Plant Circadian Networks: Methods and Protocols [1 ed.] 1493906992, 9781493906994

Citation preview

Methods in Molecular Biology 1158

Dorothee Staiger Editor

Plant Circadian Networks Methods and Protocols

METHODS

IN

M O L E C U L A R B I O LO G Y

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

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

Plant Circadian Networks Methods and Protocols

Edited by

Dorothee Staiger Molecular Cell Physiology, University of Bielefeld, Bielefeld, Germany

Editor Dorothee Staiger Molecular Cell Physiology University of Bielefeld Bielefeld, Germany

ISSN 1064-3745 ISSN 1940-6029 (electronic) ISBN 978-1-4939-0699-4 ISBN 978-1-4939-0700-7 (eBook) DOI 10.1007/978-1-4939-0700-7 Springer New York Heidelberg Dordrecht London Library of Congress Control Number: 2014937283 © Springer Science+Business Media New York 2014 This work is subject to copyright. All rights are reserved 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. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. 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. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is a brand of Springer Springer is part of Springer Science+Business Media (www.springer.com)

Preface As many organisms, plants evolved an endogenous timekeeper, the biological or “circadian” clock (from Latin circa diem, about a day), to synchronize their life with environmental cycles of light and dark and associated temperature cycles. The circadian clock generates an internal time structure, causing major processes in the cell to occur at regular 24-h intervals. This rhythmic component of physiological, biochemical, and molecular processes in the plant has long been mostly neglected when conceiving experiments or analyzing data. The volume “Plant Circadian Networks” provides a collection of protocols that describe how to monitor circadian rhythms at the molecular, biochemical, and physiological level, how to evaluate the data, and how to integrate the data to obtain an overarching picture of circadian networks in the cell. Chronobiologists, those scientists who occupy their time with studying biological timing (from Greek χρόνος, time or the God of time), have long sought to uncover the molecular underpinnings of endogenous rhythms. More recently, the question why such an endogenous timekeeper may be of benefit for an organism has been addressed. Plant chronobiology entered the molecular era when the first circadian transcript pattern was described by Klaus Kloppstech about three decades ago. Circadian gene expression experiments require a large number of data points, in other words RNA preparation, around the clock for several cycles of subjective day and night. Thus, they inherently were more laborious than an on/off situation that is measured in experiments looking at the impact of an external stimulus and required a higher precision because subtle differences in expression levels had to be disclosed. A major advance for the field was the development of the luciferase reporter as a noninvasive marker by Andrew Millar, Steve Kay, and coworkers, opening a way to automatization and large genetic screens, and thus leading to the identification of the first clock mutant in Arabidopsis thaliana. Later on, the use of microarrays and next-generation sequencing greatly advanced the field, moving from the analysis of a handful of rhythmic genes to the entire circadian transcriptome. The tight interconnection between endogenous timing and hormone signaling, responses to abiotic stress, and pathogen threat add another level of complexity. Moving from the model plant Arabidopsis thaliana to other systems allowed for identifying common design principles and peculiarities of the clock in different species that may relate to the particular requirements, e.g., seasonal control in trees. This volume provides a collection of protocols, both standard techniques and the most recent technical developments, to investigate clock-controlled parameters including transcript and small RNA levels, promoter activity using luciferase reporters, protein levels and posttranslational modification, protein–protein interaction, in vivo DNA–protein interaction and RNA–protein interaction, cellular redox state, Ca2+ levels, and innate immune responses. Other topics are seasonal processes like flowering time control. Particular emphasis is on the circadian system in the model plant Arabidopsis thaliana. In addition, techniques applied in trees, moss, and algae are covered.

v

vi

Preface

Several chapters deal with computational biology. Tools to identify transcription factor binding sites, or small RNA binding sites, and to visualize alternative splicing patterns in RNA-Seq data are covered. The use of BioDare (Biological Data repository) for data storage, data sharing, and processing as well as identification of rhythmic patterns in large data sets is described. Furthermore, it is illustrated how mathematical models can help to understand the design principles of the circadian oscillator and allow to make experimentally testable predictions, ultimately leading to refined oscillator models. The book is designed for the plant chronobiology community dealing with circadian biology. As the clock has a pervasive effect on all aspects of plant physiology, I hope that the protocols will be of general use to plant biologists. Bielefeld, Germany

Dorothee Staiger

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

v ix

1 Measurement of Luciferase Rhythms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Robertson McClung and Qiguang Xie 2 Online Period Estimation and Determination of Rhythmicity in Circadian Data, Using the BioDare Data Infrastructure . . . . . . . . . . . . . . . . Anne Moore, Tomasz Zielinski, and Andrew J. Millar 3 Global Profiling of the Circadian Transcriptome Using Microarrays. . . . . . . . . Polly Yingshan Hsu and Stacey L. Harmer 4 ChIP-Seq Analysis of Histone Modifications at the Core of the Arabidopsis Circadian Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jordi Malapeira and Paloma Mas 5 Quantitative Transcriptome Analysis Using RNA-seq . . . . . . . . . . . . . . . . . . . Canan Külahoglu and Andrea Bräutigam 6 Rapid and Parallel Quantification of Small and Large RNA Species . . . . . . . . . Corinna Speth and Sascha Laubinger 7 The RIPper Case: Identification of RNA-Binding Protein Targets by RNA Immunoprecipitation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tino Köster, Meike Haas, and Dorothee Staiger 8 A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data Using Integrated Genome Browser . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alyssa A. Gulledge, Hiral Vora, Ketan Patel, and Ann E. Loraine 9 AthaMap Web Tools for the Analysis of Transcriptional and Posttranscriptional Regulation of Gene Expression in Arabidopsis thaliana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reinhard Hehl and Lorenz Bülow 10 Analysis of mRNA Translation States in Arabidopsis Over the Diurnal Cycle by Polysome Microarray . . . . . . . . . . . . . . . . . . . . . . . Anamika Missra and Albrecht G. von Arnim 11 Immunoprecipitation-Based Analysis of Protein–Protein Interactions . . . . . . . Corinna Speth, Luis A.A. Toledo-Filho, and Sascha Laubinger 12 Comparative Phosphoproteomics to Identify Targets of the Clock-Relevant Casein Kinase 1 in C. reinhardtii Flagella . . . . . . . . . . . Jens Boesger, Volker Wagner, Wolfram Weisheit, and Maria Mittag 13 Pulsed Induction of Circadian Clock Genes in Arabidopsis Seedlings . . . . . . . . Stephen M. Knowles, Sheen X. Lu, and Elaine M. Tobin

1

vii

13 45

57 71 93

107

123

139

157 175

187 203

viii

Contents

14 The Use of Fluorescent Proteins to Analyze Circadian Rhythms . . . . . . . . . . . Ekaterina Shor, Miriam Hassidim, and Rachel M. Green 15 Measuring Circadian Oscillations of Cytosolic-Free Calcium in Arabidopsis thaliana. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timothy J. Hearn and Alex A.R. Webb 16 Circadian Life Without Micronutrients: Effects of Altered Micronutrient Supply on Clock Function in Arabidopsis . . . . . . . . . . . . . . . . . Patrice A. Salomé, Maria Bernal, and Ute Krämer 17 Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Katharina König, Helena Galliardt, Marten Moore, Patrick Treffon, Thorsten Seidel, and Karl-Josef Dietz 18 Circadian Regulation of Plant Immunity to Pathogens . . . . . . . . . . . . . . . . . . Robert A. Ingle and Laura C. Roden 19 Determination of Photoperiodic Flowering Time Control in Arabidopsis and Barley. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander Steffen, Andreas Fischer, and Dorothee Staiger 20 The Perennial Clock Is an Essential Timer for Seasonal Growth Events and Cold Hardiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mikael Johansson, Cristian Ibáñez, Naoki Takata, and Maria E. Eriksson 21 Monitoring Seasonal Bud Set, Bud Burst, and Cold Hardiness in Populus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mikael Johansson, Naoki Takata, Cristian Ibáñez, and Maria E. Eriksson 22 Transformation and Measurement of Bioluminescence Rhythms in the Moss Physcomitrella patens . . . . . . . . . . . . . . . . . . . . . . . . . . . Setsuyuki Aoki, Ryo Okada, and Santosh B. Satbhai 23 Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christoph Schmal, Jean-Christophe Leloup, and Didier Gonze Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

209

215

227

239

273

285

297

313

325

337 359

Contributors SETSUYUKI AOKI • Graduate School of Information Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan ALBRECHT G. VON ARNIM • Department of Biochemistry, Cellular and Molecular Biology, The University of Tennessee, Knoxville, TN, USA; Graduate School of Genome Science and Technology, The University of Tennessee, Knoxville, TN, USA MARIA BERNAL • Department of Plant Physiology, Ruhr University Bochum, Bochum, Germany JENS BOESGER • Institute of General Botany and Plant Physiology, Friedrich Schiller University Jena Am, Jena, Germany ANDREA BRÄUTIGAM • Plant Biochemistry Cluster of Excellence on Plant Science (CEPLAS), Heinrich Heine University of Düsseldorf, Düsseldorf, Germany LORENZ BÜLOW • Institut für Genetik, Technische Universität Braunschweig, Braunschweig, Germany KARL-JOSEF DIETZ • Biochemistry and Physiology of Plants, Faculty of Biology, Bielefeld University, Bielefeld, Germany MARIA E. ERIKSSON • Department of Plant Physiology, Umeå Plant Science Centre, Umeå University, Umeå, Sweden ANDREAS FISCHER • Department of Plant Sciences and Plant Pathology, Montana State University, Bozeman, MT, USA HELENA GALLIARDT • Biochemistry and Physiology of Plants, Faculty of Biology, Bielefeld University, Bielefeld, Germany DIDIER GONZE • Unite de Chronobiologie Theorique, Faculte des Sciences, Universite Libre de Bruxelles, Brussels, Belgium RACHEL M. GREEN • Department Plant and Environmental Sciences, Hebrew University, Edmund Safra Givat Ram Campus, Jerusalem, Israel ALYSSA A. GULLEDGE • Department of Bioinformatics and Genomics, North Carolina Research Campus, University of North Carolina at Charlotte, Charlotte, NC, USA MEIKE HAAS • Molecular Cell Physiology, Bielefeld University, Bielefeld, Germany STACEY L. HARMER • Department of Plant Biology, University of California, Davis, CA, USA MIRIAM HASSIDIM • Department Plant and Environmental Sciences, Hebrew University, Edmund Safra Givat Ram Campus, Jerusalem, Israel TIMOTHY J. HEARN • Department of Plant Sciences, University of Cambridge, Cambridge, UK REINHARD HEHL • Institut für Genetik, Technische Universität Braunschweig, Braunschweig, Germany POLLY YINGSHAN HSU • Department of Plant Biology, University of California, Davis, CA, USA CRISTIAN IBÁÑEZ • Biology Department, La Serena University, La Serena, Chile ROBERT A. INGLE • Department of Molecular and Cell Biology, University of Cape Town, Rondebosch, Cape Town, South Africa MIKAEL JOHANSSON • Molecular Cell Physiology, Bielefeld University, Bielefeld, Germany

ix

x

Contributors

STEPHEN M. KNOWLES • Department of Molecular, Cell and Developmental Biology, University of California, Los Angeles, CA, USA KATHARINA KÖNIG • Biochemistry and Physiology of Plants, Faculty of Biology, Bielefeld University, Bielefeld, Germany TINO KÖSTER • Molecular Cell Physiology, Bielefeld University, Bielefeld, Germany UTE KRÄMER • Department of Plant Physiology, Ruhr University Bochum, Bochum, Germany CANAN KÜLAHOGLU • iGrad Plant Heinrich Heine University of Düsseldorf, Düsseldorf, Germany SASCHA LAUBINGER • Center for Plant Molecular Biology (ZMBP), University of Tübingen, Tübingen, Germany; Chemical Genomics Centre (CGC) of the Max Planck Society, Dortmund, Germany; MPI for Developmental Biology, Tübingen, Germany JEAN-CHRISTOPHE LELOUP • Unite de Chronobiologie Theorique, Faculte des Sciences, Universite Libre de Bruxelles, Brussels, Belgium ANN E. LORAINE • Department of Bioinformatics and Genomics, North Carolina Research Campus, University of North Carolina at Charlotte, Charlotte, NC, USA SHEEN X. LU • Department of Molecular, Cell and Developmental Biology, University of California, Los Angeles, CA, USA JORDI MALAPEIRA • Center for Research in Agricultural Genomics (CRAG) Consortium CSIC-IRTA-UAB-UB, Parc de Recerca UAB, Edifici CRAG, Campus UAB, Bellaterra (Cerdanyola del Vallés), Barcelona, Spain PALOMA MAS • Center for Research in Agricultural Genomics (CRAG), Consortium CSIC-IRTA-UAB-UB, Parc de Recerca UAB, Edifici CRAG, Campus UAB, Bellaterra (Cerdanyola del Vallés), Barcelona, Spain C. ROBERTSON MCCLUNG • Department of Biological Sciences, Dartmouth College, Hanover, NH, USA ANDREW J. MILLAR • SynthSys, University of Edinburgh, Edinburgh, UK ANAMIKA MISSRA • Department of Biochemistry, Cellular and Molecular Biology, The University of Tennessee, Knoxville, TN, USA MARIA MITTAG • Institute of General Botany and Plant Physiology, Friedrich Schiller University, Jena Am Jena, Germany MARTEN MOORE • Biochemistry and Physiology of Plants, Faculty of Biology, Bielefeld University, Bielefeld, Germany ANNE MOORE • SynthSys, University of Edinburgh, Edinburgh, UK RYO OKADA • Graduate School of Information Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan KETAN PATEL • Department of Bioinformatics and Genomics, North Carolina Research Campus, University of North Carolina at Charlotte, Charlotte, NC, USA LAURA A. RODEN • Department of Molecular and Cell Biology, University of Cape Town, Rondebosch, Cape Town, South Africa PATRICE A. SALOMÉ • Department of Molecular Biology, Max Planck Institute for Developmental Biology, Tübingen, Germany; Department of Chemistry and Biochemistry, UCLA, Los Angeles, CA, USA SANTOSH B. SATBHAI • Graduate School of Information Science, Nagoya University Furo-cho, Chikusa-ku, Nagoya, Japan CHRISTOPH SCHMAL • Center for Biotechnology, Bielefeld University, Bielefeld, Germany THORSTEN SEIDEL • Biochemistry and Physiology of Plants, Faculty of Biology, Bielefeld University, Bielefeld, Germany

Contributors

xi

EKATERINA SHOR • Department of Plant and Environmental Sciences, Hebrew University, Edmund Safra Givat Ram Campus, Jerusalem, Israel CORINNA SPETH • Center for Plant Molecular Biology (ZMBP), University of Tübingen, Tübingen, Germany; Chemical Genomics Centre (CGC) of the Max Planck Society, Dortmund, Germany; MPI for Developmental Biology, Tübingen, Germany DOROTHEE STAIGER • Molecular Cell Physiology, University of Bielefeld, Bielefeld, Germany ALEXANDER STEFFEN • Molecular Cell Physiology, Bielefeld University, Bielefeld, Germany NAOKI TAKATA • Forest Bio-Research Center, Forestry and Forest Products Research Institute, Hitachi, Japan ELAINE M. TOBIN • Department of Molecular, Cell and Developmental Biology, University of California, Los Angeles, CA, USA LUIS A.A. TOLEDO-FILHO • Center for Plant Molecular Biology (ZMBP), University of Tübingen, Tübingen, Germany; Chemical Genomics Centre (CGC) of the Max Planck Society, Dortmund, Germany; MPI for Developmental Biology, Tübingen, Germany PATRICK TREFFON • Biochemistry and Physiology of Plants, Faculty of Biology, Bielefeld University, Bielefeld, Germany HIRAL VORA • Department of Bioinformatics and Genomics, North Carolina Research Campus, University of North Carolina at Charlotte, Charlotte, NC, USA VOLKER WAGNER • Institute of General Botany and Plant Physiology, Friedrich Schiller University Jena, Jena, Germany ALEX A.R. WEBB • Department of Plant Sciences, University of Cambridge, Cambridge, UK WOLFRAM WEISHEIT • Institute of General Botany and Plant Physiology, Friedrich Schiller University Jena, Jena, Germany QIGUANG XIE • Department of Biological Sciences, Dartmouth College, Hanover, NH, USA TOMASZ ZIELINSKI • SynthSys, University of Edinburgh, Edinburgh, UK

Chapter 1 Measurement of Luciferase Rhythms C. Robertson McClung and Qiguang Xie Abstract Firefly luciferase (LUC) is a sensitive and versatile reporter for the analysis of gene expression. Transgenic plants carrying CLOCK GENE promoter:LUC fusions can be assayed with high temporal resolution. LUC measurement is sensitive, noninvasive, and nondestructive and can be readily automated, greatly facilitating genetic studies. For these reasons, LUC fusion analysis is a mainstay in the study of plant circadian clocks. Key words Arabidopsis thaliana, Biological clocks, Circadian clock, Circadian rhythms, Firefly luciferase, Reporter gene, Transcriptional regulation

1

Introduction The circadian clock is an endogenous timekeeping mechanism that enables organisms to measure and use time to coordinate their biology with the environment. As photoautotrophs dependent on the sun for energy, plants are richly rhythmic. In plants, the circadian clock regulates many aspects of biology, including basic metabolism, and serves as a key player in the coordination of metabolic and signaling pathways [1–5]. It has also become clear that the circadian clock modulates responses to biotic and abiotic stress [6, 7]. One means by which the clock coordinates so many processes is pervasive control of gene expression at the levels of transcription, transcript processing, and transcript abundance [8–11]. Therefore, analyses of gene expression have been central to the elucidation of the timekeeping mechanism [4, 5, 7, 12]. Early studies of the role of the plant circadian clock in regulating gene expression focused on the analysis of steady-state mRNA abundance [13–18]. Time series experiments, so dear to the clock researcher, were exercises in sleep deprivation for the experimenter (for example, note the 132-h time course in Figs. 5 and 6 of Pilgrim and McClung [19]), and subsequent analysis by northern RNA blot hybridization was laborious and time consuming and

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_1, © Springer Science+Business Media New York 2014

1

2

C. Robertson McClung and Qiguang Xie

required the destruction of the experimental material. It became abundantly clear that a facile and noninvasive assay for promoter function would enhance molecular genetic analysis of clock function, permit forward genetic analyses, and greatly enhance the recruitment of graduate and postgraduate researchers to the study of plant clocks. Firefly (Photinus pyralis) luciferase [20] was demonstrated to be an effective reporter of plant promoter activity [21]. A fusion construct in which the promoter of the CHLOROPHYLL a/b BINDING PROTEIN2 (CAB2, also known as LIGHT HARVESTING CHLOROPHYLL a/b BINDING1*1 or LHCB1*1) gene fused to the firefly LUCIFERASE (LUC)-coding sequence effectively recapitulated the circadian rhythm in gene expression with LUC activity (light production) in transgenic Arabidopsis seedlings [22]. This breakthrough vastly accelerated the subsequent elucidation of the plant circadian clock network [7, 12]. Luciferase is the generic term for a class of enzymes that oxidize a substrate with the concomitant release of a photon. Luciferases have been found in a broad range of taxa, including bacteria, dinoflagellates, copepods, fireflies and click beetles, and the colonial marine cnidarians, sea pansies (Renilla spp.). Emission spectra of these diverse luciferases range from 400 to 620 nm [23]; firefly luciferase has maximal emission at 560 nm [24]. Beetle luciferases, including firefly and click beetle luciferases, mediate the oxidation of their substrate, D-luciferin, in the presence of ATP, Mg2+, and O2, with concomitant light emission. The firefly luciferase reaction occurs in two steps: first luciferin (D-LH2) reacts with ATP to yield luciferyl-adenylate (LH2-AMP), which is oxidized by molecular oxygen to form oxyluciferin (OxyLH2), CO2, and AMP [25]. Other luciferases use different substrates. For example, some luciferases from deep-sea organisms oxidize coelenterazine. Uniquely, bacterial luciferases (lux) do not require exogenous substrates for light emissions, making the lux system attractive as a reporter. However, the bacterial lux system is encoded with an operon of five genes [26], in which luxA and luxB encode luciferase. luxC, luxD, and luxE encode a reductase, transferase, and synthase, respectively, that form a complex and generate an endogenous aldehyde substrate for the bioluminescent reaction [27]. Decanal can serve as an exogenous substrate for the luxA/luxB luciferase, but the decanal level necessary for maximal lux activity damages Arabidopsis seedlings [28]. In the cyanobacterium, Synechococcus elongatus PCC 7942, the luxAB-encoded luciferase from Vibrio harveyi, has proven effective [29, 30]; lethality associated with exogenous substrate addition has been overcome with the introduction of the other genes necessary for the synthesis of the aldehyde substrate into the S. elongatus chromosome.

Luciferase Rhythms

3

The stability of firefly luciferase (LUC) mRNA, protein, and enzymatic activity is of particular relevance to the use of LUC as a reporter. LUC mRNA is relatively unstable. Curiously, the LUC protein itself appears to be rather stable in the absence of its luciferin substrate. Happily, LUC activity is unstable and becomes dependent on de novo translation and, hence, mRNA abundance, which in turn tracks de novo transcription of the LUC gene [22, 28]. Over the two decades since its introduction, firefly luciferase has become the preferred reporter for use in circadian studies in plants, and its use has been extended to include fungi, invertebrates, and vertebrates in both tissue culture and transgenic animals [31–37].

2

Materials 95 % Ethanol. Bleach. HCl concentrated (37 %). 1. One-half strength Murashige and Skoog Medium [38] (see Note 1): 1,650 mg/L NH4NO3, 440 mg/L CaCl2⋅2H2O, 370 mg/L MgSO4⋅7H2O, 170 mg/L KH2PO4, 1,900 mg/L KNO3, 6.2 mg/L H3BO3, 0.025 mg/L CoCl2⋅6H2O, 0.025 mg/L CuSO4⋅5H2O, 27.8 mg/L FeSO4⋅7H2O, 22.3 mg/L MnSO4⋅4H2O, 0.83 mg/L KI, 0.25 mg/L Na2MoO4⋅2H2O, 8.6 mg/L ZnSO4⋅7H2O, 37.2 mg/L Na2EDTA⋅2H2O plus 0.5 g/L MES free acid. The pH is adjusted to 5.8, and the medium is solidified with 7 g/L agar. After autoclaving, appropriate antibiotics (filter-sterilized) are added to select for the plasmid. We also add 500 μg/mL carbenicillin to select against residual Agrobacterium. 2. D-luciferin (potassium salt): 2.5 mM in water. We typically make a 100 mM stock solution which we store frozen and dilute 40-fold for use. 3. Hamamatsu ORCA II ER CCD camera (C4742-98 ERG; Hamamatsu Photonics, Hamamatsu City, Japan, http://www. hamamatsu.com) (see Note 2). 4. Topcount™ Microplate Scintillation Counter (Perkin Elmer) (see Note 2). 5. White or black 96-well microtiter plates (we typically use white Optiplates; Perkin-Elmer) and also clear 96-well microtiter plates (see Note 2). 6. TopSeal (Perkin-Elmer), a clear adhesive plastic sealant for the 96-well microtiter plates.

4

3

C. Robertson McClung and Qiguang Xie

Methods 1. Two alternative methods to sterilize Arabidopsis seeds can be used. (a) Surface-sterilize seeds: Soak seeds in sterile water for 30 min. Decant. Soak in 95 % ethanol for 10 min. Decant. Soak in 10 % bleach for 10 min. Decant. Rinse five times (5 min per rinse) in sterile water. (b) Alternatively, surface-sterilize seeds by chlorine vapor [39]. Place seeds in a microcentrifuge tube in a rack within a desiccator in a fume hood. Place a 250 mL beaker containing 100 mL bleach into the desiccator jar, carefully add 3 mL concentrated HCl to the bleach, and immediately seal the jar. Allow sterilization by chlorine fumes to proceed for 3–16 h (we typically allow sterilization to continue overnight). Open container in a fume hood, seal microfuge tubes or other seed containers, and remove surface-sterilized seed for use (see Note 3). 2. Plate seeds on one-half-strength MS medium, which can be supplemented with 1–2 % sucrose (see Note 4). For imaging with a CCD camera, seeds can simply be spread at the desired density or can be arrayed (square petri dishes are useful for this, and we have also placed individual seedlings in the wells of 96-well microtiter plate containing 200 μL of one-half-strength MS medium). For analysis by luminometer, plate seeds on onehalf-strength MS medium, which can be supplemented with 1–2 % sucrose and solidified with 0.7 % agar to facilitate the removal of seedlings with their roots intact. 3. Stratify for 3–4 days at 4 °C to synchronize germination. 4. Release seedlings into entraining conditions (light–dark cycles or temperature cycles) for 7–10 days or until primary leaves are emerging. 5. Prior to imaging with a CCD camera, D-luciferin must be applied (one can spray the seedlings with a 2.5 mM solution, or one can pipette luciferin onto the surface of the medium at ~30 μL per seedling) (see Note 5). Transfer seedlings into a growth chamber in continuous dark for imaging with a CCD camera supported by Metamorph software. We typically image for 30 min at 1–2-h intervals, although this can be varied according to reporter signal strength (see Note 6). Figure 1 shows examples of CCA1 and TOC1 promoter activity as measured with promoter:LUC fusions, imaging seedlings grown in 96-well microtiter plates. Rhythmic parameters (period and phase, as well as relative amplitude error, a measure of the

Luciferase Rhythms

5

Fig. 1 Time series of images of Arabidopsis seedlings taken with a CCD camera. Seedlings (one seedling per well in a 96-well microtiter plate) were entrained to a 12-h light/12-h dark cycle and released into continuous dark for imaging. Upper panel shows a time series of images of transgenic Arabidopsis seedlings carrying either CCA1pro:LUC or TOC1pro:LUC (four seedlings per transgene). The lower panel shows quantification of promoter:LUC activity. Data are presented as mean ± SEM (n = 24). Filled circles indicate CCA1pro:LUC, and open squares indicate TOC1pro:LUC. White bars indicate subjective light, and gray bars indicate subjective dark

robustness of rhythmicity) are extracted from time series data by fast Fourier transform nonlinear least squares (FFT-NLLS) [40] using the BRASS software package (http://millar.bio. ed.ac.uk/PEBrown/BRASS/BrassPage.htm). BRASS is a Microsoft Excel workbook for the analysis and display of rhythmic data series. BRASS provides a convenient user interface, allowing the user to import data from commonly used data acquisition packages (e.g., Metamorph, Night Owl, TopCount), automatically run FFT-NLLS, and export the output in MS Excel format. For analysis with a TopCount, following entrainment seedlings are transferred to 96-well microtiter plates (opaque, white, or black, to prevent light contamination between wells; we use white plates, Optiplate-96, Perkin Elmer), containing 200 μL of the same medium used for seedling growth during entrainment plus 30 μL of 2.5 mM D-luciferin. Plates are then

6

C. Robertson McClung and Qiguang Xie

Fig. 2 Quantification of CLOCK GENE promoter:LUC activity in continuous light measured with a TopCount. CCA1 (CCA1pro:LUC; red circles), PRR9 (PRR9pro:LUC; magenta squares), PRR5 (PRR5pro:LUC; cyan triangles), and TOC1 (TOC1pro:LUC; blue triangles) promoter activity as measured with promoter:LUC fusions in transgenic Arabidopsis seedlings. For each trace, data were normalized to the mean value for that promoter over the time series, which facilitates the comparison of promoters of different strengths. Data are presented as mean ± SEM for 24 seedlings (12 for TOC1pro:LUC). White bars indicate subjective light, and gray bars indicate subjective dark

sealed with a clear adhesive sealant (TopSeal; Perkin Elmer); 2–3 holes should be made in the sealant above each plant with a needle to allow gas exchange. The plates are fed into the TopCount sampling chamber from “stackers” with 20- or 40-plate capacity. The TopCount is completely automated (see Note 7). A “stop plate” must be placed after the last sample plate: this plate should have two barcode stickers on the right side to be recognized by the TopCount as the last of a stack of plates. This then triggers the restacking of the plates into the feeding stacker for a second sampling (see Note 8). Figure 2 shows examples of CCA1, PRR9, PRR5, and TOC1 promoter activity as measured with promoter:LUC fusions (see Note 9).

4

Notes 1. We buy our MS medium premixed from Phytotechnology Laboratories. 2. Luciferase activity measurement: There are chiefly two routes to high-throughput measurement of LUC activity: imaging via a low-light charge-coupled device (CCD) camera or activity measurements with a luminometer. In our lab we have relied on imaging with a Hamamatsu ORCA II ER CCD camera (C4742-98 ERG; Hamamatsu Photonics, Hamamatsu City,

Luciferase Rhythms

7

Japan, http://www.hamamatsu.com) or with a Topcount™ Microplate Scintillation Counter (Perkin Elmer) with six detectors. In general, the TopCount luminometer assay permits higher throughput analyses and so is attractive for forward genetic screens [41], although the TopCount does not image and so there is a loss of spatial resolution. Light from seedlings expressing high LUC activity can “contaminate” readings from adjacent wells containing seedlings expressing lower LUC activity. Accordingly, we use opaque (white or black) 96-well microtiter plates for our samples. For experiments in continuous light or in light–dark cycles, we interleave the sample plates with three clear microtiter plates to permit light penetration to the central wells of the sample plates (see Note 7). 3. Although we describe experiments with transgenic Arabidopsis seedlings, these protocols are readily adapted for use with protoplasts [42] or with tissue culture from other species, such as Brassica rapa [43]. 4. For experiments in continuous dark it is useful to add 1–2 % sucrose to the one-half-strength MS medium as this permits sustained rhythmicity. However, it is important to note that sucrose (3 %, or ~90 mM) affects circadian period in both continuous light and continuous dark as well as amplitude and rhythmic persistence in continuous dark [44]. 5. Because LUC protein is quite stable, it accumulates prior to introduction of the substrate luciferin and the introduction of luciferin results in a transient pulse of anomalously high light production that should be allowed to dissipate prior to measurement of de novo activity. We therefore initiate imaging typically 12–24 h after luciferin addition. With the TopCount we routinely entrain seedlings within 96-well plates to which luciferin has been added for one more entraining cycle before release into continuous conditions. 6. CCD cameras can be housed in growth chambers containing lights (LEDs are best) such that continuous light or light–dark cycles can be imposed. If the lights are computer or timer controlled, one can turn off the lights prior to imaging for automated operation. Alternatively, seedlings can be grown in a growth chamber under any desired light conditions and can be manually transferred into the dark camera chamber for imaging. Of course, this type of experiment is clearly not automated. 7. We typically entrain seedlings to a light–dark cycle imposed by fluorescent or LED lights affixed to the TopCount and controlled by a timer allowing imposition of appropriate light– dark cycles and continuous light or dark, and the TopCount is programmed to load each successive plate into the sample chamber for luminescence measurement. To permit light to reach the plates when loaded into the stacker, we have cut

8

C. Robertson McClung and Qiguang Xie

holes on all sides of the stacker without compromising its physical integrity. Between each sample plate, we interleave three clear 96-well plates to allow light to reach the seedlings; these clear plates must cycle through the sampling chamber but need not be read, which is signaled by not bar-coding them. Of course, interleaving with three clear plates reduces the capacity of the large 40-plate stacker to 11 sample plates containing 1,056 seedlings. A second potential limitation to this system is that light is not uniform across the plate, but in our experience this has only minimal effect on rhythmicity. 8. TopCounts may be purchased with 2 to 12 detectors—more detectors speed sampling proportionally. For a TopCount with 6 detectors, with a reading time for each well set to 5 s (we typically read 5 s per well, but have sampled from 1 to 10 s, depending on promoter strength), one full 96-well plate can be read in about 2 min, and a stack of 11 plates in constant light can be read in 90 min. Higher throughput, although with lower temporal resolution, can be achieved in continuous dark because it is not necessary to interleave the sample plates; many clock-regulated promoters, such as that of CAB2, are expressed at lower levels in the dark [22, 45], but there are notable exceptions, including COLD AND CIRCADIAN REGULATED 2 (CCR2, also known as GLYCINE-RICH RNA-BINDING PROTEIN 7, GRP7) [46] and CATALASE 3 (CAT3) [47]. Sucrose (1–2 %) is typically added to the medium for experiments in continuous dark because it increases amplitude and persistence of the rhythms in many genes, but see Note 4. If the LUC transgene is expressed at low levels, one can reduce background and enhance signal detection by extending the delay after the plate enters the sampling chamber before counting is initiated. This delay permits the background of seedling-delayed fluorescence (light emission following cessation of illumination) to diminish before initiating sampling. We typically use a delay of 1 min but have delayed for 3 min when measuring particularly weak promoters. This obviously extends sampling time and reduces sampling frequency. It should be noted that delayed fluorescence itself cycles with a circadian rhythm and has been developed as a useful nontransgenic assay to measure circadian rhythms [48]. 9. CLOCK GENE promoter:LUC fusion constructs: LUC+ (Promega, Madison, WI) is a modified cytoplasmic form of firefly luciferase that is 5–20-fold brighter than the native peroxisomal form initially used for plant clock research [22, 28, 41]. We cloned a promoterless LUC+ gene into pENTR1 (Invitrogen) to generate pENTR1A_luc+ (Fig. 3a). We then clone CLOCK GENE promoters into the multiple cloning site of pENTR1A_luc+ and then recombine these CLOCK GENE

Luciferase Rhythms

9

Fig. 3 Generating CLOCK GENE promoter:LUC fusion constructs. Promoters from the clock genes PRR9, PRR5, and TOC1were subcloned into a modified pENTR1 (Invitrogen) entry vector, pENTR1A_luc+ (left), and recombined into the binary vector pH2GW7 [49] from which the CaMV 35S promoter had been deleted (right) by LR recombination

promoter:LUC fusions into the binary vector pH2GW7 [49] from which the CaMV 35S promoter has been deleted (Fig. 3b). The resultant binary plasmids carrying the CLOCK GENE promoter:LUC fusions are transformed into Arabidopsis by floral dip [39] using Agrobacterium tumefaciens GV3101.

