B27 can affect the period and amplitude of bioluminescence oscillations however (Fig. 3). Also, the phase of circadian rhythms in most cell types is sensitive to acute stimulation by growth factors in serum, as well as insulin and corticosterone present in B27 [7]. For recording media that contain high levels of serum or B27 therefore, media change may synchronize cells to different circadian phases compared with serum/B27-free media, and also affect the subsequent period and amplitude of bioluminescence rhythms. In this recording, for example, cells receiving media containing 10% serum or 1
B27 showed significantly shorter circadian periods compared with controls. The amplitude and phase of bioluminescence rhythms was significantly affected for all concentrations of B27 or serum. Whereas the damping rate, reflecting a combination of progressively reducing amplitude of intracellular Fluc expression and intercellular desynchronization, was not affected by either. 2. Intracellular luciferin concentration ([luciferin]) is in equilibrium with, but not identical to, extracellular [luciferin] and may be as much as tenfold lower [5]. For Fluc to function as a fidelitous reporter of circadian gene expression, it must be present in saturating excess, which requires optimization for each cell and reporter type. If [luciferin] is not at saturating, any perturbation that affects luciferin transport into the cell could (in principle) lead to apparent changes in phase, period, or amplitude that are unrelated to circadian timing function [5]. High concentrations of luciferin (>1 mM) can also elicit nonspecific effects on oscillator function however [5]. In practice we find luciferin concentrations in the 0.1–1 mM range are appropriate for most applications [5]. In addition, luciferin should be added to cell culture medium during entrainment, to minimize “flash effects.” 3. We recommend the usage of black, transparent-bottom plates to allow the monitoring of cell growth and viability prior to recording, while minimizing the interference between wells during the recording. For particularly dim cells, white plates should be used; however the first few frames will be unusable

134

Aiwei Zeng and John S. O’Neill

due to autofluorescence. White stickers can be placed on the bottom of transparent-bottom 96-well plates during recording to maximize reflected light. 4. We strongly recommend using gas-permeable plate seals instead of using a lid to prevent “edge effects,” during both entrainment and recording stages. 5. Cells should be placed on an isothermal pad set to 37  C during any medium changes or perturbations, and while carrying cells between incubators to prevent any phase shifts from occurring due to temperature fluctuations [11]. 6. The length of entrainment time by temperature cycles should be optimized for each cell type. Alternatively, other methods of entrainment are commonly used, such as the addition of dexamethasone or forskolin [12, 13]. 7. Provided that the cells being used are subject to contact inhibition of growth once confluent, the monolayer of cells can be maintained for several weeks with regular media changes (approximately every 5 days). For this reason, our preferred cell types are human U2OS cells and adult fibroblasts lines, where we have directly confirmed contact inhibition of proliferation [14, 15]. 8. The recording medium used should be optimized for each cell or tissue type. In the ALLIGATOR, it is possible to record in standard cell culture growth medium. However, should dishes need to be sealed, then “Air medium” containing MOPS of HEPES buffer can be used to buffer media pH against acidification (see Hastings et al. [16]). 9. Depending on cell type, to minimize phase shifts and maximize cell synchronization, media should be changed at the appropriate point in the circadian. For instance, in U2OS cells we would generally change media at the transition from 32 to 37  C, while in mouse fibroblasts we would change media at the transition from 37 to 32  C. 10. Exposure time should be determined for each cell type used. To begin, set a short exposure time of 10 min without EM Gain. Check the pixel intensity by placing the cursor over the brightest spot; an ideal pixel intensity is around 10,000, while the maximum pixel intensity is around 65,000. Adjust the exposure time and EM Gain accordingly. Where possible we try to avoid using EM Gain or binning, though this may not be possible for samples with very low bioluminescence.

Using ALLIGATORs to Capture Circadian Bioluminescence

135

References 1. Balsalobre A, Damiola F, Schibler U (1998) A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93: 929–937. https://doi.org/10.1016/S00928674(00)81199-X 2. Morgan LW, Greene AV, Bell-Pedersen D (2003) Circadian and light-induced expression of luciferase in Neurospora crassa. Fungal Genet Biol 38:327–332. https://doi.org/10. 1016/S1087-1845(02)00562-5 3. Ueda HR, Chen W, Adachi A et al (2002) A transcription factor response element for gene expression during circadian night. Nature 418: 5 3 4 – 5 3 9 . h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / nature00906 4. Yoo S-H, Yamazaki S, Lowrey PL et al (2004) PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101:5339–5346. h t t p s : // d o i . o r g / 1 0 . 1 0 7 3 / P N A S . 0308709101 5. Feeney KA, Putker M, Brancaccio M, O’Neill JS (2016) In-depth characterization of firefly luciferase as a reporter of circadian gene expression in mammalian cells. J Biol Rhythm 31: 5 4 0 – 5 5 0 . h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 0748730416668898 6. Millar AJ, Carre IA, Strayer CA et al (1995) Circadian clock mutants in Arabidopsis identified by luciferase imaging. Science 267: 1161–1164 7. Crosby P, Hammett R, Putker M et al (2019) Insulin/IGF-1 drives PERIOD synthesis to entrain circadian rhythms with feeding time. Cell 177:896–909.e20. https://doi.org/10. 1016/j.cell.2019.02.017 8. Crosby P, Hoyle NP, O’Neill JS (2017) Flexible measurement of bioluminescent reporters using an automated longitudinal luciferase imaging gas- and temperature-optimized recorder (ALLIGATOR). J Vis Exp 2017: 56623. https://doi.org/10.3791/56623

9. Zhang EE, Liu AC, Hirota T et al (2009) A genome-wide RNAi screen for modifiers of the circadian clock in human cells. Cell 139: 199–210. https://doi.org/10.1016/j.cell. 2009.08.031 10. Valekunja UK, Edgar RS, Oklejewicz M et al (2013) Histone methyltransferase MLL3 contributes to genome-scale circadian transcription. Proc Natl Acad Sci U S A 110: 1554–1559. https://doi.org/10.1073/pnas. 1214168110 11. Buhr ED, Yoo SH, Takahashi JS (2010) Temperature as a universal resetting cue for mammalian circadian oscillators. Science 330: 379–385. https://doi.org/10.1126/science. 1195262 12. Balsalobre A, Brown SA, Marcacci L et al (2000) Resetting of circadian time in peripheral tissues by glucocorticoid signaling. Science 289:2344–2347. https://doi.org/10.1126/ science.289.5488.2344 13. Balsalobre A, Marcacci L, Schibler U (2000) Multiple signaling pathways elicit circadian gene expression in cultured Rat-1 fibroblasts. Curr Biol 10:1291–1294. https://doi.org/10. 1016/S0960-9822(00)00758-2 14. Hoyle NP, Seinkmane E, Putker M et al (2017) Circadian actin dynamics drive rhythmic fibroblast mobilization during wound healing. Sci Transl Med 9:eaal2774. https://doi.org/10. 1126/scitranslmed.aal2774 15. Feeney KA, Hansen LL, Putker M et al (2016) Daily magnesium fluxes regulate cellular timekeeping and energy balance. Nature 532: 3 7 5 – 3 7 9 . h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / nature17407 16. Hastings MH, Reddy AB, McMahon DG, Maywood ES (2005) Analysis of circadian mechanisms in the suprachiasmatic nucleus by transgenesis and biolistic transfection. Methods Enzymol 393:579–592

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), 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 chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Chapter 9 Studying Circadian Clock Entrainment by Hormonal Signals Violetta Pilorz, Iwona Olejniczak, and Henrik Oster Abstract In mammals, molecular circadian clocks not only exist in the suprachiasmatic nucleus (SCN) but in almost all organ systems. Intriguingly, tissue clocks can operate in both isolated tissues and cell lines with endocrine signals mediating the circadian expression of local transcriptomes. This can be demonstrated by treating tissue explants with endocrine cues in a phase- and dose-dependent manner. In this chapter we provide an overview of methods to study the effects of candidate hormonal time cues on tissue clock resetting. We propose an experimental procedure based on an in vitro setup consisting of several consecutive steps in which organotypic tissue cultures or cells can be used. Our approach targets the potential resetting mechanism at three levels: the hormone, the direct clock gene target, and the tissue clock response. Key words In vitro, Peripheral circadian clocks, Hormones, Phase resetting, Screening

1

Introduction The endogenous circadian clock consists of a network that implicates the central pacemaker of the suprachiasmatic nucleus (SCN), extra-SCN central nervous system (CNS) clocks, and peripheral oscillators [1]. It orchestrates 24-h rhythms in numerous molecular, physiological, and behavioral functions by temporal segregation of conflicting events such as feeding and sleeping [2, 3], hunger and satiety [4, 5], or glycolysis and gluconeogenesis [6–8]. In addition, this system facilitates the anticipation of external environmental conditions such as light or food availability [9–11]. Mechanistically, circadian rhythmicity is generated by interlocked transcriptionaltranslational feedback loops consisting of molecular clock genes and proteins [12]. Intriguingly, circadian clock genes were shown to be expressed not only in the cells of the pacemaker, but also in numerous other cells in the body. Similar to the SCN, peripheral cells show autonomous clock rhythms [13–17]. Lately, the traditional viewpoint about the SCN being the major pacemaker that synchronizes all peripheral rhythms to the external time has been somewhat changed [18–20]. Metabolic tissue clocks, for example,

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_9, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

137

138

Violetta Pilorz et al.

are reset by timed meal intake independent of the SCN. In this way, the multi-oscillatory circadian network responds to different environmental factors such as light/dark or feeding time called Zeitgebers (from the German for time giver) to align with a complex—but predictably rhythmic—environment [21, 22]. The regulatory pathways of this network synchronization involve several endocrine factors, i.e., hormones that act as internal time cues and, thus, as modulators of phase for specific tissue clocks [14, 23]. Deciphering the mediators of clock network synchronization will yield important points of attack for manipulating clock network function and outputs, e.g., in therapeutic settings. One approach to identify humoral circadian time cues is by screening cells, tissues, or live animals for endocrine factors capable of modulating clock gene expression rhythms in tissues such as liver or gut. Our lab has in recent studies applied such screening approaches to discover peptide hormones capable of resetting clocks in liver and hypothalamus—oxyntomodulin (OXM) [24] and adiponectin (ADIPOQ) [25]. OXM is secreted after food consumption in the gut and successively induces Period gene expression in the liver while ADIPOQ is a reporter of peripheral metabolic state resetting clocks in appetite-promoting neurons of the hypothalamus through activation of Bmal1. Such studies can use two principle methodological approaches, in vivo or in vitro. While using live organisms for screening will provide important insight into tissue-specific aspects and the interaction of different target clocks at the systemic level, in vitro approaches are better scalable and potentially allow for mediumto high-throughput screening of many candidate factors—albeit with limited target tissue coverage. Such in vitro approaches will be the focus of this chapter. The method suggested in our present chapter can be applied to all types of humoral factors including steroid and peptide hormones. We include assays covering the identification of resetting factors, the characterization of their resetting capacities, and the mechanisms by which they affect clock gene regulation at the cellular level (Fig. 1).

2

Materials

2.1 Tissue Slice Circadian Reporter Cultures

Filtrate all solutions with bottle-top filters with pore size 0.45 μm. All reagents are prepared at room temperature under a cell culture hood. Dissection buffer: 1
Hanks’ balanced salt solution (HBSS), 3 mM sodium bicarbonate solution, 10 mM HEPES buffer (pH 7.2), 100 U/ml penicillin, 100 μg/ml streptomycin. Adjust the volume to 1 l with autoclaved MilliQ water (ddH2O). Store at 4  C.

Studying Circadian Clock Entrainment by Hormonal Signals

139

Fig. 1 Schematic representation of the experimental setup. The proposed experiments investigate different aspects of hormonal clock resetting: (1) tissue slice circadian reporter cultures allow the assessment of entrainment of the tissue by the hormone of interest; (2) luciferase assays probe core clock gene expression changes after hormone stimulation; (3) ChIP-qPCR experiments provide insight into a possible promoter/ transcription factor interaction

Recording medium: DMEM low-glucose, 10 mM D-glucose, 3 mM sodium bicarbonate solution, 10 mM Hepes (pH 7.2), 100 U/ml penicillin, 100 μg/ml streptomycin, 2% B-27 supplement, 0.1 mM luciferin. Adjust the volume to 1 l with autoclaved MilliQ water (ddH2O). Store at 4  C. Stimulation reagents: Dexamethasone (Sigma Aldrich, St. Louis, MO). Fetal bovine serum (Thermo Fisher Scientific, Waltham, MA). Hardware: Nalgene bottle-top sterile filters 1000 ml with pore size 0.45 μm (Sigma Aldrich, St. Louis, MO). LumiCycle luminometer (Actrimetrics, Evanston, IL). Sterile high-vacuum grease. 40-mm Microscope glass cover slips. Scalpel.

140

Violetta Pilorz et al.

5-ml Syringes. Vibratome. 35-mm Petri dishes. MilliCell culture inserts with 30 mm diameter and pore size 0.4 μm (Merck Millipore, Darmstadt, Germany). Hot plate. 2.2 Luciferase Clock Gene Promoter Assays

LB medium: 5 g tryptone, 5 g NaCl, 2.5 g yeast extract. Add 500 ml ddH2O and autoclave at 121  C. LB agar: 5 g tryptone, 2.5 g yeast extract, 5 g NaCl, 7.5 g agar. Add 500 ml ddH2O and autoclave at 121  C. Let cool down to 55  C, then add relevant antibiotics. Pour 20 ml into a Petri dish of 10 cm diameter. Allow the plates to set and seal the plates with Parafilm. Store at 4  C. Cell culture medium: Dulbecco’s modified Eagle medium (DMEM) with phenol red, 5% fetal bovine serum (FBS), 100 U/ml penicillin, 100 μg/ml streptomycin, 3 mM glutamine. Store at 4  C. Cell culture medium for luciferase assays: DMEM without phenol red, 5% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, 3 mM glutamine. Store at 4  C. Plate coating: poly-D-lysine, 1 mg/ml, diluted 1:20 in 1
PBS. Transfection: Opti-MEM 1
reduced serum medium (Gibco Thermo Fisher). Plasmid DNA (100 ng). P3000 reagent (Thermo Fisher Scientific, Waltham, MA). Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA). Luciferase assay kit: Secrete Pair Gaussia Luciferase Assay Kit (GeneCopoeia, Rockville, MD) for analysis of the activities of Gaussia Luciferase (GLuc) and Secreted Alkaline Phosphatase (SEAP) in a dual-reporter system. Both GLuc and SEAP reporter proteins do not require cell lysis. Hardware: Petri dishes of 10 cm diameter. Parafilm (Bemis, Neenah, WI). Light-tight white 96-well plate. Luminometer plate reader. Neubauer counting chamber. One Shot TOP10 cells (Thermo Fisher Scientific, Waltham, MA). S.O.C. recovery media suspension (Thermo Fisher Scientific, Waltham, MA). DNA purification: NucleoBond Xtra Midi Kit (MachereyNagel, Du¨ren, Germany).

Studying Circadian Clock Entrainment by Hormonal Signals

2.3 Circadian Enhancer Motif Chromatin Immunoprecipitation (ChIP)

141

Glycine. A/G agarose beads. Proteinase K (min. >20 U/mg). Ethanol (100% and 70%). DNase-free water. Primary antibody (e.g., anti-CREB antibody—Invitrogen, Carlsbad, CA, Cat #MA1-083). Chelex 100 resin (Bio-Rad, Hercules, CA). Protease inhibitor cocktail (PI), EDTA-free. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 M KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, ddH2O. Crosslinking buffer: 1
PBS/PI (PBS with protease inhibitors) with 1% formaldehyde. Dilution buffer: 0.01% sodium dodecyl sulfate (SDS), 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris–HCl pH 8.1, 167 mM NaCl, PI. Cell lysis buffer: 1% SDS, 10 mM EDTA, 50 mM Tris–HCl (pH 8.1), freshly added PI (see Note 1). qPCR: qPCR reaction mix. Primers (see Note 2). Hardware: Homogenizer. Sonicator. Tabletop centrifuge with cooling, with speed up to 24,104
g. 1.5-ml Reaction tubes. Vortexer. qPCR thermocycler.

3

Methods

3.1 Tissue Slice Circadian Reporter Cultures

3.1.1 Tissue Dissection

A good model to investigate hormonal effects on the synchronization of the clock rhythm in a specific tissue are organotypic slice cultures of mouse tissues from a strain expressing a reporter gene such as luciferase or a GFP derivative under the control of the circadian clock, i.e., under control of a clock gene promoter (see Note 3). All procedures need to be carried out at room temperature unless otherwise specified. 1. A quick cervical dislocation without anesthesia followed by decapitation, or anesthetization with CO2 followed by decapitation should be performed to keep the stress at the lowest possible level (see Note 4).

142

Violetta Pilorz et al.

2. After decapitation, the eyes should be removed before tissue dissection. Importantly, tissues should be quickly isolated and kept for a short time in chilled HBSS (see Note 5). 3. One piece of tissue (1–4 mm2) free of adhered visceral tissue should be cultured on MilliCell culture inserts positioned in a 35-mm Petri dish containing a standard recording medium [16]. Importantly, luciferin at a final concentration of 0.1 nM as well as B-27 supplement (2%) should be added into recording medium prewarmed at 37  C shortly before transferring it into the culture dish containing the tissue insert (see Note 6). 4. After placing dissected tissue on the membrane, any remaining HBSS should be carefully removed around the tissue with a pipette without punching the membrane and no air bubbles should be under the membrane, as these prevent the tissue from being supplied with nutrients. 5. Cover the rim of the plate with sterile high-vacuum grease using a 5-ml syringe to cover the culture dish with a 40-mm microscope glass covers. 6. Before tissue treatment with hormones, the tissue should be recorded using climatized luminometer instruments such as the 32-channel carousel LumiCycle. A temperature of 32  C is a suitable temperature to keep many cells and tissues alive for extended times [16, 26] (see Note 7). 7. When using cell cultures cells need to be synchronized for 30 min with either 50% serum shock or with 100 nM dexamethasone to achieve a transient coherence of cellular circadian phases. Tissue does not require any synchronization. 3.1.2 Assessment of Phase Treatment

The demonstration of the time-point-dependent effect of a candidate hormone treatment (i.e., a phase response curve) shows if the hormone acts as a true Zeitgeber. As one example, when using oxyntomodulin to reset PER2::LUC phase in liver slice cultures, treatments in the first quarter (0–90 ) of the PER2::LUC luminescence rhythm will result in phase delays, later treatments at 90–180 will cause phase advances whereas modest or no effect will be seen after treatment at 270–360 [24]. In addition, increasing doses of hormone treatment will reveal a dose response of tissue clock resetting (Fig. 2). To conduct stimulation with a hormone at a specific circadian time, the phase of the next cycle can be estimated by assessing the period using three consecutive cycles with the LumiCycle Analysis software of Actimetrics. Since cell cultures often show very rapid dampening of rhythms, a compromise may involve treatment at specific times after the synchronization event. After estimating the phase of treatment from extrapolations of ongoing recordings, the dish containing the tissue is removed from the LumiCycle and

Studying Circadian Clock Entrainment by Hormonal Signals

143

Fig. 2 Dose- and phase-dependent resetting of clock genes in the liver by oxyntomodulin (OXM) treatment. (a, b) An example of luminescence traces and dose- pendent responses OXM. (c) Dose-dependent phase shift; white bars represent PBS treatment (control) and dark bars dose-pendent OXM treatment. (d) Phase response curve for OXM-induced phase resetting of PER2::LUC rhythms in liver slices. Circles depict raw data of individual slices; dashed line illustrates sine wave regression with harmonics; arrows indicate the treatment time point. (Figure adapted from ref. 24 with permission from eLife Sciences)

placed on a heating plate, which has the same temperature as the incubator, to prevent an influence by changes in ambient temperature on the phase of the tissue clock. The application of the hormone into the recording medium should be performed under a hood to avoid contaminations. After re-sealing the tissue or cell dish, the sample is returned to the initial position of the LumiCycle and recording can be continued. 3.1.3 Clock Phase Assessment

Clock phase of the tissue should be assessed as the time of the first peak in luminescence after treatment [27]. In many cases it is easier to use the zero transition of normalized data for phase estimation, though, because of the steep change in signal intensity at this phase. A comparison of phases in drug-treated samples with those of solvent-treated controls will show the effect of the hormone on the circadian clock phase (see Note 8).

3.2 Luciferase Clock Gene Promoter Assays

Once the phase resetting effect of a candidate hormone has been established, the underlying signaling mechanism may be studied. Since circadian timekeeping in mammals is based on transcriptional

144

Violetta Pilorz et al.

rhythms, a good approach would be to analyze hormone-mediated regulation of core clock gene expression. Luciferase promoter assays are an established approach to estimate if a stimulus can activate or suppress the expression of a given target gene [28] (see Note 9). When working with bacteria, perform the procedures next to a flame using a Bunsen burner to kill airborne microorganisms. When using defrosted reagents and samples, keep them on ice during the preparations. 3.2.1 Plasmid Transformation

Before cell transfection with the DNA constructs, the plasmids need to be amplified. To achieve an efficient transformation, One Shot TOP10 chemically component Escherichia coli can be used. For this, 50 ng of the plasmid (DNA) is added to 50 μl One Shot Top 10 cells, incubated for 30 min on ice, heat-shocked at 42  C for 30 s, and placed on ice before transferring to 250 μl S.O.C. recovery media followed by incubation at 37  C for 1 h while shaking at 300 rpm.

3.2.2 Plasmid Amplification

1. Spread 20 μl of the transformation mix onto a prewarmed plate (room temperature) containing an appropriate antibiotic (see Note 10) using a glass loop. Before spreading the transformation sterilize the glass loop by passing it through a flame. Allow it to cool and glide the glass loop gently over the surface of the agar plate that contains 20 μl transformation mix. 2. Pass the glass loop again through the flame to sterilize it. An additional plate containing cells without the specific antibiotic is used as control. Both plates are incubated inverted overnight at 37  C in an incubator. 3. On the following day, colonies are selected with 10-μl pipette tips. These pipette tips are then transferred into 15-ml Falcon tubes for further incubation for 6–8 h at 37  C while shaking at 235 rpm. 4. On the same day, a slurry sample is chosen and transferred to 300 ml LB medium containing antibiotics and incubated overnight at 37  C while shaking at 230 rpm. 5. On the following day, the DNA purification (see Note 11) is performed to obtain a DNA pellet that can be dissolved with 500 μl 5 mM Tris–HCl (pH 8.5).

3.2.3 Cell Transfection

1. On the day of cell transfection, the cell culture must have >90% viability at ca. 80% confluency, and the cells must be attached to the wells of the plate. 2. Use poly-D-lysine coating to achieve affective cell attachment. For plate coating 1 mg/ml stock of poly-D-lysine needs to be diluted 1:20 with sterile 1
phosphate-buffered saline (PBS).

Studying Circadian Clock Entrainment by Hormonal Signals

145

For 96-well plates use 100 μl diluted poly-D-lysine for each well and incubate it for 40 min. Then wash the 96-well plate three times, either with sterile 1
PBS or with autoclaved ddH2O, and let dry at room temperature. 3. Plate 4.5
104 HEK293T cells per well in 200 μl medium and incubate the 96-well plate for 24 h at 37  C (see Note 12). 4. After 24 h incubation, cells are transfected using a mixture of Lipofectamine 3000, plasmid DNA, Opti-MEM, and P3000 reagent. For this purpose, two mixtures must be prepared beforehand. One consists of 0.2 μl P3000 reagent, 5 μl OptiMEM 1
, and 100 ng plasmid DNA per well. The second mix contains 5 μl Opti-MEM 1
and 0.2 μl Lipofectamine 3000 per well. 5. Both mixtures need to be thoroughly mixed and incubated for 10 min at room temperature before pipetting 10 μl of it into each well containing cells. 6. After 24 h incubation at 37  C replace the medium by DMEM without phenol red (see Note 13). 7. After additional 24 h of incubation collect the medium in 1.5ml reaction tubes and store at 20  C for ca. 1 month if not used immediately for the luciferase assay. 3.2.4 Luciferase Assay

3.2.5 Gaussia Luciferase Assay

To this end, cells are lysed and transferred into a reaction tube. Using SEAP or GLuc secreted reporters has the advantage of allowing for their detection in the cell-free conditioned medium avoiding the need for cell lysis (see Note 14). 1. Defrost 10
GL-S (see Note 15) at room temperature. 2. Vortex it for 3–5 s. 3. Dilute it 1:10 with ddH2O to obtain 1
GL-S. 4. Prepare Gluc Assay working solution by adding substrate GL (see Note 15) to 1 ml 1
GL-S buffer, mix it thoroughly by inverting the tube several times. This volume is for 10 wells. 5. Cover the tube light-tight and incubate for 20 min. 6. Pipette 10 μl of the transfected samples into light-tight white 96-well plate. 7. Set up luminometer and add 100 μl of GLuc Assay working solution to the samples (see Note 15). 8. Gently tap the plate to mix the working solution with the samples and start the measurement.

3.2.6 SEAP Assay

1. Take 11 μl of the collected medium and heat up at 65  C for 15 min, then place the samples on ice. 2. Prepare 1
AP buffer at room temperature by diluting 10
AP buffer 1:10 using ddH2O.

146

Violetta Pilorz et al.

3. Prepare SEAP assay working solution by adding 10 μl substrate AP to 1 ml 1
AP buffer (see Note 15). Mix it thoroughly by inverting the tube several times. 4. Cover the tube light-tight and incubate for 10 min. 5. Pipette 10 μl of the transfected samples into light-tight white 96-well plate. 6. Set up luminometer and add 100 μl of SEAP assay working solution to the samples. 7. Gently tap the plate to mix the working solution with the samples and start the measurement. 8. Photon density is estimated in a luminometer and normalized against Secreted Alkaline Phosphatase (SEAP) readout (see Note 14). Luminescence is compared between different conditions such as ascending hormone concentrations (see Note 16). 3.3 Circadian Enhancer Motif Chromatin Immunoprecipitation (ChIP)

Once a candidate hormone and its target clock gene have been identified, further experiments may assess the signaling pathways leading to changes in the clock’s transcription-translation feedback loop (TTFL) after hormone treatment. Chromatin immunoprecipitation of transcription factors that have previously been implicated in transmitting the respective hormonal signal is a technique that allows the experimenter to assess hormone regulated promoter occupancy. ChIP may be used in cellular, explant, and in vivo systems.

3.3.1 Tissue Collection and Crosslinking

1. The source material is an organotypic slice culture after hormone stimulation, as described above. Scrape off the tissue slices of the membrane insert and homogenize (see Note 17). 2. Incubate the tissue with 1% formaldehyde in 1
PBS with protease inhibitors for 15 min at room temperature under gentle shaking. 3. Stop the crosslinking reaction with the addition of 1/10 volume of 1.25 M glycine. 4. Centrifuge (1000 rpm, 4 supernatant.



C, 5 min) and discard the

5. Wash the tissue with cold PBS/PI (see Note 18), centrifuge again (1000 rpm, 4  C, 5 min). 6. Discard the supernatant. 7. Homogenize the tissue in 500 μl cell lysis buffer. 8. Sonify in cycles of 15 s “on”/20 s “off” for 22 min at 4  C [24] (see Note 19). 9. Centrifuge (24,104
g, 10 min, 4  C) to eliminate cell debris and collect the supernatant.

Studying Circadian Clock Entrainment by Hormonal Signals

147

10. Dilute the supernatant ten times. IMPORTANT: keep a portion of the supernatant from this step at 20  C as a 1% input control to serve for normalization (see Note 20). 11. Add 5 μg of anti-CREB antibodies to 1 ml of diluted sample (see Note 21) and incubate overnight at 4  C. As a positive control, antibodies against histones can be used, while negative control antibodies should not bind any DNA-binding proteins (e.g., anti-GFP or anti-IgG). 12. Incubate the antibody-protein-chromatin complex with A/G agarose beads (60 μl) for 1 h at 4  C with gentle shaking. 13. Separate the agarose beads from the solution by centrifugation (1000 rpm, 1 min, 4  C). 14. Wash five times with prechilled PBS. 3.3.2 DNA Isolation

1. Add 100 μl of a 10% Chelex solution to the agarose beads (see Note 22 and 23). 2. Incubate the sample with 3 volumes of 100% ethanol for 2 h at 20  C and then centrifuge (24,104
g, 30 min, 4  C). At this time, the input sample saved before can be processed in parallel starting with ethanol precipitation. 3. Wash the obtained DNA pellet with 70% prechilled ethanol and centrifuge once more (24,104
g, 10 min, 4  C). 4. Discard the supernatant, dry the pellet, and dissolve in 100 μl Chelex. From that point on, all samples should be processed together. 5. Vortex the samples, then boil for 10 min, allow to cool down. 6. Treat for 30 min at 55  C with 1 μl proteinase K to digest remaining proteins, which would interfere with the following PCR reaction. 7. Inactivate the enzyme for 10 min at 95  C (see Note 24). 8. Centrifuge (17,709
g, 1 min, 4  C) the Chelex/agarose slurry to separate the water-soluble DNA. 9. Remove the aqueous phase to a clean tube. 10. To increase yields, the Chelex can be eluted once again by addition of 100 μl of water.

3.3.3 qPCR and Analysis

DNA samples are now ready to be analyzed by quantitative (q)PCR. The results are normalized to the 1% input sample (see Notes 2 and 25). The more protein is bound to chromatin, the more starting DNA template is present in the qPCR reaction. If the hormone activates transcription via the transcription factor targeted in the ChIP-qPCR experiment, a high amount of transcription factor bound to DNA can be expected. Consequently, the qPCR signal from the stimulated tissue sample should be higher than for the control (Fig. 3).

148

Violetta Pilorz et al.

Fig. 3 OXM promotes binding CREB to CRE elements at the Per1 gene promoter. PBS treatment represents control. Black bar depicts the time point zero (before treatment), gray and white bars illustrate OXM and PBS treatments for 30 min. 500 bp stands for downstream sequences (no-binding control). (Figure adapted from ref. 24 with permission from eLife Sciences)

4

Notes 1. Alternatively, a commercially available ChIP-compatible lysis buffer can be used. 2. Primers for ChIP should flank the gene promotor region and contain the binding sequence of an investigated transcription factor. Genomic sequence should be used as a template for primer design. Otherwise, the design follows standard recommendations for qPCR primer design. 3. Luminescence reporters are particularly useful for this purpose, as phototoxicity or bleaching observed with many fluorescence reporters under long-term monitoring conditions are not an issue. It has been shown that the phase of luminescence rhythms in ex vivo tissue slice cultures from Per1-luc mice (expressing the firefly luciferase mRNA from the Per1 promoter) match the rhythm of luciferase mRNA in vivo [29]. The circadian rhythm and phase of PER2 protein has also widely been used as a clock readout in PER2::LUC reporter mice carrying a luciferase knock-in into the 30 -region of the Per2 gene, thus expressing a PER2::LUCIFERASE fusion protein under control of the native Per2 promoter [16]. Such tissue slice cultures can be made from animals of a wide range of ages and a broad array of peripheral and central tissues.

Studying Circadian Clock Entrainment by Hormonal Signals

149

4. Before animal decapitation it is necessary to consider the effect of light at night on the circadian phase, as it may lead to phase changes in circadian organization. Therefore, placing the time of sacrifice in the light phase should be avoided. However, if it is experimentally necessary to work under constant dark conditions, sampling should be carried out under dim red light. 5. Various peripheral and central tissues such as liver or adrenal can be used in such experiments, though preparation protocols such as the way of dissection and slicing need to be adapted. Larger tissues such as liver or kidney yield better results when prepared using a vibratome. Small tissues such as adrenal or CNS tissues can be sliced quickly with a sterile scalpel. 6. The handling of the recording medium containing luciferin is preferably carried out under low light intensities. 7. The LumiCycle itself does not contain a temperature controller; therefore it needs to be kept in a light-tight incubator with temperature set between 32 and 37  C depending on the cell or tissue type. 8. When using 96-well-based luminometers, resetting experiments can easily be up-scaled to screen for clock resetting cues in cell samples at medium- to large-scale levels. 9. The luciferase reporter assay provides information about a functional connection between the investigated hormone and its effect on the core clock machinery. To estimate the amount of gene product that is produced in the presence of the hormone, a construct will be used in which the regulatory region of a clock target gene, e.g., Per1 or Per2, is fused to the DNA coding sequence of the firefly luciferase enzyme. A second construct contains a second reporter cDNA, e.g., GFP or Renilla luciferase, which is used as control for transfection efficiency. Cells derived from the tissue of interest may be used, but in many cases human embryonic kidney (HEK 293) cells are used for the sake of their high transfection efficiencies. 10. The prepared agar plates contain relevant antibiotics (e.g., 30 μg/ml kanamycin or 50 μg/ml ampicillin). 11. The DNA purification can be performed using DNA purification kit: Macherey-Nagel NucleoBond Xtra Midi Kit. 12. As some variability in samples can be expected, use triplicates of each condition in a 96- or 24-well plate. 13. As phenol red would impact the light output, the use of phenol red in luminescence readout culture media should be avoided. 14. Photon density is estimated in a luminometer and normalized against Secreted Alkaline Phosphatase (SEAP) readout, a secondary reporter gene when Gaussia Luciferase (GLuc) is used. Using SEAP or GLuc secreted reporters has the advantage of

150

Violetta Pilorz et al.

allowing for their detection in the cell-free conditioned medium avoiding the need for cell lysis. In addition, the cells remain intact and available for confirmatory analyses [30]. 15. Luciferase assay buffers and substrates are supplied with the luciferase assay kit; they must be stored at 20  C and thawed at room temperature before use. 16. Upregulated transcription from the clock gene promoter results in expression of luciferase, which is measured after 48 h. Known regulators of the used clock gene promoter may be added as positive controls, e.g., Bmal1 combined with Clock for Per1/2, and for interactive studies. 17. The technique can be applied to any tissue or cell culture samples, although optimization may be necessary. Tissue can be collected and stored at 80  C before use. The technique requires only a small amount of tissue as input, approximately 1–3 mm3. 18. Cold temperatures inhibit protease activity that would be detrimental to the experiment, which is why keeping samples cooled is of utmost importance. 19. One of the most sensitive steps of the procedure. Sonication requires optimization for each sonicator and tissue type. The sonication shreds the chromatin into short strands. However, because the proteins of interest are crosslinked with the DNA by formaldehyde, the sequences bound by proteins are protected from cleavage. It is advised to apply short pulses separated by a pause while samples are placed on ice or a cooling block. Long sonication intervals can lead to overheating and may damage the sample. Too short, or too few pulses result in insufficient fragmentation. The optimal chromatin fragment size is between 200 and 1000 bp. When optimizing the protocol for a specific application, sonication efficiency can be analyzed by agarose gel electrophoresis. Another challenge for a successful sonication is to avoid sample foaming. One recommendation is to keep the sonicator nozzle close to the bottom of the tube. Alternatively, water bath homogenizers can be used that also allow for homoform sonication of a group of samples from the same experiment, e.g., for time-course evaluation. 20. The supernatant from this step can be stored at 80  C for several months. 21. In this step, proteins of interest with bound DNA are precipitated from the solution using a specific antibody. Antibodies for different targets, from different producers and even from different batches vary in their binding efficiency; therefore each application may require antibody-specific optimization. In circadian biology, CREB (cAMP-response element binding

Studying Circadian Clock Entrainment by Hormonal Signals

151

protein) is an interesting transcription factor to study, as cAMP-response (CRE) elements are present in Per1 and Per2 promoters and involved in clock resetting (Fig. 3) [31–33]. 22. This is so-called fast ChIP method that involves DNA isolation using Chelex [34]. 23. The reagent is difficult to pipet and wide-bore tips should be used. 24. In this approach, the reversal of the crosslinking is not required. 25. It is recommended to include a negative sequence control for the qPCR amplification with primers flanking a DNA sequence where no binding of protein is expected. As a positive control, one can design primers for a sequence that is known to bind the protein of interest. A useful resource is ChIPprimers DB (http://chipprimers.com), a database containing validated ChIP primer sequences [35].

Acknowledgments This work was funded by grants of the German Research Foundation (DFG; OS353/7-1 and OS353/10-1) to H.O. References 1. Ralph MR, Foster RG, Davis FC, Menaker M (1990) Transplanted suprachiasmatic nucleus determines circadian period. Science 247(4945):975–978 2. Deboer T (2018) Sleep homeostasis and the circadian clock: do the circadian pacemaker and the sleep homeostat influence each other’s functioning? Neurobiol Sleep Circadian Rhythm 5:68–77 3. Pickel L, Sung HK (2020) Feeding rhythms and the circadian regulation of metabolism. Front Nutr. https://doi.org/10.3389/fnut. 2020.00039 4. Scheer FAJL, Morris CJ, Shea SA (2013) The internal circadian clock increases hunger and appetite in the evening independent of food intake and other behaviors. Obesity 21(3): 421–423 5. Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabolism, and the circadian clock. Cell 161(1):84–92 6. Greco CM, Sassone-Corsi P (2019) Circadian blueprint of metabolic pathways in the brain. Nat Rev Neurosci 20(2):71–82

7. Kalsbeek A, La Fleur S, Fliers E (2014) Circadian control of glucose metabolism. Mol Metab 3(4):372–383 8. Reinke H, Asher G (2019) Crosstalk between metabolism and circadian clocks. Nat Rev Mol Cell Biol 20:227–241 9. Van Der Veen DR, Riede SJ, Heideman PD et al (2017) Flexible clock systems: adjusting the temporal programme. Philos Trans R Soc B Biol Sci 372(1734):20160254 10. Stephan FK, Swann JM, Sisk CL (1979) Entrainment of circadian rhythms by feeding schedules in rats with suprachiasmatic lesions. Behav Neural Biol 25(4):545–554 11. Van Der Veen DR, Saaltink DJ, Gerkema MP (2011) Behavioral responses to combinations of timed light, food availability, and ultradian rhythms in the common vole (Microtus arvalis). Chronobiol Int 28(7):563–571 12. Bass J, Takahashi JS (2010) Circadian integration of metabolism and energetics. Science 330(6009):1349–1354 13. Balsalobre A, Damiola F, Schibler U (1998) A serum shock induces circadian gene expression

152

Violetta Pilorz et al.

in mammalian tissue culture cells. Cell 93(6): 929–937 14. Nagoshi E, Saini C, Bauer C et al (2004) Circadian gene expression in individual fibroblasts: cell-autonomous and self-sustained oscillators pass time to daughter cells. Cell 119(5): 693–705 15. Welsh DK, Yoo SH, Liu AC et al (2004) Bioluminescence imaging of individual fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression. Curr Biol 14(24):2289–2295 16. Yoo SH, Yamazaki S, Lowrey PL et al (2004) PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101(15):5339–5346 17. Mohawk JA, Green CB, Takahashi JS (2012) Central and peripheral circadian clocks in mammals. Annu Rev Neurosci 35:445–462 18. Davidson AJ, Poole AS, Yamazaki S, Menaker M (2003) Is the food-entrainable circadian oscillator in the digestive system? Genes Brain Behav 2(1):32–39 19. Izumo M, Pejchal M, Schook AC et al (2014) Differential effects of light and feeding on circadian organization of peripheral clocks in a forebrain Bmal1 mutant. eLife. https://doi. org/10.7554/eLife.04617 20. Tahara Y, Kuroda H, Saito K et al (2012) In vivo monitoring of peripheral circadian clocks in the mouse. Curr Biol 22(11):1029–1034 21. Mistlberger RE (2020) Food as circadian time cue for appetitive behavior. F1000Research. https://doi.org/10.12688/f1000research. 20829.1 22. Piggins HD, Bechtold DA (2015) Feeding time: a hormone released from the gut after a meal can reset clock gene activity in the liver. eLife. https://doi.org/10.7554/ eLife.08166 23. Tsang AH, Barclay JL, Oster H (2013) Interactions between endocrine and circadian systems. J Mol Endocrinol 52(1):R1–R16 24. Landgraf D, Tsang AH, Leliavski A et al (2015) Oxyntomodulin regulates resetting of the liver circadian clock by food. eLife. https://doi. org/10.7554/eLife.06253

25. Tsang AH, Koch CE, Kiehn JT et al (2020) An adipokine feedback regulating diurnal food intake rhythms in mice. eLife. https://doi. org/10.7554/eLife.55388 26. Pilorz V, Cunningham PS, Jackson A et al (2014) A novel mechanism controlling resetting speed of the circadian clock to environmental stimuli. Curr Biol 24(7):766–773 27. Yamazaki S, Takahashi JS (2005) Real-time luminescence reporting of circadian gene expression in mammals. Methods Enzymol 393:288–301 28. Carter M, Shieh J (2015) Biochemical assays and intracellular signaling. In: Guide to research techniques in neuroscience, 2nd edn. Academic, New York 29. Wilsbacher LD, Yamazaki S, Herzog ED et al (2002) Photic and circadian expression of luciferase in mPeriod1-luc transgenic mice in vivo. Proc Natl Acad Sci U S A 99(1):489–494 30. Tannous BA (2009) Gaussia luciferase assay for monitoring of biological process in culture and in vivo. Nat Protoc 4(4):582–591 31. Travnickova-Bendova Z, Cermakian N, Reppert SM, Sassone-Corsi P (2002) Bimodal regulation of mPeriod promoters by CREBdependent signaling and CLOCK/BMAL1 activity. Proc Natl Acad Sci U S A 99(11): 7728–7733 32. Lee B, Li A, Hansen KF et al (2010) CREB influences timing and entrainment of the SCN circadian clock. J Biol Rhythm 25(6):410–420 33. Ding JM, Faiman LE, Hurst WJ et al (1997) Resetting the biological clock: mediation of nocturnal CREB phosphorylation via light, glutamate, and nitric oxide. J Neurosci 17(2): 667–675 34. Nelson JD, Denisenko O, Bomsztyk K (2006) Protocol for the fast chromatin immunoprecipitation (ChIP) method. Nat Protoc 1(1): 179–185 35. Kurtenbach S, Reddy R, Harbour JW (2019) ChIPprimersDB: a public repository of verified qPCR primers for chromatin immunoprecipitation (ChIP). Nucleic Acids Res 47(D1): D46–D49

Chapter 10 In Vitro Assays for Measuring Intercellular Coupling Among Peripheral Circadian Oscillators Anna-Marie Finger Abstract Circadian clocks can be found in nearly all eukaryotic organisms, as well as certain bacterial strains, including commensal microbiota. Exploring intercellular coupling among cell-autonomous circadian oscillators is crucial for understanding how cellular ensembles generate and sustain coherent circadian rhythms on the tissue level, and thus, rhythmic organ functions. Here we describe a protocol for studying intercellular coupling among peripheral circadian oscillators using three-dimensional spheroid cultures in order to measure coupling strength within peripheral clock networks. We use cell spheroids to simulate in vivo tissue integrity, as well as to increase complexity of cell–cell interactions and the abundance of potential coupling factors. Circadian rhythms are monitored using live-cell imaging of spheroids equipped with circadian reporters over several days. Key words Circadian rhythms, Intercellular coupling, Peripheral clocks, Live-cell imaging, Luciferase reporter, Fluorescence microscopy, Bioluminescence recording, Time series analysis

1

Introduction Circadian clocks are molecular oscillators present in virtually all mammalian cell types that drive circadian (~24 h) rhythms of numerous molecular, metabolic, physiological, and behavioral functions in order to allow organisms to anticipate and adapt to cyclic environmental conditions (Zeitgebers). The circadian system is essential for health [1]. Its dysregulation and misalignment with environmental cycles has been associated with the development of many pathologies, including metabolic syndrome, cardiovascular and neurodegenerative diseases, mental disorders, and cancer. In our modern 24/7 societies, which pose constant threads to our circadian system and its alignment with the natural light-dark cycle, understanding the basic mechanisms underlying circadian rhythm regulation is crucial for the development of health care strategies.

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_10, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

153

154

Anna-Marie Finger

On the organismal level, mammalian circadian clocks are organized hierarchically, with the suprachiasmatic nucleus (SCN), the “pacemaker clock,” on top. The SCN clock is entrained to the environmental light-dark cycle, as it receives photic information through melanopsin-expressing retinal ganglion cells [2]. On the cellular level, circadian clocks are cell-autonomous and selfsustained [3–5], i.e., single-cell oscillators continuously cycle with their own periods and phases driven by delayed negative transcriptional-translational feedback loops between so-called core clock genes and their own protein products—the molecular clock machinery. Within tissue ensembles, cellular oscillators display normally distributed periods and phases ranging from about 20 to 28 h. Though beautiful in its design, autonomous molecular rhythm generation entails the need for intercellular coupling between single-cell oscillators within tissue networks. Without intercellular coupling, single-cell oscillators within tissue ensembles would get out of sync with each other resulting in dampening and eventual loss of the tissue rhythm (Fig. 1). Even more importantly, under natural conditions, intercellular coupling strength, by modulating rhythm amplitude and amplitude relaxation [6–8], is related to the robustness of circadian tissue rhythms and likely governs important features of circadian tissue physiology, including acute and transient phase responses to endogenous and exogenous Zeitgeber signals, as well as entrainment to cyclic timing cues from the environment [9]. The human SCN consists of roughly 100,000 individual neuronal oscillators that couple intercellularly and in a wave-like pattern [10, 11] to sustain a coherent and robust tissue rhythm. As pacemaker clock, the SCN constitutes a strongly coupled network of single-cell oscillators, which ensures entrainment to the natural light-dark cycle within a narrow entrainment range and protects the mammalian circadian clock system against perturbation by aberrant Zeitgeber signals. The particular strength of intercellular coupling within the SCN becomes clear when looking at published studies: SCN explants display robust and non-damped tissue rhythms over long durations ex vivo, which are maintained even if single-cell oscillators within the tissue ensemble are dysfunctional [12– 15]. When intercellular communication between SCN neurons is disturbed by blocking action potentials, abnormal lighting regimens, or dispersing SCN neurons, SCN network rhythms are severely dampened and become susceptible to perturbation by Zeitgeber pulses [3, 16, 17]. Mechanistically, cell-to-cell communication and coupling among SCN neurons is not fully understood but, due to the importance of synaptic transmission, appears to depend on rhythmic neurotransmitter signaling (although gap junctions may be involved [18, 19]). Vasoactive intestinal polypeptide (VIP), the most prevalent neurotransmitter in the SCN core, is released rhythmically from SCN core neurons and strongly implicated in SCN coupling. VIP inputs to the molecular clock

Peripheral Coupling

155

Fig. 1 Intercellular coupling governs tissue rhythms. Networks of cell-autonomous circadian oscillators govern circadian rhythms of the tissue ensemble (black lines). Thus, intercellular coupling is crucial for the maintenance of coherent, high-amplitude, and robust tissue rhythms. Traditionally, the distributions of single-cell periods, phases, and amplitudes (gray lines) are used to deduced information about intercellular coupling strength within the oscillator network. Narrowing of period and phase distributions, as well as amplitude expansion are related to increased coupling strength. Additionally, synchronization of coupled single-cell oscillators will result in phase- and period-pulling, amplitude expansion, and reduced damping on the population level. Thereby, depending on the intercellular coupling strength, such pulling effects can lead to partial or complete synchronization

machinery [20], modulates light-induced phase resetting [21], and phase-shifts circadian oscillations of the SCN in vitro and in vivo [22–24]. In addition, depletion of VIP or its receptor VPAC2 weakens locomotor activity and neuronal firing rhythms [25–27] and promotes desynchronization of SCN neurons [28, 29], while SCN synchrony can be restored by daily application of VIP agonist [30]. Besides VIP, other neurotransmitters, e.g., GRP, GABA, AVP, and neurotensin, have been suggested to play a role in SCN coupling by regulating neuronal clock gene expression, phaseshifts, or synchronization [12, 13]. In contrast to the SCN, intercellular coupling within peripheral clock tissues is less well-understood and even debated in the field of chronobiology [31]. Largely that stems from the fact that, opposite to the SCN, ensemble rhythms of autonomous peripheral

156

Anna-Marie Finger

oscillators have been described to show substantial dampening of network rhythms over time. In addition, other studies report a lack of phase and period coupling using conventional (co-) culture of fibroblasts in vitro [5, 32, 33] and of rhythmic clock genes expression measured from population samples of animals housed under constant conditions [34–36]. More recently however, evidence has accumulated supporting the hypothesis that peripheral oscillators couple intercellularly: (1) peripheral circadian clocks display persistent low-amplitude rhythms independently of the SCN or external entrainment signals in vivo [37–39], (2) adult peripheral tissue slices, embryonic tissue explants cultured in toto, and organoid models of peripheral tissues display robust circadian rhythms ex vivo (although with reduced amplitudes; [40–42]), and (3) ensembles of cultured peripheral cells display density dependence, local coupling, and phase cross talk in vitro [43–47]. Interestingly, weak rhythmicity of sparsely cultured peripheral oscillators can be rescued by conditioned medium from high-density cultures [43], suggesting that, similarly to central neurotransmitters, secreted signaling factors may mediate intercellular communication and coupling within peripheral clock networks. Thus, overall, published findings indicate that peripheral circadian oscillators are indeed able to couple intercellularly, however also that, compared to the SCN pacemaker clock, the intercellular coupling within peripheral tissue networks is much weaker. Nevertheless, due to the intertwined relationship of intercellular coupling with oscillator robustness and entrainment, even weak coupling may promote partial synchronization of autonomous single-cell oscillators and control how peripheral body clocks regulate tissue-specific circadian outputs in response to intrinsic or extrinsic Zeitgeber and entrainment signals. Under natural conditions, when the SCN is intact, this may constitute an essential mechanism for regulating the balance between peripheral clock precision and plasticity in response to incoming timing cues. Exploring intercellular coupling among peripheral circadian oscillators will increase the knowledge about the multidimensional circadian clock system in mammals and improve the understanding of pathophysiological mechanisms contributing to circadian disruption and associated disease. Experimentally, intercellular coupling (strength) within ensembles of circadian oscillators can be assessed by analyzing period, amplitude, and phase distributions of singe cells (single-cell imaging) or by measuring synchronization/ pulling effects between oscillator populations that differ in circadian parameters (population imaging). Simulations of oscillator networks, as well as SCN coupling studies, show that period and phase distributions become narrower, and amplitudes expand with increasing coupling strength [7]. Similarly, coupled oscillator populations are expected to display period- and phase-pulling toward the period/phase of the dominant population (e.g., the numerically more abundant population), as well as expansion of

Peripheral Coupling

157

Fig. 2 Experimental workflow. Graphical workflow of studying intercellular coupling using 3D cell spheroids. The 15-day protocol described here affords only 3 days of actual hands-on time and includes free weekends. Formation of three-dimensional cell aggregates from hanging drop cultures should be monitored regularly to adapt the incubation period required to attain solid spheroids for imaging

the mean amplitude as the network synchrony increases [7, 48]. Because, compared to neuronal oscillator networks, intercellular coupling strength within peripheral oscillator ensembles is expected to be weak, traditional 2D monolayer cultures are not suitable to study peripheral coupling. Additionally, molecular mechanisms of peripheral coupling are unknown, rendering the targeted disruption of intercellular coupling pathways, comparable to blocking of neuronal firing or VIP signaling in the SCN, an unfeasible approach when studying peripheral coupling. Thus, we have developed an unbiased approach for assessing intercellular coupling among peripheral oscillator populations using real-time bioluminescence imaging of three-dimensional spheroids (Fig. 2) that simulate in vivo tissue microenvironments by increasing the complexity of cell–cell and cell–matrix interactions, as well as of local concentrations of hypothetical coupling factors.

2

Materials Prepare all cell culture models, reagents, and media under aseptic conditions under laminar flow. Follow all waste disposal and biosafety regulations when disposing/decontaminating materials. Store all media, buffers, and reagents at 4
C. Store fetal bovine serum (FBS), Penicillin/Streptomycin, trypsin/EDTA, dexamethasone, luciferin, and matrix gel at 20
C. Additionally, luciferin

158

Anna-Marie Finger

should be stored in aliquots in light-tight boxes as some papers report sensitivity of D-Luciferin to light, oxygen, and moisture as powder and in solution. The described protocol was established with the indicated (reporter) cell lines; however, other primary or immortalized (reporter) cells might be used as well (see Note 1). 2.1 Cell Lines (See Note 1)

1. U-2 OS wildtype non-reporter cells (ATCC HTB-96). 2. U-2 OS Bmal1:Luc wildtype reporter cells [49, 51]. 3. U-2 OS Bmal1:Luc long period CRY2/ reporter cells [51]. 4. U-2 OS Bmal1:Luc short period TNPO1/ reporter cells [52].

2.2

Reagents

1. Complete culture medium: 500 mL High Glucose DMEM, 50 mL FBS, 12.5 mL 1 M HEPES pH 7.3, 5 mL 10,000 U/ mL Penicillin/Streptomycin. 2. Reporter medium: 500 mL High Glucose DMEM phenolredfree, 5 mL 10,000 U/mL Penicillin/Streptomycin. 3. 1 mM dexamethasone solubilized in EtOH. 4. Matrix gel (e.g., Cultrex growth factor reduced BME or Matrigel). 5. 1 Phosphate-Buffered Saline (PBS), pH 7.2. 6. Trypsin/EDTA. 7. 25 mM D-Luciferin solubilized in High Glucose DMEM phenolred-free.

2.3

Equipment

1. Cell culture incubator. 2. Bioluminometer (see Note 2). 3. Incubator for bioluminometer. 4. Nunc Delta 100  17-mm dishes. 5. Nunc Delta 35  10-mm dishes. 6. Nunc Delta 12-well plate. 7. Dow Corning High Vacuum Grease. 8. Parafilm M.

2.4 Software ChronoStar

ChronoStar is a program for analysis of time series data describing damped oscillations overlaid by a nonlinear trend [53, 54]. A typical but not exclusive application is the analysis of biological time series with damped circadian oscillations, e.g., bioluminescence data. The software runs with a graphical user interface with visualization and real-time adjustment of time series data and fitting parameters. The program is available for download free of charge for academic institutions at: www.achim-kramer-lab.de.

Peripheral Coupling

3

159

Methods The described method is intended to measure period-pulling among populations of U-2 OS cells, an established model of peripheral circadian oscillators, that differ in their intrinsic circadian periods due to genetic perturbation of the clock machinery. Oscillator populations are expected to display bidirectional periodpulling if their intercellular coupling strength allows for (partial) synchronization. Even though we have not previously tested other applications of this assay, we are convinced that it may also be used to assess phase-pulling among oscillator populations differing in circadian phase or to study distributions of circadian parameters within cell spheroids using single-cell imaging (see Notes 1 and 2).

3.1 ThreeDimensional Spheroid Cultures: General Considerations

To this day, cell-based assays using two-dimensional monolayer cells cultured on hard plastic are traditionally used to study cellular behavior. However, in vivo, most cell types are surrounded by neighboring cells and extracellular matrix (ECM) in a threedimensional fashion. This alters spatial and morphological organization, communication, as well as physical constraints experienced by cells. Indeed, cell-autonomous circadian oscillators have been found to respond to their extracellular mechano-environments in a cell-type specific fashion [55]. Thus, 3D culture systems simulate more accurately the actual microenvironment where cells reside in tissues. An important difference between 2D and 3D culture systems is the access to nutrients and oxygen: while monolayer cells have uniform access, cells in 3D aggregates are subjected to a gradient of oxygen and assay reagents. Cell spheroids, comparable to in vivo tissues, display cellular heterogeneity and are composed of an inner apoptotic/hypoxic/necrotic, an intermediate quiescent, and an outer proliferative cell layer. Thus, when using spheroids as assay model, researchers should be aware that the added complexity of 3D cultures also presents challenges to experimental designs [56]. For example, reagents may not penetrate to the center of the spheroid, different cells may have different growth characteristics leading to variability in spheroid size, morphology, and composition, and cell viability and proliferation rates will differ compared to 2D monolayer cultures. 3D spheroids can be grown from cells using scaffold-based (e.g., hydrogels or inert matrices) or scaffold-free (e.g., low-adhesion plates, micropatterned surfaces, or hanging drops) methods, each of which comes with different advantages and disadvantages. The culture methods to be chosen will depend on the cell type and research question. Here, we describe three-dimensional spheroid cultures using the easy and cost-effective hanging drop method, in which cells grown in suspension form their own ECM and accumulate as spheroids. This

160

Anna-Marie Finger

method includes the advantages to control the size of the spheroid, experimental reproducibility, and the ability for automated dispensing. 3.1.1 Week 1: Hanging Drop Cultures

1. Wash confluent U-2 OS cell lines 1 with PBS and trypsinize for 5–10 min under standard cell culture conditions to detach cells. 2. Harvest cells in fresh complete culture medium and count in hemocytometer. Adjust the concentration of the cellular suspensions to 3.0  106 cells/mL. 3. Prepare control- and co-cultures of oscillator populations by mixing cellular suspensions in a 1:4 ratio (see Note 3). (a) Controls: U-2 OS Bmal1:Luc wildtype (5 parts), U-2 OS Bmal1:Luc CRY2/ (5 parts), and U-2 OS Bmal1:Luc TNPO1/ (5 parts). (b) Co-cultures: U-2 OS Bmal1:Luc wildtype (1 part) + U2 OS wildtype (4 parts), U-2 OS Bmal1:Luc CRY2/ (1 part) + U-2 OS wildtype (4 parts), and U-2 OS Bmal1: Luc TNPO1/ (1 part) + U-2 OS wildtype (4 parts). 4. Per culture condition (from step 3), prepare one 100-mm dish with 10 mL PBS in the bottom of the dish. Seed 10 μL drops of each culture conditions on the inverted lid of the culture dishes (see Note 4). A multichannel pipette can be used for seeding. Each lid should hold about 30–40 10 μL drops. Make sure that drops are placed sufficiently apart so as to not touch. 5. Quickly re-invert the lids holding the droplets and place them back onto the PBS-filled bottom of the dishes. The surface tension of the droplets will prevent their disruption upon lid inversion. Incubate the hanging drop cultures in a standard tissue culture incubator (37
C, 95% humidity, 5% CO2) to allow cells to aggregate and form spheroids (see Note 4).

3.1.2 Week 2: Prepare Spheroid Cultures

1. Prepare matrix gel for seeding of spheroids according to the manufacturer’s instructions (see Note 4). 2. When solid spheroids have formed, detach hanging drops with a 1000 μL pipette by gently applying a small volume of complete culture medium. Medium suspension containing the 3D spheroids can be collected in the inverted lid of the dish by holding it slightly tilted. 3. To wash spheroids, transfer the harvested suspensions to a 12-well plate containing 1 mL/well PBS (one well per culture condition). Pipette up and down 3–5 using a 1000 μL pipette.

Peripheral Coupling

161

4. Remove as much medium/PBS as possible without aspirating the cell spheroids. Following aspiration of the supernatant, work quickly to avoid attachment of the spheroids to the wells. 5. Suspend spheroids in ~300 μL matrix gel (volume may need to be increased if residual volume of medium/PBS is high, see Note 4). Seed 50–75 μL gel suspension as droplet into 35-mm dishes (suspensions can be split to seed multiple replicates, see Note 5). 6. Place 35-mm dishes into standard tissue culture incubators and solidify gel suspension for 30–60 min. 7. Overlay gel droplets containing spheroids with 2 mL/dish pre-warmed complete culture medium (see Note 4). Incubate in a standard tissue culture incubator overnight. 3.1.3 Week 2: Bioluminescence Imaging of Spheroid Cultures

1. Prepare imaging medium by diluting the 25 mM D-Luciferin stock 1:100 in the required volume of pre-warmed reporter medium (250 μM final concentration). 2. Following overnight incubation, synchronize spheroid cultures by diluting the 1 mM dexamethasone stock 1:1000 in the culture supernatant (1 μM final concentration). Incubate for 20–30 min in a standard tissue culture incubator. 3. Wash spheroid cultures twice with pre-warmed PBS, add 2 mL/dish imaging medium, and seal dishes with grease and parafilm (see Note 4). 4. Place sealed 35-mm dishes containing the 3D spheroids into the bioluminometer located in a tissue culture incubator (37
C, non-humidified, 5% CO2) and monitor bioluminescence oscillations for 4–7 days (see Note 6).

3.2 Time Series Analysis Using ChronoStar

A detailed description of time series analysis with ChronoStar software can be found in [53]. Briefly, bioluminescence raw data can be uploaded to ChronoStar (see Note 7). ChronoStar will display raw data (top panel, black line) and corresponding trends (top panel, blue line), as well as trend-eliminated data (bottom panel, black line) and corresponding sine wave fits (bottom panel, blue line). Trend elimination is performed by dividing raw time series data by its 24-h running average and circadian parameters are extracted by fitting a cosine wave function including an exponential decay term (see Note 7). Data analysis settings can be modified to adjust boarders of the imaging window, scale of the y-axis, reference period for detrending, and reference timepoint for amplitude extraction. Raw and analyzed data (trends, detrended data, fits, circadian parameters) can be exported and will be saved in .xml file format for further processing or plotting. Period-pulling effects are calculated by determining the difference in extracted periods of the co-culture conditions and their respective controls.

162

4

Anna-Marie Finger

Notes 1. In our laboratory we have established the 3D spheroid culture model of peripheral oscillators using U-2 OS cells harboring firefly luciferase reporters because of following features: (a) U-2 OS cells are a well-described model of human peripheral oscillators commonly used in circadian research, (b) U-2 OS cells display high transfection and transduction efficiency, allowing for easy genetic manipulation, (c) U-2 OS cells show suitable growth characteristics for spheroid culture and form regular, solid aggregates, as well as (d) bioluminescence imaging allows for long-term recordings with high sensitivity, favorable dynamics, and no risk for phototoxicity or -bleaching. Additionally, the described cell lines have been chosen due to their intrinsic period differences, making it simple to study period-pulling/synchronization between distinct peripheral oscillator populations upon co-culture. However, despite these advantages, studying intercellular coupling in peripheral oscillator networks using 3D spheroid cultures is not limited to the specific cell types and cell lines presented here. For example, cells harboring fluorescent reporters may be used to assess circadian parameters of single cells from live-cell fluorescence recordings (see Note 2) or cells expressing transgenes enabling selective phase-shifting of one of the oscillator populations (e.g., light-entrainable melanopsin-expressing PER2::LUC fibroblasts [57]) may be used to study phase-synchronization as additional measure of peripheral coupling. Ultimately, the selection and design of (reporter) cells used will depend on the research question and available tools (e.g., phase versus period synchronization, mean circadian parameters versus single-cell distributions, bioluminescence versus fluorescence imaging, scaffold-based versus scaffold-free formation of cell aggregates). 2. Our experimental setup consists of a LumiCycle 32 (32-channel luminometer for 35-mm dishes) with 4 photon-counting photomultiplier tubes (PMTs) located in a standard tissue culture incubator. However, other devices and formats might be used to study peripheral coupling using 3D spheroids. For example, spheroid gel suspensions may be seeded into 96-well plates to increase the experimental throughput and perform bioluminescence imaging using standard microplate luminometers or the NXT TopCount. Cells expressing circadian fluorescence reporters may be used to measure spheroid circadian rhythms with a fluorescence microscope (caution: media formulations that reduce background fluorescence while supplying required nutrients for routine cell culture should be used), which, if different fluorescent proteins are used, will

Peripheral Coupling

163

even allow to distinguish between cellular populations. Moreover, both fluorescence and bioluminescence imaging can be used to study coupling via the assessment of single-cell distributions of circadian parameters rather than populations dynamics upon co-culture. Nevertheless, such advanced imaging techniques of three-dimensional specimens set new challenges and imposes specific requirements concerning the choice of suitable reporter systems, as well as fluorescence microscopes or bioluminescence cameras [58, 59]. 3. To study period-synchronization between our peripheral oscillators models, we chose a co-culture ratio of 1:4. When studying ensembles of weakly coupled oscillators (as expected for peripheral oscillator cells) it appears to be important to have one oscillator population in excess as this will enhance pulling/ synchronization effects on the other population. However, variable co-cultivation ratios can be used, also to test ratiodependent changes in intercellular coupling strength and to explore limits of peripheral oscillator synchronization. When the reporter cell population is used to measure period changes upon co-culture, the minimal abundance of the reporter cells required to observe pulling effects, while still being able to measure bioluminescence signals reliably, will be limited by the sensitivity of the luminometer. In addition, when establishing 3D spheroids from co-cultured cells, especially when using cell types or cell lines with different growth properties, researchers should be careful that the low abundant population is not overgrown by the high abundant population. 4. The protocol presented here describes the scaffold-free hanging drop culture to establish 3D spheroids that will be embedded in a matrix gel for imaging. Some practical aspects should be taken into consideration when using this method: larger or smaller petri dishes can be used for seeding of hanging drops depending on the desired experimental throughput. Depending on the size and growth properties of the cell type used, seeding volume of the hanging drops, concentration of the seeded cell suspension, or incubation time of the hanging drops may need to be optimized to allow for the formation of tight spheroids. Formation of cell aggregates should be checked regularly by visual inspection and/or inspection with a stereo or inverted microscope (when assessing aggregate formation, care should be taken when handling the 100-mm dishes to not disrupt the hanging drops). When harvesting spheroids with a pipette, the tip of the 1000 μL pipette tip can be cut off to avoid disruption of larger spheroids. Matrix gel should be thawed on ice and at 4
C overnight to ensure dispersion of the gel. Additionally, when suspending spheroids in matrix gel, the volume of residual medium/PBS should not

164

Anna-Marie Finger

exceed 10% of the gel to ensure polymerization of the matrix. Following embedding of the spheroids, all media/buffers coming in contact with the gel must be pre-warmed to 37
C to avoid disruption of the solidified matrix. Depending on the initial concentration of the matrix gel, it may be diluted 1:1 in reporter medium (as long as polymerization of the matrix still works). If the chosen cell type allows, we recommend using serum-free imaging medium when studying intercellular coupling since serum will contain unknown protein factors that may influence coupling behavior. Lastly, independent of the format or platform used for long-term spheroid imaging, dishes/plates should be sealed to avoid evaporation of the imaging medium. 5. In our experimental setup spheroids were suspended evenly in matrix gel and seeded as 50–75 μL droplets into 35-mm dishes, i.e., one droplet contained multiple spheroids (about 5–10/ droplet). Because spheroids are synchronized prior to the start of bioluminescence imaging, recording luminescent signal from multiple spheroids in one droplet does not pose a problem. However, serial dilutions of the matrix gel-spheroid suspension (in matrix gel) can be performed to seed as few as one spheroid/droplet/dish. In such cases, the brightness of the reporter system and the sensitivity of the imaging device used will be crucial for determining the minimum number of spheroids that can be embedded per droplet. 6. Real-time bioluminescence measurements were performed at ambient humidity levels as high humidity will promote corrosion of electronic components of the luminometer. We further recommend a final sampling interval of 30 min to obtain high resolution, which may be especially important when extracting phase information. The minimum sampling interval is critically influenced by the number of PMTs for parallel sampling, integration time (time of PMT on one sample/interval), PMT sensitivity/noise level, signal intensity of the samples, and number of plates monitored in parallel. 7. Raw data files that are to be uploaded to ChronoStar software should be in tab delimited .txt format (ASCII code). First row: column headers; first column: time in days; columns 1–n: bioluminescent counts; number format: rational, positive elements with point as decimal separator. ChronoStar view can be customized and individual panels may be resized, rearranged, or dragged to convenient places within the software applications. Data columns may be rearranged or individually selected using the customize table option. When performing phase shift or stimulation experiments, circadian parameters should be extracted separately for pre- and be post-treatment imaging periods, i.e., raw data must be cut accordingly (otherwise the

Peripheral Coupling

165

fitting algorithm may introduce errors when setting the boarders). Additionally, to allow for phase comparison across conditions and samples, we recommend normalizing the extracted phase to the extracted period on a 24-h scale. Visual inspection of the detrended data and fitted curve is always favorable. ChronoStar performs trend-elimination by dividing the raw data by its a 24-h running average and rhythm parameters are estimated by fitting a cosine wave function including an exponential decay term (Eq. 1). x ðt Þ ¼ edt  A  cos ðωt  24  ω  ϕÞ

ð1Þ

A ¼ amplitude d ¼ damping constant ω ¼ 2π/period [h] ϕ ¼ phase [h] t ¼ time [h]

Acknowledgments Work in the host laboratory is supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)— grants 278001972—TRR 186 and KR1989/12—and the Joachim Herz Stiftung (Add-On Fellowship for Interdisciplinary Life Sciences). Figures were created with BioRender.com. References 1. Rijo-Ferreira F, Takahashi JS (2019) Genomics of circadian rhythms in health and disease. Genome Med 11:82 2. Panda S (2002) Melanopsin (Opn4) requirement for normal light-induced circadian phase shifting. Science 298:2213–2216 3. Webb AB, Angelo N, Huettner JE et al (2009) Intrinsic, nondeterministic circadian rhythm generation in identified mammalian neurons. Proc Natl Acad Sci U S A 106:16493–16498 4. Leise TL, Wang CW, Gitis PJ et al (2012) Persistent cell-autonomous circadian oscillations in fibroblasts revealed by six-week single-cell imaging of PER2::LUC bioluminescence. PLoS One 7:e33334 5. Welsh DK, Yoo S-H, Liu AC et al (2004) Bioluminescence imaging of individual fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression. Curr Biol 14:2289–2295

6. Abraham U, Granada AE, Westermark PO et al (2010) Coupling governs entrainment range of circadian clocks. Mol Syst Biol 6:438 7. Schmal C, Herzog ED, Herzel H (2018) Measuring relative coupling strength in circadian systems. J Biol Rhythm 33:84–98 8. Bordyugov G, Granada AE, Herzel H (2011) How coupling determines the entrainment of circadian clocks. Eur Phys J B 82:227–234 9. Finger A-M, Dibner C, Kramer A (2020) Coupled network of the circadian clocks: a driving force of rhythmic physiology. FEBS Lett 594: 2734–2769 10. Evans JA, Leise TL, Castanon-Cervantes O et al (2011) Intrinsic regulation of spatiotemporal organization within the suprachiasmatic nucleus. PLoS One 6:e15869 11. Yamaguchi S, Isejima H, Matsuo T et al (2003) Synchronization of cellular clocks in the suprachiasmatic nucleus. Science 302:1408–1412

166

Anna-Marie Finger

12. Mohawk JA, Takahashi JS (2011) Cell autonomy and synchrony of suprachiasmatic nucleus circadian oscillators. Trends Neurosci 34: 349–358 13. Welsh DK, Takahashi JS, Kay SA (2010) Suprachiasmatic nucleus: cell autonomy and network properties. Annu Rev Physiol 72:551–577 14. Liu AC, Welsh DK, Ko CH et al (2007) Intercellular coupling confers robustness against mutations in the SCN circadian clock network. Cell 129:605–616 15. Nakamura W, Honma S, Shirakawa T et al (2002) Clock mutation lengthens the circadian period without damping rhythms in individual SCN neurons. Nat Neurosci 5:399–400 16. Buhr ED, Yoo S-H, Takahashi JS (2010) Temperature as a universal resetting cue for mammalian circadian oscillators. Science 330: 379–385 17. Ohta H, Yamazaki S, McMahon DG (2005) Constant light desynchronizes mammalian clock neurons. Nat Neurosci 8:267–269 18. Long MA, Jutras MJ, Connors BW et al (2005) Electrical synapses coordinate activity in the suprachiasmatic nucleus. Nat Neurosci 8: 61–66 19. Rash JE, Olson CO, Pouliot WA et al (2007) Connexin36 vs. connexin32, “miniature” neuronal gap junctions, and limited electrotonic coupling in rodent suprachiasmatic nucleus. Neuroscience 149:350–371 20. Nielsen HS, Hannibal J, Fahrenkrug J (2002) Vasoactive intestinal polypeptide induces per1 and per2 gene expression in the rat suprachiasmatic nucleus late at night. Eur J Neurosci 15: 570–574 21. Hughes AT (2004) Aberrant gating of photic input to the suprachiasmatic circadian pacemaker of mice lacking the VPAC2 receptor. J Neurosci 24:3522–3526 22. Reed HE, Meyer-Spasche A, Cutler DJ et al (2001) Vasoactive intestinal polypeptide (VIP) phase-shifts the rat suprachiasmatic nucleus clock in vitro. Eur J Neurosci 13:839–843 23. Watanabe K, Vanecek J, Yamaoka S (2000) In vitro entrainment of the circadian rhythm of vasopressin-releasing cells in suprachiasmatic nucleus by vasoactive intestinal polypeptide. Brain Res 877:361–366 24. Piggins H, Antle M, Rusak B (1995) Neuropeptides phase shift the mammalian circadian pacemaker. J Neurosci 15:5612–5622 25. Colwell CS, Michel S, Itri J et al (2003) Disrupted circadian rhythms in VIP- and PHI-deficient mice. Am J Physiol Integr Comp Physiol 285:R939–R949

26. Harmar AJ, Marston HM, Shen S et al (2002) The VPAC2 receptor is essential for circadian function in the mouse suprachiasmatic nuclei. Cell 109:497–508 27. Brown TM, Colwell CS, Waschek JA et al (2007) Disrupted neuronal activity rhythms in the suprachiasmatic nuclei of vasoactive intestinal polypeptide-deficient mice. J Neurophysiol 97:2553–2558 28. Hughes ATL, Guilding C, Lennox L et al (2008) Live imaging of altered period1 expression in the suprachiasmatic nuclei of Vipr2 /  mice 1. J Neurochem 106:1646–1657 29. Maywood ES, Reddy AB, Wong GKY et al (2006) Synchronization and maintenance of timekeeping in suprachiasmatic circadian clock cells by neuropeptidergic signaling. Curr Biol 16:599–605 30. Aton SJ, Colwell CS, Harmar AJ et al (2005) Vasoactive intestinal polypeptide mediates circadian rhythmicity and synchrony in mammalian clock neurons. Nat Neurosci 8:476–483 31. Finger AM, Kramer A (2021) Peripheral clocks tick independently of their master. Genes Dev 35(5–6):304–306. https://doi.org/10.1101/ gad.348305.121. PMID: 33649161; PMCID: PMC7919411 32. Nagoshi E, Saini C, Bauer C et al (2004) Circadian gene expression in individual fibroblasts: cell-autonomous and self-sustained oscillators pass time to daughter cells. Cell 119:693–705 33. Noguchi T, Ikeda M, Ohmiya Y et al (2012) A dual-color luciferase assay system reveals circadian resetting of cultured fibroblasts by co-cultured adrenal glands. PLoS One 7: e37093 34. Guo H, Brewer JM, Lehman MN et al (2006) Suprachiasmatic regulation of circadian rhythms of gene expression in hamster peripheral organs: effects of transplanting the pacemaker. J Neurosci 26:6406–6412 35. Koronowski KB, Kinouchi K, Welz P-S et al (2019) Defining the independence of the liver circadian clock. Cell 177:1448–1462.e14 36. Welz P-S, Zinna VM, Symeonidi A et al (2019) BMAL1-driven tissue clocks respond independently to light to maintain homeostasis. Cell 177:1436–1447.e12 37. Saini C, Liani A, Curie T et al (2013) Real-time recording of circadian liver gene expression in freely moving mice reveals the phase-setting behavior of hepatocyte clocks. Genes Dev 27: 1526–1536 38. Tahara Y, Kuroda H, Saito K et al (2012) In vivo monitoring of peripheral circadian clocks in the mouse. Curr Biol 22:1029–1034

Peripheral Coupling 39. Sinturel F, Gos P, Petrenko V, Hagedorn C, Kreppel F, Storch KF, Knutti D, Liani A, Weitz C, Emmenegger Y, Franken P, Bonacina L, Dibner C, Schibler U (2021) Circadian hepatocyte clocks keep synchrony in the absence of a master pacemaker in the suprachiasmatic nucleus or other extrahepatic clocks. Genes Dev 35(5–6):329–334. https://doi.org/10. 1101/gad.346460.120PMID: 33602874; PMCID: PMC7919413 . PMID: 33649161; PMCID: PMC7919411 40. Landgraf D, Achten C, Dallmann F et al (2015) Embryonic development and maternal regulation of murine circadian clock function. Chronobiol Int 32:416–427 41. Yoo S-H, Yamazaki S, Lowrey PL et al (2004) PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 101:5339–5346 42. Moore SR, Pruszka J, Vallance J et al (2014) Robust circadian rhythms in organoid cultures from PERIOD2::LUCIFERASE mouse small intestine. Dis Model Mech 7:1123–1130 43. Noguchi T, Wang LL, Welsh DK (2013) Fibroblast PER2 circadian rhythmicity depends on cell density. J Biol Rhythm 28:183–192 44. O’Neill JS, Hastings MH (2008) Increased coherence of circadian rhythms in mature fibroblast cultures. J Biol Rhythm 23:483–488 45. Guenthner CJ, Luitje ME, Pyle LA et al (2014) Circadian rhythms of PER2::LUC in individual primary mouse hepatocytes and cultures. PLoS One 9:e87573 46. Rougemont J, Naef F (2007) Dynamical signatures of cellular fluctuations and oscillator stability in peripheral circadian clocks. Mol Syst Biol 3:93 47. Finger AM, Ja¨schke S, Del Olmo M, Hurwitz R, Granada AE, Herzel H, Kramer A (2021). Intercellular coupling between peripheral circadian oscillators by TGF-β signaling. Sci Adv 7 (30):eabg5174. https://doi.org/10.1126/ sciadv.abg5174. PMID: 34301601; PMCID: PMC8302137 48. Strogatz SH (2000) From Kuramoto to Crawford: exploring the onset of synchronization in populations of coupled oscillators. Phys D Nonlinear Phenom 143:1–20

167

49. Maier B, Wendt S, Vanselow JT et al (2009) A large-scale functional RNAi screen reveals a role for CK2 in the mammalian circadian clock. Genes Dev 23:708–718 50. Maier B, Lorenzen S, Finger A-M et al (2021) Searching novel clock genes using RNAi-based screening, Methods in molecular biology. Springer, New York, pp 103–114 51. Bo¨rding T, Abdo AN, Maier B et al (2019) Generation of human CRY1 and CRY2 knockout cells using duplex CRISPR/Cas9 technology. Front Physiol 10:577 52. Korge S, Maier B, Bru¨ning F et al (2018) The non-classical nuclear import carrier Transportin 1 modulates circadian rhythms through its effect on PER1 nuclear localization. PLoS Genet 14:e1007189 53. Maier B, Lorenzen S, Finger A-M et al (2021) Circadian clocks. Springer, New York 54. Maier B, Lorenzen S, Finger AM, Herzel H, Kramer A (2021) Searching novel clock genes using RNAi-based screening. Methods Mol Biol 2130:103–114. https://doi.org/10. 1007/978-1-0716-0381-9_8. PMID: 33284439 55. Streuli CH, Meng Q-J (2019) Influence of the extracellular matrix on cell-intrinsic circadian clocks. J Cell Sci 132:jcs207498 56. Edmondson R, Broglie JJ, Adcock AF et al (2014) Three-dimensional cell culture systems and their applications in drug discovery and cell-based biosensors. Assay Drug Dev Technol 12:207–218 57. Pulivarthy SR, Tanaka N, Welsh DK et al (2007) Reciprocity between phase shifts and amplitude changes in the mammalian circadian clock. Proc Natl Acad Sci U S A 104: 20356–20361 58. Cevenini L, Calabretta MM, Lopreside A et al (2017) Bioluminescence imaging of spheroids for high-throughput longitudinal studies on 3D cell culture models. Photochem Photobiol 93:531–535 59. Pampaloni F, Ansari N, Stelzer EHK (2013) High-resolution deep imaging of live cellular spheroids with light-sheet-based fluorescence microscopy. Cell Tissue Res 352:161–177

Chapter 11 Circadian Control of Transcriptional and Metabolic Rhythms in Primary Hepatocytes Sung Kook Chun and Selma Masri Abstract Isolation of primary hepatocytes and culturing these cells ex vivo provides a powerful platform to model liver physiology in vivo. Primary hepatocytes can be cultured for several days, the circadian clock can be synchronized, and these primary cells can be utilized for functional gene regulation analysis and metabolic studies. In this chapter, we describe detailed methodology for isolation of viable primary hepatocytes, techniques for culturing these cells, methods for synchronization of the circadian clock, transfection and luciferase reporter analysis, as well as glucose production assays as a functional readout of metabolic state. Key words Primary hepatocyte isolation, Primary hepatocyte cell culture, Luciferase reporter assays, Synchronization of the circadian clock, Clock-controlled gene expression, Glucose production assays

1

Introduction The isolation and culture of primary hepatocytes provides a unique tool to study physiologically relevant questions in an ex vivo setting. The ability to manipulate primary hepatocytes and perform functional assays has placed this primary cell culture model as a superior system to in vitro immortalized or transformed cell culture, because a variety of metabolic, pharmacological, and toxicological studies can be addressed using molecular approaches [1–4]. Though the longterm viability of primary cells is a limiting factor, these hepatocytes do maintain properties ex vivo that model the physiological environment of the liver in vivo [4–6]. Additionally, the oscillation of the circadian clock is very robust in the liver and in hepatocytes. The hepatic clock is controlled by a hierarchical circadian structure as well as the tissuespecific pacemaker in the liver that drives the expression of ~15% of rhythmic transcripts [7–9]. Furthermore, the liver clock is highly entrained by additional Zeitgebers or time-givers, such as food intake [10–13]. The composition of food [14–17], the amount of food [18–20], and the timing of food intake [10, 14, 21, 22] all impinge

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_11, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

169

170

Sung Kook Chun and Selma Masri

Fig. 1 Transient transfection and luciferase reporter assays using primary mouse hepatocytes. (a) Ectopic expression of EGFP in primary hepatocytes. (b) Results of luciferase reporter assay in primary mouse hepatocytes by transient transfection. Hepatocytes were transfected with pGL4-Basic control plasmid or Bmal1-Luc reporter, with or without ectopic expression of the transcriptional repressor, Rev-erbα. Luminescence units were normalized to LacZ and labeled as RLU. Data represent the mean
SEM. *** indicates p < 0.001 by Student’s t test

on circadian gene expression in the liver. Therefore, primary hepatocytes provide an ideal cell culture model to study the core circadian clock machinery as well as the metabolic clock system. In this chapter, we provide detailed methodology on isolation of viable primary hepatocytes from mouse liver in addition to cell culture conditions that provide a healthy environment for these cells. We describe a protocol for the synchronization of the circadian clock in primary hepatocyte cells. Also, we provide a straightforward method for ectopic expression of genes or luciferase reporter systems utilizing a transient transfection approach with minimal toxicity (Fig. 1). Lastly, the major advantage of primary hepatocytes is their ability to maintain metabolic rhythms in culture. Therefore, we provide methodology for performing functional glucose production assays in synchronized primary hepatocytes (Fig. 2). Overall, this chapter details important technical steps required for isolation and culture of viable primary

Circadian Analysis in Mouse Primary Hepatocytes

171

Fig. 2 Measurement of glucose production from mouse primary hepatocytes. Hepatocytes were maintained in glucose-free collection buffer for up to 8 h and glucose concentration was measured using the hexokinase (HK) assay kit (Sigma Aldrich)

hepatocytes, as well as useful protocols for dissecting mechanistic questions related to the circadian clock and rhythmic metabolic pathways in the liver.

2

Materials Prepare all solutions using purified deionized water and analytical grade reagents. 1. Collagenase from Clostridium histolyticum. 2. Collagen from rat tail. To make collagen coating solution, dissolve 1 mg/mL of collagen in 0.1% glacial acetic acid. Store at 4  C. 3. Bovine serum albumin (BSA). 4. Perfusion buffer A: 25 mM HEPES, 115 mM NaCl, 5 mM KCl, 1 mM KH2PO4, 2.5 mM MgSO4, 0.5 mM EGTA. Adjust to pH 7.4 with 10 M NaOH, autoclave, and store at 4  C. 5. Perfusion buffer B: 25 mM HEPES, 115 mM NaCl, 5 mM KCl, 1 mM KH2PO4, 2 mM CaCl2. Adjust to pH 7.4 with 10 M NaOH, autoclave, and store at 4  C. Add 0.1 mg/mL of collagenase immediately prior to use.

172

Sung Kook Chun and Selma Masri

6. Perfusion buffer C: 25 mM HEPES, 115 mM NaCl, 5 mM KCl, 1 mM KH2PO4, 2.5 mM MgSO4, 4 mM CaCl2. Adjust to pH 7.4 with 10 M NaOH, autoclave, and store at 4  C. Add 10 mg/mL of BSA prior to use. 7. Dissection tools: forceps, surgical scissors, perforated plate with tray. Important: at least one of the forceps should have a curved and sharp tip. 8. Microvascular clamp (bulldog clamp). 9. Surgical suture. 10. Plasticware: 15 and 50 mL plastic conical tubes, 100 mm diameter petri dishes, cell culture dishes. 11. Water bath set to 42  C. For glucose production assay, water bath should be equipped with a stand for keeping culture dishes or plates above water. Also, water bath lid is required for maintaining temperature and humidity. 12. Insulin syringe. 13. 70 μm Cell strainer. 14. Sodium barbiturate (50 mg/mL) dissolved in saline. 15. 70% Ethanol. 16. Peristaltic pump. 17. Tubing for peristaltic pump. 18. Intravenous cannula with needle (20 G). 19. Intravenous administration device with drip chamber and outlet that can be assembled with cannula. 20. Refrigerated centrifuges for 50 mL conical tubes and 1.5 mL microtubes. 21. Hemocytometer. 22. 0.4% Trypan blue solution. 23. Cotton swabs. 24. Hepatocyte culture media: DMEM media with 10% fetal bovine serum (FBS) and penicillin/streptomycin. 25. Wash buffer for culture: 25 mM HEPES, 115 mM NaCl, 5 mM KCl, 1 mM KH2PO4, 2.5 mM MgSO4, 2 mM CaCl2. Adjust to pH 7.4 with 10 M NaOH, and sterilize by autoclaving. 26. 1 mM Dexamethasone in 100% ethanol. 27. Cell culture incubator set at 37  C temperature, 5% CO2. 28. Lipofectamine 3000. 29. Glucose (HK) assay. 30. Lysis buffer for luciferase assays: 25 mM Tris–HCl (pH 7.8), 2 mM EDTA, 1 mM DTT, 10% glycerol, and 1% Triton X-100.

Circadian Analysis in Mouse Primary Hepatocytes

173

31. Luciferase reaction buffer: 20 mM Tris–HCl (pH 7.8), 1 mM MgCl2, 2.5 mM MgSO4, 0.1 mM EDTA, 33.3 mM DTT, 10 mM beetle luciferin, 100 mM ATP, and 10 mM coenzyme A. 32. Z buffer for β-galactosidase assay: 60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, 50 mM β-mercaptoethanol. 33. Collection buffer for glucose production assay: Add 10 mg/ mL of BSA, 20 mM of sodium lactate, and 2 mM of sodium pyruvate to wash buffer. 34. Glucose-free media (for glucose production assay): Glucosefree DMEM F-12 media supplemented with 10% FBS and penicillin/streptomycin.

3

Methods

3.1 Isolation and Culture of Mouse Primary Hepatocytes

1. Prepare collagen-coated dishes at least a day before isolation. Spread 0.1% collagen solution evenly on the surface area of culture dishes and leave them in the biosafety cabinet for 1 h to dry. Wash the coated dishes with sterile water, aspirate water completely, and dry again. Store coated dishes in 4  C until use. 2. On the day of hepatocyte isolation, prepare perfusion buffer A, buffer B, and buffer C in autoclaved glass bottles. For a single mouse, 100 mL of each buffer is sufficient. Pre-warm buffer A and B in a 42  C water bath and keep buffer C on ice. 3. Prior to start of hepatocyte isolation, clear the tubing for perfusion. Pump out any pre-existing liquids from the tubing and fill the tubing with ddH2O. Allow water flow for 5 min. Empty the tubing again and replace it with pre-warmed perfusion buffer A. Make sure that the tubing is filled with the buffer A without any residual bubbles inside the tubing. Adjust the speed of pump to 4 mL per minute (see Note 1). 4. Add 10 mg of collagenase to 100 mL of pre-warmed perfusion buffer B. Gently shake the bottle to dissolve completely. 5. Add 1 g of BSA to 100 mL of perfusion buffer C and leave the bottle undisturbed until the powder dissolves completely. After BSA is dissolved, pour approximately 20 mL of buffer C into a petri dish on ice. Place the rest of buffer C and two empty 50 mL conical tubes on ice. 6. Anesthetize the mouse by injecting 100 mg/kg of pentobarbital solution by intraperitoneal (IP) injection. Pinch the tail or hindlimb to ensure that the mouse is fully anesthetized before proceeding.

174

Sung Kook Chun and Selma Masri

7. Sterilize the abdomen of the mouse with 70% ethanol. Open the body cavity by making a U-shaped incision extending across the length and width of the abdomen. Expose the portal vein and inferior vena cava (IVC) by carefully repositioning intestines and other organs with cotton swab. 8. Place a suture around the IVC a few millimeters posterior to the renal vein by using sharp-tipped curved forceps. Tie the suture loosely, and importantly, do not make a tight knot with the suture (see Note 2). 9. Connect the cannula to the outlet of the tubing and fill it with buffer A. Make sure both needle and cannula are completely filled with buffer and disconnect from the tubing. 10. Insert both needle and cannula into the IVC. Gently press the IVC with a cotton swab to dilate the IVC while placing cannula. Carefully retract the needle and remove it. Cannula will be filled with the blood as you remove the needle. After removing the needle, secure the cannula inside of the IVC by tightening the knot of the suture (see Note 3). 11. Carefully sever the portal vein with a scissor. This will be the exit for perfusion liquids. Do not hurt the liver or other organs during this step since it will severely affect the outcome. 12. Connect the cannula to the outlet of the tubing. Allow perfusion buffer A to flow for 5 min. The liver should change color from dark red/brown to a uniform light tan color as the blood is flushed out (see Note 4). 13. Immediately after you start perfusing the liver, open the diaphragm to expose the thoracic cavity and superior vena cava. Cut through the ribs alongside the sternum if you need better exposure to locate the superior vena cava. Clamp the superior vena cava between the heart and diaphragm with the bulldog clamp. 14. After 5 min of perfusion with buffer A, change to perfusion buffer B with collagenase. Perfuse the liver for 15–20 min. Liver should be enlarged and become porous with a spongy texture due to digestion. 15. Turn off the pump and excise the liver after perfusion is complete. Remove the gall bladder before collecting the liver. Make sure not to damage the gall bladder to avoid leaking bile as this is highly detrimental to hepatocytes. 16. Transfer the liver into the petri dish with ice-cold perfusion buffer C with 1% BSA. Gently peal the outside membrane of the liver with sharp-tipped forceps from all lobes. Gently shake the liver until most of the hepatocytes have fallen into the buffer. During shaking, volume of the liver will be reduced gradually while the buffer will become cloudy and brownish in color.

Circadian Analysis in Mouse Primary Hepatocytes

175

17. Set up the 70 μm cell strainer on a pre-chilled 50 mL conical tube. Filter the entire suspension into a petri dish through the cell strainer. Pour an additional 5 mL of buffer C into the petri dish, gently shake 2–3 times, and pour through the cell strainer again to collect remaining cells. 18. Place the hepatocytes on ice for 15–20 min. From this point on, the cell suspension should be kept on ice as much as possible and treated as sterile. Cells will precipitate to the bottom of the tube after 15 min. 19. Carefully aspirate supernatant. Be careful not to aspirate any cells in bottom of the tube. Resuspend the cell pellet with 20 mL of ice-cold buffer C and gently invert the tube. After resuspending completely, pour half of the cell suspension (about 10 mL) into a pre-chilled empty 50 mL tube. Adjust the amount of cell suspension for centrifugation. 20. Centrifuge the suspension for 1 min at 30  g at 4  C. Aspirate the supernatant in the hood. At this step, try to remove as much supernatant as you can since the supernatant contains dead cells and non-hepatocyte cells. Resuspend each pellet in 10 mL of buffer C. Centrifuge tubes for 2 min at 30  g in 4  C. Aspirate the supernatant in the hood and resuspend each pellet in 10 mL of buffer C. Centrifuge tubes for 2 min at 30  g in 4  C and aspirate the supernatant in the hood. Resuspend the pellet in 5 mL of buffer C for first tube by gently shaking, then transfer cell suspension to second tube and resuspend cells again by gently shaking. 21. For counting cells, mix cell suspension thoroughly by gently shaking and take 10 μL of cell suspension to 1.5 mL Eppendorf tube. Mix it with 40 μL of 0.4% trypan blue staining solution (5 dilution). Place the mix on a hemocytometer and count both viable and dead hepatocytes. Calculate viability and total cell counts in suspension. Viability should be over 90% for a young mouse without any pathological condition (see Note 5) Viability : ½ðLive cellsÞ=ðLive cells þ dead cellsÞ  100 ¼ %of viability 22. Seed the cells onto collagen-coated dishes or plates with culture media. Use cold media from the refrigerator or pre-chilled media on ice. Do not pre-warm the media to avoid extreme changes of temperature which could be harmful to hepatocytes. For a 35 mm culture dish, seed one million (1  106) cells in 2 mL culture media. Adjust cell numbers based on culture surface you want to use. After seeding cells, place cells inside of a cell culture incubator at 37  C with 5% CO2. Incubate for 4 h to allow hepatocytes to stabilize and attach to the bottom of the dish.

176

Sung Kook Chun and Selma Masri

23. Pre-warm the wash buffer and fresh culture media in a water bath at 37  C. After 4 h of stabilization, wash hepatocytes twice with pre-warmed wash buffer and replace with fresh culture media at 37  C. Place cells inside of cell culture incubator at 37  C and 5% CO2 overnight. 24. For synchronization of the circadian clock, wash hepatocytes thoroughly with wash buffer and treat cells with DMEM media containing 100 nM of dexamethasone for 45–60 min. Dexamethasone media should be prepared in DMEM media with antibiotics (penicillin and streptomycin) but without FBS. After synchronization, wash cells thoroughly with wash buffer and replace with fresh culture media and collect hepatocytes at desired circadian time points. 25. For transient transfection, prepare mixture of plasmid and lipofectamine 3000 according to manufacturer’s recommendation. Add this mixture to the hepatocytes for 4 h. Harvest cells for further experiments at least 24 h after transfection. 26. For luciferase reporter assays, primary hepatocytes can be transfected with desired luciferase reporter constructs and additional plasmids such as LacZ for normalization purposes. 24 h post-transfection, cells are harvested with luciferase lysis buffer. Cell lysates are mixed with luciferase reaction buffer and luminescence is measured using a luminometer. For normalization using a β-galactosidase assay, cell lysates are mixed with Z buffer and incubated at 37  C for up to 1 h until yellow coloration appears. Absorbance is measured at 450 nm and these values are used to normalize luciferase units to obtain relative light units (RLU). 3.2 De Novo Glucose Production Assay with Synchronized Mouse Primary Hepatocytes

1. Prepare collection buffer before start of experiment with the addition of 1% BSA, 20 mM of sodium lactate, 2 mM of sodium pyruvate, and penicillin/streptomycin and store at 4  C until use. 2. Isolate mouse primary hepatocytes and plate them on a 6-well plate with 1  106 cells/well. Stabilize cells for 4 h in the incubator. Wash cells thoroughly with pre-warmed wash buffer and replace media with fresh glucose-free media. Incubate cells for 24 h in incubator. 3. Wash cells thoroughly with pre-warmed wash buffer. To synchronize the circadian clock, add 100 nM dexamethasone to glucose-free DMEM F-12 media without serum for 45 min. Wash cells with pre-warmed wash buffer and replace with fresh glucose-free DMEM F-12 media with serum. Incubate cells for 12 h. 4. Wash cells thoroughly with wash buffer and replace media with collection buffer (1 mL/well). Transfer hepatocytes in culture

Circadian Analysis in Mouse Primary Hepatocytes

177

plates to a water bath with stand. Close the lid and incubate for 1–8 h in water bath, depending on desired collection times (see Note 6). 5. After desired time point, collect buffer from each well. Centrifuge at 2000  g, 4  C to remove residual cells and debris and save supernatant. Store supernatant at 20  C if needed. 6. Measure the glucose concentration from supernatant via glucose (HK) assay kit based on manufacturer’s protocol. Use unused collection buffer as a negative control.

4

Notes 1. During the perfusion, buffers will gradually lose heat and become cold, which can affect the isolation outcome. Measure the temperature of the buffer when you set up the hepatocyte isolation. The temperature of the buffer should be close to 37  C. Minimize tubing length to prevent heat loss. Also, the temperature of water bath that you use for warming the buffer can be adjusted. 2. Make sure not to cause unnecessary damages to any blood vessels or organs when you sacrifice the mouse and prepare the perfusion. Excessive bleeding before starting perfusion will make the procedure difficult and affect quality of hepatocytes. 3. Visceral fat or other connective tissues can be removed to better expose IVC or portal vein. When working with aged or obese mice, this is an essential step. However, be extremely careful not to damage any blood vessels or organs during this procedure. Causing unwanted damage to IVC will affect yield and viability of hepatocytes because of incomplete perfusion. 4. Make sure there are no bubbles inside of the tubing and cannula when you start perfusion. Generation of some bubbles may be unavoidable while changing buffers during perfusion or connecting cannula with outlet. Use drip chamber to trap the bubbles. If any bubbles are seen inside the cannula, empty it and carefully refill with fresh buffer. If bubbles enter the liver during perfusion, it will block the microvessels inside of the liver and cause incomplete perfusion, leading to poor yield and viability of hepatocytes. 5. Using viable hepatocytes is critical for downstream experiments. Hepatocyte viability of 90% is commonly accepted. Reduced viability will impact transfection efficiency of primary hepatocytes. To ensure high viability of primary hepatocytes, always use fresh perfusion buffers. Also, extreme changes of pH can cause severe damage to hepatocytes. It is strongly recommended to adjust pH of perfusion buffers.

178

Sung Kook Chun and Selma Masri

6. For de novo glucose production assays, ensure that the ambient temperature inside the water bath is 37  C. Since the collection buffer does not contain a bicarbonate-based buffering system, use of a normal incubator at 37  C and 5% CO2 may cause severe changes of pH.

Acknowledgments This work was supported by the National Institutes of Health (NIH) and the National Cancer Institute (NCI) through grants, K22CA212045 and R01CA244519 to SM. In addition, the Masri Laboratory is supported by grants through the Concern Foundation, the V Foundation for Cancer Research, the Cancer Research Coordinating Committee, and the Chao Family Comprehensive Cancer Center at the University of California, Irvine. References 1. Hirota T, Lee JW, St John PC, Sawa M, Iwaisako K, Noguchi T, Pongsawakul PY, Sonntag T, Welsh DK, Brenner DA, Doyle FJ III, Schultz PG, Kay SA (2012) Identification of small molecule activators of cryptochrome. Science 337:1094–1097 2. Krishnaiah SY, Wu G, Altman BJ, Growe J, Rhoades SD, Coldren F, Venkataraman A, Olarerin-George AO, Francey LJ, Mukherjee S, Girish S, Selby CP, Cal S, Er U, Sianati B, Sengupta A, Anafi RC, Kavakli IH, Sancar A, Baur JA, Dang CV, Hogenesch JB, Weljie AM (2017) Clock regulation of metabolites reveals coupling between transcription and metabolism. Cell Metab 25:961–974 3. Chun SK, Lee S, Flores-Toro J, Rebecca YU, Yang MJ, Go KL, Biel TG, Miney CE, Pierre Louis S, Law BK, Law ME, Thomas EM, Behrns KE, Leeuwenburgh C, Kim JS (2018) Loss of sirtuin 1 and mitofusin 2 contributes to enhanced ischemia/reperfusion injury in aged livers. Aging Cell 17:e12761 4. Godoy P, Hewitt NJ, Albrecht U, Andersen ME, Ansari N, Bhattacharya S, Bode JG, Bolleyn J, Borner C, Bo¨ttger J, Braeuning A, Budinsky RA, Burkhardt B, Cameron NR, Camussi G, Cho CS, Choi YJ, Craig Rowlands J, Dahmen U, Damm G, Dirsch O, Donato MT, Dong J, Dooley S, Drasdo D, Eakins R, Ferreira KS, Fonsato V, Fraczek J, Gebhardt R, Gibson A, Glanemann M, Goldring CE, Go´mez-Lecho´n MJ, Groothuis GM, Gustavsson L, Guyot C, Hallifax D, Hammad S, Hayward A, Ha¨ussinger D, Hellerbrand C, Hewitt P, Hoehme S,

Holzhu¨tter HG, Houston JB, Hrach J, Ito K, Jaeschke H, Keitel V, Kelm JM, Kevin Park B, Kordes C, Kullak-Ublick GA, LeCluyse EL, Lu P, Luebke-Wheeler J, Lutz A, Maltman DJ, Matz-Soja M, McMullen P, Merfort I, Messner S, Meyer C, Mwinyi J, Naisbitt DJ, Nussler AK, Olinga P, Pampaloni F, Pi J, Pluta L, Przyborski SA, Ramachandran A, Rogiers V, Rowe C, Schelcher C, Schmich K, Schwarz M, Singh B, Stelzer EH, Stieger B, Sto¨ber R, Sugiyama Y, Tetta C, Thasler WE, Vanhaecke T, Vinken M, Weiss TS, Widera A, Woods CG, Xu JJ, Yarborough KM, Hengstler JG (2013) Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and non-parenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Arch Toxicol 87: 1315–1530 5. Bell CC, Hendriks DF, Moro SM, Ellis E, Walsh J, Renblom A, Fredriksson Puigvert L, Dankers AC, Jacobs F, Snoeys J, Sison-Young RL, Jenkins RE, Nordling A˚, Mkrtchian S, Park BK, Kitteringham NR, Goldring CE, Lauschke VM, Ingelman-Sundberg M (2016) Characterization of primary human hepatocyte spheroids as a model system for drug-induced liver injury, liver function and disease. Sci Rep 6:25187 6. Dominy JE Jr, Lee Y, Jedrychowski MP, Chim H, Jurczak MJ, Camporez JP, Ruan HB, Feldman J, Pierce K, Mostoslavsky R, Denu JM, Clish CB, Yang X, Shulman GI, Gygi SP, Puigserver P (2012) The deacetylase

Circadian Analysis in Mouse Primary Hepatocytes Sirt6 activates the acetyltransferase GCN5 and suppresses hepatic gluconeogenesis. Mol Cell 48:900–913 7. Akhtar RA, Reddy AB, Maywood ES, Clayton JD, King VM, Smith AG, Gant TW, Hastings MH, Kyriacou CP (2002) Circadian cycling of the mouse liver transcriptome, as revealed by cDNA microarray, is driven by the suprachiasmatic nucleus. Curr Biol 12:540–550 8. Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, Schultz PG, Kay SA, Takahashi JS, Hogenesch JB (2002) Coordinated transcription of key pathways in the mouse by the circadian clock. Cell 109:307– 320 9. Koike N, Yoo SH, Huang HC, Kumar V, Lee C, Kim TK, Takahashi JS (2012) Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science 338:349–354 10. Damiola F, Le Minh N, Preitner N, Kornmann B, Fleury-Olela F, Schibler U (2000) Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev 14:2950–2961 11. Vollmers C, Gill S, DiTacchio L, Pulivarthy SR, Le HD, Panda S (2009) Time of feeding and the intrinsic circadian clock drive rhythms in hepatic gene expression. Proc Natl Acad Sci U S A 106:21453–21458 12. Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabolism, and the circadian clock. Cell 161:84–92 13. Stokkan KA, Yamazaki S, Tei H, Sakaki Y, Menaker M (2001) Entrainment of the circadian clock in the liver by feeding. Science 291: 490–493 14. Hatori M, Vollmers C, Zarrinpar A, DiTacchio L, Bushong EA, Gill S, Leblanc M, Chaix A, Joens M, Fitzpatrick JA, Ellisman MH, Panda S (2012) Time-restricted feeding

179

without reducing caloric intake prevents metabolic diseases in mice fed a high-fat diet. Cell Metab 15:848–860 15. Kohsaka A, Laposky AD, Ramsey KM, Estrada C, Joshu C, Kobayashi Y, Turek FW, Bass J (2007) High-fat diet disrupts behavioral and molecular circadian rhythms in mice. Cell Metab 6:414–421 16. Eckel-Mahan KL, Patel VR, de Mateo S, Orozco-Solis R, Ceglia NJ, Sahar S, DilagPenilla SA, Dyar KA, Baldi P, Sassone-Corsi P (2013) Reprogramming of the circadian clock by nutritional challenge. Cell 155:1464–1478 17. Tognini P, Murakami M, Liu Y, Eckel-Mahan KL, Newman JC, Verdin E, Baldi P, SassoneCorsi P (2017) Distinct circadian signatures in liver and gut clocks revealed by ketogenic diet. Cell Metab 26:523–538 18. Sato S, Solanas G, Peixoto FO, Bee L, Symeonidi A, Schmidt MS, Brenner C, Masri S, Benitah SA, Sassone-Corsi P (2017) Circadian reprogramming in the liver identifies metabolic pathways of aging. Cell 170:664– 677 19. Kinouchi K, Magnan C, Ceglia N, Liu Y, Cervantes M, Pastore N, Huynh T, Ballabio A, Baldi P, Masri S, Sassone-Corsi P (2018) Fasting imparts a switch to alternative daily pathways in liver and muscle. Cell Rep 25: 3299–3314 20. Patel SA, Velingkaar N, Makwana K, Chaudhari A, Kondratov R (2016) Calorie restriction regulates circadian clock gene expression through BMAL1 dependent and independent mechanisms. Sci Rep 6:25970 21. Chaix A, Zarrinpar A, Miu P, Panda S (2014) Time-restricted feeding is a preventative and therapeutic intervention against diverse nutritional challenges. Cell Metab 20:991–1005 22. Chaix A, Lin T, Le HD, Chang MW, Panda S (2019) Time-restricted feeding prevents obesity and metabolic syndrome in mice lacking a circadian clock. Cell Metab 29:303–319

Chapter 12 Electrophysiology of the Suprachiasmatic Nucleus: Single-Unit Recording Martha U. Gillette and Jennifer W. Mitchell Abstract Oscillatory output from the suprachiasmatic nuclei (SCN) of the hypothalamus communicates time-of-day information to the brain and body. The SCN’s intrinsic ~24-h rhythm can be measured in the neuronal firing rate both in vivo and in vitro, where it continues unperturbed. This robust reporter of endogenous physiology in the SCN brain slice can be widely used to study dynamic changes in SCN physiology, its changing sensitivity to phase-altering signals, and underlying mechanisms. To provide relevant and reproducible data, care must be taken to ensure health of the SCN brain slice. The methods detailed here have been proven to produce healthy, long-lived brain slices. Key words Suprachiasmatic nucleus, Circadian rhythm, Oscillator, Phase, Electrophysiology, Single unit

1

General Introduction A variety of cellular and molecular methods have been applied to study circadian rhythms in a range of cell types and tissues in a phylogenetically diverse array of plants, fungi, and animals. For the central circadian clock in the brain, spontaneous electrical activity is a faithful reporter of rhythmicity. The ability to keep 24-h time is expressed directly as an oscillation in neuronal firing rate within the suprachiasmatic nuclei (SCN), which comprise the primary circadian pacemaker in mammals. This oscillation is an endogenous property of SCN neurons that survive within an in vitro brain slice preparation. The SCN in these hypothalamic brain slice preparations continue unperturbed in their circadian oscillation with a subsequent peak in single-unit activity that occurs at a circadian time (CT) similar to that observed in rats in vivo implanted with multiunit electrodes [1]. Additionally, the phase of the oscillation matches the expected phase based on the lighting schedule from the donor colony from which the animal was obtained.

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_12, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

181

182

Martha U. Gillette and Jennifer W. Mitchell

1.1 Electrophysiology of SCN

2

Over the course of each daily cycle, SCN undergoes rhythmic changes at all levels of organization: molecular and biochemical states [2], neuropeptide and neurotransmitter composition and release [3, 4], cell-signaling pathways [5], metabolic states [6], and neuronal electrical activity [1, 7, 8]. The robust circadian rhythm in electrical activity persists in vivo in a slice of hypothalamus bearing the SCN maintained in minimal salt and glucose [9– 13]. The sign and shape of the phase-response relationship for resetting the phase of the isolated oscillator are very similar to that for intact animals [13, 14] except that, in the absence of feedback from other brain regions, the isolated SCN oscillator undergoes large amplitude shifts in phase during the cycle after the phase-shifting stimulation. This demonstrates that underlying oscillator mechanisms, including phase readjustments, are endogenous to the SCN and can be probed in the brain slice. Accessibility en slice offers the opportunity to examine directly what imbues the SCN tissue with coherent pacemaker properties—a conundrum that remains to be solved. Protocols for preparing the hypothalamic brain slice (s) containing the SCN and for studying single-unit recording of SCN electrophysiology are detailed here. Understand that the challenge is to not subject naked, unsupported brain tissue to any tugging, bending, or shear stress—in other words, conditions not present in situ. You have less than 5 min between the time the head is severed from the vascular system and a
500 μm slice of brain is placed in the “life-support system” of the brain slice chamber before hypoxia causes irreversible neuronal cell death/brain damage. Achieving this critical time will require practice, but is necessary to study the healthy living brain in vitro.

Materials 1. The brain slice chamber is a design developed by Glenn Hatton at Michigan State to provide support for the very soft hypothalamic tissue and also a high rate of perfusion of the brain slices (61). Our chamber is custom built in the Life Sciences Machine Shop at the University of Illinois containing a notched ring to secure a nylon mesh support (see Fig. 1). Requirements of this Plexiglas chamber are described herein. 2. 70% Ethanol. 3. Deionized water. 4. For hypothalamic brain slice electrophysiology, we use a standard electrophysiology rig. This includes a Faraday cage to shield electrical noise and a stainless-steel base, a micromanipulator, a Narishige MO-8 hydraulic Microdrive, and a fiber optics light source. The setup also requires an oscilloscope,

Recording Method to Study SCN

183

Fig. 1 A hypothalamic brain slice chamber containing three SCN brain slices within the inner chamber. The brain slices are laying on a membrane in the inner well, where the media rises around and bathes the slices. In this example, the modified Earle’s Balanced Salt Solution (EBSS) contains phenol red, a pH indicator. Visible to the upper left is the inner-chamber lid with glass microscope slide resting over the inner hole used for accessing brain slices with microelectrodes. Droplets of condensate from bubbles due to aeration of the warm water bath are apparent beneath the lid of the outer well of the chamber. The silver tubes at the top are part of the heating system for the water bath. Inset: A fresh hypothalamic brain slice from rat with paired SCN nestled ventrally in the optic chiasm. The open third ventricle appears medially in the brain slice [12]

small computer, water bath large enough to accommodate a 500 ml flask of Earle’s Balanced Salt Solution (EBSS) media, dual peristaltic pumps to drive inflow and outflow, and a water bath that provides constant-temperature recycled water pumped through steel coils to heat the outer chamber of the brain slice dish, and an aquarium airstone (air diffuser) attached via an air pipe to a tank of condensed 95% O2:5% CO2 gas through a regulator that enables fine control of gas flow. 5. Dental wax. 6. Parafilm. 7. Nylon mesh or microfilament cloth. 8. Coffee filter paper. 9. Microscope slide. 10. Subjects are typically 2–5-month-old Long-Evans rats or C57BL/6J mice bred and reared in our laboratory to reduce individual differences. Animals are housed on a 12 h:12 h lightdark cycle with food and water available ad libitum. They are handled at random intervals to reduce handling stress.

184

Martha U. Gillette and Jennifer W. Mitchell

11. Modified Earle’s Balanced Salt Solution (EBSS) media: 1 mM CaCl2, 0.8 mM MgSO4, 5 mM KCl, 117 mM NaCl, 1 mM NaH2PO4, 6 mM Glucose, 0.028 mM Phenol Red supplemented with NaHCO3 (26.2 mM) and glucose (19.0 mM), pH 7.2 adjusted with HCl and saturated with 95% O2:5% CO2. 12. 11 N Hydrochloric acid. 13. 95% O2:5% CO2 gas tank. 14. Pasteur Pipets. 15. Whatman Filter Paper. 16. Rongeurs. 17. Dissecting scissors. 18. 35 mm Petri dishes. 19. Iridectomy scissors. 20. 95% Ethanol. 21. Single-edge razor. 22. Camel-hair brush (size #1). 23. A chopper fitted with a straight-edge razor and a mechanical micrometer head was fabricated in the Life Sciences Machine Shop at the University of Illinois (see Fig. 2). 24. 5 M Sodium chloride. 25. Electrodes are pulled from glass capillary (1.0 mm  0.5 mm, 4 in.) using a Flaming/Brown micropipette puller, and backfilled via syringe fitted with a nonmetalic syringe.

3

Methods

3.1 Setting Up the Hypothalamic Brain Slice Chamber

Steps #1–13 can be done a day prior to preparing the brain slices.

3.1.1 Chamber Setup

2. Rinse the chamber with 70% ethanol, then several times with deionized water.

1. Select the correct chamber size (small or large) for your experiment. Check that the lid fits snugly.

3. Mount the chamber on a round base on the floor of the Faraday cage where electrophysiological recordings will take place. Make sure that the chamber is very stable. Add dental wax to stabilize, as necessary. 4. Insert a coil that will enable warm water to heat the water bath within the chamber. Position within indentations in the rim of the dish. Use dental wax to seal tightly. 5. Insert an air pipe and seal the passageway into the chamber with wax.

Recording Method to Study SCN

185

Fig. 2 Blocking the brain to prepare hypothalamic SCN brain slices. (a) The locations of caudal and rostral cuts (in black dashed lines) and sagittal cuts (blue dashed lines) recommended for blocking the brain for sectioning in the coronal plane are indicated on the ventral-side up brain. (b) Blocked rat brain. (c) Placement of the block, ventral-side down for obtaining hypothalamic SCN brain slices. (d) Sequential 500 μm section isolated from the block. The center image contains the SCN demarcated with a red box. C caudal, R rostral

6. Attach the airstone to the air pipe and secure it in the heating coil, so that it will be submerged when water is added. 7. Wash ends of outflow and inflow tubing with 70% ethanol and insert outflow tubing with siphon. Seal with parafilm. 8. Insert inflow tubing and seal onto the chamber with parafilm. 9. Fill the outer chamber with deionized water until the airstone is submerged. 10. Seal the top into the chamber with dental wax. You can seal further with an extra layer of parafilm, if necessary. 11. Clean a notched plastic ring with 70% ethanol and then deionized water. Cut a ring of nylon mesh a little bigger than the plastic ring.

186

Martha U. Gillette and Jennifer W. Mitchell

12. Place the mesh in the center chamber, with notched ring on top, so that the notch straddles the outflow port. This mesh forms the webbed platform on which the brain slices will rest. They will be exposed to the continuous flow of media rising from beneath. Use a small piece of coffee filter (3 mm  15 mm) as a wick through the outflow port. 13. Place the chamber lid on top, and cover over the opening above the actual slice chamber with a microscope slide that has been cleaned with 70% ethanol. 3.1.2 On the Day of Slicing

14. Start heating the circulator, medium water bath, to 37  C for rats and 35  C for mice, and oxygenate the chamber at least 1 h before slicing. 15. Prepare modified EBSS medium, adjust with drops of 11 N HCl to pH 7.25 to be in the appropriate physiological range for the brain slice, and aerate immediately, allowing it to warm and oxygenate at least 1 h before beginning the media pumps. 16. Start pumps. (Optional: Fill inner well of chamber using a Pasteur pipet and warmed media, to save time.) 17. Wet Whatman filter paper on the chopper with modified EBSS medium and smooth air bubbles out from underneath with Pasteur pipet.