Acknowledgements This work was supported by grants from the National Science Foundation (IOS-0923752 and IOS-1025965) to C.R.M. References 1. Fukushima A, Kusano M, Nakamichi N et al (2009) Impact of clock-associated Arabidopsis pseudo-response regulators in metabolic coordination. Proc Natl Acad Sci U S A 106: 7251–7256 2. Robertson F, Skeffington A, Gardner M et al (2009) Interactions between circadian and hormonal signalling in plants. Plant Mol Biol 69:419–427 3. Graf A, Schlereth A, Stitt M et al (2010) Circadian control of carbohydrate availability

for growth in Arabidopsis plants at night. Proc Natl Acad Sci U S A 107:9458–9463 4. Pruneda-Paz JL, Kay SA (2010) An expanding universe of circadian networks in higher plants. Trends Plant Sci 15:259–265 5. McClung CR, Gutiérrez RA (2010) Network news: prime time for systems biology of the plant circadian clock. Curr Opin Genet Dev 20:588–598 6. Hotta CT, Gardner MJ, Hubbard KE et al (2007) Modulation of environmental

10

7. 8.

9. 10. 11.

12. 13.

14. 15.

16.

17.

18. 19.

20.

C. Robertson McClung and Qiguang Xie responses of plants by circadian clocks. Plant Cell Environ 30:333–349 McClung CR (2011) The genetics of plant clocks. Adv Genet 74:105–138 Covington MF, Maloof JN, Straume M et al (2008) Global transcriptome analysis reveals circadian regulation of key pathways in plant growth and development. Genome Biol 9:R130 Doherty CJ, Kay SA (2010) Circadian control of global gene expression patterns. Annu Rev Genet 44:419–444 Filichkin SA, Priest HD, Givan SA et al (2010) Genome-wide mapping of alternative splicing in Arabidopsis thaliana. Genome Res 20:45–58 Sanchez SE, Petrillo E, Beckwith EJ et al (2010) A methyl transferase links the circadian clock to the regulation of alternative splicing. Nature 468:112–116 Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22:R648–R657 Kloppstech K (1985) Diurnal and circadian rhythmicity in the expression of light-induced nuclear messenger RNAs. Planta 165: 502–506 Nagy F, Kay SA, Chua N-H (1988) A circadian clock regulates transcription of the wheat Cab-1 gene. Genes Dev 2:376–382 Paulsen H, Bogorad L (1988) Diurnal and circadian rhythms in the accumulation and synthesis of mRNA for the light-harvesting chlorophyll a/b-binding protein in tobacco. Plant Physiol 88:1104–1109 Tavladoraki P, Kloppstech K, ArgyroudiAkoyunoglou J (1989) Circadian rhythm in the expression of the mRNA coding for the apoprotein of the light-harvesting complex of photosystem II. Plant Physiol 90:665–672 Fejes E, Pay A, Kanevsky I et al (1990) A 268 bp upstream sequence mediates the circadian clock-regulated transcription of the wheat Cab-1 gene in transgenic plants. Plant Mol Biol 15:921–932 Millar AJ, Kay SA (1991) Circadian control of cab gene transcription and mRNA accumulation in Arabidopsis. Plant Cell 3:541–550 Pilgrim ML, McClung CR (1993) Differential involvement of the circadian clock in the expression of genes required for Ribulose-1,5bisphosphate carboxylase/oxygenase synthesis, assembly, and activation in Arabidopsis thaliana. Plant Physiol 103:553–564 de Wet JR, Wood KV, DeLuca M et al (1987) Firefly luciferase gene: structure and expression in mammalian cells. Mol Cell Biol 7:725–737

21. Ow DW, Wood KV, DeLuca M et al (1986) Transient and stable expression of the firefly luciferase gene in plant cells and transgenic plants. Science 234(4778):856–859 22. Millar AJ, Short SR, Chua N-H et al (1992) A novel circadian phenotype based on firefly luciferase expression in transgenic plants. Plant Cell 4:1075–1087 23. Ozawa T, Yoshimura H, Kim SB (2013) Advances in fluorescence and bioluminescence imaging. Anal Chem 85:590–609 24. Aflalo C (1991) Biologically localized firefly luciferase: a tool to study cellular processes. Int Rev Cytol 130:269–323 25. Vieira J, Pinto da Silva L, Esteves da Silva JCG (2012) Advances in the knowledge of light emission by firefly luciferin and oxyluciferin. J Photochem Photobiol B 117:33–39 26. Belas R, Mileham A, Cohn D et al (1982) Bacterial bioluminescence: isolation and expression of the luciferase genes from Vibrio harveyi. Science 218:791–793 27. Close D, Xu T, Smartt A et al (2012) The evolution of the bacterial luciferase gene cassette (lux) as a real-time bioreporter. Sensors 12:732–752 28. Millar AJ, Short SR, Hiratsuka K et al (1992) Firefly luciferase as a reporter of regulated gene expression in higher plants. Plant Mol Biol Rep 10:324–337 29. Liu Y, Golden SS, Kondo T et al (1995) Bacterial luciferase as a reporter of circadian gene expression in cyanobacteria. J Bacteriol 177:2080–2086 30. Kondo T, Strayer CA, Kulkarni RD et al (1993) Circadian rhythms in prokaryotes: luciferase as a reporter of circadian gene expression in cyanobacteria. Proc Natl Acad Sci U S A 90:5672–5676 31. Stanewsky R (2007) Analysis of rhythmic gene expression in adult Drosophila using the firefly luciferase reporter gene. Methods Mol Biol 362:131–142 32. Morgan LW, Greene AV, Bell-Pedersen D (2003) Circadian and light-induced expression of luciferase in Neurospora crassa. Fungal Genet Biol 38:327–332 33. Gooch VD, Mehra A, Larrondo LF et al (2008) Fully codon-optimized luciferase uncovers novel temperature characteristics of the Neurospora clock. Eukaryot Cell 7:28–37 34. 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 99:489–494 35. Welsh DK, Yoo S-H, Liu AC et al (2004) Bioluminescence imaging of individual

Luciferase Rhythms

36.

37.

38.

39.

40.

41. 42.

fibroblasts reveals persistent, independently phased circadian rhythms of clock gene expression. Curr Biol 14:2289–2295 Yamaguchi S, Mitsui S, Miyake S et al (2000) The 5′ upstream region of mPer1 gene contains two promoters and is responsible for circadian oscillation. Curr Biol 10:873–876 Yamazaki S, Numano R, Abe M et al (2000) Resetting central and peripheral circadian oscillators in transgenic rats. Science 288:682–685 Murashige TR, Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue culture. Physiol Plant 15:473–497 Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16:735–743 Plautz JD, Straume M, Stanewsky R et al (1997) Quantitative analysis of Drosophila period gene transcription in living animals. J Biol Rhythms 12:204–217 Southern MM, Millar AJ (2005) Circadian genetics in the model higher plant Arabidopsis thaliana. Methods Enzymol 393:23–35 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

11

43. Xu X, Xie Q, McClung CR (2010) Robust circadian rhythms of gene expression in Brassica rapa tissue culture. Plant Physiol 153:841–850 44. Dalchau N, Baek SJ, Briggs HM et al (2011) The circadian oscillator gene GIGANTEA mediates a long-term response of the Arabidopsis thaliana circadian clock to sucrose. Proc Natl Acad Sci U S A 108:5104–5109 45. Millar AJ, Kay SA (1996) Integration of circadian and phototransduction pathways in the network controlling CAB gene transcription in Arabidopsis. Proc Natl Acad Sci U S A 93:15491–15496 46. Strayer C, Oyama T, Schultz TF et al (2000) Cloning of the Arabidopsis clock gene TOC1, an autoregulatory response regulator homolog. Science 289:768–771 47. Michael TP, McClung CR (2002) Phasespecific circadian clock regulatory elements in Arabidopsis thaliana. Plant Physiol 130:627–638 48. Gould PD, Diaz P, Hogben C et al (2009) Delayed fluorescence as a universal tool for the measurement of circadian rhythms in higher plants. Plant J 58:893–901 49. Karimi M, Inzé D, Depicker A (2002) GATEWAY vectors for Agrobacteriummediated plant transformation. Trends Plant Sci 7:193–195

Chapter 2 Online Period Estimation and Determination of Rhythmicity in Circadian Data, Using the BioDare Data Infrastructure Anne Moore, Tomasz Zielinski, and Andrew J. Millar Abstract Circadian biology is a major area of research in many species. One of the key objectives of data analysis in this field is to quantify the rhythmic properties of the experimental data. Standalone software such as our earlier Biological Rhythm Analysis Software Suite (BRASS) is widely used. Different parts of the community have settled on different software packages, sometimes for historical reasons. Recent advances in experimental techniques and available computing power have led to an almost exponential growth in the size of the experimental data sets being generated. This, together with the trend towards multinational, multidisciplinary projects and public data dissemination, has led to a requirement to be able to store and share these large data sets. BioDare (Biological Data repository) is an online system which encompasses data storage, data sharing, and processing and analysis. This chapter outlines the description of an experiment for BioDare, how to upload and share the experiment and associated data, and how to process and analyze the data. Functions of BRASS that are not supported in BioDare are also briefly summarized. Key words Circadian clock, Period estimation, Data repository, Data sharing

1

Introduction The field of chronobiology has been in existence since the eighteenth century and has been a continuously active area of research since the 1950s. Circadian clocks have been identified across both prokaryotes and eukaryotes in organisms ranging from cyanobacteria [1, 2] through mammals [3, 4]. They control myriad different processes, many of which are only now being elucidated. In order to improve understanding of clock mechanisms and their significance, models of the circadian clock have been developed for many organisms and these models tested through both experimentation and simulation [5]. A key aspect of data analysis is to determine whether a time series (derived either from experimental data or from simulations) is rhythmic and, if so, to determine the underlying period to a sufficient level of accuracy. There are many techniques and algorithms for doing this [6–9], all with

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_2, © Springer Science+Business Media New York 2014

13

14

Anne Moore et al.

different assumptions and of differing levels of complexity. Some of these algorithms are available as part of software packages, either proprietary with an associated charge, or free online, for example from labs which have developed them for their own needs. Choosing which algorithm, which implementation and which measures of accuracy to use to determine rhythmicity and the identification of an associated period are all time-consuming, and can offer many pitfalls, especially for the nonexpert. Furthermore, recent advances in technology and computing power have led to an almost exponential growth in the complexity of experiments and the size of the resulting data sets. Molecular genetic experiments using reporter genes such as luciferase can now routinely generate total volumes of data that were once the preserve of animal activity monitors or electrophysiological readings. However, these molecular data are relatively short, sparse, and noisy from many samples, in excess of 200 time series in one experiment [10]. As another example, delayed fluorescence experiments [11] can produce at least 500 time series each containing at least 120 time points. During an individual project or grant many such data sets will be generated and there is a need to keep the data both safely, for instance backed up, and correctly curated; such data curation is a broad challenge for biology. Another trend is towards multi-lab and multidisciplinary projects. A key requirement for such projects is data sharing among collaborators. Here version control and backed up data is crucial for the successful outcome of the project. Research funders increasingly demand public access to the resulting data: this must take minimal effort from the researcher, or it tends not to happen. To this end BioDare (Biological Data Repository) has been developed under the ROBuST project [12] to facilitate data exchange in this multisite project. BioDare is an online system which allows data sharing (including public dissemination), data processing and analysis, with the main focus on time series data produced in circadian experiments from various model species. One of the most important aspects of the data processing capability is its period analysis facility and in this respect it can be regarded as the successor to the Biological Rhythms Analysis Software Suite (BRASS), the Visual Basic embedded application used to analyze circadian time series data, developed by Paul E. Brown in the Millar group [7, 13]. BRASS development [7, 14] was inspired by the Steve Kay lab’s I & A, an earlier suite of Excel macros. Both suites helped users to analyze free-running circadian rhythms with the FastFourier Transform-Non Linear Least Squares algorithm [6]. BRASS v3.0 introduced David Rand’s mFourfit analysis for entrained rhythms [13], and is available online from www.amillar.org. The FFT-NLLS software is also available with this distribution. However, these packages were Windows-specific; BRASS was

Circadian Period Estimation and Rhythmicity

15

developed using mainly Excel 2000 and Windows 2000. Changing Windows or Excel versions regularly disrupted their operation, though usually this could be corrected by the user. BioDare is much more widely compatible and offers a larger range of services. A small number of BRASS functions not yet supported in BioDare are described in Subheading 3.16. The main features of BioDare are as follows:

2

–

Repository for experimental data accompanied by extensive metadata including details about environmental conditions and biological material used.

–

Rhythm analysis and period estimation using four different algorithms.

–

Generation of secondary data (normalized, detrended, averaged …).

–

Graphical output of data, secondary data and rhythm analysis.

–

Simple text-based search throughout metadata.

–

Biology- and conditions-aware search for data.

–

Data aggregation and export.

Materials 1. Access to a Web browser (there are currently no known issues of compatibility with any browsers). 2. Access to BioDare (login available from biodare@staffmail. ed.ac.uk, or use the “public” account) (see Note 1). 3. Pedro Application (contact [email protected] for details of how to obtain this). This requires Java version 1.5 or higher. 4. Numerical data file, typically an .xls file in a tabular form, with one time column followed by columns of data reads (see Note 2). 5. For Subheading 3.16, the BRASS v3.0 application, available from http://www.amillar.org/PEBrown/BRASS/BrassPage. htm for MS Windows and Excel (see Note 3).

3

Methods The basic entity in BioDare is an experiment. Although a time series is the basic data unit to be processed and analyzed, many time series are typically collected in parallel in contemporary circadian research. The behavior of the circadian system is greatly influenced by the environmental conditions under which the data were

16

Anne Moore et al.

generated, which typically vary among experiments. Thus it is important to store in BioDare not only the time series data but also the associated experimental conditions. To ensure that the “experiment” which is captured and uploaded to BioDare is complete and comprehensive, BioDare provides metadata templates for different types of experiment (see Note 4). Processing circadian data with BioDare is a three-stage process: 1. Describing the experiment and the associated data using Pedro (producing metadata). 2. Creating the new experiment in BioDare using the produced metadata, and associating the time series data. 3. Processing and analyzing time series data with BioDare. Before giving more detail for each of these stages it is necessary to summarize some conventions used in this methods section and to describe the “Display Experiment” page which can be regarded as the launch pad for many of the activities and methods which follow the creation of a new experiment in BioDare and so are relevant to Subheading 3.5 and beyond. Whenever possible, methods are illustrated here using publicly accessible data that users could access from the “public” or “demo” accounts. 3.1

Conventions

●

●

3.2 “Display Experiment” Screen

Text in quotes, e.g., “Display Experiment,” refers to titles of screens or sections of screens in BioDare. Text in bold and in quotation marks, e.g., “Add experiment,” refers to comments, links or menu options found on BioDare screens, which trigger specific actions.

The “Display Experiment” screen is the launch pad for most of the activities undertaken in BioDare. It is available once an experiment has been uploaded (see Subheading 3.2, below) or whenever the experiment has been selected from results of a search. When performing experiment-related operations, it is possible to return to the “Display Experiment” screen from other screens by clicking on “Return” link at various points on the screens. The “Display Experiment” screen is shown in Fig. 1 and comprises several different sections. The Top section “General information” provides an overall summary of the experiment of interest. The fields in this section are not directly editable and are populated from the metadata file generated by Pedro (see Subheading 3.3 below). This “General information” section is also repeated at the top of all other BioDare pages. The second section of the screen “Actions” is the main control center for BioDare. It lists the key actions which can be undertaken once an experiment has been created (see Note 5). To undertake one of the actions it is necessary to click on the green text, e.g., “update.”

Circadian Period Estimation and Rhythmicity

17

Fig. 1 The “Display experiment” screen showing the “General information” section, the “Actions section,” and subsequent sections with detailed descriptions of the provenance and growth conditions. A “show arrow” is highlighted; this can be used to expand the hidden sections. On the left hand side Quick links to other areas of BioDare are shown

Beneath the Actions section there are several sections with detailed descriptions of the provenance and growth conditions underlying the experiment. The detail in each of these sections can be seen by clicking on the “show arrow” label, see Fig. 1. Note that this is not directly editable and is populated from the metadata file generated by Pedro. Finally the bottom section of the “Display experiment” screen lists all the files associated with that particular experiment. Clicking on the show arrow next to each filename lists the versions of that file which are available and clicking on “Download” allows the files to be downloaded, see Subheading 3.12 for more detail.

18

Anne Moore et al.

3.3 Using Pedro to Describing and Export the Experiment

The experimental description (metadata) is in XML. XML is a flexible, machine-readable format. Although the metadata are held as text, the XML is not conveniently human-readable. Pedro provides concise and customized metadata creation, based on the XML schema [15]. 1. Start the Pedro application (see Note 2) and click on “make new experiment.” 2. Enter the relevant experiment description (see Note 6). 3. Go to “samples” in Pedro and click “fill-up” (see Note 7). 4. Describe the samples according to the Excel file of raw data: e.g., first sample in first column, second sample in second column, etc. 5. Save the Pedro file. 6. Chose “View/Show Errors” to check if all mandatory information has been provided. 7. Once the file is error-free export the experiment using “File/ Export to final submission format.” This will generate the .xml file with the metadata for the experiment.

3.4 Creation of a New Experiment in BioDare

1. In your Web browser navigate to the BioDare home page (www.biodare.ed.ac.uk) and login (see Note 8). 2. Click “Add experiment,” this will load the page shown in Fig. 2. 3. Populate the “Description file” field by navigating to the Pedro output file to be uploaded; this will be in an XML format (see Note 9). Once the experiment has been uploaded the “Display experiment” screen will appear with the message “New experiment has been successfully added to the repository” at the top, see Fig. 2. 4. Add the numerical data by clicking on “add raw data” in the Actions section. The “Append raw data” screen will appear, see Fig. 3. Populate the “raw data” field by navigating to the data file to be uploaded; this will typically be an .xls file. Depending on the experiment type and data format some additional information may be required in order to import the data. Typically it will be necessary to describe the location of the data in the file (i.e., first data row, first data column), or to give conversion parameters (for example the interval between pictures in imaging experiments). Once this has been done clicking on “attach” will result in the “Display experiment” screen being displayed with the message “Raw data file added successfully” displayed near the top. 5. To add any additional files associated with the experiment (see Note 10), click on “add files” in the “Actions” section. This will result in the “Append user files” screen being displayed, see Fig. 4. Click on “Browse” in the “Attach new user files” section of this screen and navigate to the file to be uploaded. When all

Circadian Period Estimation and Rhythmicity

19

Fig. 2 The “Add new experiment” screen after a new experiment has been successfully added. Inset shows the description file field to be populated prior to adding the experiment

files to be attached have been selected click on attach. This will refresh the “Append user files” screen and the message “Files appended successfully” will appear at the top of the screen (see Note 11). 3.5

Granting Access

This allows the user to specify which other user groups (including “public”) can view and write to the experiment, (see Note 12). 1. Click on “grant” which opens the “Grant access to experiment” screen, see Fig. 5. 2. Use the “Set access to the experiment section” of this screen to specify who can read the experiment and who can write to it (see Note 13). The read or write permissions are set by selecting appropriate tick box next to group name for which permission should be granted.

20

Anne Moore et al.

Fig. 3 The “Append raw data” screen showing the additional information required in order to import the data

3. Once the desired boxes have been selected, or deselected, click on “Grant” to refresh the “Grant access to experiment” screen. The message “Access permissions updated successfully” will appear near the top. 3.6

Removing Data

Sometimes it is desirable to remove time series before further analysis (see Note 14). 1. Click on “remove samples” which will open the “Removing data” screen, see Fig. 6 (see Note 15). 2. Select the preferred way of presenting the data. Typically in order to spot irregularities, it is easier to look at secondary data (for example normalized) than raw data. 3. Select any time series to be removed from the analysis by checking the checkbox next to their labels. When all time series to be removed have been checked click on the box “Remove selected” at the bottom of the screen. 4. The selected data to be deleted are displayed once more, and final confirmation is done by clicking on “Remove selected.” The “Display experiment” screen will now be displayed, the top of which will record how many samples have been removed.

Circadian Period Estimation and Rhythmicity

21

Fig. 4 The “Append user files” screen showing the “Attach new user files” section and the sections where raw data and other files can be downloaded 3.7 Display Experimental Data

1. Click on “display.” The “Display experimental data” screen is shown in Fig. 7, this shows plots of the experimental data selected using the default criteria. The graphs are presented in the bottom section of the screen after the title “Graphs (N in total).” 2. The selection of data that are displayed and how they are grouped together can be modified by using “Change display criteria.” The data can be selected, or deselected, according to genotype, marker and line. The type of data to be plotted, for example, raw data, averaged data or normalized data (see Note 16), can be selected from the pull-down menu. The number of traces per plot can be selected from the pull-down menu and there is also the option to add error bars to the data. The data can be grouped,

22

Anne Moore et al.

Fig. 5 The “Grant access to experiment” screen showing the read and write options for different groups

and hence plotted, by genotype, marker, genotype–marker, or genotype–marker and line. The experimental conditions can be selected according to code or name (see Note 7). Depending on which of these is selected there will be tick boxes for the experimental conditions to plot. Once the modifications to the display criteria have been made clicking on “Replot” data will result in the data being displayed according to the modified criteria. The expandable section, “Data series details,” lists the details of the series which have been plotted together with their experimental conditions. In this section it is also possible to export the numerical values of all the plotted time series to a csv (commaseparated values, see Note 17) file. This can be achieved by clicking on “Export data” or “Export data&errors,” depending whether the exported data should contain values of standard errors or not (n.b. errors are only applicable to averaged data, that is data which have been grouped based on their biological origin and experimental conditions and then averaged over that group, or biological replicate).

Circadian Period Estimation and Rhythmicity

23

Fig. 6 The “Removing data” screen, inset shows the Change display criteria section where different preprocessing regimes can be applied

3.8

Period Analysis

1. Click on “Period analysis:display” which opens the “Period analysis” screen. From there click on “Start new analysis” which is located below the General Information Section. This will open the “New period analysis” screen, see Fig. 8. 2. Populate the Data Window “from” and “to” fields to specify the time limits between which the time series will be analyzed (see Notes 18 and 19). 3. Populate the Expected period “from” and “to” fields. The default values are 18 and 30, respectively and signify the range of periods which are considered to be circadian or to be of interest to the experimentalist (see Note 20).

24

Anne Moore et al.

Fig. 7 The “Display experimental data” screen. The insets show the selection of data that are displayed, the type of secondary data to be used, and how the data can be grouped together

4. Select a method from the pull-down Method menu (see Note 21). 5. Select the type of input data (see Note 22). 6. Click on the “analyse” box at the bottom of the panel. This will result in the appearance of the “Period analysis” screen, see Fig. 9. 3.9 Results of Period Analysis

1. The second section of the “Period analysis” screen, “Set display parameters” allows the user to choose certain display parameters. If “Individual results” is selected it is possible to see all the different potential periods for each trace (at the bottom of the screen). The next two fields determine the way

Circadian Period Estimation and Rhythmicity

25

Fig. 8 The “Start new period analysis” screen showing the Data window and Expected period fields which should be populated with user selected values

phase is reported, it allows the user to select the phase estimation method and whether the phase should be related to time zero or to the start of the analysis window (see Note 23). The checkbox ‘weighted by GOF’ determines if the statistics summary should include averaged values or weighted by results GOF values. After changing the display setting, the “refresh” button must be pressed. 2. The third section of the “Period analysis” screen, “N jobs submitted, m still queuing,” lists the number of analyses (or analysis jobs) submitted for this experiment. Clicking on the

26

Anne Moore et al.

Fig. 9 (a) The “Period analysis” screen showing the form to adjust display properties, number, and status of the analyses submitted for this experiment and the Refresh link used to check if a job has been completed and the link to edit the results which are included in the statistical analysis. (b) The “Period analysis” screen (bottom) showing the summary statistics from each analysis job for all traces grouped by their genotypes (see Note 26) plus the collection of individual results

show arrow shows the history of period analysis carried out on this experiment with the most recent at the bottom of the list (see Note 24). This screen does not refresh automatically, so it is necessary to click on “refresh” to check if a job has been completed. There is also the option “Clean failed jobs” which removes any failed analyses from the list. 3. If not all results could be automatically included in the summary results then a section of text may appear in red, “N jobs

Circadian Period Estimation and Rhythmicity

27

Fig. 9 (continued)

need manual editing” (see Note 25), this is described more fully in Subheading 3.10 below. 4. Below the job status there is the “Analysis statistics” section, which shows the summary statistics from each analysis job for all traces grouped by their genotypes (see Note 26). It lists the results of all period analysis for individual data traces, and is composed of period, phase (see Note 23), amplitude, Goodness of Fit (GOF) (see Note 27), and global error (ERR) for each trace (see Note 28). It also identifies how many samples were not included, gives the option to “select periods manually” and to “show goodness of fit plots,” see Fig. 10.

28

Anne Moore et al.

Fig. 10 An example of a secondary data versus an original curve. (a) Shows a typical plot of raw data from a luciferase experiment where interpretation is difficult due to the presence of trends and traces of different orders of magnitude. (b) Shows a plot for two traces from the same experiment now constructed using secondary data (normalized, detrended, and averaged over replicates). This clearly shows a long period phenotype for the mutant (blue trace) compared to the wild type (red trace). (c) An individual data trace and its theoretical fit derived from the FFT-NLLS algorithm. (d) Shows the Goodness of Fit (GOF) plot for the same experiment, again clearly showing long period phenotype for mutant (red) compared to the wild type (blue ) (color figure online)

5. To export the results click on the show arrow next to “export results” and then click on export data. This will export a zip file containing Excel spreadsheets (see Note 29). 6. If ‘individual results’ is selected the last section contains results for individual results, including period, phase, amplitude, GOF and global error. The entry rows in Italics highlight results which were ignored by the user and not included in summary statistics. The entries in red are those that were not automatically included in statistics and require user decision. The failed analysis will be marked with failed with a description of the possible cause. It is also possible to see graphically the fit of the model (based on the proposed period) against the original data. Clicking on “edit” gives a list of the periods selected for each time series and allows the user to “ignore” or remove individual traces as required. As well as the individual fits it is also possible to view the “combined fit” which shows the fits from all analyses carried out so far. 3.10 Ambiguous Data and Selection of Correct Period

1. If the message “N jobs need manual editing” appears, clicking on the show arrow shows the job number and gives the number of samples where multiple periods have been identified or the periods found are outside the defined circadian range and which need manual selection. Clicking on “select” opens the “Edit analysis results” window. The second section repeats details of the job number and the number of samples where multiple periods have been identified. Those time series which yielded a period outside the specified circadian range or have

Circadian Period Estimation and Rhythmicity

29

multiple periods will be listed first (in red). The proposed period will be given together with the associated phase, amplitude, goodness of fit and error. It is possible to view the period analysis fit graph which shows the original time series, the fit of the best model generated by the algorithm and the cosines of the proposed periods. Once the period to be manually selected have either been selected or ignored, clicking on the “select” at the bottom of the screen will apply the changes. NB all periods are listed, even those which do not require manual selection. Thus it is also possible to deselect data series which have been automatically selected. Once “select” has been clicked the screen will refresh and the “Result of PPA analysis” screen will reappear. If the algorithm used was FFT-NLLS then the biological data from any ambiguous time series are shown together with the possible periods that can be assigned to those time series. One of the periods can be selected, or all period can be dismissed (see Note 30). As with the other methods, the samples which were not already included in the statistics are listed first, followed by those that were included in the analysis. 3.11 Display GOF Plots

Scatter plots of the period and the GOF of each time series conveniently summarize the circadian properties of a group of samples. GOF plots group biological replicates on one plot, and for each individual biological sample the period against GOF value is plotted. For robust period estimates, this will result in points grouped in area cluster surrounding the mean period and GOF. In other case the points may be spread over wide range of periods indicating huge biological variance or bad period estimates, also points with higher GOF values should be regarded as less important that those with smaller values. Note that GOF graphs can only be plotted after period analysis has been carried out. 1. Click on “GOF Graphs: display,” this will result in the appearance of the “Display GOF Graphs” screen. 2. Select which period analysis to use to generate the GOF graphs, see Fig. 10d, by clicking on “Show GOF” next to the job of interest in the “Select analysis with results to plot” section of the screen (see Note 31).

3.12 Exporting Data and Results

It is possible to download files, data and results for any experiment for which the user has been granted read access. 1. To export all files associated with an experiment go to the bottom of the “Display experiment” screen, see Fig. 4. 2. Click on the show arrow to list the available versions of the file of interest. The desired version can then be downloaded by clicking on “download.” 3. To export analysis results see Subheading 3.9, step 6.

30

Anne Moore et al.

Fig. 11 The “Search experiments” screen. The top section allows users to search for a specific phrase either in the title, in the samples or anywhere within a file. The search can be refined using “Advanced filters” shown in the middle section. The bottom of the screen shows the links to the experiments which match the search conditions

3.13 Finding Experiments

The “Find Experiment” link in the “Quick links” on the left hand side of the page can be used to search the text in the database for experiments (see Note 32). 1. Click on “Find Experiment” to go to the “Search experiments” screen, see Fig. 11 (see Note 33). 2. The top section of this screen allows the user to search either in the title, in the samples or anywhere within a file for a search phrase entered by the user. Enter the text string of interest and

Circadian Period Estimation and Rhythmicity

31

click on “Search.” This will refresh the “Search experiments” screen and display the results. Typing * as the search phrase yields all visible experiments. 3. The search can be refined using “Advanced filters.” Click on the show arrow to reveal the filter options. The search can be filtered by owner, temperature, light or chemicals. When a particular filter is selected either a drop down menu is activated, for “filter by owner,” or subcategories are activated, for “filter by temperature” or “filter by light.” If “filter by chemicals” is checked the chemical name must be entered in the search box. 4. When the correct search filters have been selected click on “Search.” This will result in the “Search experiments” screen being refreshed and the refined list of experiments being displayed. 3.14

Searching Data

The Search data function can be used to search and aggregate time series data throughout all visible experiments (see Note 32). The functionality used is very similar to that used on the “Display experiment” screen. 1. Access the “Search data” function either by clicking on “Search data” from the login landing page or by using the “Search data” Quick Link on the left hand side of any page. This will result in the “Search data” screen being displayed. 2. In the first screen select the desired genotypes and click on “Select genotype.” This will result in a second selection screen being displayed. 3. On the second screen select the markers of interest and click on “Select markers.” This will result in a third selection screen being displayed. 4. On the third screen select the desire lines, choose how the data should be displayed and specify how the data should be grouped together. The type of data to be plotted, for example raw data, averaged data or normalized data, can be selected from the pull-down menu. The number of traces per plot can also be selected from a pull-down menu and there is also the option to add error bars to the data. 5. You can further narrow your search by limiting data to those gathered during specific experimental conditions or originating from particular experiments. The filtering is done by ticking boxes next to the condition labels or experiment names. 6. Once all the options have been selected click on “Plot data.” This will result in graphs of the retrieved time series being plotted. 7. The “Data series details” section contains information about all the traces plotted on the graphs with links to the experiments from which the particular trace originated.