3.2 Preparing the Hypothalamic Brain Slice

1. Maintain each animal donating a brain slice under its entrained lighting condition until initiating the brain slice procedure (see Note 1). At that moment, gently and casually take the donor from the cage, into laboratory lighting, and decapitate by guillotine. 2. To remove the brain for slicing, carefully peel back the dorsal skull bones using Rongeurs, starting at the base of the skull where the spinal cord enters. This procedure exposes the dorsal surface of the brain. Hold the head upside-down so the brain falls away slightly from the rest of the skull. You will be able to see the still-attached myelinated optic nerves. Using a dissecting scissors, carefully sever the optic nerves close to where they emerge from the rostral inner skull, taking care not to stretch them. Now, the brain should be loose in the skull. 3. Gently roll the brain out onto a rectangle of filter paper wetted with modified EBSS in a 35 mm Petri dish. This prevents the brain from sliding around during dissection. Position the brain ventral-side up. The hypothalamus will be upper-most (see Fig. 2). Gently rinse the brain by expelling modified EBSS from a Pasteur pipette to keep it moist. 4. Using the tips of a sharp iridectomy scissors, snip through the surface meningeal membrane on the four sides of the hypothalamus. This will enable you to slice through the brain to block the hypothalamic region without tugging the brain tissue.

Recording Method to Study SCN

187

Then, lay this blocked tissue on its side and block the dorsalventral aspect taking care not to cut into the third ventricle. For the blocking cuts, use a fresh, 95% ethanol-cleaned singleedged razor blade. 5. With a camel-hair paint brush cleaned previously with 70% ethanol and dried, gently move the blocked hypothalamusbearing piece of brain to the stage of the nearby mechanical chopper. For coronal slices, position it with long axis perpendicular to the cutting blade in the tissue chopper with the meningeal side down. We cut caudal to rostral. 6. To slice the hypothalamus, wet the slicing blade and the brain surface to be cut with modified EBSS using the camel-hair brush. Repeat with each slicing action. This is necessary to keep the tissue moist and to minimize sticking on the blade and pulling on the tissue. Drop the chopper arm from only a couple inches above the platform bearing the block of brain. Dropping it from too great a height causes chattering/vibration of the tissue against the blade, which negatively impacts the health of the slice of brain tissue. 7. As each slice is cut, very carefully roll it onto the moist camelhair brush and transfer it into a Petri dish containing fresh modified EBSS that has been oxygenated by bubbling for a few minutes with 95% O2:5% CO2. Slices will remain here, slightly submerged until transfer into the brain slice chamber. Advance the stage micrometer by 500 μm and prepare to make the next slice by wetting with modified EBSS-dipped camelhair brush. Location with respect to the rostral-caudal SCN can be deduced by the landmark-shape of the optic chiasm (Fig. 2). Only one or two of the 500 μm slices will contain SCN, which is football-shaped with a maximum length ¼ 800 μm. 8. Hypothalamic brain slices should be transferred immediately to the prepared brain slice chamber and allowed to recover for 30–40 min from the spreading depression of excitability induced by K+ release during the slicing procedure (see Note 2). 9. Clean up the dissection/slice area during this period so that it is ready for future use. 3.3 Extracellular Recording of SingleUnit Activity

1. Two to four hypothalamic brain slices containing the SCN are allowed to equilibrate for 1 h in a Hatton-style brain slice chamber (Fig. 1) before commencing electrical recording (see Note 3). 2. Continuously perfuse slices with modified EBSS. Media should be as high as possible around the slices resting on the central membrane at the interface with the moist 95% O2:5% CO2 atmosphere. Under these conditions, active SCN neurons can be recorded for 24–36 h (see Note 4).

188

Martha U. Gillette and Jennifer W. Mitchell

3. Extracellular electrical activity of single neurons is detected with glass microelectrodes pulled to a tip-diameter of 5 μm in a Flaming/Brown micropipette puller, and backfilled via syringe with filtered 5 M NaCl (see Note 5). 4. Each electrode is slowly advanced using a Narishige MO-8 hydraulic microdrive until a large stable electrical deflection (an action potential or spike ¼ a “unit”) is encountered. Individual units are isolated with a WPI 121 window discriminator and activity allowed to stabilize for at least 2 min before counting the frequency of the unit. Our criterion for adequate differentiation of a unit for counting is a signal:noise ratio of >2: 1. 5. The pattern and rate of action potential firing of electrically isolated, individual neurons together with the location and depth within the SCN and circadian time (CT) are noted. 6. Data thus generated from randomly encountered neurons from a single SCN in a single hypothalamic slice are analyzed by a small computer. They are plotted against circadian time of the donor (CT, where the clock starts at “lights on” and continues for 24 h). Data is meaned by 2-h bins and smoothed in order to determine the time of peak neuronal electrical activity. 7. When the effect of time of slice preparation from animals housed on a reversed lighting cycle for a minimum of 2 weeks is compared with that of animals on daytime lights-on, the results for a given time are completely overlapping. Thus, the time of peak electrical activity is not dependent on the environmental light-dark cycle. Rather, it intrinsic to the SCN.

4

Notes 1. To ensure research accuracy and integrity, record information about experimental procedure and date of birth, sex, age, and lighting schedule of the animal. Also, record concentrations of solutions used and time and duration of treatment. Trends or patterns in data, initially inexplicable, may be apparent when looking back in the data, if detailed records are kept. 2. Make a drawing of the slice and its position in the dish. With each cell, note which part of the SCN you are recording from. 3. Let the slices recover 1 h from a treatment before you start recording. 4. Don’t forget to put the ground wire into the recording chamber media. 5. Take care to keep syringe needles that are used for backfilling the electrodes from plugging with high salt solutions. As you finish backfilling, pull back on the syringe so that no salt

Recording Method to Study SCN

189

remains in the needle; this will ensure that precipitated salt crystals to not block the needle. At the end of an experiment, rinse the needle with distilled water and store it dry.

Acknowledgments Preparation of this chapter was facilitated by NIH and NSF grants STC CBET-0939511, R61HL159948, and R21MH117277 (MUG). We thank Rhanor Gillette for contributions to the development of methods and this manuscript. References 1. Inouye ST, Kawamura H (1979) Persistence of circadian rhythmicity in a mammalian hypothalamic “island” containing the suprachiasmatic nucleus. Proc Natl Acad Sci U S A 76(11): 5962–5966. https://doi.org/10.1073/pnas. 76.11.5962 2. Hastings MH, Maywood ES, Brancaccio M (2018) Generation of circadian rhythms in the suprachiasmatic nucleus. Nat Rev Neurosci 19(8):453–469. https://doi.org/10.1038/ s41583-018-0026-z 3. Atkins N Jr, Ren S, Hatcher N, Burgoon PW, Mitchell JW, Sweedler JV, Gillette MU (2018) Functional peptidomics: stimulus- and timeof-day-specific peptide release in the mammalian circadian clock. ACS Chem Neurosci 9(8): 2001–2008. https://doi.org/10.1021/ acschemneuro.8b00089 4. Southey BR, Lee JE, Zamdborg L, Atkins N Jr, Mitchell JW, Li M, Gillette MU, Kelleher NL, Sweedler JV (2014) Comparing label-free quantitative peptidomics approaches to characterize diurnal variation of peptides in the rat suprachiasmatic nucleus. Anal Chem 86(1): 4 4 3 – 4 5 2 . h t t p s : // d o i . o r g / 1 0 . 1 0 2 1 / ac4023378 5. Gillette MU, Mitchell JW (2002) Signaling in the suprachiasmatic nucleus: selectively responsive and integrative. Cell Tissue Res 309(1): 99–107. https://doi.org/10.1007/s00441002-0576-1 6. Wang TA, Yu YV, Govindaiah G, Ye X, Artinian L, Coleman TP, Sweedler JV, Cox CL, Gillette MU (2012) Circadian rhythm of redox state regulates excitability in suprachiasmatic nucleus neurons. Science 337(6096): 839–842. https://doi.org/10.1126/science. 1222826 7. Colwell CS (2011) Linking neural activity and molecular oscillations in the SCN. Nat Rev

Neurosci 12(10):553–569. https://doi.org/ 10.1038/nrn3086 8. Mei L, Zhan C, Zhang EE (2018) In vivo monitoring of circadian clock gene expression in the mouse suprachiasmatic nucleus using fluorescence reporters. J Vis Exp (137):5676. https://doi.org/10.3791/56765 9. Gillette MU (1986) The suprachiasmatic nuclei: circadian phase-shifts induced at the time of hypothalamic slice preparation are preserved in vitro. Brain Res 379(1):176–181. https://doi.org/10.1016/0006-8993(86) 90273-8 10. Green DJ, Gillette R (1982) Circadian rhythm of firing rate recorded from single cells in the rat suprachiasmatic brain slice. Brain Res 245(1):198–200. https://doi.org/10.1016/ 0006-8993(82)90361-4 11. Groos G, Hendriks J (1982) Circadian rhythms in electrical discharge of rat suprachiasmatic neurones recorded in vitro. Neurosci Lett 34(3):283–288. https://doi.org/10.1016/ 0304-3940(82)90189-6 12. Hatton GI, Doran AD, Salm AK, Tweedle CD (1980) Brain slice preparation: hypothalamus. Brain Res Bull 5(4):405–414. https://doi. org/10.1016/s0361-9230(80)80010-4 13. Mei L, Fan Y, Lv X, Welsh DK, Zhan C, Zhang EE (2018) Long-term in vivo recording of circadian rhythms in brains of freely moving mice. Proc Natl Acad Sci U S A 115(16): 4276–4281. https://doi.org/10.1073/pnas. 1717735115 14. Mitchell JW, Atkins N Jr, Sweedler JV, Gillette MU (2011) Direct cellular peptidomics of hypothalamic neurons. Front Neuroendocrinol 32(4):377–386. https://doi.org/10.1016/j. yfrne.2011.02.005

Chapter 13 Anatomical Methods to Study the Suprachiasmatic Nucleus Eric L. Bittman Abstract The mammalian suprachiasmatic nucleus (SCN) functions as a master circadian pacemaker. In order to examine mechanisms by which it keeps time, entrains to periodic environmental signals (zeitgebers), and regulates subordinate oscillators elsewhere in the brain and in the periphery, a variety of molecular methods have been applied. Multiple label immunocytochemistry and in situ hybridization provide anatomical insights that complement physiological approaches (such as ex vivo electrophysiology and luminometry) widely used to study the SCN. The anatomical methods require interpretation of data gathered from groups of individual animals sacrificed at different time points. This imposes constraints on the design of the experiments that aim to observe changes that occur with circadian phase in free-running conditions. It is essential in such experiments to account for differences in the periods of the subjects. Nevertheless, it is possible to resolve intracellular colocalization and regional expression of functionally important transcripts and/or their peptide products that serve as neuromodulators or neurotransmitters. Armed with these tools and others, understanding of the mechanisms by which the hypothalamic pacemaker regulates circadian function is progressing apace. Key words Suprachiasmatic nucleus, Circadian rhythm, In situ hybridization, Immunocytochemistry

1

Introduction A variety of molecular methods have been applied to study circadian rhythms in a range of cell types and tissues in a phylogenetically diverse array of plants, fungi, and animals. Particular focus on the suprachiasmatic nucleus (SCN) is justified by its unique role as a master pacemaker in mammals. This emphasis was initially based on evidence gathered in a few rodent and primate species that destruction of the SCN eliminates free-running rhythms [1, 2]. However, such experiments don’t eliminate the possibility that the SCN could mediate the output, or otherwise play a permissive role in the function of a pacemaker localized elsewhere. More powerfully, subsequent experiments showed that arrhythmicity induced by SCN lesions can be reversed by grafts of the SCN, but not other

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_13, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

191

192

Eric L. Bittman

structures [3]. Most compelling is the demonstration, made possible by the availability of suitable mutants, that the period of rhythms restored by SCN transplants to animals rendered arrhythmic by either surgical [4] or genetic [5] lesions matches that of the donor. This has been achieved only in mice and hamsters, and only for behavioral rhythms. Immunocytochemistry and in situ hybridization provide insights into SCN heterogeneity and organization and thus complement physiological approaches. I refer the reader to previously published articles describing aspects and further modifications of anatomical and physiological techniques useful in elucidation of SCN function [6–15]. Not only have multiple peptidergic cell types been identified in the SCN, but inputs of classical neurotransmitters and additional neuropeptides from the retina, intergeniculate leaflet, and midbrain influence its function [16–19]. The importance of many of these peptides in circadian function is unclear. They may merely mark functional populations, they may play a physiological role, possibly as neuromodulators of small classical transmitters (principally GABA; [20]) within the SCN, or they may act as efferent signals to drive rhythmicity or entrain subordinate circadian oscillators elsewhere in the brain [21, 22]. Whatever molecular techniques are used to study the SCN, it is essential to evaluate not just effects of light exposure or changes with time of day, but to examine circadian function. The usefulness of experiments performed only in a light:dark cycle, or including only a few sampling times per cycle, is limited. Endogenous oscillations retain the capacity to free run when animals are maintained in appropriate constant conditions. These considerations have several implications for experimental design: l

In order to establish the circadian basis of diurnal variations in SCN function, animals may be released into constant conditions and sampled over the second or third free-running cycle (e.g., 24–72 h after the last LD cycle; Fig. 1). It is preferable to record an independent measure of oscillator phase, e.g., by monitoring locomotor activity, body temperature, or some other measure, so that the dependent variable can be expressed relative to circadian time. If this is not possible, zeitgeber time may be assigned by referencing the prior LD cycle, keeping in mind that the problem of phase dispersion gets worse as survival time after release into DD increases.

l

It is particularly important to sample with sufficient frequency (number of points per cycle) to identify peak and nadir phase with reasonable accuracy. While it is difficult to estimate period with any precision in such experiments, sampling should continue for at least one complete cycle. Assessment of period and phase is easier to accomplish in physiological preparations, in which a brain slice can be followed continuously and sampled at

Anatomy of SCN

193

Fig. 1 Design for experiments to examine changes in peptide or mRNA distribution in the SCN. (a) Collection of tissue during exposure to a light:dark cycle is inadequate to document circadian fluctuations. It is necessary to release animals into constant darkness and to sacrifice animals on at least the second free-running cycle (arrows). Note that samples must be taken at sufficiently short intervals to resolve phase well enough to obtain an accurate approximate estimate of the timing of peaks and nadirs of the oscillation. (b) It is necessary to record an output of the oscillator in order to express results with respect to circadian time. This is especially important when comparing effects of treatments that might alter phase or period. Double plotted actograms illustrate 2 days of locomotor activity in a 12L:12D cycle followed by 3 days of constant darkness. Wild type, Cry1- and Cry2-deficient mice have free-running periods of 23.9 h, 22.2 h, and 24.4 h, respectively. If all are sacrificed as depicted in a but results are plotted with respect to zeitgeber time, circadian phase at these sampling times will differ between genotypes and thus effects of genetic deletion on peptide or mRNA content may be misinterpreted. Note that the difference of phase between genotypes increases with each succeeding cycle after release into DD

194

Eric L. Bittman

frequent intervals, than in anatomical studies in which each animal contributes only a single data point and circadian properties must be inferred from grouped data. When groups of animals are studied under free-running conditions, differences in period can provide the impression of damping or arrhythmicity even though robust rhythms may persist in each individual. Rhythmicity of the group data fades as phase coherence between rhythmic individuals is lost. l

If anatomical studies are to be quantitative, a sufficient number of subjects must be examined at each time point to allow meaningful statistical analysis. Non-parametric methods should be used when the group size is insufficient to evaluate normality of distribution. Unfortunately, this has not always been the case even in some widely cited studies [23], and many papers inappropriately use parametric statistical methods for analysis when normality of distribution cannot be assessed due to insufficient size of the groups sacrificed at each circadian phase. Use of non-parametric methods, such as the Kruskal-Wallis test, is often more appropriate.

This chapter describes two anatomical approaches to study of the SCN. Both in situ hybridization to detect RNA transcripts of genes that are functionally relevant to the operation of the circadian pacemaker and/or its output, and immunocytochemistry to detect the protein products of these genes, are useful tools. In situ hybridization allows localization of the immediate response to entraining stimuli without the complication of background staining in neuropil. It may be particularly useful when antibodies to peptides of interest are not sufficiently specific or are inadequate for other reasons. Immunocytochemistry allows visualization of peptides that more directly contribute to function. There are multiple examples of rhythmic transcripts whose corresponding peptide products do not oscillate, and vice versa. This may be due at least in part to post-transcriptional regulation, including the function of microRNAs and post-translational modifications. Experimenters should consider both of these methods, and perhaps use them in combination, to answer their questions about circadian mechanisms. 1.1 Triple Label Immunocytochemistry

Resolution of the peptidergic organization of the SCN can be readily accomplished through use of antibodies directed at several of the peptides that mark cell types, various products of the core clock genes that participate in the core transcriptional-translational feedback loops, or markers of cell activation. Typically this includes staining for VIP, AVP, or GRP; BMAL1, PER1 or PER2, CRY1 or CRY2, NPAS2 or CLOCK; or c-FOS, c-JUN, or pERK, respectively. The latter category has recently expanded to include use of genetically encoded Calcium indicators (principally GCaMP), sometimes utilizing Cre-Lox approaches and FLEX switches to

Anatomy of SCN

195

target particular cell types and optical fibers or gradient-index (GRIN) lenses that allow anatomical resolution through endoscopy [24]. These methods allow continuous monitoring in freely moving animals but are beyond the scope of this review. There are many commercial antibodies that may be purchased to detect each of these products, but their specificity and sensitivity varies considerably. Table 1 contains a partial list of antibodies that we and others have found useful in such experiments. When staining for multiple peptides in order to co-localize or assess anatomical relationships, it should be kept in mind that the requirements for fixation or denaturation may vary between antibodies so that not all combinations are practicable; the experimenter must optimize procedures for each individual stain and then work to reconcile protocols. Labeling for multiple peptides is easiest when antibodies against the target peptides are raised in different species (see Fig. 2). For example, a mouse monoclonal antibody for c-FOS may be used in conjunction with a rabbit antiVIP and a guinea pig anti-Bmal1. This allows cellular colocalization through use of anti-mouse, anti-rabbit, and anti-guinea pig second antibodies conjugated to fluorophores that emit at different wavelengths. It is possible, however, to use two antibodies against different proteins that are raised in the same species if tyramide signal amplification (TSA) is applied [38–40]. This is sometimes necessary as most primary antibodies are raised in rabbits. In this case, staining for the antigen that is recognized by the higher dilution antibody is completed first with the use of TSA, followed by detection of the second peptide using the less dilute antibody [41]. Cellular colocalization is facilitated if the peptide targets are localized to different cellular compartments. For example, core clock proteins are mostly (but not exclusively) found in the cell nucleus, as are c-FOS and c-JUN, while phenotypic markers (AVP, VIP, GRP, etc.) are cytoplasmic. Practical considerations arise from (1) the need to sample with sufficient frequency to resolve phase-specific expression, and (2) the importance of including in each run sections from multiple animals that are collected at each of several zeitgeber or circadian times. Although double or triple staining may be necessary to gain detailed insight into phase- and cellular phenotype-specific events, it can lead to mechanical disruption of even well-fixed sections. In order to minimize variability within the experiment, and to reduce deterioration of the condition of the sections, tissue can be loaded into a multi-well rack (Fig. 3) that can be moved between solutions with minimal handling. This is especially practical for denaturation (where necessary) and buffer rinses, although the expense and limited quantities of primary and secondary antibodies requires that brain sections be incubated in lower volumes rather than in a common bath. The problem of tissue disruption may be further minimized by combining multiple primary antibodies in a cocktail

196

Eric L. Bittman

Table 1 Antibodies used for immunostaining in SCNa

a

Peptide

Raised in

Supplier

Catalog #

Reference

AVP

Rabbit

Immunostar

20069

[25]

AVP

Guinea pig

Peninsula

T-5048

[25]

AVP-neurophysin

Mouse

Hal Gainer

PS45

[26]

AVP

Mouse

Santa Cruz

Sc-390723

[27]

AVP

Rabbit

Millipore

AB1565

[28]

VIP

Rabbit

Immunostar

20077

[25]

VIP

Sheep

Millipore

AB1581

[29]

c-FOS

Rabbit

Santa Cruz

Sc-52-G

[30]

p-ERK1/2

Rabbit

Cell Signaling Technology

9101L

[31]

GRP

Rabbit

Immunostar

20073

[32]

Substance P

Mouse Rabbit

Santa Cruz Millipore

Sc-25347

[33]

CCK

Rabbit

Sigma

C258I

[34]

Calretinin

Goat

Millipore

1550

[24]

Calbindin

Mouse

Sigma

CB-955

[32, 35]

Enkephalin

Rabbit

Immunostar

20065

[25]

Somatostatin

Rabbit

PER1

Goat Rabbit

Bethyl Millipore

AB2201

[31] [27]

PER2

Rabbit

Millipore

AB2202

[27]

BMAL1

Guinea pig Rabbit

Millipore Novus

AB2204 NB100-2288

[27] [36, 37]

CRY1

Rabbit

Novus

NBP1-69080

[58]

NMS

Rabbit

Bachem

4033397.0001

[37]

CLOCK

Rabbit

Rabbit

AB2203

[27]

[15]

This is a partial and incomplete listing of commercial primary antibodies some investigators have used with success in immunocytochemical studies. Note that not all may still be available from the suppliers, and some investigators who have developed antibodies in their own labs may distribute them on request. Note also that several companies have been known to change the source of the antibody without changing the catalog number, so that investigators sometimes purchase antibodies that do not work as expected from previously published studies. Secondary antibodies may be obtained from Jackson Immuno Research, West Grove, PA. These include Alexa Fluor 488-conjugated AffiniPure Donkey anti Guinea Pig (used at 1:300 in PBS with 0.5% BSA and 0.4% Triton-X 100); Cy3 Donkey anti-Rabbit (1:500); Alexa Fluor 488 Donkey anti-Goat (1:300); and Cy5 Donkey anti-mouse (1:300). Alexa Fluor 633 Donkey anti-Goat (DAG; 1:300) may be obtained from Invitrogen (Eugene, OR)

Anatomy of SCN

197

Fig. 2 Immunocytochemical staining of hamster SCN showing cellular and/or regional colocalization of peptides. (a) PER1-ir (red) is detected in cell nuclei while VIP-ir (green) is found in soma and processes. (b) cFos expression (green) is found lateral to PER1-ir (red) and ventral to AVP-ir (blue) 1 h after exposure to light at CT15. Note that vasopressin staining extends beyond the cytoarchitectonic boundaries of the SCN. III, third ventricle. See ref. [26] for details

Fig. 3 A lucite tray containing 60 wells serves to allow simultaneous incubations or washes of brain sections from multiple animals perfused at different circadian or zeitgeber times. Characterization of rhythmic patterns of peptide content requires the experimenter to stain sections from enough animals sacrificed at multiple phases. Use of such a rack not only allows rapid processing but also minimizes damage that might otherwise result from repeated handling of sections. The wells are 2.5 cm in diameter and the bottom of the rack is lined with nylon mesh to allow buffer to reach sections when the rack is lowered into a Pyrex rectangular baking dish that can be placed in a water bath for gentle agitation

198

Eric L. Bittman

so that simultaneous labeling of the target peptides can be achieved. The same strategy can also be used for secondary antibodies in order to minimize handling of sections. Sections from all animals and groups (genotypes, phases, treatments, etc.) that are to be compared in the experiment should be included in the same run, in order to further minimize variability. 1.2 Dual Label In Situ Hybridization

Expression patterns of multiple genes may be documented and quantified by in situ hybridization histochemistry (ISHH). Initially, this technique was applied in SCN using isotopic probes including 35 S-labeled cDNA or cRNA [42]. Early efforts using film autoradiograms with quantification by optical density gave way to use of emulsions (e.g., NTB, Kodak) that provided somewhat better anatomical resolution. Although isotopic detection remains an option, non-isotopic ISHH using cRNA probes prepared by in vitro transcription digoxygenin- or fluoroscein-labeled bases to prepare cRNA probes provides improved anatomical and temporal resolution of transcripts. These can be detected through use of antidigoxygenin or directly by fluorescent microscopy (with amplification by anti-fluoroscein; see Fig. 4). Both isotopic and non-isotopic

Fig. 4 Dual label in situ hybridization in hamster SCN showing AVP (green) and Per1 (red) transcripts revealed by Alexafluor 488 and HNPP/FR, respectively, at four circadian times. Scale bar, 100 μm

Anatomy of SCN

199

ISHH methods for study of SCN slices were described by de la Iglesia [8] in a previous volume in this series. In order to provide further insight into co-expression of genes within the SCN, we have combined multiple fluorescent non-isotopic probes that can be detected using confocal microscopy. As for the immunocytochemical detection of peptides described above, we focus on this approach in order to gain insight into function of SCN cell types. It is also possible to combine immunocytochemical targeting of SCN peptides with labeling of mRNAs by ISHH with fluorescent probes [43–46]. This approach could be expanded by combination with 35S or 33P-labeled probes, although this excludes use of the confocal microscope and raises stereological concerns. The targeted transcripts can include products of core clock genes, peptides that mark SCN regions, or combinations. The procedure below follows the method of Watakabe et al. [47] as modified by Mahoney et al. [48] to detect Per1 mRNA using digoxygenin-labeled probe in combination with AVP mRNA using fluorescein in hamster SCN (Fig. 4). However, in recent years the sensitive RNAscope technology has become increasingly popular for sensitive detection of multiple transcripts (multiplexing; [49]). This approach, which is marketed by Advanced Cell Diagnostics (https://acdbio.com), utilizes probe pair sets comprised of short cRNA sequences that target contiguous stretches of the transcript sequence and base tail regions that are exponentially amplified in sequential steps. It requires investment in a specific hybridization oven and reliance on proprietary reagents including proprietary “ZZ” probe sets. While this technology is useful for detection of colocalization of rare transcripts, it is likely unnecessary for visualization of most SCN peptides or activity markers. Another recent advance is BaseScope, which offers the ability to discriminate single nucleotide polymorphisms and noncoding, circular RNAs [50, 51]. The reader is advised to consider these approaches, and to consider whether the added sensitivity and multiplexing capabilities justify the additional expense and reliance on proprietary commercial services. Furthermore, these newer techniques limit the number of samples that can be processed simultaneously at any reasonable cost, so that increased sensitivity may come at the expense of the ability to resolve circadian phase for the reasons outlined above. If limited sensitivity due to low transcript abundance and/or the need for expanded multiplex capacity presents a problem, it is worthwhile to use TSA amplification as described above for immunocytochemical detection of peptides. As the digoxigenin label is more sensitive, amplification of the anti-FITC should be attempted before resorting to the ultra-sensitive (and expensive) RNA-Scope method.

200

2

Eric L. Bittman

Materials

2.1 Triple Label Immunocytochemistry

1. Phosphate buffer (0.1 M, pH 7.3): 3.5 g sodium phosphate monobasic dihydrate (NaH2PO4; FW 155.99), 10.9 g sodium phosphate dibasic anhydrous (Na2HPO4; FW 141.96). Bring to 1 L. Adjust pH with NaOH if necessary. 2. 4% Paraformaldehyde (PFA) in PB: 40 g paraformaldehyde PRILLS (Electron Microscopy Sciences 19202), 10.9 g sodium phosphate dibasic anhydrous, NaOH (3 beads/L). Heat to 55
C with stirring until no longer cloudy. Add 3.5 g sodium phosphate monobasic dihydrate, filter, and adjust to 1 L. Adjust pH to 7.3 by addition of HCl or NaOH (see Note 1). 3. Cryoprotectant solution (30% ethylene glycol, 1% polyvinylpyrrolidone, 30% sucrose in PBS; [52]). 4. O.C.T. Compound (Tissue-Tek 4583, Sakura). 5. Sucrose (20% in Phosphate buffer). 6. PBS+: Phosphate buffer (as above), 9 g NaCl/L, 0.4% Triton X-100, 0.1% BSA (Bovine serum albumin, Fraction V). 7. Blocking solution: 4% normal serum from the species in which the secondary antibody was raised, in PBS+ containing 0.02% sodium azide (see Note 2). 8. Primary and secondary antibodies (see Table 1 for a partial list). 9. Superfrost “plus” slides (positively charged). 10. Coverslips. 11. Xylene. 12. HistoPrep. 13. CitriSolv or other clearing compound. 14. Krystalon or Aqua-Poly/Mount. 15. Anti-fading agent (1,4-diazabicyclo[2.2.2]octane, 50 mg/ mL). If non-fluorescent detection is used: 16. Vector ABC Elite kit (PK-6101). 17. 3,30 -Diaminobenzidine (DAB). 18. 30% H2O2 (CVS or other drug store brand works fine!).

2.2 Dual Label In Situ Hybridization

19. Transcription buffer (5, supplied with polymerase). 20. Nuclease-free water (used for all solutions). Treat distilled, deionized water with 1 mL. 21. Diethylpyrocarbamate (DEPC), let stand overnight, boil for 15 min, and autoclave.

Anatomy of SCN

201

22. Dithiothreitol (DTT), 100 mM. 23. RNase inhibitor (RNAsin). 24. ATP, CTP, GTP mixture (10 mM each). 25. UTP (1.0 mM). 26. Digoxigenin-11-UTP (10 mM). 27. Fluoroscein-N6-ATP or fluorescein-N4-CTP. 28. Linearized cDNA template (1 μg/μL). 29. RNA polymerase (SP6, T3, or T7 as appropriate for template). 30. DNase (Ambion or New England Biolabs). 31. Tris-EDTA: 10 mM Tris–HCl, 1 mM EDTA, pH 7.6. 32. TNT buffer: Tris-NaCl pH 7.5, 0.05% Tween 20. 33. 4 mM NaCl. 34. tRNA (25 mg tRNA in 10 mL nuclease-free water). 35. 1 M MgCl2. 36. 0.1% SDS, autoclaved. 37. 100% EtOH. 38. Maleate buffer (5): 58 g maleic acid in 850 mL distilled water, add NaOH to bring to pH 7.5; add 43.8 g NaCl and bring to 1 L. 39. Blocking reagent Roche 11096176001 (distributed by Millipore Sigma) in 1 maleate buffer. 40. Anti-digoxygenin-alkaline phosphatase. 41. Anti-FITC-HRP (Jackson ImmunoResearch #200-0232037). 42. HNPP-Fast Red (for detection of alkaline-phosphatase labeled anti-digoxigenin; Roche Diagnostics). 43. anti DNP-Alexa488 (Molecular Probes). 44. Hoechst 33342. 45. 3,30 -Diaminobenzidine (DAB).

3

Methods

3.1 Triple Label Immunocytochemistry 3.1.1 Tissue Preparation

1. Animals are transcardially perfused with 0.1 M PB followed by 4% phosphate-buffered paraformaldehyde (PFA; both pH 7.3–7.4; [53]). Sodium nitrite (0.1%) may be added to the phosphate buffer to vasodilate in order to maximize perfusion of small vessels. When fully anesthetized, the blunted perfusion needle is inserted through the apex of the heart and guided to advance it by a few millimeters into the ascending aorta. In order to avoid perfusing into a closed system, the right atrium must be opened before the perfusion is begun.

202

Eric L. Bittman

Note that if the animal is in DD or is sacrificed during the dark phase of the LD cycle, it is necessary to prevent exposure to light during the procedure. This may be facilitated by covering the eyes with black masking tape after anesthetization and before the thorax is opened. The transcardial perfusion volumes vary with species; we use 50 mL of PBS and 100 mL of 4% PFA for mice and 100 mL of PB and 300 mL of PFA for hamsters. A perfusion rate of about 10 mL/min is suitable. 2. The brain is removed from the skull, taking care not to damage the ventral surface. The frontal pole may be reflected up and the optic nerves cut with a scalpel in order to avoid damage to the suprachiasmatic region. The brain is post-fixed overnight in 4% PFA, then transferred to 20% phosphate-buffered sucrose (pH 7.3) until it is infiltrated and sinks to the bottom of the vessel. 3. The brain is blocked and is typically mounted on the freezing stage of a microtome using several drops of O.C.T .compound. We typically section at a thickness of 40 μm (see Note 3). Sectioning is most typically done in the coronal plane, but sagittal or horizontal planes may be used to gain insight into SCN organization and spatial relationships [54]. Sections are removed from the microtome knife using a fine paint brush and placed in buffer. When cut on a freezing microtome, sections that are not to be stained immediately are stored in cryoprotectant at 4
C for 12–24 h, and then moved to 20
C freezer. The procedures below are performed on free-floating sections that are moved between solutions using a fine paint brush or glass hook fashioned from a Pasteur pipette. Thinner sections may be thaw-mounted directly onto slides, which can be accomplished most easily when cut on a cryostat. 3.1.2 Staining

4. All subsequent steps are performed at room temperature. Before use of the primary antibody it is necessary to rinse the tissue (4  5 min) in PBS+ to remove excess aldehydes that may compromise antibodies (see Note 4). 5. In order to block nonspecific signal, sections must be incubated in normal serum from the species in which the secondary antibody was raised. This incubation should be carried out for a minimum of 1 h. Blocking serum, primary and secondary antibodies are all made up to the appropriate concentration in PBS+. The concentration should be determined empirically, but in our hands 4% normal goat or normal donkey serum is generally sufficient. Inclusion of detergent (Triton-X) in the PBS+ insures that the cell membrane is permeabilized to allow penetration of serum and antibodies, and addition of sodium azide prevents bacterial or fungal growth during storage at room temperature or 4
C.

Anatomy of SCN

203

6. Sections are placed in primary antibody, made up in PBS containing BSA (fraction V) or 2% normal serum (from the species in which the secondary antibody was raised)/0.02% Triton-X. Incubation is done overnight at room temperature or for 48 h at 4
C on a gentle rocker or rotator. Concentration of antibody must be empirically determined. As noted above, for colocalization studies, two or three different antibodies (raised against various peptides in different species) may be combined in a cocktail. 7. Sections are rinsed 4  5 min in phosphate buffer, followed by incubation for 1 h or longer in secondary antibody. All steps starting with incubation in solutions containing fluorescent antibodies should be carried out in subdued light in order to reduce photobleaching. Incubations in fluorescent antibody solutions take place in microfuge tubes or multichamber racks covered in aluminum foil. Fluorescent detection may be accomplished directly (a) or indirectly (b): (a) Fluorescent 2
antibody (e.g., Cy2, Cy3, or Cy-5 conjugated goat or donkey anti-rabbit, anti-mouse, or antiguinea pig), 1:300 in PBS+. This may be visualized without further incubation steps (after rinsing and cover slipping—go to step 10). (b) As an alternative method for indirect immunocytochemistry, sections are incubated in biotinylated 2
antibody diluted 1:300 in PBS buffer containing 0.02% Triton-X and BSA or 1% normal serum. Either a fluorescent or non-fluorescent signal may be generated, with detection as described in step 9a or 9b, respectively. 8. Rinse 4  5 min in PBS. 9. Detection of 2
antibody (a) If biotinylated 2
antibody was used and fluorescent detection is desired (e.g., for confocal analysis), incubate in Streptavidin Alexaflour of appropriate wavelength. We use a 1:300 concentration, made up in buffer containing Triton-X and BSA or normal serum, for at least 1 h. Rinse 2  50 in PBS and 2  50 in PB. Counterstaining with Hoechst (10 mg/mL) or DAPI (1 μg/mL) to label nuclei, or FluoroNissl to label somata, may be done after completion of immunostaining. (b) If staining is to be detected using conventional (i.e., non-fluorescent) microscopy, avidin-labeled peroxidase may be used. The Vectastain Elite kit is most often used to detect the biotinylated secondary antibody. Incubate sections for 1 h in ABC reagent (5 μL of A and B in

204

Eric L. Bittman

490 mL of 0.1 M PB or Tris-NaCl). Rinse 4  5 min in buffer, then incubate in DAB (0.05%/0.003% H2O2 in PBS or Tris-NaCl). Watch the sections carefully as they can quickly overdevelop. The progress of the reaction may be slowed by use of glucose oxidase. Stop reaction with buffer rinses. 3.1.3 Mounting

10. Using a fine paint brush, maneuver stained sections onto a subbed slide. Be sure to avoid folding and tearing of tissue. Allow to dry several hours or overnight. 11. Dehydrate through graded alcohols (70%-85%-95%-absolute). Clear in xylenes or substitute (e.g., Citri-Solv, Deacon laboratories; or histosol, National Diagnostics, etc.). Anti-fading agent may be added if extensive exposure to confocal laser is anticipated. 12. Cover slip with suitable medium. If using Alexafluor or fluorescent secondary antibody, dehydration and clearing can be omitted and cover slipping can be done using Aqua-Poly. Otherwise, Krystalon is used for fluorescent sections. Fluorescent labeling is compatible with confocal microscopy and analysis/quantification using NIH Image software with FIJI [55]. If tissue is DAB stained (non-fluorescent, as described in step 9b), use Permount.

3.2 Dual Label In Situ Hybridization 3.2.1 Tissue Preparation

1. Rapidly decapitate using a guillotine. This must be done in darkness or under dim red light (>620 nm, 20 min. (c) Place the beaker at 4  C overnight, in order to allow the Chelex® to pellet. Day 2 (a) Carefully discard the water and add fresh ddH2O to obtain a total volume of 4 l. (b) Adjust pH to 7.4 by using 10 M HCl, while continuing stirring until the pH remains stable for >20 min. (c) Place the beaker at 4  C for 1 h, in order to allow the Chelex® to pellet. (d) Carefully discard the water. (e) Add 1 l of characterized FBS and slowly stir at 4  C for 1 h, avoiding the formation of bubbles. (f) Place the beaker at 4  C for 1 h, in order to allow the Chelex® to pellet. (g) Filter the obtained serum through a bottle-top filter of 0.45 μm under sterile conditions and aliquot it in falcons to be stored at 20  C (see Note 3). 2.3

Other Reagents

1. ddH2O. 2. PBS. 3. Trizol. 4. FBS. 5. Trypan blue stain. 6. Dispase: Dissolve 5 g of dispase in PBS to achieve a final concentration of 10 mg/ml. Aliquot and freeze at 20  C. Dilute 1:10 in PBS prior to its use.

Use of Interfollicular Stem Cells for Circadian Studies

3

247

Methods

3.1 Processing of Mouse Skin

1. Sacrifice mice by following the national or regional animal welfare regulations. Sacrifice mice at a specific ZT, noting the specific time of sacrifice. As one mouse can only be used for one time point, it is recommended to include at least four replicates per time point to avoid high inter-animal variability. 2. Shave the entire back and belly area up to the arms and down to the legs, by shaving against the hair direction. 3. Remove the skin with dissection tools and place it on ice inside a 100 mm petri dish (see Notes 4 and 5). 4. Place the tissue upside down, with the dermal layer facing up (Fig. 2a). 5. Remove the fat bumps with scissors (see Note 6 and Fig. 2a). 6. Scrape off the hypodermis to work with dermis and epidermis by removing fat and muscle using a sterile scalpel (see Note 7 and Fig. 2b–d). Additionally, remove possible black spots (anagen patches) on the skin, if necessary, as they will disturb the circadian profile of the cells. 7. Float the skin tissue on 0.25% trypsin without EDTA (see Notes 8 and 9) on a clean petri dish, with the dermal layer facing down and in contact with the liquid (Fig. 2e, f). Incubate at 37  C for approximately 1 h (see Notes 10 and 11).

3.2 Isolation of Epidermal Cells from Mouse Skin

The following steps should be carried out on ice. Cool down a centrifuge to 4  C. 1. Transfer the skin tissue to the petri dish lid (Fig. 2g). 2. Gently scrape off the epidermal layer with the hairs using a sterile scalpel (see Note 12 and Fig. 2h). 3. Discard the dermis and stop the trypsin with 0.5 ml of chelated FBS. 4. Chop the epidermis with the tissue chopper until reducing the tissue into very small homogenous pieces (see Note 13 and Fig. 2i). 5. Transfer the chopped epidermis into a 50 ml Falcon tube and add 10 ml of cold PBS. 6. Shake vigorously (Fig. 2j). 7. Filter through a 100 μm strainer into a new 50 ml Falcon tube (Fig. 2k). 8. Repeat this straining step by transferring the clumps left in the strainer into the previously used 50 ml Falcon tube and adding further 10 ml of wash buffer (Fig. 2l).

248

Paloma Sola´ and Valentina M. Zinna

Fig. 2 Epidermal layer isolation from adult murine skin. (a) Hypodermal view of dissected skin. Arrows are pointing to adipose clumps present after dissection. (b) Removal of fat clumps with scissors. (c) Scraping off adipose layer using a sterile scalpel. Appearance of the dermis while adipose layer is being removed. The arrow is marking the dermal fraction still containing adipose cells. (d) Back skin after completely removal of adipose tissue. (e) Epidermal view of dissected skin. Arrow points at the spot where trypsin is being pipetted into the dish and below the skin tissue. (f) Trypsin incubation in petri dish with dermal side down. (g) Skin is placed in a new petri dish after trypsin incubation, with dermal side down (see Note 7). (h) Scraping off epidermal layer using a sterile scalpel. (i) Chopping of epidermal cells. (j) Epidermal cells in suspension after being scraped off from the skin tissue. (k) Filtering of epidermal cells. (l) Epidermal cells in suspension after being filtered. Appropriate institutional regulatory board permission was obtained to use the mouse that was the source of tissue for this and subsequent figures

9. Filter through 40 μm strainer; discard the hair leftovers and the strainer. 10. Centrifuge at 300  g and 4  C for 20 min (see Note 14). 11. Carefully aspirate the supernatant and resuspend the pelleted cells in 500 μl of cold PBS. 12. Assess cell count and viability by mixing 50 μl of cells with 50 μl of Trypan blue stain, and check under the microscope with a Neubauer chamber.

Use of Interfollicular Stem Cells for Circadian Studies

3.3

Stem Cell Sorting

249

1. Divide the cell suspension in order to add the antibody of interest and the single controls into each separate tube (see Notes 15 and 16). The single controls are: (a) Unstained including DAPI only (1 mg/ml). (b) α-CD34-FITC (1:50, clone RAM34). (c) α-α6-Integrin/CD49f-PE (1:200, clone NKI-GoH3). 2. Spin down at 300  g and 4  C for 5 min. Discard supernatant in all tubes. 3. Add α-CD34-FITC (1:50) and α-α6-Integrin/CD49f-PE (1: 200) to the samples and single controls. Add approximately 1 ml of sorting buffer with the antibody mix per 107 cells, and 200 μl for the unstained and single controls (see Note 2). In this step, the unstained sample must be resuspended only in staining buffer without antibodies, and the single controls in only one antibody each (see Notes 17 and 18). 4. Incubate for 30–40 min on ice in darkness and mix by mildly pipetting after the initial 15 min. 5. Add 2 ml of wash buffer. Centrifuge at 300  g and 4  C for 5 min. 6. Remove supernatant and resuspend all pellets in 300 μl sorting buffer supplemented with DAPI (1 mg/ml) (see Note 19). 7. Make sure the FACS is calibrated and ready to use. The sorting must be carried out using a nozzle of 100 μm into 1.5 Eppendorf tubes, and with a temperature of 4  C throughout the sorting (see Notes 20 and 21). 8. The FACS gating strategy for isolation and analysis of epidermal and hair follicle stem cells is displayed in Fig. 3. Sort for single cells, live DAPI cells, CD34 and CD49f+. 9. For RNA extraction, centrifuge the sorted cells at 300  g and 4  C for 5 min after sorting, resuspend in 500 μl of Trizol, and store at 80  C until further use. 10. Extract the RNA following the manufacturer’s instructions. 11. Check RNA quality and use for gene expression analysis such as RNA-sequencing or RT-qPCR.

3.4 Identification of Rhythmic Genes

In order to detect rhythmic transcripts from gene expression data, there are several algorithms that can be used. One of the most widely accepted and used is the Jonckheere-Terpstra-Kendall (JTK_CYCLE) algorithm, which is able to detect and characterize oscillations in large-scale datasets [8]. Nonetheless, there are also algorithms available for this same purpose [9–11].

Paloma Sola´ and Valentina M. Zinna

250

b

All events

Single Cells

250K

250K

200K

200K

150K

150K

FSC–W

SSC–A

a

100K

100K

50K

50K

Cells 84.8%

Single cells 72.4%

0

0 0

50K

100K

150K

200K

250K

0

50K

FSC–A

c

100K

150K

200K

250K

FSC–A

d

Single Cells

EpSCs

5

105

104

104 HFSCS 27.1%

CD34

DAPI

10

103

103 Live cells 71.6%

0 –103

EpSCs 47.7%

0 –103

0

50K

100K

150K

FSC–A

200K

250K

–103

0

103

104

105

CD49f (D6 Integrin)

Fig. 3 FACS isolation strategy and gene expression controls. Representative examples of the FACS strategy and gating scheme for epSC isolation from mice are shown. (a) Sorting of cells, (b) single cells, (c) live cells, (d) CD34 CD49f+ cells

4

Notes 1. Example of conjugated antibody panel shown. Other fluorochrome-conjugated antibodies recognizing CD34 and CD49f can be used after ensuring the lowest spectral overlap possible. 2. Alternatively, solely PBS at 4  C can also be used. 3. The usage of chelated FBS prevents cell differentiation. It can be aliquoted and stored at 20  C for 6 months to 1 year. Check manufacturer’s instructions for storage recommendations.

Use of Interfollicular Stem Cells for Circadian Studies

251

4. If kept at 4  C, it is possible to keep the skin tissue on ice for up to 6 h without affecting the quality of the whole cell preparation procedure. 5. Make sure samples are correctly and consistently labeled at all times to avoid sample mixing. An example of sample labeling is depicted (Fig. 2g–i). 6. An optional step is to take a 1 cm2 from the caudal back-skin area for histology assays before removing fat. Transfer the section into a tissue-embedding cassette with the dermis side in contact with a piece of Whatman paper, and incubate it in fixing solution (10% neutral buffered formalin), either at room temperature for 3 h or at 4  C overnight. Transfer into 4  C PBS until embedding in paraffin blocks (transfer into PBS with Azide if stored for longer time prior to embedding). 7. This step is critical and should be performed properly and carefully for two main reasons: (a) Skin tissue can get ripped easily, especially when working with females whose skin is generally much thinner than in males. (b) The muscle and fat layers should be completely removed, in order for the trypsin to properly detach the epidermal layer from the dermis. 8. If the cells will be cultured in vitro, an additional sterilization step is recommended prior to the trypsin treatment. Incubate the skin with the dermis side up in an Antibiotic (100 U/ml Penicillin, 100 μg/ml Streptomycin) and Antimycotic (0.5 μg/ ml) solution for 5 min at room temperature in a new petri dish. 9. The usage of trypsin without EDTA allows preserving cell viability. 10. Trypsin incubation times can vary depending on skin thickness. Skin isolated from females usually requires shorter incubation times (30–45 min) than males (1 h). Check for optimal incubation time by lightly scraping a patch of the epidermis with a sterile scalpel while being incubated in trypsin. Stop incubation when epidermis is easily separated from dermis, as prolonged trypsin incubation may affect cell viability. 11. Trypsin can be replaced by dispase (1 mg/ml). This can be done if the antibodies used for FACS are trypsin-sensitive. Check the antibody manufacturers recommendations. 12. The epidermis should come off very easily while scraping. If this is not the case, it is likely that the fat on the dermal side was not scraped properly before the tissue digest step. An option is to leave the skin samples in trypsin for longer; however, longer incubation times can affect cell viability.

252

Paloma Sola´ and Valentina M. Zinna

13. Chopping can be done manually using scissors instead of using the automatic tissue chopper. 14. Alternatively, the centrifugation step can be carried out at 200  g in case the sample is rich in fat. Centrifugation times throughout the protocol depend on the final volume used. 15. Antibody dilution can vary depending on the antibody used. Check antibody datasheet for recommended dilutions. 16. Small volumes of cell suspension can be used for the single controls if sorting a high number of cells is required. 17. The final volume of sorting buffer can vary depending on cell number. 18. If more than two antibodies are included in the sorting, fluorescence minus one (FMO) control must be included for each fluorochrome. This control is based on including all fluorochromes in a panel except for the one being measured. This control ensures that any possible spread of the fluorochromes into the channel of interest is considered for the gatings. 19. The volume of sorting buffer used in this step can vary, but it is recommended to follow the relation of maximum 1  106 cells per 100 μl in order to sort efficiently. 20. If the cells will be cultured in vitro, make sure that the sorter maintains sterility. 21. When sorting, it is advisable to take note of the number of sorted cells. However, it is recommended to count cells in a Neubauer chamber if a more precise cell count is needed for the downstream application.

Acknowledgments P.S. is supported by the Spanish Ministry of Economy and Development (ID BES-2017-081279). V.M.Z. has received financial support through the “la Caixa” INPhINIT Fellowship Grant for Doctoral Studies at Spanish Research Centers of Excellence, “la Caixa” Foundation, Barcelona (ID 100010434). The fellowship code is LCF/BQ/IN17/11620018. In addition, V.M.Z. has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie SkłodowskaCurie grant agreement no. 713673. Figure 1 was created with BioRender.com.

Use of Interfollicular Stem Cells for Circadian Studies

253

References 1. Solanas G, Benitah SA (2013) Regenerating the skin: a task for the heterogeneous stem cell pool and surrounding niche. Nat Rev Mol Cell Biol 14(11):737–748 2. Janich P et al (2011) The circadian molecular clock creates epidermal stem cell heterogeneity. Nature 480(7376):209–214 3. Janich P et al (2013) Human epidermal stem cell function is regulated by circadian oscillations. Cell Stem Cell 13(6):745–753 4. Solanas G et al (2017) Aged stem cells reprogram their daily rhythmic functions to adapt to stress. Cell 170(4):678–692.e20 5. Geyfman M et al (2012) Brain and muscle Arnt-like protein-1 (BMAL1) controls circadian cell proliferation and susceptibility to UVB-induced DNA damage in the epidermis. Proc Natl Acad Sci U S A 109(29): 11758–11763 6. Jensen KB, Driskell RR, Watt FM (2010) Assaying proliferation and differentiation

capacity of stem cells using disaggregated adult mouse epidermis. Nat Protoc 5(5): 898–911 7. Turksen K (2019) Skin stem cells: methods and protocols. Springer, New York 8. Hughes ME, Hogenesch JB, Kornacker K (2011) JTK_CYCLE: an efficient non-parametric algorithm for detecting rhythmic components in genome-scale datasets. J Biol Rhythms 4(164):1687–1697 9. Yang R, Su Z (2010) Analyzing circadian expression data by harmonic regression based on autoregressive spectral estimation. Bioinformatics 26(12):i168–i174 10. Thaben PF, Westermark PO (2014) Detecting rhythms in time series with RAIN. J Biol Rhythm 29(6):391–400 11. Hutchison AL et al (2015) Improved statistical methods enable greater sensitivity in rhythm detection for genome-wide data. PLoS Comput Biol 11(3):e1004094

Chapter 17 Detecting Circadian Rhythms in Human Red Blood Cells by Dielectrophoresis Andrew D. Beale, Fatima H. Labeed, Stephen J. Kitcatt, and John S. O’Neill Abstract Dielectrophoresis (DEP) enables the measurement of population-level electrophysiology in many cell types by examining their interaction with an externally applied electric field. Here we describe the application of DEP to the measurement of circadian rhythms in a non-nucleated cell type, the human red blood cell. Using DEP, population-level electrophysiology of ~20,000 red blood cells can be measured from start to finish in less than 3 min, and can be repeated over several days to reveal cell-autonomous daily regulation of membrane electrophysiology. This method is amenable to the characterization of circadian rhythms by altering entrainment and free-run conditions or through pharmacological perturbation. Key words Dielectrophoresis, Erythrocyte, Red blood cell, Circadian rhythm, TTFL

1

Introduction Circadian rhythms have been detected in the absence of TTFLbased mechanisms in a number of contexts [1–4]. Most recently, focus has turned to the human red blood cell as a model of a TTFLless circadian clock [5–8]. However, unlike the vast majority of studies into cellular circadian rhythms, the longitudinal measurement of circadian activity by bioluminescence reporters is not available for circadian studies of red blood cells since transcription and translation do not occur in this minimal cell type. Instead, serial sampling methods, such as immunoblotting of lysed cells, have been deployed to study aggregated population-level rhythmicity [5, 6]. Dielectrophoresis (DEP) has emerged as a rapid and simple alternative to immunoblotting for the study of circadian rhythms in red blood cells. DEP measures the electrophysiological properties of a population of cells, such as membrane conductance and cytoplasmic conductivity, by examining the interaction of cells with externally applied electrical fields. DEP has been used to in a range

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_17, © The Author(s) 2022

255

256

Andrew D. Beale et al.

of cellular applications including the study of apoptosis, malarial infection, and the cellular basis of cardiac arrhythmias [9–11]. More recently, DEP has been applied to the study of circadian rhythms in red blood cells, revealing a critical function for dynamic potassium transport and casein kinase 1 activity [7, 8]. The device used in those studies, the 3DEP, is a benchtop instrument consisting of an optical detector and a disposable chip which measures the electrical properties of ~20,000 cells in 10 s [12]. Its ease of use and rapid measurement time makes it ideal for use over circadian timescales with multiple biological replicates and conditions. Here we describe the materials and method for measuring the circadian dielectric properties of red blood cells as used in Henslee et al. [7] and Beale et al. [8], measuring RBC rhythms from two human donors using a single thermocycler. Increased thermocycler capacity allows increased experimental throughput, and it is quite feasible to measure RBC rhythms from four human donors in response to two different pharmacological treatments in a single experimental run [7].

2

Materials

2.1 Sample Isolation and Preparation

1. 6 mL Lithium heparin Vacutainer. 2. Insulated box to carry Vacutainers. 3. Dulbecco’s phosphate-buffered saline (DPBS). 4. Histopaque-1077. 5. 15 and 50 mL conical centrifuge tubes. 6. Benchtop centrifuge. 7. Modified Krebs-Henseleit buffer (KHB): Krebs-Henseleit buffer (K3753, Sigma-Aldrich), 2.54 mM CaCl2 dihydrate, 25 mM NaHCO3, 1 g/L bovine serum albumin, 100 U/L penicillin, and 100 μg/L streptomycin, pH 7.6, 290–299 mOsm. Add approximately 900 mL dH2O to a 1 L graduated cylinder or glass beaker. Add Krebs-Henseleit powdered medium to the cylinder, rinsing the original vial with dH2O and adding to the cylinder, and stir for 20 min (see Note 1). Add 0.393 g CaCl2 dihydrate (to a final concentration of 2.54 mM) to the cylinder and stir until completely dissolved, approximately 5 min. Add 2.1 g NaHCO3 and stir until completely dissolved, approximately 5 min. Add 1 g bovine serum albumin, and 10 mL of 100
penicillin and streptomycin and stir until completely dissolved, approximately 5 min. Bring to 1 L with dH2O. Adjust pH to 7.2 (see Note 2) using 1 N HCl and 290–299 mOsm using 5 M NaCl. Sterilize by filtration using 0.22 μm pore membrane. Store medium at 4  C for up to 1 month. 8. 0.5 mL Sterile PCR tubes.

Circadian Rhythms in Red Blood Cells by Dielectrophoresis

257

Fig. 1 The 3DEP reader setup with computer running 3DEP analysis software

2.2 Entrainment and Treatment of Cells

1. Extended capacity thermocycler/PCR machine (able to take 0.5 mL tubes). 2. Drugs and vehicle controls at 10
concentration in KHB.

2.3 Dielectrophoresis

1. 3DEP 3D Dielectrophoresis Cell Analysis System (3DEP reader, Labtech; Fig. 1). 2. 3DEP chips (Labtech, Heathfield, East Sussex, UK; Fig. 2a). 3. 18
18 mm #1 square coverslips. 4. 1 mL Luer syringes. 5. 25 G 16 mm stainless steel needles, tips removed using surgical scissors and bent to approximately 45  C (Fig. 2b). 6. 1 mL Microcentrifuge tubes. 7. Iso-osmotic DEP medium: 248 mM sucrose, 16.7 mM dextrose, 250 μM MgCl2, 100 μM CaCl2, 290–299 mOsm (see Note 3). Dissolve 85 g of sucrose and 5 g of dextrose in 850 mL of dH2O stir until completely dissolved, approximately 5 min. Adjust to 290–299 mOsm using NaCl and bring up to 1 L with dH2O. Adjust to 0.043 S/m using DPBS (see Note 4). Sterilize by filtration using 0.22 μm pore membrane. Store medium at 4  C for up to 1 month. 8. 70% Ethanol. 9. Non-scratch, no-lint paper wipes (e.g., Kimtech Precision Wipes). 10. Paper towelling.

258

Andrew D. Beale et al.

Fig. 2 (a) 1 mL Luer with attached syringe clipped and bent to 45 angle. (b) The 3DEP chip showing small meniscus after addition of cell suspension

3 3.1

Methods Sample Isolation

1. Under the supervision of a qualified phlebotomist, draw venous blood from a human donor into 6 mL lithium heparin Vacutainers. One Vacutainer/donor is sufficient. Place in an insulated box (Fig. 3). 2. Transfer Vacutainers from donor facility to laboratory for processing. Spray external surfaces of Vacutainer with 70% ethanol and place on a rack in a laminar flow/cell culture hood. 3. Add 3 mL of whole blood to 5 mL of DPBS in a 15 mL conical centrifuge tube (1:1.67). Mix gently by inversion. 4. Layer blood-DPBS mixture on top of 3 mL Histopaque-1077 (Sigma-Aldrich) in a new 15 mL conical centrifuge tube, and

Circadian Rhythms in Red Blood Cells by Dielectrophoresis

259

Fig. 3 Schematic of steps for the isolation of red blood cells from human donors. (Created with BioRender. com)

spin in a benchtop centrifuge at 50
g for 30 min with brake and acceleration set at their lowest settings. Purified RBCs are pelleted at the bottom of this mixture. 5. Carefully remove supernatant and higher layers and dispose. Wash pelleted RBCs in 10 mL DPBS (250
g for 15 min). 6. Remove supernatant and wash pellet in modified KrebsHenseleit buffer (250
g for 10 min). 7. Take 330 μL of washed and packed RBCs to 40 mL of KHB in a 50 mL Falcon tube (see Note 5). 8. Count cells using a hemocytometer to check cell concentration is approximately 6
106 cells/mL. Check cell morphology for damage, cell swelling or shrinking, and purity of RBC prep. 9. Aliquot 450 μL RBC/KHB suspension to sterile 0.5 mL PCR tubes. Prepare 1 PCR tube/donor/treatment/timepoint + 3 additional tubes/donor/treatment/timepoint (see Note 6). Label tubes with donor ID and treatment condition. 3.2 Entrainment and Sampling

1. Transfer PCR tubes to thermocycler preheated to 37  C. 2. Program the thermocycler to cycle between 37  C and 32  C for two complete cycles—i.e., 12 h 37  C, 12 h 32  C, 12 h 37  C, 12 h 32  C. Alternative entrainment conditions may be tested by altering the duration of each temperature step (T-cycles).