32

3.15

Anne Moore et al.

Sharing Data

In addition to controlled access, which can vary from the individual group to the public (see Subheading 3.5), BioDare generates permanent hyperlinks (URLs) that unambiguously locate specific experiments or files. 1. To share an experiment go to the “Display experiment” screen of the experiment of interest and copy the whole address from the navigation bar of the Web browser. This hyperlink can then be sent to the person/people with whom the user wishes to share the experiment. 2. To share a file associated with an experiment scroll to the bottom of the “Display experiment” screen of the experiment of interest to the section where the attached files are listed. Right click on the download link next to the file of interest and select Copy link location or the equivalent from your Web browser’s context menu. This hyperlink can then be sent to the person/ people with whom the user wishes to share the file(s). 3. Users must log on to BioDare to access the information. So even if the link is sent, only users who are granted read permission will be able to access the experiment or files, so BioDare links are a secure way of exchanging information.

3.16 Distinctive Functions of the Legacy BRASS Package

BRASS v3.0 features in two areas are not yet reproduced in BioDare, because they were optional features for a minority of experiments. In data import, direct import of data files from a range of hardware is supported in BRASS, including the NightOwl camera system and TopCount scintillation counter. Support is planned in BioDare. However, any means to produce a set of time-stamped, contiguous data series (for example using BRASS) will yield data that can readily be imported to BioDare. In data analysis, BRASS offers two functions that depend upon real-time interaction with many or all of the time series in a data set. Communication lags make these features hard to reproduce satisfactorily online, so they are not offered in BioDare. 1. Customisation of data windows for each time series (BRASS v3.0 manual, pages 29–32). In BioDare, the same data window is analyzed for all time series. When samples produce optimal rhythmic signals at different times within the experiment, there may not be any single data window that is optimal to analyze for all time series. Arabidopsis leaf movement studies, for example, only produce rhythms in a developmental interval when that leaf’s petiole is elongating. The many leaves in an experiment give optimal rhythms at different times. (a) After starting an FFT-NLLS or mFourfit analysis, choose the “Select Best ZT Window for each” option.

Circadian Period Estimation and Rhythmicity

33

(b) The entire time series will be plotted for each sample, and the user will be able to select a data window from the graphs individually. The selected time window will be highlighted in red, and the user will be asked to confirm the selection. (c) Change the times in the Lower and Upper fields to tune the start and end time points of the data window, and click Include to record the window and advance to the next time series, or Skip to exclude that time series. This procedure allows customized analysis of moderately large sets of time series, for example in Edwards et al. [16] and Salathia et al. [17]. 2. Phase estimation by semi-automated peak-picking (see BRASS v3.0 manual, page 51–54). When the duration of individual time series is very limited, visual peak identification may be the only viable approach to analyze the data. BRASS automated the recording of peak times. (a) Select Phase Values ->Manual Selection Analysis section of the BRASSv3.0 menu, and select the worksheet to analyze. (b) The peak identification routine graphs a three-point moving average of the data within the specified data window and highlights in red the largest signal value that is not at a window border. (c) Use the Higher/Lower buttons to adjust the time point selected as the peak, if required. (d) Use the Accept button to record the peak time point and advance to the next time series, or the Skip button to move to the next series and record ‘Skipped’ instead of a peak time. This procedure allows rapid inspection of moderately large sets of short time series data, for example in the acute light responses or single peaks of CAB2 expression in DD, as in McWatters et al. [18]. The compatibility issues of BRASS’s Visual Basic for Applications undoubtedly limit the appeal of this package. However, the Brass v3.0 manual explains in detail which Excel Visual Basic libraries and (for mFourfit only) Matlab Runtime Libraries are required. In our experience, ensuring that these are available should allow some or all of BRASS v3.0 functions to run with many versions of Windows and Excel. The Matlab component runtime library for mFourfit analysis is an exception, as this absolutely requires a 32-bit operating system. BioDare is greatly superior.

34

Anne Moore et al.

3.17 Future Developments

4

As mentioned previously, BRASS already has compatibility issues, and it is not our intention to continue to support it. Within BioDare there is a series of planned upgrades. These include, but are not limited to: implementation of additional period estimation algorithms; comprehensive phase output for all period estimation algorithms; modification of period estimation algorithms to allow the analysis of ultradian data; improved feedback to user on status of processing; additional secondary data (e.g., different detrending options).

Notes 1. It is possible to access BioDare without an individual account and many data sets in BioDare are available to the public, to access them login into BioDare using login: public and password: public. Also a demo account can be used to run an analysis (credentials demo/demo); however, this account is available to anyone and so should not be used for confidential data. 2. The numerical data format of the input file can be customized to the needs of a particular group to match the outputs from their specific experiments. 3. The BRASS manual in the distribution file describes the Windows, Excel and Matlab libraries or add-ons required. BRASS development ended in 2006 and compatibility with more recent software may be reduced. 4. Pedro can create entry forms based on a data model in an XML schema that defines a particular group’s biological experiments constrained to controlled vocabularies. BioDare administrators can customize forms that cater for the needs of particular research groups and types of experiments. This generally speeds up data entry and experiment description. 5. The Actions which are displayed in the “Display Experiment” screen will vary depending on whether a user has read access or read and write access. 6. Within Pedro there are a few important things to remember: Bold font in fields labels mark mandatory fields; an asterisk (*) next to a label indicates the presence of an option list for that parameter, the list of choices is opened by pressing right click (cmd + click on mac) on the label; do not use the simple forms for temperature and light settings or describing samples, look for the “Fill” or “FillUp” button and choose appropriates settings from the dialogue box; Growth and experimental conditions must be described before describing the biological samples.

Circadian Period Estimation and Rhythmicity

35

7. The current genotypes and conditions were developed for plant circadian experiments, describing the genetic background, genotype, the reporter (“marker”), assay, and the independent transgenic line (where relevant, otherwise “unknown”). Conditions describe the detailed light or temperature cycles and are sometimes summarized as “codes.” 8. Logins need to be requested from the BioDare support team (e-mail: [email protected]). Be aware that passwords are case sensitive. 9. When creating a new experiment a common mistake is to upload the Pedro temporary file, i.e., the .pdx file instead of the .xml file produced during export to final submission format. 10. Additional files uploaded to an experiment will be stored along with the experiment in their native form but will not be analyzed by BioDare. 11. When uploading files only three files can be appended at a time, but more than three files in total can be appended by repeating the “Append user files” operation several times. 12. The person who uploads an experiment becomes its owner. The owner has full control over the experiment and the associated files and data. The owner’s supervisor can always see the experiment and associated data and can also change ownership of the experiment and assign it to someone else (this can be useful for continuity if a user leaves a lab). 13. Granting read permission to a group will mean that all members of that group can access this experiment description, visualize the data, look at period analysis or download any attached files. Granting write permission will allow members of that group to update the experiment, modify its attachments, remove samples from the raw data or run new period analysis. There is also an additional group; “public.” All users of BioDare are members of the group public, and hence, granting read permission to this group allows all users of BioDare to see those data. 14. Sometimes time series have been included in an experiment which may contain unwanted artifacts, for example due to a temporary equipment failure or due to a slow but consistent migration in one of the environmental conditions. To avoid misleading results which could possibly arise from erroneous data, it can be beneficial to remove such time series. 15. When considering which time series to remove it can be helpful to look at the data displayed after different preprocessing regimes have been applied. For example in the left hand side of Fig. 12, where the raw data are plotted, all data look equally rhythmic/arrhythmic and there would be no justification for

36

Anne Moore et al.

Fig. 12 Representative time series viewed as raw data and as detrended normalized data. By viewing the secondary data (in this case normalized detrended data), it becomes apparent that there was a technical problem in the generation of time series 296 (blue) and this series should be omitted from further analysis

excluding any of the time series. However, if the same data are plotted using the “Detrended normalized” option the figure on the right hand side of Fig. 12 is obtained. Here it becomes clear that it would be beneficial to omit series 296 from the analysis as there is an underlying trend which will mask any periodicity within the data. 16. The available data sets are: raw data—the original raw data that were uploaded, n.b. it is possible that these were preprocessed (for example have background noise removed) prior to uploading; (linear) detrended raw—the detrended raw data; normalized raw—normalized raw data; smoothed raw—smoothed raw data; averaged raw—averaged raw data (group by biological replicas and averaged); averaged raw (no lines)—averaged raw data but the information about different lines is ignored; detrended normalized—first data are normalized to the mean, then detrended; smoothed normalized—first data are normalized to the mean, then smoothed; averaged normalized—first data are normalized, then averaged over biological replicas; averaged normalized (no lines)—first data are normalized, then averaged over biological replicas but ignoring differences in lines; smoothed detrended—first data are detrended, then smoothed; averaged detrended—first data are detrended, then averaged over biological replicates; averaged detrended (no lines)—first data are detrended, then averaged over biological replicates but ignoring differences in lines; smoothed det.

Circadian Period Estimation and Rhythmicity

37

norm.—first data are normalized to the mean, then detrended and then smoothed; averaged det. norm.—first data are normalized to the mean, then detrended and then averaged over biological replicates; averaged det. norm. (no lines)—first data are normalized to the mean, then detrended and then averaged over biological replicates but the information about different lines is ignored. 17. A comma-separated values (csv) file stores tabular data (numbers and text) in plain-text form. A csv file consists of a number of records, separated by line breaks of some kind; each record consists of fields, separated by some other character or string, most commonly a comma or tab. 18. When populating the Data Window “from” and “to” fields prior to carrying out the period analysis if either field is left blank or set to zero it will default to the minimum and maximum time values, respectively of the time series under consideration. 19. Increasing the number of data-points usually leads to an improved period estimate with reduced standard errors. There are, of course, exceptions to this. In some cases, data are gathered during entrainment and usually it is desirable to exclude this part of data from analysis or to process these data separately to the data obtained under free running conditions. As an example samples could be subject to an initial period of 14 h light–14 h dark cycles (14:14) and then be transferred to constant light conditions. Data from the whole duration of the experiment would then take the form shown in Fig. 13. In this case two period analyses should be run; one with the data window set to be 0-84, to capture the first 3 sets of light–dark cycles and a second with the window set to be 96-0 (which is interpreted as max), here the first 12 h after the last dark to light switch have been omitted. The results from the initial 14:14 long days would be used to examine the phase responses, e.g., the acute light response at dawn and dusk, whereas the results from the later 3 days would be used to examine the free-running behavior of the different genetic mutants. 20. The time series data processing and analysis currently implemented within BioDare was designed primarily for circadian data. Therefore, we do not currently recommend using the period analysis tools for time series where the underlying period is expected to be 8 h or less. 21. The periodogram is one of the older techniques used for period estimation. The Enright Periodogram [8] steps through a series of proposed periods. For each proposed period, data are cut into period-length chunks of which are aligned 'on top of

38

Anne Moore et al.

Fig. 13 Representative time-series from an experiment running over 6 days. Initially the plants were in 14:14 (28 h) conditions. After 3 days the plants were transferred to constant darkness conditions. This illustrates conditions in which it would be beneficial to run two period analyses, one with the data window set to be 0-84, to capture the first 3 sets of light–dark cycles and a second omitting the first 12 h after the last dark to light switch with the window set to be 96-0 (which is interpreted as max)

each other' and averaged. This yields an averaged, model waveform. The period corresponding to the waveform with the biggest amplitude is then selected. One fundamental problem with Enright’s periodogram is that it does not include a method of testing the significance of any peaks. So a further modification of this approach by Sokolove and Bushell [19] introduces a measure for the period uncertainty. mFourfit was originally developed specifically for the analysis of rhythms under entrained conditions and will always return a value for the period, even if the underlying data are arrhythmic. It tries to fit one, model waveform of period-length to the whole time series and selects the period for which the best fit was obtained.

Circadian Period Estimation and Rhythmicity

39

By contrast, FFT-NLLS is not constrained to a single shaped waveform for the duration of the data, but can reproduce complicated shapes. Indeed, it was developed for the examination of free-running systems which tend to change their characteristics over the time. It is based on the principle of trying to fit a series of cos functions to reproduce the shape of the data shape, selecting the period one of a single cos in the circadian range. Spectrum resampling [20] was developed to improve period estimation in the case of non-sinusoidal data. It is based on the knowledge that any periodic signal, even if it does not comprise sinusoidal signals, will produce a spectrum with a dominant peak at the frequency corresponding to the period of the inputs signal. A spectrum is generated and its deviation from the smoothed spectrum is calculated. Boot-strapping is then applied to estimate the optimal spectrum and also to give confidence levels on the period estimate. As with mFourfit, spectrum resampling will always return a period, even if the underlying data are arrhythmic. Given the low processing times for all the algorithms our recommendation is to use at least two algorithms (we recommend FFT-NLLS and mFourfit as a minimum) to analyze data series and then to compare the output periods, phases and associated standard deviations. A more detailed assessment of the performance of the four period estimation algorithms is provided in the following reference [21]. 22. The available data sets are as follows: raw data—the original raw data that were uploaded, n.b. it is possible that these were preprocessed (for example have background noise removed) prior to uploading and data with different detrending applied (for example linear detrending, amplitude detrending or baseline detrending) Amplitude detrending, as the name suggests, removes any trends in amplitude whilst allowing the baseline to follow the experimental data. Baseline detrending can also be applied to remove any drift in the baseline and both amplitude and baseline detrending can be applied. Applying detrending can introduce substantially more ambiguities into the period estimation. This is particularly true if there is an overall damping effect. If amplitude detrending removes this effect then small/low level oscillations become more apparent and may lead to period estimates which do not reflect the original data. This is illustrated in Fig. 14. Also if there is substantial damping towards the end of the time series, the amplitude detrending algorithm can introduce spurious spikes in this region, see Fig. 15, which again can lead to spurious period estimates. In all of these cases the best way to proceed is to examine the ambiguous periods and select the most appropriate period according to whether the period lies within the expected Circadian range and has appropriate amplitude and GOF values.

40

Anne Moore et al. x 105

2.5

input amp detrended 2

1.5

1

0.5

0

−0.5

0

50

100

150

Fig. 14 Example of biological time series showing input data (blue line) and data after amplitude detrending (green line). Removing the amplitude trend has revealed low level oscillations in the later stages of the experiment which could potentially introduce spurious period estimates

1.8

x 108 input amp detrended

1.6

1.4 1.2 1

0.8 0.6

0.4

0

10

20

30

40

50

60

70

80

90

100

Fig. 15 Example of biological time series showing input data (blue line) and data after amplitude detrending (green line). The data are severely damped towards the end of the experiment, and application of the amplitude detrending algorithm introduces spurious oscillations which can lead to false period estimates

Circadian Period Estimation and Rhythmicity

41

23. The different options for reporting phase are as follows: Phase by fit: this fits a single cosine using the period found by whichever method the user selected. The phase of the single cosine is reported. Phase by first peak: this finds the first peak of the curve fitted by the user’s chosen method and reports the phase associated with that first peak. Phase by average peak: uses all peaks of the curve fitted by the user’s chosen method and reports the average value. Phase by method: If FFT-NLLS is used this will report the cosine of the main component. If there is only a single cosine, the phase will be the same as reported by “Phase by fit”; however, if there is more than one cosine, the phase reported will be slightly different. If mFourfit is used “Phase by method” will report the first/average peak of the best fit curve. This should be the same as “Phase by first peak” and “Phase by average peak.” If the periodogram is used “Phase by method” will report the first/average peak of the best fit curve. This should be the same as “Phase by first peak” and “Phase by average peak.” If spectrum resampling is used “Phase by method” will report the phase of the cos component used to create the fit. In all cases the phase lies within the range [0, period], or [data_window_start, data_window_ start + period] if phase related to data window was selected in the output options. Such a way of reporting phase matches intuitively the observed peak in the data. For example if the data window was [50-max] and the phase by fit is 20, one can expect a peak in the data around time 70 (50 + 20). 24. For each period analysis job number, the following are listed: the choice of algorithm; the input data used, data window and the expected period range (the circadian range). 25. It is possible that the best fit period is one which lies outside the circadian range. In this case it is up to the user to decide whether to include or exclude this period. Furthermore, when the FFT NLLS algorithm is used to carry on the period analysis there may be situations where it is not possible to determine automatically which period should be considered as the best circadian period for a specific time series (as more than one circadian component was found). In this case the FFT the various possible periods together with their phase, amplitude and RAE will be displayed. 26. The results given in the “Analysis statistics” section are: period estimate (arithmetic mean over all traces for that genotype); phase estimates (again arithmetic mean over all traces for that genotype); the (arithmetic) mean amplitude; the (arithmetic) mean Goodness of Fit (GOF) and the (arithmetic) mean of global error (ERR) for each genotype. If the ‘Average value weighted by GOF) was selected then the means are GOF

42

Anne Moore et al.

weighted, so that the results with smaller GOF will contribute more to the mean. 27. Goodness of Fit (GOF) is defined as the ratio of two errors: the method fit error, i.e., the error between the original time series and the curve predicted by the user-selected algorithm, and the “polynomial fit error,” i.e., the error between the original time series and a polynomial (cubic) curve fitted to the time series. The ratio can vary from 0 (model provides a perfect fit) to a large number, indicating that the model is no better than (or is worse than) a cubic fit to the data. 28. Unlike the GOF, the global error (ERR) is method specific and represents the error measure defined for a given method. In case of FFT it is RAE (relative amplitude error), the ratio between estimated error in amplitude and amplitude value, for mFourFit it is AIC (which describes the trade-off between the accuracy of the fit and the complexity of the model), for the periodogram it is the critical value of a chi squared distribution with P degrees of freedom (where P is the period of interest), for spectrum resampling it is standard deviation in period values found by the bootstrap. 29. When exporting the analysis results the zip file contains one individual spreadsheet per analysis job and one summary spreadsheet. The individual spread sheets give the following parameters: GOF-weighted mean; GOF-weighted standard error; GOF-weighted standard deviation; mean; standard error; standard deviation; median; variance; kurtosis; skewness; minimum; maximum; and range for each of the period, the phase, and the range. Variance-weighted statistics were introduced for this analysis in the following reference [22]. 30. Usually the best guideline about which, if any, period to select is to choose the one which lies inside the circadian range and which has the smaller or smallest error or the larger or largest amplitude. Plotting the fit from selection screen may also help, as it will show not only generated fit but also cosine waves of corresponding to the estimated periods, the period of the cosine wave that follows the original data the best should be selected. If there is still uncertainty and there are many different time series from the same genotype, marker, and line, then it is probably best to omit this time series from the analysis. 31. The final section of the “Display GOF” screen shows the graphs of GOF versus period (in hours). All results from the same line are displayed on a single graph, so if there were 12 data-series from the same genetic line in the experiment and they were all included in the analysis there will be 12 points on the scatter plot. First the graphs for all genetic lines with a wild-type background are displayed, and these are followed by

Circadian Period Estimation and Rhythmicity

43

the different mutant backgrounds with the wild-type shown as a comparison. 32. Only experiments for which the user has read access will be found. 33. The first visit to the “Search experiments” screen after logging in will list all the experiments owned (submitted) by the user. Searching without any criteria will also list the experiments owned (i.e., uploaded) by the user.

Acknowledgements We are grateful for much helpful discussion, feedback and programming input from Drs. Martin Beaton and Eilidh Troup (Edinburgh), Paul Brown (Warwick), from members of our collaborating projects listed below, and from the community via the UK Circadian Clock Club. BioDare development was funded by the BBSRC and EPSRC systems biology projects ROBuST (award BB/F005237) and SynthSys (award BB/D019621), and by the EU FP7 Integrated Project TiMet (award 245143). References 1. Mackey SR, Golden SS (2007) Winding up the cyanobacterial circadian clock. Trends Microbiol 15(9):381–388 2. Dong G, Golden SS (2008) How a cyanobacterium tells time. Curr Opin Microbiol 11(6):541–546 3. Ukai H, Ueda HR (2010) Systems biology of mammalian circadian clocks. Ann Rev Physiol 72:579–603 4. Lowrey PL, Takahashi JS (2011) Genetics of circadian rhythms in mammalian model organisms. Adv Genet 74:175–229 5. Bujdoso N, Davis SJ (2013) Mathematical modelling of an oscillating gene circuit to unravel the circadian clock network of Arabidopsis thaliana. Frontiers Plant Syst Biol 4:2013.00003 6. Plautz JD, Straume M, Stanewsky R et al (1997) Quantitative analysis of Drosophila period gene transcription in living animals. J Biol Rhythms 12:204–217 7. Locke JCW, Southern MM, Kozma-Bognar L, Hibberd V, Brown PE, Turner MS, Millar AJ (2005) Extension of a genetic network model by iterative experimentation and mathematical analysis. Mol Syst Biol 1:13 8. Enright JT (1965) The search for rhythmicity in biological time-series. J Theor Biol 8: 426–468

9. Refinetti R (1992) Laboratory instrumentation and computing: comparison of six methods for the determination of the period of circadian rhythms. Physiol Behav 54:869–875 10. Gould PD et al (2013) Network balance via CRY signalling controls the Arabidopsis circadian clock over ambient temperatures. Mol Syst Biol 9:650 11. Gould PD, Diaz P, Hogben C et al (2009) Delayed fluorescence as a universal tool for the measurement of circadian rhythms in higher plants. Plant J 58:893–901 12. http://hallidaylab.bio.ed.ac.uk/ROBuST.html 13. Edwards KD, Akman OE, Knox K, Lumsden PJ, Thomson AW, Brown PE, Pokhilko A, Kozma-Bognar L, Nagy F, Rand DA, Millar AJ (2010) Quantitative analysis of regulatory flexibility under changing environmental conditions. Mol Syst Biol 6:424 14. Southern MM, Millar AJ (2005) Circadian genetics in the model higher plant, Arabidopsis thaliana. Methods Enzymol 393:23–35 15. Jameson D, Garwood K et al (2008) Data capture in bioinformatics: requirements and experiences with Pedro. BMC Bioinformatics 9:183 16. Edwards KD, Lynn JR, Gyula P, Nagy F, Millar AJ (2005) Natural allelic variation in the temperature compensation mechanisms of the

44

Anne Moore et al.

Arabidopsis thaliana circadian clock. Genetics 170:387–400 17. Salathia N, Lynn JR, Millar AJ, King GJ (2007) Detection and resolution of genetic loci affecting circadian period in Brassica oleracea. Theor Appl Genet 114:683–692 18. McWatters HG, Bastow RM, Hall A, Millar AJ (2000) The ELF3 zeitnehmer regulates light signalling to the circadian clock. Nature 408: 716–720 19. Sokolove PG, Bushell WN (1978) The chi square periodogram: its utility for analysis

of circadian rhythms. J Theor Biol 72: 131–160 20. Costa MJ et al (2013) Inference on periodicity of circadian time series. Biostatistics 2013: 1–15 21. Zielinski T, Moore A et al (2014) Strengths and limitations of period estimation methods for circadian data. PLoS ONE (in press) 22. Millar AJ, Straume M, Chory J, Chua N-H, Kay SA (1995) The regulation of circadian period by phototransduction pathways in Arabidopsis. Science 267:1163–1167

Chapter 3 Global Profiling of the Circadian Transcriptome Using Microarrays Polly Yingshan Hsu and Stacey L. Harmer Abstract Circadian rhythms are biological cycles with a period length of approximately 24 h that are generated by endogenous clocks. The application of microarrays for high-throughput transcriptome analysis has led to the insight that substantial portions of the transcriptomes of both humans and many model organisms are clock-regulated. In a typical circadian time course microarray experiment, samples are collected from organisms maintained in constant environmental conditions, gene expression at each time point is determined using microarrays, and finally clock-regulated transcripts are identified using statistical algorithms. Here, we describe how to design the experiment, process RNA, determine expression profiles using ATH1 microarrays, and use a nonparametric statistical algorithm named JTK_CYCLE in order to identify circadian-regulated transcripts in Arabidopsis. This basic procedure can be modified to identify clockregulated transcripts in different organisms or using different expression analysis platforms. Key words Transcriptome, Microarray, Circadian rhythms, Cycling genes, JTK_CYCLE

1

Introduction Circadian rhythms are physiological and behavior cycles with a period of approximately 24 h. Numerous circadian microarray experiments have shown that a substantial portion of the transcriptomes of cyanobacteria, Arabidopsis, Drosophila, and mammals is controlled by the clock [1–5]. The examination of global transcript abundance over circadian time allows for the identification of both components of the circadian oscillator as well as genes and pathways under circadian regulation. Microarray analysis is a well-established technique that enables the simultaneous, genome-wide monitoring of essentially all transcripts in a sample. It involves the hybridization of biotin-labeled copy RNA (cRNA) (or DNA, depending upon the labeling procedure) molecules to thousands, or even millions of different probe sequences that are immobilized on a solid support to form a microarray. For Affymetrix gene expression arrays, these probes are

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_3, © Springer Science+Business Media New York 2014

45

46

Polly Yingshan Hsu and Stacey L. Harmer

25 nucleotide—long oligonucleotides complementary to RNA transcripts. A group of probes complementary to the same transcript is called a probeset; to help control for local differences in hybridization efficiency, probes belonging to a single probeset are usually distributed across the entire array. After biotin-labeled nucleic acids are hybridized to an array, the arrays are washed, stained with streptavidin–phycoerythrin, and then scanned to determine the patterns of hybridization. This raw image data is then preprocessed to reduce technical sources of variation, such as uneven hybridization efficiency among different microarrays. Here we describe the robust multi-array average (RMA) [6] method for this low-level processing. RMA allows for background correction, quantile normalization, and summarization using a robust, multi-array linear model [6]. This low-level analysis requires a .cdf (chip description file) file which provides a “map” indicating the position of each probe on the array and generates expression values for individual probesets and thus transcripts. RMA analysis of each array in a time course therefore results in an average expression value for each probeset on the array for every time at which a sample was harvested. After obtaining expression values for individual transcripts over such a time course, clock-regulated transcripts are commonly identified as those with expression patterns that fit a cosine wave with a near-circadian period (typically 20–28 h). Many statistical algorithms have proven useful in these studies [7–9]. We focus here on the use of a nonparametric algorithm, JTK_CYCLE [10], to identify circadian-regulated transcripts. JTK_CYCLE is a computationally efficient algorithm that has been implemented in the statistical language and environment R, making it relatively fast and easy to use. In this chapter, we describe the steps involved in identifying circadian-regulated transcripts using microarrays. These steps are as follows: (1) collection of samples over a circadian time course, (2) RNA extraction and quality control, (3) labeling of RNA and hybridization to microarrays, (4) low-level analysis of microarray data, and (5) identification of circadian-controlled genes using JTK_CYCLE. Since Step 3 is usually carried out in core facilities, we will focus our attention on the other steps, which are commonly carried out in individual labs.

2

Materials

2.1 Circadian Time Course Sample Collection

1. Growth media: 1× Murashige and Skoog, 0.7 % agar, 3 % sucrose, pH 5.7. 2. Petri dishes: 100 × 100 × 15 mm disposable petri dishes with grids (Fisher Scientific, cat. number 0875711A). 3. Growth chamber: Conviron TC 16 growth chamber.

Profiling of the Circadian Transcriptome using Microarrays

47

4. 1.5 mL screw-cap tubes. 5. 2 mm zirconia beads (BioSpec Products, Cat. number 11079124zx). 6. Liquid nitrogen. 7. Forceps for collecting plant samples. 2.2 RNA Extraction and Quality Control

1. Liquid nitrogen. 2. Mini-beadbeater-96 (BioSpec Products, Cat. number 1001) and 2.0 ml vial rack (BioSpec Products, cat. number 702VH45). 3. Ice bucket with ice. 4. Trizol (Invitrogen). 5. Chloroform. 6. 4 °C Microfuge. 7. Isopropanol. 8. 70 % Ethanol. 9. 95 % Ethanol. 10. RNase-Free DNase Set (Qiagen): resuspend the DNase I powder in nuclease-free water and aliquot according to the manufacturer’s manual. Store at −20 °C. RDD buffer is included in kit. 11. RNeasy MinElute Cleanup Kit (Qiagen). 12. NanoDrop ND 1000 (NanoDrop Technologies). 13. DNase-, RNase-free pipet tips and eppendorf tubes.

2.3 Low-Level Analysis of Microarray Data

1. R: R is a free software environment for statistical computing and graphics. It compiles and runs on a wide variety of operating systems, including UNIX, Windows, and Mac OS platforms. You can download R here: http://www.r-project.org/. 2. Practice microarray dataset: GSE8365 data can be downloaded from here: http://www.ncbi.nlm.nih.gov/geo/query/acc. cgi?acc=GSE8365 (In the bottom of the page, click (http) under Download, unzip the file, place all of the .CEL files (raw data for each arrays) in your working folder).

2.4 Detection of Circadian-Controlled Genes Using JTK_CYCLE

1. JTK_CYCLE: a nonparametric test procedure that detects cycling components from a large dataset, developed by Hughes et al. [10]. You can download JTK_CYCLE here: http://openwetware.org/wiki/HughesLab:JTK_Cycle (click “Image:JTK Cycle (2-19-13).zip” to download, unzip the file). 2. Microsoft Excel.

48

3

Polly Yingshan Hsu and Stacey L. Harmer

Methods 1. Arabidopsis seeds are sterilized and plated on growth media (~30 seeds for each sample) (see Note 1) and stratified at 4 °C in the dark for 2 days before release to growth chamber.