260

Andrew D. Beale et al.

3. Program the thermocycler to transfer into constant temperature conditions after two cycles of temperature entrainment (see Note 7). 4. At the point of transition into constant conditions, remove PCR tubes from thermocycler and place on isothermal pads. In laminar flow hood, add 50 μL of 10
drug or vehicle as required to cells and replace in thermocycler. 5. At 3-hourly intervals from transfer into constant conditions, remove from the thermocycler a single 500 μL PCR tube of RBCs at each time point for each donor for each treatment condition. 6. Process each tube for DEP (see Note 8). 3.3 DEP Using the 3DEP

1. Prepare 3DEP chips by flushing the wells with 70% EtOH (see Notes 9 and 10). Clean glass base with low-lint towel in 70% EtOH. Dry upside down on paper towelling. 2. Carefully remove KHB medium from cell pellet (see Note 11) by pipette. Take care not to disturb cell pellet. 3. Resuspend pellet in 500 μL iso-osmotic DEP medium and spin for 2 min at 500
g. 4. Remove supernatant and resuspend in 200 μL DEP medium in order to reduce ionic carryover from the high ionic strength KHB medium. Take 50 μL of resuspended cells and add to 950 μL of DEP medium in 1 mL microcentrifuge tubes to achieve working cell concentration of 1
106 cells/mL. 5. Take 10 μL of cell suspension and place in a hemocytometer. Using a bright-field microscope, take picture for the subsequent measurement of the radius of red blood cells in software such as Image J (Fig. 4). 6. Using a bent needle attached to a 1 mL Luer syringe, take up ~400 μL of diluted cell suspension from the microcentrifuge tubes. Inject ~100 μL into the 3DEP chip by injecting through various electrode rings of the chip, ensuring that no bubbles remain in the chip well (see Note 12). A small meniscus should be present on the surface of the chip (Fig. 2b). 7. Gently place a coverslip on top of the 3DEP chip over the wells and cells. Ensure that no bubbles are trapped between coverslip and wells. If the amount of cell resuspension is correct, very little should escape from the sides of the coverslip. 8. Place coverslip-topped chip into 3DEP reader. Lock chip in place using software controllers (see Note 13), shut the door, and energize chip for 10 s at 10 Vp–p, at 20 frequencies between 10 kHz and 20 MHz.

Circadian Rhythms in Red Blood Cells by Dielectrophoresis

261

Fig. 4 Example of image taken of RBCs suspended in DEP medium immediately before 3DEP analysis. Field of view is equal to 250 μm

9. Repeat measurement three to four times for each donor and condition: rinse chip with DEP medium, ensure that no solution remains by blotting chip on tissue, before injecting cell suspension. 10. Using 3DEP analysis software, fit Clausius-Mossotti model to overlayed data, using radius measurement obtained by analysis of hemocytometer images. Example data is shown in Fig. 5.

4

Notes 1. KHB needs to be fully dissolved before adding any of the other components, after filtration, adjust to pH 7.4 at 37  C with 1 M HCl. 2. The bicarbonate buffering in the modified KHB is most compatible with red blood cells, which contain carbonic anhydrase at high abundances. 3. The divalent cations supplied by MgCl2 and CaCl2 are provided from stock solutions in dH2O. The divalent cations are included to prevent potential membrane function degradation [10]. 4. DPBS is isosmotic with the sucrose/dextrose solution that is the base of DEP medium. Its purpose here is solely to adjust the conductivity of the suspension solution for dielectrophoresis using the 3DEP reader and chip.

Andrew D. Beale et al. 12000

Geff (S/m2)

10000 8000 6000 4000 2000

(S/m)

0.25

0.20

cyt

262

0.15

0.10 0

12

24

36

48

Time (h)

Fig. 5 Isolated RBCs from four human donors were cultured at constant 37  C after 2 days of temperature entrainment. DEP measurements of membrane conductance (Geff) (top) and cytoplasmic conductivity (σ cyt) (bottom) are reported (mean  SEM, n ¼ 4). Damped cosine curves were fitted to the data using non-linear regression comparison of fits (h1 ¼ damped cosine model, h0 ¼ straight line) extra sum-of-squares F test. The cosine fit was preferred over straight line fit if p < 0.05. Geff and σ cyt oscillate in antiphase, as previously reported (Henslee et al. [7])

5. The resuspended RBCs in KHB should appear as a translucent red suspension. Lysis of red blood cells will be apparent as the solution will be a transparent red, similar to a rose´ wine. 6. Prepare 1 PCR tube/donor/treatment/timepoint + 3 additional tubes/donor/treatment. Since this procedure relies on serial sampling, additional tubes are used as backups where cell quality is not sufficient, or mistakes occur. Using this approach, a standard extended capacity PCR machine with 96 wells can be used to test two donors with one treatment and one control over 2 days of constant conditions (16 timepoints at a 3 h interval). 7. The objective of this protocol is to assess bona-fide circadian rhythmicity by sampling in constant conditions—a hallmark of circadian studies. The protocol can be adapted to assess entrainment by sampling under cycling conditions. Rather

Circadian Rhythms in Red Blood Cells by Dielectrophoresis

263

than programming the thermal cycler to stay at a fixed temperature after two cycles, simply continue the cycling program for more time—sampling for an additional two cycles will be sufficient. Care must be taken to not extend the time in KHB much longer than 4 days as RBC viability will start to decrease after this time. 8. DEP sampling—our experience is that preparing a single donor and condition in turn is the most reliable method. Electrophysical properties of cells change with increased time in low ionic strength media such as DEP medium (Hughes et al., in submission), and therefore minimizing time spent in DEP medium is critical. 9. DEP medium is a sugar solution and workstations or DEP chips can therefore become sticky when it evaporates. Rinsing the DEP chips with copious amounts of EtOH, and frequent spray downs of benchtops are therefore highly recommended. This is particularly important for the surface electrodes and the glass well bottom of the DEP chip. 10. DEP chips “improve” with age up to a point. When new, DEP chips need to be “conditioned” by rinsing with both EtOH and DEP medium, with a final wash in EtOH before use. 11. RBCs will settle to the bottom of PCR tubes over time. The settled RBCs are referred to as the cell pellet in this paper. 12. We find that injecting ¼ of the volume into the 4 corner rings of the chip is a reliable way to fill the chip without trapping bubbles. 13. Locking the chip in place is a critical step; at this point, electrical contacts are made between the 3DEP and the chip. The locking mechanism of the 3DEP is positioned so that these contacts are made automatically, but we find a small amount of horizontal force from a thumb on the outward-facing edge of the chip improves the frequency of true contacts at all 20 electrodes. References 1. O’Neill JS, Van Ooijen G, Dixon LE et al (2011) Circadian rhythms persist without transcription in a eukaryote. Nature 469:554–558. https://doi.org/10.1038/nature09654 2. Lakin-Thomas PL (2006) Transcriptional feedback oscillators: maybe, maybe not. . . . J Biol Rhythm 21:83–92. https://doi.org/10. 1177/0748730405286102 3. Tomita J, Nakajima M, Kondo T, Iwasaki H (2005) No transcription-translation feedback in circadian rhythm of KaiC phosphorylation.

Science 307:251–254. https://doi.org/10. 1126/science.1102540 4. Woolum JC (1991) A re-examination of the role of the nucleus in generating the circadian rhythm in Acetabularia. J Biol Rhythm 6:129– 1 3 6 . h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 074873049100600203 5. O’Neill JS, Reddy AB (2011) Circadian clocks in human red blood cells. Nature 469:498– 503. https://doi.org/10.1038/nature09702

264

Andrew D. Beale et al.

6. Cho C-S, Yoon HJ, Kim JY et al (2014) Circadian rhythm of hyperoxidized peroxiredoxin II is determined by hemoglobin autoxidation and the 20S proteasome in red blood cells. Proc Natl Acad Sci U S A 111:12043–12048. https://doi.org/10.1073/pnas.1401100111 7. Henslee EA, Crosby P, Kitcatt SJ et al (2017) Rhythmic potassium transport regulates the circadian clock in human red blood cells. Nat Commun 8:1978. https://doi.org/10.1038/ s41467-017-02161-4 8. Beale AD, Kruchek E, Kitcatt SJ et al (2019) Casein kinase 1 underlies temperature compensation of circadian rhythms in human red blood cells. J Biol Rhythm 34:144–153. https://doi. org/10.1177/0748730419836370 9. Gascoyne P, Pethig R, Satayavivad J et al (1997) Dielectrophoretic detection of changes in erythrocyte membranes following malarial infection. Biochim Biophys Acta Biomembr

1323:240–252. https://doi.org/10.1016/ S0005-2736(96)00191-5 10. Fry CH, Salvage SC, Manazza A et al (2012) Cytoplasm resistivity of mammalian atrial myocardium determined by dielectrophoresis and impedance methods. Biophys J 103:2287– 2294. https://doi.org/10.1016/j.bpj.2012. 10.023 11. Chin S, Hughes MP, Coley HM, Labeed FH (2006) Rapid assessment of early biophysical changes in K562 cells during apoptosis determined using dielectrophoresis. Int J Nanomedicine 1:333–337 12. Hoettges KF, Henslee EA, Torcal Serrano RM et al (2019) Ten–second electrophysiology: evaluation of the 3DEP Platform for highspeed, high-accuracy cell analysis. Sci Rep 9: 1–13. https://doi.org/10.1038/s41598019-55579-9

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), 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 chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Chapter 18 Measuring Circadian Neutrophil Infiltration in Tissues by Paired Whole-Mount Tissue Clearing and Flow Cytometry Tommaso Vicanolo, Andres Hidalgo, and Jose M. Adrover Abstract Neutrophils infiltrate most tissues in the organism in the steady state, often following circadian patterns. Neutrophil infiltration is also key to immune defense under inflammatory conditions. In all cases, accurate measurements of the absolute number of infiltrated cells and of their localization are important to understand steady-state or inflammatory migration patterns and kinetics. Here we present a method to obtain accurate information on both neutrophil number and distribution that can be successfully applied to circadian studies of neutrophil (or any other cell of interest) migration in vivo. Moreover, this method can be also used to obtain information on activation states or effector functions, for example, by measurement of neutrophil extracellular trap formation in tissues. Key words Circadian, NETs, TRALI, Lung injury, Tissue clearing

1

Introduction The immune system is tightly controlled by circadian oscillations [1] as shown, for example, in granulocytes and monocytes which exhibit circadian oscillations in numbers and phenotype while in circulation, both in humans [2] and mice [3]. Neutrophils are the most abundant leukocyte in the human blood and their estimated half-life in the bloodstream is 6–12 h, before being eliminated mainly in the bone marrow, spleen, and liver [4]. Their short life span in circulation demands constant production and release of neutrophils [5], and notably their release from the bone marrow follows circadian fluctuations [4], as do their counts in blood, with cycles that suggests a period of active release around ZT17 (i.e., 17 h after the onset of light in a 12 h light:12 h dark regime), and a period of clearance (i.e., migration out of blood and into the tissues) between ZT5 and 13 [6]. Remarkably, these circadian oscillations also encompass changes in the phenotype, a property

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_18, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

265

266

Tommaso Vicanolo et al.

referred to as neutrophil aging, which is controlled cell-intrinsically by the core molecular clock machinery [7]. We have recently shown that neutrophils, which have been typically studied in the context of infection or inflammation, naturally migrate to virtually every tissue (including the intestine, lung, white-adipose tissue (WAT), skin, skeletal muscle, lymph nodes, kidneys, and heart) even in the absence of injury [8, 9]. Notably, infiltration of neutrophils into most naı¨ve tissues follows circadian patterns with increased migration in the evening, with some exceptions in the intestine, liver, and WAT in which no circadian oscillations are detected [8]. Oscillations in most tissues were in anti-phase with those in the circulation such that numbers within tissues increased as they decreased from the blood. Further evidence suggests that the circadian accumulation of neutrophils within tissues may serve physiological purposes [9, 10]; for example, they influence circadian transcription patterns in the lungs [8], promote angiogenesis when infiltrating the lung and the intestine [9], and in the bone marrow infiltrating neutrophils control the size and activity of the hematopoietic niche [6]. The spatial location of neutrophils in tissues remain elusive as of today, especially in the context of circadian time. Studies of human spleens have demonstrated the presence of at least two populations of neutrophils in the perifollicular zone [11], and in mouse spleens neutrophils locate in the marginal zone and red pulp [12]. In the intestinal mucosa neutrophils do not distribute homogeneously, but rather cluster in discrete areas enriched in B lymphocytes that resemble isolated lymphoid follicles (ILFs) [8]. By contrast, in heavily vascularized tissues, such as the liver and lungs, neutrophils seem to localize preferentially within vessels. For instance, intravital microscopy studies in the murine lung micro-vasculature revealed a substantial number of neutrophils marginated within the network of small capillary vessels that are rapidly mobilized by the CXCR4 antagonist Plerixafor in both mice and primates [13, 14], or which are actively crawling on small pulmonary capillaries in mice [15]. The precise infiltration and spatial distribution patterns of neutrophils inside tissues are still poorly defined. Hence, we developed a method to allow paired and precise evaluation of neutrophil numbers and neutrophil localization in multiple tissues at different timepoints of the day. We employ flow cytometry and whole-mount cleared-tissue immunofluorescence to obtain both quantitative and qualitative information of the patterns of neutrophil clearance in the different tissues. This method will be valuable when analyzing both steady-state and inflammationguided neutrophil migration in tissues.

Analysis of Neutrophil Infiltration Dynamics

2

267

Materials 1. Sterile PBS: NaCl 137 mM, KCl 2.7 mM, Na2HPO4 10 mM, KH2PO4 1.8 mM in distilled water; pH 7.4. 2. PEB: 2 mM EDTA, 0.5% heat-inactivated FBS and 0.1% Sodium Azide in PBS. 3. Counting beads. 4. Sterile HBSS: KCl 5.33 mM, KH2PO4 0.44 mM, NaHCO3 4.16 mM, NaCl 137.9 mM, Na2HPO4 0.33 mM in distilled water; pH 7.4. 5. 36% Percoll in HBSS. 6. 10
Red Blood Cell Lysis Buffer: 1.5 M KH4Cl, 0.1 M KHCO2 and 0.1 M Na2-EDTA in distilled H2O. Dilute 10 times with distilled water before use. 7. Digestion media: 1 U/mL Liberase and 1.2 U/mL DNase I in HBSS. 8. Anti CD45-PerCP-CY5.5 (30-F11) from BioLegend. 9. Anti CD11b-VB510 (M1/70) from BioLegend. 10. Anti Ly6G-APC (1A8) from BioLegend. 11. 0.1 mg/mL DAPI stock solution. 12. Tissue clearing blocking buffer: 0.3% Triton X100, 0.2% BSA, 5% DMSO, 0.1% Azide, 5% serum (serum should match the host of the secondary antibodies) in PBS. 13. Tissue clearing fixing buffer: 4% PFA and 30% sucrose in PBS. 14. Tissue clearing washing buffer: 0.2% Triton X100, 3% NaCl in PBS. 15. BABB: benzyl alcohol, benzyl benzoate (mix 1:2, store at 4  C protected from light). Use always glass containers for BABB. 16. Anti-Citrullinated histone-3: cit-H3 (R2 + R8 + R17), from Abcam. 17. Anti-MPO-biotin (Polyclonal), from R & D (biotinylated with EZ-Link Sulfo-NHS-LC-Biotin from Thermo Fisher). 18. Anti-CD31 (2H8) from Thermo Fisher. 19. Goat-anti-rabbit-AF647 (Thermo Fischer). 20. Goat-anti-hamster-AF405 (Jackson ImmunoResearch). 21. Streptavidin-AF488 (BioLegend). 22. Propidium iodide.

268

3

Tommaso Vicanolo et al.

Methods Below we first describe how to collect the organs from the mice at different zeitgeber times. The organs will then be split in two so that half of the collected organ will be used for flow cytometry and the other half for whole-mount tissue imaging.

3.1 Obtaining Tissue Samples for Paired Analyses

In a typical circadian experiment, we sacrifice a group of mice every 4 h for a total of 6 time points (ZT1, ZT5, ZT9, ZT13, ZT17, and ZT21) in 1 day. Alternatively, mice can be housed in inverted light cycle cabinets so that opposite ZTs can be processed at the same time (see Note 1). It is advisable to use at the very least four mice per group. If the dynamics of neutrophil infiltration to naı¨ve tissues are important for your experiment, consider setting up parabiotic pairs in which most partner-derived neutrophils in the tissues will be neutrophils cleared from blood (see Note 2). Collect organs for FACS analysis and whole-mount imaging in the following steps: 1. Collect blood by cheek-bleeding into an anticoagulant blood collection tube (e.g., EDTA-K blood collection tubes). 2. Euthanize mouse by CO2 asphyxiation. 3. Open the thoracic cavity and expose the heart. Puncture either the right atria or the cava vein and perfuse the mice with 30 mL of PBS or saline using a 30-mL syringe via the left ventricle (see Notes 3 and 4 on how to better discriminate vascular, marginated, and tissue neutrophils, if this is important in your experimental setting). Perfuse with more PBS if needed until the wash out is clear and free of red blood cells. 4. Heart: Remove the heart after perfusion, divide it in two parts (one will be used for flow cytometry and the other for wholemount staining). Weight the section that will be used for flow cytometry and place it in cold PBS (4  C). Keep all tubes on ice unless otherwise specified. Store the half for whole-mount staining in cold PFA 4% in PBS. 5. Lung: Excise two lung lobes, one will be used for FACS analysis and the other for whole-mount staining. Weight the lobe used for FACS analysis and store it in cold PBS. Store the lobe for whole-mount staining in cold PFA 4% in PBS. 6. Liver: Excise two liver lobes, one for FACS analysis and one for whole-mount staining. Weight the lobe used for FACS analysis and store it in HBSS at room temperature. Store the lobe for whole-mount staining in cold PFA 4% in PBS. 7. Spleen: Excise the spleen and cut in half. Weight the half used for FACS analysis and store it in cold PBS. Store the other half for whole-mount staining in cold PFA 4% in PBS.

Analysis of Neutrophil Infiltration Dynamics

269

8. Intestine: Excise the colonic region and flush it from the luminal side with PBS to remove any fecal pellets. Divide it in two parts. Weight the part used for FACS analysis and place it in cold PBS. Store the other part for whole-mount staining in cold PFA 4% in PBS. 9. Kidney: Obtain the two kidneys. Weight the one used for FACS analysis and store in cold PBS. Store the other kidney for whole-mount staining in cold PFA 4% in PBS. 10. Skin: For skin sample we will use the skin of the ear. Sever both ears and remove as much hair as possible, as well as and the proximal region of the ear to avoid fat and cartilage contamination. Weight the ear used for FACS analysis and separate the two skin layers from the proximal part of the ear. Put both skin layer used for FACS analysis in cold PBS and the ones used for whole-mount staining in cold PFA 4% in PBS. 11. Bone marrow and skeletal muscle: Remove skin of both hindlegs. Carefully use a pair of scissors to remove the muscles and cartilages around the femur head until you can dislocate the femur head and remove the leg. Then, carefully separate the femur from the rest of the leg. For that, you can either dislocate the joint or cut between the femur and the tibia, but in that case be careful not to cut the femur. Remove the thigh muscles around the femur and store both the muscle and femur used for FACS analysis in cold PBS. Repeat the procedure for the remaining leg and store the other muscle and femur for wholemount staining in cold PFA 4% in PBS. 3.2 Measuring Neutrophil Infiltration into Tissues by Flow Cytometry

For this method, we first prepare a common antibody mix for all the samples, as well as a final mix for flow cytometry that includes counting beads to estimate absolute number of cells in the tissues, as well as DAPI for viability. We then show how to process and stain all tissues, and how to analyze them in the flow cytometer.

3.2.1 Before You Begin

Prepare the antibody mix for all the samples before starting the experiment and keep on ice protected from light. We use 50 μL of the mix per sample. Prepare the antibody mix using the antibodies (items 8–10) diluted 1/200 in PEB (item 2). To obtain absolute numbers of neutrophils in the tissues by flow cytometry, we make use of counting beads in the final buffer. By diluting your cells in a buffer containing a known concentration of beads, you can estimate the absolute number of a given cell population per organ or per gram of tissue, thereby allowing comparison between samples. We dilute the counting beads in the PEB cytometry buffer (see Subheading 2). We also include DAPI in this final buffer to account for viability.

270

Tommaso Vicanolo et al.

1. Prepare the amount of PEB-DAPI-Beads solution for all the samples by diluting the beads to a final concentration of 10,000 beads/mL in PEB. Add DAPI for viability assessment (DAPI is diluted 1:10,000 from a 0.1 mg/mL stock solution). We also use 36% Percoll in HBSS for leukocyte enrichment in some tissues. We need 7 mL of 36% Percoll per sample. Prepare the volume needed in advance and keep it at room temperature. The next step is to prepare single cell suspensions of all the tissues for subsequent staining. This step takes around 4 h. 3.2.2 Processing the Liver

1. Place the liver onto a 100 μm cell strainer on a 50 mL conical tube. 2. Smash liver through mesh with a syringe plunger and use 10 mL of HBSS to periodically wash the strainer. 3. Centrifuge at 500
g for 5 min at RT. 4. Discard supernatant. 5. Resuspend the pellet in 10 mL of HBSS. We use 5% of the liver suspension (500 μL) for flow cytometry. 6. Add 5% of the liver suspension to 7 mL of the Percoll-HBSS solution in a 15 mL conical tube and mix thoroughly. 7. Spin all samples at 800
g for 30 min at room temperature (with no brake and no acceleration set in the centrifuge). 8. Discard supernatant carefully. Note that hepatocytes will be located at the top in a thin layer and are to be discarded. We only keep the pellet. 9. Resuspend the pellet in 1 mL of PEB. Transfer it to a 1.5 mL Eppendorf tube. 10. Keep on ice ready to be stained. We stain all tissues at once.

3.2.3 Processing the Lung

1. Chop the organs into small pieces with scissors inside 2-mL round-bottom tubes and add 1 mL of digestion solution (see Subheading 2). 2. Incubate samples for 30 min at 37  C to allow enzymatic digestion. 3. Strain the suspension through a 100 μm cell strainer into 50 mL conical tubes and use 10 mL cold PEB to periodically wash the strainer and recover any cells attached to it. Mash any remaining tissue with a syringe plunger adding 5 mL PEB. 4. Centrifuge at 436
g for 5 min at 4  C. 5. Add 500 μL RBC Lysis Buffer to resuspend the pellet and incubate for 5 min on ice. 6. Add 1 mL of PEB to stop the RBC lysis. 7. Centrifuge at 436
g for 5 min at 4  C, discard supernatant.

Analysis of Neutrophil Infiltration Dynamics

271

8. Resuspend the pellet in 1 mL of PEB. 9. Take the fraction used for staining and transfer to a 1.5 mL Eppendorf tube. The fraction used for staining is typically 30% of the starting volume (300 μL). Keep on ice until staining. 3.2.4 Processing the Skin

1. Chop vigorously the two skin layers in a 2-mL round-bottom tube with 1.5 mL of digestion solution. 2. Incubate samples for 1 h and 30 min at 37  C. 3. Use a syringe with a 1.2 mm inner diameter needle to dissociate the digested tissue until no obvious pieces are present. 4. Pass the suspension through a 100 μm cell strainer into a 50-mL conical tube. Mash the remaining tissue with a syringe plunger and use 10 mL of PEB to wash the strainer. 5. Centrifuge at 436
g for 5 min at 4  C. 6. Discard the supernatant and resuspend the pellet with 1 mL of PEB. 7. Take the fraction used for staining and transfer it to a 1.5 mL Eppendorf. The fraction used for staining is typically 10–30% (100–300 μL). Keep samples on ice until staining. Once your cells are ready and prior to feed them to the flow cytometer, filter them again in a 20 μm mesh to remove any possible hair fragments in the sample that may clog the fluidics of the cytometer.

3.2.5 Processing the Spleen

1. Place a 100 μm cell strainer in a 50-mL conical tube, and rinse with 2 mL of PBS. 2. Place the spleen in the cell strainer and grind with a syringe plunger. Use 10 mL of cold PBS to wash the strainer from time to time to recover all cells. 3. Centrifuge at 436
g for 5 min at 4 supernatant.



C, discard the

4. Add 500 μL of RBC Lysis Buffer to resuspend the pellet and incubate for 5 min at room temperature. 5. Add 1 mL of PEB to stop RBC lysis. 6. Centrifuge at 436
g for 5 min at 4  C. 7. Discard the supernatant and resuspend the pellet with 1 mL of PEB. 8. Take the fraction used for staining and transfer it to a 1.5 mL Eppendorf. Usually, the fraction used for staining is 5–10% (50–100 μL). Keep on ice until staining. 3.2.6 Processing the Intestine

1. Cut the colon longitudinally and then in 3–4 pieces with dissecting scissors. Wash them gently in a petri dish with clean HBSS and transfer them to a 50 mL Falcon tube with 20 mL of HBSS + 200 μL of EDTA 0.5 M.

272

Tommaso Vicanolo et al.

2. Incubate the sample for 20 min at 37  C. 3. Vigorously shake the tubes in a vortex to detach most epithelial cells, then transfer colon pieces to a new tube with 5–10 mL cold PBS 1
. Remove all epithelial cells by vortexing as many times as needed (up to 3 times may be required) to get rid of all the floating epithelial cells, changing the samples to a new tube in each wash. Repeat this process 2–3 times. 4. Remove colon pieces from the last wash and place samples in a 2-mL round-bottom tube with 1 mL digestion buffer. Chop the sample until you get a thin paste. 5. Digest samples for 45 min at 37  C. 6. Pass the suspension through a 100 μm cell strainer into a 50-mL tube. Mash the remaining tissue with a syringe plunger and use 10 mL of PEB wash the strainer. 7. Centrifuge at 436
g for 5 min at 4  C. 8. Discard the supernatant and resuspend the pellet with 1 mL PEB. 9. Take the fraction used for staining and transfer it to a 1.5 mL Eppendorf. The usual fraction for staining is 10% of the staring volume (100 μL). Keep on ice until staining. 3.2.7 Processing the Bone Marrow

1. Cut the femoral bone epiphysis. 2. Hold the femur with forceps over a 15 mL tube. 3. Introduce a syringe loaded with 1 mL PEB into the femur through the opening made when cutting the epiphysis. Flush the PEB inside the femur (the bone marrow will fall into the 15 mL tube). 4. Repeat step 3 once or twice to recover as many cells as possible. If interested in cells attached firmly to the endosteal surfaces (such as osteoblasts), see Note 13. 5. Centrifuge at 436
g for 5 min at 4 supernatant.



C, discard the

6. Add 500 μL of RBC Lysis Buffer to resuspend the pellet and incubate for 5 min on ice. 7. Add 1 mL PEB to stop the lysis. 8. Centrifuge at 436
g for 5 min at 4  C. 9. Discard the supernatant and resuspend the pellet with 1 mL PEB. 10. Take the fraction used for staining and transfer it to a 1.5 mL Eppendorf. The usual fraction for staining is 10% (100 μL). Keep on ice until staining.

Analysis of Neutrophil Infiltration Dynamics 3.2.8 Processing the Blood

273

1. Place the blood in a 15 mL tube. Use the same volume of blood for all samples, typically 30 μL. 2. Add 5 mL of RBC Lysis Buffer and incubate for 5 min on ice. 3. Add 5 mL of PEB to stop the lysis. 4. Centrifuge at 436
g for 5 min at 4  C, discard supernatant. 5. Repeat steps 2 to 4 if needed until no red blood cells are obvious in the pellet. 6. Discard the supernatant and resuspend the pellet with 1 mL of PEB and transfer it to a 1.5 mL Eppendorf. Keep samples on ice.

3.2.9 Processing the Heart, Skeletal Muscle and Kidney

1. Place the tissue in 2-mL round-bottom tubes with 1 mL of digestion buffer solution and chop them to little pieces with scissors 2. Digest the sample for 40 min at 37  C. Ensure that large tissue aggregates are digested by manually agitating the mixture every 10 min. 3. After incubation, disaggregate tissues by pipetting up and down using a 1 mL micropipette and transfer the cell suspension to a 40 or 70 μm strainer and into a 50 mL conical tube. 4. To stop the digestion reaction, add 3–4 mL ice-cold PEB while smashing remaining large pieces of tissue with a plunger. Add more PEB to wash the strainer. 5. Centrifuge at 436
g for 5 min at 4 supernatant.



C, discard the

6. Resuspend in 1 mL of PEB. 7. Add the cells to 7 mL of the Percoll-HBSS solution in a 15 mL conical tube and mix thoroughly. 8. Spin all samples at 800
g for 30 min at room temperature (with no brake and no acceleration set in the centrifuge). 9. Discard supernatant carefully. We will only keep the pellet. 10. Add 500 μL of RBC Lysis Buffer to resuspend the pellet and incubate for 5 min on ice. 11. Add 1 mL of PEB to stop the lysis. 12. Centrifuge at 436
g for 5 min at 4  C. 13. Discard supernatants and resuspend the pellet with 1 mL of PEB and transfer it to a 1.5 mL Eppendorf, keep on ice. 3.2.10

Staining

This step details how to stain the single cell suspensions with the previously prepared antibody cocktail for neutrophil quantification: 1. Take the 1.5 mL Eppendorf containing the fractions used for staining.

274

Tommaso Vicanolo et al.

2. Centrifuge all of them, 436
g for 5 min at 4  C, discard the supernatant. 3. Resuspend the pellets in 50 μL of antibody mix as prepared earlier. 4. Incubate for 15 min at 4  C. 5. Wash remaining antibodies by adding 1 mL PEB. 6. Centrifuge at 436
g for 5 min at 4 supernatant.



C, discard the

7. Resuspend pellets in 400 μL PEB-DAPI-Beads (as prepared earlier). 8. Cover the samples from light and immediately analyze them in the flow cytometer. 3.2.11 Data Collection and Analysis

Collect data on a flow cytometer. Unstained samples and singlestained samples are needed to set appropriate photomultiplier voltages and for compensation. Neutrophil population is gated as Cells/Single cells/DAPIneg/CD45+/CD11b+/Ly6G+ (as shown in Fig. 1, red gating route). Counting beads are found in the lower FSC region, so make sure to set the threshold accordingly (see Note 16). To gate them, select the region of the beads (using a tube with no cells, only beads) and from this gate, select the double positive events for APC and FITC (as these beads fluoresce in several channels, see the manufacturer instructions for more information). Record at least 300 beads (gated as shown in Fig. 1, blue gating route) for all samples.

Fig. 1 Gating strategy for flow cytometric analysis of neutrophil numbers in tissues. The red gating route shows the strategy to obtain the number of neutrophils. The blue gating route shows the strategy to obtain the number of counting beads

Analysis of Neutrophil Infiltration Dynamics

275

Counting beads can be used to estimate the number of a given cell population per organ or per g tissue. The use of this approach to estimate changes in a given population instead of using the percentage of the same population within a previous gate (i.e., neutrophils within CD45+ positive cells) avoids artifacts caused by simultaneous changes in other cells within the same previous gate (e.g., macrophages). Once the count for both neutrophils and counting beads is obtained, the number of neutrophils per milligram of tissue can be calculated as follows: Cells Neutrophil count 10, 000 beads ¼  mg mL Bead count 

0:4 mL 1  Fraction of tissue analyzed Organ weight ðmgÞ

For blood and bone marrow, the absolute number of cells is presented as cells per mL of blood or per femur (considering and correcting for the fraction used for analysis). When comparing circadian infiltration to the tissues, it may be useful to normalize the values in the tissues by the number of neutrophils in blood, to account for differences in blood counts, and when comparing different tissues, it may also be useful to normalize the highly different values (i.e., there are orders of magnitude more neutrophils in the lung or bone marrow than in the skin, for instance) to ZT5 or any other Zeitgeber Time of choice to allow for easier comparison between tissues. 3.3 Whole-Mount Tissue Clearing and Immunofluorescence Staining

With half of the collected tissues we can perform whole-mount immunofluorescence staining and tissue clearing. This will allow to examine the spatial distribution of those neutrophils at different times of day. Here, we will stain for neutrophils, neutrophil extracellular traps (NETs, as an example of a neutrophil effector function) and blood vessels, to quantify neutrophils infiltrating the tissue and deploying NETs, but the technique can be adapted to assess other neutrophil effector functions, other immune cells, or the localization of structures other than blood vessels. See Fig. 2 for sample images of different tissues with specific projection methods. Our clearing process has been successfully tested in different organs, such as lung, brain, fat, heart, intestine, kidney, liver, bone marrow, or skeletal muscle.

3.3.1 Tissue Immunofluorescence Staining and Clearing: Day 1—Tissue Permeabilization and Blocking

1. Fix the samples in tissue clearing fixing buffer (see Subheading 2) for 2 h with shaking (see Note 5). 2. Wash three times in PBS at room temperature for 1 h with shaking. For the bone marrow, the bone must be decalcified at this step (see Note 6). All the other organs can continue immediately.

276

Tommaso Vicanolo et al.

Fig. 2 Tissue clearing image samples. (a) Maximum intensity projection of a heavily inflamed lung stained for neutrophils (cyan), DNA (green), citrullinated histone 3 (red), and blood vessels (gray). (b) Blended projection of a lung stained for blood vessels (gray), MPO (cyan), and DNA (blue). (c) Maximum intensity projection of a cleared colon stained for laminin (gray) and neutrophils (green). (d) Blended projection of a cleared skeletal muscle stained for vessels (red) and neutrophils (green)

3. Permeabilize in Methanol gradients in PBS, all at room temperature with shaking. (a) 30 min in Methanol 50% in PBS. (b) 30 min in Methanol 80% in PBS. (c) 30 min in Methanol (100%). (d) 30 min in Methanol with 20% DMSO (see also Note 7). (e) 30 min Methanol 80% in PBS. (f) 30 min Methanol 50% in PBS. (g) 30 min in PBS. 4. Block in tissue clearing blocking solution overnight at 4  C with shaking. 3.3.2 Tissue Immunofluorescence Staining and Clearing: Day 2 and 3—Staining with Primary Antibodies

1. Stain the tissues in tissue clearing blocking solution with 1:100 primary antibodies (approximately 2 μg/mL) for 48 h (see Note 8). Volume needed will depend on the size of the tissue fragment. It should be completely covered for at least twice its height. We stained NETs with antibodies against

Analysis of Neutrophil Infiltration Dynamics

277

myeloperoxidase (MPO, will also stain intact neutrophils) and citrullinated histone 3 (citH3), and further stained CD31 for blood vessels (see Subheading 2). The latter can be changed to reveal any other structure of interest. On the selection of antibodies (see Note 9). 3.3.3 Tissue Immunofluorescence Staining and Clearing: Day 4 and 5—Washing and Staining with Secondary Antibodies

1. Wash three times in tissue clearing washing buffer for 2 h at room temperature.

3.3.4 Tissue Immunofluorescence Staining and Clearing: Day 6—Washing

1. Add propidium iodide (1:1000) or any preferred and compatible DNA dye (such as sytox red, for instance, depending on your secondary antibodies) to the tubes and let it stain for 60 min with shaking.

2. Stain for 48 h with secondary antibodies (see Subheading 2) 1: 500 in tissue clearing blocking solution without DMSO at 4  C, with shaking (see also Note 10).

2. Wash twice for 15 min in tissue clearing washing buffer (change buffer every time). 3. Leave washing overnight in tissue clearing washing buffer. 3.3.5 Tissue Immunofluorescence Staining and Clearing: Day 7—Tissue Clearing

1. Clear the tissue always in glass containers (we use 4 mL glass vials, see Note 11). For that we dehydrate the tissue first in Methanol gradients as in step 1d, except now gradients should be made in deionized water instead of PBS. Finally, we will clear the tissue using BABB (Benzyl-Alcohol:Benzyl-Benzoate, see Subheading 2). (a) 30 min in Methanol 50% in dH2O. (b) 30 min in Methanol 70% in dH2O. (c) 30 min in Methanol 95% in dH2O. (d) Three times 30 min in Methanol (100%). (e) 30 min in BABB 50% in Methanol. (f) 30 min in BABB. 2. At this point the tissue should be transparent and ready to image (otherwise see Note 12).

3.3.6 Image Capture

1. To capture microscopy images, samples should be first mounted for imaging. Keep in mind that BABB must not spill to the microscope. We use 35 mm glass bottom dishes for imaging in an inverted microscope. (a) Place the cleared tissue in the well of the dish. (b) Add enough BABB to cover the sample and carefully but firmly place a coverslip on top to keep the tissue in place.

278

Tommaso Vicanolo et al.

(c) Confirm that no air bubbles are left in the imaging well. If bubbles are formed, remove the coverslip, add more BABB to the well. and then place the coverslip again. 2. We recommend capturing a large volume of the cleared tissue. For visualization of NETs we used an 20
objective. A large working-distance objective would be preferable to capture the whole depth of the tissue, but if none is available, then just capture a tile-scan of the desired area with the maximum Z-depth possible in the system. 3.3.7 Quantification Pipeline: Quantification of Total Neutrophils and Infiltrating Neutrophils

Here, we use Bitplane Imaris for analysis, but the pipeline can be adapted to the imaging software of choice. Open Imaris and load the first image dataset. 1. First, visually inspect the image carefully. Depending on the microscope setup, one common problem is the artificial z-axis elongation (i.e., the z dimension is elongated compared to the x and y dimensions). This must be taken into account for otherwise it may lead to serious overestimation of the quantifications. To assess if that is the case, and easy way is to locate a neutrophil in the image, zoom in on it and then turn the camera to visualize it sideways. If it looks reasonably spherical, then no further action is needed. If it looks elliptical, highly enlarged in the Z dimension compared to the X and Y dimensions, then for all subsequent quantifications using the Spots tool in Imaris, PSF-elongation modeling along the Z-axis must be used. 2. To quantify the total number of neutrophils, use the Spots tool on the MPO channel, with an estimated x–y diameter around 7 μm. 3. To quantify infiltrating neutrophils, first create a Surface of the whole vasculature using the CD31 channel. Make sure it encompasses all the vessels (see Note 14). 4. Then, mask the MPO channel using the newly created vasculature surface, setting all pixels inside the surface to zero. 5. Finally, use the Spots tool to quantify the number of neutrophils present outside blood vessels (also see Note 15). 6. If you wish to compare different conditions or mice, these numbers should be normalized by the total volume of the image, in case it differs from mouse to mouse.

3.3.8 Quantification Pipeline: Quantification of NETs

1. To quantify bona fide NETs, we recommend using the triple colocalization criteria of the MPO, citH3 and DNA channels. To do this in Imaris, we first create a colocalization channel of citH3 with MPO.

Analysis of Neutrophil Infiltration Dynamics

279

2. Next, perform a second colocalization of the channel output from the previous step and the channel for DNA. With that, we obtain a triple-colocalization channel of MPO + citH3 + DNA containing the NETs. 3. Using the Spots tool of Imaris, locate the NETs in the threedimensional space. 4. Finally, note the number of detected NETs and normalize it by the total number of neutrophils present in that same volume of tissue. 3.3.9 Quantification Pipeline: Measuring Spatial Relationships

1. To measure the distance between objects in our images we can also use Imaris. We exemplify here the procedure measuring the distance between neutrophils and endothelial cells; however this procedure can be used to calculate the distance between any two cell types in out images. Start by opening the dataset in Imaris. 2. Convert the image to a 32-bit floating-point dataset: select Edit > Change data type and select 32-bit float (this will make sure floating-point numbers, i.e., non integer numbers, are allowed). 3. Create a surface of the object(s) to which you wish to measure the distance. In our case the CD31+ cells. Use the Surfaces tool of Imaris to create a surface that encompasses your target structures. 4. Use the Distance Transformation tool on your surface (located in the “Tools” tab when you select the target surface). Choose “outside surface” as we want to measure distances in this region. This creates a new channel whose “intensity” values are the distance to the surface in the three-dimensional space. 5. Use the Spots tool in Imaris to localize the neutrophils in the dataset. For each calculated spot we can now inspect the mean “intensity” of the newly created distance transformation channel. This value will be the mean distance of the spot to the surface, which can be exported and further analyzed.

3.4 Amplitude vs. Zero Test to Statistically Validate Circadian Behavior

Once circadian data is gathered for different time points, one common hurdle is how to analyze whether the parameter under study shows bona fide circadian behavior or not. It is common to see data analyzed comparing two arbitrarily chosen ZT points (usually peak and trough) but this approach is problematic as it takes into account only those arbitrarily chosen points and not the overall shape of the circadian curve. To avoid this bias, we propose a simple statistical method referred to as “amplitude vs. zero test” [7, 16] that compares the COSINOR-based [17] curve-fitted amplitude of the data and its standard error, with a hypothetical zero-amplitude curve that assumes that both curves have identical errors. See Fig. 3 for mock data as an example. Note that both data sets in Fig. 3 are

280

Tommaso Vicanolo et al.

Fig. 3 Amplitude-vs.-zero test examples. Both circadian plots (a and b) show significant differences ( p < 0.001) when comparing just peak and trough. When comparing the amplitude with our proposed test, data distribution in (a) is also significative ( p < 0.001) while the data in (b) is not

highly significant when comparing only peak and trough, even though the rest of the data points in Fig. 3b are arrhythmic. Hence, we propose that this method is more relevant to statistically define the circadian behavior of a dataset. For the COSINOR non-linear curve fitting we will use the following equation: y ¼ B þ A  cos ðF  x þ P Þ where B is the baseline defined as the average of the y maximum (Ymax) and minimum (Ymin) value; A is the amplitude defined as 0.5
(Ymax  Ymin); F is the frequency fixed as 2
π 24 ¼ 0:2618; and P is the phase-shift defined as the value of x at Ymax.This is easily performed using GraphPad Prism, and that is the software used below, though the method can be easily adapted to the statistical software of choice. 1. In GraphPad Prism, get the data and all its replicates in an XY table, setting the different times sampled in the X column and the replicates for each timepoint in the Y column.

Analysis of Neutrophil Infiltration Dynamics

281

2. On this data we perform a non-linear regression using an userdefined explicit equation named COSINOR that we’ll define as: Y ¼ Baseline + Amplitude*cos (Frequency*X + Phaseshift). With the “Rules for initial values” defined as: (a) Baseline: 1*YMID. (b) Amplitude: 0.5*(YMAX  YMIN). (c) Frequency: 0.2618. (d) Phaseshift: 1*(Value of X at YMAX). And with the following “Default constrains”: (a) Phaseshift: “Constant or equal to” 0.2618. 3. The output of the fitting provides the best-fit values for the amplitude (see Note 17), as well as its standard error. 4. Perform an unpaired t-test comparing the best-fit value of amplitude and its standard error to a hypothetical curve with 0 amplitude but with the same error as your original data. The obtained p-value would be the “amplitude vs. zero test” p-value (also see Note 18).

4

Notes 1. Light cabinets have mice exposed to inverted/shifted light cycles compared with the light regime in the rest of the animal facility. After 3 weeks of adaptation to the new light regime, mice adjust to the new light-dark cycle and adjust their circadian rhythms. The use of light cabinets reduces the time of the circadian experiment to half, as it allows organ extraction from mice at the opposite ZTs (ZT1 and ZT13; ZT5 and ZT17; ZT9 and ZT21) at the same time. 2. To assess the dynamics of neutrophils infiltration in multiple tissues, parabiotic strategy could be exploited. Set up parabiotic pairs [18] of congenic CD45.1 and CD45.2 WT mice and analyze by flow cytometry the frequency of partner-derived neutrophils in tissues. Partner-derived neutrophils can be distinguished by host neutrophils with the use of specific antibodies recognizing CD45.1 and CD45.2. Another possibility is the use neutrophil reporter mice (such as Lyz2tm1.1Graf with GFP-expression in granulocytes and macrophages [19]) for the parabiotic pair in order to analyze by both cytometry and imaging the frequency and location of partner-derived neutrophils in tissues. 3. It is critical to perfuse all blood out, otherwise there will be contamination by circulating blood cells in the tissue.

282

Tommaso Vicanolo et al.

4. In order to better discriminate intravascular and tissueinfiltrating neutrophils by flow cytometry, additional steps can be performed before proceeding with the protocol (adapted from [8]): (a) Inject fluorochrome-conjugated antibody against Ly6G (e.g., anti-Ly6G APC-conjugated) intravenously in WT mice and allow recirculating for 5 min. (b) Sacrifice mice and perfuse to remove blood. (c) Recover and process the tissues as explained in the protocol and label the neutrophils for flow cytometry with another anti-Ly6G conjugated antibody (e.g., with an anti-Ly6G antibody phycoerythrin-conjugated). Neutrophils with the highest levels of APC labeling (or the conjugated antibody used for in vivo labeling of blood neutrophils) can be considered intravascular, whereas those with intermediate levels are featured by tissues with a fenestrated endothelium and are determined to be mostly extravascular by alternative methods. Neutrophils negative for APC are considered extravascular in this setting [8]. 5. The tissues can be safely left to fix overnight, but this is not mandatory. 6. For the bone marrow, femurs must be decalcified in a PBS solution with EDTA 0.25 M at 4  C for 15 days, renewing the media every 2 days. Once femurs are decalcified, they must be cut longitudinally in half under a binocular dissection microscope using a thin razor (we used cryostat razors) in a single, continuous cut. Alternatively, small transversal cuts can also be performed if transversal visualization is preferred. In any case, this ensures that the antibodies penetrate the bone marrow. 7. If the sample shows excessive autofluorescence, this step can be replaced with Dent’s bleach, a solution of 15% H2O2, 16.7% DMSO in MetOH. Keep the samples for 1 h with shaking in this solution. 8. All incubation times noted should work, but incubation time depends on the size of the tissue and how compact it is. In general, if penetrance of the antibody was not complete (which will be seen as a bright staining on the outer regions and lack of staining deeper into the tissue), then the incubation times should be increased. 9. In our experience, the trickiest part of the procedure is to find good primary and secondary antibodies that work in the conditions noted. Generally speaking, it is a good idea to do a battery of tests prior to any experiment to ensure coherent staining of every antibody. We recommend to extract lungs

Analysis of Neutrophil Infiltration Dynamics

283

from one mice and after fixation chop the lungs in several small pieces to stain each one with all possible combinations of primary and secondary antibodies available, as well as secondary staining controls, and so on. 10. Secondary antibodies can be also tricky in tissue-cleared samples. One possibility to keep in mind if the secondary antibodies are not working correctly is to use a Biotinylating kit with the most problematic primary antibody, and then use a streptavidin instead of a secondary antibody on that one. 11. BABB is highly toxic, please use caution and read the safety data sheets of both compounds. BABB may break plastic away, so always use glass vials. Store protected from light at 4  C. As a less toxic alternative, ethyl cinnamate can be used instead of BABB. 12. If the tissue is not clear enough (it should be totally transparent, not necessarily colorless, but transparent), then keep it in BABB 100% for longer, changing the medium every 30 min until clear. Opacity could be due to residual water in the sample. If it still does not clear upon 3 BABB changes, go back to the step with 100% Methanol and repeat from there to ensure complete dehydration of the sample. 13. If interested in cell populations that may be attached to the endosteal surfaces, such as osteoblasts, once the bone marrow is ejected from the bone, incubate the bone with digestion buffer for 30 min at 37  C and then flush again with PEB to recover those cells. To this end, get the bone inside a 1.5 mL Eppendorf tube and cover it with digestion buffer, then using a 1 mL syringe, carefully load more digestion buffer inside the bone and incubate at 37  C. 14. If the CD31 signal is weak or the automated surface generation process is unable to correctly select the vasculature, manual surface creation may be necessary. Refer to Imaris manual for help on how to perform manual surface creation. 15. A similar procedure can be used to isolate the neutrophils inside blood vessels (i.e., in Subheading 3.3.7, step 4, set the pixels outside the surface to zero). 16. True count beads are in the lower FSC region. If your cytometer has a threshold function for FSC, you may need to lower the threshold in order to detect the beads. 17. Note that COSINOR will try to force-fit the data to a cosine wave function (unless this is clearly impossible). Thus, a COSINOR fit alone is not an indication that the parameter has a circadian behavior. 18. The fitted curve is also useful to show in representations. We recommend plotting individual values for each timepoint

284

Tommaso Vicanolo et al.

(or alternatively the mean with the standard error of the mean) together with the calculated COSINOR curve. In our experience this adds clarity to the circadian plot. References 1. Scheiermann C, Gibbs J, Ince L, Loudon A (2018) Clocking in to immunity. Nat Rev Immunol 18(7):423–437 2. Haus E, Smolensky MH (1999) Biologic rhythms in the immune system. Chronobiol Int 16:581–622 3. Scheiermann C, Kunisaki Y, Lucas D, Chow A, Jang JE, Zhang D, Hashimoto D, Merad M, Frenette PS (2012) Adrenergic nerves govern circadian leukocyte recruitment to tissues. Immunity 37:290–301 ´ vila JA, Hidalgo A 4. Adrover JM, Nicolás-A (2016) Aging: a temporal dimension for neutrophils. Trends Immunol 37:334–345 5. Summers C, Rankin SM, Condliffe AM, Singh N, Peters AM, Chilvers ER (2010) Neutrophil kinetics in health and disease. Trends Immunol 31:318–324 6. Casanova-Acebes M, Pitaval C, Weiss LA, Nombela-Arrieta C, Che`vre R et al (2013) Rhythmic modulation of the hematopoietic niche through neutrophil clearance. Cell 153: 1025–1035 7. Adrover JM, del Fresno C, Crainiciuc G, Cuartero MI, Casanova-Acebes M et al (2019) A neutrophil timer coordinates immune defense and vascular protection. Immunity 50:390– 402.e10 ´ vila JA, Li JL, 8. Casanova-Acebes M, Nicolás-A Garcı´a-Silva S, Balachander A et al (2018) Neutrophils instruct homeostatic and pathological states in naive tissues. J Exp Med 215:2778– 2795 9. Ballesteros I, Rubio-Ponce A, Genua M, Lusito E, Kwok I et al (2020) Co-option of neutrophil fates by tissue environments. Cell 183:1282–1297.e18 10. Aroca-Creville´n A, Adrover JM, Hidalgo A (2020) Circadian features of neutrophil biology. Front Immunol 11:576

11. Puga I, Cols M, Barra CM, He B, Cassis L et al (2012) B cell-helper neutrophils stimulate the diversification and production of immunoglobulin in the marginal zone of the spleen. Nat Immunol 13:170–180 12. Deniset JF, Surewaard BG, Lee WY, Kubes P (2017) Splenic Ly6Ghigh mature and Ly6Gint immature neutrophils contribute to eradication of S. pneumoniae. J Exp Med 214:1333–1350 13. Athens JW, Haab OP, Raab SO, Mauer AM, Ashenbrucker H, Cartwright GE, Wintrobe MM (1961) Leukokinetic studies. IV. The total blood, circulating and marginal granulocyte pools and the granulocyte turnover rate in normal subjects. J Clin Invest 40:989–995 14. Devi S, Wang Y, Chew WK, Lima R, A-González N et al (2013) Neutrophil mobilization via plerixafor-mediated CXCR4 inhibition arises from lung demargination and blockade of neutrophil homing to the bone marrow. J Exp Med 210:2321–2336 15. Yipp BG, Kim JH, Lima R, Zbytnuik LD, Petri B et al (2017) The lung is a host defense niche for immediate neutrophil-mediated vascular protection. Sci Immunol 2:eaam8929 16. Adrover JM, Aroca-Creville´n A, Crainiciuc G, Ostos F, Rojas-Vega Y et al (2020) Programmed ‘disarming’ of the neutrophil proteome reduces the magnitude of inflammation. Nat Immunol 21(2):135–144 17. Cornelissen G (2014) Cosinor-based rhythmometry. Theor Biol Med Model 11:16 18. Kamran P, Sereti KI, Zhao P, Ali SR, Weissman IL, Ardehali R (2013) Parabiosis in mice: a detailed protocol. J Vis Exp (80):50556 19. Faust N, Varas F, Kelly LM, Heck S, Graf T (2000) Insertion of enhanced green fluorescent protein into the lysozyme gene creates mice with green fluorescent granulocytes and macrophages. Blood 96:719–726

Chapter 19 In Vivo Imaging of Circadian NET Formation During Lung Injury by Four-Dimensional Intravital Microscopy Alejandra Aroca-Creville´n, Andres Hidalgo, and Jose M. Adrover Abstract Neutrophil extracellular traps (NETs) are toxic extracellular structures deployed by neutrophils in response to pathogens and sterile danger signals. NETs are circadian in nature as mouse and human neutrophils preferentially deploy them at night or early morning. Traditionally, NETs have been quantified using a plethora of methods including immunofluorescence and ELISA-based assays; however few options are available to visualize them in vivo. Here we describe a method to directly visualize and quantify NET formation and release in the microvasculature of the lung using intravital imaging in a model of acute lung injury. The method allows four-dimensional capture and quantification of NET formation dynamics over time and should be a useful resource for those interested in visualizing neutrophil responses in vivo. Key words Circadian, NETs, TRALI, Lung, Injury, Intravital microscopy, Microvasculature

1

Introduction Neutrophil extracellular traps (NETs) are filamentous web-like structures of DNA coated with histones and granule-derived cytotoxic proteins, deployed by neutrophils in response to supernumerary [1] or oversized [2] pathogens, in most cases causing the lytic death of the neutrophil in a type of cell death termed NETosis [3]. NETs ensnare and kill bacteria [1] and fungi [4], and also can inactivate viruses [5]. In addition, these structures have become widely recognized for their capacity to harm healthy tissues if deployed in the wrong place or at the wrong time, as the granulederived proteins they contain (such as cationic antimicrobial peptides (CAMPs), myeloperoxidase (MPO) and neutrophil elastase (NE)) are highly cytotoxic [6], as are the histones [7–9] that are released with the NET. Thus, the uncontrolled deployment of NETs is harmful to the host and is involved in a whole array of pathological states [10, 11], including systemic lupus erythematosus [12–14], diabetes [15], liver damage during sepsis [16],

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_19, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

285

286

Alejandra Aroca-Creville´n et al.

endothelial damage and thrombosis [17–20], atherosclerosis [21– 23], cancer [24, 25], and acute lung injury (ALI)—acute respiratory distress syndrome (ARDS) [26–30]. Neutrophils display marked circadian patterns both in number [31], phenotype and function [32–34] that are thought to anticipate infections by synchronizing the peak of neutrophil activity with the highest likelihood of pathogen encounter, which is likely a major driving force for immune evolution [35]. In mice, neutrophils are produced in the bone marrow and released to the bloodstream at night [32, 34]. This freshly released (sometimes termed “fresh” or “young”) neutrophils change over the course of the day to an “aged” or “senescent” phenotype under control of the molecular clock [34]. In this process, neutrophils change their trafficking properties, as well as their transcriptomic and proteomic profiles [26, 34]. The notion of circadian changes in neutrophil activity is recent, and much remains to be explored on its impact pathophysiology. For example, we have recently reported that one of the neutrophil properties that shows circadian fluctuations is their capacity to form NETs: “fresh” neutrophils (sampled at ZT13, or Zeitgeber Time 13, in a 12 h:12 h light/dark cycle) form NETs more efficiently than their “aged” counterparts (sampled at ZT5), both in ex vivo and in vivo models of ischemia/reperfusion and transfusion-related acute lung injury (TRALI) [26]. In these studies, we developed a method that allowed us to quantify circadian NET formation through direct visualization of the pulmonary microvasculature at different times of the day by intravital imaging. In our studies, we used sterile TRALI as a model of rapid and efficient pulmonary inflammatory responses mediated by neutrophils; however the method should be adaptable to different injury models of ALI/ARDS, infection with both bacterial and viral pathogens, or cancer including metastatic spread in the lung. It should also allow to interrogate the dynamics and spatial distribution of NETs formation, and allow quantification of different NET-formation modalities, such as NETs that are rapidly washed away by the bloodstream upon release (“flowing NETs” [26]), which cannot be scored by static immunofluorescence of fixed tissues.

2

Materials 1. 7–12 weeks old male BALB/C mice from Charles River. 2. LPS: Resuspend LPS to obtain a 5 mg/mL stock solution. From this, prepare a fresh working solution to inject 0.1 mg/ kg to the animals. 3. 25 G Needles and 1 mL syringes.

Intravital Imaging of NET Formation in the Lung

287

4. Anti-H2d antibody (Ref. BE0180, BioXcell): Resuspend 2 mg of antibody in 1 mL of sterile PBS for a 2 mg/mL stock solution and store it at
80  C in small aliquots. Avoid freezing and thawing. Prepare a fresh working solution to inject 1 mg/ kg. 5. 0.5 mL Insulin syringes; 0.33 mm (29G)  12.7 mm. 6. Timer. 7. Sterile PBS: NaCl 137 mM, KCl 2.7 mM, Na2HPO4 10 mM, KH2PO4 1.8 mM in distilled water; pH 7.4. 8. Anesthetics cocktail: 5 mg/mL of ketamine + 0.05 mg/mL of medetomidine hydrochloride in sterile PBS. 9. Electric shaving razor. 10. Custom surgical board. 11. 25 mm  9.1 m Surgical tape. 12. Antiseptic: chlorhexidine. 13. Surgical suture thread (4/0). 14. Sterile dissecting scissors. 15. Sterile Castroviejo scissors. 16. Surgical tweezers. 17. Surgical tweezers with curved tips. 18. Cauterizer. 19. 20 G catheter. 20. Intravital thoracic window. 21. Spinning disk confocal microscope. In our case, this is a VIVO system built by 3i (Intelligent Imaging Innovations, Dever, CO) upon an Axio Examiner Z.1 workstation (Zeiss, Oberkochen, Germany) and mounted on a 3D motorized stage (Sutter Instrument, Novato, CA). Our system is equipped with a CoolLED pE widefield fluorescence LED light source (CoolLED Ltd. UK) and a quad pass filter cube with Semrock Di01-R405/488/561/635 dichroic and FF01-446/523/ 600/677 emitter. A plan-apochromat 20 W NA1.0 objective (Zeiss) and a CoolSnap HQ2 camera (Photometrics, Tucson, AZ) are used for image capture. Our system runs on a Dell Precision T7500 computer system using the SlideBook software (3i). 22. Chest drain with wet suction. 23. 0.5 mg/mL Anti-mouse Ly6G AF647 (Ref. 127610, BioLegend) for neutrophil labeling. 24. 0.2 mg/mL Anti-mouse CD41 PE (Ref. 12-0411-83, eBioscience) for platelet labeling. 25. 5 mM Sytox Green.