3.1 Collection of Samples to Generate a Circadian Time Course

2. For a circadian time course, after stratification, the seeds are germinated and plants are entrained in 12 h light–12 h dark cycles (light: we use ~55 μEi, cool white fluorescent bulbs) at 22 °C for a week and released to free-run (constant light: 55 μEi white fluorescent light and constant temperature: 22 °C), and samples are collected starting after 24 h in free-run (24 h after the last light to dark transition) at a 4-h interval for 2 days (see Fig. 1). It has been reported that shorter time intervals and/or a longer time course increase both the resolution of the time course and the statistical power for identifying circadian-regulated genes [1, 11]. However, we have found our usual experimental design allows for the identification of thousands of clock-regulated transcripts. 3. Prepare and label screw-cap tubes each containing 3–4 beads. At each time point, quickly harvest plants using forceps, place plants in a screw-cap tube, freeze samples immediately in liquid nitrogen, and store the samples at −80 °C. 3.2 RNA Extraction and Quality Control

●

●

●

If not otherwise specified, RNA samples should be kept on ice at all times. All centrifuge steps should use a microfuge cooled to 4 °C. Consult your core facility first to determine the quantity and quality of RNA required for your microarray analysis. 1. Pre-cool the microfuge to 4 °C. Take samples from −80 °C and place them in liquid nitrogen. Slowly pour liquid nitrogen on the vial rack of the beadbeater to cool it down. Place the samples in the cooled vial rack (put tubes in the center of the rack if possible so the samples can keep cold), and shake for 2 min using the beadbeater. Quickly place the samples back in the liquid nitrogen container (see Note 2). free-run

entrainment

7 days

1 day

ZT24 28 32

…

64 68

Fig. 1 Design of a circadian time course experiment. Arabidopsis seedlings are typically grown under light–dark cycle for a week and then released to free-run (constant light and temperature). Samples are collected after 24 h in free-run at 4-h intervals over two consecutive days

Profiling of the Circadian Transcriptome using Microarrays

49

2. Prepare ice bucket with ice in the fume hood. In the fume hood, move 1–2 tubes from liquid nitrogen and place in the ice bucket. Open the screw caps and add 0.7–1 ml Trizol (see Note 3) per samples. Quickly place the cap back on and vortex the samples until the powdered tissue is completely resuspended (see Note 4). Leave these processed samples on ice while working on the rest of the samples. 3. Add 1/5 volume of chloroform (e.g., if you used 700 μl Trizol, add 140 μl chloroform) and invert the tube several times until the solution is thoroughly mixed. Centrifuge the samples in a microfuge at 18,000 × g for 15 min. 4. Use 200 μl pipet tips (pipetman is set to 150 μl), carefully transfer the aqueous layer to a new 1.5 ml eppendorf tube (avoid disturbing the aqueous–organic interface; it is okay to leave a small volume behind); it typically takes 3 times to transfer the aqueous layer per sample. Add 1 volume of isopropanol, usually about 450 μl. Invert the tube several times until the solution is thoroughly mixed. Centrifuge the samples in a microfuge at 18,000 × g for 15 min. 5. Remove the supernatant using 1 ml pipet tips. Wash the pellet with 700 μl 70 % ethanol. Invert the tube several times until the pellet is released from the bottom of the eppendorf tube. Centrifuge the samples in a microfuge at 18,000 × g for 5 min. 6. Remove the supernatant using 1 ml pipet tips. Wash the pellet with 700 μl 95 % ethanol. Invert the tube several times until the pellet is released from the bottom of the eppendorf tube. Centrifuge the samples in a microfuge at 18,000 × g for 5 min. 7. Remove the supernatant using 1 ml pipet tips. Briefly spin again. Use 20 μl tips to remove as much as liquid as possible. Air-dry the pellet by setting the tube upside-down on a piece of paper towel for 10–15 min at room temperature (see Note 5). 8. Add 87.5 μl nuclease-free water and allow the sample to sit on the bench at room temperature for 5 min. While waiting, combine DNase I solution and RDD buffer following the manufacturer’s protocol. 9. Add 12.5 μl DNase I/RDD solution and pipet up and down to mix. Allow it to sit on the bench at room temperature for 10 min. 10. Clean RNA samples using Qiagen mini-elute cleanup kit following the manufacturer’s protocol (see Note 6). 11. Examine the quality and quantity of RNA using the NanoDrop. Select “Nucleic Acids” application module; select “RNA-40” sample type. Use 2 μl of nuclease-free water to blank the machine and 2 μl of each sample for measurement. The 260/280 and 260/230 ratios (ratios of sample absorbance at

50

Polly Yingshan Hsu and Stacey L. Harmer

260 and 280 nm, or 260 and 230 nm) are measures of nucleic acid purity. A ratio of ~2.0 for 260/280 is generally accepted as “pure” for RNA (a lower ratio may indicate contamination with protein or phenol). The ratio for 260/230 for “pure” nucleic acid is often higher than its respective 260/280 ratio (around 2.2) [12]. Sample concentration is presented in ng/μl (see Note 7). 12. Aliquot and dilute RNA samples as recommended by your core facility to avoid repeated freeze-thaw cycles (see Note 8). 13. RNA should be stored in −80 °C until ready for the next steps. 3.3 Labeling and Microarray Hybridization

These steps are commonly performed by a core facility affiliated with a university or research institute and are not discussed here.

3.4 Low-Level Analysis of Microarray Data

This and the next data analysis step assume some basic familiarity with the statistical language and software environment R, in that we expect you are able to import data to R and carry out basic data manipulations. Although to those not familiar with the command line this may seem daunting, a small investment of time will give you enough experience with R to carry out the analytical steps described below. Working through the early chapters of either Dalgaard’s Introductory Statistics with R [13] or Adler’s R in a Nutshell [14] should provide sufficient background for these analysis steps. We will use a published dataset generated by Covington et al. [15] using Affymetrix ATH1 arrays (GSE8365) to demonstrate how to analyze microarray data. In this dataset, Arabidopsis seedlings were grown in light–dark cycles for 7 days and then moved to constant light and temperature. After 24 h in this free-running condition, plants were harvested at 4-h intervals over 2 days. These samples were labeled ZT24–ZT68, corresponding to time of harvesting (ZT0=the last dark to light transition). RNA was extracted, labeled, and then hybridized to ATH1 arrays in a core facility. Arrays were scanned to generate the raw image data (.CEL files) that were processed as described below. ●

●

●

Follow Method (item 2 of Subheading 2.3) to download the .CEL files from GSE8365. The machine used in this demonstration is Mac OS X 10.7.4 with processor: 1.8 GHz Intel Core i5, and memory: 4GB 1600 MHz DDR3. The software is R 2.15.2 and JTK_CYCLE updated on 2013/02/19. Commands typed in R are presented in courier font. 1. In R, change the working directory to where the microarray raw data (.CEL files) are stored by clicking Misc ->Change Working Directory, and selecting the correct folder.

Profiling of the Circadian Transcriptome using Microarrays

51

2. Install the “affy” package: source("http://bioconductor.org/biocLite.R") biocLite("affy") Load the “affy” package: library(affy) 3. Read all of the microarray raw data in the current working directory into memory; the CDF file corresponding to the ATH1 microarray platform will be automatically loaded (see Note 9). rawData=ReadAffy() Check information about the raw data and the microarray platform: rawData You should see the following information: AffyBatch object size of arrays=712x712 features (21 kb) cdf=ATH1-121501 (22810 affyids) number of samples=12 number of genes=22810 annotation=ath1121501 notes= 4. Evaluate hybridization of each individual array: image(rawData[,1]) This produces an image for the first array (takes about 30 s). For arrays with good hybridization results, you should see a fairly even signal across the entire chip. Change the value of x in rawData[,x] to different numbers (1–12 for these 12 arrays) to examine different arrays. 5. Rename the sample names in rawData: sampleNames(rawData)=c ("ZT24", "ZT28", "ZT32", "ZT36", "ZT40", "ZT44", "ZT48", "ZT52", "ZT56", "ZT60", "ZT64", "ZT68") 6. To determine whether some arrays have overall higher or lower hybridization values than the others, examine the distribution of the signal intensities across all of the arrays in the experiment: boxplot(rawData, main="Before "log2 intensity", las=2)

RMA",

ylab=

main=“Before RMA” generates a title for the plot, las =2 indicates that text in x-axis is vertical (use las=1 to specify horizontal text). It is common for the distribution of signal intensities to vary between individual arrays (Fig. 2a), suggesting better

52

Polly Yingshan Hsu and Stacey L. Harmer

hybridization on some arrays than others. To correct for this, it is necessary to normalize the data by adjusting the distribution of the signal intensity among all of the arrays in the experiment. This allows for the comparison of expression values for the same transcript over different arrays (see Note 10). 7. Normalize the arrays using RMA: CovingtonRMA=rma(rawData) This generates an object (CovingtonRMA) that has undergone RMA processing. RMA carries out background correction, normalization and calculation of expression values for each probeset (transcript). 8. Extract the expression values after RMA e=exprs(CovingtonRMA) Examine the expression values by checking the first six lines of the data head(e) 9. Evaluate normalized expression after RMA boxplot(e, main="After intensity", las=2)

RMA",

ylab="log2

Compare this plot with that generated before RMA normalization (Fig. 2a). After RMA processing, the distribution of signal intensities is very similar among the samples (Fig. 2b). 10. Output the normalized expression values

Fig. 2 (a) Distributions of signal intensities in log2 scale among individual arrays before RMA normalization, (b) distributions of signal intensities in log2 scale among individual arrays after RMA normalization

Profiling of the Circadian Transcriptome using Microarrays

53

write.csv(e, file="Covington_RMA.csv") You have now completed low level analysis and are ready to identify clock-regulated transcripts using JTK_CYCLE. 3.5 Detection of Circadian-Controlled Genes Using JTK_CYCLE

1. Prepare the data formats: –

Open the data file (Covington_RMA.csv) above using Excel and saved as a text delimited file, i.e., Covington_ RMA.txt file.

–

Copy the annotation column and paste to a new Excel sheet, giving it a header line such as “probeset,” and then save as a text delimited file, i.e., Covington_RMA_annot. txt file.

–

Place the data file (Covington_RMA.txt) and the annotation file (Covington_RMA_annot.txt) in the JTK_CYCLE folder (see Note 11).

2. Open the file “run_JTK_CYCLE (Example1).R” using R and modify the following: –

Line 4: change the project name to something that makes sense to you.

–

Line 7: change the file name to ”Covington_RMA_annot. txt.”

–

Line 8: change the file name to ”Covington_RMA.txt.”

–

Line 12: indicate the number of time points and replicates (no change needed for Covington’s data; see Note 12).

–

Line 15: specify the spacing of the time points; for Covington’s dataset, the sampling interval is 4 h, change to “jtk.init(periods, 4).”

–

Line 14: specify the period length you are interested in. The numbers in Line 14 are based on the number of time points per cycle. In this case, the time points are spaced 4 h apart. Change lines 14 to: “periodsChange Working Directory, and selecting the correct folder. 4. After modifying the ‘Run_JTK_CYCLE_Covington.R’ script, select all and paste into R’s console. The first half of script will run quickly, but the other half takes a while, dependent on your computer and the size of your data (takes about 1 min using the machine and softwares described here).

54

Polly Yingshan Hsu and Stacey L. Harmer

5. JTK_CYCLE outputs two files: (1) a .txt with all the important cycling statistics, and (2) an .rda file. The .txt includes BH.Q (Benjamani–Hochberg q-value), ADJ.P (adjusted p-value), PER (period), LAG (phase), and AMP (amplitude), along with the raw data. The .rda file contains R objects and will not typically be needed by most users. 6. The BH.Q and ADJ.P values include a correction for multiple testing. ADJ.P < 0.05 is often used as a cutoff to identify clockregulated transcripts [16], but you may want to compare several ADJ.P values (or BH.Q for more stringent criteria) and empirically determine if they are reasonable cutoffs for your data by examining gene expression patterns.

4

Notes 1. For 1-week-old Arabidopsis seedlings, follow the Trizol extraction described here, which typically yields about 20–30 μg total RNA per sample. Consult your core facility first about the quantity and quality of RNA that you’ll need for your microarray experiment. 2. Make sure that the samples are still frozen. The samples should consist of a pale green powder when removed from the beadbeater. 3. For 1-week-old Arabidopsis seedlings, 700 μl Trizol is sufficient for 30 plants. 4. When mixing the tissue with Trizol, some tissue may be stuck on the screw cap; therefore, shake the tubes up and down as well as vortex to make sure that they are mixed well. 5. Watch the pellets to make sure that they are dry enough—they should look transparent rather than whitish. Any residual ethanol in the sample will cause a lower 260:230 nm ratio and may interfere with downstream enzymatic application, like cDNA synthesis. 6. The Qiagen mini-elute cleanup kit procedure is done at room temperature. It is best to split the samples into several batches (like 6 or 8 samples per batch) to reduce the time that the RNA sits at room temperature and to ensure highquality RNA. 7. We typically obtain 260/280 ratios of ~2.1 and 260/230 ratios of ~2.3 using the Trizol extraction method described here. Your core facility may perform additional quality control procedures, for example, using a Bioanalyzer (Agilent) machine to determine the size distribution and degradation level of the RNA samples.

Profiling of the Circadian Transcriptome using Microarrays

55

8. The concentration and volume of RNA needed for RNA labeling/hybridization vary depending on the procedures used and should be discussed with your core facility first. 9. If you are using custom Affymetrix microarrays, the corresponding .cdf file will not be loaded automatically. You will have to obtain the .cdf from the manufacturer and set the cdf environment using the bioconductor packages “altcdfenvs” [17] and “makecdfenv” [18] or follow the recommendations of your manufacturer. 10. There are three main assumptions underlying this normalization procedure: (1) the majority of genes are not differentially expressed, (2) mis-regulated genes are equally likely to be upor down-regulated, and (3) the differential expression signal is independent of the average gene expression levels [19]. 11. If you collect samples in replicate, make sure that the replicate samples are in adjacent columns when preparing the data file for JTK_CYCLE (see JTK_User’s Guide “Example2_data.txt”). 12. If you collect time points with replicates, for example, if you have 13 time points, and 2 replicates per time point, change Line 12 to read: “jtkdist(13,2).

Acknowledgement We thank Hsin-Yen Wu for reading and testing the commands in this manuscript, and Michael Hughes for helpful discussions about JTK_CYCLE. This work was supported by the National Institutes of Health (NIGMS) (http://www.nigms.nih.gov/) [GM069418] and the Taiwan Merit Scholarship (http://web1.nsc.gov.tw/) [NSC-095-SAF-I-564-014-TMS]. References 1. Covington MF, Maloof JN, Straume M, Kay SA, Harmer SL (2008) Global transcriptome analysis reveals circadian regulation of key pathways in plant growth and development. Genome Biol 9:R130 2. McDonald MJ, Rosbash M (2001) Microarray analysis and organization of circadian gene expression in Drosophila. Cell 107:567–578 3. Kucho K, Okamoto K, Tsuchiya Y, Nomura S, Nango M et al (2005) Global analysis of circadian expression in the cyanobacterium Synechocystis sp. strain PCC 6803. J Bacteriol 187:2190–2199 4. Akhtar RA, Reddy AB, Maywood ES, Clayton JD, King VM et al (2002) Circadian cycling of the mouse liver transcriptome, as revealed by

cDNA microarray, is driven by the suprachiasmatic nucleus. Curr Biol 12:540–550 5. Panda S, Antoch MP, Miller BH, Su AI, Schook AB et al (2002) Coordinated transcription of key pathways in the mouse by the circadian clock. Cell 109:307–320 6. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ et al (2003) Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4:249–264 7. Straume M (2004) DNA microarray time series analysis: automated statistical assessment of circadian rhythms in gene expression patterning. Methods Enzymol 383: 149–166

56

Polly Yingshan Hsu and Stacey L. Harmer

8. Levine JD, Funes P, Dowse HB, Hall JC (2002) Signal analysis of behavioral and molecular cycles. BMC Neurosci 3:1 9. Wichert S, Fokianos K, Strimmer K (2004) Identifying periodically expressed transcripts in microarray time series data. Bioinformatics 20:5–20 10. Hughes ME, Hogenesch JB, Kornacker K (2010) JTK_CYCLE: an efficient nonparametric algorithm for detecting rhythmic components in genome-scale data sets. J Biol Rhythms 25:372–380 11. Hughes ME, DiTacchio L, Hayes KR, Vollmers C, Pulivarthy S et al (2009) Harmonics of circadian gene transcription in mammals. PLoS Genet 5:e1000442 12. NanoDrop, Technologies, Inc. (2007) ND-1000 Spectrophotometer User’s Manual 13. Dalgaard P (2009) Introductory statistics with R. Springer, New York

14. Adler J (2010) R in a Nutshell. O’Reilly Media, Sebastopol, CA 15. Covington MF, Harmer SL (2007) The circadian clock regulates auxin signaling and responses in Arabidopsis. PLoS Biol 5:e222 16. Hsu PY, Harmer SL (2012) Circadian phase has profound effects on differential expression analysis. PLoS One 7:e49853 17. Gautier L, Moller M, Friis-Hansen L, Knudsen S (2004) Alternative mapping of probes to genes for Affymetrix chips. BMC Bioinformatics 5:111 18. Irizarry RA, Gautier L, Huber W, Bolstad B (2006) makecdfenv: CDF environment maker. R package version 1360 19. Gohlmann H, Talloen W (2010) Gene expression studies using Affymetrix microarrays. CRC Press, Boca Raton, FL

Chapter 4 ChIP-Seq Analysis of Histone Modifications at the Core of the Arabidopsis Circadian Clock Jordi Malapeira and Paloma Mas Abstract Over the past years, chromatin modification has emerged as a key regulator of gene expression. A very useful method for chromatin analysis is chromatin immunoprecipitation (ChIP), which allows the quantification and localization of specific histone modifications. The basic steps of the ChIP protocol include cross-linking of histones and DNA, chromatin isolation, shearing the DNA into smaller fragments, immunoprecipitation with specific antibodies, and enrichment analysis by several methods including real-time quantitative PCR (ChIP-qPCR), microarray hybridization (ChIP-chip), or sequencing (ChIP-seq). Here, we describe how to use ChIP-qPCR to analyze histone modifications at the core of the Arabidopsis thaliana circadian clock. We also briefly discuss a number of protocol adjustments to be considered in ChIP-seq experiments. Key words Chromatin immunoprecipitation (ChIP), ChIP-qPCR, Circadian clock, Histone modifications, Arabidopsis thaliana

1

Introduction Regulation of chromatin structure plays a central role in the epigenetic control of gene expression. A key mechanism for chromatin modulation is the post-translational modification of histones. The N-terminal tail of histones can be covalently modified by acetylation, methylation, phosphorylation, ubiquitination, and ADP-ribosylation. The combination of these modifications is essential to regulate gene expression [1]. Chromatin immunoprecipitation (ChIP) and the subsequent analysis of the immunoprecipitated chromatin through qPCR (ChIP-qPCR) [2, 3], microarray hybridization (ChIP-chip) [4, 5], or sequencing (ChIPseq) [6, 7] has become a very useful tool in the field of epigenetics. This technique allows the detection of DNA–histone interactions at a specific site of interest or throughout the genome. A widespread use of ChIP consists in the detection and mapping of specific N-terminal tail modifications of histones, which is a key step

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_4, © Springer Science+Business Media New York 2014

57

58

Jordi Malapeira and Paloma Mas

for improving our understanding of the so-called “histone code”: combinations of histone modifications that ultimately determine the transcription state of chromatin [8, 9]. To perform chromatin immunoprecipitation, the plant material is first treated with a buffer containing formaldehyde, which cross-links the chromatin complexes [10–12]. The fixation reaction is stopped with glycine, followed by tissue grinding in liquid nitrogen and extraction of chromatin. The chromatin is subsequently sheared by sonication in order to obtain smaller fragments of DNA [13]. Specific antibodies against the histone modification or the transcription factor of interest are used to immunoprecipitate the chromatin, which is then washed, eluted and reverse crosslinked to release the DNA [11, 14]. This DNA will be enriched in those sequences associated with the chromatin element of interest (the histone modification or transcription factor) in comparison to a negative control where no antibody is used. Further analysis of this DNA includes qPCR (ChIP-qPCR) [2, 3] to screen a specific locus or genome-wide analysis such as microarray hybridization (ChIP-chip) [4, 5] or sequencing (ChIP-seq) [6, 7]. Here, we describe in detail the ChIP-qPCR protocol to analyze the accumulation of two well-known histone modifications H3K56 acetylation (H3K56ac) and H3K4 trimethylation (H3K4me3) at the core of the Arabidopsis thaliana circadian clock. Different time points, conditions, or genetic backgrounds can be used to compare the accumulation of the histone marks. We routinely start our assays with two time points, e.g., Zeitgeber Time 3 (ZT3) and Zeitgeber Time 15 (ZT15). Further studies include more complete time-course analysis, with samples taken every 4 h over the 24-h circadian cycle. For the particular assays described in this article, the sheared chromatin is immunoprecipitated with antibodies against H3K56ac and H3K4me3 in samples taken at ZT3 and ZT15. The purified DNA is analyzed by qPCR with specific primers that amplify the 5′ end region of the circadian clock genes including CIRCADIAN CLOCK ASSOCIATED 1 (CCA1), LATE ELONGATED HYPOCOTYL (LHY), PSEUDORESPONSE REGULATOR 9 (PRR9), PSEUDO-RESPONSE REGULATOR 7 (PRR7), LUX ARRITHMO (LUX), and TIMING OF CAB EXPRESSION 1 (TOC1). If no knowledge about the positional enrichment of the specific mark is available, it is advisable to design primers that cover the entire gene.

2 2.1

Materials Equipment

1. Refrigerated centrifuge. 2. Laboratory vacuum chamber and pump. 3. Vortex.

ChIP-Seq Analysis of Histone Modifications at the Core…

59

4. Rotating wheel. 5. Branson digital sonifier. 6. pH meter. 7. Desiccator jar. 8. qPCR machine. 9. Mixing block at 65 °C. 10. Mortar and pestle. 11. 50 and 15 ml tubes. 12. 1.5 and 2 ml tubes. 13. Petri dishes. 14. Filter paper. 15. Aluminum foil. 16. Balance. 17. Tweezers. 18. Miracloth. 19. Micropore tape. 20. 96-Well plate and transparent film. 2.2

Reagents

1. Murashige and Skoog (MS) medium. 2. MES hydrate (2-(N-Morpholino) ethanesulfonic acid hydrate). 3. Agar. 4. Bleach (NaClO 5 %). 5. HCl 37 %. 6. Milli-Q water. 7. Liquid nitrogen. 8. Sepharose beads (see Note 1). 9. Antibodies: anti-H3K56ac and anti-H3K4me3 (see Note 2). 10. DNA purification kit.

2.3 Medium and Buffers

1. 3 % MS medium: 0.44 % MS, 0.05 % MES, 3 % sucrose, 0.8 % agar. To prepare 1 l of 3 % MS medium dissolve 4.4 g of MS, 0.5 g of MES hydrate and 30 g of sucrose into 800 ml of Milli-Q water by stirring. Using a pH meter, titrate the solution with 2 N KOH until pH 5.7 is reached. Add 8 g of agar and adjust the volume to 1 l in a graduated cylinder with Milli-Q water. Autoclave and store at room temperature until further use. To prepare the MS plates completely melt the MS bottle in a microwave and pour in the plates under sterile conditions. We used 3 % sucrose MS plates although different plate conditions and medium can be used depending on the experimental design.

60

Jordi Malapeira and Paloma Mas

2. PMSF stock solution: 100 mM (17.42 mg/ml) PMSF. Dissolve in ethanol and store at −20 °C freezer. For 10 ml of buffer add 100 μl of PMSF stock. 3. Antipain stock solution: 3.7 mM (2.5 mg/ml) antipain. Dissolve in Milli-Q water and store at −20 °C freezer. For 10 ml of buffer add 30 μl of antipain stock. 4. Chymostatin stock solution: 4.1 mM (2.5 mg/ml) chymostatin. Dissolve in dimethyl sulfoxid (DMSO) and store at −20 °C freezer. For 10 ml of buffer add 30 μl of chymostatin stock. 5. Leupeptin stock solution: 21 mM (10 mg/ml) leupetin. Dissolve in Milli-Q water and store at −20 °C freezer. For 10 ml of buffer add 10 μl of leupeptin stock. 6. Aprotinin stock solution: 1.5 mM (10 mg/ml) aprotinin. Dissolve in Milli-Q water and store stock at −20 °C freezer. For 10 ml of buffer add 10 μl of aprotinin stock. 7. Pepstatin stock solution: 1.45 mM (1 mg/ml) pepstatin. Dissolve in ethanol and store at −20 °C freezer. For 10 ml of buffer add 10 μl of pepstatin. 8. E64 stock solution: 2.8 mM (1 mg/ml) E64. Dissolve in Milli-Q water and store at −20 °C freezer. For 10 ml of buffer add 10 μl of E64 stock. 9. Sodium butyrate stock solution: 25 mM (2.77 mg/ml) sodium butyrate. Dissolve in Milli-Q water and store at −20 °C freezer. For 10 ml of buffer add 200 μl of sodium butyrate stock (see Note 3). 10. Fixation buffer: 400 mM sucrose, 10 mM Tris–HCl pH 8.0, 0.05 % Triton X-100, 1 % formaldehyde (see Note 4 and Table 1). 11. Glycine buffer: 2 M glycine (see Note 4 and Table 1). 12. Extraction buffer 1: 400 mM sucrose, 10 mM Tris–HCl pH 8.0, 1 mM EDTA, 5 mM β-mercaptoethanol, protease inhibitors, histone deacetylase inhibitor (see Note 5 and Table 1). 13. Extraction buffer 2: 250 mM sucrose, 10 mM Tris–HCl pH 8.0, 1 mM EDTA, 10 mM MgCl2, 1 % Triton X-100, 5 mM β-mercaptoethanol, protease inhibitors, histone deacetylase inhibitor (see Note 5 and Table 1). 14. Nuclei lysis buffer: 50 mM Tris–HCl pH 8.0, 10 mM EDTA, 1 % SDS, protease inhibitors, histone deacetylase inhibitor (see Note 6 and Table 1). 15. ChIP dilution buffer: 15 mM Tris–HCl pH 8.0, 150 mM NaCl, 1 mM EDTA, 1.1 % Triton X-100, protease inhibitors, histone deacetylase inhibitor (see Note 6 and Table 1).

ChIP-Seq Analysis of Histone Modifications at the Core…

61

Table 1 List of ChIP buffers ChIP buffer

Reagents

Fixation buffer (250 ml)

0.4 M Sucrose: 34.225 g 10 mM Tris–HCl pH 8.0: 2.5 ml (1 M stock) 0.05 % Triton X-100: 625 μl (20 % stock) 1 % formaldehyde: 6.8 ml (37 % stock)

Glycine buffer (100 ml) Glycine 2 M: 15 g Extraction buffer 1 (200 ml)

0.4 M Sucrose: 27.34 g 10 mM Tris–HCl pH 8.0: 2 ml (1 M stock) 1 mM EDTA: 400 μl (0.5 M stock) 5 mM β-mercaptoethanol: 70 μl (14 M stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

Extraction buffer 2 (10 ml)

0.25 M Sucrose: 0.85 g 10 mM Tris–HCl pH 8.0: 100 μl (1 M stock) 1 mM EDTA: 20 μl (0.5 M stock) 10 mM MgCl2: 100 μl (1 M stock) 1 % Triton X-100: 500 μl (20 % stock) 5 mM β-mercaptoethanol: 3.5 μl (14 M stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

Nuclei lysis buffer (50 ml)

50 mM Tris–HCl pH 8.0: 2.5 ml (1 M stock) 10 mM EDTA: 1 ml (0.5 M stock) 1 % SDS: 2.5 ml (20 % stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

ChIP dilution buffer (50 ml)

15 mM Tris–HCl pH 8.0: 750 μl (1 M stock) 150 mM NaCl: 1,670 μl (5 M stock) 1 mM EDTA: 120 μl (0.5 M stock) 1.1 % Triton X-100: 2,750 μl (20 % stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

Low-salt wash buffer (50 ml)

150 mM NaCl: 1.5 ml (5 M stock) 0.1 % SDS: 250 μl (20 % stock) 1 % Triton X-100: 2.5 ml (20 % stock) 2 mM EDTA: 200 μl (0.5 M stock) 20 mM Tris–HCl pH 8: 1 ml (1 M stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3) (continued)

62

Jordi Malapeira and Paloma Mas

Table 1 (continued) ChIP buffer

Reagents

High-salt wash buffer (50 ml)

500 mM NaCl: 5 ml (5 M stock) 0.1 % SDS: 250 μl (20 % stock) 1 % Triton X-100: 2.5 ml (20 % stock) 2 mM EDTA: 200 μl (0.5 M stock) 20 mM Tris–HCl pH 8.0: 1 ml (1 M stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

LiCl wash buffer (50 ml)

0.25 M LiCl: 3,125 μl (4 M stock) 1 % NP-40: 2.5 ml (20 % stock) 1 % sodium deoxycholate: 0.5 g 1 mM EDTA: 100 μl (0.5 M stock) 10 mM Tris–HCl pH 8.0: 0.5 ml (1 M stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

TE buffer (50 ml)

10 mM Tris–HCl pH 8.0: 0.5 ml (1 M stock) 1 mM EDTA: 100 μl (0.5 M stock) Protease inhibitors: 1 mM PMSF, 0.01 mM antipain, 0.01 mM chymostatin, 0.02 mM leupeptin, 1.45 μM pepstatin, 1.5 μM aprotinin, 2.8 μM E64 Histone deacetylase inhibitor: 25 μM sodium butyrate (see Note 3)

Elution buffer (20 ml)

1 % SDS: 1 ml (20 % stock) 0.1 M NaHCO3: 0.168 g

16. Low-salt wash buffer: 150 mM NaCl, 0.1 % SDS, 1 % Triton X-100, 2 mM EDTA, 20 mM Tris–HCl pH 8.0, protease inhibitors, histone deacetylase inhibitor (see Note 6 and Table 1). 17. High-salt wash buffer: 500 mM NaCl, 0.1 % SDS, 1 % Triton X-100, 2 mM EDTA, 20 mM Tris–HCl pH 8.0, protease inhibitors, histone deacetylase inhibitor (see Note 6 and Table 1). 18. LiCl wash buffer: 250 mM LiCl, 1 % Nonidet-P40, 1 % Sodium deoxycholate, 1 mM EDTA, 10 mM Tris–HCl pH 8.0, protease inhibitors, histone deacetylase inhibitor (see Note 6 and Table 1). 19. TE wash buffer: 10 mM Tris–HCl pH 8.0, 1 mM EDTA, protease inhibitors, histone deacetylase inhibitor (see Note 6 and Table 1). 20. Elution buffer: 1 % SDS, 0.1 M NaHCO3 (see Note 7 and Table 1).

ChIP-Seq Analysis of Histone Modifications at the Core…

3

63

Methods

3.1 Sterilization, Stratification, and Germination of Arabidopsis thaliana Seeds

1. Sterilize the Arabidopsis seeds by placing them in 1.5 ml tubes. 2. Open the seeds tubes and put them inside a desiccator jar together with a 250 ml beaker containing 35 ml of bleach. 3. Carefully add 2.5 ml of HCl (37 %) into the beaker and immediately seal the desiccator jar with Parafilm. 4. Allow sterilization for at least 4 h. 5. Prepare the MS plates and place a sterile filter paper on top to avoid medium contamination when collecting the seedlings. 6. Spread the sterile seeds onto the filter paper, seal the plates with micropore tape and wrap them with aluminum foil. 7. Place the plates in the fridge at 4 °C for 4 days. 8. Incubate the plates under the appropriated experimental conditions for 2 weeks in a growth chamber. For the particular assays described in this article, the experimental conditions are 12:12 light–dark (LD) cycles at 22 °C.

3.2 Fixation of Plant Material

1. Collect 1 g of 2-week-old seedlings for each time point (ZT3 and ZT15) (see Note 8). Add the seedlings into a 200 ml bottle containing 60 ml of fixation buffer (see Note 4 and Table 1). 2. Vacuum-infiltrate for 5 min at room temperature. 3. Add 15 ml of glycine buffer to stop the cross-linking reaction, mix well, and vacuum-infiltrate for additional 10 min (see Note 4 and Table 1). 4. Rinse the seedlings three times with 50 ml of Milli-Q water. 5. Dry the seedlings thoroughly with paper towels. 6. Wrap the fixed material with foil and immediately freeze in liquid nitrogen. 7. The frozen samples can be directly used for chromatin isolation or stored at −80 °C freezer until further use.

3.3 Chromatin Isolation

1. Cool the mortar by adding liquid nitrogen and grind the frozen seedlings to a fine powder. 2. Suspend the powder in 25 ml of cold extraction buffer 1 in 50 ml tubes (see Note 5 and Table 1). 3. Filtrate the suspension through two layers of miracloth into new 50 ml tubes. 4. Centrifuge at 2,900 × g for 15 min at 4 °C. 5. Remove the supernatant, suspend the pellet in 2 ml of cold extraction buffer 2 and transfer to 2 ml tubes (see Note 5 and Table 1). 6. Centrifuge at 12,000 × g for 10 min at 4 °C.

64

Jordi Malapeira and Paloma Mas

7. Remove the supernatant and suspend the pellet in 1 ml of cold nuclei lysis buffer (see Note 6 and Table 1). 8. Sonicate the chromatin to a DNA size of approximately 400– 800 bp. The sonication settings and the duration should be determined experimentally because it changes considerably depending on the sonication devices (see Notes 9 and 10). 9. Centrifuge the sonicated chromatin at 16,000 × g for 10 min and transfer supernatant into new 1.5 ml tubes. 10. The supernatant can directly be used for immunoprecipitation or can be shock frozen in liquid nitrogen and stored at −80 °C until further use. 3.4 Immunoprecipitation

1. For the immunoprecipitation step, label 1.5 ml tubes with the name of your samples including always a negative control (no antibody). For example: ZT3 negative control, ZT3 H3K56ac, ZT3 H3K4me3, ZT15 negative control, ZT15 H3K56ac, ZT15 H3K4me3. 2. Add 100 μl of sepharose beads into each tube containing 900 μl of cold ChIP dilution buffer to equilibrate the beads. 3. Centrifuge at 2,000 × g for 5 min at 4 °C and discard the supernatant. 4. Add 900 μl of cold ChIP dilution buffer, 100 μl of the sonicated chromatin and 2 μl of the antibody (anti-H3K56ac or anti-H3K4me3) for each condition. As mentioned above, a tube without antibody should be used as a negative control. 5. Incubate overnight on a rotator at 4 °C. 6. Wash the sepharose bead pellet with 900 μl of each wash buffer (see the list below). For each buffer, incubate the beads with the buffer for 5 min on a rotator at 4 °C, then centrifuge for 5 min at 2,000 × g at 4 °C, discard the supernatant, and add the next buffer. (a) Low-salt wash buffer (see Note 6 and Table 1). (b) High-salt wash buffer (see Note 6 and Table 1). (c) LiCl wash buffer (see Note 6 and Table 1). (d) TE wash buffer (see Note 6 and Table 1). (e) TE wash buffer (see Note 6 and Table 1).