288

Alejandra Aroca-Creville´n et al.

26. Cl-amidine (Ref. 10599, Cayman NET-inhibition, used 12 mg/kg.

Chemical)

for

27. Small animal ventilator. 28. Sterile petroleum jelly.

3

Methods To visualize circadian NET formation in the lung microvasculature, we need a model of injury and a means of visualizing the target organ. We will first describe here the model of lung injury we chose (TRALI) and then we will discuss how to perform the intravital imaging. We finally detail how to visualize and quantify NETs in this setting.

3.1 TRALI as a Model of Acute Lung Injury Dependent on Neutrophils and NETs

TRALI is a neutrophil-driven lung injury syndrome that is caused by transfused blood products typically containing antibodies to human leukocyte antigens (HLAs) [36]. Here we use a two-step model of injury [37] that produces rapid neutrophil infiltration, NETs formation, and edema in the lungs [26–28, 38], leading to frequent death in mice over the next 2 h post-induction (see Note 1). The first step consists of baseline immune priming with low dose LPS 24 h prior to the disease induction, which is triggered upon challenge of the animals with an anti-MHC Class I monoclonal antibody (BE0180, BioXcell). We designed this protocol to identify circadian differences in the outcome of the disease by analyzing the survival of mice upon TRALI induction at different circadian times: ZT5 (11 am CET, daytime) and ZT13 (7 pm CET, early nighttime) (see Note 2). Survival is higher at daytime, when the ability of neutrophils to produce NETs is reduced, whereas the increased NET formation found during nighttime accounted for the increased mortality observed at ZT13 [26]. Male BALB/c mice are used for this protocol [39]. We house them in specific pathogen-free (SPF) conditions. All experimental procedures involving mice are approved by the Animal Care and Ethics Committee of CNIC and the regional authorities. Animals must be continuously watched for 2 h upon anti-MCH-I antibody injection and any mouse showing excess distress should be immediately sacrificed. At the end of the 2 h, all mice subject to TRALI should be sacrificed if they are not to be observed for longer, as this is a severe procedure. Please, consult your local veterinarian for guidance if necessary. What follows is the procedure to obtain the survival curves of the mice in response to TRALI performed at different times of the day. This should be done prior to the imaging experiments to confirm that circadian differences are found, or the effect of any test compound.

Intravital Imaging of NET Formation in the Lung

289

1. Treat mice with an intraperitoneal injection of 0.1 mg/kg LPS using a 25 G needle 1 mL syringe 24 h prior to the administration of 1 mg/kg anti-MHC-I antibody, to induce immune priming (see Note 3). 2. Distribute and position the mice in such a way that their behavior is easy to observe, in cages by groups. Prepare the timer. 3. Administer intravenously 1 mg/kg of anti-H2d antibody using an insulin syringe (see Notes 4 and 5). 4. Follow the behavior of the mice up to 2 h after the antibody injection. Whenever a mouse shows signs of distress or suffering, immediately euthanize the animal according to the method determined by the veterinarian authority guidelines (see Note 6). 5. Record the time of death and obtain Kaplan-Meyer survival curves (see Note 7). This can be easily done in GraphPad Prism or similar software. 3.2 Method for Intravital Imaging of the Lung

3.2.1 Surgical Procedure for Tracheotomy

To visualize the lung microvasculature in vivo we employ a reported model of stabilized lung intravital imaging [40]. Using this protocol, we observe circadian differences in NET-formation in the lung microvasculature by high-speed multichannel four-dimensional intravital imaging [26]. For that, we induce TRALI as explained in the previous section, in an anesthetized and intubated mouse that was previously prepared for intravital imaging of the lung, with a window open to the thorax though which the lung is readily observable. As mentioned above, the neutrophil’s NET-formation capacity is higher at ZT13, correlating with higher inflammatory injury in the lungs of TRALI-induced mice. Observation of NETs in this model by intravital microscopy will therefore be more evident at nighttime (see Note 2). Male BALB/c mice maintained in specific pathogen-free conditions are used for this protocol. All experimental procedures involving mice were approved by the Animal Care and Ethics Committee of CNIC and the regional authorities. 1. Anesthetize mice by injecting intraperitoneally 10 μL of the anesthetic cocktail per gram of body weight using a 1 mL syringe with a 25 G needle. 2. Ensure that the animal is deeply anesthetized and shave the upper abdomen, neck, right forelimb, and part of the back with an electric razor so that there is no hair covering the trachea or the right ribcage area. 3. Place the animal in supine position on the surgical board.

290

Alejandra Aroca-Creville´n et al.

Fig. 1 Steps for 4D-IVM of murine lungs. Schematic representation of different sequential steps for visualization of NET formation in the lung microvasculature: (a) positioning of the anesthetized mouse in the surgical board; (b) longitudinal incision in the neck; (c) removal of the salivary glands; (d) surgical exposition of the trachea; (e) insertion of the surgical thread; (f) sectioning of the trachea to create a hole; (g) intubation with a catheter; (h) repositioning of the anesthetized mouse in the surgical board; (i) incision in the chest; (j) exposition of the ribs; (k) sectioning of the ribs to create a hole in the thorax; (l) surgical window placement

4. Fix the forelimbs, the hindlimbs, and the tail with surgical tape as shown in Fig. 1a. Use a surgical silk thread behind the frontal teeth (pull and fix the silk thread with tape to the surgical board) to fix the head and expose the neck for surgery, as seen in Fig. 1a. Apply an antiseptic solution to the neck area. 5. Perform a longitudinal incision on the neck from just below the mandible to just above the collar bone (Fig. 1b).

Intravital Imaging of NET Formation in the Lung

291

6. Move the skin and the submandibular salivary glands to the sides to expose the area of the trachea (see Note 8 and Fig. 1c). 7. Separate the tracheal muscle and pull it to the sides with the tweezers. This will expose the trachea (Fig. 1d). 8. Pass a surgical silk thread of approximately 5 cm behind the trachea using surgical tweezers with curved tips and pull from both sides until it is placed in the lower part of the exposed trachea. Allow at least half the length of the thread free on each side (Fig. 1e). 9. Hemisect the cricoid cartilage with a pair of scissors without completely cutting the trachea and insert a 20 G catheter through the hole (see Note 9 and Fig. 1f). 10. Tie the catheter and trachea with the previously laid suture thread by making a double knot. Make sure to knot the thread over the catheter to ensure that it will not disengage, as we will need to reposition the mouse later (Fig. 1g). 11. Confirm that the mouse is breathing normally. 3.2.2 Surgical Procedure for Lung Exposure

1. Reposition the animal and place it in left lateral decubitus position. Fix the forelimbs together near the upper edge of the surgical customized board and the paws apart with surgical tape (see Note 10 and Fig. 1h). Apply antiseptic solution to the skin surrounding the ribs. Make sure that the animal is properly anesthetized, and make sure to maintain deep anesthesia for the duration of the procedure. Consult your local veterinarian if in doubt. 2. Perform an incision in the skin from the lower thorax toward the right forelimb, as shown in Fig. 1i (white dashed line). 3. Open the skin and dissect the underlying muscle to expose the ribs (see Note 11). When pulling the skin, try to localize and cauterize the main vessels in the area. When removing the muscle layers below the skin, make sure to find and cauterize the blood vessels prior to removing the muscle (Fig. 1j). When you are ready to cut the muscle out to expose the ribs, have the cauterizer at hand and ready in order to stop any potential bleeding. 4. If the small animal ventilator is located in the microscope area, transport the surgical board with the animal to the microscope. We will need to connect the mouse to the ventilator now as the thoracic cavity will be open next. If using a mobile ventilator system just connect the mouse to the ventilator now. 5. Connect the catheter installed after the tracheotomy to a small animal ventilator with an established breath ratio of 115 breaths per minute to maintain breathing while opening the thoracic cavity.

292

Alejandra Aroca-Creville´n et al.

6. The lung intravital window should now be connected to a vacuum pump providing 20–25 mmHg suction. This will help maintain the lung in place and minimize motion artifacts. This suction can be achieved with a vacuum pressure reduction device, or with a surgical chest drain apparatus. In the next step, we will open the chest and the window should be ready to be quickly secured to the area. 7. Hold the third rib with fine pointed tweezers and cut on both sides with fine tipped scissors. Cut the intercostal muscle between the severed rib and its neighbors to create a 4–5 mm square hole in the thorax (Fig. 1k, and see Notes 12 and 13). This step is critical, and the lung should not be punctured. Use an angle almost parallel to the ribs when using the scissors. 8. Place the surgical window parallel to the opening in contact with the skin (Fig. 1l) and ensure the return of the thoracic negative pressure by adding sterile petroleum jelly around the surgical window covering any possible air gap between the window and the skin (completely covering the yellow-shaded area in Fig. 1l, also see Note 14). When the vacuum in the thoracic cavity is restored, you should see the lung come in contact with the glass of the window, and should stay in place for the whole procedure. Make sure that any air gaps are sealed with petroleum jelly. 9. The mouse is now ready for lung intravital imaging. Add a drop of water on top of the glass of the surgical window and focus on the lung microvasculature (we used a 20/1.0 water immersion objective). 3.3 In Vivo Staining and Quantification of NET Formation in the Lungs Upon Acute Lung Injury

This method allows in vivo visualization of NETs being formed in real time in the lung microvasculature of mice subject to a model of acute lung injury performed at different Zeitgeber times. It allows the measurement of the dynamics of NET formation over time as disease progresses and to quantify different modalities of NET-formation, which would be lost in a static scenario such as fixed-tissue immunofluorescence. One example of this are “flowing NETs” [26] that are deployed by neutrophils but rapidly washed away by the blood flow, and may have important implications in the development of immune thrombosis, and may additionally disseminate damage to remote areas. One caveat of this technique is that one cannot use a “gold-standard” triple-colocalization criteria (NETs are usually defined as triple-colocalization events of citrullinated histone 3, MPO and DNA) to define the NETs in intravital microscopy. Thus, we rely on the quantification of “NET-like” structures, whose NET identity can be confirmed by using NET-inhibitory agents (like Cl-amidine) which should eliminate the formation of these structures.

Intravital Imaging of NET Formation in the Lung

293

As stated in the sections above, we treat mice with LPS 24 h before TRALI induction, and this induction can be performed at different Zeitgeber times to analyze the circadian pattern of NET release from neutrophils in response to injury. The day of the experiment we first set the intravital imaging as shown above, then we inject an antibody to stain neutrophils (and optionally platelets, as they can be useful to help delineate the vessels) together with a dye for the visualization of DNA being deployed to the extracellular space. We then record multi-position, four-dimensional baseline time-lapse videos to quantify the basal levels of neutrophils in the lung microvasculature. Finally, we induce TRALI and record the same positions for 2 h to quantify the response. We subdivide this part in three sections: capture, quantification, and analysis. 3.3.1 Capture

1. With the mouse pre-treated with LPS 24 h before (see TRALI Subheading 3.1 above), first set up the mice for lung intravital microscopy (see lung intravital Subheading 3.2 above). 2. To confirm the nature of the NET-like structures, it is important to treat a set of mice with Cl-amidine, a PAD4 inhibitor that reduces NET formation [41] (or any alternative compounds able to block NET formation in vivo). Thus, pre-treat a group of mice with 12 mg/kg of Cl-amidine in saline intravenously 1 h before TRALI induction (i.e., 1 h before injection of antibody against MHC-I). We will use this group to confirm the specificity of the method to measure NETs. 3. Prepare a mix of AF647-conjugated Ly6G antibody to stain neutrophils, PE-conjugated CD41 antibody to stain platelets (see also Note 15) and Sytox-green to stain extracellular DNA: dilute 1–1.5 μg of the antibodies against Ly6G and CD41 and 2 μL of Sytox-green in a final volume of 100 μL of sterile saline. Other combination of dyes and fluorophores are possible depending on your microscope features. 4. Inject intravenously the antibody mix and wait 5 min for complete staining of circulating neutrophils. 5. In the intravital microscope software, set up multipoint captures in five randomly chosen areas of the lung (see Note 16). 6. Capture basal recordings on all the positions using the timelapse interval of choice (see Note 17). 7. Carefully inject the antibody against MHC-I (see Subheading 3.1 above) intravenously to initiate lung injury. 8. Immediately resume the captures in the same positions for as long as needed. We found the peak of NET production 30–35 min after TRALI induction (see also Note 18).

294

Alejandra Aroca-Creville´n et al.

Fig. 2 Imaging NETs in the pulmonary microvasculature. (a) Example from a time lapse movie of a neutrophil (Ly6g + cell, cyan) forming a NET-like structure (Sytox-green, red). (b) Quantification of NETs being formed over time in the time lapse videos. (c) Area-under-curve comparison of the curves shown in (b). Data shown as mean  SEM. N ¼ 5 random fields of view from 1 mouse

9. This setup allows for continuous capture of at least 2 h per mouse, if needed. See Fig. 2a for sample images showing a NET being formed in the lung microvasculature. Ensure proper anesthesia and proper temperature control during the whole procedure. 3.3.2 Quantification

In the captured videos, NET-like structures are revealed as distinct events in which extracellular DNA is extruded from Ly6G-stained neutrophils. To quantify them, we construct a DNA-Ly6G colocalization channel. This new channel serves a dual purpose: it enables automatic quantification of NET-like structures, and aids in their localization for manual inspection if interested in analyzing NET-formation modalities. We use Bitplane Imaris here for analysis, but the pipeline can be adapted to the imaging software of choice. 1. Open Imaris and load the file. If you have more than one file for each capture (i.e., captures had to be stopped and resumed) (see Note 19).

Intravital Imaging of NET Formation in the Lung

295

2. Manually inspect the footage to locate one event of DNA being extruded from a Ly6G positive neutrophil and stay around that frame. Generate a colocalization channel of Ly6G and sytox green using the colocalization tool in Imaris, using the localized event as a guide. 3. Once the colocalization channel has been generated, we proceed to quantification of NET-like events. For that, use Imaris spot tool on the new channel. Set an appropriate size for the spot detection to detect a single spot per event. 4. Save the number of NET-like events per frame data. This will be used to analyze the dynamics of NET formation over time. 5. Additionally, perform another spots quantification of the number of neutrophils on the Ly6G channel (and optionally platelets on the CD41 channel) and save the outcome tables (see also Note 20). 3.3.3 Analysis

For statistical analysis, we use GraphPad Prism, but the same procedures can be adapted to the software of choice. 1. First, get the data on neutrophils, platelets, and NET-like structures into XY tables. Add the different mice of each group as replicates. For NETs, you can use the per-frame counts normalized by the number of neutrophils in that same frame as Y values, and the frame numbers as X values. This will better estimate NET formation than using the absolute number of detected NETs, as it will take the number of neutrophils present in that frame into account. 2. Given the interval chosen at the moment of capture, convert the frame numbers into a timescale (i.e., minutes). 3. Plot the number of NET-like events as a function of time and observe the resulting curve. The Cl-amidine treated group should show very few of these structures. See Fig. 2b for a typical curve at ZT5 and ZT13, with Cl-amidine as negative control (one mouse shown per condition). 4. To compare multiple curves, such as treated and non-treated mice, we can calculate the area under the curve (AUC) for the different conditions and use its value and error to perform a simple statistical test, such as t-test or ANOVA (depending on the number of conditions to be compared). See Fig. 2c for a typical example of NET quantification.

4

Notes 1. Because TRALI may cause rapid suffocation, for this procedure mice must be artificially ventilated. We have found that this reduces drastically the likelihood of death during image

296

Alejandra Aroca-Creville´n et al.

capture, but it does not eliminate it. This should be considered when planning the experimental procedure. 2. Mice are usually kept under a 12 h light/12 h dark schedule (in our facility, lights are on at 7:00 am and off at 7:00 pm). If an inverted light cycle cabinet is available, mice kept in inverted light cycle for at least 3 weeks allow to perform the experiment with the different zeitgeber times of choice at similar times of the day. Keep in mind that correct cycle inversion should be assured, and can be easily measured quantifying neutrophils numbers in blood with a hematological counter, for instance. 3. LPS priming is important especially in SPF-housed animals and may not be needed in animals housed in conventional conditions. Priming may also be affected by the mouse origin, means of transportation, facility conditions, and so on, so preliminary experiments should be conducted before running the final experiment. In our experience, mice from Charles River Laboratories respond with a 50% death rate at ZT5 with the conditions stated in our protocol. 4. The anti-H2d antibody dose may also be changed depending on the specific conditions of the animal facility. Preliminary experiments should be performed to find the dosage that causes around 50% mortality in 2 h upon TRALI induction at ZT5. The doses stated in the current protocol have worked for us in different laboratories with both conventional and SPF conditions but may need to be increased or decreased depending on the actual animal facility in which the mice are housed. 5. All the mice in one cage should be injected in less than 1–2 min. Please, note the time of injection of each cage in that case. For slower operators, you may need to annotate the time of injection for each individual mouse. 6. Mice can be sacrificed by cervical dislocation or using the CO2 chamber. Note that to allow blood perfusion if organs are going to be collected, CO2 is preferable. 7. For survival experiments, we recommend that no less than 10 mice per group should be analyzed. The experiment should be repeated at least twice, so in the end at least 20 mice per condition are analyzed. 8. Cauterize quickly in case of rupture of blood vessels during exposure of the trachea. 9. To ensure proper breathing of the animal, the catheter should not be fully inserted, as this may cause damage or suffocation. It should not be left too close to the incision site either, as it may come out. Both proper ventilation and firm adhesion of the catheter must be ensured, as we need to reposition the mouse to continue the procedure and the catheter must remain in place at all times.

Intravital Imaging of NET Formation in the Lung

297

10. The orientation of the forelimbs should be adjusted so that the right chest area is well exposed and that the skin is not too tight. 11. Cauterize the main blood vessels when pulling the skin to make the following steps easier. Hydrate the tissue throughout the surgical procedure to prevent organ dehydration, using sterile saline. From this point on, be careful and avoid direct contact of the tweezers or the scissors with the ribs and the lung below. 12. Prior to opening the thoracic cavity, make sure that the area is flat. You can insert the hood of a needle or some pieces of paper towel under the animal to raise the thoracic cavity and keep it flat. 13. This is one of the most critical steps as it is possible to damage the underlying lung while cutting the ribs. To avoid this, grab the rib tightly and pull it up to detach it as far as possible from the lung surface. Do not insert the scissors perpendicular to the ribcage, but almost in parallel to avoid touching the lung immediately underneath. 14. Petroleum jelly filling should be done as quickly as possible to avoid it being sucked into and contacting with the glass of the surgical window, as this could obstruct visibility. If this happens, removing the window and cleaning it prior to positioning it again could be useful. As the operator gets more comfortable with the procedure, this should not happen. 15. If you are not interested in platelets, alternatively a fluorescent Dextran may be used to delineate blood vessels (Rhodamineconjugated Dextran, for instance). This may be useful in some scenarios. High molecular weight Dextrans will have less tendency to extravasate, although it will eventually in edemaprone TRALI lungs. 16. When selecting random positions to capture, the main thing to keep in mind is to confirm that the blood flow is normal in the selected area. If vessel shows signs of backflow or halted flow, you should avoid this area and its surroundings, as the vacuum suction may be impeding correct flow and the measurements would be affected. 17. We used an interval of 15 s between captures, but this could be increased or decreased depending on your specific needs. The width of capture in the Z axis should be set to the extent allowed by the microscope, but we found it useful to acquire wider depths in the Z axis (i.e., well above and below the in-focus area) as this will help in case the Z coordinates change over the course of capture, which is a common problem. 18. Note that the timing of basal capture and between injection of the anti-MHC-I antibody and resuming the captures should be

298

Alejandra Aroca-Creville´n et al.

kept consistent between mice, as we analyze all mice of each experimental group together. Also, keep in mind that anesthesia should be carefully maintained for the whole imaging time; please consult with your Institution’s veterinarian for help on how to maintain proper anesthesia for longer periods of time depending on how long you need to image. Additionally, edema formation during TRALI can cause death of the animal even if intubated. Captures should be stopped in such an event. The easiest way to confirm the animal is alive is to carefully inspect the blood flow from time to time as capture progresses. 19. Sometimes there might be a need to stop the capture and resume it afterward, for instance if the Z axis is changing and needs to be adjusted. If so, stopping the capture is no problem if Z recalibration is done quickly and the capture is resumed at the same positions as soon as possible. In that case, you will end up with several time lapse videos instead of just one which can be later concatenated. We found FIJI (a distribution of ImageJ) the easiest software to concatenate time lapse videos. For that, open all the videos in FIJI and choose Tools > Concatenate. Confirm the correct order and click ok. Then export the concatenated video to OME-TIFF using the BioFormats exporter (Plugins > Bio-formats > Bio-formats exporter and choose OME-TIFF as type). Videos concatenated this way can just be dragged and dropped to Imaris to continue analysis. 20. Using the Spot tool of Imaris it is also possible to track neutrophils as they crawl in the pulmonary microvasculature. This yields distance and speed calculations that may be useful. For that, activate tracking in the Imaris Spots tool, and confirm by manual inspection that the tracks accurately reflect the movements of the selected neutrophils in the microvasculature. You may need to apply some restrains to reduce artifacts and improve track quality. References 1. Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss DS, Weinrauch Y, Zychlinsky A (2004) Neutrophil extracellular traps kill bacteria. Science 303: 1532–1535 2. Branzk N, Lubojemska A, Hardison SE, Wang Q, Maximiliano G, Brown GD, Papayannopoulos V (2015) Neutrophils sense microbial size and selectively release neutrophil extracellular traps in response to large pathogens. Nat Immunol 15:1017–1025 3. Fuchs TA, Abed U, Goosmann C, Hurwitz R, Schulze I, Wahn V, Weinrauch Y, Brinkmann V, Zychlinsky A (2007) Novel cell death program

leads to neutrophil extracellular traps. J Cell Biol 176:231–241 4. Urban CF, Reichard U, Brinkmann V, Zychlinsky A (2006) Neutrophil extracellular traps capture and kill Candida albicans and hyphal forms. Cell Microbiol 8:668–676 5. Saitoh T, Komano J, Saitoh Y, Misawa T, Takahama M et al (2012) Neutrophil extracellular traps mediate a host defense response to human immunodeficiency virus-1. Cell Host Microbe 12:109–116 6. Poon IKH, Baxter AA, Lay FT, Mills GD, Adda CG et al (2014) Phosphoinositide-mediated

Intravital Imaging of NET Formation in the Lung oligomerization of a defensin induces cell lysis. elife 2014:1–27 7. Saffarzadeh M, Juenemann C, Queisser MA, Lochnit G, Barreto G, Galuska SP, Lohmeyer J, Preissner KT (2012) Neutrophil extracellular traps directly induce epithelial and endothelial cell death: a predominant role of histones. PLoS One 7:e32366 8. Silvestre-Roig C, Braster Q, Wichapong K, Lee EY, Teulon JM et al (2019) Externalized histone H4 orchestrates chronic inflammation by inducing lytic cell death. Nature 569:236–240 9. Abrams ST, Zhang N, Manson J, Liu T, Dart C et al (2013) Circulating histones are mediators of trauma-associated lung injury. Am J Respir Crit Care Med 187:160–169 10. Aroca-Creville´n A, Adrover JM, Hidalgo A (2020) Circadian features of neutrophil biology. Front Immunol 11:1–9 ´ vila JA ´ , Adrover JM, Hidalgo A 11. Nicola´s-A (2017) Neutrophils in homeostasis, immunity, and cancer. Immunity 46:15–28 12. Knight JS, Zhao W, Luo W, Subramanian V, Dell AAO et al (2013) Peptidylarginine deiminase inhibition is immunomodulatory and vasculoprotective in murine lupus. J Clin Invest 123:2981–2993 13. Hakkim A, Fu¨rnrohr BG, Amann K, Laube B, Abed UA, Brinkmann V, Herrmann M, Voll RE, Zychlinsky A (2010) Impairment of neutrophil extracellular trap degradation is associated with lupus nephritis. Proc Natl Acad Sci U S A 107:9813–9818 14. Leffler J, Martin M, Gullstrand B, Tyden H, Lood C, Truedsson L, Bengtsson AA, Blom AM (2012) Neutrophil extracellular traps that are not degraded in systemic lupus erythematosus activate complement exacerbating the disease. J Immunol 188:3522–3531 15. Wong SL, Demers M, Martinod K, Gallant M, Wang Y, Goldfine AB, Kahn CR, Wagner DD (2015) Diabetes primes neutrophils to undergo NETosis, which impairs wound healing. Nat Med 21:815–819 16. McDonald B, Urrutia R, Yipp BG, Jenne CN, Kubes P (2012) Intravascular neutrophil extracellular traps capture bacteria from the bloodstream during sepsis. Cell Host Microbe 12: 324–333 17. Go´mez-Moreno D, Adrover JM, Hidalgo A (2018) Neutrophils as effectors of vascular inflammation. Eur J Clin Investig 48:1–14 18. Fuchs TA, Brill A, Duerschmied D, Schatzberg D, Monestier M et al (2010) Extracellular DNA traps promote thrombosis. Proc Natl Acad Sci U S A 107:15880–15885

299

19. Massberg S, Grahl L, Von Bruehl ML, Manukyan D, Pfeiler S et al (2010) Reciprocal coupling of coagulation and innate immunity via neutrophil serine proteases. Nat Med 16: 887–896 20. Jime´nez-Alca´zar M, Rangaswamy C, Panda R, Bitterling J, Simsek YJ et al (2017) Host DNases prevent vascular occlusion by neutrophil extracellular traps. Science 358: 1202–1206 21. Do¨ring Y, Soehnlein O, Weber C (2014) Neutrophils cast NETs in atherosclerosis. Circ Res 114:931–934 22. Knight JS, Luo W, O’Dell AA, Yalavarthi S, Zhao W et al (2014) Peptidylarginine deiminase inhibition reduces vascular damage and modulates innate immune responses in murine models of atherosclerosis. Circ Res 114: 947–956 23. Borissoff JI, Joosen IA, Versteylen MO, Brill A, Fuchs TA et al (2013) Elevated levels of circulating DNA and chromatin are independently associated with severe coronary atherosclerosis and a prothrombotic state. Arterioscler Thromb Vasc Biol 33:2032–2040 24. Albrengues J, Shields MA, Ng D, Park CG, Ambrico A et al (2018) Neutrophil extracellular traps produced during inflammation awaken dormant cancer cells in mice. Science 361(6409):eaao4227 25. Park J, Wysocki RW, Amoozgar Z, Maiorino L, Fein MR et al (2016) Cancer cells induce metastasis-supporting neutrophil extracellular DNA traps. Sci Transl Med 8(361):361ra138 26. Adrover JM, Aroca-Creville´n A, Crainiciuc G, Ostos F, Rojas-Vega Y et al (2020) Programmed ‘disarming’ of the neutrophil proteome reduces the magnitude of inflammation. Nat Immunol 21:135–144 27. Thomas GM, Carbo C, Curtis BR, Martinod K, Mazo IB et al (2012) Extracellular DNA traps are associated with the pathogenesis of TRALI in humans and mice. Blood 119: 6335–6343 28. Caudrillier A, Kessenbrock K, Gilliss BM, Nguyen JX, Marques MB, Monestier M, Toy P, Werb Z, Looney MR (2012) Platelets induce neutrophil extracellular traps in transfusion-related acute lung injury. J Clin Invest 122:2661–2671 29. Barnes BJ, Adrover JM, Baxter-Stoltzfus A, Borczuk A, Cools-Lartigue J et al (2020) Targeting potential drivers of COVID-19: neutrophil extracellular traps. J Exp Med 217:1–7 30. Mikacenic C, Moore R, Dmyterko V, West TE, Altemeier WA, Liles WC, Lood C (2018)

300

Alejandra Aroca-Creville´n et al.

Neutrophil extracellular traps (NETs) are increased in the alveolar spaces of patients with ventilator-associated pneumonia. Crit Care 22:1–8 31. Summers C, Rankin SM, Condliffe AM, Singh N, Peters AM, Chilvers ER (2010) Neutrophil kinetics in health and disease. Trends Immunol 31:318–324 32. Casanova-Acebes M, Pitaval C, Weiss LA, Nombela-Arrieta C, Che`vre R et al (2013) Rhythmic modulation of the hematopoietic niche through neutrophil clearance. Cell 153: 1025–1035 ´ vila JA, Hidalgo A 33. Adrover JM, Nicola´s-A (2016) Aging: a temporal dimension for neutrophils. Trends Immunol 37:334–345 34. Adrover JM, del Fresno C, Crainiciuc G, Cuartero MI, Casanova-Acebes M et al (2019) A neutrophil timer coordinates immune defense and vascular protection. Immunity 50: 390–402.e10 35. Martinez-Bakker M, Helm B (2015) The influence of biological rhythms on host-parasite interactions. Trends Ecol Evol 30:314–326 36. Fung YL, Silliman CC (2009) The role of neutrophils in the pathogenesis of transfusion-

related acute lung injury. Transfus Med Rev 23:266–283 37. Boe E, Usfbunfou B, Njdf Q, Uxp JOB, Npefm F et al (2009) Platelet depletion and aspirin treatment protect mice in a two-event model of transfusion-related acute lung injury. J Clin Invest 119(11):3450–3461 38. Sreeramkumar V, Adrover JM, Ballesteros I, Cuartero MI, Rossaint J et al (2014) Neutrophils scan for activated platelets to initiate inflammation. Science 346:1234–1238 39. Looney MR, Su X, Van Ziffle JA, Lowell CA, Matthay MA (2006) Neutrophils and their Fcγ receptors are essential in a mouse model of transfusion-related acute lung injury. J Clin Invest 116:1615–1623 40. Looney MR, Thornton EE, Sen D, Lamm WJ, Glenny RW, Krummel MF (2011) Stabilized imaging of immune surveillance in the mouse lung. Nat Methods 8:91–96 41. Li P, Li M, Lindberg MR, Kennett MJ, Xiong N, Wang Y (2010) PAD4 is essential for antibacterial innate immunity mediated by neutrophil extracellular traps. J Exp Med 207: 1853–1862

Chapter 20 Real-Time Measurement of Energy Metabolism Over Circadian Time Using Indirect Calorimetry-Enabled Metabolic Cages Kevin B. Koronowski and Paolo Sassone-Corsi Abstract Indirect calorimetry probes the relationship between fuel consumed and energy produced, and in doing so provides an estimation of whole-body energy expenditure and fuel preference. When assayed continuously in real-time, rhythms appear and illuminate the temporal regulation of energy metabolism by the circadian clock. Here we describe a method for recording circadian energy metabolism in mice using indirect calorimetry-enabled metabolic cages, encompassing mouse entrainment, experimental design, data acquisition and analysis, troubleshooting of common problems, and important considerations. This method is adaptable to the end user’s equipment and serves as an effective tool to study, for example, mutant mice, dietary interventions, drug treatments, or circadian disruption. Key words Circadian clock, Circadian rhythm, Indirect calorimetry, Energy metabolism, Metabolic cage, In vivo recording, Real-time recording

1

Introduction Life thrives on energy metabolism, the ability of organisms to transform chemical energy stored in molecules into the versatile energy currency adenosine tri-phosphate (ATP). In animals, the production of energy derives largely from oxidative phosphorylation, wherein the primary substrates glucose, fatty acids, amino acids, and oxygen (O2) are recombined to give ATP, water, and carbon dioxide (CO2) [1]. By the end of the nineteenth century, scientists were able to link this process with the heat it generates, that is the relationship between fuel consumed and heat produced (law of conservation of energy) [2]. As a result, direct calorimetry was born—a method to measure energy expenditure directly through whole-body heat production. Further work showed energy expenditure can be estimated by measuring O2

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_20, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

301

302

Kevin B. Koronowski and Paolo Sassone-Corsi

consumption, and that fuel preference can be calculated with the additional measurement of CO2 production (indirect calorimetry) [2, 3]. Current metabolic cage systems designed for rodents, with built-in calorimeters, enable the continuous recording of energy expenditure and related behaviors like feeding. Studies using these systems beautifully demonstrate the circadian rhythmicity of wholebody energy metabolism [4–6]. The circadian clock, which is expressed in most cells of the body, drives ~24 h rhythms of physiology and aligns them with geophysical time (i.e., the solar cycle). Temporal regulation of energy pathways across organs is essential for metabolic homeostasis and is a quintessential task for the clock system [7, 8]. In mice, maximal energy expenditure is observed in the middle of nighttime (during the active phase), and coincides with the peaks of VO2 and VCO2 that are tied to cycles of locomotor activity and food intake, respectively [5]. The fuel switching that occurs between the daytime (inactive, fasting) and nighttime (active, feeding) is reflected in the respiratory exchange ratio (RER ¼ VCO2/VO2), which alternates from more lipid oxidation in the light phase to more carbohydrate oxidation during the dark phase [9]. Here, we describe a method for measuring energy expenditure and related parameters in mice over several circadian cycles using a metabolic cage system with indirect calorimetry. Our protocol includes equipment specifications, circadian entrainment, experimental setup, data acquisition and analysis, common troubleshooting, and several important considerations. Although tailored to our system, this method can be readily adapted to the user’s equipment and serve as an effective tool to investigate organism-wide energy metabolism over circadian time in mutant mice, across diets or under drug treatments, to give a few examples.

2

Materials Materials will vary depending upon the metabolic cage system being used. The following list pertains to the indicated hardware. Please review the user manual of your specific system and adjust accordingly.

2.1 Equipment (Fig. 1)

1. Comprehensive Lab Animal Monitoring System (CLAMS) Hardware (Columbus Instruments). This is an open-circuit calorimeter with additional technology to monitor physiological and behavioral parameters like food intake, water intake, urine collection, and feces collection (urine and feces are not analyzed in this protocol). In this negative flow ventilation system, the mouse inspires and expires from the stream of atmospheric air passing through the cage. The air supply line

Measuring Circadian Energy Metabolism in Real-Time

303

Fig. 1 Main components of an indirect calorimetry-enabled metabolic cage system. (a) Main features of the metabolic cage system are highlighted. (b) Features of individual cages. I—Mouse chamber; II—drinking unit; III—feeding unit; IV—feces collection; V—urine collection

then directs air flow into either the sample conditioning unit (primes air for presentation to the gas sensors) or the ventilation pump (return air to the atmosphere). 2. Oxymax for Windows Software (Columbus Instruments). 3. Standard temperature and humidity monitor for the room. 2.2

Consumables

1. Autoclaved tap water for drinking water. 2. Standard normal chow diet (e.g., Teklad global soy proteinfree extruded, 2020X). To ensure accurate feeding recordings, pellets must be ground into a relatively fine powder. A standard kitchen blender on the “grind” setting can serve this purpose. 3. Desiccant-anhydrous indicating DRIERITE (e.g., stock# 23025) for system moisture absorption. 4. Calibration gas: 5000 PPM CO2, 20.5% O2, balance nitrogen (may vary with gas sensor and calorimeter types).

3

Methods The precise order and details of the following steps will depend on the user’s specific equipment. Here, we list a general framework for execution of the experiment.

3.1

Entrainment

1. Individually house mice in an investigator maintained room, that is a room in which animal husbandry (i.e., changing of cages) does not occur, for at least 1 week prior to experimentation. This allows for minimal disturbance of the mice during

304

Kevin B. Koronowski and Paolo Sassone-Corsi

the daytime and begets robustness of sleep-wake and feedingfasting circadian cycles. Set entrainment and experimental room conditions as follows: 12 h light:12 h dark schedule— ~300–400 lux, room temperature (20–26  C, 68–79  F), ideally 40–60% humidity (see Notes 1 and 2). 3.2 Experimental Setup

1. Turn on all system equipment in order: (a) Main system power; (b) CI-Bus; (c) CO2 sensor; (d) all other devices. If available, use a dedicated emergency outlet to avoid loss of data in the event of a power outage. Be sure to cover all light-emitting devices (see Notes 3 and 4). 2. Open Oxymax software, checking that all hardware is recognized by the computer and status is “OK.” Go to “File” -> “Save Experiment Configuration” to open the default experiment configuration file that has previously been saved. Set sampling frequency equal to 10 min by going to “Experiment” -> “Properties.” Set flow rate equal to 0.5 L/min by going to “Experiment” -> “Set Up.” Consult the user manual for a description of all parameters. 3. Open calibration gas tank. Go to “Experiment” -> “Calibrate” and calibrate system gases. When prompted, adjust oxygen gain to 20.48%. Don’t forget to close the calibration gas tank when finished. 4. In your records, record mouse information including weight, age, sex, original cage #, mouse number (i.e., ear numbering), and parental breeding cage. 5. Go to “Experiment” -> “Set Up” and enter mouse identifier and weight for cages 1–4 (4 cages, 1 mouse per cage). 6. Fill food containers with powdered chow and place them in feeding units, then connect feeding units to the cages. Ensure feeders are enabled by going to “Tools” -> “Feeding” -> “Refill Load Feeders.” 7. Check that drinking nozzles are functioning properly by tapping nozzle on the back of your hand and checking for water droplets. Drinking lines can be activated if needed—go to “Tools” -> “Drinking” -> “Pump Status.” 8. Place mice in cages and close lids. 9. Go to “Experiment” -> “Run” to start recording. 10. Check that first gas measurements are within the expected range (1000–6000 mL/kg/h).

3.3

Data Acquisition

1. During data acquisition, it is important to monitor physiological and behavioral measurements. Once daily, in the data window go to “Graph” -> “Select Data” icon and apply “VO2,” “VCO2,” “Feed ACC,” and “Drink ACC” data for all cages.

Measuring Circadian Energy Metabolism in Real-Time

305

Ensure proper functionality of the system in the graphed data. If drinking lines fail, for example, data cannot be used since lack of hydration will impact energy homeostasis. See Notes 4–7 regarding the troubleshooting of common problems. 2. Run cages for a minimum of 3 full days (1 acclimation day + 2 usable data days), preferably 4–5 days (see Note 8). 3.4 Experiment Shutdown

1. Go to “Experiment” -> “Stop” to stop recording. 2. Go to “File” -> “Export” -> “Subject CSV Files” to export data for each mouse/cage to a CSV file. 3. Exit the software. 4. Remove mice from the cages. 5. Turn off all system equipment in the reverse order of the start sequence. 6. Deconstruct cages, clean thoroughly with soap and water and animal facility approved disinfectant spray. Rinse and dry (see Note 9). 7. Build cages for the next run.

3.5

Data Analysis

While some parameters are in their final form as exported from the software, others need to be processed or filtered to eliminate noise and erroneous data points resulting from faulty readings. Setting up a template file with the following formulas will save time and minimize errors when analyzing the data. 1. Extract data from exported CSV files and save in one Excel file. 2. For all parameters, delete at least the first full day of recording (acclimation period for the mouse). For example, if you have 3 days of data, delete day 1 and start analysis from day 2. Also, be sure to note your sampling frequency (i.e., measurements taken every 10 min) when summing or averaging values for each hour. Recording from the same number of mice each time ensures sampling frequency need not be adjusted during analysis. 3. Food Intake: (a) using an “IF” formula, make negative values ¼ 0; (b) using an “IF” formula, make values > 0.7 ¼ 0 (at the sampling rate, a feeding event greater than this value is highly unlikely to be true and is most likely to come from removal or loss of an unground piece of chow through the grid, as mice often dig in the food containers); (c) sum values within each hour, such that you now have 1 value per hour; (d) sum day values (ZT0–ZT12, light phase, e.g., 7:30 AM–7: 30 PM); (e) sum night values (ZT12–ZT24, dark phase, e.g., 7:30 PM–7:30 AM); (f) average day vs. night values from each day.

306

Kevin B. Koronowski and Paolo Sassone-Corsi

4. VO2 and VCO2: (a) divide values by 1000 to go from mL/kg/ h to L/kg/h; (b) average values within each hour, such that you now have 1 value per hour; (c) average day values (ZT0– ZT12); (d) average night values (ZT12–ZT24); (e) average day vs. night values from each day. 5. Energy expenditure (heat) and respiratory exchange ratio (RER, VCO2/VO2): (a) average values within each hour, such that you now have 1 value per hour; (b) average day values (ZT0–ZT12); (c) average night values (ZT12–ZT24); (d) average day vs. night values from each day. RER can differentiate energy production from carbohydrate vs. fat because the ratio of CO2 produced to O2 consumed is different for the oxidation of the two substrates: 6O2 + C6H12O6 (glucose) ¼ 6CO2 + 6H2O + 38 ATP (6CO2/6O2 ¼ 1.0); 23O2 + C16H32O2 (fatty acid) ¼ 16CO2 + 16H2O + 129 ATP (16CO2/23O2 ¼ 0.7). 6. For all parameters, plot hourly data from individual mice to check for any interruptions in the recording. Plot group averages for hourly data and day vs. night values, as shown for 1 example mouse in Fig. 2. Apply appropriate statistical tests across groups depending upon your experimental design (see Note 10).

Fig. 2 Measurements of energy metabolism in an adult female C57Bl/6J mouse over 2 circadian cycles. Data was acquired under a 12 h light:12 h dark schedule. Gray backdrops indicate the dark phase. Zeitgeber time (ZT) is the standard unit of circadian time under light-dark conditions, where ZT0 ¼ lights on (7:30 AM geophysical time) and ZT12 ¼ lights off (7:30 PM geophysical time). ZT0 and ZT24 are interchangeable. Line graphs display hourly values. Bar graphs display light phase (ZT0–ZT12) vs. dark phase (ZT12–ZT24) averaged values. Notice the units for each parameter. Rhythmicity analysis using BioDare2 is shown to the right of each metabolic feature—period in hours (with Benjamini Hochberg-corrected p-value), phase, and amplitude (Tau)

Measuring Circadian Energy Metabolism in Real-Time

307

7. Rhythmicity analysis—using specific tools to determine features of circadian rhythmicity for each individual mouse complements the aforementioned analysis. BioDare2 (biodare2.ed. ac.uk) is one such online tool that can calculate circadian rhythmicity p-value, period, amplitude/power (tau) and peak and trough phase [10]. The values presented in Fig. 2 were determined using the following settings in BioDare2: rhythmicity test—“data window” ¼ 0–48 h, input data—“linear dtr,” test method—“BD2 eJTK,” analysis preset—“eJTK Classic”; Period analysis—“data window” ¼ 0–48 h, input data— “linear dtr,” expected periods ¼ 23–25 (can extend if expecting wider range in mutant animals), analysis method—“MFourFit” (see Note 11).

4

Notes 1. Light schedule—one should consider certain points when choosing the appropriate light schedule. Most animal facilities house mice under standard 12 h light:12 h dark cycles and accompanying experiments for a study are performed under these conditions. Thus, this may be appropriate to compare with other past and present data. However, it is worth noting that 12 h light:12 h dark cycles are unnatural for mice in the wild. For this reason, circadian and other studies may also be conducted under a skeleton photoperiod, that is a light schedule that mimics a mouse’s more natural exposure to light at dawn and dusk (e.g., 1 h light:10 h dark, 1 h light:12 h dark) [11]. Still, this is less common in the current literature. Lastly, in the circadian research field, truly circadian rhythms are demonstrated in the absence of external synchronizing factors (i.e., light-dark cycle, temperature changes). As such, many experiments are conducted under constant darkness. Please consult these references for further experimental design considerations [12, 13]. 2. Temperature—as do humans, mice prefer to live within their thermo-neutral zone, expending the least amount of energy to maintain core body temperature [14]. Room temperature set to 20–26  C (68–79  F) is below the mouse’s thermo-neutral zone of ~30  C. As a result, even in typical animal holding and experimental rooms, mice may experience cold stress and activation of brown fat metabolism at the lower end of the room temperature range [14]. Consider if this is an important aspect of your hypothesis. 3. Light-emitting devices—even very low intensity light (~5 lux) can perturb certain metabolic rhythms [15, 16]. Thus, all lightemitting devices with on/off indicator or monitoring lights

308

Kevin B. Koronowski and Paolo Sassone-Corsi

should be covered. Turn off the computer monitor or cover the screen when it is not being used. It is pertinent to avoid these unwanted sources of light during nighttime. Do not enter the room during the dark phase if you have no way of blocking hallway light from entering the room with you. If necessary, consider hanging a blackout curtain that completely encompasses the room door, allowing you to enter without any light disturbance. 4. Troubleshooting—computer not recognizing hardware. The order in which pieces of hardware are turned on may be important for system functionality. Similarly, all hardware may need to be active prior to launching software. Establish an activation sequence for hardware and software for the start of each run. If problems arise, restarting the computer is one potential fix. 5. Troubleshooting—extremely high or low VO2 or VCO2. Calibration failure may indicate a gas sensor issue. Even if system calibration is successful, inaccurate gas readings can occur from several issues including sensors. The life span of O2 and CO2 sensors may be shortened by heavy use of the equipment. Consult your user manual for typical replacement timeframes. Excess moisture can result in abnormal readings. Be sure to monitor desiccant color (DRIERITE changes from blue to pink upon absorption of moisture) and replace it prior to complete loss of dryness. Moisture and other build up in filters of gas lines may also affect readings negatively. Replace these filters if problems persist. 6. Troubleshooting—drinking supply failure. If drinking graphs indicate a lack of activity for an abnormally long period of time, go to “Experiment” -> “Drinking” -> “Pump Status”—cages with line issues will be indicated (e.g., “Overpump” or “Underpump”). Re-enable troublesome lines and confirm water is again available at the nozzle tip. 7. Troubleshooting—food intake recordings. Long periods of 0 values or negative values may indicate a feeder unit problem. Feeder units are prone to the buildup of hardened food particles, among other biological materials, which can reduce the sensitivity of readings or cause complete loss of function. It may be necessary to occasionally take apart the feeders to clean them thoroughly. During a run, acutely removing debris around certain areas may temporarily fix this problem. 8. Stress—it is well documented that certain metabolic cages— wherein mice rest upon wire grids or mesh, the cages are devoid of bedding and nesting material and also lack environment enrichment—induce behavioral and molecular markers of stress in mice [17, 18]. Hence, time in these types of cages should be limited to the shortest window that includes a

Measuring Circadian Energy Metabolism in Real-Time

309

habituation period (novelty of new environment wears off, ~1 day) and accurate data acquisition period (~2 days). Newer metabolic cages circumvent this added stress by keeping mice in a home cage environment which features bedding and nesting materials, as well as the option for environment enrichment. 9. Cage cleaning and odors—it is important to clean cages well to eliminate odor from the previous mouse in the subsequent run. It has also been shown that human smells can induce markers of stress in rodents [19]. 10. Assumptions, data interpretation, and limitations—although outside the scope of this methods description, it is important to understand the assumptions and potential limitations of indirect calorimetry in the mouse metabolic cage, as to correctly interpret results [2, 20]. For example, a key assumption for calculating energy expenditure is that the loss of substrates in feces and urine is negligible [2]. However, mutations or interventions may alter intestinal absorption of substrates substantially. Concerning interpretation, total energy expenditure is an amalgamation of resting (basal) expenditure, thermoregulation, locomotor activity, and the thermic effect of food [3]. Distinguishing between these components requires additional tests. 11. Rhythmicity analysis—we find that 2 days of continuous data can give accurate measurements in terms of circadian rhythmicity parameters. The performance of certain rhythmicity algorithms may improve with more data points, so consider 3 or 4 days of recording if feasible and ethical [21]. A clear picture of results can be obtained from both rhythmicity analysis tools and standard statistical tests comparing parameter averages between day and night or waveform fitting of hourly data points. References 1. Wilson DF (2017) Oxidative phosphorylation: regulation and role in cellular and tissue metabolism. J Physiol 595(23):7023–7038. https:// doi.org/10.1113/JP273839 2. Mtaweh H, Tuira L, Floh AA, Parshuram CS (2018) Indirect calorimetry: history, technology, and application. Front Pediatr 6:257. https://doi.org/10.3389/fped.2018.00257 3. Speakman JR (2013) Measuring energy metabolism in the mouse - theoretical, practical, and analytical considerations. Front Physiol 4:34. https://doi.org/10.3389/fphys.2013.00034 4. Chaix A, Lin T, Le HD, Chang MW, Panda S (2019) Time-restricted feeding prevents

obesity and metabolic syndrome in mice lacking a circadian clock. Cell Metab 29(2): 303–319.e304. https://doi.org/10.1016/j. cmet.2018.08.004 5. Adamovich Y, Ladeuix B, Sobel J, Manella G, Neufeld-Cohen A, Assadi MH, Golik M, Kuperman Y, Tarasiuk A, Koeners MP, Asher G (2019) Oxygen and carbon dioxide rhythms are circadian clock controlled and differentially directed by behavioral signals. Cell Metab 29(5):1092–1103.e1093. https://doi.org/ 10.1016/j.cmet.2019.01.007 6. Welz PS, Zinna VM, Symeonidi A, Koronowski KB, Kinouchi K, Smith JG, Guillen IM,

310

Kevin B. Koronowski and Paolo Sassone-Corsi

Castellanos A, Furrow S, Aragon F, Crainiciuc G, Prats N, Caballero JM, Hidalgo A, Sassone-Corsi P, Benitah SA (2019) BMAL1-driven tissue clocks respond independently to light to maintain homeostasis. Cell 177(6):1436–1447.e1412. https:// doi.org/10.1016/j.cell.2019.05.009 7. Greco CM, Sassone-Corsi P (2019) Circadian blueprint of metabolic pathways in the brain. Nat Rev Neurosci 20(2):71–82. https://doi. org/10.1038/s41583-018-0096-y 8. Panda S (2016) Circadian physiology of metabolism. Science 354(6315):1008–1015. https://doi.org/10.1126/science.aah4967 9. Koronowski KB, Kinouchi K, Welz PS, Smith JG, Zinna VM, Shi J, Samad M, Chen S, Magnan CN, Kinchen JM, Li W, Baldi P, Benitah SA, Sassone-Corsi P (2019) Defining the Independence of the liver circadian clock. Cell 177(6):1448–1462.e1414. https://doi.org/ 10.1016/j.cell.2019.04.025 10. Zielinski T, Moore AM, Troup E, Halliday KJ, Millar AJ (2014) Strengths and limitations of period estimation methods for circadian data. PLoS One 9(5):e96462. https://doi.org/10. 1371/journal.pone.0096462 11. Dallmann R, DeBruyne JP, Weaver DR (2011) Photic resetting and entrainment in CLOCKdeficient mice. J Biol Rhythm 26(5):390–401. https://doi.org/10.1177/0748730411414345 12. Hughes ME, Abruzzi KC, Allada R, Anafi R, Arpat AB, Asher G, Baldi P, de Bekker C, BellPedersen D, Blau J, Brown S, Ceriani MF, Chen Z, Chiu JC, Cox J, Crowell AM, DeBruyne JP, Dijk DJ, DiTacchio L, Doyle FJ, Duffield GE, Dunlap JC, Eckel-Mahan K, Esser KA, FitzGerald GA, Forger DB, Francey LJ, Fu YH, Gachon F, Gatfield D, de Goede P, Golden SS, Green C, Harer J, Harmer S, Haspel J, Hastings MH, Herzel H, Herzog ED, Hoffmann C, Hong C, Hughey JJ, Hurley JM, de la Iglesia HO, Johnson C, Kay SA, Koike N, Kornacker K, Kramer A, Lamia K, Leise T, Lewis SA, Li J, Li X, Liu AC, Loros JJ, Martino TA, Menet JS, Merrow M, Millar AJ, Mockler T, Naef F, Nagoshi E, Nitabach MN, Olmedo M, Nusinow DA, Ptacek LJ, Rand D, Reddy AB, Robles MS, Roenneberg T, Rosbash M, Ruben MD, Rund SSC, Sancar A, Sassone-Corsi P, Sehgal A, Sherrill-Mix S, Skene DJ, Storch KF, Takahashi JS, Ueda HR, Wang H, Weitz C, Westermark PO, Wijnen H, Xu Y, Wu G, Yoo SH, Young M, Zhang EE, Zielinski T, Hogenesch JB (2017) Guidelines for genome-scale analysis of biological

rhythms. J Biol Rhythm 32(5):380–393. h t t p s : // d o i . o r g / 1 0 . 1 1 7 7 / 0748730417728663 13. Eckel-Mahan K, Sassone-Corsi P (2015) Phenotyping circadian rhythms in mice. Curr Protoc Mouse Biol 5(3):271–281. https://doi. org/10.1002/9780470942390.mo140229 14. Ganeshan K, Chawla A (2017) Warming the mouse to model human diseases. Nat Rev Endocrinol 13(8):458–465. https://doi.org/ 10.1038/nrendo.2017.48 15. Cisse YM, Peng J, Nelson RJ (2017) Effects of dim light at night on food intake and body mass in developing mice. Front Neurosci 11: 294. https://doi.org/10.3389/fnins.2017. 00294 16. Borniger JC, Maurya SK, Periasamy M, Nelson RJ (2014) Acute dim light at night increases body mass, alters metabolism, and shifts core body temperature circadian rhythms. Chronobiol Int 31(8):917–925. https://doi.org/10. 3109/07420528.2014.926911 17. Sahin Z, Solak H, Koc A, Ozen Koca R, Ozkurkculer A, Cakan P, Solak Gormus ZI, Kutlu S, Kelestimur H (2019) Long-term metabolic cage housing increases anxiety/depression-related behaviours in adult male rats. Arch Physiol Biochem 125(2):122–127. https:// doi.org/10.1080/13813455.2018.1441314 18. Kalliokoski O, Jacobsen KR, Darusman HS, Henriksen T, Weimann A, Poulsen HE, Hau J, Abelson KS (2013) Mice do not habituate to metabolism cage housing—a three week study of male BALB/c mice. PLoS One 8(3):e58460. https://doi.org/10.1371/jour nal.pone.0058460 19. Sorge RE, Martin LJ, Isbester KA, Sotocinal SG, Rosen S, Tuttle AH, Wieskopf JS, Acland EL, Dokova A, Kadoura B, Leger P, Mapplebeck JC, McPhail M, Delaney A, Wigerblad G, Schumann AP, Quinn T, Frasnelli J, Svensson CI, Sternberg WF, Mogil JS (2014) Olfactory exposure to males, including men, causes stress and related analgesia in rodents. Nat Methods 11(6):629–632. https://doi.org/10.1038/ nmeth.2935 20. Lighton JR (2017) Limitations and requirements for measuring metabolic rates: a mini review. Eur J Clin Nutr 71(3):301–305. https://doi.org/10.1038/ejcn.2016.265 21. Mei W, Jiang Z, Chen Y, Chen L, Sancar A, Jiang Y (2020) Genome-wide circadian rhythm detection methods: systematic evaluations and practical guidelines. Brief Bioinform. https:// doi.org/10.1093/bib/bbaa135

Chapter 21 Untargeted and Targeted Circadian Metabolomics Using Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) and Flow Injection-Electrospray Ionization-Tandem Mass Spectrometry (FIA-ESI-MS/MS) Anna Artati, Cornelia Prehn, Dominik Lutter, and Kenneth Allen Dyar Abstract A diverse array of 24-h oscillating hormones and metabolites direct and reflect circadian clock function. Circadian metabolomics uses advanced high-throughput analytical chemistry techniques to comprehensively profile these small molecules (4  SD (standard deviation) are considered as outliers, and are thus removed from the dataset. For a comprehensive overview on different filtering methods for untargeted metabolomics, see ref. 45. 4. Scaling: In some cases, scaling may be applied to the data. However, there are multiple scaling methods, each with a different purpose. The most popular scaling methods for metabolomics data are autoscaling, range scaling, and Pareto scaling. An overview and a comparison of different scaling methods is described in [46, 47].

Circadian Metabolomics

321

5. Imputation: Data imputation is applied for analysis approaches that cannot deal with missing values. Several causes for missing values exist, which can be generally divided into three different types: missing not at random (MNAR), missing at random (MAR), and missing completely at random (MCAR). MAR and MCAR originate either from unexpected stochastic errors (MCAR) or from an unknown or unobserved determining factor or variable. MNAR based missing values are usually caused by censoring or limits of quantification (LOQ) [48]. Whereas the latter is often imputed by adding the minimum of the detected metabolite, the former can sometimes be replaced with raw or baseline signals. In general, more accurate methods are recommended here. We usually apply k-nearest neighbor imputation. A detailed comparison of different imputation methods is given by [48]. 6. Tests for periodicity: A variety of tools can be used to determine 24-h periodicity of fluctuating metabolites, as well as their amplitude of oscillation and phase characteristics. Widely used tests we recommend include the nonparametric statistical algorithm JTK_CYCLE [49] and the and deep learning method BIO_CYCLE [50].

4

Notes 1. All reagents, extraction solvent, and LC-MS/MS solvents are prepared using ultrapure water (deionized water, sensitivity: of 18 MΩ cm at 25  C) and HPLC grade reagents. Prepare and store all solvents at room temperature. 2. Standard compounds are added into the extraction solvent to enable the monitoring of extraction efficiency. The extraction solvent is stable for about 2 weeks at 4  C. 3. The reconstitution solvents are stable for about 2 weeks at 4  C. 4. It is best to store the kit plate and the vial box at 80  C, at least at 20  C. When ready to use, wait until the kit plate and the vials have reached room temperature before opening. 5. Always prepare the pre-mix and the derivatization reagent fresh. Add the PITC directly before use. Buy very small units of PITC and relatively small units of pyridine, because both can degrade, especially when frequently opening the bottles. 6. Prepare stock of extraction solvent by dissolving 386 mg ammonium acetate in 1000 mL methanol. The extraction solvent is stable for about 2 months. Original ammonium acetate bottles should be discarded about 1 year after first opening. If preparing stock solutions, always print your name, preparation date, and expiration date.