3.5 Elution and Reverse Cross-linking

1. Release the bead-bound complexes by adding 300 μl of fresh elution buffer to the sepharose beads (see Note 7 and Table 1). 2. Vortex briefly to mix and incubate at 65 °C for 1 h with 0.2 × g agitation. 3. Centrifuge at 16,000 × g for 5 min and transfer the supernatant into new 1.5 ml tubes.

ChIP-Seq Analysis of Histone Modifications at the Core…

65

Table 2 List of primers used in Fig. 1 Name

Sequence

CCA1-F

TCC AGA TAA GAA GTC ACG CTC AGA

CCA1-R

CAT TAA GCC AAT GAA GAT GAG AAC A

LHY-F

AAT CTA AAG AGG TTA TCA CAA CGG C

LHY-R

GCT GCT TCA AAT CCT CTC TAA CAA G

PRR9-F

TCT CGG TAG ATT AAG ATC TAA AGC TCG TTG

PRR9-R

CAA CAC TTG GTA AAA CCA ACA AAG CCT A

PRR7-F

GCA ATA ATC GAA ATT AGG GTT TAT GGC T

PRR7-R

TTA GCA TTC ATC ACA CCA ACT CTG CTT

LUX-F

AGC TTC GAA GAG CTC AAT CTC TAA CTG AA

LUX-R

TCG TAA TCG CTC ATT TGT ACT TCC TCT C

TOC1-F

TGT TAA GGG GAT AAA TTA GGC GAC

TOC1-R

GCT ATG ATA CTT CCA TGG CCA AA

AT5G55480-F

GAT TCT GCT TCT CAC CAA

AT5G55480-R

ATT CAG CAA TAG CCA CAA

4. Add 13 μl of 5 M NaCl to the eluate and incubate overnight at 65 °C to revert the cross-linking. 5. To prepare the input chromatin, label two 1.5 ml tubes as ZT3 input and ZT15 input. Add 100 μl of the sonicated chromatin, 300 μl of elution buffer and 13 μl of 5 M NaCl into each tube and incubate overnight at 65 °C. 6. Purify the DNA using a DNA purification kit (see Note 11). 7. The immunoprecipitated and purified DNA is ready to be used in qPCR reactions to amplify the examined target sequences (see Note 12). 3.6 Real-time Quantitative PCR

Use 1 μl of the purified DNA sample to amplify by qPCR the regions of interest. Amplification of a control gene (if known) is also required to normalize the data (see Note 13). Take into account that each experimental condition must be analyzed by triplicate. In the particular experiment described in this report we use the primers listed in Table 2 to analyze the 5′ end region of CCA1, LHY, PRR9, PRR7, LUX, and TOC1, we also use AT5G55480 as a negative control gene (see Fig. 1).

Jordi Malapeira and Paloma Mas

a Relative ChIP abundance

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Ab

b

ZT 3 ZT 15

H3K56Ac

- + - + - + - + - + - + - + - + - + - + - + - + CCA1 LHY PRR9 PRR7 LUX TOC1

4.5

Relative ChIP abundance

66

4 3.5

ZT 3 ZT 15

H3K4Me3

3 2.5 2 1.5 1 0.5 0 Ab - + - + - + - + - + - + - + - + - + - + - + - + CCA1 LHY PRR9 PRR7 LUX TOC1

Fig. 1 H3K56ac and H3K4me3 accumulation at the core of the circadian clock. ChIP assays of H3K56ac (a) and H3K4me3 (b) were performed in WT plants analyzed at ZT3 (white) and ZT15 (black ). Primers listed in Table 1 were used to amplify the promoter region of CCA1, LHY, PRR9, PRR7, LUX, and TOC1. Values were normalized to the control gene AT5G55840 and represented as means ± SEM

To perform the qPCR analysis, proceed with the following steps: 1. Prepare a mix containing 1 μl of DNA, 10 μl of qPCR Master Mix 2×, and 7 μl of Milli-Q water for each reaction. 2. Add 18 μl of the previous mix to the corresponding well in a 96-well plate. 3. Prepare a mix containing the forward and reverse primers to a final concentration of 5 μM each. 4. Add 2 μl of the primer mix to the corresponding well. 5. Seal the 96-well plate with transparent film, briefly vortex and spin the plate. 6. Run the qPCR program (see Notes 14 and 15). 7. Normalize the data against the input chromatin and the control gene (see Note 16).

ChIP-Seq Analysis of Histone Modifications at the Core…

4

67

Notes 1. For the experiments described in this article, we used the Gammabind Plus Sepharose beads from GE Healthcare. 2. We also used the H3K56ac and the H3K4me3 antibodies from Millipore. 3. Sodium butyrate is only required to analyze histone acetylation modifications to avoid histone deacetylation. 4. Prepare the buffer 1 day in advance and keep it at 4 °C with agitation overnight. 5. The buffer must be freshly prepared and used at 4 °C. Protease inhibitors and histone deacetylase inhibitor must be added just before use. 6. The buffer must be used at 4 °C. Protease inhibitors and histone deacetylase inhibitor must be added just before use. 7. The buffer must be freshly prepared. 8. If you want to perform a ChIP-seq more material will be required. We recommend using five samples for each time point, process them in parallel, and finally elute all five samples in the same 60 μl of Milli-Q water. Also the use of non-stick tubes is advisable to prevent loss of DNA sample. 9. In our assays, we use a Branson Digital Sonifier and we sonicate the samples 6 times using the following settings: 10 % amplitude, 0.5 s pulse on, 1 s pulse off. 10. If you want to perform ChIP-seq sonicate the chromatin to a DNA size of approximately 100–500 bp. If you are using a Bioruptor Next Generation we recommend sonicating for 2 h using the following settings: high position, 30 s pulse on, 30 s pulse off. We also recommend using TPX tubes in order to improve the sonication. 11. We routinely use a QIAGENE gel extraction kit and the following protocol is used: (a) Add 960 μl of QG buffer to the eluates, mix well, and transfer into the QIAGEN columns. (b) Centrifuge for 1 min at 16,000 × g and discard flow-through. (c) Add 750 μl of PE buffer. (d) Centrifuge for 1 min at 16,000 × g and discard flow-through. (e) Centrifuge for 1 min at 16,000 × g and transfer the columns into new 1.5 ml tubes. (f) Add 60 μl of Milli-Q water into the columns. (g) Centrifuge for 1 min at 16,000 × g.

68

Jordi Malapeira and Paloma Mas

12. If you want to perform ChIP-seq check that the size of the DNA fragments is correct with a high sensitivity bioanalyzer. The DNA size should be of approximately 100–500 bp. 13. In this particular experiment, we used AT5G55480 as a control gene. We have verified that AT5G55480 does not show changes at its promoter in H3K56ac and H3K4me3 accumulation during the circadian cycle. 14. To perform the qPCR analysis described in this article we use a LightCycler 480 II from Roche. 15. The following qPCR program is used in this particular experiment: Taq activation: 95 °C 1 min, PCR: (45×) 95 °C 10 s, 60 °C 30 s, melting curve: 95 °C 2 min. 16. The Cp values obtained in the qPCR reaction must be normalized to their input Cp value. The normalized data indicates the ratio between immunoprecipitated DNA and the initial amount of DNA. Data can be also normalized relative to a control gene that does not change in the particular experimental conditions. This last normalization step corrects technical problems such as pipetting errors when preparing the qPCR mix.

Acknowledgements The work in our laboratory is supported by grants from the Ramón Areces Foundation, the Spanish Ministry of Science and Innovation (MICINN) (BIO2010-16483), the EMBO YIP program and from EUROHORCS (European Heads of Research Councils) and the European Science Foundation (ESF) through the EURYI Award to Paloma Mas. We thank Rossana Henriques for critical comments on the manuscript. References 1. Bannister AJ, Kouzarides T (2011) Regulation of chromatin by histone modifications. Cell Res 21:381–395 2. Das PM, Ramachandran K, vanWert J et al (2004) Chromatin immunoprecipitation assay. Biotechniques 37:961–969 3. Collas P, Dahl JA (2008) Chop it, ChIP it, check it: the current status of chromatin immunoprecipitation. Front Biosci 13:929–943 4. Kim TH, Barrera LO, Zheng M et al (2005) A high-resolution map of active promoters in the human genome. Nature 436:876–880 5. Lee TI, Jenner RG, Boyer LA et al (2006) Control of developmental regulators by Polycomb in human embryonic stem cells. Cell 125:301–313

6. Kharchenko PV, Tolstorukov MY, Park PJ (2008) Design and analysis of ChIP-seq experiments for DNA-binding proteins. Nat Biotechnol 26:1351–1359 7. Valouev A, Johnson DS, Sundquist A et al (2008) Genome-wide analysis of transcription factor binding sites based on ChIP-Seq data. Nat Methods 5:829–834 8. Jenuwein T, Allis CD (2001) Translating the histone code. Science 293:1074–1080 9. Strahl BD, Allis CD (2000) The language of covalent histone modifications. Nature 403: 41–45 10. Bowler C, Benvenuto G, Laflamme P et al (2004) Chromatin techniques for plant cells. Plant J 39:776–789

ChIP-Seq Analysis of Histone Modifications at the Core… 11. Orlando V (2000) Mapping chromosomal proteins in vivo by formaldehyde-crosslinkedchromatin immunoprecipitation. Trends Biochem Sci 25:99–104 12. Solomon MJ, Varshavsky A (1985) Formaldehydemediated DNA-protein crosslinking: a probe for in vivo chromatin structures. Proc Natl Acad Sci U S A 82:6470–6474

69

13. Reneker JS, Brotherton TW (1991) Discrete regions of the avian beta-globin gene cluster have tissue-specific hypersensitivity to cleavage by sonication in nuclei. Nucleic Acids Res 19:4739–4745 14. Haring M, Offermann S, Danker T et al (2007) Chromatin immunoprecipitation: optimization, quantitative analysis and data normalization. Plant Methods 3:11

Chapter 5 Quantitative Transcriptome Analysis Using RNA-seq Canan Külahoglu and Andrea Bräutigam Abstract RNA-seq has emerged as the technology of choice to quantify gene expression. This technology is a convenient accurate tool to quantify diurnal changes in gene expression, gene discovery, differential use of promoters, and splice variants for all genes expressed in a single tissue. Thus, RNA-seq experiments provide sequence information and absolute expression values about transcripts in addition to relative quantification available with microarrays or qRT-PCR. The depth of information by sequencing requires careful assessment of RNA intactness and DNA contamination. Although the RNA-seq is comparatively recent, a standard analysis framework has emerged with the packages of Bowtie2, TopHat, and Cufflinks. With rising popularity of RNA-seq tools have become manageable for researchers without much bioinformatical knowledge or programming skills. Here, we present a workflow for a RNA-seq experiment from experimental planning to biological data extraction. Key words Next-generation sequencing, RNA-seq, Gene expression quantification, Circadian rhythm

1

Introduction High-throughput RNA-sequencing (RNA-seq) enables the researcher to quantify gene expression, discover new splice variants and (polyadenylated) transcripts within a single assay. It has been successfully applied to a variety of plants [1–3]. Unlike an analysis with microarrays, RNA-seq detected differential splicing events along the developmental gradient in maize [2]. Knowledge about absolute expression levels provided by RNA-seq enables the identification of abundant genes in contrast to just identifying relative changes [4, 5]. RNA-seq data can be combined with microarrays to both identify abundant genes and to characterize their relative expression pattern [6]. Compared to microarrays the RNA-seq experiment poses unique challenges. The RNA-seq experiment starts with planning a suitable strategy. After sampling total RNA is isolated from the samples in sequencing grade quality and contaminating DNA is removed. RNA is converted to cDNA libraries for sequencing.

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_5, © Springer Science+Business Media New York 2014

71

72

Canan Külahoglu and Andrea Bräutigam

Fig. 1 Overview of RNA-seq data analysis and biological information extraction. Software used in this review is written in parentheses

After the wet-lab procedure of preparing RNA and the sequencing itself, analysis of the sequenced reads requires a bioinformatic pipeline of read cleaning, read mapping, statistics, and biological interpretation. In this protocol we introduce a pipeline, which requires minimal prior bioinformatic knowledge, since it relies on public domain software packages. A selection of possible methods to extract biological data from the experiment is also presented (summarized in Fig. 1).

2

Materials

2.1 Sampling of Plant Material

1. Liquid nitrogen. 2. Dewar. 3. Scissors or scalpel. 4. Aluminum foil bags or tubes with the lid punctured.

2.2

Grinding

1. Mortar. 2. Pestle. 3. Styrofoam grinding platform. For easier grinding, the lid of a conventional Styrofoam shipping box can be hollowed out to accommodate the mortar, thus relieving the person of grinding by holding the mortar in place.

RNA-seq

73

4. Liquid nitrogen transfer vessel. To pre-cool the mortar efficiently and to transfer fresh liquid nitrogen during grinding, a transfer vessel can be fashioned out of a plastic 10 mL pipet and a 50 mL tube using tape. RNA Extraction

In order to have optimal sequencing results it very important to avoid RNA degradation by RNAses (see Note 1). When preparing solutions for the described protocols, make sure to use RNAse-free water (see Note 2).

2.3.1 RNA Extraction Method I: Guanidinium Isothiocyanate Extraction Method

This standard method is modified from [7, 8]. It delivers around 200 ng/µl clean total RNA per 100 mg fresh weight from tissues, which do not contain high phenolic, carbohydrate, or lipid contents. The main chaotropic reagent is guanidinium isothiocyanate, which effectively disrupts tissues and inactivates RNAses, while keeping the RNA intact.

2.3

1. Water-saturated phenol: dissolve 100 g phenol crystals in distilled water at 65 °C. Aspirate the upper water phase and store up to 1 month at 4 °C (see Note 3). Acidic phenol is also available commercially. 2. 1 N acetic acid: dilute 2.8 mL of 17.4 M acetic acid with 47.2 mL DEPC-treated water. 3. 3 M sodium acetate: dissolve 20.41 g sodium acetate trihydrate (Mr = 136.08 g/mol) in 30 mL water, adjust the pH to 6.0 with 3 M acetic acid, fill up with water to the final volume of 50 mL, add 0.1 % DEPC, and sterilize by autoclaving. 4. 5 M lithium chloride: dissolve 10.59 g lithium chloride (Mr = 42.39 g/mol) in 50 mL water, add 0.1 % DEPC, and sterilize by autoclaving. 5. RNase-ALL stock solution: 4 M Guanidine isothiocyanate, 25 mM Na-acetate, 0.5 % (w/v) N-laurosylsarcosine, 0.7 % (v/v) mercaptoethanol, pH 7.0. Dissolve 47.2 g guanidinium isothiocyanate (Mr = 118.16 g/mol) and 0.5 g N-laurosylsarcosine sodium salt (Mr = 293.4 g/mol) in 30 mL water at 65 °C. Then add 830 μL 3 M Sodium acetate (pH 5.2). Adjust the pH to 7.0 with 2 N NaOH and fill up with water to the final volume to 100 mL. 6. Chloroform–Isoamyl alcohol 24/1 (v/v) solution: mix 24 mL of chloroform with 1 mL of isoamyl alcohol. 7. DEPC-water: mix H2O with 0.1 % DEPC, stir and autoclave [alternative: purchase RNAse- and DNAse-free water]. 8. 80 % p.a. EtOH.

74

Canan Külahoglu and Andrea Bräutigam

2.3.2 RNA Extraction Method II: Modified CTAB Protocol with Silica Columns

This protocol is recommended for recalcitrant plant tissues with high secondary metabolites, starch, and lipid composition. It is a cetyltrimethylammonium bromide (CTAB)-based method in combination with silica columns of RNAeasy Plant Mini kit (Qiagen, Germany) for effective RNA isolation [9]. 1. CTAB-PVP buffer: 2 % (w/v) CTAB, 2 % (w/v) Polyvinylpyrrolidone (PVP-40), 100 mM Tris–HCl (pH 8.0), 25 mM EDTA, 2 M NaCl. Add 2 % β-mercaptoethanol (BME) before usage. 2. Chloroform–Isoamyl alcohol 24/1 (v/v) solution: mix 24 mL of chloroform with 1 mL of isoamyl alcohol. 3. 96 % EtOH: use fresh p.a. EtOH. 4. Silica column-based commercial kit (e.g., Qiagen RNeasy Plant Kit, Qiagen, Germany).

2.4 RNA Quality Control

1. Photometer or NanoDrop (Thermo Scientific). 2. Access to Bioanalyzer (Agilent). 3. PCR reagents (PCR-machine, PCR-microtubes, dNTPs, TAQ-Polymerase, Primer, suitable reaction buffer).

2.5

Sequencing

2.6

Read Analysis

Presumed to be outsourced to intramural or commercial supplier. 1. Sequencing reads in .FASTQ file format. 2. Standard desktop computer >8 GB RAM, Linux environment preferred. 3. Programs of Table 1 installed.

2.7 Read Mapping and Quantification

1. Cleaned reads from Subheading 3.7. 2. Fasta file of the reference sequence. 3. .gff/.gtf file of annotation for the reference sequence. 4. Standard desktop computer with Linux-based operating system and >2 GB RAM. 5. Programs from Table 2 installed. Table 1 Overview of read analysis and cleaning software recommended in this review Program

Version

Environment

Source

FASTQC

v0.10.1

Install all dependencies

www.bioinformatics. babraham.ac.uk/projects

FASTX

v0.0.13

Install all dependencies

http://hannonlab.cshl.edu/ fastx_toolkit/

RNA-seq

75

Table 2 Overview of software needed for read mapping and expression statistics Program

Version

Environment

Source

BAMtools

v1.0.2

Install all dependencies

https://github.com/pezmaster31/bamtools

Bowtie

v0.12.8

Install all dependencies

http://bowtie-bio.sourceforge.net

Cufflinks

v2.0.2

Install all dependencies

http://cufflinks.cbcb.umd.edu/downloads/

TopHat

v1.4.1

Install all dependencies

http://tophat.cbcb.umd.edu/downloads/

2.8 Extracting Biological Information

1. Tab delimited text file (or Excel File) containing the gene identifier in the first column and the resulting read counts in subsequent columns with one header row. 2. Standard desktop computer system with >8 GB RAM (for hierarchical clustering analysis of genes >64 GB RAM). 3. Microsoft Office Excel. 4. MapMan Software. 5. MultiExperiment Viewer Software. 6. VirtualPlant Web access.

3

Methods Designing a RNA-seq experiment is similar in principle to designing a microarray experiment. Each sampling point needs to be analyzed in biological replicates. The statistical methods require replicates to assess variation prior to determination of differentially expressed genes. While some tools estimate variation based on different samples (and use this assessment to guide their decisions), most tools require replication for proper assessment. The large number of reads is no substitute for replication. For some RNAseq technologies such as Illumina HiSeq 2000 samples can be multiplexed up to 12-fold, that is, up to 12 samples can be barcoded and run on a single lane enabling replication at almost the same sequencing cost. Sampling time points for circadian experiments will depend on the analysis to be done post-sequencing and hence the individual experiment. If a model with predictive qualities or transcription factor network analysis is the goal, sampling time points should be defined with the help of the statistician or bioinformatician who will partake in the modeling effort. Finally, the strategy ought to be aligned with the available computer power. RNA-seq experiments create vast amounts of data to be stored and analyzed and hence require the infrastructure to support the experiment.

76

Canan Külahoglu and Andrea Bräutigam

3.1 Sampling Plant Material for RNA-Seq

1. Prepare the aluminum foil bags or puncture the lids of tubes. 2. Set up a dewar with liquid nitrogen. 3. Cut off the tissue, transfer it to the bag or tube and immediately shock-freeze. Any delay in freezing or thawing after initial freezing will lead to RNA degradation. 4. Either proceed to grinding immediately or store at −80 °C.

3.2 Grinding the Samples

1. Set up the mortar and pestle for grinding. 2. Set up the dewar with clean liquid nitrogen. 3. Pre-cool the mortar and pestle by twice evaporating liquid nitrogen. 4. Grind the tissues under liquid nitrogen to a fine powder. Alternative grinding methods such as bead mills are suitable if the samples remain deep frozen at all times. 5. Store powder at −80 °C or proceed immediately to RNA extraction.

3.3

RNA Extraction

3.3.1 Method I: Guanidinium ThiocyanateBased Extraction

We will describe two alternative RNA extraction methods. The guanidinium isothiocyanate-based extraction method delivers sequencing quality RNA for most tissue types. For tissues with a high secondary-metabolite, carbohydrate, and fatty-acid content such as seed tissue, a CTAB-based extraction is recommended. Conduct all procedures at room temperature unless specified otherwise. Keep RNA samples on ice during all processes. When working with toxic, volatile chemicals, such as chloroform and phenol, work under a fume hood. 1. Pre-cool centrifuge to 10 °C. 2. Prepare RNase-ALL working solution by mixing acidic phenol with RNAase-ALL solution 1:1 (v:v). When using phenol, make sure the phenolic phase beneath the aqueous phase is used. 3. Homogenization: Transfer 100 mg of ground plant material to 2 mL microtube and add immediately 1 mL RNase-ALL working solution (add 1 mL RNase-ALL working solution per 100 mg fresh tissue) and vortex tube 5 s. Incubate tubes on ice for at least 15 min, while vortexting sample every 5 min for 5 s. 4. Protein extraction: Add 300 μL of chloroform–isoamyl alcohol 24:1 (v/v) solution and vortex tube for 10 s. Incubate on ice for at least 5 min. Centrifuge at 10,000 × g for 10 min at 10 °C. Transfer the upper aqueous phase (~700 μL to 1 mL) into a fresh 2 mL microtube. Add 700 μL water-saturated Phenol solution

RNA-seq

77

(acidic phenol). Invert several times and add subsequently 300 μL of chloroform–isoamyl alcohol 24:1 (v/v) solution to remove residual phenol. Vortex for 10 s and centrifuge at 10,000 × g for 10 min at 10 °C. Transfer the upper aqueous phase (~700 μL to 1 mL) into a fresh micro 2 mL tube. Add 300 μL of chloroform– isoamyl alcohol 24:1 (v/v) solution to remove residual phenol. 5. First RNA precipitation: Transfer the upper (aqueous) phase (~700 μL) containing the extracted RNA into a fresh 1.5 mL tube. Add 1/20 volume of 1 N acetic acid (~35 μL), invert several times, add 0.7–1.0 volume of 100 % ethanol (~490–700 μL) and vortex. Centrifuge at 10,000 × g for 20 min at 10 °C and carefully remove the supernatant. Resuspend the RNA pellet with 1 mL of 3 M sodium acetate (pH 6.0). Centrifuge at 12,000 × g for 10 min at 10 °C and remove the supernatant. Add 1 mL of 80 % ethanol and resuspend the RNA pellet. Centrifuge at 12,000 × g for 10 min at 10 °C and remove the supernatant. Add again 1 mL of 80 % ethanol and resuspend the RNA pellet. Centrifuge at 12,000 × g for 10 min at 10 °C and remove the supernatant. Air-dry the pellet for 10–15 min at RT (set the tube upside down on a paper towel). Resuspend the pellet in 500 μL RNase-free water (in case of pooling: resuspend in 250 μL RNase-free water and combine after heat treatment). 6. Second precipitation: Incubate at 60 °C for 5 min to ensure complete solubilization. Centrifuge at 12,000 × g for 1 min at 10 °C. Transfer supernatant into a fresh tube (combine your two samples). Add an equal volume of 5 M lithium chloride (~500 μL), mix well, and incubate overnight at 4 °C. Centrifuge at 12,000 × g for 10 min at 10 °C and remove the supernatant. Resuspend the pellet of RNA with 1 mL 80 % ethanol. Centrifuge at 12,000 × g for 10 min at 10 °C and remove the supernatant. Resuspend the pellet again with 1 mL 80 % ethanol. Centrifuge at 12,000 × g for 10 min at 10 °C and remove the supernatant. Air-dry the pellet for 10–15 min at RT (set the tube upside down on a paper towel). Add 50 μL DEPCtreated water. Incubate at 60 °C for 10–15 min to ensure complete solubilization. Centrifuge shortly. 7. Store RNA at −80 °C for long term storage or on ice if continuing with Subheading 3.4. 3.3.2 Method II: Modified CTAB Protocol with Silica Columns

If the material is recalcitrant during standard RNA isolation, a modified CTAB protocol combined with silica columns may yield intact, clean RNA: 1. Prepare CTAB buffer working solution: 50 % (v/v) CTAB buffer stock solution, 2 % (v/v) BME, and 50 % (v/v) acidic phenol in a 50 mL polypropylene tube (Falcon). Pre-heat the CTAB working solution to 65 °C in water bath and make sure not to close the tube completely to prevent phenol spills.

78

Canan Külahoglu and Andrea Bräutigam

2. Homogenization: Add 1 mL pre-heated CTAB buffer working solution per 100 mg frozen ground tissue in 2 mL microtube. Vortex tube for 5 s and be careful opening tube letting gases evaporate. Incubate samples in water bath at 65 °C for 15–30 min. Vortex samples every 5 min to help tissue disruption. 3. Protein extraction: Add an equal volume of chloroform–isoamyl alcohol (24:1) to each sample and vortex each sample for at least 10 s. Centrifuge samples at 10,000 × g at 10 °C for 20 min. Transfer the aqueous (upper) supernatant (1 mL per tube) to a new 2 mL microtube and add an equal volume of chloroform–isoamyl alcohol (24:1). Vortex samples for 10 s and centrifuge at 10,000 × g 4 °C for 10 min. Take care not to touch the interphase separating the aqueous (upper) and nonaqueous (lower) phase, when transferring the supernatant to a new 1.5 mL tube. 4. RNA precipitation: add to the supernatant 0.5 vol 96 % EtOH and invert tube immediately. Load supernatant–ethanol mixture quickly onto RNA binding silica columns (Qiagen RNAeasy plant kit or similar kit; column max volume 0.75 mL). Spin loaded column at 10,000 × g for 30 s. Leftover supernatant–ethanol mixture are loaded onto the same column, processing the entire sample. Follow the kit protocol for the subsequent washing and desalting steps (see Note 4). Furthermore, we recommend performing the on-column DNAse digest available for the Qiagen RNA extraction kit. 5. RNA elution: add 50 μL of RNAse-free water on column and incubate sample for 1 min at room temperature. Put column in fresh RNAse-free tube and spin it for 1 min at 10,000 × g. Store RNA on ice or at −80 °C until continuing with Subheading 3.4. 3.4 Quality Assessment of RNA

1. Determine the RNA concentration, the protein contamination and the carbohydrate and ethanol contamination using a spectrophotometer. The 260/230 nm absorbance represents carbohydrate or ethanol contaminations and the 260/280 nm represents protein contamination. Clean RNA ranges above 1.8–2.0 in both ratios. Values of the 260/230 nm and 260/280 nm ratios below 1.6 indicate a contamination of the extraction and it is not recommended to use this material for library preparation and subsequent sequencing. 2. Test the integrity of the RNA using a 2100 Bioanalyzer (Agilent) [10]. The electropherograms with intact RNA should show clear distinct peaks for the 28S and 18S ribosomal RNA and almost no peaks in the smaller RNA size range. Intact RNA is also indicated by the RNA Integrity Number (RIN), which describes numerically the overall RNA quality beyond the standard method of the ratio of the two ribosomal peaks [11].

RNA-seq

79

Fig. 2 Bioanalyzer RNA 6000 Nanochip electropherograms of RNA samples with different qualities. Upper panel shows non-degraded high-quality RNA with characteristic 18 and 28s rRNA peaks before DNAseI treatment with gDNA peak indicated by arrow (a) and after the digest (b) with a RIN of 9. Lower panel shows two samples of degraded low concentrated RNA (c, d) with a RIN of below 3

For successful sequencing library preparation, it is recommended not to use RNA samples with a RIN below 8. On the RNA 6000 Nanochip 1–5 μL of RNA concentrations between 25 and 250 ng/μL can be loaded. A typical electropherogram is shown in Fig. 2. 3. If DNAse treatment of the RNA samples is required dilute the RNA to 100 ng/μL in 20 μL volume and add 2 μL 10× reaction buffer and 1 μL of DNAse. Incubate reaction at 37 °C for 10 min. Inactivate DNAse by adding 2 μL RNAse-free (50 mM) EDTA and incubate tube at 75 °C for 5 min (see Note 5). 4. Test the DNA contamination by PCR on diluted RNA using RNA spiked with DNA as the control. DNA contamination is detected as an additional peak in the electropherogram (Fig. 2a). A PCR using primers designed on genomic DNA on the RNA itself as the template will also detect DNA contamination when compared with a genomic DNA spiked control (see Note 6). 5. Retest the integrity of the RNA after DNAse digest using a 2100 Bioanalyzer chip. 3.5

Sequencing

Different sequencing technologies are currently available that differ in output, read length and price. The preparation protocols for libraries vary depending on the technology. The pipeline described below is customized for Illumina HiSeq 2000 sequencing data.

80

3.6

Canan Külahoglu and Andrea Bräutigam

Read Analysis

3.6.1 Read Quality Assessment

The pipeline for read analysis is modular. Different modules can be replaced with alternative software if desired. The steps are: (1) read quality assessment, (2) read trimming and filtering to remove low quality reads, (3) read mapping, (4) parsing the read mapping to extract quantitative information, (5) statistical analysis and (6) extraction of biological information (Fig. 1). If the experiment was conducted in a species without a sequenced genome, read quantification is prefaced with transcriptome assembly [12, 13]. For the pipeline described below six programs running on the command line are required (Table 1). The programs are well documented and described. Thus, only minimal bioinformatics expertise is required to successfully run the programs (see Note 7). For read quality analysis run FastQC (http://www.bioinformatics. babraham.ac.uk/projects/fastqc/). The FastQC tool delivers a quality assessment of your data. Check read quality to estimate parameter settings for read cleaning: 1. Input: Read output for Illumina data; for each sample read files are divided in four Mio reads sub-files in the FASTQ format. This format also encodes the quality information for each read by the Q score, which is similar to the Phred score known from SANGER sequencing [14, 15]. The Phred quality scoring schemes estimates the error probability of the individual sequenced bases. So a Q score of 30 (Q30) assigned to a base is equivalent to one erroneous base call in 1,000 times. 2. Load data and run analysis via FastQC program with the following terminal command fastqc -f format -o directory read_library. fastq/.fasta FastQC generates its output as HTML file containing all image files mentioned in the section below. For viewing the FastQC output open it in a Web browser. 3. Assess HTML file with regard to the listed attributes of the reads. Figure 3 shows the different quality traits of sequencing data. (a) Per sequence quality: boxplot of base sequence quality generated by FastQC (Fig. 3a). Usually for RNA-seq data the read quality drops toward the sequence end. Bases with a Q score below 20 should be trimmed by the FASTX toolkit. (b) Quality score per sequence: distribution of mean quality score of all bases per sequence, which should display one high peak at the x-axis end, meaning the average Q scores within the reads are equally distributed (Fig. 3b). If you see multiple peaks, your sample has different pools of read quality. Instead of trimming, these reads can be subsequently filtered be the FASTX quality filter.