322

Anna Artati et al.

7. Use glass-handling safety gloves and protective glasses. The FIA running solvent is stable for about 2 months. 8. Always fill the bottles first with water or acetonitrile and add formic acid last. Mix the solvents vigorously. Make sure that the bottle for acetonitrile is completely dry. Formic acid can degrade. Use small original units of formic acid only and replace within latest 1 year. 9. There are different types of pooled QC samples commonly used in untargeted metabolomics assays. The preparation of pooled QC samples depends on the type of the study and the amount of each sample available [51]. Whenever possible, it is recommended to choose option “A” first, followed by option “B.” Option “C” can be applied in cohort studies with large numbers of test samples: A. A pooled QC sample created from all of the samples in the study. B. A pooled QC sample created from a representative subset of the samples in the study. C. A pooled sample created from the same biological matrix type of the study samples but it is not taken from any the test sample in the study. D. A pooled sample created from the processed sample solution for all or a representative subset of the test samples, e.g., tissue homogenate, cell extracts. E. An artificial QC sample created with authentic chemical standards and a dummy sample matrix. 10. Sample preparation and LC-MS/MS measurements of the AbsoluteIDQ™ p180 Kit are comprehensively described in the manufacturers manual UM-P180, which comes with the kit. The method has been proven to be in conformance with the European Medicines Agency (EMA) guidelines [52], which implies proof of reproducibility within a given error range. Analytical specifications for LODs and evaluated quantification ranges, further LOD for semiquantitative measurements, identities of quantitative and semiquantitative metabolites, specificity, potential interferences, linearity, precision and accuracy, reproducibility and stability are described in Biocrates analytical specifications manual AS-P180. The LODs were set to three times the values of the zero samples (PBS). 11. Frozen tissue samples are homogenized and extracted using homogenization tubes containing ceramic beads (1.4 mm) and a certain amount of extraction solvent. Tissue pieces are weighed frozen and placed in pre-cooled ( 80  C) tubes. A certain amount of dry ice cooled extraction solvent is added to each mg of frozen tissue. The sample is homogenized using a

Circadian Metabolomics

323

Precellys24 homogenizer with an integrated cooling unit (PEQLAB Biotechnology GmbH, Germany) and centrifuged. The homogenate supernatant is taken for the actual measurements. More detailed information about tissue sample preparation, the optimized extraction solvent, and tissue-to-solvent ratio are given here [42]. Ready-to-use methods for tissue analysis are provided for skeletal muscle, adipose tissue, liver, brain, kidney, pituitary gland, lung, bone, adrenal gland, testis, and ovary. 12. If many samples must be collected in a short time, collecting dried blood spot (DBS) samples on specific filter paper can be a useful approach. No special sample preparation is required, and only about 80 μL blood is sufficient. In this case, full blood markers can be analyzed. A 3-mm diameter disk, corresponding to 3 μL blood, is punched out of the center of the DBS and used for the measurements. Targeted metabolomics measurements of DBS are described in full detail in [40]. 13. Make system suitability tests to ensure that the LC-MS/MS is performing well before starting with sample preparation. 14. After vortexing, tap the tube hard on the lab bench to ensure no drops of liquid remain at the lid. 15. Zero sample wells contain only internal standards, no analytes. The results of the “empty” samples are used to calculate the LOD. For plasma, PBS buffer is recommended. 16. If the number of study samples exceeds the kit plate capacity (in this case 82 samples), the samples must be measured in more than one batch. In addition to the study samples, five aliquots of a pooled reference sample should be analyzed on each kit plate. These results can be used for calculation of potential batch effects and data normalization between batches. Plasma samples should be stored at 80  C until measurement and be frozen again directly after taking out the aliquot for the preparation of the kit plate. 17. All pipetting steps of one study (even if there is more than one batch) should be done by the same person. Always use the same pipette for the same pipetting volume. 18. If the sample preparation cannot be accomplished in one run, after this step the plate can wait dry up to 2 h or packed air-proof at 20  C for up to 1 week. 19. Prepare the derivatization reagent and dry the samples under a fume hood. 20. The bottom of the upper 96-well plate consists of a semipermeable membrane. During centrifugation, the sample solution passes through the membrane to the lower deep well plate. Captured samples after extraction can be stored in the sealed plate for maximum 2 days at 4  C.

324

Anna Artati et al.

21. It is recommended to start measuring the LC plate, because the LC metabolites are less stable. Ensure that the LC-MS/MS is running properly before diluting the LC plate. Samples diluted for LC and FIA measurements are stable in the auto sampler (10  C) for maximum 48 h. FIA diluted samples are stable up to 7 days at 4  C. Never store the capture plate, the LC plate, or the FIA plate below 0  C. 22. The assay can also be measured with other triple quadrupole mass spectrometers and using other HPLC platforms. 23. For the LC measurements, mass spectrometric compound identification and quantification are based on scheduled multiple reaction monitoring measurements (sMRM). For FIA measurements, compound identification and quantification are based on multiple reaction monitoring (MRM). 24. An additional comment on primary data evaluation: In the p180 kit all metabolites are always quantified for each sample. Nevertheless, for some metabolites, the measurement data may contain a result of “0.” In this case, the concentration of the metabolite is so low that it cannot be detected. Therefore, the value is not missing, but the concentration is below the LOD. 25. Initial statistical tests can be performed directly in the MetIDQ™ software. 26. In plasma samples of healthy donors, about three quarters of the analyzed metabolites are found above LOD in reasonable concentrations. In more detail, according to metabolite groups: carnitine, some acylcarnitines, all amino acids, about half of the biogenic amines, sum of hexoses, most of the glycerophospholipids, and all sphingolipids [53]. In an interlaboratory ring trial, the reproducibility of the p180 kit has been verified [54]. Additionally, the long-time stability of plasma metabolites during storage at 80  C and the longtime performance of the p180 measurements have been evaluated [55]. References 1. Eckel-Mahan K, Sassone-Corsi P (2013) Metabolism and the circadian clock converge. Physiol Rev 93(1):107–135 2. Storch KF, Lipan O, Leykin I, Viswanathan N, Davis FC, Wong WH, Weitz CJ (2002) Extensive and divergent circadian gene expression in liver and heart. Nature 417(6884):78–83 3. Dyar KA, Lutter D, Artati A, Ceglia NJ, Liu Y, Armenta D, Jastroch M, Schneider S, de Mateo S, Cervantes M, Abbondante S, Tognini P, Orozco-Solis R, Kinouchi K, Wang C, Swerdloff R, Nadeef S, Masri S,

Magistretti P, Orlando V, Borrelli E, Uhlenhaut NH, Baldi P, Adamski J, Tscho¨p MH, Eckel-Mahan K, Sassone-Corsi P (2018) Atlas of circadian metabolism reveals system-wide coordination and communication between clocks. Cell 174(6):1571–1585.e11 4. Yang X, Downes M, Yu RT, Bookout AL, He W, Straume M, Mangelsdorf DJ, Evans RM (2006) Nuclear receptor expression links the circadian clock to metabolism. Cell 126(4): 801–810

Circadian Metabolomics 5. Panda S (2016) Circadian physiology of metabolism. Science 354(6315):1008–1015 6. Dyar KA, Hubert MJ, Mir AA, Ciciliot S, Lutter D, Greulich F, Quagliarini F, Kleinert M, Fischer K, Eichmann TO, Wright ˜ a Paz MI, Casarin A, Pertegato V, LE, Pen Romanello V, Albiero M, Mazzucco S, Rizzuto R, Salviati L, Biolo G, Blaauw B, Schiaffino S, Uhlenhaut NH (2018) Transcriptional programming of lipid and amino acid metabolism by the skeletal muscle circadian clock. PLoS Biol 16(8):e2005886 7. Vaze KM, Sharma VK (2013) On the adaptive significance of circadian clocks for their owners. Chronobiol Int 30(4):413–433 8. Held NM, Wefers J, van Weeghel M, Daemen S, Hansen J, Vaz FM, van Moorsel D, Hesselink MKC, Houtkooper RH, Schrauwen P (2020) Skeletal muscle in healthy humans exhibits a day-night rhythm in lipid metabolism. Mol Metab 37:100989 9. Wang H, Wang X, Qi D, Sun M, Hou Q, Li Y, Jiang H (2020) Establishment of the circadian metabolic phenotype strategy in spontaneously hypertensive rats: a dynamic metabolomics study. J Transl Med 18(1):38 10. Honma A, Revell VL, Gunn PJ, Davies SK, Middleton B, Raynaud FI, Skene DJ (2020) Effect of acute total sleep deprivation on plasma melatonin, cortisol and metabolite rhythms in females. Eur J Neurosci 51(1): 366–378 11. Koronowski KB, Kinouchi K, Welz PS, Smith JG, Zinna VM, Shi J, Samad M, Chen S, Magnan CN, Kinchen JM, Li W, Baldi P, Benitah SA, Sassone-Corsi P (2019) Defining the independence of the liver circadian clock. Cell 177(6):1448–1462.e14 12. Sato S, Basse AL, Scho¨nke M, Chen S, Samad M, Altıntas¸ A, Laker RC, Dalbram E, Barre`s R, Baldi P, Treebak JT, Zierath JR, Sassone-Corsi P (2019) Time of exercise specifies the impact on muscle metabolic pathways and systemic energy homeostasis. Cell Metab 30(1):92–110.e4 13. Sun R, Huang J, Yang N, He J, Yu X, Feng S, Xie Y, Wang G, Ye H, Aa J (2019) Purine catabolism shows a dampened circadian rhythmicity in a high-fat diet-induced mouse model of obesity. Molecules 24(24):4524 14. Kervezee L, Cermakian N, Boivin DB (2019) Individual metabolomic signatures of circadian misalignment during simulated night shifts in humans. PLoS Biol 17(6):e3000303 15. Augustijn D, de Groot HJM, Alia A (2019) A robust circadian rhythm of metabolites in Arabidopsis thaliana mutants with enhanced

325

growth characteristics. PLoS One 14(6): e0218219 16. Skene DJ, Skornyakov E, Chowdhury NR, Gajula RP, Middleton B, Satterfield BC, Porter KI, Van Dongen HPA, Gaddameedhi S (2018) Separation of circadian- and behavior-driven metabolite rhythms in humans provides a window on peripheral oscillators and metabolism. Proc Natl Acad Sci U S A 115(30):7825–7830 17. Rhoades SD, Nayak K, Zhang SL, Sehgal A, Weljie AM (2018) Circadian- and light-driven metabolic rhythms in Drosophila melanogaster. J Biol Rhythm 33(2):126–136 18. Isherwood CM, Van der Veen DR, Johnston JD, Skene DJ (2017) Twenty-four-hour rhythmicity of circulating metabolites: effect of body mass and type 2 diabetes. FASEB J 31(12): 5557–5567 19. Krishnaiah SY, Wu G, Altman BJ, Growe J, Rhoades SD, Coldren F, Venkataraman A, Olarerin-George AO, Francey LJ, Mukherjee S, Girish S, Selby CP, Cal S, Er U, Sianati B, Sengupta A, Anafi RC, Kavakli IH, Sancar A, Baur JA, Dang CV, Hogenesch JB, Weljie AM (2017) Clock regulation of metabolites reveals coupling between transcription and metabolism. Cell Metab 25(4):961–974.e4 20. Loizides-Mangold U, Perrin L, Vandereycken B, Betts JA, Walhin JP, Templeman I, Chanon S, Weger BD, Durand C, Robert M, Paz Montoya J, Moniatte M, Karagounis LG, Johnston JD, Gachon F, Lefai E, Riezman H, Dibner C (2017) Lipidomics reveals diurnal lipid oscillations in human skeletal muscle persisting in cellular myotubes cultured in vitro. Proc Natl Acad Sci U S A 114(41):E8565–E8574 21. Abbondante S, Eckel-Mahan KL, Ceglia NJ, Baldi P, Sassone-Corsi P (2016) Comparative circadian metabolomics reveal differential effects of nutritional challenge in the serum and liver. J Biol Chem 291(6):2812–2828 22. Aviram R, Manella G, Kopelman N, NeufeldCohen A, Zwighaft Z, Elimelech M, Adamovich Y, Golik M, Wang C, Han X, Asher G (2016) Lipidomics analyses reveal temporal and spatial lipid organization and uncover daily oscillations in intracellular organelles. Mol Cell 62(4):636–648 23. Adamovich Y, Rousso-Noori L, Zwighaft Z, Neufeld-Cohen A, Golik M, Kraut-Cohen J, Wang M, Han X, Asher G (2014) Circadian clocks and feeding time regulate the oscillations and levels of hepatic triglycerides. Cell Metab 19(2):319–330 24. Eckel-Mahan KL, Patel VR, de Mateo S, Orozco-Solis R, Ceglia NJ, Sahar S, Dilag-

326

Anna Artati et al.

Penilla SA, Dyar KA, Baldi P, Sassone-Corsi P (2013) Reprogramming of the circadian clock by nutritional challenge. Cell 155(7): 1464–1478 25. Eckel-Mahan KL, Patel VR, Mohney RP, Vignola KS, Baldi P, Sassone-Corsi P (2012) Coordination of the transcriptome and metabolome by the circadian clock. Proc Natl Acad Sci U S A 109(14):5541–5546 26. Dyar KA, Artati A, Cecil A, Adamski J (2019) Skeletal muscle metabolomics for metabolic phenotyping and biomarker discovery. In: Burniston J, Chen YW (eds) Omics approaches to understanding muscle biology, Methods in physiology. Springer, New York 27. Hughes ME, Abruzzi KC, Allada R, Anafi R, Arpat AB, Asher G, Baldi P, de Bekker C, BellPedersen D, Blau J, Brown S, Ceriani MF, Chen Z, Chiu JC, Cox J, Crowell AM, DeBruyne JP, Dijk DJ, DiTacchio L, Doyle FJ, Duffield GE, Dunlap JC, Eckel-Mahan K, Esser KA, FitzGerald GA, Forger DB, Francey LJ, Fu YH, Gachon F, Gatfield D, de Goede P, Golden SS, Green C, Harer J, Harmer S, Haspel J, Hastings MH, Herzel H, Herzog ED, Hoffmann C, Hong C, Hughey JJ, Hurley JM, de la Iglesia HO, Johnson C, Kay SA, Koike N, Kornacker K, Kramer A, Lamia K, Leise T, Lewis SA, Li J, Li X, Liu AC, Loros JJ, Martino TA, Menet JS, Merrow M, Millar AJ, Mockler T, Naef F, Nagoshi E, Nitabach MN, Olmedo M, Nusinow DA, Ptacek LJ, Rand D, Reddy AB, Robles MS, Roenneberg T, Rosbash M, Ruben MD, Rund SSC, Sancar A, Sassone-Corsi P, Sehgal A, Sherrill-Mix S, Skene DJ, Storch KF, Takahashi JS, Ueda HR, Wang H, Weitz C, Westermark PO, Wijnen H, Xu Y, Wu G, Yoo SH, Young M, Zhang EE, Zielinski T, Hogenesch JB (2017) Guidelines for genome-scale analysis of biological rhythms. J Biol Rhythm 32(5):380–393 28. Wirz-Justice A (2007) How to measure circadian rhythms in humans. Medicographia 29:1 29. Wishart DS (2019) Metabolomics for investigating physiological and pathophysiological processes. Physiol Rev 99(4):1819–1875 30. Malik DM, Paschos GK, Sehgal A, Weljie AM (2020) Circadian and sleep metabolomics across species. J Mol Biol 432(12):3578–3610 31. Dyar KA, Eckel-Mahan KL (2017) Circadian metabolomics in time and space. Front Neurosci 11:369 32. Liang X, FitzGerald GA (2017) Timing the microbes: the circadian rhythm of the gut microbiome. J Biol Rhythm 32(6):505–515 33. Rhoades SD, Sengupta A, Weljie AM (2017) Time is ripe: maturation of metabolomics in

chronobiology. Curr Opin Biotechnol 43:70– 76 34. Gooley JJ (2016) Circadian regulation of lipid metabolism. Proc Nutr Soc 75(4):440–450 35. Brown SA (2016) Circadian metabolism: from mechanisms to metabolomics and medicine. Trends Endocrinol Metab 27(6):415–426 36. Loizides-Mangold U, Petrenko V, Dibner C (2021) Circadian lipidomics: analysis of lipid metabolites around the clock. Methods Mol Biol 2130:169–183 37. Aviram R, Wang C, Han X, Asher G (2021) A lipidomics view of circadian biology. Methods Mol Biol 2130:157–168 38. Brown SA, Sinues P (2021) Circadian metabolomics from breath. Methods Mol Biol 2130: 149–156 39. Dyar KA, Ciciliot S, Wright LE, Biensø RS, Tagliazucchi GM, Patel VR, Forcato M, Paz MI, Gudiksen A, Solagna F, Albiero M, Moretti I, Eckel-Mahan KL, Baldi P, SassoneCorsi P, Rizzuto R, Bicciato S, Pilegaard H, Blaauw B, Schiaffino S (2013) Muscle insulin sensitivity and glucose metabolism are controlled by the intrinsic muscle clock. Mol Metab 3(1):29–41 40. Zukunft S, Sorgenfrei M, Prehn C, Mo¨ller G, Adamski J (2013) Targeted metabolomics of dried blood spot extracts. Chromatographia 76:1295–1305 41. Ro¨misch-Margl W, Prehn C, Bogumil R, Ro¨hring C, Suhre K, Adamski J (2012) Procedure for tissue sample preparation and metabolite extraction for high-throughput targeted metabolomics. Metabolomics 8:133–142 42. Zukunft S, Prehn C, Ro¨hring C, Mo¨ller G, Hrabeˇ de Angelis M, Adamski J, Tokarz J (2018) High-throughput extraction and quantification method for targeted metabolomics in murine tissues. Metabolomics 14(1):18 43. Chong J, Xia J (2018) MetaboAnalystR: an R package for flexible and reproducible analysis of metabolomics data. Bioinformatics 34(24): 4313–4314 44. Aggio R, Villas-Boˆas SG, Ruggiero K (2011) Metab: an R package for high-throughput analysis of metabolomics data generated by GC-MS. Bioinformatics 27(16):2316–2318 45. Schiffman C, Petrick L, Perttula K, Yano Y, Carlsson H, Whitehead T, Metayer C, Hayes J, Rappaport S, Dudoit S (2019) Filtering procedures for untargeted LC-MS metabolomics data. BMC Bioinformatics 20(1):334 46. van den Berg RA, Hoefsloot HC, Westerhuis JA, Smilde AK, van der Werf MJ (2006) Centering, scaling, and transformations: improving

Circadian Metabolomics the biological information content of metabolomics data. BMC Genomics 7:142 47. Gromski PS, Xu Y, Hollywood KA, Turner ML, Goodacre R (2015) The influence of scaling metabolomics data on model classification accuracy. Metabolomics 11:684–695 48. Wei R, Wang J, Su M, Jia E, Chen S, Chen T, Ni Y (2018) Missing value imputation approach for mass spectrometry-based metabolomics data. Sci Rep 8(1):663 49. Hughes ME, Hogenesch JB, Kornacker K (2010) JTK_CYCLE: an efficient nonparametric algorithm for detecting rhythmic components in genome-scale data sets. J Biol Rhythm 25(5):372–380 50. Agostinelli F, Ceglia N, Shahbaba B, SassoneCorsi P, Baldi P (2016) What time is it? Deep learning approaches for circadian rhythms. Bioinformatics 32(12):i8–i17 51. Broadhurst D, Goodacre R, Reinke SN, Kuligowski J, Wilson ID, Lewis MR, Dunn WB (2018) Guidelines and considerations for the use of system suitability and quality control samples in mass spectrometry assays applied in

327

untargeted clinical metabolomic studies. Metabolomics 14(6):72 52. Committee for Medicinal Products for Human Use (CHMP). Guideline on bioanalytical method validation. EMEA/CHMP/EWP/ 192217/2009 Rev. 1 Corr. 2, 21 July 2011 53. Breier M, Wahl S, Prehn C, Fugmann M, Ferrari U, Weise M, Banning F, Seissler J, Grallert H, Adamski J, Lechner A (2014) Targeted metabolomics identifies reliable and stable metabolites in human serum and plasma samples. PLoS One 9(2):e89728 54. Siskos AP, Jain P, Romisch-Margl W, Bennett M, Achaintre D, Asad Y, Marney L, Richardson L, Koulman A, Griffin JL et al (2017) Interlaboratory reproducibility of a targeted metabolomics platform for analysis of human serum and plasma. Anal Chem 89: 656–665 55. Haid M, Muschet C, Wahl S, Ro¨mischMargl W, Prehn C, Mo¨ller G, Adamski J (2018) Long-term stability of human plasma metabolites during storage at 80 C. J Proteome Res 17(1):203–211

Chapter 22 Time-Restricted Feeding and Caloric Restriction: Two Feeding Regimens at the Crossroad of Metabolic and Circadian Regulation Amandine Chaix Abstract In addition to diet quality and quantity, the “timing” of food intake recently emerged as a third key parameter in nutritional and metabolic health. The link between nutrition timing and metabolic homeostasis is in part due to the regulation of daily feeding:fasting cycles and metabolic pathways by the circadian clock. Preclinical feeding regimen studies in rodents are invaluable to further define the modalities of this relationship and get a better understanding of its mechanistic underpinnings. Time-restricted feeding (TRF) and caloric restriction (CR) are examples of feeding regimen at the crossroads of metabolic and circadian regulation. Here we propose methods to implement TRF and CR highlighting the parameters that are relevant to the study of circadian and metabolic health. We also provide methods to determine their impact on the output of the circadian clock by analyzing diurnal expression profiles using 24 h time-series collection as well as their impact on metabolic homeostasis using a glucose tolerance test (GTT). Key words Nutrition, Metabolism, Intermittent fasting, Circadian clock, Circadian rhythms

1

Introduction A feeding regimen is a plan that specifies a diet type, amount, and schedule of nutritional intake. Beyond their role in cardiometabolic health, the quality, quantity, and timing of dietary intake can also affect the activity of the circadian clock and circadian rhythms. In this chapter, we highlight this relationship and propose methods to investigate it. The first section of the introduction will provide a short overview of the interaction between nutrition, metabolic health, and the circadian clock. In the second section of the introduction, we will describe experimental parameters that influence feeding regimen studies while highlighting the ones that can specifically affect the circadian clock. Then we will discuss different approaches to implementing feeding regimen studies. Subheading 3 describes possible protocols to implement manually two feeding

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_22, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

329

330

Amandine Chaix

regimens that lie at the interface between metabolic homeostasis and the circadian clock, namely time-restricted feeding (TRF) and calorie restriction (CR). We also describe methods to evaluate the activity of the circadian clock and metabolic health under these feeding interventions. In particular, we suggest a protocol for performing a glucose tolerance test that takes into account the challenges associated with the comparisons between groups under different feeding regimen. 1.1 Nutrition, Metabolic Health, and the Circadian Clock

Nutrition plays a crucial role in overall health and well-being. Good dietary habits are associated with lower risks of many chronic diseases that affect a large fraction of the World’s population [1, 2]. Reasonably, dietary guidelines emphasize the importance of the quality and composition of the food we consume as well as the quantity needed to maintain one’s caloric balance (https:// www.dietaryguidelines.gov [3]). The “timing” of food intake recently emerged as a third key parameter [4–7]. The temporal aspect of nutrition encompasses both the duration of the daily eating window (number of hours between first and last caloric intake of the day) and the regularity of one’s eating habits. The link between nutrition timing and metabolic health is in part due to the regulation of daily feeding:fasting cycles and metabolic pathways by the circadian clock [8, 9]. There is a bidirectional relationship between the circadian clock and metabolic health. On the one hand, perturbations to the circadian clock function are associated with increased risk of developing metabolic disease. Indeed, epidemiological studies have shown that populations with disrupted circadian rhythms such as shift workers have increased risk of metabolic disease [10]. In addition, genetic disruption of the clock in a variety of animal models leads to metabolic dysfunction [11]. On the other hand, metabolic disease can be associated with alterations in the clock function. Landmarks studies have shown that one of the most widely used mouse preclinical model of obesity and metabolic disease, the dietinduced obesity (DIO) model—display dampened daily activity rhythms and altered daily feeding rhythms [12]. Further studies using timed feeding and time-restricted feeding (TRF) have then demonstrated that mistimed food consumption is associated with increased weight gain and metabolic disorders [13–16]. Rodent preclinical models of obesity, diabetes, and cardiometabolic disease have been vastly used in nutrition and metabolism research to understand the mechanistic underpinnings of the diseases and explore therapeutics approaches. Now they are also used to study the interaction between food timing, circadian clock, and metabolic health. In this chapter, we focus on time-restricted feeding (TRF) and caloric restriction (CR), two feeding regimens that are at the crossroad of metabolic and circadian regulation. TRF is a feeding regimen wherein food intake is limited to a consistent

Feeding Regimen and Circadian Rhythms

331

8–10 h daily window without changes in nutritional quality or quantity. TRF of a high fat diet has been shown to protect and reverse obesity and associated metabolic disorders [13–15, 17– 21]. From the clock standpoint, TRF prevents dampening of the circadian clock associated with the consumption of an energy-dense diet and restores high amplitude diurnal rhythms in clock gene expression and physiological output of the circadian clock [13, 15, 22]. CR is the most established and well-characterized intervention to slow down aging and prevent chronic disease. For example, a 15–40% daily caloric reduction can delay age-related diseases and increase life span in worms, flies, rodents, and non-human primates [23]. Notably, CR in mice can be seen as an extreme form of TRF since mice usually eat all their food within 3 h. This temporal restriction on eating could be in part responsible for the health benefits of CR. CR can also affect the activity of both the central and peripheral clocks [24–26]. Whether the clock is necessary for the benefits of CR is still under investigation but at least in C57Bl/6 mice, BMAL1, a master component of the clock, has been shown to be required for the benefits of CR [27]. Thus, CR and TRF are two feeding regimens that have strong metabolic and health benefits but also impinge on the activity of the circadian clock. 1.2 Feeding Regimen Studies

As for any research using mouse models, studies that involve modifications in the feeding regimen should mention the strain, genotype, age, and sex of the animals being used. In addition, when conducting feeding regimen intervention studies, other parameters that need to be specified are: (1) composition and energy content of the diet, (2) duration of the feeding intervention, (3) age of the mice upon initiation, (4) housing information such as animal density and bedding materials, and (5) timing of food access relative to the light:dark (L:D) cycle when applicable. The latest point is key when studying feeding-clock interaction since the timing of food intake can affect the activity of the circadian clock. This will be discussed in further details with the example of TRF and CR in Subheading 3. There are a variety of methods to implement feeding regimen studies and monitor feeding behaviors in rodents that can broadly be divided into automatic versus manual approaches. Automatic systems to control food availability, monitor food consumption or both are available from commercials vendors (Columbus Instruments, TSE Systems, Sable Systems, Research Diets BioDAQ, etc.) and have also been developed in research laboratories lab [26, 28]. Commercial systems are usually integrated with other technologies such as activity or calorimetry analysis into full “metabolic phenotyping units/cages” which allow for high frequency and simultaneous measurements of food consumption, activity, and energetics patterns (CLAMS, PhenoMaster, Promethion).

332

Amandine Chaix

These systems are relatively expensive and often require single housing which make them not easily amendable to large-scale studies. Some laboratories have developed custom-made food dispensers that are much cheaper and allow the delivery of a specific amount of food at a given time and frequency (without recording the actual amount of food eaten). On the other hand, manual approaches can also be used to perform feeding regimen studies. They are labor intensive when aiming to monitor body weight and food consumption and might be less precise that automated approaches but they are much easier to adopt with large cohort studies. In this chapter we provide possible methods to implement TRF and CR manually in mice as well as to evaluate the activity of the circadian clock and metabolic health under these feeding interventions. There are many ways of implementing feeding regimen studies and the optimal design and approach depends on the research question and the feasibility at one’s research institution. Our goal is to highlight some of the parameters that are especially relevant to TRF and CR in the context of investigating feedingclock interactions. Since these parameters might influence the study outcomes, they should ideally be controlled as much as possible in the approach and precisely described in Subheading 3. We feel that providing a detailed description of the protocol implemented goes a long way in allowing the reproduction of research findings between research institutions. Ultimately, we hope that this description will help researchers in the rigorous design and performance of these types of studies as well as their reproducibility between different research teams, for the purpose of scientific advancement.

2

Materials

2.1 Manual Implementation of TRF and CR

When implementing TRF or CR during the dark phase, having the mice on a reverse L:D cycle can dramatically increase the feasibility of a long-term intervention while minimizing the disruption to the experimenters own circadian rhythms among others. This is achievable by housing the animal cohort either in a reverse L:D cycle housing room or in a light-tight cabinet in which the lighting schedule can be controlled automatically (“circadian cabinets”). Other materials needed to implement TRF and CR manually are: 1. Diet. 2. Scale to record body weight and food weight. 3. Diet containers.

Feeding Regimen and Circadian Rhythms

2.2 Twenty-four Hour Time-Series Collection (See Note 1)

333

The materials needed for a 24 h time-series collection are: 1. Dissection tools and board. 2. Liquid nitrogen for fast freezing of the collected organs. 3. Pre-labeled collection tubes. 4. Light-proof cage/animal container to prevent light exposure during the dark phase if the mice are being relocated for collection.

2.3

ipGTT

The materials needed to perform an ipGTT are: 1. Clean cages for fasting—paper bedding recommended. 2. Scale to record body weight. 3. Sharp razor blade to make a small cut on the tail. 4. Digital glucose reader and strips to measure blood glucose. 5. D-Glucose to prepare glucose solution to inject. 6. Insulin syringes to inject glucose bolus. 7. Timer to monitor blood sampling time.

3

Methods

3.1 Manual Implementation of TRF Studies in Rodents (See Note 2)

1. Set the duration of the daily feeding window. Classically, mice on TRF have ad libitum access to food for a period of 8–12 h daily 7 days a week. The amount of food provided every day should be ample enough that the animals don’t run out of food during the feeding interval. 2. Set the timing of the feeding window relative to the light:dark cycle in the animal housing room. Per convention, Zeitgeber Time 0 (ZT0) denotes the time at which lights turn on in the housing room (see Note 3). For TRF during the light phase, food will be present in the cage for a chosen number of hours (step 1) between ZT0 and ZT12. For TRF during the dark phase, food will be present in the cage between ZT12 and ZT24. Classically, for 9 h TRF during the dark phase, food is provided between ZT13 and ZT21 [13, 17]. 3. Prepare the food. If monitoring food consumption is part of the experimental design, an accurate measure of weekly intake can be obtained by weighing the amount of food placed in the cage and the amount of food left at the end of a week. The amount of food required for 1 week can be evaluated from ad libitum cages on the same diet and should be adapted to the housing density.

334

Amandine Chaix

4. Implement TRF daily. There are different methods to implement TRF. Two of them are presented below: Method 1: (a) Add food in the hopper at the set time-of-day (see step 2). (b) Take out food at the end of the chosen feeding duration (see step 1). (c) When monitoring food consumption, use food from a container dedicated to each cage to take food in and out from the cage. Prepare the food container for each cage that is being monitored with at least a week worth of food accurately weighed and recorded. Weigh the food left in every container after 1 week. Weekly food consumption can be measured in grams by subtracting the amount of food left to the amount of food put in. Daily food consumption in kcal per mouse can be calculated based on the diet caloric density, the number of animal in the cage, and the actual number of days of food monitoring. Method 2 (see Notes 4 and 5): (a) Place mice in a cage that contains food at the set time-of-day (see step 2), the “feeding cage.” (b) At the end of the chosen feeding duration (see step 1), transfer mice in a cage without food, the “fasting cage.” (c) When monitoring food consumption, place at least a week worth of food accurately weighed and recorded in the “feeding cage” and weigh the food left in the cage after 1 week. (d) This method involves twice-a-day animal handling that should be reflected in the controls. 3.2 Manual Implementation of Caloric Restriction (CR) in Rodents (See Note 2)

1. Set the percentage of caloric restriction. Classically, mice on CR are fed every day with 60–70% of the total daily caloric intake of mice fed ad libitum. The amount of food to be provided every day can be evaluated prior to the start of the intervention from monitoring food consumption in similar mice fed ad libitum with the same diet. The exact quantity of food to be provided should be adapted to the housing density. When switching an ad libitum fed cohort to CR, adopting a weekly 10% reduction approach over 2 or 3 weeks can be helpful in preventing an abrupt drop in body weight. If an ad libitum control cohort is run in parallel, weekly monitoring of food consumption in this group can allow to fine

Feeding Regimen and Circadian Rhythms

335

tune the amount of food to give to the CR group—with a 1 week delay. A decrease in the caloric intake is expected as the mice age. 2. Set the timing and frequency of feeding. Mice on CR can be fed the entire amount of food at once daily—this is the most current approach—or by portions across the day. The later usually requires the use of automated food dispensers. These two approaches are associated with very distinct feeding patterns but also activity profiles reflecting that they act differently on the activity of the central clock. When feeding the mice at once, the time-of-day of food addition is also critical, since food anticipatory activity (FAA) is different when the food bolus is provided during the light or the dark phase (see Note 6 for more details). In any case, most mice on CR eat the entire amount of food provided within 2–3 h, leading to an extended period of fasting [26]. 3. Prepare the food. If monitoring food consumption is part of the experiment, an accurate measure of daily intake can be obtained by weighing the amount of food placed in the cage and the amount of food left at the end of a day. In most cases, it should be 0. 4. Implement CR daily with a one-time daily addition of food (a) Add set amount of food (see step 1) in the hopper at the set time-of-day (see step 2). (b) Take out food left—if any—after 24 h (see Note 7). (c) When monitoring food consumption, use food from a container dedicated to each cage to take food in and out from the cage. Record the amount of food left daily—if any—to compute weekly food consumption for each cage. Daily food consumption in kcal per mouse can be calculated further based on the diet caloric density, the number of animals in the cage, and the actual number of days of food monitoring. 3.3 Evaluation of the Activity of the Molecular Circadian Clock in Rodents from 24 h Time-Series Collection (See Note 8)

To assess the activity of the clock at the molecular levels, samples are collected at specific times across a day—a 24 h time-series collection. These samples can then be analyzed by targeted or untargeted omics approaches (see Chapter 21). Guidelines for the experimental design to obtain and analyze biological rhythms in genome-scale datasets are available in the landmark consortium paper from Michael E. Hugues [29]. In particular, this manuscript highlights the complexity of choosing “the right combination of replicates and temporal resolution for their intended application.” This manuscript is a must read when deciding on the parameters described below.

336

Amandine Chaix

1. Set the sampling frequency. Most circadian time-series studies in rodents are performed with a sampling interval of 3–4 h (i.e., 6 or 8 time points per day respectively). Biological replicates at each time point are also recommended. For example, for a discovery study comparing the effect of ad libitum to TRF feeding on the diurnal transcriptome in various clock mutants, we chose to use a sampling interval of 3 or 4 h with biological duplicates at each time point [17]. That is 12–16 animals per feeding group per genotype. Thus the number of animals required for a circadian collection is significant, usually exceeding the number required for a steady state analysis. 2. Choose the approach to cover 24 h. There are two approaches for covering a 24-h cycle: (1) all in 1 day, (2) in multiple shifts. The numbers of mice and organs to be collected as well as available personnel are key parameters to inform this choice. Regardless of the approach, 24 h timeseries collections are strenuous and require careful planning. 3. Collect. Perform traditional collection and flash freezing of the samples collected at each time point. Limit mice exposure to light prior to collection during the dark phase time points. 3.4 Performing an Intra-peritoneal Glucose Tolerance Test (ipGTT) in Mice Under Different Feeding Regimen

Most feeding regimen studies include the measurements of readouts of metabolic health such as serum glycemia, insulinemia, and lipidemia as well as kinetics of glucose processing and insulin action using glucose or insulin tolerance tests, respectively. The results of these assays are highly dependent on the fasting state of the mice. Recommendations exist to guide the investigators when performing these assays in order to standardize them and increase reproducibility [30]. However, especially when comparing different groups of mice undergoing chronic fasting intervention, the standard procedure might not be applicable (see Note 9). For GTT assays, 14–18 h long “overnight fast” is often used. In most cases the overnight fast occurs during the dark phase, which is the phase at which mice consume the most of their food, thus resulting in the depletion of energy stores, especially hepatic glycogen. In the case of TRF, mice undergo 15–16 h long fast daily, of which 12 h are happening during the light phase. For mice under CR, most animals undergo >20 h of fasting daily. Here we propose an ipGTT protocol to compare mice under ALF and TRF groups. 1. Fast the mice. We recommend fasting the ALF group at the same time that the TRF group is fasted daily so both groups experience the same fasting duration prior to the assay. (a) Transfer the mice to a clean cage with water and without food.

Feeding Regimen and Circadian Rhythms

337

A clean cage is preferred to the home cage without food to prevent coprophagy. Non-corncob bedding (that could be eaten by the mice during fasting) is also preferred. (b) Properly identify the fasting cages according to local institutional animal care and use committee (IACUC) requirements. 2. Set the time-of-day of the experiment. Since insulin sensitivity varies according to the time-of-day, it’s highly recommended to perform all metabolic studies at the same ZT time especially if longitudinal measurements are presented (see Note 3). The time chosen should be indicated in the methods section. For GTT we recommend setting the time of the glucose bolus at a time where the TRF group is usually fed. 3. Perform GTT at the set time-of day. (a) Weigh the mice and make them identifiable (e.g., tail tag with a marker). (b) Measure basal glycemia with a strip glucometer from tail blood obtained after a tail tip cut with a sharp sterile blade. (c) Inject glucose at a dose of 1–2 g/kg body weight. (d) At 15, 30, 60, and 120 min following the IP injection, blood glucose levels are measured. Between sampling times mice will be returned to their cage and monitored for any changes in behavior.

4

Notes 1. For nighttime collection, we recommend exceptional deviation from standard laboratory practice with ample access to food and entertainment and a place to rest for the experimenters. 2. Water bottle is provided at all time in all cages. Cages are changed to clean cages set up based on local IACUC policies. 3. Zeitgeber Time or ZT is a way to define time relative to a circadian entrainment cue. Here, light is the main cue and thus ZT0 is defined as lights ON. Referencing ZT rather than clock time of an experiment is fundamental since different vivarium have different timings for turning lights on and off. Whether the animals are housed in 12 h light:12 h dark cycle is also relevant since the lengths of the L:D cycles can affect the activity of the clock. 4. If the experimenter observes a significant amount of food on the floor in the cage, consider implementing method 2 to prevent unwarranted food consumption during the fasting interval. This typically depends on the diet and strains being

338

Amandine Chaix

used. Generally, the harder the diet and the younger the mice, the more diet will be found on the cage floor from mice grinding the food pellet. 5. Compared to method 1, method 2 leads to an additional bout of activity from exploratory behavior in the fasting cage when old cages are switched to clean ones. It usually disappears after a day when mice are acclimatized to their new home cage. 6. Mice on CR display a peak of activity prior to the addition of food, a phenomenon called food anticipatory activity (FAA). CR during the light phase has been used in the circadian field to try to identify a food-entrainable oscillator outside of the SCN—the hypothalamic nucleus where the brain master clock is located—responsible for this behavioral response. This bout of activity is different in mice on CR fed during the light or the dark phase, resulting in differences in both the phase and the diurnal profile of activity between those two. 7. Presence of leftover food in the hopper in a CR cage is a strong indicator that either the calculations were incorrect or that the health of the animal(s) is compromised. In our experience, very occasional leftover food in a cage 1 day has been observed but almost certainly points at a sick animal when occurring 2 days in a row. 8. Monitoring daily rhythms in activity:rest is the gold standard method to evaluate the activity of the central circadian clock. The amplitude and the phase of these rhythms can be evaluated using various techniques such as wheel-running, video tracking, infrared beam breaks, and telemetry, allowing to compare the effect of various feeding regimens on a physiological output of the activity of the central clock. As discussed previously, both TRF and CR have been shown to modulate activity:rest rhythms in mice. 9. To our knowledge, there is no consensus on recommended fasting duration and phase relative to the L:D cycle for different metabolic assays under chronic fasting interventions. Thus, we will just echo the recommendations of the MMPC: “It is therefore crucial for these factors to be described accurately in methods sections.” [30].

Acknowledgments I would like to acknowledge the numerous outstanding scientists from every stage of the scientific career path in the Panda and Chaix’s lab who have made it possible for me to run TRF experiments for almost 10 years. I also would like to thank Patrick-Simon Welz for careful and helpful reading and editing of this chapter. This chapter is dedicated to Drs. Paolo Sassone-Corsi and Michael

Feeding Regimen and Circadian Rhythms

339

E. Hughes for their influential role in my scientific training as a chronobiologist. AC is supported by grants from the National Institutes of Health (NIH) R01 AG065993 and from the American Heart Association (AHA) 18CDA34110292. References 1. GBD 2017 Diet Collaborators (2019) Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 393(10184):1958–1972 2. Micha R, Shulkin ML, Penalvo JL, Khatibzadeh S, Singh GM, Rao M, Fahimi S, Powles J, Mozaffarian D (2017) Etiologic effects and optimal intakes of foods and nutrients for risk of cardiovascular diseases and diabetes: systematic reviews and meta-analyses from the Nutrition and Chronic Diseases Expert Group (NutriCoDE). PLoS One 12(4):e0175149 3. DeSalvo KB, Olson R, Casavale KO (2016) Dietary guidelines for Americans. JAMA 315(5):457–458 4. Mattson MP, Allison DB, Fontana L, Harvie M, Longo VD, Malaisse WJ, Mosley M, Notterpek L, Ravussin E, Scheer FA, Seyfried TN, Varady KA, Panda S (2014) Meal frequency and timing in health and disease. Proc Natl Acad Sci U S A 111(47): 16647–16653 5. Beccuti G, Monagheddu C, Evangelista A, Ciccone G, Broglio F, Soldati L, Bo S (2017) Timing of food intake: sounding the alarm about metabolic impairments? A systematic review. Pharmacol Res 125(Pt B):132–141 6. Manoogian ENC, Chaix A, Panda S (2019) When to eat: the importance of eating patterns in health and disease. J Biol Rhythm 34(6): 579–581 7. Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabolism, and the circadian clock. Cell 161(1):84–92 8. Bass J, Lazar MA (2016) Circadian time signatures of fitness and disease. Science 354(6315): 994–999 9. Panda S (2016) Circadian physiology of metabolism. Science 354(6315):1008–1015 10. Pan A, Schernhammer ES, Sun Q, Hu FB (2011) Rotating night shift work and risk of type 2 diabetes: two prospective cohort studies in women. PLoS Med 8(12):e1001141 11. Zarrinpar A, Chaix A, Panda S (2016) Daily eating patterns and their impact on health and

disease. Trends Endocrinol Metab 27(2): 69–83 12. Kohsaka A, Laposky AD, Ramsey KM, Estrada C, Joshu C, Kobayashi Y, Turek FW, Bass J (2007) High-fat diet disrupts behavioral and molecular circadian rhythms in mice. Cell Metab 6(5):414–421 13. Chaix A, Zarrinpar A, Miu P, Panda S (2014) Time-restricted feeding is a preventative and therapeutic intervention against diverse nutritional challenges. Cell Metab 20(6):991–1005 14. Sherman H, Genzer Y, Cohen R, Chapnik N, Madar Z, Froy O (2012) Timed high-fat diet resets circadian metabolism and prevents obesity. FASEB J 26(8):3493–3502 15. Hatori M, Vollmers C, Zarrinpar A, DiTacchio L, Bushong EA, Gill S, Leblanc M, Chaix A, Joens M, Fitzpatrick JA, Ellisman MH, Panda S (2012) Time-restricted feeding without reducing caloric intake prevents metabolic diseases in mice fed a high-fat diet. Cell Metab 15(6):848–860 16. Arble DM, Bass J, Laposky AD, Vitaterna MH, Turek FW (2009) Circadian timing of food intake contributes to weight gain. Obesity (Silver Spring) 17(11):2100–2102 17. Chaix A, Lin T, Le HD, Chang MW, Panda S (2019) Time-restricted feeding prevents obesity and metabolic syndrome in mice lacking a circadian clock. Cell Metab 29(2):303–319.e4 18. Woodie LN, Luo Y, Wayne MJ, Graff EC, Ahmed B, O’Neill AM, Greene MW (2018) Restricted feeding for 9h in the active period partially abrogates the detrimental metabolic effects of a Western diet with liquid sugar consumption in mice. Metabolism 82:1–13 19. Delahaye LB, Bloomer RJ, Butawan MB, Wyman JM, Hill JL, Lee HW, Liu AC, McAllan L, Han JC, van der Merwe M (2018) Time-restricted feeding of a high-fat diet in male C57BL/6 mice reduces adiposity but does not protect against increased systemic inflammation. Appl Physiol Nutr Metab 43(10):1033–1042 20. Sundaram S, Yan L (2016) Time-restricted feeding reduces adiposity in mice fed a highfat diet. Nutr Res 36(6):603–611

340

Amandine Chaix

21. Duncan MJ, Smith JT, Narbaiza J, Mueez F, Bustle LB, Qureshi S, Fieseler C, Legan SJ (2016) Restricting feeding to the active phase in middle-aged mice attenuates adverse metabolic effects of a high-fat diet. Physiol Behav 167:1–9 22. Damiola F, Le Minh N, Preitner N, Kornmann B, Fleury-Olela F, Schibler U (2000) Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev 14(23):2950–2961 23. Fontana L, Partridge L, Longo VD (2010) Extending healthy life span—from yeast to humans. Science 328(5976):321–326 24. Solanas G, Peixoto FO, Perdiguero E, Jardi M, Ruiz-Bonilla V, Datta D, Symeonidi A, Castellanos A, Welz PS, Caballero JM, Sassone-Corsi P, Munoz-Canoves P, Benitah SA (2017) Aged stem cells reprogram their daily rhythmic functions to adapt to stress. Cell 170(4):678–692.e20 25. Challet E (2010) Interactions between light, mealtime and calorie restriction to control daily timing in mammals. J Comp Physiol B 180(5):631–644 26. Acosta-Rodriguez VA, de Groot MHM, RijoFerreira F, Green CB, Takahashi JS (2017) Mice under caloric restriction self-impose a temporal restriction of food intake as revealed by an automated feeder system. Cell Metab 26(1):267–277.e2 27. Patel SA, Chaudhari A, Gupta R, Velingkaar N, Kondratov RV (2016) Circadian clocks govern calorie restriction-mediated life span extension through BMAL1- and IGF-1-dependent mechanisms. FASEB J 30(4):1634–1642 28. Greenwell BJ, Trott AJ, Beytebiere JR, Pao S, Bosley A, Beach E, Finegan P, Hernandez C,

Menet JS (2019) Rhythmic food intake drives rhythmic gene expression more potently than the hepatic circadian clock in mice. Cell Rep 27(3):649–657.e5 29. Hughes ME, Abruzzi KC, Allada R, Anafi R, Arpat AB, Asher G, Baldi P, de Bekker C, BellPedersen D, Blau J, Brown S, Ceriani MF, Chen Z, Chiu JC, Cox J, Crowell AM, DeBruyne JP, Dijk DJ, DiTacchio L, Doyle FJ, Duffield GE, Dunlap JC, Eckel-Mahan K, Esser KA, FitzGerald GA, Forger DB, Francey LJ, Fu YH, Gachon F, Gatfield D, de Goede P, Golden SS, Green C, Harer J, Harmer S, Haspel J, Hastings MH, Herzel H, Herzog ED, Hoffmann C, Hong C, Hughey JJ, Hurley JM, de la Iglesia HO, Johnson C, Kay SA, Koike N, Kornacker K, Kramer A, Lamia K, Leise T, Lewis SA, Li J, Li X, Liu AC, Loros JJ, Martino TA, Menet JS, Merrow M, Millar AJ, Mockler T, Naef F, Nagoshi E, Nitabach MN, Olmedo M, Nusinow DA, Ptacek LJ, Rand D, Reddy AB, Robles MS, Roenneberg T, Rosbash M, Ruben MD, Rund SSC, Sancar A, Sassone-Corsi P, Sehgal A, Sherrill-Mix S, Skene DJ, Storch KF, Takahashi JS, Ueda HR, Wang H, Weitz C, Westermark PO, Wijnen H, Xu Y, Wu G, Yoo SH, Young M, Zhang EE, Zielinski T, Hogenesch JB (2017) Guidelines for genome-scale analysis of biological rhythms. J Biol Rhythm 32(5):380–393 30. Ayala JE, Samuel VT, Morton GJ, Obici S, Croniger CM, Shulman GI, Wasserman DH, McGuinness OP, N.I.H.M.M.P.C. Consortium (2010) Standard operating procedures for describing and performing metabolic tests of glucose homeostasis in mice. Dis Model Mech 3(9–10):525–534

Chapter 23 Chromatin Immunoprecipitation and Circadian Rhythms Kenichiro Kinouchi, Kazutoshi Miyashita, and Hiroshi Itoh Abstract Organisms exhibit daily changes of physiology and behavior to exert homeostatic adaptations to day-night cycles. The cyclic fluctuation also takes place at transcriptional levels, giving rise to rhythmic gene expression. Central to this oscillatory transcription is the core clock machinery which constitutes a circuit of transcriptional-translational feedback and achieves circadian functions accordingly. Chromatin immunoprecipitation provides understanding of such mechanisms that clock and non-clock transcription factors along with co-regulators and chromatin modifications dictate circadian epigenome through cyclic alterations of chromatin structures and molecular functions in a concerted fashion. Besides, innovation of highthroughput sequencing technology has broadened our horizon and renewed perspectives in circadian research. This article summarizes the methodology of a chromatin immunoprecipitation experiment in light of circadian rhythm research. Key words Dual cross-linking, DNA shearing, Transcription factors, Histone marks, Circadian rhythms

1

Introduction In synchrony with light-dark cycles, organisms possess inherent rhythms of sleep, feeding, and hormone levels, among others, which are predominantly driven by the circadian clock [1, 2]. Mammalian clocks comprise central pacemaker in the hypothalamus and peripheral oscillators. Light is a major environmental cue to reset the central clock, while food is a zeitgeber (time giver) for peripheral clocks. Remarkably, light signals are also conveyed to peripheral tissues such as the skin and muscle, directing circadian oscillation of cell cycles and insulin sensitivity, respectively [3, 4]. Circadian disruption such as rotating shift working and mistimed eating is associated with a wide array of pathologies such as metabolic syndrome and cancer [5, 6]. The molecular clock components consist of transcription factors (TFs) and co-regulators [7]. In nocturnal mammals, circadian activators CLOCK (Circadian Locomotor Output Cycles Kaput) and BMAL1 (Brain and Muscle ARNT-like

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_23, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

341

342

Kenichiro Kinouchi et al.

protein-1) form a heterodimer and induce transcription of their target genes during the day [8]. These genes include circadian repressors such as Per and Cry, which in turn inhibit CLOCKBMAL1-mediated transcription during the night [8]. Circadian circuit is also present in orphan nuclear receptors ROR (retinoic acid receptor-related orphan receptors) and REV-ERB, which alternately activate and repress Bmal1 gene expression, respectively [8]. This complex transcriptional-translational feedback loop along with daily activation of non-clock transcription factors and co-regulators give rise to a number of cyclic gene expression with distinct peak phases [5, 9]. Strikingly, circadian gene expression is highly tissue-specific, and the number of oscillatory genes differ across tissue-types or cell-types [10]. Crucially, cyclic expression can be reprogramed by metabolic perturbations [5, 9, 11– 15]. For instance, high-fat diets rewire circadian transcriptome and metabolome partly through the de novo cyclic activation of SREBP and PPARs [11, 15–17]. Similarly, ketogenic diets lead to activation of PPARα, whereas chronic alcohol intake dampens diurnal cycling of SREBP1 [12, 14]. Hence, gain or loss of daily activation of TFs, so-called metabolic clocks, contribute to remodeling of circadian rhythms [18–20]. Additionally, metabolites seem to impinge on clock functions [21]. Specifically, acetyl-CoA is used as a substrate for acetylation of histones and clock proteins, while nicotinamide dinucleotide (NAD+) serves as a co-enzyme of sirtuins to deacetylate proteins and thereby modifies circadian transcriptional regulation [22–26]. Likewise, S-adenosyl-L-methionine (SAM) is a substrate for methylation of histones, and histone methyltransferase MLL1 and MLL3 modulate circadian transcription [27, 28]. On the other hand, S-adenosyl-L-homocysteine (SAH) is converted from SAM by methyltransferase, and hydrolase of SAH interacts clock proteins to couple methionine metabolism to circadian gene expression and chromatin remodeling [29]. These notions underscore robust yet flexible and receptive nature of circadian transcriptional regulations in response to metabolic inputs. Nucleosome is a unit of chromatin, a highly ordered structure of DNA wrapped around histone octamers [30–32]. Nucleosome positioning dictates accessibility of transcriptional regulators [30– 32]. Specifically, open spans of DNA with nucleosome-depleted regions enable transcriptional complex to interact with chromatin and regulate gene expression [30–32]. Nucleosome-free genomic regions often contain histone marks such as H3K4me1 and H3K27ac [30–32]. Conversely, compaction of DNA with nucleosome-dense regions prevents transcriptional machinery from binding to chromatin [30–32]. TF binding events can also modulate nucleosome positioning. Importantly, nucleosome repositioning, chromatin remodeling, and TF occupancy are subject to developmental and environmental signals [32]. Besides, binding of

Chromatin Immunoprecipitation

343

Fig. 1 Workflow of chromatin immunoprecipitation. Tissues or cells are homogenized and cross-linked, which allows transcription factors (TF) to be fixed with DNAs. Such TF-DNA complex is sheared by sonication. Subsequently, TF-DNA complex is immunoprecipitated with specific antibody against TFs and beads, followed by reverse cross-linking to degrade TFs and precipitation of DNA

TFs to chromatin can be highly dynamic, with the kinetic of seconds [33]. Widely used chromatin immunoprecipitation (ChIP) enables us to investigate genomic binding of TFs and enrichment of histone marks which signify transcriptional state at the chromatin level, thereby providing insights and understanding as to how such rhythmic transcription is established [34, 35] (Fig. 1). Indeed, ChIP followed by deep sequencing (ChIP-seq) have identified genome-wide mapping of rhythmic occupancy and distinct binding sites of clock TFs. These epigenetic fluctuations can only be observed by sampling tissues at several time points over the circadian cycle from animals under strictly controlled lighting conditions, or by harvesting cells undergoing synchronization of the circadian clock using serum, glucocorticoid, and so on. This chapter lays out detailed methodology of how to carry out ChIP in light of circadian research.

2

Materials Prepare all solutions using distilled water. Prepare and store all reagents at room temperature unless otherwise specified.

2.1

Cross-linking

1. PBS (137 mmol/l NaCl, 2.68 mmol/l KCl, 10.1 mmol/l Na2HPO4, 1.76 mmol/l KH2PO4, pH 7.4) with inhibitors: protease inhibitor cocktail, 0.5 mM PMSF, 20 mM sodium fluoride, 10 mM nicotinamide, 330 nM trichostatin A. Add inhibitors on the day of experiments.

344

Kenichiro Kinouchi et al.

2. Electric homogenizer. 3. 25 mM Magnesium chloride (MgCl2). Store MgCl2 in
20  C. 4. 260 mM Disuccinimidyl glutarate (DSG) in dimethyl sulfoxide (DMSO). Prepare DSG on the day of experiments. DSG should not be either stored or reused once it is resolved in DMSO. 5. 37% Formaldehyde for cross-linking. 6. 1.25 M Glycine. 2.2

DNA Shearing

1. Cell lysis buffer: 5 mM HEPES [pH 8.0], 85 mM KCl, 0.5% NP-40. 2. SDS lysis buffer: 50 mM Tris–HCl [pH 8.0], 10 mM EDTA [pH 8.0], 1% SDS. 3. FACS buffer: PBS with 2% bovine serum, 0.05% NaN3. Store at 4  C. 4. Sonicator or Bioruptor. 5. Dilution buffer: 50 mM Tris–HCl [pH 8.0], 167 mM NaCl, 1.1% Triton X-100, 0.11% Sodium Deoxycholate. 6. RIPA buffer: 50 mM Tris–HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA [pH 8.0], 1% Triton X-100, 0.1% Sodium Deoxycholate, and 0.1% SDS.

2.3 Immunoprecipitation

1. Tube rocker. 2. 50% Slurry Protein G Sepharose beads in low-salt RIPA buffer: 50 mM Tris–HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA [pH 8.0], 1% Triton X-100, 0.1% Sodium Deoxycholate, and 0.1% SDS. The Protein G Sepharose beads should be preblocked with low-salt RIPA buffer, containing salmon sperm DNA for ChIP-qPCR or 5 mg/ml BSA for ChIP-seq at 4  C O/N and stored at 4  C. 3. Antibodies for immunoprecipitation. 4. Low-salt RIPA buffer: 50 mM Tris–HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA [pH 8.0], 1% Triton X-100, 0.1% Sodium Deoxycholate, and 0.1% SDS. 5. High-salt RIPA buffer: 50 mM Tris–HCl [pH 8.0], 500 mM NaCl, 1 mM EDTA [pH 8.0], 1% Triton X-100, 0.1% Sodium Deoxycholate, 0.1% SDS. 6. LiCl buffer: 10 mM Tris–HCl [pH 8.0], 250 mM LiCl, 1 mM EDTA [pH 8.0], 0.5% NP-40, 0.5% Sodium Deoxycholate. 7. TE buffer: 10 mM Tris–HCl [pH 8.0], 1 mM EDTA [pH 8.0].

2.4 Reverse Crosslink

1. Elution buffer: 10 mM Tris–HCl [pH 8.0], 300 mM NaCl, 5 mM EDTA [pH 8.0], 0.5% SDS. 2. 2 mg/ml RNase A.

Chromatin Immunoprecipitation

345

3. 10 mg/ml Proteinase K. 4. 20 mg/ml Glycogen. 5. Phenol:chloroform:isoamyl alcohol solution (100 ml Phenol, 96 ml Chloroform, 4 ml Isoamyl alcohol, 50 mg 8-hydroxyquinoline). Store at 4  C. 6. TE buffer (10 mM Tris–HCl [pH 8.0], 1 mM EDTA [pH 8.0]) with 200 mM NaCl. 7. 100% and 70% ethanol. 2.5 DNA Fragment Check

3

1. 3 M Sodium acetate (NaOAc) pH 5.