RNA-seq

81

Fig. 3 FastQC report of Illumina reads. Sequence characteristics used for read quality check

(c) Per base sequence content: plot of base sequence content. All bases should be equally present in a non-biased library. However, it is typical for RNA-seq data to have a base content bias in the first 1–9 bases due to the sequencing primers. A really biased library which should be not used would show peaks for the individual bases that go up to 80 % (Fig. 3c). (d) Per base GC content: similar to the base sequence content plot, though just plotting the GC content. This is biased as well at the sequencing start (1–10th bases), due to the primers used for sequencing (Fig. 3d). (e) Per sequence GC content: distribution of GC content across all reads per sample. This graph should show a normal distribution. The blue curve is a theoretical normal distribution calculated from the mean and standard deviation of the loaded data, the red line plots the actual sample GC distribution. An indicator of sample contamination is a secondary peak in the sample GC distribution. These samples should be analyzed with caution and should be checked for contaminating sequenced DNA by a UNIREF blast (Fig. 3e). (f) Per base N content: in this figure the uncalled bases per library are plotted. For high quality data the line should be flat. Peaks indicate N insertion in the sequence, which can be trimmed or filtered out by the artifact filter from the FASTX toolkit (Fig. 3f). (g) Length distribution: plot of the library length distribution. This should be ranging depending on the number of sequencing cycles around 50 or 100 bp (Fig. 3g).

82

Canan Külahoglu and Andrea Bräutigam

(h) Sequence duplication level: display of library uniqueness. FastQC usually warns the user of RNA-seq data, since sequences occur more than once in a non-normalized RNA-seq library. Over-sequencing of the highly expressed genes is needed to ensure capture of the lowly expressed ones (Fig. 3h). (i) Overrepresented sequences: in this plot the library is analyzed for individual sequences that are overly represented, e.g., adapter primer contamination. For an over-representation tag the sequences have to represent more than 0.1 % of the library (not shown). (j) Kmer Plot/content: this analysis helps the user to spot unusual enrichments of sequences, which are not aligned with the reads, e.g., adapter sequences that start at variable points within the reads (not shown). 3.6.2 Read Cleaning

The FASTX-Toolkit contains a set of command line tools for read file FASTA/FASTQ preprocessing [16]. Satisfactory Illumina sequence runs will have less than 20 % of reads removed by this pipeline. To access all functionalities of each program type in the terminal e.g.: fastx_clipper –h This will print the help screen on the terminal, where all functions and tags are briefly explained. 1. Input: read files from Sequencing Center in .FASTQ format 2. OPTIONAL: If adaptor sequences are present, run the fastx_ clipper to remove them from the reads. fastx_clipper -v –Q33 -a ADAPTER -i-o The –a tag indicates the adapter used for this sample, -i stands for input (.fastq file) and –o for output. Use the –v (verbose) for more explanatory program output. If you have Illumina generated reads use –Q33. Output: The Fastx clipper will deliver a .fastq read file as named by the user. 3. Run the fastx_trimmer to trim bases from the reads, which have an overall low PHRED score and thereby high error rate (Fig. 3a). fastx_trimmer -v -Q33 -f N -l N –i read_ library.fastq –o output_directory/trimmed_ read_library.fastq The flags –f indicate the first base –l the last base to keep from your reads, the –Q33 indicates that you have Illumina NextGen reads.

RNA-seq

83

Output: .fastq file containing trimmed reads. Use this file for the next step. 4. Proceed with the fastx_artifacts_filter to filter out reads with more than three uncalled bases in a stretch (Fig. 3g). fastx_artifacts_filter -v -Q33 -i output_ directory/trimmed_read_library.fastq -o output_directory/filtered_read_library.fastq The flags –i stands for input file and –o for output file. With Illumina data the –Q33 needs to be added to the command. Output: filtered read file in .fastq format, named as indicated by user after –o flag. 5. Run fastq_quality_trimmer to cut bases from the end of the reads depending on the quality threshold set by the user. fastq_quality_trimmer -v –Q33 -t N -l N -i output_directory/filtered_read_library.fastq -o output_directory/filtered2_read_library. fastq It removes also reads, which are shorter than a certain length or which have a lower overall base PHRED score than at 50 % of the bases. Specify with –t the desired PHRED quality threshold of your reads and with –l the minimum read length tolerated by the program. Reads that are shorter than this length will be discarded after trimming. 6. Output: quality trimmed read file in .fastq format, named as indicated by user after –o tag. 7. Concatenate cleaned output fastq files from step 5 to one sample file for each library (see Note 8). cat subreadfiles>output_sample_file 3.7 Read Mapping and Quantification 3.7.1 Read Mapping

Read mapping using TopHat requires the reference genome/ transcriptome, an annotation file in .gff/.gtf format and the read files. TopHat was especially customized for the needs of short RNA-seq reads longer than 75 bp from the Illumina Genome Analyzer. It aligns the reads in two consecutive steps, using first the Bowtie “engine” to map reads that match continuously to the reference sequence followed by a second step which identifies gapped alignment [17, 18]. You can use either paired-end or single reads for a TopHat run, mixing both read types is not supported yet. TopHat delivers several output files. The mapping output file accepted_hits.bam is in the bam format. The splicing junctions reported by TopHat are stored in the junctions.bed file, which is a UCSC BED track (http://genome.ucsc.edu/FAQ/FAQformat. html) of the junctions with each of it consisting of two connected BED blocks. Each BED block is as long as the maximal overhang of reads spanning the junction. The number of alignments spanning the junctions is described by the score. TopHat also reports the

84

Canan Külahoglu and Andrea Bräutigam

insertions and deletion sites of the alignment. These are saved in the insertions.bed and deletions.bed UCSC BED tracks. The unmapped reads are stored in the unmapped_reads.fa.gz file. 1. Create Bowtie index from assembly or reference genome. This is usually a .FASTA file. bowtie-build -f The reference index comprises of several files with the .ebwt file ending. Bowtie employs a technique derived from datacompression called Burrows–Wheeler transformation [19]. 2. Run TopHat with genome .gff/-gtf, index and reads (default settings). tophat -p N -G *.GTF/GFF -o output_directory bowtie_index read_file.FASTA/FASTQ Per default TopHat allows up to two base mismatches and three indels per read alignment. In case you are working with a polymorphic or heterozygous population you can increase the allowed mismatched bases up to five per read by using the –N flag. The number of CPUs used by TopHat can be set by the –p flag. The annotation file provided for the genome is loaded by using the –G flag in your code. If the first run using a pre-built index fails restart for another time before compiling the index yourself. 3. Use the terminal and bam_tools to read out the mapping statistics [20] bamtools count -in accepted_hits.bam counting all mapped reads bamtools filter -in accepted_hits.bam -tag "NH:1"| bamtools count counting the uniquely mapped reads gunzip unmapped_left.* grep -c '^@HWI' unmapped_left.* counting all unmapped reads 3.7.2 Read Quantification

The Cufflinks package contains four tools for RNA-seq analysis: (1) Cufflinks assembles and quantifies transcripts, (2) Cuffcompare compares transcript assemblies to annotation, (3) Cuffmerge merges two or more transcript assemblies into one, and (4) Cuffdiff detects differentially expressed genes and transcripts, including different isoforms and promoter usage [21, 22]. In order to calculate the expression level of each transcript, Cufflinks counts the reads mapping to each transcript and normalizes these to the length of the transcript. To make expression counts comparable from one sequencing run to the other the expression is normalized to millions of mapped reads. Both normalization steps are incorporated in Cufflinks. To quantify paired

RNA-seq

85

end reads, Cufflinks uses a probabilistic model resulting in fpkm output (expressed fragments per kilobase per million); single end reads are quantified as rpkm (reads per kilobase per million). Cufflinks assembles the reads for isoform detection, with the output stored in the following files isoforms.fpkm_tracking, skipped. gtf, and transcripts.gtf. Cuffdiff is part of the Cufflinks package and it first calculates the expression of two or more samples and then tests statistically whether gene expression changes across the samples are significant. Additionally, it is able to identify differentially spliced isoforms. 1. Quantify reads by running Cufflinks with default settings on the terminal with following command: cufflinks -p 2 –G *.GTF/GFF -o output_directory accepted_hits.bam This step requires the TopHat output accepted_hits.bam files of the libraries and annotation file in the .gtf/gff format. The –p flag indicates the number of CPUs Cufflinks is allowed to use. Output: The software output delivers for expression quantification fragments per kilobase of transcript per million mapped reads (FPKM or RPKM for single end reads) in the genes.fpkm_tracking file. Export this file tab delimited file to Excel. This is the output needed for Subheading 3.9. 2. Estimate significantly differentially expressed reads by cuffdiff -o output_directory *.GTF/GFF sample1_replicate1.bam[,…,sample1_replicateM. bam] sample2_replicate1.bam[,…,sample2_replicateM.bam] Cuffdiff requires the .gtf/.gff reference annotation file and accepted_hits.bam files from the TopHat mapping step. Biological replicates of each condition are separated by comma, and each different condition is separated by a white space. The samples can be loaded as .bam or .sam files. 3. Export output files to Excel or any preferred tabular calculation software, e.g. output file for differential expression analysis gene_exp.diff. 3.8 Extracting Biological Information

Only a brief introduction into a limited selection of possible analyses can be given due to space constraints. All software developed to analyze microarray generated data can be adapted to use with RNA-seq experiments. The programs were chosen since they can be operated via graphical user interfaces familiar to biologists.

3.8.1

Even though programming languages are a more useful way to operate with large datasets, Excel is a versatile program which is familiar to most biologists.

Excel

86

Canan Külahoglu and Andrea Bräutigam

Converting File Formats

Excel is capable of opening files in which values are separated by a common delimiter, in most cases this will be a tabulator. Conversely, by choosing “Save as” Excel can save files using different delimiters as needed for downstream programs.

Transferring Information from Different Sources

Given two sources of information, for example (1) a file with expression values each headed by a gene identifier and (2) a list of genes known to be involved in circadian processes, these information can be combined in one file for easy access and interpretation. The combination of large files takes significant computing time (see also Note 9). 1. Open both files in Excel, combine into one file on different tabs. 2. Ensure that the gene identifiers have the same format in each table (e.g., capital and small letters, gene identifier given as the locus or the gene isoform); given a difference, alter all identifiers at once using functions such as “replace,” or “LEFT” or “RIGHT”. 3. Use the VLOOKUP function to match the information. This function requires (1) the common identifier between both tables, (2) the table from which to gather the new information with the common identifier in the first column, (3) the number of the column from which the information is to be returned, and (4) FALSE to disable searching for similar but not identical matches. 4. After the function was applied to match the new information to each identifier, copy the columns and reinsert as values only to save computing power. Using this simple function, the expression information can be augmented with descriptions from TAIR (or similar), with localization information, with expression information from previous experiments, essentially with any information that is given as a table with the same gene identifier format. Information from different sources, such as rpkm files and significance tests can be combined.

Testing Enrichment

For many experiments, it is desirable to assess whether a list of genes (i.e., those that are differentially regulated in the experiment, or those that are members of one expression pattern cluster, see below; now called group A) is enriched in genes from a second list (i.e., a previous experiment, or a list of genes known to be involved in a trait from a publication, now called group B). Excel can be used to make this assessment although this analysis can be carried out faster using programming languages such as R [23]. 1. Input: a table that indicates membership in group A and group B for each gene.

RNA-seq

87

2. Using the COUNTIFS function, count the number of genes that belong to (1) both A and B, (2) to A, and (3) to B. 3. Calculate a contingency table with the four fields in A and B, not in A/in B, in A/not in B, and not in A/not in B. 4. Calculate whether there is significant enrichment using a Fishers Exact Test or a chi square test (if the numbers are large) [24]. 5. If multiple tests are carried out at the same time, the alpha level of significance requires adjustment to multiple hypothesis testing. The spectrum of possible corrections ranges from Bonferroni (very conservative) [25] to Benjamini and Hochberg False Discovery Rate (FDR, rather permissive) [26]. 3.8.2 Mapman

Mapman is a tool to visualize data on previously determined pathways [27, 28]. 1. Download the software from http://mapman.gabipd.org/web/ guest/mapman; download the Mapman mapping for your species of interest from the MapMan store. The Mapman program requires three input files to run: (1) the mapping in which each gene is associated with a function, provided by the software or the Mapman store, (2) the pathway file in which the pathway image is associated with the genes in the pathway, provided by the software, and (3) the experiment file provided by the user. 2. Produce your input file by opening the file generated by the Cufflinks-pipeline. Make sure it has row identifiers identical to those in the Mapman mapping file which can be inspected in the text version of the mapping file. Remove all information except the expression information. Consider whether you want to transform the expression information into ratios or normalize it for display. 3. Display a pathway of interest by clicking at the mapping file, the pathway file and the experiment file. Mapman provides numerous pathways build into the software which can all be inspected for each dataset. It is mostly a visualization tool. If given a subset of significantly changed genes, it can calculate enrichment.

3.8.3 MultiExperiment Viewer

MultiExperiment Viewer is a tool to analyze large datasets in both supervised and unsupervised methods [29]. The software is Javabased and freely available. 1. Download the software from http://www.tm4.org/mev/ 2. Open the batch file and adjust the RAM available for the application to 2/3 of the RAM available on your computer.

88

Canan Külahoglu and Andrea Bräutigam

For example, if you run on 8 GB of RAM alter the parameter –Xmx to –Xmx6000 in the last line of the script. Save the alteration. 3. Run MeV. 4. Load the data by clicking the appropriate button. 5. Run a principle component analysis on the samples. The data reduction tool PCA will capture only the variation between samples and thus replicates will appear closer in a PCA compared to different samples. Based on the closeness of points in the PCA, similarity and difference between samples can be visualized. 6. Run a hierarchical clustering on the samples. Hierarchical clustering will sort similar samples close together on a tree based on similarity matrices. Euclidean and Pearson distance are most frequently used and reflect absolute differences (Euclidean) and trends (Pearson). Biological replicates are expected to cluster closer together than different samples. 7. Run a k-means clustering on the genes. For this clustering, a predetermined number of expected clusters have to be entered by the user. The program will fit the genes into clusters to achieve the most homogenous cluster appearance possible given the number of possible clusters available to the program. The MeV is a powerful software suite that can be used to analyze data beyond simple analyses of whether biological replicates are indeed similar or visualize the general trends in the data. 3.8.4 VirtualPlant

The VirtualPlant Platform http://virtualplant.bio.nyu.edu/cgi-bin/ vpweb/ was designed to extract information from large datasets [30]. 1. Upload gene list(s) to VirtualPlant. 2. Click analyze, the menu will open and you can choose your type of analysis. 3. Use the functions “union,” “intersect,” and “symmetric difference” to identify membership and membership overlap. 4. Use the function Biomap to determine whether functions and/or categories are overrepresented in one of the lists. 5. VirtualPlant is connected to Cytoscape to visualize the data on known networks [31]. To analyze data that includes more conditions that a comparison of two states clustering algorithms are required to identify the genes that have similar expression patterns [32]. Hierarchical clustering and k-means clustering are two of the early tools for this purpose. Recently, a range of additional algorithms have been

RNA-seq

89

developed [33, 34]. It is advisable to determine the tools needed for analysis of the data while planning the experiment to ensure that the requirements of the algorithms to be used are met.

4

Notes 1. Use RNAse-free plasticware. Wear gloves at all times when handling with RNA and all solutions and equipment for RNA work. Treat all working surfaces with RNAse inhibiting reagents such as RNAseExitus (AppliChem). Bake glassware and spatulas for RNA work at 180 °C for >2 h prior to use or use RNAseExitus solution (AppliChem). 2. Prepare all RNA extraction solutions and buffer in RNA-free water treated previously with 0.1 % (v/v) Diethylpyrocarbonate (DEPC). DEPC inactivates RNases. When adding DEPC to the solution, make sure to use a syringe, which is submerged in the solution. Before autoclaving, it is important to stir the solution on a magnetic stirrer for at least 2 h, without closing the container tightly. After autoclaving, DEPC is broken down to CO2 and H2O. 3. The acidic pH is critical factor to ensure the separation of RNA from DNA and proteins. Never use buffered phenol instead of water-saturated phenol. 4. After adding Wash Buffer 1 invert closed microtube with column several times to ensure that all remnants from prior steps are washed away. 5. During DNAse treatment do not exceed the incubation time of 30 min at 37 °C. Before DNAse inactivation at elevated temperature make sure to add 5 mM EDTA (end concentration). 6. In case of persistent gDNA contamination of your RNA sample, do multiple DNAse treatments. 7. View basic help page by typing -h. 8. Sub-read files, which had a bad FastQC report are re-checked by FastQC again after the FASTX cleaning. If the sub read files now pass the FastQC pipeline, include them when concatenating the high quality cleaned sub read files to one sample. 9. Excel is a suboptimal tool to analyze large datasets. Scripts will do the same job much faster. However, given the fact that most biologists are not skilled programmers, Excel delivers the same results but will take up to minutes to compute. Once computation is finished it is advisable to copy the resulting columns and re-insert the information as values to prevent Excel from crashing.

90

Canan Külahoglu and Andrea Bräutigam

References 1. Jiao YL, Tausta SL, Gandotra N et al (2009) A transcriptome atlas of rice cell types uncovers cellular, functional and developmental hierarchies. Nat Genet 41:258–263 2. Li PH, Ponnala L, Gandotra N et al (2011) The developmental dynamics of the maize leaf transcriptome. Nat Genet 42:1060–1067 3. Davidson RM, Gowda M, Moghe G et al (2012) Comparative transcriptomics of three poaceae species reveals patterns of gene expression evolution. Plant J 71:492–502 4. Bräutigam A, Kajala K, Wullenweber J et al (2011) An mrna blueprint for c4 photosynthesis derived from comparative transcriptomics of closely related c3 and c4 species. Plant Physiol 155:142–156 5. Gowik U, Brautigam A, Weber KL et al (2011) Evolution of c4 photosynthesis in the genus flaveria: how many and which genes does it take to make C4? Plant Cell 23:2087–2105 6. Pick TR, Bräutigam A, Schlüter U et al (2011) Systems analysis of a maize leaf developmental gradient redefines the current c4 model and provides candidates for regulation. Plant Cell 23:4208–4220 7. Chomczynski P, Sacchi N (1987) Single-step method of rna isolation by acid guanidinium thiocyanate phenol chloroform extraction. Anal Biochem 162:156–159 8. Chomczynski P, Sacchi N (2006) The singlestep method of rna isolation by acid guanidinium thiocyanate-phenol-chloroform extraction: twenty-something years on. Nat Protoc 1: 581–585 9. Sangha JS, Gu K, Kaur J et al (2010) An improved method for rna isolation and cdna library construction from immature seeds of jatropha curcas l. BMC Res Notes 3:126 10. Mueller O, Hahnenberger K, Dittmann M et al (2000) A microfluidic system for highspeed reproducible DNA sizing and quantitation. Electrophoresis 21:128–134 11. Schroeder A, Mueller O, Stocker S et al (2006) The rin: an rna integrity number for assigning integrity values to rna measurements. BMC Mol Biol 7:3 12. Braeutigam A, Gowik U (2010) What can next generation sequencing do for you? Next generation sequencing as a valuable tool in plant research. Plant Biol 12:831–841 13. Schliesky S, Gowik U, Weber APM et al (2012) Rna-seq assembly - Are we there yet? Front Plant Sci 3:220

14. Ewing B, Green P (1998) Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res 8:186–194 15. Ewing B, Hillier L, Wendl MC et al (1998) Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res 8:175–185 16. Blankenberg D, Gordon A, Von Kuster G et al (2010) Manipulation of fastq data with galaxy. Bioinformatics 26:1783–1785 17. Langmead B, Trapnell C, Pop M et al (2009) Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol 10:R25 18. Trapnell C, Pachter L, Salzberg SL (2009) Tophat: discovering splice junctions with rnaseq. Bioinformatics 25:1105–1111 19. Ferragina P, Manzini G (2001) An experimental study of a compressed index. Inf Sci 135: 13–28 20. Li H, Handsaker B, Wysoker A et al (2009) The sequence alignment/map format and samtools. Bioinformatics 25:2078–2079 21. Trapnell C, Williams BA, Pertea G et al (2010) Transcript assembly and quantification by rna-seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol 28:511–515 22. Trapnell C, Roberts A, Goff L et al (2012) Differential gene and transcript expression analysis of rna-seq experiments with tophat and cufflinks. Nat Protoc 7:562–578 23. Gentleman RC, Carey VJ, Bates DM et al (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5:R80 24. Fisher RA (1922) On the interpretation of χ2 from contingency tables, and the calculation of p. J R Stat Soc 85:87–94 25. Dunn OJ (1961) Multiple comparisons among means. J Am Stat Assoc 56:52–64 26. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate - A practical and powerful approach to multiple testing. J R Stat Soc Ser B Methodol 57:289–300 27. Thimm O, Blasing O, Gibon Y et al (2004) Mapman: a user-driven tool to display genomics data sets onto diagrams of metabolic pathways and other biological processes. Plant J 37:914–939 28. Usadel B, Nagel A, Thimm O et al (2005) Extension of the visualization tool mapman to allow statistical analysis of arrays, display of

RNA-seq corresponding genes, and comparison with known responses. Plant Physiol 138: 1195–1204 29. Howe E, Holton K, Nair S et al (2010) Mev: multiexperiment viewer 30. Katari MS, Nowicki SD, Aceituno FF et al (2010) Virtualplant: a software platform to support systems biology research. Plant Physiol 152:500–515 31. Shannon P, Markiel A, Ozier O et al (2003) Cytoscape: a software environment for

91

integrated models of biomolecular interaction networks. Genome Res 13:2498–2504 32. Langfelder P, Horvath S (2012) Fast r functions for robust correlations and hierarchical clustering. J Stat Softw 46:1–17 33. Allen JD, Xie Y, Chen M et al (2012) Comparing statistical methods for constructing large scale gene networks. PLoS One 7:e29348 34. Jay JJ, Eblen JD, Zhang Y et al (2012) A systematic comparison of genome-scale clustering algorithms. BMC Bioinformatics 13:7

Chapter 6 Rapid and Parallel Quantification of Small and Large RNA Species Corinna Speth and Sascha Laubinger Abstract Quantitative real-time PCR (qRT-PCR) is a common technique for mRNA quantification. Several methods have been developed in the past few years in order to adapt qRT-PCR also for small non-coding RNAs (sRNA). We here provide a simple and sensitive protocol that allows quantification of mRNAs, selected sRNAs, and long non-coding RNAs (lncRNA) in one cDNA sample by qRT-PCR. Key words Small RNA (sRNA), microRNA (miRNA), Small interfering RNA (siRNA), Trans-acting siRNA (tasiRNA), Long non-coding RNA (lncRNA), Primary-miRNA (pri-miRNA), mRNA, qRTPCR, Arabidopsis

1

Introduction A large number of noncoding RNAs (ncRNAs), coming in two major flavors, i.e., small RNAs (sRNAs) and long ncRNAs (lncRNAs), have been found in eukaryotes. Plant endogenous sRNAs are basically divided into three main classes: (a) small interfering RNAs (siRNAs), (b) microRNAs (miRNAs), and (c) transacting small interfering RNA (tasiRNA). All sRNAs are processed from longer double-strand RNA (dsRNA) precursors into 20–24 nt long sRNA duplexes by one of the four Arabidopsis DICER-LIKE (DCL) proteins. sRNAs associate with one of the ten Arabidopsis ARGONAUTE proteins to fulfill their function. siRNAs ensure transcriptional silencing of heterochromatic loci, whereas miRNAs and tasiRNAs mostly control target mRNA levels by posttranscriptional silencing mechanisms (cleavage or translational inhibition). sRNAs have been described as regulators of diverse developmental processes as well as regulators of adaptive responses to abiotic and biotic stress [1, 2]. miRNAs are also implicated in circadian clock function in animals, and a number of Arabidopsis miRNAs have been reported to cycle in a diurnal or a circadian manner [3–5]. Comparatively little is known about the function of lncRNAs, but

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_6, © Springer Science+Business Media New York 2014

93

94

Corinna Speth and Sascha Laubinger

also lncRNAs have the potential to cycle (see below), and it would be not surprising if they fulfill important regulatory functions for clock-regulated gene expression in plants as well. Because of the emerging importance of ncRNAs in regulating gene expression, especially sRNA quantification has become a more and more important tool to understand the processes of gene expression. There are four common techniques available for sRNA quantification: (1) For northern blot analysis, total RNA is size-separated by electrophoresis and subsequently transferred on a nylon membrane to detect sRNAs with labeled antisense probes [6]. (2) Detection of sRNAs by qRT-PCR comprises reverse transcription of sRNAs followed by quantification of resulting cDNA via PCR [7, 8]. (3) sRNAs can be labeled and hybridized to sRNA microarrays, which represent known sRNAs by complementary probes [9, 10]. (4) Analysis of sRNA populations using sRNA sequencing (sRNA-seq) requires ligation of RNA-adaptor molecules and reverse transcription, which is followed by next-generation sequencing [10]. Importantly, one should not insist on any of the above-described alternatives, because all the methods have advantage and disadvantage and the experimenter has to select the best suited technique for a specific question. Also, comparisons between the different available platforms revealed some inconsistencies, which are likely to be attributable to specific biases that are introduced during miRNA priming or adaptor ligation [11–13]. All techniques can be used to quantify the abundance of known sRNAs (quantitative information), with the advantage of sRNA-seq and sRNA microarray analysis allowing sRNA quantification globally. sRNA-seq additionally provides qualitative information due to its single-nucleotide resolution and offers the unique opportunity to identify novel sRNAs. Both methods, however, are relatively expensive, and data analyses require some informatics for data evaluation. In contrast, northern blot analyses and qRT-PCR techniques are fast and easy solutions to quantify selected sRNAs. This is especially of importance when a large number of samples are subjected to sRNA analyses, such as in circadian experiments for which different time points and mutants are investigated. qRT-PCR analyses, however, only provide quantitative information and are solely suitable for the detection of known sRNAs (correctly processed miRNAs, in phase-cleaved tasiRNAs). Nevertheless, a great advantage of quantifications by qRT-PCR is its high sensitivity. Down to 10 ng total RNA per sample is enough to quantify several sRNAs at the same time. In contrast, northern blot analyses require 5–50 μg total RNA per sample and allow only the detection of one sRNA per sample at once. This protocol (adapted from [14]) provides an easy, highly sensitive, and fast method to simultaneously quantify mRNA, sRNA, and lncRNA by quantitative real-time PCR using one cDNA sample [14]. It is a three-step protocol composed of (1) total RNA extraction and (2) reverse transcription followed by (3) quantitative real-time PCR (Fig. 1).

Quantification of Small and Large RNAs

95

(i) total RNA extraction

(ii) cDNA synthesis using oligo(dT)18 and sRNA specific stem-loop oligonucleotides poly(A)-tail transcript AAAAAAAAAAAAAAAAA TTTTTTTTTTTTTTTTTT

oligo(dT)18

sRNA stem-loop oligonucleotide

reverse transcription poly(A)-tail transcript AAAAAAAAAAAAAAAAA TTTTTTTTTTTTTTTTTT

oligo(dT)18 sRNA stem-loop oligonucleotide

(iii) quantification of sRNAs and poly(A)-tail transcripts by quantitative real-time PCR poly(A)-tail transcript forward oligonucleotide

reverse oligonucleotide TTTTTTTTTTTTTTTTTT

forward oligonucleotide

sRNA

universal reverse oligonucleotide

Fig. 1 Flow chart for sRNA quantification This protocol comprises three steps: (i ) total RNA extraction, (ii) reverse transcription using oligo(dT)18 and sRNA-specific oligonucleotides, and (iii) quantification of sRNA and poly(A)-tail transcripts by qRT-PCR

(1) Total RNA is isolated using TRIzol reagent based on the single-step RNA isolation method developed by Chomczynski and Sacchi [15] (see Note 1). (2) In order to use sRNAs as a template for PCR, an individual sRNA is primed with a specific stem–loop oligonucleotide. This oligonucleotide is reverse complementary to the 3′ end of the sRNA. During cDNA synthesis, the stem–loop

96

Corinna Speth and Sascha Laubinger

oligonucleotide is extended which results in a longer cDNA template. By adding up to four sRNA-specific stem–loop oligonucleotides and an oligo(dT) oligonucleotide to the same cDNA synthesis reaction one can reverse transcribe poly(A)-tail transcripts as well as up to four sRNAs in one reaction (Fig. 1). (3) After reverse transcription, poly(A)-tail transcripts (mRNAs, lncRNAs) and sRNAs can be analyzed by PCR (e.g., to identify lncRNAs) and quantified by standard qRT-PCR techniques (e.g., to analyze the expression of sRNA under diurnal condition, Figs. 1 and 2). Depending on the number of samples, sRNA analyses can be performed within 12–16 h (see Note 2). As a proof of concept, we analyzed the expression of various mRNAs, mature miRNAs, a pri-miRNA, a tasiRNA, and lncRNAs across the day (Fig. 2). We show that a single reverse transcription reaction per time point to be analyzed provides information about the expression behavior of all these RNA species under diurnal conditions. In addition, we found the approach also useful to verify the existence of lncRNAs by simple PCR amplification followed by gel analysis (Fig. 2f ).

2 2.1

Materials Plant Material

2.2 Total RNA Extraction

Plants are grown on half-strength MS medium for 10 days in light– dark cycles, harvested at the indicated time points, and snap-frozen in liquid nitrogen. 1. TRIzol® (Life Technologies) or guanidinium thiocyanate– phenol solutions from other companies (e.g., TRI Reagent®, Sigma-Aldrich). 2. Isopropanol. 3. Chloroform. 4. Nuclease-free water or DEPC-treated water. 5. To prepare DEPC-treated water, add 0.1 % (v/v) DEPC to demineralized water. Stir overnight, and autoclave to degrade DEPC. 6. 80 % (v/v) Ethanol (prepared with nuclease-free water or DEPC-treated water). 7. Nuclease-free 1.5 mL safe-lock reaction tubes. 8. Nuclease-free tips (optional: filter tips). 9. Gloves. 10. Hood. 11. Bench-top centrifuge with cooling function. 12. System to cast/run agarose gels.

97

Quantification of Small and Large RNAs

d 1.2 TOC1 CCA1 1

1.2

Relative expression

0.8 0.6 0.4 0.2 0

0

4

8

12

16

20

0.8 0.6 0.4 0.2 0

24

siR1511

1

0

4

8

12

12h

12h

20

24

20

24

12h

12h

b

e WT miR156 se-1 miR156

1.2 At5NC029980

WT pri-miR156 se-1 pri-miR156

Relative expression

20 16 12 8 4

1 0.8 0.6 0.4 0.2 0 0

4

8

12

8

12

16

20

24

Zeitgeber

12h

12h

99

50 02 5N

C

06

5N

C

41

17

8

79

20 40 2G

5N

AT

4

AT

0

At

1N

C

0.2

00

06

49

60

f

0.4

AT

0.6

80

12h

12h

0.8

0

16

Zeitgeber

1

AT

Relative expression

16

Zeitgeber

Zeitgeber

C

Relative expression

a

250 bp > 100 bp >

Relative expression

c 1.2 miR159 1 0.8 0.6 0.4 0.2 0 0

4

8

12

16

20

24

Zeitgeber

12h

12h

Fig. 2 Expression analysis of different sRNAs and poly(A) transcripts of 9-day-old seedlings grown under 12-h light/12-h dark conditions on ½ MS plates were collected every 4 h. Expression analyses were conducted as described in this protocol. All experiments were performed using only one cDNA sample for each data point. (a–e) qRT-PCR analyses of several mRNAs, lncRNAs, and sRNAs in WT and se-1. Error bars denote the range of three independent biological experiments. (f) Amplification of five lncRNAs by standard PCR

98

Corinna Speth and Sascha Laubinger

Table 1 Oligonucleotide design for small RNAs miR156

Sequence (5′ → 3′)

Sense sequence

TGACAGAAGAGAGTGAGCAC

RT oligonucleotide

GTCGTATCCAGTGCAGGGTCCGAGGTATTC GCACTGGATACGACGTGCTC

Forward oligonucleotide

GCGGCGG TGACAGAAGAGAGT

Universal reverse oligonucleotide

GTGCAGGGTCCGAGGT

13. Agarose. 14. 10× TBE buffer: 890 mM tris base; 890 mM boric acid; 20 mM EDTA; has pH 8.3. 15. Photometer for RNA quantification. 2.3 Oligonucleotide Design

Oligonucleotides for quantification of sRNAs were designed as described by Varkonyi-Gasic et al. [14]. The oligonucleotide design for miR156 is shown below as an example. 1. RT oligonucleotides: sRNA-specific stem–loop oligonucleotide anneal with specific sRNAs. The following stem–loop is identical for all stem–loop oligonucleotides (bold italic letters mark complementary base pairing regions): 5 ′ - G T C G TAT C C A G T G C A G G G T C C G A G G TAT TC GCACTGGATACGAC-3′. The specificity for specific sRNAs is obtained by adding six nucleotides (bold letters) to the 3′ end of the stem–loop oligonucleotide, which are reverse complementary to the 3′ end of the sRNA (see Table 1; Fig. 1). 2. Quantitaive real-time PCR (qPCR) oligonucleotide: The forward oligonucleotide is specific for each sRNA. The whole sRNA sequence excluding the last six nucleotides (underlined regular letters) is extended by a randomly chosen 5–7 nt long sequence (italic characters) at the 5′ end (see Table 1; Fig. 1). The sequence should be designed in a way that the overall GC content of the primer is between 40 and 60 % and melting temperature Tm is 60 ± 5 °C. A universal oligonucleotide complementary to the invariant stem–loop region of the RT oligonucleotide is used as a reverse primer (see Note 3).

Quantification of Small and Large RNAs

2.4 Reverse Transcription

99

1. DNase (RNase free)/EDTA (Thermo Scientific, # EN0521) including: DNaseI, 10× DNaseI buffer, EDTA solution (25 mM) for inactivation. 2. RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, # K1622) including: dNTPs, 5× reaction buffer, oligo(dT)18, Ribo-LOCK, RevertAid, RNAse-free water. 3. Thermo cycler. 4. Nuclease-free tips (optional: filter tips). 5. Gloves. 6. sRNA-specific stem–loop oligonucleotide. 7. Purified total RNA.

2.5 Quantitative Real-Time PCR

1. Maxima SYBR Green qPCR Master Mix (2×), ROX solution provided (Thermo Scientific, #K0252). 2. Oligonucleotides. 3. 96- or 384-well plate suitable for qRT-PCR (e.g., PCR FramePlate; VWR). 4. Optical adhesive seal for PCR plates (e.g., Microseal “B” Film; Bio-Rad). 5. qRT-PCR thermo cycler (e.g., CFX384 system, Bio-Rad). 6. System to cast/run agarose gels. 7. Agarose. 8. 10× TBE buffer: 890 mM tris base; 890 mM boric acid; 20 mM EDTA; has pH 8.3.

3

Methods

3.1 Total RNA Extraction

When working with RNA it is important to follow general lab techniques: wear gloves, use nuclease-free tips (optional: filter tips), and prepare solutions with DEPC-treated water to protect RNA from degradation by RNases. Work under the hood when pipetting the toxic and harmful reagents TRIzol and chloroform! 1. Harvest plant material, and grind it into a fine powder using liquid nitrogen, mortar, and pistil. For small amounts of tissues, grind the plant material in a reaction tube to a fine powder using a plastic pestle (Sigma-Aldrich, Z359947-100EA). 2. Transfer approximately 50–100 μg (100–200 μL) of the ground plant material to a 1.5 mL safe-lock reaction tube. It is important to prevent thawing of the powder.

100

Corinna Speth and Sascha Laubinger

3. Add 1 mL of the TRIzol reagent. Subsequently resuspend the plant material by vigorous vortexing. 4. After 10-min incubation at RT collect the plant debris by centrifugation (16,000 × g, 10 min, 4 °C), and transfer the supernatant to a new 1.5 mL safe-lock reaction tube. Discard the pellet. 5. Add 200 μL chloroform to the supernatant, and mix each sample by vigorous vortexing (15 s). 6. Separate the organic (lower phase) and the aqueous (upper phase) phase by centrifugation (16,000 × g, 5 min, 4 °C). The aqueous phase contains total RNA. 7. Carefully transfer the upper phase to a new tube without disturbing the interphase. 8. Add 500 μL chloroform to each sample, mix by vigorous vortexing (15 s), and centrifuge at 16,000 × g for 10 min at 4 °C. 9. Transfer the upper phase without disturbing the interphase to a new 1.5 mL safe-lock reaction tube. 10. Repeat steps 8 and 9 at least twice or until the interphase is clear. 11. Add 1 vol of isopropanol to the upper phase. Mix each sample well by inverting several times (see Note 4). 12. Incubate the sample for 1 h at −80 °C or overnight at −20 °C. After that, RNA is precipitated by centrifugation (16,000 × g, 30 min, 4 °C). 13. Carefully remove the supernatant by pipetting. 14. Add 500 μL 80 % ethanol to each sample without disturbing the pellet and centrifuge at 16,000 × g for 10 min at 4 °C. 15. Carefully remove supernatant by pipetting. 16. Make a short spin to collect the residual liquid, and remove it by pipetting. 17. Air-dry pellet for approximately 3 min at RT. 18. Preheat an aliquot of nuclease-free water or DEPC-treated water to 65 °C. 19. Resuspend the pellet in 50 μL preheated nuclease-free water or DEPC-treated water. 20. Incubate each sample on ice (up to 1 h). If the pellet has not completely dissolved heat RNA to 65 °C for 1–5 min. Do not vortex RNA to avoid shearing! If your pellet does not dissolve, the RNA pellet was overdried. In that case, freeze the sample and try to dissolve the pellet again. 21. Determine RNA concentration using a photometer. This RNA extraction method yields usually 20–100 μg total RNA dependent on the plant tissue used (RNA amount: rosette leaves < seedlings < inflorescences).

Quantification of Small and Large RNAs

101

22. Analyze the integrity of 0.5 μg total RNA on a 1 % (w/v) agarose gel using TBE buffer. For this step it is not necessary to use DEPC-treated solutions. Clean DNA gel chamber, gel tray, and comb with detergent before use and rinse with water. 3.2 Reverse Transcription (RT)

cDNA synthesis is performed with the RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific). To set up the reaction use filter tips, nuclease-free PCR tubes, and gloves and pipette everything on ice to protect RNA from degradation. We are performing all reactions in 8-strip PCR tubes with separate 8-strip caps. After each incubation step, briefly spin down the tubes using a mini-centrifuge and use new caps for the PCR tubes. For cDNA synthesis use 0.25 ng–2 μg total RNA, depending on the abundance of the sRNA to be quantified. 1 μg total RNA was used for the analyses shown in Fig. 2. Prepare the RT-reaction as follows: 1. First residual DNA has to be removed by DNaseI (RNase-free) treatment. Prepare the reaction mixture in nuclease-free PCR tubes: x μL

0.25 ng–2 μg total RNA

1 μL

10× DNaseI buffer

1 μL

DNaseI (RNase free)

x μL

Up to 10 μL nuclease free (comes with the cDNA Synthesis Kit)

Mix well, and incubate for 30 min at 37 °C in a thermocycler. 2. To inactivate DNAseI add 1 μL of EDTA, 25 mM (comes with the DNAseI), to each sample. Mix well, and incubate for 10 min at 65 °C. 3. Add 1 μL of 100 mM oligo-dT18 (see Note 5) and 0.5 μL of the stem–loop oligonucleotide mixture (up to four different oligonucleotides with a concentration of 2 μM each; see Table 2). 4. Mix everything well, and denature the samples for 5 min at 65 °C. Immediately place samples on ice. 5. Complete the RT reaction mixture by combining the components listed below: 12.5 μL

RNA with oligonucleotides

4 μL

5× RT buffer

2 μL

10 mM dNTPs

1 μL

Ribo-LOCK

1 μL

RevertAid

102

Corinna Speth and Sascha Laubinger

Table 2 List of oligonucleotides used in this chapter Sequence (5′ → 3′)

Reference

miR156-RT

GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCAC TGGATACGACGTGCTC

[14]

miR159a-RT

GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCAC TGGATACGACTAGAGC

siR1511-RT

GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCAC TGGATACGACAAGTAT

[14]

oligo(dT)18

TTTTTTTTTTTTTTTTTT

Thermo Scientific

TOC1-F

TCTTCGCAGAATCCCTGTGAT

[17]

TOC1-R

GCTGCACCTAGCTTCAAGCA

[17]

CCA1-F

GATGATGTTGAGGCGGATG

[17]

CCA1-R

TGGTGTTAACTGAGCTGTGAAG

[17]

ACTIN2-F

CTTGCACCAAGCAGCATGAA

[18]

ACTIN2-R

CCGATCCAGACACTGTACTTCCTT

[18]

At1NC064960-F

CGGAGATGAAGATTGGTTGG

[19]

At1NC064960-R

GGACCGGATCGGTACAATAGA

[19]

AT5NC004020-F

TGCAAATGCTACATGGTTCG

[19]

AT5NC004020-R

GAGCAACTGGTTATGATTGG

[19]

AT2G41178-F

GTCTTGACCGGAAGGAGAGTG

[19]

AT2G41178-R

TGTCATGGAGACACGATGAAC

[19]

AT5NC067950-F

AGAGGGTGGATGATACTGTCG

[19]

AT5NC067950-R

CCACTGACCATTGAGACACAC

[19]

AT5NC029980-F

CAAGGAGGAGGAAGTTTCCAG

[19]

AT5NC029980-R

GTCACAAAAGAAGCCGTGTTG

[19]

pri-miR156a-F

ggacaagagaaacgcaaagaaaCTGAC

[20]

pri-miR156a-R

AGTGAGCACGCAAGAGAAGCAAG

[20]

Universal sRNA-R

GTGCAGGGTCCGAGGT

[14]

miR156-F

GCGGCGGTGACAGAAGAGAGT

[14]

miR159-R

GCGGCGTTTGGATTGAAGGGA

[14]

si1511-R

GGTCGTCCAAGCGAATGATG

[14]

Reverse transcription

qRT-PCR

Quantification of Small and Large RNAs

103

6. Mix well, and incubate in a thermocycler using the following program: (a)

16 °C

30 min

(b)

30 °C

30 s

(c)

42 °C

30 s

(d)

50 °C

1s

Repeat steps b–d 59 times (e)

3.3 Quantitative Real-Time PCR

85 °C

5 min

The setup for a qRT-PCR reaction is comparable to a standard PCR (see Subheading 3.3, step 1), which includes a DNA polymerase, dNTPs, oligonucleotides, and additionally SYBR Green to quantify the amplification of double-stranded DNA photometrically at 530 nm. When pipetting a qRT-PCR at least two or three technical replicates should be prepared for each sample/template. The technical replicates get averaged for analysis. Each experiment should be repeated three times using independent biological replicates. It also is recommended to include a “no-template control” for each oligonucleotide combination to exclude contaminations or unspecific amplified products. The expression of a sRNA or a mRNA of interest is calculated relative to that of a reference gene (e.g., ACTIN2, which was used for Fig. 2), which exhibits invariant expression under the experimental conditions. 1. Standard qRT-PCR reaction mixture: 5 μL

Maxima SYBR Green qPCR Master Mix (2×)

0.25 μL

Forward oligonucleotide (10 μM)

0.25 μL

Reverse oligonucleotide (10 μM)

2 μL

Template (different dilutions)

2.5 μL

Nuclease-free water (is included in the qPCR Kit)

2. Use a dilution of the cDNA template for highly abundant transcripts. In this study, a 1:10 dilution was prepared for all qRTPCR reactions except for the quantification of low-abundance pri-miRNAs, from which mature miRNAs are processed (Fig. 2). 3. For the “no-template control,” add nuclease-free water instead of the template. 4. Perform an oligonucleotide standard curve with a serial 1:5 dilution of wild-type cDNA with at least four dilutions to calculate PCR amplification efficiency.

104

Corinna Speth and Sascha Laubinger

5. It is recommended to prepare a master mix for each oligonucleotide combination without template by scaling up a standard qRT-PCR mixture and adding 5–10 % for pipetting errors. 6. First distribute the master mix into the PCR plate, and then add template DNA. 7. Seal the PCR plate with an optical adhesive seal. 8. Collect the samples at the bottom of the well by centrifugation (short spin) and incubate in a qRT-PCR thermocycler using the following three-step program with a photometric measurement at 530 nm step after each amplification cycle followed by a melting curve at the end of the amplification steps (see Note 6). The melting curves should peak homogenously at one temperature. More than one peak suggests formation of oligonucleotide dimers or production of unspecific PCR products during amplification: (a)

95 °C

5 min

(b)

95 °C

10 s

(c)

55 °C

30 s

(d)

72 °C

20 s

Photometric measurement at 530 nm Repeat steps b–d 39 times Melting curve: (e)

95 °C

(f)

55–95 °C: Each step + 1 °C; 5 s Photometric measurement at 530 nm

30 s

Data evaluation using a relative quantification method [16]. 9. c(t) values were determined using the threshold cycle method. Threshold and resulting c(t) values generated by the Bio-Rad CFX384 program were used in this study (Fig. 2). 10. Calculate the average c(t) value for all technical replicates. 11. PCR amplification efficiency was determined by blotting the logarithm of the dilution series versus the corresponding c(t) values. The relationship between c(t) value and log (dilution) usually is linear. By adding a trend line one can calculate the PCR amplification efficiency, which should be around 100 % assuming a doubling of the DNA during each cycle. Normal PCR efficiency is 90–110 %:

Quantification of Small and Large RNAs

E = 10(

105

-1/ slope )

.

12. Relative expression was determined using the ∆∆c(t) method without real-time efficiency correction: - Dc (t ) sample - Dc (t ) reference ) R=2( .

13. Calculate the standard error of mean (SEM), which denotes the range of the independent biological of replicates used in the experiment. 14. Check the size of PCR products on a 1–3 % (w/v) agarose gel.

4

Notes 1. It is also possible to isolate RNA with an extraction kit, but it is highly important to use RNA extraction kits which are suitable for sRNA isolation. Traditional glass-fiber filter purification kits (e.g., RNeasy® Plant Mini Kit, Qiagen) recover only larger RNA species. 2. When quantifying miRNAs and pri-miRNAs, we include a genetic control (like se-1 mutants, which are available from the NASC Stock Centre, N3257). Mutants like se-1 accumulate low levels of miRNAs and high levels of miRNA precursors (Fig. 2b). 3. Whether oligonucleotides are specific for an sRNA can be tested by cloning the resulting PCR product once into a TA-cloning vector and performing some Sanger sequencing reactions. This ensures that the correct sRNA is amplified. 4. Small amounts of precipitated RNA can be visualized by adding 20 μg RNase-free glycogen (e.g., Thermo Scientific). 5. Some RNAs might not contain a poly(A)-tail. In this case, priming with random hexamers is required. 6. Depending on the PCR settings, it is also possible to use a twostep protocol for qRT-PCR.

Acknowledgements This work was supported by the DFG (LA2633-1/2) and the Max Planck Society (MPG)—Chemical Genomics Centre (CGC) through its supporting companies AstraZeneca, Bayer CropScience, Bayer Healthcare, Boehringer-Ingelheim, and Merck-Serono.

106

Corinna Speth and Sascha Laubinger

References 1. Chen X (2010) Small RNAs – secrets and surprises of the genome. Plant J 61(6):941–958. doi:10.1111/j.1365-313X.2009.04089.x 2. Voinnet O (2009) Origin, biogenesis, and activity of plant microRNAs. Cell 136(4):669–687, http://dx.doi.org/10.1016/j.cell.2009.01.046 3. Staiger D, Koster T (2011) Spotlight on posttranscriptional control in the circadian system. Cell Mol Life Sci 68(1):71–83. doi:10.1007/ s00018-010-0513-5 4. Sire C, Moreno AB, Garcia-Chapa M, LopezMoya JJ, San Segundo B (2009) Diurnal oscillation in the accumulation of Arabidopsis microRNAs, miR167, miR168, miR171 and miR398. FEBS Lett 583(6):1039–1044. doi:10.1016/j.febslet.2009.02.024 5. Hazen SP, Naef F, Quisel T, Gendron JM, Chen H, Ecker JR, Borevitz JO, Kay SA (2009) Exploring the transcriptional landscape of plant circadian rhythms using genome tiling arrays. Genome Biol 10(2):R17. doi:10.1186/ gb-2009-10-2-r17 6. Khraiwesh B (2012) Use of northern blotting for specific detection of small RNA molecules in transgenic plants. In: Dunwell JM, Wetten AC (eds) Transgenic plants, vol 847, Methods in molecular biology. Humana Press, Totowa, NJ, pp 25–32. doi:10.1007/978-1-61779-558-9_3 7. Chen C, Ridzon DA, Broomer AJ, Zhou Z, Lee DH, Nguyen JT, Barbisin M, Xu NL, Mahuvakar VR, Andersen MR, Lao KQ, Livak KJ, Guegler KJ (2005) Real-time quantification of microRNAs by stem–loop RT–PCR. Nucleic Acids Res 33(20):e179. doi:10.1093/ nar/gni178 8. Chiang RSVL (2005) Facile means for quantifying microRNA expression by real-time PCR. Biotechniques 39(4):19–525 9. Liu H-H, Tian X, Li Y-J, Wu C-A, Zheng C-C (2008) Microarray-based analysis of stressregulated microRNAs in Arabidopsis thaliana. RNA 14(5):836–843. doi:10.1261/ rna.895308 10. Lu C, Souret F (2010) High-throughput approaches for miRNA expression analysis. Methods Mol Biol 592:107–125. doi:10.1007/978-1-60327-005-2_8 11. Adhikari S, Turner M, Subramanian S (2013) Hairpin priming is better-suited than in vitro polyadenylation to generate cDNA for plant

12.

13.

14.

15.

16.

17.

18.

19.

20.

miRNA qPCR. Mol Plant 6:229–231. doi:10.1093/mp/sss106 Feldmesser E, Leshkowitz D, Parmet Y, HornSaban S (2012) Major differences in microRNA quantification are platform and sequence dependent. BMC Proc 6(Suppl 6):P7 Zhang W, Gao S, Zhou X, Chellappan P, Chen Z, Zhou X, Zhang X, Fromuth N, Coutino G, Coffey M, Jin H (2011) Bacteria-responsive microRNAs regulate plant innate immunity by modulating plant hormone networks. Plant Mol Biol 75(1–2):93–105. doi:10.1007/ s11103-010-9710-8 Varkonyi-Gasic E, Wu R, Wood M, Walton E, Hellens R (2007) Protocol: a highly sensitive RT-PCR method for detection and quantification of microRNAs. Plant Methods 3(1):12 Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162(1):156– 159. doi:10.1016/0003-2697(87)90021-2 Pfaffl MW (2006) Relative quantification. In: Dorak T (ed) Real-time PCR. International University Line, San Diego, CA, pp 63–82 Gould PD, Locke JCW, Larue C, Southern MM, Davis SJ, Hanano S, Moyle R, Milich R, Putterill J, Millar AJ, Hall A (2006) The molecular basis of temperature compensation in the Arabidopsis circadian clock. Plant Cell Online 18(5):1177– 1187. doi:10.1105/tpc.105.039990 Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible W-R (2005) Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol 139(1):5–17. doi:10.1104/pp. 105.063743 Liu J, Jung C, Xu J, Wang H, Deng S, Bernad L, Arenas-Huertero C, Chua NH (2012) Genome-wide analysis uncovers regulation of long intergenic noncoding RNAs in Arabidopsis. Plant Cell 24(11):4333–4345. doi:10.1105/tpc.112.102855 Pant BD, Musialak-Lange M, Nuc P, May P, Buhtz A, Kehr J, Walther D, Scheible W-R (2009) Identification of nutrient-responsive Arabidopsis and rapeseed microRNAs by comprehensive real-time polymerase chain reaction profiling and small RNA sequencing. Plant Physiol 150(3):1541–1555. doi:10.1104 /pp. 109.139139

Chapter 7 The RIPper Case: Identification of RNA-Binding Protein Targets by RNA Immunoprecipitation Tino Köster, Meike Haas, and Dorothee Staiger Abstract Control at the posttranscriptional level emerges as an important layer of regulation in the circadian timing system. RNA-binding proteins that specifically interact with cis-regulatory motifs within premRNAs are key elements of this regulation. While the ability to interact with RNA in vitro has been demonstrated for numerous Arabidopsis RNA-binding proteins, a full understanding of posttranscriptional networks controlled by an RNA-binding protein requires the identification of its immediate in vivo targets. Here we describe differential RNA immunoprecipitation in transgenic Arabidopsis thaliana plants expressing RNA-binding protein variants epitope-tagged with green fluorescent protein. To control for RNAs that nonspecifically co-purify with the RNA-binding protein, transgenic plants are generated with a mutated version of the RNA-binding protein that is not capable of binding to its target RNAs. The RNA-binding protein variants are expressed under the control of their authentic promoter and cis-regulatory motifs. Incubation of the plants with formaldehyde in vivo cross-links the proteins to their RNA targets. A whole-cell extract is then prepared and subjected to immunoprecipitation with an antibody against the GFP tag and to mock precipitation with an antibody against the unrelated red fluorescent protein. The RNAs coprecipitating with the proteins are eluted from the immunoprecipitate and identified via reverse transcription-PCR. Key words Circadian, RNA immunoprecipitation, Posttranscriptional, RNA-binding protein, Formaldehyde cross-linking, Green fluorescent protein, GFP trap beads

1

Introduction Posttranscriptional control emerges as an important layer of regulation in the circadian timing system. The endogenous circadian clock allows organisms to time biochemical and physiological processes to the optimal phase of the day–night cycle. Towards this end, the clock causes large parts of the transcriptome to fluctuate with a 24-h rhythm [1, 2]. It is well documented that global changes in transcription rate throughout the day widely account for circadian transcript oscillations [3]. Moreover, the Arabidopsis circadian clock itself represents a multilevel regulatory network consisting of at least three interwoven

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_7, © Springer Science+Business Media New York 2014

107

108

Tino Köster et al.

transcriptional/translational feedback loops that generate the 24-h rhythm in clock gene expression [4, 5]. With time, it was observed that transcription rates and circadian oscillations in mRNA steady-state abundance did not match, raising an awareness for an additional layer of control at the posttranscriptional level in the circadian system [6–8]. Furthermore, RNA degradation rates were found to change over the circadian cycle [9, 10]. RNA-binding proteins that undergo high-amplitude circadian oscillations were obvious candidates to participate in these processes [11–14]. Additionally, clock phenotypes were observed in mutants of predicted posttranscriptional regulators. A mutation in Arabidopsis protein arginine methyltransferase 5 (PRMT5) lengthens the period of transcript oscillations [15]. Among proteins methylated by PRMT5 are spliceosomal proteins that are required for the biogenesis of small nuclear ribonucleoproteins that participate in spliceosome function [16, 17]. Most notably, the mutant causes missplicing of the transcript encoding the clock component PSEUDORESPONSE REGULATOR 9, linking alternative splicing to the core oscillator mechanism. Subsequently, a mutant defective in the putative RNAbinding protein SPLICEOSOMAL TIMEKEEPER LOCUS1 that shows homology to a spliceosomal protein in humans and yeast was shown to have a long-period phenotype [18]. The heterogeneous nuclear ribonucleoparticle protein (hnRNP)-like Arabidopsis thaliana glycine-rich RNA-binding protein 7 (AtGRP7) is another splicing factor that has been shown to play a regulatory role in the circadian system through reverse genetics [19]. AtGRP7 binds to its own pre-mRNA and negatively autoregulates by alternative splicing linked to degradation of the alternative splice form via nonsense-mediated decay [20, 21]. Furthermore, it regulates alternative splicing of other target transcripts, and this was shown to occur by direct binding of AtGRP7 to some of these targets [22, 23]. Independently, it was observed that alternative splicing plays a prominent role in circadian timekeeping and response of the clock to temperature changes [24–27]. In mammals, the clock-regulated deadenylase nocturnin plays a role in transcript stability [28] and the heterogeneous nuclear ribonucleoprotein hnRNPQ regulates translation of the clock protein PERIOD1 by binding to its 5′ untranslated region [29]. To unravel the complete posttranscriptional networks that are controlled by RNA-binding proteins it is critical to capture the full complement of mRNAs bound by the protein of interest in vivo. To this end, RNA immunoprecipitation (RIP) has been adapted for the use in plants. Endogenous messenger ribonucleoprotein (mRNP) complexes assembled by an RNA-binding protein and its target RNAs are precipitated by virtue of an antibody against the RNA-binding protein itself or an epitope tag, and subsequently the RNAs are identified, e.g., by reverse transcription (RT)-PCR.

RNA Immunoprecipitation

109

Fig. 1 Schematic outline of the RIP strategy. Transgenic plants expressing a GFP-tagged RBP (right ) and plants expressing a GFP-tagged mutant version of the RBP impaired in the RNA-binding activity (left ) are subjected to formaldehyde cross-linking. Whole-cell extracts are prepared and incubated with GFP-Trap® beads for immunoprecipitation of cross-linked RBP–RNA complexes (and in parallel with RFP-Trap® beads for mock immunoprecipitation as a negative control, not depicted). RNA is isolated from the input fraction (IN), the mock immunoprecipitate (IP−), and the immunoprecipitate (IP+). The RNA is reverse transcribed, and the cDNA is analyzed by real-time PCR using candidate primers. RIP of the GFP-tagged RBP results in an enrichment of target transcript, whereas no enrichment is detectable in RIP with the mutated GFP-RBP

The first step in this procedure is an in vivo cross-linking step to stabilize transient and weak RNA–protein interactions (Fig. 1). Thus, more stringent washing conditions can be chosen and the rearrangement of RNA–protein complexes upon cell lysis is

110

Tino Köster et al.

reduced [30]. In Arabidopsis, recovery of the small spliceosomal U2 snRNA upon immunoprecipitation of the cognate spliceosomal U2B″ protein has been improved by almost three orders of magnitude upon chemical cross-linking through formaldehyde fixation compared to immunoprecipitation without prior fixation [31]. After cell lysis, mRNP particles are precipitated from native cellular extracts by virtue of a bead-coupled antibody. Subsequently, target mRNAs are identified in the RNA fraction recovered from the immunoprecipitate through RT-PCR. The first version of this protocol compared RIP performed with transgenic plants expressing an epitope-tagged RNA-binding protein and control plants expressing the epitope tag by itself side by side [22, 32]. The use of GFP as a tag allows the use of a highaffinity GFP antibody immobilized on agarose beads. These GFP-Trap® beads contain a covalently linked GFP-binding protein representing the GFP-recognizing domain of a heavy-chain antibody raised in alpaca (Lama pacos) [33, 34]. RFP-Trap® beads generated in an analogous way with antibodies directed against red fluorescent protein conveniently serve as the unrelated antibody. To increase the stringency, we describe here differential RIP that relies on an RNA-binding protein and a mutant version that has lost its RNA-binding capability. Such a strategy that discriminates between specific and unspecific nucleic acid–protein interaction through mutagenesis has been used to purify RNA-binding proteins or DNA-binding proteins [35, 36]. To create a mutant version of AtGRP7 that would have reduced RNA-binding affinity we focused on a conserved arginine residue within the RNA recognition motif, R49. Mutation of R49 into Q reduces the in vitro binding affinity of recombinant AtGRP7 [37]. In plants, the mutation abolishes the negative autoregulation of AtGRP7 and regulation of downstream targets [22, 37, 38]. Notably, the very arginine is targeted by the Pseudomonas syringae type III effector protein HopU1 as part of its virulence strategy. ADP ribosylation of R49 by HopU1 interferes with the RNAbinding capability of AtGRP7 in vitro and in vivo and impairs plant defense responses [39–42]. To closely match the spatial and temporal expression pattern of the endogenous RNA-binding protein under study, the constructs are driven by the endogenous promoter as well as the authentic cis-regulatory sequences within the transcribed part of the gene, i.e., 5′ UTR, 3′ UTR, and introns. Retention of in vivo functionality is demonstrated by showing that the wild-type fusion protein complements a loss-of-function mutant. In contrast, the mutant version is expected to be nonfunctional and to restore the phenotype to a lesser extent or not at all. This indeed was seen for AtGRP7 R49Q [40, 41]. Specific binding of an mRNA to the RNA-binding protein is indicated by an enrichment in the immunoprecipitate (IP+) relative

RNA Immunoprecipitation

Relative transcript level

a 2

2

PP2A IN

1

1

IPIP+

0

b Relative transcript level

FYD

111

2

0

FYD

2

PP2A

IN 1

1

IPIP+

0

0

Fig. 2 In vivo interaction of AtGRP7-GFP but not the mutant AtGRP7 R49Q-GFP with a target transcript. qRT-PCR analysis of transgenic plants expressing AtGRP7-GFP (a) or AtGRP7 R49Q-GFP (b). RNA from the input fraction (IN), RIP with GFP-Trap® beads (IP+), and mock RIP with RFP-Trap beads® (IP−) are reverse transcribed and amplified using specific FYD (At3g12570) primers. In parallel, the level of the PP2A reference transcript is determined in each fraction. Data are based on three biological replicates

to the mock precipitate (IP−). However, processing of the sample (IP+) and mock precipitation (IP−) may vary, and the two different antibodies may generate different levels of background. Furthermore, the use of different primer sets may result in different background levels strongly influencing fold enrichment. Thus, as discussed for ChIP [43], just presenting RIP signals as fold enrichment over signals in mock precipitation is not recommended. In our opinion the use of indispensible controls and a combination of data normalization methods lead to the most reliable data. Although it is likely that IP and mock IP levels are not directly related to input level caused by differences in sample processing, transcript levels in IP and mock IP relative to input give us important information for data interpretation [22, 31, 44]. Therefore, we decided to display input, IP, and mock IP side by side to a control transcript for both plants (Fig. 2). Also data normalization to background signal of a reliable control transcript is applicable. The RIP-qPCR method can be extended to the whole-genome level by preparing probes for hybridization of microarrays or tiling arrays RIP-chip and by constructing RNA libraries for nextgeneration sequencing (RIP-seq) (“ribonomics”) [45–50].

112

2

Tino Köster et al.

Materials

2.1 Growth of Transgenic Plants

1. Seeds of transgenic plants expressing the RNA-binding protein of interest and the mutated version, respectively, fused to GFP (see Notes 1 and 2). 2. Bleach: Half-strength commercial bleach (final concentration 6 % sodium hypochlorite) in 0.01 % Triton X-100. 3. 70 % (v/v) Ethanol. 4. Autoclaved H2O. 5. Half-strength Murashige–Skoog (MS) medium (1 L): 2.2 g MS powder (Duchefa), 0.5 g morpholinoethane sulfonate (MES), 50 g sucrose, adjust to pH 5.7 using 1 M KOH, add 10 g plant growth agar (Sigma), and autoclave.

2.2 Formaldehyde Cross-Linking

1. Formaldehyde (37 %). 2. 125 mM Glycine (sterilized by autoclaving). 3. Ice-cold autoclaved water. 4. Desiccator. 5. Water-jet vacuum pump. 6. Erlenmeyer flasks (250 mL). 7. Absorbent paper tissue (e.g., Kleenex).

2.3 Preparation of Beads

1. Sepharose beads (e.g., IBA). 2. GFP-Trap® agarose beads (Chromotek) (see Note 3). 3. RFP-Trap® agarose beads (Chromotek). 4. DNA LoBind® tubes (Eppendorf). 5. Complete® Protease inhibitor tablets EDTA-free (Roche), phenylmethylsulfonyl fluoride (PMSF) 1 M in isopropanol. 6. RNA immunoprecipitation lysis buffer without RNase inhibitors (RIP-LB-): 50 mM Tris–HCl pH 7.5, 150 mM NaCl, 4 mM MgCl2, 0.25 % IGEPAL, 0.25 % sodium deoxycholate, 1 % SDS (see Note 4). Immediately before use, add 5 mM DTT, 1 mM PMSF, and Complete® Protease inhibitor tablets EDTA-free (Roche) according to the instructions of the supplier.

2.4 Preparation of Total Extract

1. Liquid N2. 2. RiboLock™ (Thermo Scientific). 3. Vanadylribosyl complex (VRC) RNAse inhibitor (New England Biolabs). 4. Heparin sodium salt (Roth). 5. Complete® Protease inhibitor tablets EDTA-free (Roche), PMSF 1 M in isopropanol.