Methods Perform all procedures at room temperature unless otherwise specified. Use personal protective equipment and work under a fume hood particularly when formaldehyde, phenol, and chloroform are handled.

3.1 Cross-linking for the Liver Tissue

1. Take 150 mg of frozen liver tissue on dry ice and mince in 1 ml PBS with inhibitors on ice using scissors (see Note 1). 2. Homogenize tissue chunks with electric homogenizer to eliminate visible tissue pieces. 3. Add 3 ml of PBS with inhibitors, 260 mM DSG (for a final concentration of 2 mM), and 25 mM MgCl2 (for a final concentration of 1 mM) to 1 ml of liver homogenate (final volume is approximately 4 ml). 4. Place tubes on a rocker at room temperature for 30 min to cross-link (see Note 1). 5. Add 37% formaldehyde (for a final concentration of 1%) to the suspension. 6. Rock tubes for an additional 5–10 min at room temperature (see Note 1). 7. Add 1.25 M glycine for a final concentration of 0.125 M to stop the cross-link. 8. Place tubes on a rocker for 10 min at room temperature. 9. Centrifuge samples at low speed (1000  g) for 5 min at 4  C. 10. Discard supernatant ice-cold PBS.

and

resuspend

pellets

in

5

ml

11. Pellet cells by low-speed spin (1000  g) for 10 min at 4  C. 12. Discard supernatant, resuspend in 1 ml cold cell lysis buffer, and incubate for 10 min on ice. 13. Spin at 1000  g for 5 min at 4  C.

346

Kenichiro Kinouchi et al.

14. Resuspend pellet in 600 μl cold SDS lysis buffer, incubate for 10 min on ice. 15. Pipet up and down. 16. Incubate on ice for 20 min. Isolated chromatin can be frozen at this point in
80  C. 3.2 Cross-linking for Cells

1. Using 10 cm dish, wash dishes with PBS twice at room temperature (see Note 1). 2. Add 1% formaldehyde in PBS, incubate, and shake for 10 min at room temperature (see Note 1). 3. Add 1.25 M glycine for a final concentration of 0.125 M. 4. Incubate dishes for 10 min at room temperature. 5. Wash dishes with ice-cold PBS twice. 6. Harvest cells in 700 μl of FACS buffer. 7. Spin samples at low speed (1000  g) for 5 min at 4  C. 8. Resuspend in 500 μl cold cell lysis buffer, and incubate for 10 min on ice. 9. Spin at 1000  g for 5 min at 4  C. 10. Resuspend pellet in 600 μl cold SDS buffer, incubate for 10 min on ice. 11. Pipet up and down. 12. Incubate on ice for 20 min, samples can be frozen at this point in
80  C.

3.2.1 Sonication for the Liver Tissues

1. Sonication: Sonicater (10 s on, 50 s off, Amp 50), or Bioruptor (High, 30 s on, 30 s off) (see Note 2). Optimize the number of cycles based on DNA shearing to aim at DNA fragment size from 200 to 1000 bp (see Note 3). 2. Spin at max speed for 10 min at 4  C. 3. Collect supernatant, add 2.5 ml ChIP dilution buffer to make final volume 3.1 ml. 4. Save 100 μl for input, and 1 ml for all samples including IgG. Store in
80  C.

3.2.2 Sonication for Cells

1. Sonication: Sonicator (10 s on, 50 s off), or Bioruptor (30 s on, 30 s off) (see Note 2). Optimize the number of cycles based on DNA shearing to aim at DNA fragment size from 200 to 1000 bp (see Note 3). 2. Spin at max speed for 10 min at 4  C. 3. Collect supernatant, add RIPA buffer to make final volume 1 ml. 4. Save 100 μl for input, and 900 μl for all samples including IgG. Store in
80  C.

Chromatin Immunoprecipitation

3.3 Immunoprecipitation

347

1. Add 50 μl of preblocked, 50% Slurry Protein G Sepharose beads in low-salt RIPA buffer to pre-clear 1 ml of sheared lysates. 2. Rock tubes for 2 h at 4  C. 3. Spin at 12,000  g for 10 s. 4. Collect supernatant in fresh tubes. 5. Add antibody, incubate O/N, rocking, 4  C (2 μg of BMAL1, H3K27ac, H3K4me1, and IgG). 6. Next day, add 50 μl of preblocked, 50% Slurry Protein G Sepharose beads in low-salt RIPA buffer. 7. Incubate for 3 h, 4  C rocking. 8. Spin at 12,000  g for 10 s. Discard supernatant using a pipet or an aspirator. Wash with 1 ml of RIPA/150 mM NaCl once by rocking for 5 min at 4  C. 9. Wash: RIPA/500 mM NaCl once by rocking for 5 min at 4  C. 10. Wash: LiCl wash buffer once by rocking for 5 min at 4  C. 11. Wash: TE (pH 8.0) twice by rocking for 5 min at 4  C.

3.4 Reverse-Crosslinking

1. Thaw input samples to get ready for reverse cross-linking. 2. Add 200 μl ChIP direct elution buffer to beads and input samples. 3. Reverse cross-link by heating 65  C for O/N, IP and input samples. 4. Add 2 μl RNaseA, incubate for 30 min at 37  C. 5. Add 1 μl proteinase K, incubate for 1 h at 55  C. 6. Add 1–2 μl Glycogen. 7. Spin at 12,000  g for 10 s, and collect supernatant in fresh tubes. 8. Add 210 μl phenol:chloroform:isoamyl alcohol to IP and input samples, and vortex vigorously. 9. Spin down tubes at max speed for 3 min at room temperature. The centrifugation separates top (aqueous) and bottom (organic) phases. 10. Collect the top phase in fresh tubes. 11. Add 180 μl TE buffer to the bottom phase and vortex vigorously. 12. Spin at max speed for 3 min at room temperature. 13. Collect and combine the upper phase with previously obtained upper phase. 14. Vortex and spin at 12,000  g for 10 s.

348

Kenichiro Kinouchi et al.

15. Add 900 μl ethanol to the IP and input samples, and invert the tubes manually five times. 16. Incubate O/N at
20  C. 17. Spin at max speed for 30 min at 4  C. 18. Discard supernatant and wash pellet with 1 ml 70% ethanol. 19. Spin at max speed for 10 min at 4  C. 20. Discard supernatant and air-dry pellet for 15 min. 21. Resuspend DNA precipitates in TE buffer (100 μl for qPCR, 25 μl for sequencing). Store in
20  C (see Note 4). 3.5 DNA Fragment Check

In order to establish optimal ChIP conditions, fragment size checking of sheared DNA should be conducted before proceeding with immunoprecipitation (see Note 3). 1. Aliquot 50 μl of sheared lysates for DNA fragment check. 2. Add 1 μl RNase. Incubate for 1 h at 37  C. 3. Add 3.2 μl of 5 M NaCl (0.3 M final) and 2 μl proteinase K (stock ¼ 10 mg/ml) to reverse cross-link and incubate at 65  C for at least 5 h. 4. Add 150 μl TE buffer and 210 μl phenol:chloroform:isoamyl alcohol. 5. Vortex thoroughly and centrifuge at max speed for 5 min at 4  C. 6. Take aqueous phase and add 400 μl 95% ethanol, 1/10 volume (~55 μl) 3 M NaOAc pH 5, and 1 μl of 20 mg/ml glycogen. Vortex. 7. Centrifuge at max speed for 5 min at 4  C and discard supernatant. 8. Add 300 μl of 70% ethanol at room temperature to wash the pellet. Spin at max speed for 1 min at room temperature. 9. Air-dry pellet for 15 min. 10. Resuspend in 12 μl TE buffer and 3 μl 6 loading buffer. Run on 1% gel to check fragment size.

4

Notes 1. The amount of tissue or cells as well as duration of cross-linking need to be optimized based upon experimental conditions. 2. Sonication rises temperature of samples. Intermittent sonication is typically conducted to keep samples cold. Pay attention to heat during sonication and use ice if needed.

Chromatin Immunoprecipitation

349

MEF Sheared DNA

RIPA

mix

SDS

1kbp 500bp

Fig. 2 DNA shearing. Sheared chromatin on agarose electrophoresis. Mouse embryonic fibroblasts were cultured in 10 cm dish, fixed with 1% formaldehyde for 15 min, and resuspended in 200 μl of either RIPA buffer (first lane), 1:1 mixture of RIPA buffer and SDS lysis buffer (second lane), and SDS lysis buffer (third lane). The samples were sonicated (30 s ON, 30 s OFF, 10 cycles), followed by reverse cross-link and DNA purification

3. The size of DNA fragments should range from 200 to 1000 bp (Fig. 2). 4. Enrichment of immunoprecipitated DNAs is validated by qPCR with negative controls, including IgG immunoprecipitation, and primers designed to amplify outside putative enriched sites.

Acknowledgments The authors thank all members of the Sassone-Corsi laboratory and Itoh laboratory for discussions and support. This study was funded by Japan Society for the Promotion of Science (JSPS) KAKENHI (C) 19K09012 for K.K.; Takeda Science Foundation; The NOVARTIS Foundation (Japan) for the Promotion of Science; The Japan Diabetes Society; Yamaguchi Endocrine Research Foundation; Mochida Memorial Foundation for Medical and Pharmaceutical Research; The Sumitomo Foundation.

350

Kenichiro Kinouchi et al.

Competing Interests The authors declare no competing interests. References 1. Kinouchi K, Sassone-Corsi P (2020) Metabolic rivalry: circadian homeostasis and tumorigenesis. Nat Rev Cancer 20:645–661 2. Schibler U, Sassone-Corsi P (2002) A web of circadian pacemakers. Cell 111:919–922 3. Aras E, Ramadori G, Kinouchi K, Liu Y, Ioris RM, Brenachot X, Ljubicic S, Veyrat-DurebexC, Mannucci S, Galie M, Baldi P, SassoneCorsi P, Coppari R (2019) Light entrains diurnal changes in insulin sensitivity of skeletal muscle via ventromedial hypothalamic neurons. Cell Rep 27:2385–2398.e3 4. Welz PS, Zinna VM, Symeonidi A, Koronowski KB, Kinouchi K, Smith JG, Guillen IM, Castellanos A, Crainiciuc G, Prats N, Caballero JM, Hidalgo A, Sassone-Corsi P, Benitah SA (2019) BMAL1-driven tissue clocks respond independently to light to maintain homeostasis. Cell 177:1436–1447.e12 5. Kinouchi K, Magnan C, Ceglia N, Liu Y, Cervantes M, Pastore N, Huynh T, Ballabio A, Baldi P, Masri S, Sassone-Corsi P (2018) Fasting imparts a switch to alternative daily pathways in liver and muscle. Cell Rep 25: 3299–3314.e6 6. Masri S, Kinouchi K, Sassone-Corsi P (2015) Circadian clocks, epigenetics, and cancer. Curr Opin Oncol 27:50–56 7. Asher G, Sassone-Corsi P (2015) Time for food: the intimate interplay between nutrition, metabolism, and the circadian clock. Cell 161:84–92 8. Reppert SM, Weaver DR (2002) Coordination of circadian timing in mammals. Nature 418: 935–941 9. Pastore N, Vainshtein A, Herz NJ, Huynh T, Brunetti L, Klisch TJ, Mutarelli M, Annunziata P, Kinouchi K, Brunetti-Pierri N, Sassone-Corsi P, Ballabio A (2019) Nutrientsensitive transcription factors TFEB and TFE3 couple autophagy and metabolism to the peripheral clock. EMBO J 38:e101347 10. Zhang R, Lahens NF, Ballance HI, Hughes ME, Hogenesch JB (2014) A circadian gene expression atlas in mammals: implications for biology and medicine. Proc Natl Acad Sci U S A 111:16219–16224 11. Dyar KA, Lutter D, Artati A, Ceglia NJ, Liu Y, Armenta D, Jastroch M, Schneider S, de Mateo S, Cervantes M, Abbondante S,

Tognini P, Orozco-Solis R, Kinouchi K, Wang C, Swerdloff R, Nadeef S, Masri S, Magistretti P, Orlando V, Borrelli E, Uhlenhaut NH, Baldi P, Adamski J, Tschop MH, Eckel-Mahan K, Sassone-Corsi P (2018) Atlas of circadian metabolism reveals system-wide coordination and communication between clocks. Cell 174:1571–1585.e11 12. Gaucher J, Kinouchi K, Ceglia N, Montellier E, Peleg S, Greco CM, Schmidt A, Forne I, Masri S, Baldi P, Imhof A, Sassone-Corsi P (2019) Distinct metabolic adaptation of liver circadian pathways to acute and chronic patterns of alcohol intake. Proc Natl Acad Sci U S A 116:25250–25259 13. Masri S, Papagiannakopoulos T, Kinouchi K, Liu Y, Cervantes M, Baldi P, Jacks T, SassoneCorsi P (2016) Lung adenocarcinoma distally rewires hepatic circadian homeostasis. Cell 165:896–909 14. Tognini P, Murakami M, Liu Y, Eckel-Mahan KL, Newman JC, Verdin E, Baldi P, SassoneCorsi P (2017) Distinct circadian signatures in liver and gut clocks revealed by ketogenic diet. Cell Metab 26:523–538.e5 15. Tognini P, Samad M, Kinouchi K, Liu Y, Helbling JC, Moisan MP, Eckel-Mahan KL, Baldi P, Sassone-Corsi P (2020) Reshaping circadian metabolism in the suprachiasmatic nucleus and prefrontal cortex by nutritional challenge. Proc Natl Acad Sci U S A 117(47): 29904–29913 16. Eckel-Mahan KL, Patel VR, de Mateo S, Orozco-Solis R, Ceglia NJ, Sahar S, DilagPenilla SA, Dyar KA, Baldi P, Sassone-Corsi P (2013) Reprogramming of the circadian clock by nutritional challenge. Cell 155:1464–1478 17. Guan D, Xiong Y, Borck PC, Jang C, Doulias PT, Papazyan R, Fang B, Jiang C, Zhang Y, Briggs ER, Hu W, Steger D, Ischiropoulos H, Rabinowitz JD, Lazar MA (2018) Dietinduced circadian enhancer remodeling synchronizes opposing hepatic lipid metabolic processes. Cell 174:831–842.e12 18. Chaix A, Lin T, Le HD, Chang MW, Panda S (2019) Time-restricted feeding prevents obesity and metabolic syndrome in mice lacking a circadian clock. Cell Metab 29:303–319.e4 19. Quagliarini F, Mir AA, Balazs K, Wierer M, Dyar KA, Jouffe C, Makris K, Hawe J,

Chromatin Immunoprecipitation Heinig M, Filipp FV, Barish GD, Uhlenhaut NH (2019) Cistromic reprogramming of the diurnal glucocorticoid hormone response by high-fat diet. Mol Cell 76:531–545.e5 20. Vollmers C, Gill S, DiTacchio L, Pulivarthy SR, Le HD, Panda S (2009) Time of feeding and the intrinsic circadian clock drive rhythms in hepatic gene expression. Proc Natl Acad Sci U S A 106:21453–21458 21. Katada S, Imhof A, Sassone-Corsi P (2012) Connecting threads: epigenetics and metabolism. Cell 148:24–28 22. Masri S, Rigor P, Cervantes M, Ceglia N, Sebastian C, Xiao C, Roqueta-Rivera M, Deng C, Osborne TF, Mostoslavsky R, Baldi P, Sassone-Corsi P (2014) Partitioning circadian transcription by SIRT6 leads to segregated control of cellular metabolism. Cell 158:659–672 23. Nakahata Y, Kaluzova M, Grimaldi B, Sahar S, Hirayama J, Chen D, Guarente LP, SassoneCorsi P (2008) The NAD(+)-dependent deacetylase SIRT1 modulates CLOCK-mediated chromatin remodeling and circadian control. Cell 134:329–340 24. Nakahata Y, Sahar S, Astarita G, Kaluzova M, Sassone-Corsi P (2009) Circadian control of the NAD+ salvage pathway by CLOCKSIRT1. Science 324:654–657 25. Ramsey KM, Yoshino J, Brace CS, Abrassart D, Kobayashi Y, Marcheva B, Hong HK, Chong JL, Buhr ED, Lee C, Takahashi JS, Imai S, Bass J (2009) Circadian clock feedback cycle through NAMPT-mediated NAD+ biosynthesis. Science 324:651–654 26. Sato S, Solanas G, Peixoto FO, Bee L, Symeonidi A, Schmidt MS, Brenner C, Masri S, Benitah SA, Sassone-Corsi P (2017) Circadian reprogramming in the liver identifies metabolic pathways of aging. Cell 170:664– 677.e11 27. Katada S, Sassone-Corsi P (2010) The histone methyltransferase MLL1 permits the

351

oscillation of circadian gene expression. Nat Struct Mol Biol 17:1414–1421 28. Valekunja UK, Edgar RS, Oklejewicz M, van der Horst GT, O’Neill JS, Tamanini F, Turner DJ, Reddy AB (2013) Histone methyltransferase MLL3 contributes to genome-scale circadian transcription. Proc Natl Acad Sci U S A 110:1554–1559 29. Greco CM, Cervantes M, Fustin JM, Ito K, Ceglia N, Samad M, Shi J, Koronowski KB, Forne I, Ranjit S, Gaucher J, Kinouchi K, Kojima R, Gratton E, Li W, Baldi P, Imhof A, Okamura H, Sassone-Corsi P (2020) S-adenosyl-l-homocysteine hydrolase links methionine metabolism to the circadian clock and chromatin remodeling. Sci Adv 6:eabc5629 30. Calo E, Wysocka J (2013) Modification of enhancer chromatin: what, how, and why? Mol Cell 49:825–837 31. Shlyueva D, Stampfel G, Stark A (2014) Transcriptional enhancers: from properties to genome-wide predictions. Nat Rev Genet 15: 272–286 32. Spitz F, Furlong EE (2012) Transcription factors: from enhancer binding to developmental control. Nat Rev Genet 13:613–626 33. McNally JG, Muller WG, Walker D, Wolford R, Hager GL (2000) The glucocorticoid receptor: rapid exchange with regulatory sites in living cells. Science 287:1262–1265 34. Koike N, Yoo SH, Huang HC, Kumar V, Lee C, Kim TK, Takahashi JS (2012) Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science 338:349–354 35. Koronowski KB, Kinouchi K, Welz PS, Smith JG, Zinna VM, Shi J, Samad M, Chen S, Magnan CN, Kinchen JM, Li W, Baldi P, Benitah SA, Sassone-Corsi P (2019) Defining the independence of the liver circadian clock. Cell 177: 1448–1462.e14

Chapter 24 Fluorescent Reporters for Studying Circadian Rhythms in Drosophila melanogaster Kathyani Parasram, Daniela Bachetti, Vania Carmona-Alcocer, and Phillip Karpowicz Abstract Circadian rhythms are daily oscillations in physiology and gene expression that are governed by a molecular feedback loop known as the circadian clock. In Drosophila melanogaster, the core clock consists of transcription factors clock (Clk) and cycle (cyc) which form protein heterodimers that activate transcription of their repressors, period (per) and timeless (tim). Once produced, protein heterodimers of per/tim repress Clk/cyc activity. One cycle of activation and repression takes approximately (“circa”) 24-h (“diem”) and repeats even in the absence of external stimuli. The circadian clock is active in many cells throughout the body; however, tracking it dynamically represents a challenge. Traditional fluorescent reporters are slowly degraded and consequently cannot be used to assess dynamic temporal changes exhibited by the circadian clock. The use of rapidly degraded fluorescent protein reporters containing destabilized GFP (dGFP) that report transcriptional activity in vivo at a single-cell level with ~1-h temporal resolution can circumvent this problem. Here we describe the use of circadian clock reporter strains of Drosophila melanogaster, ClockPER and ClockTIM, to track clock transcriptional activity using the intestine as a tissue of interest. These methods may be extended to other tissues in the body. Key words Circadian rhythms, Circadian clock, Regeneration, Intestine, Drosophila melanogaster, Transcriptional reporters

1

Introduction

1.1 The Circadian Clock

Circadian rhythms are endogenous daily oscillations in biochemical processes and gene expression that allow an organism to coordinate behavior and physiology through the day and night [1–3]. In animals, these endogenous rhythms are manifested at the cellular level by a molecular mechanism called the circadian clock, which at its core comprises a transcription–translation feedback loop (TTFL) [4, 5]. The study of the mechanism that drives the circadian oscillations started with the characterization of drosophila mutants that shortened, lengthened, or eliminated circadian behavioral rhythms. All these mutations were mapped to the gene period (per) [6]. Over

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_24, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

353

354

Kathyani Parasram et al.

Fig. 1 The circadian clock in the Drosophila midgut. (a) A schematic showing the circadian clock in Drosophila melanogaster. (b) The Drosophila midgut is divided into multiple regions (R0–R5) from anterior to posterior. The foregut or cardia is the R0 region, followed by the midgut which consists of R1–R5 regions, followed by the hindgut. Each region can be tested for circadian clock activity separately (see Fig. 2b). Scale bar indicates 500 μm, DAPI marks nuclei. Midgut is from a 5-day-old female ClockPER fly. (c) The Drosophila midgut lineage consists of Intestinal Stem Cells that divide into Enteroblasts, who differentiate into either Enterocytes or Enteroendocrine Cells. Differentiation is directed by Notch signaling with high levels resulting in ECs and low levels in EEs. (d) The ClockPER and ClockTIM strains report Clk/cyc activity using a destabilized GFP (dGFP) molecule downstream of four tandem copies of a minimal promoter of period and timeless, respectively [8, 9]

the years the clock has been characterized [7], and at its core, the mechanism consists of positive and negative elements that activate and repress gene transcription on a daily basis (Fig. 1a). In Drosophila, Clock (Clk) and cycle (cyc) form a heterodimer that bind to E-box (50 -CAC GTG -30 ) regions in the per and timeless (tim) loci

Fig. 2 ClockPER and ClockTIM reporters of circadian clock activity. (a) Representative images taken at ZT0 (7 am, lights-on) from ClockPER (left) and ClockTIM (right) strains. Scale bar indicates 500 μm, DAPI marks nuclei. (b) Magnified images of ClockPER taken from each region of the midgut (shown in Fig. 1b) from the same timepoint, ZT0. Reporter activity is present in the epithelial cells present. Scale bar indicates 50 μm, DAPI marks nuclei. (c) Representative images for a ClockPER timeseries, with samples taken every 3 h from lights-on (ZT0) to lights-off (ZT12). Scale bar indicates 500 μm, midgut is outlined. (d) Table showing the fluorescence measurements exported from ZenBlue (Zeiss) software and used to calculate the GFP/DAPI ratio for one experimental replicate of n ¼ 7 midguts. (e) GFP/DAPI ratio for a timeseries using the ClockPER reporter shows a peak at ZT0 and a trough at ZT12/15

356

Kathyani Parasram et al.

to activate their transcription [8, 10, 11]. The per and tim proteins form heterodimers, which repress their own transcription by inhibiting their transactivators, Clk/cyc [12–14]. This feedback loop interlocks with secondary TTFL loops to confer additional precision to the clock, where Clk/cyc drive the sequential expression of the transcription activator PAR-domain protein 1 (Pdp1) and the repressor vrille (vri), two additional factors that modulate the expression of the Clk gene [15–18]. The circadian clock has the ability to persist in constant conditions, a characteristic known as free-running, with a timing of approximate 24 h [19]. This occurs in part through the rate of phosphorylation and degradation of the repressors, per/tim, which ensure cycles of accumulation and heterodimer formation [20–22]. However, with few exceptions, most animals are subjective to environment cycles such as photoperiod or temperature changes during the 24-h day. These environmental cycles adjust the period of the molecular clock to match the external cycle in a process known as entrainment [3]. Light, temperature, and food availability are called zeitgebers (from the German word “time-giver”) and can synchronize endogenous circadian clocks to match the environment [3, 23]. In Drosophila, a gene called cryptochrome (cry) produces a photoreceptive protein that induces tim degradation that frees Clk/cyc heterodimers to reset the next circadian cycle [24, 25] (Fig. 1a). The clock is present across different tissues in the animal body, including the intestine [26–28], where it coordinates 24-h daily changes in many functions, including motility, absorption, digestion, and permeability [29–33]. In other tissues of the body, the clock also plays a role in the timing of daily functions. In this paper, we will focus on the intestine as an example of a tissue where circadian clock activity is present, but the methods we outline can be applied to other tissues in the body depending on the research question. 1.2 Drosophila as a Model to Understand Intestinal Stem Cell Function

The intestinal biology of Drosophila is highly conserved with other animals and its biology regulated by similar pathways, which makes Drosophila an attractive genetic model to elucidate intestinal function [34, 35]. The Drosophila intestinal epithelium forms a tube lined with a luminal peritrophic matrix, to protect cells from damage due to ingested food and pathogens, and has basal visceral muscles, nerves, and oxygen transporting trachea [36–38]. Like that of other animals, the Drosophila intestinal epithelium is subdivided into functionally distinct regions from anterior to posterior, depending on their physiological function. At the anterior, the foregut and crop descend into the midgut regions R0–R5, and are then followed by the hindgut [39, 40] (Fig. 1c). The midgut is analogous to the mammalian stomach and small intestine and it is the main site of nutrient digestion and absorption [35, 41, 42]. This tissue is a pseudostratified epithelium with four main

Using Fluorescent Reporters for Drosophila Rhythms

357

cell types: undifferentiated Intestinal Stem Cells (ISCs), progenitors known as Enteroblasts (EBs), and differentiated cells that include Enterocytes (ECs) and Entero-Endocrine cells (EEs) [43, 44]. Similar to mammals, the midgut of Drosophila is renewed by the activity of ISCs [43–45], which divide into a daughter ISC that maintain the pool of stem cells and an EB [46, 47] that differentiates into ECs or EEs [48–50] (Fig. 1b). The ECs are large polyploid cells responsible for nutrient absorption and gut barrier functions, whereas EEs secrete peptides to regulate digestion and metabolism [34, 35]. ISC proliferation and EB differentiation is mediated by conserved pathways, including Jak/Stat, EGF, Wnt, Notch, Hippo, Hedgehog, and Bmp [51–65]. The intestinal epithelium has a highly renewal capacity, and recent studies show that circadian clock regulates the 24-h timing of tissue renewal [66–68]. 1.3 Reporters of Circadian Rhythms

2

The intestine is an example of a peripheral tissue that consists of cells containing endogenous circadian clocks [69–71]. Our understanding of the function of these peripheral clocks was first enabled by the development of bioluminescent reporters to track oscillations of the molecular clock at the tissue level in both Drosophila [72, 73] and mice [74–76]. The use of these Luciferase-based reporters in combination with imaging techniques allowed researchers to test differences in the temporal organization of clock function with a certain degree of spatial resolution [72, 77]. However, resolution at the single-cell level was limited and required specialized equipment. An alternative to track molecular oscillations with improved spatial resolution was made possible through the use of the fluorescent reporters using dGFP, which has a short half-life (1–2-h) [78], making possible to capture hourly changes in mammals [79, 80] and Drosophila [67, 81]. Here, we describe the use of in vivo Clock reporter strains that contain dGFP, ClockPER and ClockTIM [67], to study circadian rhythms in the Drosophila midgut peripheral clock.

Materials

2.1 Fly Strains and Food

1. A fluorescent circadian clock reporter strain of Drosophila melanogaster ClockPER or ClockTIM [67]. These strains contain the minimal promoters of per [8] and tim [9] to report circadian transactivation rhythms. The reporters were designed to use dGFP that is rapidly degraded to capture hourly changes to track Clk/cyc activity in the midgut [67] (see Note 1). Reporter strains are usually maintained at 25  C in vials containing media (see Subheading 3.2 below for details), under light/dark photoperiod.

358

Kathyani Parasram et al.

2. Fly food: 1.2% w/v inactive dry yeast (Genesee Scientific), 0.7% w/v soy flour (Caldic), 5% w/v cornmeal (Caldic), 0.4% w/v malt (Caldic), 0.4% w/v Drosophila agar type II (Genesee Scientific), 0.6% v/v glucose syrup (Caldic), 0.24% w/v tegosept (Apex Bioresearch Products), 0.027% w/v propionic acid (Genesee Scientific), 0.94% v/v 95% ethanol, 0.14% v/v 2 M hydrochloric acid, ~4 L of distilled water. If different suppliers are used, the grains may have different textures and the recipe (e.g., water content, cooking time) may need to be adjusted according to strain needs. (a) 1 M Propionic Acid Stock Solution: Mix 96 g of propionic acid powder in 500 mL of distilled water, then add 500 mL of 2 M hydrochloric acid and mix well (pH ¼ 4.6). (b) 1.7 M Tegosept Stock Solution: Mix 127 g powder in 500 mL 95% ethanol with heat at 80  C and stir at 120 rpm until dissolved. 3. Fly food vials and stoppers, the recipe provided in Subheading 3.2 will make approximately 400 vials. 2.2

Solutions

1. 10 PBS (Modified Gibco/Cold Spring Harbor): Dissolve 4.8 g (0.02 M) of potassium phosphate monobasic, 28.8 g (0.10 M) of sodium phosphate dibasic, and 180 g (1.54 M) of sodium chloride in 1.2 L of distilled H2O, then add distilled H2O to 2 L and autoclave. 2. 1 PBS: To make 1 PBS, dilute the 10 PBS stock solution using distilled water (1:10, 10 PBS and distilled water) and then adjust the pH to 7.4 using sodium hydroxide or hydrochloric acid. 3. 70% v/v (12 M) Ethanol: Mix 700 mL of ethanol (>95% purity) with 300 mL of distilled water. 4. 4% (1.25 M) Paraformaldehyde (PFA): Mix 125 μL of 32% PFA and 875 μL of 1 PBS. You will need ~1 mL of 4% PFA for each group of flies (approximately 7–15 midguts). 5. 0.2% PBS-T: Use 200 mL of 10 PBS, 4 mL (8.5 mM) of Triton X-100, and 1800 mL distilled water, mix well, and then adjust pH to 7.4 using sodium hydroxide or hydrochloric acid.

2.3

Dissection Tools

1. #5 Fine Forceps, two pairs. 2. Transfer pipettes 3.5 mL Graduated. 3. Transfer pipettes 1 mL Ultra Fine Tip. 4. Pyrex spot plate nine depressions, Glass. 5. P200 pipette, take one disposable pipette tip and cut off the edge to widen the mouth; this will be used to move a group of midguts without damaging them.

Using Fluorescent Reporters for Drosophila Rhythms

2.4 Immunostaining Reagents

359

1. 1% w/v BSA: 100 mL 0.2% PBS-T, 1 g of BSA (Albumin Fraction V) (see Note 2). Add 20 μL of sodium azide (make 0.1% w/v solution in distilled water), and adjust pH to 7.3. 2. 40 ,6-Diamidino-2-phenylindole (DAPI) (5 mg/mL). 3. Primary antibody in 1% BSA. Primary antibodies used are antiDelta (1:50, DSHB C594.9B-s) to mark intestinal stem cells, and anti-Prospero (1:50, DSHB MR1A) to mark enteroendocrine cells. Delta+ cells contain cytoplasmic puncta, and Prospero+ positive cells have strong nuclear staining. Enterocytes and enteroblasts are both Delta and Prospero negative. However, enterocytes can be identified by their large polyploid nuclei and enteroblasts can be identified as small cells negative for both markers. In this manner, all intestinal epithelial cells can be identified. 4. Secondary antibody with DAPI (1:5000) counterstain in 1% BSA. The ClockPER and ClockTIM reporters use GFP; therefore secondary antibodies such as Alexa goat anti-mouse 555 (2 mg/mL), 594 (2 mg/mL), or 647 (2 mg/mL) may be used at 1:2000 dilution. 5. Microscope slides, 25  75  1.0 mm (Fischer Scientific, 12-550-15) and microscope cover glass (Fischer Scientific, 12545E). The microscopy settings, Subheading 3.6 (for example, z-stack #6c), may need to be adjusted if different slides or cover glass with a different thickness are used. 6. Mounting reagent, such as ProLong Gold Antifade Reagent (Invitrogen).

3

Methods

3.1 Clock Reporter Design

Two circadian clock reporter Drosophila strains (ClockPER and ClockTIM) are used in the assays outlined below. However, these methods should be suitable for the analysis of any pathway reporter provided the fluorophore presence is short enough to vary over a period of 24 h. If the fluorophore’s stability is >24 h, it may be difficult to detect HTT93Q flies (HTT) vs. control flies (ctrl) carrying, but not expressing HTT93Q (UAS-HTT93Q). Flies were grown in standard conditions. (b) Total proteins visualized with Ponceau S solution

13. Bradford reagent. 14. BSA—2 mg/mL protein standard. 15. Spectrophotometer and cuvettes. 16. 4 non-reducing Laemmli sample buffer: 300 mM Tris–HCl pH 6.8, 66% glycerol, 2.4% SDS, 0.02% bromophenol blue. Add 750 μL of 2 M Tris–HCl pH 6.8, 3.3 mL of 100% glycerol, 600 μL of 20% SDS (20% SDS stock: 20 g of SDS in 100 mL final volume of H2O), and 1 mg of bromophenol blue. Bring to 5 mL with H2O. Store at room temperature (RT) (see Notes 2 and 3). 17. Thermo block. 18. General molecular biology plasticware.

376

Laura Delfino et al.

2.2 Agarose Gel for Protein Separation

19. Agarose gel casting tray and comb (see Note 4). 20. Agarose gel electrophoresis tank. 21. Agarose gel: 2% agarose (see Note 5), 375 mM Tris–HCl, pH 8.8, 0.1% SDS. Mix 2 g of agarose and 25 mL of 1.5 M Tris–HCl pH 8.8 in a heatproof glass bottle. Bring the volume to 100 mL with H2O and microwave to dissolve the agarose. To prevent boiling over, turn the power off as soon as the solution starts boiling and mix by gentle swirl. Repeat until the agar is fully dissolved. Cool the solution in a 50  C water bath for about 20 min, then add 500 μL of 20% SDS, mix by gentle swirl, and pour on the casting tray (see Notes 3 and 6). 22. Gel running buffer: 25 mM Tris–HCl, 192 mM glycine, 0.1% SDS (see Note 7).

2.3

Immunoblotting

23. PVDF membrane with 0.2 μm size pore for high binding immunoblotting, cut to size (see Note 8). 24. Western blot apparatus. 25. Blotting paper sheets, cut to size. 26. Glass rod. 27. Transfer buffer: 25 mM Tris–HCl, 192 mM glycine, 20% methanol (see Note 9). 28. Ponceau S solution. 29. TBST buffer: 20 mM Tris–HCl, 150 mM NaCl, 0.1% Tween20 (see Note 10). 30. Blocking buffer: 5% skimmed milk powder (analytical grade is not required) in TBST. 31. Primary antibody: α-HTT (MAB5374, Millipore), 1:2000 (diluted in blocking buffer) (see Note 11). 32. Secondary antibody: α-mouse HRP (PI-2000 Vector), 1: 10,000 (diluted in blocking buffer). 33. Chemiluminescent substrate for Western blotting.

2.4 Locomotor Activity Analysis

3

34. DAM2 (TriKinetics, Waltham, Mass., USA) Drosophila activity monitors with glass tubes (5 mm in diameter  80 mm in length, 1 mm in thickness).

Methods

3.1 Flies: Experimental Set Up

Flies are grown under a 12 h light, 12 h dark regime (LD 12:12) at 25  C on a standard fly medium. We used the pan-clock driver timGAL4 to express HTT93Q, a pathogenic variant of the first exon of human HTT containing 93 glutamines [12, 13] in all canonical

Visualising mHTT Aggregates from Fly Clock Neurons

377

Table 1 Period, power (with SD) and rhythmicity of experimental flies according to the peaks identified by Lomb-Scargle, CLEAN, autocorrelation, cosinor and Morlet wavelet obtained with BeFly3 Analysis

Genotype

N

n

% rhythmic Mean period, h (SD) Mean power (SD)

L-S

PdfGAL4>HTT93Q 30 14 46.7 PDF control 32 31 96.9 HTT93Q control 30 27 90.0

23.6 (3.1) 24.2 (0.5) 24.0 (0.6)

9.3 (3.6) 34.3 (12.1) 43.6 (12.5)

L-S (stringent)

PdfGAL4>HTT93Q 30 5 16.7 PDF control 32 32 100 HTT93Q control 30 30 100

22.2 (1.2) 24.2 (0.5) 24.0 (0.5)

12.9 (3.6) 36.2 (11.4) 43.4 (11.9)

CLEAN

PdfGAL4>HTT93Q 30 15 50 PDF control 32 32 100 HTT93Q control 30 29 96.7

22.4 (3.2) 24.2 (0.5) 24.1 (0.5)

2.9 (1.3) 6.2 (2.2) 10.2 (2.7)

Autocorrelation PdfGAL4>HTT93Q 30 5 16.7 PDF control 32 31 96.9 HTT93Q control 30 29 96.7

21.6 (1.8) 24.1 (0.7) 23.8 (1.3)

0.23 (0.04) 0.42 (0.12) 0.47 (0.14)

Cosinor

PdfGAL4>HTT93Q 30 15 50 PDF control 32 32 100 HTT93Q control 30 30 100

23.3 (2.8) 24.1 (0.5) 23.9 (0.50)

6.8 (2.9) 12.7 (4.4) 20.6 (5.5)

Morlet

PdfGAL4>HTT93Q 30 17 56.7 PDF control 32 31 96.9 HTT93Q control 30 30 100

21.0 (2.8) 23.7 (0.6) 24.0 (0.6)

– – –

N ¼ number of animal tested; n ¼ number of flies rhythmic (see ref. 16)

clock neurons of the brain as well as clock competent organs and tissues (notably the eyes), which allow easier detection of aggregates from head extracts. Our preliminary experiments have shown that in timGAL4>HTT93Q flies, aggregates start becoming visible within 3 days after eclosion. Conversely, the PdfGAL4 driver expresses in only 16 neurons of the brain [14], which is not sufficient in PdfGAL4>HTT93Q to visualize any aggregation with AGERA even though this genotype generates a reduction in freerunning rhythmicity levels and a shortening of period (Table 1 and Figs. 2 and 3). Other drivers/poly-Q length can lead to different rates of aggregation; hence the age of collection as well as the amount of protein per AGERA lane must be optimized according to the flies’ genetic background. 3.2 Flies: Locomotor Activity Analysis

1. For locomotor activity analysis, flies (1 day old) are anesthetized with CO2 and placed individually into glass tubes containing standard food. The tubes are then inserted into the DAM2 monitors and locomotor activity is recorded for 3–5 days under LD 12:12 and for at least 5 days under constant darkness (DD) conditions [15].

378

Laura Delfino et al.

Fig. 2 Interface of BeFLY3 in Matlab

2. We use our own in-house software, BeFly3, which is implemented as a Matlab app (tested with Matlab 2016a and above). BeFly3 implements a range of techniques to identify rhythmic patterns in activity records and generates estimates of statistical significance through a randomization procedure [16]. The BeFly3 installation package and example files can be downloaded from https://gitlab.com/EdGreen21/befly3. The BeFly3 graphical user interface is shown in Fig. 2. The core input file for BeFly3 is the TriKinetics “channel file” formatted output, as this allows the multiple genotypes of flies to be multiplexed within the 32 channels of a single TriKinetics activity monitor, thereby limiting the potential impact of per-monitor batch effects. Channel files from flies of the same genotype or experimental condition are placed in named subdirectories of a top-level directory representing the entire experiment. This top-level directory is the main input for the BeFly3 user interface. Other variables can be entered, such as truncating records between a “Start bin #” and “end bin #”—a feature we typically use to divide the LD and DD parts of a record—and automated

Visualising mHTT Aggregates from Fly Clock Neurons

379

Fig. 3 Examples of the “compact row” style graphical output generated by BeFly3 for (a) rhythmic and (b) arrhythmic fly of the Pdf>HTT93Q genotype. The different analyses show good concordance: panels left to right show a histogram of activity (with Morlet instantaneous period plotted in red and cosinor best-fit in orange), a double-plotted actogram (48 h per row), and periodograms for Lomb-Scargle, CLEAN, autocorrelation and cosinor rhythmicity analyses. Where appropriate, the conventionally calculated confidence limits for each analysis are plotted on each periodogram in blue ( p ¼ 0.05 dotted line, p ¼ 0.01 solid line). Confidence limits based on comparison of the real data to the same data randomized in time are plotted in red for LombScargle and CLEAN (n ¼ 200 randomizations in this example, 5% threshold dotted line, 1% threshold solid line). For Lomb-Scargle analysis the randomization based confidence limits are less stringent than the conventionally calculated limits—see ref. 16 for a detailed discussion and also Table 1

identification and trimming of activity in channels where the fly has died during the experiment. The analysis is started by clicking “Run”; depending on the number of randomizations performed the analysis may take around 30 s per fly in a typical 6 days DD experiment. BeFly3 generates several outputs; an “Analysis_results.xlsx” file containing tabulated results for the whole experiment, and a graphical summary of the results each channel analyzed. This graphical summary can be presented in two formats: a large format for highresolution analysis and validation of individual flies, or a compact row-based format suitable for publication, as shown in Fig. 3. The “Analysis_results.xlsx” file is divided into multiple tabsheets, including a “Settings” tab recording the variables used for the analysis, and “Summary” sheets for each experimental condition analyzed. “Summary” sheets include the mean, median, standard deviation, and count for flies under this experimental condition (i.e., a format suitable for further statistical analysis), and per-fly tables of rhythmicity peaks identified by the CLEAN, LombScargle, autocorrelation, cosinor, and Morlet wavelet analyses, respectively. Finally, the XY coordinates to redraw the results figures for any fly are included should the pre-formatted graphs be insufficient.

380

3.3

Laura Delfino et al.

Gel Preparation

The volumes provided below are for one 15  11 cm gel. 1. Assemble the casting tray as per manufacturer instructions. 2. Pour the gel into the tray up to a thickness of 0.8 cm (see Notes 12 and 13). Leave undisturbed until solid.

3.4 Protein Extraction and Gel Electrophoresis

1. Flies (2–5 days old) are anesthetized with CO2 and moved into a 15 mL tube to be flash frozen in liquid nitrogen. Sample can be stored at
80  C until processing. 2. Collect 100 flies in a 15 mL polypropylene tube and freeze by immersion in liquid nitrogen (see Note 14). 3. To separate the heads from the bodies, vortex the tube in short burst while keeping the animals frozen by dipping the tube in liquid nitrogen periodically. Open the content on a frozen metallic sieve that is resting on a metallic plate on top of a bed of dry ice (in a tray). Shake the sieve until all heads go through; the bodies remain trapped. The occasional body that has gone through can be removed with a pair of frozen tweezers. Use the frozen funnel to transfer the heads into a 1.5 mL microfuge tube. 4. Move the heads on ice and add 70 μL of chilled protein extraction buffer. 5. Homogenize the heads on ice with the electric pestle for about 1.5 min. 6. Centrifuge the samples at 12,000  g for 10 min at 4  C, then transfer the supernatant into a chilled 1.5 mL tube. 7. Determine the protein concentration using the Bradford assay according to standard procedures (see Note 15). 8. Dilute 150 μg of protein extracts 3:1 with 4 non-reducing Laemmli sample buffer. Heat for 5 min at 95  C (see Note 16). Centrifuge briefly and leave at RT. 9. Load the samples on the gel and run at 4.5 V/cm at RT until the blue dye reaches the end of the gel (see Notes 16–18).

3.5 Immunoblotting and Visualization

All steps are carried out at RT, unless otherwise specified. 1. Activate the PVDF membrane by submerging it in methanol for 5 min and then move it in blotting buffer. 2. Equilibrate gel, membrane, sponges, blotting paper, in blotting buffer for 5 min. 3. Assemble the sandwich in the following order: sponge, 3 sheets of blotting paper, gel, membrane (laid on the side opposite the wells), 3 sheets of blotting paper, sponge. Use a glass rod to roll out any bubble between layers.

Visualising mHTT Aggregates from Fly Clock Neurons

381

4. Proteins are transferred at 300 mA for 3 h at 4  C (see Note 19). 5. To visualize total proteins, submerge the membrane in Ponceau S solution for 5 min soon as it comes off the transfer. Rinse the membrane with H2O until the background is clear. 6. Block the membrane in blocking buffer for 1 h (see Note 20). 7. Incubate the membrane with the primary antibody diluted in blocking buffer for 1 h (see Note 20). 8. Wash the membrane in TBST 3 for 15 min. 9. Incubate the membrane with the HRP-conjugated, secondary antibody diluted in blocking buffer for 1 h. 10. Wash the membrane 3 for 15 min in TBST. 11. Visualize the signal with a chemiluminescent substrate for Western blotting, following the manufacturer’s instruction. Figure 1 shows a representative result.

4

Notes 1. To make a 25 solution of protease inhibitor cocktail, dissolve one tablet in 2 mL water. Store as 50 μL aliquots at
20  C. 2. To pipette viscous solution like glycerol or Tween-20 cut pipette tips with scissors. 3. Use a hotplate stirrer to help solubilization. The stock is stored at RT and does not require autoclaving. SDS powder is highly irritant and should be weighed in a controlled environment (we suggest using a fume cabinet with the fan turned OFF) wearing standard personal protective equipment (PPE, laboratory coat, examination gloves, eye protection). Wipe all surfaces with wet paper after use and dispose of it safely. 4. We used a casting tray of 15  11 cm. The teeth dimensions of the combs were 7 mm  1.5 mm. 5. Use molecular grade agarose with high gel strength and Low EEO (Electroendosmosis). 6. Exert caution when melting the agarose in a microwave as the solution can easily boil over. Use heat-resistant gloves in addition to standard PPE. 7. Prepare a 10 Tris-glycine (TG) stock: 144.4 g glycine, 30.3 g Tris; bring to 1 L with H2O and dissolve on a hotplate stirrer. Use in combination with the 20% SDS stock at the right dilutions to prepare the gel running buffer. 8. A nitrocellulose membrane may also be used; however, we recommend optimizing the protocol.

382

Laura Delfino et al.

9. To prepare 1 L of transfer buffer, use 100 mL of 10 TG stock and add 700 mL of H2O and 200 mL of methanol (see also Note 7). Methanol should be handled under a fume hood. Always follow local regulations for waste disposal. 10. Prepare 10 TBS solution: 24.23 g Tris–HCl and 87.66 g of NaCl in H2O to a final 1 L volume. Store at RT after autoclaving. To prepare TBST, dilute the required amount of 10 TBS and add Tween-20 (see also Note 2). 11. Once diluted, this antibody can be stored at
20  C and reused several times. 12. If bubbles form, quickly “drive” them toward the end of the gel with the help of a pipette tip. 13. Using ruler and pencil, mark a line on the sides of the casting tray that is 0.8 cm above the bottom. 14. About 30 fly heads will provide sufficient proteins for loading one lane. It is not advisable to attempt the experiment with any less. We recommend using 0.75 μL of extraction buffer per head. 15. Expected protein concentration is about 10–30 mg/mL; hence, the appropriate concentration for the BSA standards is within the range of 0.1–1 mg/mL. 16. It is recommended that the loading volume is 60% of the maximum well capacity. The optimum loading volume for our setup was 30 μL. 17. Under these conditions it takes approximately 4 h for the gel to run. 18. Gel can alternatively be run at 0.45 V/cm O/N. 19. Leveling the corners of the gel with a sharp blade can help evening it out and making it easier placing the gel on top of the blotting sheets. 20. Alternatively, the membrane can be incubated O/N at 4  C.

Acknowledgments LD, CPK, and ER were funded by BBSCR grant BB/P010121/1 and SC and FG by MRC grant MR/M013847/1. References 1. Mattis J, Sehgal A (2016) Circadian rhythms, sleep, and disorders of aging. Trends Endocrinol Metab 27:192–203 2. He Q, Binbin W, Price JL, Zhao Z (2017) Circadian rhythm neuropeptides in Drosophila:

signals for normal circadian function and circadian neurodegenerative disease. Int J Mol Sci 18:886–900 3. Kadener S, Villella A, Kula E, Palm K, Pyza E, Botas J, Hall JC, Rosbash M (2006)

Visualising mHTT Aggregates from Fly Clock Neurons Neurotoxic protein expression reveals connections between the circadian clock and mating behavior in Drosophila. Proc Natl Acad Sci U S A 103:13537–13542 4. Koh K, Evans JM, Hendricks JC, Sehgal A (2006) A Drosophila model for age-associated changes in sleep:wake cycles. Proc Natl Acad Sci U S A 103:13843–13847 5. Sheeba V, Fogle KJ, Kaneko M, Rashid S, Chou YT, Sharma VK, Holmes TC (2008) Large ventral lateral neurons modulate arousal and sleep in Drosophila. Curr Biol 18: 1537–1545 6. Means JC, Venkatesan A, Gerdes B, Fan JY, Bjes ES, Price JL (2015) Drosophila spaghetti and doubletime link the circadian clock and light to caspases, apoptosis and tauopathy. PLoS Genet 11:e1005171 7. Delfino L, Mason RP, Kyriacou CP, Giorgini F, Rosato E (2020) Rab8 promotes mutant HTT aggregation, reduces neurodegeneration, and ameliorates behavioural alterations in a Drosophila model of Huntington’s disease. J Huntingtons Dis 9:253–263 8. Scherzinger E et al (1997) Huntingtinencoded polyglutamine expansions form amyloid like protein aggregates in vitro and in vivo. Cell 90:549–558 9. Kushnirov VV, Alexandrov IM, Mitkevich OV, Shkundina IS, Ter-Avanesyan MD (2006) Purification and analysis of prion and amyloid aggregates. Methods 39:50–55

383

10. Weiss A et al (2008) Sensitive biochemical aggregate detection reveals aggregation onset before symptom development in cellular and murine models of Huntington’s disease. J Neurochem 104:846–858 11. Greaser ML, Warren CM (2019) Electrophoretic separation of very large molecular weight proteins in SDS agarose. Methods Mol Biol 1855:203–210 12. Campesan S, Green EW, Breda C, Sathyasaikumar KV, Muchowski PJ, Schwarcz R, Kyriacou CP, Giorgini F (2011) The kynurenine pathway modulates neurodegeneration in a Drosophila model of Huntington’s disease. Curr Biol 21:961–966 13. Steffan JS et al (2001) Histone deacetylase inhibitors arrest polyglutamine-dependent neurodegeneration in Drosophila. Nature 413:739–743 ¨ zkaya O ¨ , Hemsley M, 14. Dissel S, Hansen CN, O Kyriacou CP, Rosato E (2014) The logic of circadian organization in Drosophila. Curr Biol 24:2257–2266 15. Rosato E, Kyriacou CP (2006) Analysis of locomotor activity rhythms in Drosophila. Nat Protoc 1:559–568 16. Kyriacou CP, Dowse HB, Zhang L, Green EW (2020) A computational error and restricted use of time-series analyses underlie the failure to replicate period-dependent song rhythms in Drosophila. J Biol Rhythm 35:235–245

Chapter 26 Methods for Delivery of dsRNAi Against Canonical Clock Genes and Immunocytodetection of Clock Proteins in Crustacea David C. Wilcockson, Lin Zhang, and Charalambos P. Kyriacou Abstract The use of nonclassical model organisms for biological rhythm research has become popular in the last two decades. Here we describe techniques for delivery of dsRNAi molecules to knock down clock gene transcripts in a small intertidal crustacean, Eurydice pulchra, as well as our method for immunodetection of clock proteins in the brain. These methods can be generalized for gene knockdown in any small crustacean or arthropod in which mutagenesis by other methods is neither practical nor possible. Key words Crustacea, Clock genes, dsRNAi, Microinjection, Immunocytochemistry

1

Introduction In the last two decades, research in biological rhythms has moved from focusing on a few well-established model animals, plants, and microbes that are genetically and molecularly tractable, the fly, mouse, Cyanobacteria, Neurospora, Arabidopsis, and zebrafish, to include more exotic and less easily genetically manipulated organisms. An impetus for working on “non-models” is that not only might they be more interesting in an ecological context, but also that the advent of “omics” can, in part, bypass the type of classical genetics that have facilitated molecular manipulations in flies and mice. Furthermore, the development of dsRNAi and siRNA methods permit the experimenter to manipulate levels of RNA of their favorite gene assuming that they can deliver the interfering molecule at the right place and the right time. We have developed such a non-model system to study circatidal rhythmicity, the 12.4 h cycles associated with the ebb and flow of the tides [1]. The organism in question is the specked sea louse, Eurydice pulchra, an isopod that unlike many animals that can be

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_26, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

385

386

David C. Wilcockson et al.

collected on the shoreline, shows extremely robust circatidal cycles of locomotion [1]. We cannot rear these animals in the laboratory, so we catch them on spring tides at a beach near Bangor in Wales, where this species is plentiful during the summer months from May to September. We bring them back to the laboratory where we can maintain them for up to 2 months and study their genome/transcriptome [2], canonical clock genes [3, 4] as well as manipulating them genetically and pharmacologically [4]. A major focus of our work is to knockdown the canonical circadian clock genes by dsRNAi and observe whether these have any effects on 24 h circadian and 12.4 h circatidal rhythms [4]. Our methods are generic and can be used in small crustaceans and other invertebrates where it is difficult or impractical to use CRISPR/Cas9 gene editing.

2

Materials All solutions are made with ultrapure ddH2O treated with 0.1% diethyl pyrocarbonate (DEPC) for 4 h and autoclaved prior to use. All other reagents should be molecular grade. The following method was developed to ensure rapid and reliable delivery of dsRNA to Eurydice but will equally work with other small arthropods and other invertebrates [4].

2.1 Design, Synthesis, Storage, and Delivery of dsRNA

1. Ambion MEGAscript RNAi kit (Thermofisher). Use as directed in the manufacturer’s instructions. 2. Standard polymerase chain reaction reagents and thermocycler for amplification of target gene(s) and agarose gel electrophoresis rigs. 3. Microcon YM-30 centrifugation filter (Millipore). 4. Injection medium (vehicle)-crustacean physiological saline: Physiological saline for crustaceans: 433 mM NaCl, 12 mM KCl, 12 mM CaCl2
2H2O, 20 mM MgCl2
6H2O, 10 mM HEPES, adjusted to pH 7.60 with 2 N NaOH [5]. 5. 152 mm standard borosilicate glass capillaries without filament, pulled to a fine point in an electrode puller (such as Suter P-97). 6. Phenol red 0.05%. 7. Scotch tape. 8. 100% Ethanol. 9. RNaseZap solution/wipes. 10. Agarose gel electrophoresis. 11. Nanoliter gas operated injector such as World Precision Instruments PV830 PicoPump.

dsRNAi and Immunodetection in Crustacea

387

12. Perspex or Plexiglas block- (150 mm  150 mm  10 mm), or similar. 13. NanoDrop Spectrophotometer. 14. Microcentrifuge. 15. Filter tips and tubes (Nuclease free). 16. Fine iris scissors. 17. Micromanipulator and stereomicroscope. 18. Fluorescein isothiocyante (FITC) in DMSO (1 mg/mL). 2.2 Immunocytochemistry

1. Bouin Hollande fixative (25 g cupric acetate, 40 g picric acid, 100 mL 40% formaldehyde, 15 mL glacial acetic acid, to 1 L in distilled water) Filtered. 2. Ethanol series (50%, 70%, 80%, 95%, and 100% in distilled water). 3. Methyl benzoate. 4. Toluene. 5. Xylene. 6. Histological paraffin wax (Leica Paraplast wax embedding medium) and warm oven (to 65  C). 7. Microtome for wax sectioning. 8. Phosphate-buffered saline (PBS) made from tablets (Thermo Fisher Scientific) pH 7.5. 9. Small (10 mL) weighing boats or histology embedding molds (depending on size of tissues) for embedding tissues in paraffin wax prior to sectioning. 10. Standard glass laboratory desiccator with desiccant. 11. Wax cutting microtome. 12. 10% Normal goat serum containing 0.001% sodium azide in PBS. 13. PBST (PBS containing 0.01% Tween-20). 14. Disposable Pasteur pipettes. 15. Positively charged glass slides (Thermo Scientific™ SuperFrost Plus™). 16. Primary antisera: anti-Eurydice pulchra PER (SG1993 or SG1994) (raised in rabbits) [4]; Drosophila anti-CLK (GP204) and anti-CYC (GP122) antibodies (raised in guineapigs [6, 7]). 17. Secondary antibodies (Goat anti-guineapig IgG 488 fluor and goat anti-rabbit IgG 568 fluor, Molecular probes, UK). 18. Humid chamber (we use a plastic sandwich box containing saturated absorbent tissues, or a bespoke slide humid chamber made in-house).

388

David C. Wilcockson et al.

19. Antifade slide mounting medium (Vectorlabs, Vectashield). 20. Glass coverslips (22  50 mm). 21. Clear nail lacquer.

3

Methods

3.1 Design, Synthesis, and Storage of dsRNA

1. Double-stranded RNAs for the Epperiod circadian gene, for example, was 758 bp and targeted toward the 50 end of the mRNA (nucleotides 170–928). For knockdown of Eurydice clock genes clk, bmal1, and cry2, 30 regions of the coding sequences were used, guided by the e-RNAi web service to design the dsRNAi (http://www.e-rnai.org [8]). We have used the yellow fluorescent protein (YFP) gene from pEYFP-N1 (Clonetech, UK) as a dsRNAi control. 2. Prepare 4  50 μL PCR reactions on ice using Bioline Reddymix 2 polymerase master-mix containing 1 μL (10 μM) of each primer. Two of these reactions should contain a forward primer with a T7 promoter sequence flanking the gene-specific sequence and a non-T7 flanked reverse primer. The remaining two reactions should be primed with a non-T7 flanked forward primer and T7-flanked reverse primer (This will generate sense and antisense strands of cRNA). 1 μL cDNA generated from brain total RNA (we use 1 μg total RNA in 20 μL cDNA reaction using Superscript III reverse transcriptase (Thermo, UK)) is used in each PCR reaction. Alternatively, any suitable plasmid containing T7 promoter sequences DNA and the target insert can be used as the PCR template according to manufacturer’s guidelines. 3. Run the PCR reactions for 40 cycles with the first five cycles at the annealing temperature for the gene-specific primers without T7 flanking sequences and the remaining 35 cycles at the annealing temperature of the T7 flanked primers, with an extension time of 60 s. 4. Pool the complete PCR reactions, keeping the forward and reverse T7 primed reactions separate. 5. Clean and concentrate the reactions by centrifuging the PCRs through a Microcon YM-30 filter at 14,000  g for 2 min. Add 20 μL DEPC water to the filter and reverse it in a clean 1.5 mL microfuge tune and spin for a further minute at 14,000  g. 6. Run 2.5 μL of each product on a 1% agarose gel to check for specificity and successful amplification. 7. Quantify the PCR product by NanoDrop spectrophotometry. 8. Use the clean T7-flanked PCR products as template in an Ambion MegaScript RNAi kit in vitro transcription reaction

dsRNAi and Immunodetection in Crustacea

389

according to the manufacturer’s guidelines and with 1 μg template PCR reaction from step 7. Each reaction is incubated at 37  C for 2–4 h. 9. Following transcription of each strand of cRNA, heat to 75  C for 10 min, combine sense and antisense strands and cool over 60 min to RT to anneal (we used a heated block and after heating switched it off and allowed to cool naturally to RT). After annealing, the double-stranded RNA can be checked for successful transcription and formation on a standard 1%, non-denaturing agarose gel. 10. dsRNA should be stored at 80  C until use. 11. When preparing for injection, dilute to working concentration in saline supplemented with phenol red at a final concentration of 0.05% (phenol red is optional to help visualize reagent when filling capillaries and injecting). 3.2 Preparation and Loading of Capillaries

1. Glass capillaries (1 mm diameter without filament) are pulled to a fine point on a standard capillary/electrode puller. Beveling may benefit some applications depending on the animal/ target being injected but we found that it was not necessary for delivery of dsRNA to Eurydice. After pulling capillaries (settings for different pullers will vary but we aimed to draw to a tip less than 10 μm or so), use sharpened watch-makers forceps to break the pulled end and leaving a point at about 30–50 μm (see Note 1). 2. Capillaries are best front-loaded. Pipette 2.0 μL dsRNA (or saline vehicle) onto a piece of laboratory film and maneuver the pulled end of the capillary, held in a micromanipulator, so the tip is immersed in the liquid. Filling should occur by capillary action (see Note 2) and calibrated by marking off 100 nL intervals with a fine laboratory marker (see Note 6).

3.3 Injection Delivery of dsRNA

1. Chill-anesthetize animals in seawater held on ice for 15 min or until locomotor activity has ceased. Chilling slows animals down for handling and prevents movement during injection (see Note 3). 2. We discovered that single injections at any one time were the most effective and minimized mortality. Remove an individual animal from the chilled water using a soft paintbrush and dry carefully but briefly on absorbent paper to remove excess water. 3. Carefully position the animal on the Plexiglas plate and secure to the plate using Scotch tape leaving the last thoracic segment visible and accessible (see Note 4). If the animal is too mobile, re-chill for 15 min. 4. Position the Plexiglass plate on the microscope stage and advance the capillary from the posterior direction close to the last thoracic tergite (cuticular plate) in the latero-dorsal region

390

David C. Wilcockson et al.

of the animal. At this stage the tergites should be slightly lifted to allow the capillary to slip forward easily between the final thoracic and first abdominal segments. Move the capillary carefully forward until some resistance from the arthrodial membrane is registered by a slight deflection of the capillary (see Note 1). Continue gradual pressure on the arthrodial membrane by moving the capillary forward until it punctures and the tip of the capillary enters the hemocoel (see Note 5). The tip should penetrate into the hemocoel by about 200 μm and remain shallow to the dorsal tergites so as not to damage any internal structures. 5. Deliver 200 nL (equivalent to 200 ng) of dsRNA or equivalent for control, using compressed nitrogen regulated via a PV830 PicoPump set at 40 psi and delivering pulses at 50 ms (see Note 6) over 60 s (see Note (video) 7). 6. Following injection, slowly and carefully withdraw the capillary and release the animal from restraint to avoid positive pressure created by injection from forcing reagent back through the puncture wound and leading to excessive bleeding. Hold the animal on tissue paper saturated with seawater for 2 min before returning to water for full recovery. 7. Up to five animals may be injected with each capillary before replacing to avoid blockage or blunting (see Note 8). 8. We periodically check our injection protocol and effective delivery by injecting a fluorescent marker, such as fluoresceine isothiocyante (FITC) in DMSO (1 mg/mL) and visualizing the animals under UV light, such as a standard UV transilluminator (Fig. 1).

Fig. 1 Eurydice injected with approximately 200 nL FITC dye to demonstrate uptake and circulation of delivered reagents. Images were taken within 1 min of injection over a standard UV transilluminator. (a) Lateral view, (b) FITC injected and saline injected animals for comparison

dsRNAi and Immunodetection in Crustacea

3.4 Immunohistochemistry 3.4.1 Tissue Preparation and Sectioning

391

1. Remove heads using fine iris scissors by cutting transverse behind the second thoracic segment and immediately place in Bouin-Hollande fixative in 1.5 or 5 mL Eppendorf tubes and leave overnight (12–16 h) at 4  C. 2. Wash heads repeatedly in 50% ethanol on a rotating wheel or shaking table until all leaching of picric acid has stopped. 3. Dehydrate for 5 min each step through an ethanol series, including 50%, 70%, 80%, 90%, 95%, and absolute ethanol. 4. Clear heads in methyl benzoate until heads have sunk from the surface (usually about 30–60 min) and then replace with toluene (see Note 9). Soak in toluene for 30 min to replace methyl benzoate prior to infiltration and embedding in wax. 5. Infiltrate heads in molten wax for at least 1 h for each of three changes of wax. Then embed heads in wax by placing them cut-side down in small (10 mL) weighing boats or histology embedding molds. Four heads can be embedded in one block of wax in 10 mL boats. Solidify wax in cold running water for 15 min and store in desiccator at RT. 6. Cut ribbons of 6 μm sections on a standard wax microtome. Float serial sections (as ribbons—it is possible to fit an entire Eurydice head on one slide) on water (~500 μL per head) on positively charged glass slides, heated to about 65  C until they have stretched to remove any compression lines from cutting. At this stage, remove excess water by carefully aspirating with a disposable Pasteur pipette so the sections settle in place and adhere to the slides. Dry overnight on a warm plate (45  C).

3.4.2 Immunodetection

1. Dewax slides in xylene (see Note 9) for 5 min and repeat with fresh xylene before re-hydrating in 50:50 xylnene:ethanol followed by 100% (x2), 95%, 90%, 70%, and 50% ethanol and PBS. 2. Wash 2 for 5 min in PBS. 3. Incubate slides in a humid chamber in 10% normal goat serum containing 0.001% sodium azide (NaN3; take care, this preservative is highly toxic) for 60 min at RT (see Note 10). 4. Do not wash slides at this stage but proceed directly to primary antiserum incubation. We used anti-PER SG1994 and SG1993 (raised in rabbits) at 1:500 and same for Drosophila anti-CLK GP204 or anti-CYC GP122 antibodies (raised in guineapigs) diluted in PBST. We used Anti-PER antisera raised in rabbit at the same time as anti-CLK or anti-CYC with no issues of cross reactivity. Incubate at 4  C overnight in a humid chamber (see Note 11). Take care to ensure slides are level (a small spirit level helps here). 5. Wash slides 3 for 5 min in PBS.

392

David C. Wilcockson et al.

6. Apply second antisera Goat anti guineapig-488 and Goat antirabbit-564 at 1:500 in PBST and incubate for 60 min at RT in a humid chamber. 7. Wash five times for 5 min each time in PBST. 8. Apply two drops antifade slide mounting medium and apply coverslip. Seal the edges of the coverslip with clear nail lacquer. 9. Visualize on a fluorescence microscope or laser-confocal microscope within two weeks. Store slides in the dark at 4  C.

4

Notes 1. We found that for Eurydice a capillary tip too elongated and fine merely deflected the arthrodial membrane of the animal, between the cuticular plates and ran the risk also of breaking and blocking. Conversely a large bore can result in sudden and excessive delivery, potentially damaging internal organs and tissues. Some preliminary experimentation is necessary to define the correct tip thickness and taper. 2. Capillaries can be filled in several ways. We found that capillary action works well for volumes ~1 μL. Solutions were aspirated onto laboratory film and the capillary tip manipulated into the resulting droplet at an angle steep enough to prevent the droplet moving between the film and capillary, alongside the capillary. For increased loading speed we also used negative pressure, either generated by the Picopump vacuum or a 50 mL syringe with a Leuer fitting and fine silicone tubing continuous with the top of the capillary. Following filling of required volume (1–2 μL), the capillary can be graduated by measuring the distance between the tip and meniscus and marking off 100 nL intervals. 3. A cooled alloy block or similar can be used to keep animals immobilized. This is particularly useful when refining the process and injections might take longer. 4. We have trialed a number of restraining devices for small invertebrates and found Scotch tape to be effective, cheap, and easy to use. The amount of pressure applied to the individual animals can be easily regulated to avoid damaging delicate appendages while immobilizing. An additional benefit of this method is that the tergites are lifted slightly, enabling easy access to the arthrodial membrane. Care should be taken to avoid generating excessive pressure in the hemocoel that will affect delivery of the vehicle and reagent. 5. The tip of the capillary will be obscured by the tergites when advanced to meet with the arthrodial membrane. Puncture of the membrane can be observed in two ways. First by a sudden

dsRNAi and Immunodetection in Crustacea

393

advance of the capillary and loss of flexion in the tip and, second, in some cases by slight ingress of blood and liberation of hemocytes into the capillary tip. 6. The number and duration of pulses required will vary according to the bore of the drawn tip, so it is important to properly prime and test each capillary to avoid over-delivery and excessive bleeding-out of the reagent. The experimenter should observe carefully the advance of the meniscus to ensure delivery of the correct volume. We recommend injecting a volume of 200 nL over 60 s for Eurydice. 7. Video clip of injection—NB the time interval between delivery pulses has been reduced for illustration purposes. Video 1 Demonstration of a pulled capillary loaded with saline and delivering approximately 20 nL per pulse of nitrogen. The clicking sound is the Picopump nitrogen delivery. https://drive.google.com/file/d/1Lgl_7sJjg7a9oj0 SGMBYYKjyXBDE-dbC/view Video 2 Restrained Eurydice receiving dsRNA. Note, the video is for demonstrations purposes and the interval between pulses of gas delivery are reduced. Each pulse equates to about 20 nL volume of vehicle or dsRNA. h t t p s : // d r i v e . g o o g l e . c o m / fi l e / d / 1 2 SBiDzIm5dCg9FCTScqZLCgWJaUisXRK/view 8. Should the capillary become blocked, attempt to clear by pulsing outside of the animal. If this fails, a quick emersion in DEPC treated water may help. Avoid drawing water into the capillary and diluting the injection reagent. If all attempts fail, replace the capillary. Blockages can be avoided by (a) optimally sized bore of capillary and (b) limiting the interval between injections and allowing clotting of blood. 9. Care: Observe local safety guidelines—methyl benzoate, toluene, and xylene are harmful by inhalation and skin contact. Use appropriate PPE and perform steps in a fume hood. 10. It is critically important to ensure that drying does not occur at any stage after rehydration of sections and blocking in normal goat serum. If drying occurs unacceptable nonspecific labeling will result. 11. It is important to ensure slides are level to avoid drying or loss of reagent. If they are being held in a fridge, use a small spirit level to check that the humid chamber is level in all directions.

Acknowledgments CPK and DCW acknowledge BBSRC grants BB/E000835/1, BB/K009702/1, and BB/R01776X/1.

394

David C. Wilcockson et al.

References 1. Wilcockson DC, Zhang L (2008) Circatidal clocks. Curr Biol 18:R753–R755 2. O’Neill JS, Lee KD, Zhang L, Feeney K, Webster SG, Blades MJ, Kyriacou CP, Hastings MH, Wilcockson DC (2015) Metabolic molecular markers of the tidal clock in the marine crustacean Eurydice pulchra. Curr Biol 8:R326–R327 3. Wilcockson DC, Zhang L, Hastings MH, Kyriacou CP, Webster SG (2011) A novel form of pigment-dispersing hormone in the intertidal marine isopod, Eurydice pulchra (Leach). J Comp Neurol 519:562–575 4. Zhang L, Hastings MH, Green EW, Tauber E, Sladek M, Webster SG, Kyriacou CP, Wilcockson DC (2013) Dissociation of circadian and circatidal timekeeping in the marine crustacean Eurydice pulchra. Curr Biol 19:1863–1873 5. Saver MA, Wilkens JL, Syed NI (1999) In situ and in vitro identification and characterization of

cardiac ganglion neurons in the crab, Carcinus maenas. J Neurophysiol 81:2964–2976 6. Kim EY, Ko HW, Yu W, Hardin PE, Edery I (2007) A DOUBLETIME kinase binding domain on the Drosophila PERIOD protein is essential for its hyperphosphorylation, transcriptional repression, and circadian clock function. Mol Cell Biol 13:5014–5028 7. Bae K, Lee C, Hardin PE, Edery I (2000) dCLOCK is present in limiting amounts and likely mediates daily interactions between the dCLOCK-CYC transcription factor and the PER-TIM complex. J Neurosci 5:1746–1753 8. Horn T, Boutros M (2010) E-RNAi: a web application for the multi-species design of RNAi reagents—2010 update. Nucleic Acids Res 38(Web Server issue):W332–W339

Chapter 27 In Vivo Bioluminescence Analyses of Circadian Rhythms in Arabidopsis thaliana Using a Microplate Luminometer Masaaki Okada and Paloma Mas Abstract Our understanding of the circadian clock function in plants has been markedly assisted by studies with the model species Arabidopsis thaliana. Molecular and genetics approaches have delivered a comprehensive view of the transcriptional regulatory networks underlying the Arabidopsis circadian system. The use of the luciferase as a reporter allowed the precise in vivo determination of circadian periods, phases, and amplitudes of clock promoter activities with unprecedented temporal resolution. An increasing repertoire of finetuned luciferases together with additional applications such as translational fusions or bioluminescence molecular complementation assays have considerably expanded our view of circadian protein expression and activity, far beyond transcriptional regulation. Further applications have focused on the in vivo simultaneous examination of rhythms in different parts of the plant. The use of intact versus excised plant organs has also provided a glimpse on both the organ-specific and autonomy of the clocks and the importance of long distance communication for circadian function. This chapter provides a basic protocol for in vivo highthroughput monitoring of circadian rhythms in Arabidopsis seedlings using bioluminescent reporters and a microplate luminometer. Key words Circadian rhythms, Luciferase, Promoter activity, Gene expression, Arabidopsis thaliana, Bioluminescence, Reporter assays, Microplate luminometer

1

Introduction Circadian rhythms are orchestrated by a timing mechanism or circadian clock that is present in nearly all organisms examined to date. The circadian function ensures that the organism’s biology is in sync with the external environmental conditions [1]. Due to their sessile nature, the perception and adjustment to these environmental conditions is particularly important for plants [2]. Over the years, the components and mechanisms of the plant circadian clock have been well studied in the model system Arabidopsis thaliana [3]. Research has been lately considerably expanded to crops, opening interesting opportunities for improving plant productivity and resilience to environmental insults [4].

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0_27, © The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

395

396

Masaaki Okada and Paloma Mas

As in some other organisms, circadian rhythms in Arabidopsis rely on the rhythmic expression of clock genes, which form transcription-translation reciprocal regulations at the core of the central oscillator [3]. A number of core clock components, whose expression oscillates in a circadian manner and with specific phases during the day-night cycle, have been characterized [5–7]. Analyses of circadian rhythms of core gene expression have uncovered fundamental properties of the Arabidopsis circadian system including its robustness and flexibility under different environmental conditions [8–10]. One crucial step for our understanding of the circadian system in plants was the fine time-resolved elucidation of the circadian waveforms of the main oscillating genes. Gene expression analyses by time courses of RNA abundance require sampling of plants and the subsequent RNA extraction at the different time points. Although this procedure is amenable and produces reliable results, it is destructive and the time-series data are necessarily obtained from different samples. In vivo long-term analyses using reporter genes provide an advantage for circadian studies. The approach is based on the use of a reporter gene fused to a promoter, gene, or protein of interest that is exogenously introduced into cells, and literally “reports” its expression through the signal readout without the need of sample extraction. In plant cells, chlorophyll and carotenoids in plastids, and lignin and other phenolic compounds in cell walls cause autofluorescence, which often interferes and complicates the detection of the fluorescent reporters [11]. In addition, fluorescent reporters that require excitation by light impose restrictions on the experimental conditions and might affect the circadian outputs as circadian systems are sensitive to light. In vivo bioluminescence-based assays, on the other hand, rely on the chemical reaction between the enzyme luciferase and a luminogenic substrate. Thus, monitoring bioluminescence does not require excitation light and provides a sensitive and reliable way for long-term monitoring of circadian rhythms. Luciferase genes cloned from bacteria, beetles, deep sea shrimp, and Renilla are generally used as bioluminescent reporters. Each one of these luciferases has specific properties related to the size of the enzyme, the luminogenic substrate, wavelength and intensity of luminescence, half-life, etc. High-resolution bioluminescence assays have been achieved both by optimizing reporter genes and by improving the sensitivity of the detectors for luminescence. Native luciferases have been optimized for bioluminescent assays by genetic modifications. Adjusting the codon usage increased luciferase expression up to several hundred-fold in some instances [12]. Monitoring luminescence using photo-multiplier tubes (PMTs) or the highly sensitive electron-multiplying chargecoupled-device (EM-CCD) camera enables to minimize the duration of the photon count or exposure time, thus reducing the time

Monitoring Bioluminescence Rhythms in Arabidopsis

397

in which samples are under darkness. The microplate luminometers use PMTs and can be easily controlled by appropriate software that precisely program the measurements. In this way, microplate luminometers can automatically and simultaneously record bioluminescence rhythms within appropriate time intervals. This highthroughput monitoring system for bioluminescence is suitable for screening clock-related mutants or chemical compounds affecting circadian rhythms in Arabidopsis [13]. In this chapter, we describe how to monitor in vivo circadian rhythms in Arabidopsis seedlings using bioluminescent reporters and a microplate luminometer. The protocol has been successfully used to follow circadian rhythms under a variety of light and temperature conditions, and will likely be useful for researchers not only with an interest in chronobiology but in many other areas of plant biology.

2 2.1

Materials Equipment

1. Orbital incubator. 2. Electroporation cuvette (Bio-Rad). 3. Bio-Rad Gene Pulser (Bio-Rad). 4. 50 mL Sterile Falcon tube (FALCON). 5. Disposable plastic bag. 6. 1.5 mL microcentrifuge tubes (Eppendorf). 7. Tube rotator. 8. Sterile filter paper. 9. Surgical paper tape (3M). 10. Aluminum foil. 11. Laminar flow cabinet. 12. 96-Well Microplate: white, sterile with lid (Berthold Technologies). 13. Cover film (Ratiolab). 14. Tweezers. 15. Hypodermic needle (BD). 16. Luminometer LB-960 (Berthold Technologies). 17. Software Microwin (Mikrotek Laborsysteme). 18. BioDare2 software. 19. Environmentally controlled chamber (Inkoa Sistemas).

2.2

Reagents

1. YEB media: beef extract 5 g/L, yeast extract 1 g/L, peptone 5 g/L, sucrose 5 g/L, MgCl2 0.5 g/L. 2. 1 mM HEPES, pH 7.0.

398

Masaaki Okada and Paloma Mas

3. 10% Glycerol. 4. Luria-Bertani (LB) plate: yeast extract 5 g/L, bacto-tryptone 10 g/L, NaCl 10 g/L, agar 15 g/L. 5. Floral dip solution: 5% sucrose solution containing 0.03% of Silwet L-77 (Lehle Seeds). 6. Ethanol. 7. Sterilization solution: 70% ethanol (v/v), 0.1% Triton X-100 (v/v). 8. Murashige and Skoog (MS) medium. 9. Luciferin solution: 1.4 mM D-Luciferin acid form (Biothema) in 2.6 mM 2-(N-morpholino)ethanesulfonic acid (MES) pH 5.8. Filter the solution under sterile conditions. 2.3

Bacteria

Agrobacterium tumefaciens GV2260.

2.4

Plants

Transgenic Arabidopsis thaliana lines expressing a bioluminescent reporter [14].

3

Methods

3.1 Luciferase Vectors and Bioluminescent Reporter Constructs

1. Choose the appropriate luciferase vector for your studies and generate transcriptional or translational fusions of the promoters, genes, or proteins of interest fused to the chosen luciferase. 2. Use standard molecular cloning techniques for the generation of the constructs [15]. A number of bioluminescent reporters with different luciferases, such as the Firefly luciferase, the Renilla luciferase, or the NanoLUC luciferase, are currently available (Promega). Each one of these luciferases has specific properties related to the size of the enzyme, the luminogenic substrate, wavelength and intensity of luminescence, half-life, etc. (please see Table 1). Choosing the appropriate luciferase depends on the particular analysis to be performed. In general, instability (shorter half-life) of the luciferase activity is preferred for monitoring promoter activities fused to the reporter, as the luminescence reflects the actual dynamics of the promoter (not interfered by a stable reporter). Recent studies have reported that NanoLUC has longer half-life compared to firefly luciferase [16]. The NanoLUC can be a useful tool for tracking protein accumulation and dynamics in plants (see Notes 1 and 2). The luc+ gene has been modified to improve its function as a genetic reporter. These modifications include the removal of the peroxisomal translocation sequence, resulting in the transport of luciferase to the cytoplasm and the removal of glycosylation sites.

Monitoring Bioluminescence Rhythms in Arabidopsis

399

Table 1 List of common luciferases used for circadian studies in plants [34]

a

Luciferase

Substrate

Size (kDa)

Wavelength (nm)

Vector seriesa

Firefly

D-Luciferin

61

560

pGL3 (luc+) pGL4 (luc2)

NanoLUC

Furimazine

19

460

pNL

Renilla

Coelenterazine

36

480

pGL4 (hRluc)

Available from Promega

Together these changes produce a several fold increase in the light signal [17]. The cryptic regulatory sequences in the luc gene that may adversely affect transcription, resulting in anomalous expression of the reporter, were removed without changing the amino acid sequence to create the luc2 gene. The generated luc2 gene displays higher expression than the luc+ gene. In addition, sequences resembling splice sites, poly(A) addition sequences, Kozak sequences (translation start sites for mammalian cells), E. coli promoters or E. coli ribosome-binding sites were removed wherever possible. This process has led to a reduced number of cryptic regulatory sequences and therefore a reduced risk of anomalous transcription. A similar process was performed using Renilla luciferase to produce the hRluc gene. (Promega website: “Reporter Genes and their Applications”). 3.2 Transformation of Arabidopsis thaliana

Bioluminescent reporter constructs can be used to transform Arabidopsis plants to generate stable transgenic lines following the methods described in this section. Please see Note 3 for alternative analyses with transient expression.

3.2.1 Transformation of Agrobacterium

This section is adapted from ref. 18. 1. Inoculate 10 mL of YEB media with the Agrobacterium strain GV2260 (rifampicin, kanamycin resistant). Incubate overnight at 28  C in an orbital incubator. 2. Use the overnight culture to inoculate 300 mL of YEB and incubate in an orbital incubator at 28  C until an optical density at 600 nm (OD600) of 0.5 is reached. Chill bacteria on ice and harvest by centrifugation. Wash bacteria with 10 mL of 1 mM HEPES, pH 7.0, three times and once with 10% glycerol, and finally suspend them in 3 mL 10% glycerol. Snap-freeze in 200 μL aliquots (competent Agrobacterium) and store at 80  C. 3. Transform competent Agrobacterium by using the electroporation method with the constructs generated in Subheading 3.1. Thaw competent cells on ice and add approximately

400

Masaaki Okada and Paloma Mas

200 ng of vector. Put them in a 0.2-cm electroporation cuvette and place it in the Bio-Rad Gene Pulser. Apply a single electric pulse of 2.5 kV initial voltage using the 25 μF capacitor. After the pulse add 500 μL YEB media into the cuvette and pipet gently. Transfer to a microcentrifuge tube and incubate for 1 h at 28  C. 4. Spread aliquots on selective LB plates and incubate for 48 h at 28  C. 3.2.2 AgrobacteriumMediated Transformation of Arabidopsis thaliana by the Floral Dip Method

This section is adapted from ref. 19. 1. Prepare Arabidopsis plants, which are flowering. Optimal plants have many immature flower clusters and only few fertilized siliques. As an option, first bolts can be clipped to encourage proliferation of many secondary bolts. 2. Select an Agrobacterium colony from selective LB plates generated in Subheading 3.2.1, and resuspend bacteria in 10 μL H2O. Spread half of the volume immediately on a selective YEB plate, and incubate at 28  C for 2–3 days. Use the other half to verify the presence of transformed DNA construct by PCR analysis. 3. Collect the densely grown bacteria from the plate by scraping, and resuspend them in 30 mL YEB in a sterile Falcon tube. OD600 should be about 2.0. 4. Prepare 120 mL of floral dip solution per transformation, and pour solution into a disposable plastic bag and add the bacteria. 5. Dip the inflorescence of the plants into the Agrobacterium solution for 10 s under gentle agitation. A film of liquid coating the plants should be observed. The bacteria are distributed to all plant parts by gently pressing the outside of the bag. 6. To maintain high humidity, cover dipped plants for 16–24 h. Avoid exposure to excessive sunlight, otherwise the temperature inside can get high. 7. Water and grow the plants as normal, tying up loose bolts. Stop watering as seeds become mature. 8. Harvest dry seeds. 9. Select for transformants using appropriate antibiotics or herbicides. Transformation of luciferase constructs can be confirmed by measurements of bioluminescence. Homozygous lines containing a single copy of the reporter gene should be selected.

3.3 Seed Sterilization, Seedling Growth, and Entrainment

1. For sterilization, seeds are placed in 1.5 mL microcentrifuge tubes and surface sterilized by soaking them in sterilization solution for 10 min on a rotator, followed by 2 washes with 70% ethanol and 1 wash with 100% ethanol (Fig. 1) (see Note 4).

Monitoring Bioluminescence Rhythms in Arabidopsis

Triton X-100

401

EtOH

1. Seed sterilization in the dark (2 ~ 3 days)

1

2

3

1

2

3

3. Seedling growth and synchronization in a growth chamber (7 ~ 12 days)

A B

A B 4. Transfer of seedlings to the 96-well plate under sterile conditions

5. Monitoring bioluminescence using a microplate luminometer placed on the growth chamber

Fig. 1 Schematic drawing depicting the main steps for the in vivo bioluminescence analyses of circadian rhythms in Arabidopsis thaliana seedlings using a microplate luminometer

2. Seeds are then dried on a sterile filter paper in a laminar flow cabinet for 1 h and sown on plates containing MS medium (3% sucrose, 0.8% Agar, see Note 5). 3. Plates are sealed with surgical paper tapes and covered with aluminum foil. 4. After 2–3 days of stratification at 4  C in the dark, plates are then transferred to an environmentally controlled plant growth chamber. 5. Seedlings are grown for about 7–12 days in the chamber at the appropriate entraining (synchronizing) cycles, such as lightdark cycles and/or high-low temperature cycles, before timecourse bioluminescence assay. 3.4 Preparation of 96-Well Plates

All procedures should be done in a laminar flow cabinet. 1. Use 160 μL of agar (or liquid) MS medium and add 40 μL of a luciferin solution into each well of the 96-well plate.

402

Masaaki Okada and Paloma Mas

2. The 7–12-day-old seedlings expressing the luciferase and grown on the MS agar medium as specified above should be then removed from the plates one by one using sterile tweezers and placed into the wells of the 96-well plates (one seedling per well). Each seedling should be picked carefully to avoid damaging the shoots, hypocotyls, or breaking the roots. The seedling should be placed in the center of the well. Bioluminescence rhythms of different organs can be also examined [10, 20, 21]. Luminescence from distinct parts of seedlings (e.g., shoots and roots) can be also measured simultaneously [10, 21]. For this method, one side of the walls of the 96-wells should be slightly serrated to allow communication between two adjacent wells. Seedlings are then horizontally positioned such that the shoot can be placed in one well and the roots in the contiguous well [10, 21]. 3. Cover the plate tightly with a cover film. Make 4–6 small holes on the seal at each well with a sterile hypodermic needle. 3.5 Setting the Luminometer and Chamber Conditions

Install a microplate luminometer (e.g., LB-960, Berthold Technologies) into a growth chamber, in which conditions of illumination and temperature can be programmed (see Note 6). Set the chamber conditions under which the measurement should be done, i.e., continuous light, light/dark cycles, high/low temperature cycles, etc.

3.6 Setting Measurement Conditions

Set the measurement conditions using the software Microwin (Mikrotek Laborsysteme). To that end, click on settings and: 1. Select the wells of the plate to be measured (normally the 96 wells). 2. Click on “measurements” and select 5 s per well. 3. Select the option “unload” to let the plate out of the luminometer after every time point. Introduce the seconds (time interval) between two measurements in “use delay time”. 4. Set the number of repeats, e.g., 168 repeats for the 1-week measurements with 1-h intervals or 84 repeats for 1-week measurements every 2 h. 5. Introduce the name of your experiment including the relevant information to be easily identified. Start measurements.

3.7

Data Analyses

The circadian analysis of luminescence data can be achieved by using the BioDare2 software [22]. BioDare2 is a repository for circadian, biological data, providing a platform for data sharing and rhythmic analysis. You can estimate periodic functions, such as period lengths, phases, and amplitudes, of your luminescence data by uploading it to the BioDare2 website.

Monitoring Bioluminescence Rhythms in Arabidopsis 3.7.1 Upload Data Sets

403

1. After measurements, export bioluminescence data as an excel file. The excel file should be with one of the columns holding the measurement time and the other columns recording measured values. 2. To analyze data with BioDare2, go to the New Experiment page on the BioDare2 website and create an experiment with descriptions. (Registration is required the first time). 3. To upload the data file, choose Excel Table as File Format, and select data layout, such as data in rows or columns, depending on your file layout. 4. After selecting Time unit, Time offset, and information about rows and columns of the data file, timeseries can be imported.

3.7.2 Analysis of Data

After importing timeseries, individual bioluminescence rhythms are shown, and those are ready to be analyzed. Choose periodic functions, which you need to analyze, and set conditions, such as Data window and Analysis method. Results of analyses can be shown on the webpage, and also can be downloaded to your local disk.

3.7.3 Further Information About Using BioDare2

To ensure that the data are in the correct format for BioDare2, you should follow the instruction described in the Documentation page on the BioDare2 website. Furthermore, descriptions about analyses performed in BioDare2 are available in the page.

4

Notes 1. Several possible luciferases isolated from different species are available for studies in plants [23]. These luciferases emit bioluminescence with different wavelengths (colors), which can be separated with appropriate interference filters. Thus, using different bioluminescent reporters, multiple gene expression levels are simultaneously monitored. 2. Clock-gene promoter activities in specific tissues can be monitored by the tissue-specific luciferase assay (TSLA) developed by Endo et al. [24]. In TSLA, the carboxy- and amino-terminal fragments of firefly luciferase are driven by tissue-specific and clock promoters, respectively. Fragmented N- and C-terminal luciferase can be also used for bioluminescence complementation assays in which physical interactions between genes of interest in the circadian system can be assayed [25]. 3. Transient expression of the bioluminescent reporter can be also assayed by particle bombardment with detached leaves or polyethylene glycol (PEG)-mediated transfection of protoplasts [26, 27]. The transient methods allow to analyze circadian rhythms with available Arabidopsis mutant/overexpression

404

Masaaki Okada and Paloma Mas

lines without generating stable reporter lines. Also, these methods are useful for plants/crops, in which the generation of stable transgenic lines is very difficult or not possible [28, 29]. By using the particle bombardment method, circadian rhythms in individual cells on leaves can be analyzed, and rhythms can be genetically modified with effector constructs, such as RNAi, CRISPR/Cas9, and overexpression. By PEG-mediated transfection with protoplasts, cell autonomy on circadian rhythms can be assessed by transient expression of amiRNAs targeting clock genes. 4. Bleach solution can be used for sterilizing the seeds. To dilute bleach, add 100 mL of bleach to 100 mL of distilled water. Add 50 μL of Tween 20 detergent to the bleach solution. In a laminar flow hood, add 500 μL of the bleach solution to the microcentrifuge tube containing seeds. Shake the tube for 10 min with a platform shaker to keep the seeds suspended. Remove the bleach solution from the microcentrifuge tube using a pipette or an aspirator fitted with a pipette tip on the end. Add 500 μL of sterile distilled water to the tube. Close the tube and invert to mix. Allow seeds to settle to the bottom of the tube. Once seeds have settled to the bottom of the tube, carefully remove the bleach solution by pipetting. Repeat this rinsing process six times. Add 1 mL of autoclaved distilled water to the tube to suspend the seeds [30]. 5. The properties of circadian rhythms, such as robustness and/or period length, are affected by the presence of exogenous sucrose [8]. 6. To observe spatiotemporal patterns of clock-genes expression, the microscope or the highly sensitive EM-CCD camera can be also used [31, 32]. Controlling devices by an imaging software, such as HOKAWO (Hamamatsu photonics), enables us to capture a series of bioluminescence images with appropriate intervals automatically [33]. Relevant web links: Promega website: “Reporter Vectors” h t t p s : //w w w.p r o m e g a . e s / e n / r e s o u rc e s / v e c t o rsequences/reporter-vectors/ Promega website: “Reporter Genes and their Applications” https://www.promega.es/en/resources/guides/cell-biol ogy/bioluminescent-reporters/ BioDare2 website https://biodare2.ed.ac.uk BioDare2 website: “TimeSeries data and formats” https://biodare2.ed.ac.uk/documents/timeseries-data

Monitoring Bioluminescence Rhythms in Arabidopsis

405

Acknowledgments The Mas laboratory is funded with a research grant (PID2019106653GB-I00) from the MCIN/AEI/10.13039/ 501100011033, from the Ramon Areces Foundation and from the Generalitat de Catalunya (AGAUR). P.M. laboratory also acknowledges financial support from the CERCA Program/Generalitat de Catalunya and by the Spanish Ministry of Economy and Competitiveness through the “Severo Ochoa Program for Centers of Excellence in R & D” 2016–2019 (CEX2019-000902-S). M.O. is funded with a “Severo Ochoa” Internationalization Postdoctoral Program. References 1. Young MW, Kay SA (2001) Time zones: a comparative genetics of circadian clocks. Nat Rev Genet 2:702–715 2. McClung CR (2006) Plant circadian rhythms. Plant Cell 18:792–803 3. Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22:R648–R657 4. McClung CR (2013) Beyond Arabidopsis: the circadian clock in non-model plant species. Semin Cell Dev Biol 24:430–436 5. Wang ZY, Tobin EM (1998) Constitutive expression of the CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) gene disrupts circadian rhythms and suppresses its own expression. Cell 93:1207–1217 6. Matsushika A, Makino S, Kojima M, Mizuno T (2000) Circadian waves of expression of the APRR1/TOC1 family of pseudo-response regulators in Arabidopsis thaliana: insight into the plant circadian clock. Plant Cell Physiol 41: 1002–1012 7. Nusinow DA, Helfer A, Hamilton EE, King JJ, Imaizumi T, Schultz TF, Farre´ EM, Kay SA (2011) The ELF4-ELF3-LUX complex links the circadian clock to diurnal control of hypocotyl growth. Nature 475:398–402 8. Haydon MJ, Mielczarek O, Robertson FC, Hubbard KE, Webb AAR (2013) Photosynthetic entrainment of the Arabidopsis thaliana circadian clock. Nature 502:689–692 9. Mizuno T, Nomoto Y, Oka H, Kitayama M, Takeuchi A, Tsubouchi M, Yamashino T (2014) Ambient temperature signal feeds into the circadian clock transcriptional circuitry through the EC night-time repressor in Arabidopsis thaliana. Plant Cell Physiol 55:958–976 10. Chen WW, Takahashi N, Hirata Y, Ronald J, Porco S, Davis SJ, Nusinow DA, Kay SA, Mas P

(2020) A mobile ELF4 delivers circadian temperature information from shoots to roots. Nat Plants 6:416–426 11. Shaw SL, Ehrhardt DW (2013) Smaller, faster, brighter: advances in optical imaging of living plant cells. Annu Rev Plant Biol 64:351–375 12. Kirkpatrick A, Xu T, Ripp S, Sayler G, Close D (2019) Biotechnological advances in luciferase enzymes. In: Bioluminescence - analytical applications and basic biology. IntechOpen, Rijeka, Croatia, p 13 13. Uehara TN, Mizutani Y, Kuwata K, Hirota T, Sato A, Mizoi J, Takao S, Matsuo H, Suzuki T, Ito S et al (2019) Casein kinase 1 family regulates PRR5 and TOC1 in the Arabidopsis circadian clock. Proc Natl Acad Sci U S A 116: 11528–11536 14. Millar AJ, Short SR, Chua NH, Kay SA (1992) A novel circadian phenotype based on firefly luciferase expression in transgenic plants. Plant Cell 4:1075–1087 15. JoVE Science Education Database (2020) Basic methods in cellular and molecular biology. Molecular cloning. JoVE, Cambridge, MA. https://www.jove.com/science-educa tion/5074/molecular-cloning. Accessed 6 May 2020 16. Urquiza-Garcı´a U, Millar AJ (2019) Expanding the bioluminescent reporter toolkit for plant science with NanoLUC. Plant Methods 15:68 17. Hall A, Brown P (2007) Monitoring circadian rhythms in Arabidopsis thaliana using luciferase reporter genes. In: Circadian rhythms. Humana Press, Totowa, NJ, pp 143–152 18. Mattanovich D, Ru¨ker F, da Ca¨mara Machado A, Laimer M, Regner F, Steinkeliner H, Himmler G, Katinger H (1989) Efficient transformation of

406

Masaaki Okada and Paloma Mas

Agrobacterium spp. by eletroporation. Nucleic Acids Res 17:6747–6747 ¨ lker B, Soms19. Logemann E, Birkenbihl RP, U sich IE (2006) An improved method for preparing Agrobacterium cells that simplifies the Arabidopsis transformation protocol. Plant Methods 2(1):16 20. Kim H, Kim Y, Yeom M, Lim J, Nam HG (2016) Age-associated circadian period changes in Arabidopsis leaves. J Exp Bot 67: 2665–2673 21. Takahashi N, Hirata Y, Aihara K, Mas P (2015) A hierarchical multi-oscillator network orchestrates the Arabidopsis circadian system. Cell 163:148–159 22. Moore A, Zielinski T, Millar AJ (2014) Online period estimation and determination of rhythmicity in circadian data, using the BioDare data infrastructure. In: Staiger D (ed) . Springer, New York, pp 13–44 23. Ogura R, Matsuo N, Wako N, Tanaka T, Ono S, Hiratsuka K (2005) Multi-color luciferases as reporters for monitoring transient gene expression in higher plants. Plant Biotechnol 22:151–155 24. Endo M, Shimizu H, Nohales MA, Araki T, Kay SA (2014) Tissue-specific clocks in Arabidopsis show asymmetric coupling. Nature 515: 419–422 25. Xie Q, Wang P, Liu X, Yuan L, Wang L, Zhang C, Li Y, Xing H, Zhi L, Yue Z et al (2014) LNK1 and LNK2 are transcriptional coactivators in the Arabidopsis circadian oscillator. Plant Cell 26:2843–2857 26. Kanesaka Y, Okada M, Ito S, Oyama T (2019) Monitoring single-cell bioluminescence of Arabidopsis leaves to quantitatively evaluate

the efficiency of a transiently introduced CRISPR/Cas9 system targeting the circadian clock gene ELF3. Plant Biotechnol 36:187– 193 27. Kim J, Somers DE (2010) Rapid assessment of gene function in the circadian clock using artificial microRNA in Arabidopsis mesophyll protoplasts. Plant Physiol 154:611–621 28. Muranaka T, Oyama T (2016) Heterogeneity of cellular circadian clocks in intact plants and its correction under light-dark cycles. Sci Adv 2:e1600500 29. Okada M, Muranaka T, Ito S, Oyama T (2017) Synchrony of plant cellular circadian clocks with heterogeneous properties under light/ dark cycles. Sci Rep 7:317 30. Lindsey BE, Rivero L, Calhoun CS, Grotewold E, Brkljacic J (2017) Standardized method for high-throughput sterilization of Arabidopsis seeds. J Vis Exp 2017:1–7 31. Wenden B, Toner DLK, Hodge SK, Grima R, Millar AJ (2012) Spontaneous spatiotemporal waves of gene expression from biological clocks in the leaf. Proc Natl Acad Sci U S A 109: 6757–6762 32. Fukuda H, Ukai K, Oyama T (2012) Selfarrangement of cellular circadian rhythms through phase-resetting in plant roots. Phys Rev E 86:041917 33. Muranaka T, Kubota S, Oyama T (2013) A single-cell bioluminescence imaging system for monitoring cellular gene expression in a plant body. Plant Cell Physiol 54:2085–2093 34. England CG, Ehlerding EB, Cai W (2016) NanoLuc: a small luciferase is brightening up the field of bioluminescence. Bioconjug Chem 27:1175–1187

INDEX A

C

Acetylome ........................................................................ 88 Actogram .............................................193, 213, 214, 379 Acute lung injury (ALI)............. 286, 288–289, 292–295 Acute respiratory distress syndrome (ARDS).............. 286 Adenoviral vector .......................219, 223, 224, 232, 236 Agarose Gel Electrophoresis for Resolving Aggregates (AGERA) .................................................. 373–382 Amplitude ................................................... 35, 37, 41–45, 47–49, 52, 59, 60, 62, 63, 66, 68–76, 82–84, 87, 89, 90, 96, 97, 99, 101, 122, 127, 129, 130, 132, 133, 154–157, 161, 165, 182, 213, 215, 220, 228, 229, 233, 237, 279–281, 306, 307, 313, 321, 331, 338, 402 Amplitude-phase models ...................... 68–70, 73–74, 76 Anaconda ...................................................................36–38 Anesthesia ................................................... 141, 204, 219, 222, 223, 226–230, 239, 291, 294, 298 Arabidopsis thaliana ............................................ 395–405 Arnold tongue ...........................................................72–74 Artificial light.............................................................1, 3–5 Autodock tools ........................................................18, 23–26, 30 Vina .................................................. 18, 23, 25–26, 29 Automated longitudinal luciferase imaging gas-and temperature-optimized recorder (ALLIGATOR) ........................................ 125–134

C57Bl/6J mice...........................118, 183, 234, 306, 331 Caloric restriction (CR) ....................................... 329–339 Cell based circadian assays ................................................ 96 cycle ........................................................105–123, 341 Cellular colocalization .................................................. 195 CellulaRhythm ....................................................... 97, 100 CHARMM ...................................................17–19, 22, 27 Chimera ........................................................................... 17 Chromatin immunoprecipitation (ChIP) followed by deep sequencing (ChIP-seq) .... 343, 344 ChronoStar ..........................................158, 161, 164, 165 Circadian epigenome ............................................................... 341 metabolomics ................................................. 311–324 omic datasets ............................................................. 87 omic time series...................................................82, 87 pacemaker ........................................... 42, 67, 69, 137, 154, 156, 169, 181, 182, 191, 194, 211, 218, 312, 313, 341 rhythms ............................................v, 3–5, 15, 16, 48, 55–76, 81–92, 95, 96, 99, 105–123, 125, 133, 137, 148, 153, 155, 156, 181, 211, 219, 243–252, 255–263, 302, 307, 329–339, 341–349, 353–367, 373, 395–405 time series ....................................... 35–53, 62, 69, 82, 87, 126, 128, 132, 158, 161, 333, 335–336, 396 CircadiOmics .............................................................81–92 Circatidal rhythmicity .......................................... 385, 386 Clock controlled gene expression ............... 74, 75, 218, 312 gene.................................................16, 48, 49, 59, 64, 83, 95, 96, 99, 106, 112, 118, 137–141, 143–146, 150, 154–156, 194, 199, 211, 217, 331, 359, 386, 388, 396, 403, 404 modulating compounds....................................95–102 neurons ............................................. 70, 95, 138, 154, 155, 181, 188, 373–382 reporter ................................. 357, 359–360, 365, 366 ClockLab ..................................................... 212, 213, 215 Confocal microscopy ................................. 114–118, 199, 204, 207, 287, 364

B BALB/C mice ............................................. 286, 288, 289 BeFly3 ................................................................... 377–379 Bifurcation diagrams .................................................62–64 BIO_CYCLE .................................................... 81–92, 321 Bioluminescence recording ...................... 42, 43, 45, 126 Blood ............................................... 5, 15, 174, 177, 228, 258, 265, 266, 268, 273, 275–278, 281–283, 286, 288, 291, 292, 296–298, 312, 318, 323, 333, 337, 393 Bone marrow.............................................. 265, 269, 272, 275, 282, 283, 286 Brain..........................................4, 88, 92, 181–187, 192, 195,197, 202, 204, 237, 275, 317, 323, 338, 341, 377, 388

Guiomar Solanas and Patrick-Simon Welz (eds.), Circadian Regulation: Methods and Protocols, Methods in Molecular Biology, vol. 2482, https://doi.org/10.1007/978-1-0716-2249-0, © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature 2022

407

CIRCADIAN REGULATION: METHODS AND PROTOCOLS

408 Index

Constant darkness ...................................... 193, 211, 214, 232, 307, 377 Core clock genes .............................................16, 83, 118, 139, 144, 199, 217 COSINOR non-linear curve fitting ............................. 280 Coupled oscillators.................................. 68–74, 156, 163 cRNA probes ........................................................ 198, 207 Cross-linking ........................................................ 343–348 Crustacea .............................................................. 385–393 Crystal structure................................................. 16–19, 27

D Deep neural network (DNN)...................................82–84 Degeneration................................................................. 374 Detrending .....................................36, 38–40, 43, 46–47, 50, 51, 122, 129, 161, 234 Dielectrophoresis ................................................. 255–263 Diet ......................................... 81, 88, 92, 223, 239, 302, 303, 329–335, 337, 338, 342 DNA shearing ............................................. 344, 346, 349 Docking .....................................................................15–31 Drosophila ......................................................57, 353–367, 373–382, 387, 391 Drug discovery ..........................................................15–31 dsRNAi ................................................................. 385–393

E Electrophysiology................................................. 181–189 Energy metabolism .............................................. 301–309 Ensemble analysis ......................................................42–44 Entrainment .......................................57, 68–74, 76, 129, 133, 134, 137–151, 154, 156, 218, 232, 257, 259–260, 262, 302–304, 312, 337, 356, 400–401 Epidermal interfollicular stem cell isolation ....... 243–252 Epigenetic ............................................................... 92, 343 Eurydice ................................................................ 385–393 Exercise .............................................................88, 92, 313 Ex vivo..............................................................35, 42, 148, 154, 156, 169, 218, 286

F Feeding regimen ........................................................... 329–339 Feeding:fasting cycles.................................................... 330 Firefly luciferase (Fluc)................................. 96, 125, 148, 149, 162, 218, 223, 235, 398, 403 Flow cytometry ............................................245, 265–284 Flow injection-electrospray ionization-tandem mass spectrometry (FIA-ESI-MS/MS) ........... 311–324 Fluorescent reporter ...........................162, 353–367, 396 Food consumption................ 138, 232, 239, 330–335, 337 intake .................................... 169, 302, 305, 308, 330

4-dimensional capture .................................287, 293–298 Free running conditions ...................................... 194, 232 Free-running period....................... 70, 72, 193, 211, 215 FUCCI......................................................... 107, 114, 117

G Gaussia luciferase assay ............................... 140, 145, 149 Gel electrophoresis...................... 150, 226, 373–382, 386 preparation .............................................................. 380 Glucose production assays ................. 170, 172, 173, 176–178 tolerance test (GTT) ...................................... 330, 337 Goodwin model ...........................................56, 59, 61–65

H Hanging drop cultures......................................... 157, 160 High-throughput screen ................................................................ 16, 138 sequencing technology ........................................... 341 Histone marks ...................................................... 342, 343 Hit compound validation ............................................... 99 Hormone .................................................... 4, 81, 96, 138, 139, 142–144, 146, 147, 149, 218, 312, 341 Human leukocyte antigens (HLAs) ............................. 288 Hypothalamic brain slice ..................................... 181–187 Hypothalamus .............................................. 95, 138, 186, 187, 211, 312, 341

I Igor Plus ........................................................................ 233 ImageJ................................................................... 132, 298 Immunoblotting .................................355, 376, 380–381 Immunocytochemistry............................... 192, 194–198, 200–204, 387–388 Immunocytodetection ......................................... 385–393 Immunofluorescence ................................. 113–114, 266, 275–279, 286, 292 Immunostaining......................... 196, 203, 359, 363–364 Indirect calorimetry ............................................. 301–309 Infrared motion detectors ............................................ 212 In situ hybridization ........................................... 192, 194, 198–201, 204, 207 Intercellular coupling........................................... 153–165 Interlocked feedback loops.......................................64–67 Intermittent fasting....................................................... 329 Intestine .....................................106, 110, 118, 174, 266, 269, 271–272, 275, 356, 357 Intraperitoneal injection ...................................... 173, 289 Inverted light cycle .............................................. 268, 296 In vivo recording ........................................................ 219, 228 Ischemia/reperfusio´n ................................................... 286

CIRCADIAN REGULATION: METHODS J JTK_CYCLE ..............................83, 85, 87, 90, 249, 321

K Kaplan-Meyer survival curves ....................................... 289 Kidney....................... 149, 237, 266, 269, 273, 275, 323

L Light-dark cycle................................. 153, 154, 188, 192, 211, 212, 217, 281, 286, 302, 307, 312, 333, 341, 360 Light:dark cycle .................................................... 193, 331 Light schedule ............................................................... 307 Limit cycles ..................................... 59–66, 68, 69, 73, 74 Liquid chromatography-tandem mass spectrometry (LC-MS/MS) ........................................... 311–324 Live-cell imaging ........................................................... 153 Liver ................................................. 16, 88, 92, 138, 142, 143, 149, 169–171, 174, 177, 218, 219, 229, 236, 238, 239, 265, 266, 268, 270, 275, 317, 323, 345–346, 2807 Locomotor activity.............................................. 155, 192, 193, 211–215, 220, 231, 232, 302, 309, 312, 376–379, 389 LPS..............................................286, 288, 289, 293, 296 Luciferase reporter ....................................... 96, 149, 162, 170, 176, 219, 220, 223, 237 LumiCycle ......................... 112, 139, 142, 143, 149, 162 Luminometer ............................................. 112, 139, 140, 142, 145, 146, 149, 162–164, 176, 395–404 Lung......................................................73, 237, 266, 268, 270–271, 275, 276, 282, 283, 285–298, 323 Lung intravital imaging ....................................... 289, 292

M Mammalian Huntingtin gene (mHTT).............. 373, 374 Mathematical modeling ............................................55–76 Matlab.................................. 58, 212, 213, 233, 320, 378 Metabolic Atlas .....................................................................90, 91 Cages............................................................... 301–309 networks .................................................................... 90 rhythms ...................................................169–178, 307 Metabolites .......................................16, 82, 90, 312–315, 317–322, 324, 342 Metabolomic targeted metabolomics.................. 314, 316, 318–320 untargeted metabolomics ............. 314, 315, 317–322 Mice ................................................. 16, 48, 99, 106, 177, 183, 192, 212, 219, 245, 265, 286, 302, 331, 351, 357, 387 Microinjection ............................................................... 385

AND

PROTOCOLS Index 409

Microscopy .........................................114–118, 122, 200, 203, 204, 207, 266, 277, 285–298, 359, 364–365 Microvasculature ............... 286, 288–290, 292–294, 298 Midgut dissection ................................................ 361–362 Mini-osmotic pump ...................220, 222, 228, 232, 239 Mouse ................................................... 16–18, 66, 87, 99, 106–118, 122, 134, 141, 170, 171, 173–177, 195, 196, 203, 212–214, 220–222, 226–230, 234, 236, 238, 239, 244, 247–248, 266, 268, 278, 287–291, 293–296, 302–307, 309, 312, 314, 330, 331, 334, 335, 349, 360, 385 MultiCycle ......................................................97, 100, 101

N NAMD................................................................ 17–18, 22 NETosis ......................................................................... 285 Neuronal firing rate ...................................................... 181 Neurotransmitters ................................. 70, 154–156, 182 Neutrophil aging ........................................................................ 266 extracellular traps (NETs)............................. 275, 276, 278–279, 285–298 migration ................................................................. 266 Nonlinear dynamics ........................................... 55, 56, 59 Non-stationary signals .................................................... 52 Numerical simulations .................................................... 56 Nutrition................................................ 92, 314, 329–331

O Optomorph ................................................................... 131 Ordinary differential equations (ODEs).................57–59, 63, 64, 66, 67, 69, 76 Organoids ............................................106, 108, 112, 156 Oscillations ........................................................35, 36, 39, 41–43, 49–53, 56, 57, 59–66, 68–69, 74, 75, 81–83, 87, 89, 96, 105, 106, 112, 122, 127, 133, 155, 158, 161, 169, 181, 193, 217, 218, 235, 243, 249, 265, 266, 321, 341, 353, 357 Oscillator .................................................... 15, 36, 56, 95, 133, 137, 153, 182, 192, 311, 338, 341, 398

P Pacemaker................................................. 42, 67, 69, 137, 154, 156, 169, 181, 182, 191, 194, 211, 218, 312, 313, 341 Perfusion..................................................... 171–174, 177, 182, 201, 202, 268, 296 Periodic trajectories...................................................83, 84 Period pulling......................................155, 159, 161, 162 Peripheral clocks ...................................... 73, 156, 218, 331, 357 organs ............................................................. 217–240 oscillators .............................. 156, 157, 162, 169, 341

CIRCADIAN REGULATION: METHODS AND PROTOCOLS

410 Index

Phase resetting .......................................................... 143, 155 Phyre ................................................................................ 17 Polymerase chain reaction (PCR) ...................... 147, 256, 257, 259, 260, 262, 263, 388, 389, 400 Preindustrial societies........................................................ 1 Primary hepatocyte cell culture ............................................................... 169 isolation ..................................................169, 173–176 Protein aggregates ................................................................ 373 Data Bank (PDB) .................................. 17, 19, 27, 29 extraction ................................................374–375, 380 separation................................................................. 376 Proteomic ...................................................................... 286 Pulmonary inflammatory responses ............................. 286 Pymol .........................................................................17, 19 Python ........................................................ 26, 31, 36, 37, 44–51, 58, 82, 88, 320 Python-based Biological Oscillations Analysis Toolkit (pyBOAT) ......................................................36–53

R Raptor-X .......................................................................... 17 Real-time biolumicorder ................................................. 217–240 quantitative polymerase chain reaction (qPCR).... 249 recording ............................................... 125, 218, 220 Red blood cells ........................... 255–263, 267, 268, 273 Regeneration ................................................................. 106 Renilla luciferase......................................... 149, 398, 399 Respiratory exchange ratio (RER) ...................... 302, 306 Rodent .......................191, 211–215, 330, 331, 334–336

S Screening ..................... 15–31, 96, 97, 99–101, 138, 397 Secreted alkaline phosphatase (SEAP) assay ...............140, 145–146, 149 Single unit recording ........................................... 181–189 Skeletal muscle ........................................... 237, 266, 269, 273, 275, 276, 314, 317, 323 Skin .................................................... 227, 230, 237, 243, 244, 247–248, 251, 266, 269, 271, 275, 291, 292, 297, 341, 365, 393 Sleep............................................................ 1, 16, 81, 137, 231, 304, 312, 341, 375 Small animal ventilator ........................................ 288, 291 Small molecule compound ............................................. 95 Sonication ............................................150, 343, 346, 348 Specific pathogen-free (SPF) ............................... 288, 296 Spectral analysis ............................................................... 36 Spheroids ......................................................157, 159–164 Spinning disk confocal microscope .............................. 287

Spleen...................................................265, 266, 268, 271 Suprachiasmatic nucleus (SCN) ........................v, 3, 4, 42, 43, 48, 51, 70, 73, 95, 96, 137, 138, 154–157, 181–189, 191–208, 218, 232, 236, 312, 313, 338 Swiss-Model ....................................................... 17–20, 27 Synchronization ................................................42, 50, 59, 69–74, 134, 138, 142, 155, 156, 159, 162, 163, 170, 176, 218, 343, 401 Systems biology.........................................................81–92

T Temperature rhythms ................................................... 217 Thermo-neutral zone.................................................... 307 3D culture ..................................................................... 159 Time restricted feeding (TRF) ............................ 329–339 Time series analysis ................................................ 37, 161 Tissue slice ............................................. 45, 138–143, 148 Trachea ................................................289–291, 296, 356 Transcription factor ........................................................74, 139, 146, 147, 151, 341–343 reporter ..........................................112, 126, 129, 360 Transcriptomics ...................................................... 74, 286 Transfection.......................................... 99, 140, 144–145, 149, 162, 170, 176, 177, 219, 224, 403, 404 Transfusion-related acute lung injury (TRALI) .........286, 288–289, 293, 295–298 Tyramide signal amplification (TSA) .................. 195, 199

U U2OS cells........................... 96, 100, 126, 127, 129, 134 Ultra high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS)...... 314, 315

V Virtual screening ............................................................. 15 VMD.............................................................17, 19, 21–23

W Wavelet ridge ........................................................41, 47–48, 52 spectrogram .................................................. 40, 41, 49 Western blotting................................................... 376, 381 Whole-mount tissue clearing............................... 265–284

Z Zeitgeber .................................................. 3, 5, 57, 67, 69, 70, 72–74, 76, 138, 142, 153, 154, 156, 169, 192, 193, 195, 197, 211, 244, 268, 275, 286, 292, 293, 296, 306, 312, 333, 337, 341, 356, 360