RNA Immunoprecipitation

113

6. RNA immunoprecipitation lysis buffer (RIP-LB): 50 mM Tris–HCl pH 7.5, 150 mM NaCl, 4 mM MgCl2, 0.25 % IGEPAL, 0.25 % sodium deoxycholate, 1 % SDS, 1 mg/mL Heparin sodium salt. Before use, add 5 mM DTT, 10 mM VRC, 100 U RiboLock™/mL, 1 mM PMSF, and Complete® Protease inhibitor tablets according to the instructions of the supplier (see Note 5). 7. Mortar and pestle. 8. Eppendorf shaker and steel balls ø 4 mm. 9. Disposable plastic syringe (10 mL). 10. 0.45 μm sterile filter. 2.5 RNA Immunoprecipitation

1. RNA immunoprecipitation-lysis Subheading 2.4).

buffer

(RIP-LB)

(see

2. RNA immunoprecipitation-wash buffer (RIP-WB): 50 mM Tris–HCl pH 7.5, 500 mM NaCl, 4 mM MgCl2, 0.5 % IGEPAL, 0.5 % sodium deoxycholate, 1 % SDS, 2 M urea. Immediately before use, add 2 mM DTT. 3. End-over-end rotator. 4. DNA LoBind® tubes (Eppendorf). 2.6

Isolation of RNA

1. TRI reagent: 0.8 M guanidinium thiocyanate, 0.4 M ammonium thiocyanate, 4.35 % (w/v) glycerol, 0.1 M sodium acetate pH 5, 38 % (v/v) acidic phenol (see Note 6). 2. Chloroform:isoamyl alcohol (24:1). 3. Isopropanol, 70 % (v/v) ethanol pre-chilled at −20 °C. 4. Ambion® GlycoBlue™ (Life Technologies). 5. Eppendorf shaker.

2.7

DNase Treatment

2.8 RNA Precipitation with LiCl 2.9 Reverse Transcription

RQ1 DNase, RQ1 buffer, and RQ1 Stop solution (Promega). LiCl 8 M. 1. Reverse transcriptase, 5× RT buffer, and 0.1 mM DTT (Roboklon). 2. Deoxynucleotide triphosphate 5 mM stock solution. 3. Random hexamer primers p(dN)6 (e.g., Life Technologies). 4. RiboLock™ (Thermo Scientific).

2.10 Reverse Transcription and Real-Time PCR

1. SYBR Green real-time PCR kit (e.g., Bio-Rad). 2. Primers for candidate and reference transcripts (see Note 7).

114

3

Tino Köster et al.

Methods

3.1 Growth of Transgenic Plants

1. To surface-sterilize Arabidopsis seeds, put seeds into microfuge tube. We plate 100–150 seeds per petri dish (85 mm ø) and use three plates per immunoprecipitation. Add 1 mL of 70 % ethanol, mix, and remove ethanol. 2. Add 1 mL of bleach, mix by inverting several times during 5 min, microcentrifuge at 7,500 × g, and remove bleach. 3. Wash seeds with sterile H2O, spin at 7,500 × g, and remove H2O by pipetting. 4. Repeat step 3 twice. 5. Place the microfuge tube at 4 °C for 2 days in the dark. This stratification step helps to synchronize germination of the seeds. 6. Plate seeds on half-strength MS plates supplemented with 0.5 % sucrose. Germinate and grow plants in defined light– dark cycles at 20 °C.

3.2 Formaldehyde Cross-Linking

1. Harvest aerial parts of the plants using a razor blade and immediately transfer to an Erlenmeyer flask containing 200 mL of 1 % formaldehyde (see Note 8). 2. Place the flask inside a desiccator. Apply vacuum for 15 min (see Note 9). This step ensures efficient penetration of the cross-linking agent into the plant cell by displacing the gas in the intercellular space. Take care while releasing the vacuum. 3. Stop the cross-linking reaction by pouring off the formaldehyde and adding 200 mL of 125 mM glycine. Apply vacuum for 5 min, carefully release vacuum, and remove glycine solution. 4. Wash plants with ice-cold sterile H2O four times. 5. Remove residual H2O by blotting onto paper tissue. 6. Immediately transfer plants to liquid N2. Store the material at −80 °C until use.

3.3 Preparation of Beads

A major source of false-positive RNAs identified in the RNA immunoprecipitation is contaminating RNAs that bind to the beads. This nonspecific interaction is minimized by preclearing the cellular lysate with plain beads (see Note 10). 1. To prepare the plain sepharose beads for the preclearing step, cut off the end of the tips and pipette 100 μL of plain sepharose beads each (50 % slurry) in two microfuge tubes. Add 1 mL ice-cold RIP-LB, and incubate for 5 min at 4 °C. Microcentrifuge for 1 min at 500 × g at 4 °C. Repeat the washing step twice. Store at 4 °C until use.

RNA Immunoprecipitation

115

2. To prepare the antibody beads for immunoprecipitation (IP+), wash 15 μL of GFP-Trap® beads (50 % slurry) with 1 mL icecold RIP-LB buffer in a DNA LoBind® microfuge tube for 5 min at 4 °C (see Note 11). Microcentrifuge for 1 min at 500 × g. Repeat this washing step twice. 3. To prepare the antibody beads for the mock precipitation (IP−), wash 15 μL of RFP-Trap® beads (50 % slurry) with 1 mL ice-cold RIP-LB buffer in a DNA LoBind® microfuge tube for 5 min at 4 °C (see Note 11). Microcentrifuge for 1 min at 500 × g. Repeat this washing step twice. 3.4 Preparation of Total Extract

1. Pre-chill mortar and pestle with liquid N2. 2. Grind frozen plant material to a fine powder under liquid N2. 3. Puncture the lids of 2 mL microfuge tubes, and pre-chill tubes in liquid N2. 4. Transfer ground plant material corresponding to 0.5 g to prechilled 2 mL microfuge tubes. Per line use 3 × 0.5 g distributed on three 2 mL microfuge tubes. 5. Remove the microfuge tubes with ground plant material from liquid N2, and add 750 μL of buffer RIP-LB prewarmed to 60 °C to each tube. 6. Add a steel ball to each tube. Replace each punctured lid by an intact lid of a new tube. Shake tubes in an Eppendorf shaker at 40 °C until you get a viscous homogenate (see Note 12). 7. Microcentrifuge at maximal speed at 4 °C for 10 min. 8. Combine the supernatants, and filter the cell extract using a syringe and a 0.45 μm filter (see Note 13).

3.5 RNA Immunoprecipitation

An aliquot of the cellular lysate is taken before preclearing. This “input” (IN) sample represents the RNA complement employed for RIP and serves as a positive control for the presence of the transcripts under study. Furthermore, a mock precipitation is done using beads carrying the unrelated antibody (IP−). RNAs recovered in this mock IP reflect noise during processing of the extract, e.g., RNAs that nonspecifically interact with the antibody. 1. Remove 100 μL of the supernatant (see Subheading 3.4, step 8) and store at 4 °C until RNA isolation (see Subheading 3.6, step 1). This is the “input” (IN) control. 2. To preclear the extracts for immunoprecipitation (IP+) and mock immunoprecipitation (IP−), transfer 1 mL of the extract (see Subheading 3.4, step 8) to 50 μL of washed sepharose beads (see Subheading 3.3, step 1) twice. Mix by vortexing. Incubate on an end-over-end rotator for 1 h at 4 °C. 3. Microcentrifuge both precleared samples (IP+) and precleared controls (IP−) for 2 min at 500 × g and 4 °C to pellet the sepharose beads.

116

Tino Köster et al.

4. Transfer the supernatant of the precleared samples (IP+) to DNA LoBind® tubes with washed GFP-Trap® beads (Subheading 3.3, step 2) for immunoprecipitation (see Note 14). 5. Transfer supernatant of precleared controls (IP−) to microfuge tubes with washed RFP-Trap® beads (Subheading 3.3, step 3). 6. Incubate samples and controls on an end-over-end rotator for 1 h at 4 °C. 7. Microcentrifuge for 1 min at 500 × g, and discard supernatants. 8. Wash beads with 1 mL of ice-cold RIP-WB for 10 min at 4 °C and microcentrifuge for 1 min at 500 × g (see Note 15). 9. Repeat washing (step 9) two more times. 10. Transfer beads to new DNA LoBind® tubes. 11. Repeat washing (step 9) two more times. 12. Wash beads with 1 mL of ice-cold RIP-LB at 4 °C for 5 min and microcentrifuge for 1 min at 500 × g (see Note 16). 3.6

Isolation of RNA

1. RNA is purified from the immunoprecipitated RNPs (IP+) and mock immunoprecipitates (IP−) and from the input (IN), respectively, using the TRIzol® procedure (see Note 17). Add 400 μL of TRI reagent to the GFP-Trap® beads from the immunoprecipitation and the RFP Trap® beads from the control precipitation, respectively. Add 400 μL of TRI reagent to 100 μL of the input control (see Subheading 3.5, step 1). 2. Vortex for 15 s. 3. Incubate for 5 min at 65 °C and 1,400 rpm in an Eppendorf shaker (see Note 18). 4. Add 100 μL of chloroform:isoamyl alcohol, mix vigorously for 3 min, and incubate for a further 3 min at room temperature. 5. Microcentrifuge at 12,000 × g and 4 °C for 15 min. Carefully remove upper phase and transfer to a new microfuge tube (see Note 19). 6. Add 1 μL of GlycoBlue™ and 400 μL isopropanol, mix, and incubate for 30–45 min at room temperature. GlycoBlue™ acts as a carrier to enhance the recovery and improves the visibility of the RNA pellet. 7. Microcentrifuge at 12,000 × g at 4 °C for 10 min, and discard supernatant. 8. Wash RNA pellets with ethanol pre-chilled at −20 °C and microcentrifuge at 12,000 × g at 4 °C for 5 min (see Note 20). 9. Repeat the washing step. 10. Air-dry RNA pellets and resuspend in 17 μL RNase-free water (see Note 21).

RNA Immunoprecipitation

3.7

DNase Digestion

117

1. Add 1 μL of RQ1 buffer and 2 μL RQ1 DNase to 17 μL of RNA samples. Mix by flicking the tube or by pipetting up and down (see Note 22). Incubate for 15 min at 37 °C. 2. Add 2 μL of RQ1 stop solution, and incubate for 10 min at 65 °C. 3. Remove 1 μL to control success of DNase treatment by PCR (-RT control) (see Note 23).

3.8 RNA Precipitation with LiCl

1. To DNase-treated RNA, add 80 μL of RNase-free water. 2. Add 35 μL 8 M LiCl, and precipitate RNA at 4 °C overnight. 3. Microcentrifuge at 12,000 × g at 4 °C for 45 min, and discard supernatants. 4. Wash RNA pellets with ethanol pre-chilled at −20 °C and microcentrifuge at 12,000 × g at 4 °C for 5 min (see Note 20). 5. Repeat the washing step. 6. Air-dry RNA pellets and resuspend in 12.25 μL RNase-free water (see Note 20).

3.9 Reverse Transcription

1. To 12.75 μL of RNA, add 1.25 μL of random hexamer primers (200 ng/μL) and 1 μL 5 mM dNTPs and incubate for 5 min at 65 °C. 2. Add 4 μL of 5× RT buffer, 1 μL 0.1 M DTT, and 0.5 μL reverse transcriptase. 3. Incubate for 1 h at 42 °C. 4. Incubate for 10 min at 80 °C, and add 60 μL of H2O.

3.10

Real-Time PCR

1. To set up a standard real-time PCR reaction, use 2 μL of singlestranded cDNA for the analysis with SYBR Green Kits according to the manufacturer’s instruction. Include a control reaction without template to detect contaminations of the reagents and the formation of primer dimers. Perform each reaction in triplicates (see Note 24). 2. Expression levels are determined for the gene of interest and a control transcript, e.g., PP2A, in the IP fraction, the mock IP fraction, and the input fraction, respectively.

4

Notes 1. The amino acid exchange is introduced into a cDNA for the RNA-binding protein via site-specific mutagenesis. Although kits are available, we have made good experience in doing the mutagenesis with standard PCR reagents using overlapping primers to introduce the mutation and a Taq polymerase with proofreading activity. To easily identify the mutated clone it is

118

Tino Köster et al.

of use to introduce (or remove) a restriction site overlapping the amino acid(s) to be exchanged. Likely this will need exchanges of additional bases neighboring the amino acid(s) to be mutated, and care has to be taken that only silent mutations are created. Sequencing of the final mutated construct is imperative. The generation of the AtGRP7 R49Q mutation has been described in detail [37]. 2. Other epitopes used for RIP include the hemagglutinin tag [41] or FLAG tag [51]. 3. We obtain similar results by using GFP-Trap® coupled to magnetic beads. A systematic investigation of proteins binding nonspecifically to commonly used affinity matrices, designated the “bead proteome,” showed that less unspecific binding for cytoplasmic extracts occurred on sepharose or agarose beads whereas magnetic beads showed lower nonspecific binding for nucleic acid-associated proteins and, thus, nuclear extracts [52]. 4. Although we use fixed plant material we achieve better results by including MgCl2 in the buffer. Certain RNA–protein interactions may depend on Mg2+. 5. VRC has been reported to inhibit the RT reaction. If this appears to be a problem, VRC should be omitted or depleted by adding ten equivalents of EDTA before the isopropanol precipitation (Subheading 3.6, step 6). 6. Commercial versions of TRI reagent are available, e.g., TRIzol (Invitrogen). 7. Primers for real-time PCR should be designed using standard programs. We use PP2A (At1g13320) as a reference transcript. The following primer pair flanking an intron is used for amplification: PP2A_for CGATAGTCGACCAAGCGGTT and PP2A_rev TACCGAACATCAACATCTGG. The primers will amplify an 88 bp fragment from cDNA and a 210 bp fragment from contaminating genomic DNA. 8. If the RNA-binding protein under study and/or the target transcripts are expected to undergo circadian oscillations, the time of day to harvest the plants has to be chosen. If root material is harvested, care has to be taken to remove agar adhering to the roots. 9. Specificity and efficiency of formaldehyde cross-linking depend on the duration of the treatment. Thus, it is advisable to do a pilot time course for each RNA-binding protein. 10. Preparation of beads (Subheading 3.3) has to be completed before the extracts are prepared. 11. The amount of beads can be adjusted to the amount of fusion protein in the cellular extract. 12. This is a critical step: Homogenize as fast as possible to avoid RNA degradation.

RNA Immunoprecipitation

119

13. Clearing the cell extract from any cell and membrane debris is crucial for a good signal-to-noise ratio. 14. Use DNA-LoBind® tubes (Eppendorf) to avoid unspecific binding of proteins and nucleic acids to walls of the tube during IP. 15. Washing conditions have to be established carefully to reduce unspecific binding as much as possible but to avoid dissociation of the specific RNA–protein interactions and of the RBP– antibody interaction. Stringency of washing buffer can be changed by adjusting the concentration of NaCl, urea, or other chaotropic salts, e.g., LiCl. 16. After the last washing step, successful binding of the GFP fusion protein can be monitored under a fluorescence stereomicroscope [32]. 17. Recovery of RNA is most efficient by using Tri-Reagent. The use of commercial kits based on column purification of RNA is not recommended. 18. We incubate the samples at 65 °C to substitute for the heat treatment commonly used to reverse the formaldehyde crosslinking [53]. 19. The interphase contains the beads. The organic phase can be used to recover immunoprecipitated proteins, as described in the Invitrogen TRIZOL® product information, http://tools. invitrogen.com/content/sfs/manuals/trizol_reagent.pdf [54]. 20. Be careful when washing the RNA pellet and decanting the supernatant because the pellet can detach from the wall of the tube very easily. 21. The RNA concentration in the input control is around 200– 400 μg/mL. A260/280 should be ≥2. A260/230 should be between 2.0 and 2.2, but often it is very low (0.5–1.0) apparently without negative effects [48]. The RNA concentration in the immunoprecipitate is below the detection limit. 22. Do not vortex. 23. PCR products obtained at this stage indicate contamination with genomic DNA. 24. For high-abundance transcripts, a single PCR is sufficient. For low-abundance transcripts we recommend to perform two sequential rounds of amplification: After 20–24 cycles in a standard PCR cycler remove 2–4 μL of the PCR and start a second round of PCR using nested primers.

Acknowledgements We thank Kristina Neudorf for expert technical assistance. This work was supported by the DFG (STA 653).

120

Tino Köster et al.

References 1. Harmer SL, Hogenesch JB, Straume M et al (2000) Orchestrated transcription of key pathways in Arabidopsis by the circadian clock. Science 290:2110–2113 2. Michael TP, Mockler TC, Breton G et al (2008) Network discovery pipeline elucidates conserved time-of-day-specific cis-regulatory modules. PLoS Genet 4:e14 3. Michael TP, McClung CR (2003) Enhancer trapping reveals widespread circadian clock transcriptional control in Arabidopsis. Plant Physiol 132:629–639 4. Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22:R648–R657 5. Staiger D, Shin J, Johansson M et al (2013) The circadian clock goes genomic. Genome Biol 14:208 6. Kojima S, Shingle DL, Green CB (2011) Posttranscriptional control of circadian rhythms. J Cell Sci 124:311–320 7. Staiger D, Köster T (2011) Spotlight on posttranscriptional control in the circadian system. Cell Mol Life Sci 68:71–83 8. Staiger D, Green R (2011) RNA-based regulation in the plant circadian clock. Trends Plant Sci 16:517–523 9. So WV, Rosbash M (1997) Post-transcriptional regulation contributes to Drosophila clock gene mRNA cycling. EMBO J 16:7146–7155 10. Gutierrez RA, Ewing RM, Cherry JM et al (2002) Identification of unstable transcripts in Arabidopsis by cDNA microarray analysis: rapid decay is associated with a group of touchand specific clock-controlled genes. Proc Natl Acad Sci U S A 99:11513–11518 11. Staiger D (2001) RNA-binding proteins and circadian rhythms in Arabidopsis thaliana. Philo Trans R Soc Lond B Biol Sci 356: 1755–1759 12. Wang D, Liang X, Chen X et al (2013) Ribonucleoprotein complexes that control circadian clocks. Int J Mol Sci 14:9018–9036 13. Mittag M (2003) The function of circadian RNA-binding proteins and their cis-acting elements in microalgae. Chronobiol Int 20: 529–541 14. Newby LM, Jackson FR (1996) Regulation of a specific circadian clock output pathway by lark, a putative RNA-binding protein with repressor activity. J Neurobiol 31:117–128 15. Sanchez SE, Petrillo E, Beckwith EJ et al (2010) A methyl transferase links the circadian clock to the regulation of alternative splicing. Nature 468:112–116

16. Deng X, Gu L, Liu C et al (2010) Arginine methylation mediated by the Arabidopsis homolog of PRMT5 is essential for proper pre-mRNA splicing. Proc Natl Acad Sci U S A 107:19114–19119 17. Hong S, Song HR, Lutz K et al (2010) Type II protein arginine methyltransferase 5 (PRMT5) is required for circadian period determination in Arabidopsis thaliana. Proc Natl Acad Sci U S A 107:21211–21216 18. Jones MA, Williams BA, McNicol J et al (2012) Mutation of Arabidopsis spliceosomal timekeeper locus1 causes circadian clock defects. Plant Cell 24:4907–4916 19. Staiger D, Heintzen C (1999) The circadian system of Arabidopsis thaliana: forward and reverse genetic approaches. Chronobiol Int 16:1–16 20. Staiger D, Zecca L, Wieczorek Kirk DA et al (2003) The circadian clock regulated RNAbinding protein AtGRP7 autoregulates its expression by influencing alternative splicing of its own pre-mRNA. Plant J 33:361–371 21. Schöning JC, Streitner C, Meyer IM et al (2008) Reciprocal regulation of glycine-rich RNA-binding proteins via an interlocked feedback loop coupling alternative splicing to nonsense-mediated decay in Arabidopsis. Nucleic Acids Res 36:6977–6987 22. Streitner C, Köster T, Simpson CG et al (2012) An hnRNP-like RNA-binding protein affects alternative splicing by in vivo interaction with target transcripts in Arabidopsis thaliana. Nucleic Acids Res 40:11240–11255 23. Streitner C, Simpson CG, Shaw P et al (2013) Small changes in ambient temperature affect alternative splicing in Arabidopsis thaliana. Plant Signal Behav 8:e24638 24. James AB, Syed NH, Bordage S et al (2012) Alternative splicing mediates responses of the Arabidopsis circadian clock to temperature changes. Plant Cell 24:961–981 25. Filichkin SA, Priest HD, Givan SA et al (2010) Genome-wide mapping of alternative splicing in Arabidopsis thaliana. Genome Res 20:45–58 26. Sanchez SE, Petrillo E, Kornblihtt AR et al (2011) Alternative splicing at the right time. RNA Biol 8:954 27. Bartok O, Kyriacou CP, Levine J et al (2013) Adaptation of molecular circadian clockwork to environmental changes: a role for alternative splicing and miRNAs. Proc R Soc B Biol Sci 280:20130011 28. Stubblefield JJ, Terrien J, Green C (2012) Nocturnin: at the crossroads of clocks and metabolism. Trends Endocrinol Metabol 23:326

RNA Immunoprecipitation 29. Lee KH, Woo KC, Kim DY et al (2011) Rhythmic interaction between Period1 mRNA and HnRNP Q leads to circadian timedependent translation. Mol Cell Biol 32:717 30. Mili S, Steitz JA (2004) Evidence for reassociation of RNA-binding proteins after cell lysis: implications for the interpretation of immunoprecipitation analyses. Rna 10:1692–1694 31. Terzi LC, Simpson GG (2009) Arabidopsis RNA immunoprecipitation. Plant J 59: 163–168 32. Köster T, Staiger D (2014) RNA-binding protein immunoprecipitation from whole-cell extracts. Arabidopsis protocols. Methods Mol Biol 1062:679–695 33. Rothbauer U, Zolghadr K, Tillib S et al (2006) Targeting and tracing antigens in live cells with fluorescent nanobodies. Nat Methods 3: 887–889 34. Rothbauer U, Zolghadr K, Muyldermans S et al (2008) A versatile nanotrap for biochemical and functional studies with fluorescent fusion proteins. Mol Cell Proteomics 7:282–289 35. Staiger D, Becker F, Schell J et al (1991) Purification of tobacco nuclear proteins binding to a CACGTG motif of the chalcone synthase promoter by DNA affinity chromatography. Eur J Biochem 199:519–527 36. Zhao B, Schneid C, Iliev D et al (2004) The circadian RNA-binding protein CHLAMY 1 represents a novel type heteromer of RNA recognition motif and lysine homology domain-containing subunits. Eukaryot Cell 3: 815–825 37. Schöning JC, Streitner C, Page DR et al (2007) Autoregulation of the circadian slave oscillator component AtGRP7 and regulation of its targets is impaired by a single RNA recognition motif point mutation. Plant J 52:1119–1130 38. Streitner C, Hennig L, Korneli C et al (2010) Global transcript profiling of transgenic plants constitutively overexpressing the RNA-binding protein AtGRP7. BMC Plant Biol 10:221 39. Fu ZQ, Guo M, Jeong BR et al (2007) A type III effector ADP-ribosylates RNA-binding proteins and quells plant immunity. Nature 447:284–288 40. Jeong B-R, Lin Y, Joe A et al (2011) Structure function analysis of an ADP-ribosyltransferase type III effector and its RNA-binding target in plant immunity. J Biol Chem 286: 43272–43281 41. Nicaise V, Joe A, Jeong B et al (2013) Pseudomonas HopU1 affects interaction of

42. 43.

44.

45.

46.

47.

48.

49.

50. 51.

52.

53.

54.

121

plant immune receptor mRNAs to the RNAbinding protein GRP7. EMBO J 32:701–712 Staiger D, Korneli C, Lummer M et al (2013) Emerging role for RNA-based regulation in plant immunity. New Phytol 197:394–404 Haring M, Offermann S, Danker T et al (2007) Chromatin immunoprecipitation: optimization, quantitative analysis and data normalization. Plant Methods 3:11 Zhao J, Ohsumi TK, Kung JT et al (2010) Genome-wide Identification of PolycombAssociated RNAs by RIP-seq. Mol Cell 40: 939–953 Hafner M, Landthaler M, Burger L et al (2010) Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 141:129–141 Tenenbaum SA, Carson CC, Lager PJ et al (2000) Identifying mRNA subsets in messenger ribonucleoprotein complexes by using cDNA arrays. Proc Natl Acad Sci U S A 97: 14085–14090 Mukherjee N, Corcoran DL, Nusbaum JD et al (2011) Integrative regulatory mapping indicates that the RNA-binding protein HuR couples pre-mRNA processing and mRNA stability. Mol Cell 43:327–339 Galgano A, Gerber AP (2011) RNA-binding protein immunopurification-microarray (RIPChip) analysis to profile localized RNAs. Methods Mol Biol 714:369–385 Jensen KB, Darnell RB (2008) CLIP: crosslinking and immunoprecipitation of in vivo RNA targets of RNA-binding proteins. Methods Mol Biol 488:85–98 Ule J, Jensen KB, Ruggiu M et al (2003) CLIP identifies Nova-regulated RNA networks in the brain. Science 302:1212–1215 Knuckles P, Vogt MA, Lugert S et al (2012) Drosha regulates neurogenesis by controlling neurogenin 2 expression independent of microRNAs. Nat Neurosci 15:962–969 Trinkle-Mulcahy L, Boulon S, Lam YW et al (2008) Identifying specific protein interaction partners using quantitative mass spectrometry and bead proteomes. J Cell Biol 183:223–239 Niranjanakumari S, Lasda E, Brazas R et al (2002) Reversible cross-linking combined with immunoprecipitation to study RNA-protein interactions in vivo. Methods 26:182–190 Hummon AB, Lim SR, Difilippantonio MJ et al (2007) Isolation and solubilization of proteins after TRIzol extraction of RNA and DNA from patient material following prolonged storage. Biotechniques 42:467–472

Chapter 8 A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data Using Integrated Genome Browser Alyssa A. Gulledge, Hiral Vora, Ketan Patel, and Ann E. Loraine Abstract Ultrahigh-throughput sequencing of cDNA (RNA-Seq) is an invaluable resource for investigating alternative splicing in an organism. Alternative splicing is a form of posttranscriptional regulation in which primary RNA transcripts from a single gene can be spliced in multiple ways leading to different RNA and protein products. In plants and other species, it has been shown that many genes involved in circadian regulation are alternatively spliced. As new RNA-Seq data sets become available, these data will lead to new insights into links between regulation RNA splicing and the circadian system. Analyzing RNA-Seq data sets requires software tools that can display RNA-Seq read alignments alongside gene models, enabling assessment of how treatments or developmental stages affect splicing patterns and production of novel variants. The Integrated Genome Browser (IGB) software program is a free and flexible desktop tool that enables discovery and quantification of alternative splicing. In this protocol, we use IGB and a cold-stress RNA-Seq data set to examine alternative splicing of Arabidopsis thaliana LHY, a circadian clock regulator. IGB is freely available from http://www.bioviz.org. Key words Genome browser, Visualization, Visual analytics, Alternative splicing, A. thaliana, LHY, LHY1, Circadian clock

1

Introduction Most genes in higher eukaryotes contain introns, sequence segments that are removed from the primary RNA transcript either co- or posttranscriptionally. The process of removing introns, called splicing, involves stepwise assembly of a macromolecular complex called the spliceosome onto the nascent primary RNA transcript. The spliceosome catalyzes removal of the intron and ligation of flanking exonic sequences via two transesterification reactions involving the five prime and three prime splice sites of the intron, also called donor and acceptor sites, and a branch point adenosine residue upstream of the three prime splice sites. The spliceosome consists of five ribonucleoproteins (U1, U2, U5, and U4/6 snRNPs) whose activity and interactions are modulated by a

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_8, © Springer Science+Business Media New York 2014

123

124

Alyssa A. Gulledge et al.

host of accessory proteins, including RNA helicases, SR proteins, hnRNP proteins, and others. Differential expression and activity of these accessory proteins influence splice site selection and also allow for different sites to be selected, a phenomenon known as alternative splicing (AS). AS can lead to production of alternative isoforms with distinct and sometimes antagonistic activities. Changes in the mRNA through differential inclusion of sequences can alter transcript localization, transcript longevity, or protein sequences. In animals, AS has been recruited as a regulatory mechanism enabling cellular diversity and differentiation, such as sex determination in insects and neuronal cell differentiation in the mammalian brain. According to conservative estimates based on annotated gene models [1], around 20 % of multi-exon genes in higher plants are thought to produce multiple transcript isoforms through AS. However, prior to the development of high-throughput sequencing of cDNA (RNA-Seq), it was difficult to estimate the prevalence and thus the biological significance of AS in plants. Studies that attempted to address the degree to which AS occurs in plants mainly used ESTs or tiling arrays to identify and quantify AS [2–4]. Our analysis of Arabidopsis thaliana ESTs found that for most genes annotated as producing multiple variants, one isoform tended to predominate, and the number of ESTs supporting the minor form was typically less than one in ten [1]. This result suggested that although many genes are capable of producing multiple isoforms through AS, the frequency with which this occurs is low. However, the study was limited by the heterogeneity of EST libraries as well as relatively small number of Arabidopsis ESTs (around 1.5 million) that were available at the time. We later repeated the analysis using new RNA-Seq data sets from Arabidopsis pollen and seedling and found essentially the same result: although around 15–20 % of alternatively spliced genes were found to produce the less abundant isoform in significant amounts, most annotated AS events were rarely observed [5]. Although this later RNA-Seqbased study involved many more sequences than the EST study that preceded it and thus had greater power to detect AS events, it should nonetheless be considered preliminary as only three libraries were sequenced. There will no doubt be many more RNA-Seq data sets published in the future that will yield more information about AS, including its prevalence and function in plant species. To help researchers take advantage of the existing and future RNA-Seq data sets, we have added new features to the Integrated Genome Browser (IGB) [6] that enable visual analysis of splicing patterns embedded in RNA-Seq data. This protocol explains how to use IGB to perform visual analysis of AS using Arabidopsis LHY as an example. LHY encodes an MYB transcription factor that together with CCA1 drives the morning loop of the circadian oscillator, a network of transcriptional regulators that activate or

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

125

repress gene expression according to the time of day. The LHY locus, similar to several other clock genes, undergoes extensive alternative splicing (for review, see [7]). LHY has been annotated by the Arabidopsis Information Resource (TAIR) as producing five distinct alternative splicing variants arising from alternative splicing in the 5′ UTR and from an exon-skipping event affecting the coding region. Previous analyses of RNA-Seq data observed that LHY also produces splicing variants in which introns 4 and 9 are sometimes retained [8, 9]. However, the relatively short read lengths of this early data set may have resulted in some AS events being missed. Another study that used a high-throughput qPCR panel to assess AS in circadian clock genes found additional novel splicing events in LHY, including an alternative acceptor site affecting intron 8 [10]. By reexamining LHY using RNA-Seq data with longer read lengths and by examining this data in IGB, we can recapitulate previous findings as well as report new aspects of LHY alternative splicing.

2

Materials The RNA-Seq data used in this protocol chapter were from two libraries prepared from cold-treated and control Arabidopsis thaliana seedlings. The data have not been published before now, and so we describe in detail how they were generated. Plants used in the experiment were sown onto soil in 4 in. pots and incubated for 7 days in a Percival incubator set to 22 °C, 45 % relative humidity, under long-day (16-h/8-h light/dark) illumination. At zt4 (zeitgeber time, 4 h after lights on) on the seventh day, pots selected at random to undergo a cold stress treatment were transferred to a similarly configured Percival incubator set to 4 °C. Relative humidity (RH) was adjusted to 75 % for each incubator as the colder incubator RH was difficult to maintain below this level. Following the transfer, nine samples from control and cold-treated samples were collected 45 min later, at zt7, zt10, and zt16 on the first day of treatment; zt7 on the second day; zt4, zt11, and zt17 on the third day; and zt2 on the fourth day. The above ground parts (shoots) were collected from two pots per treatment at each time point. Samples were flash-frozen on liquid nitrogen and stored at −80 °C prior to RNA extraction. Frozen samples from all time points were pooled, RNA was extracted, and cDNA libraries were prepared for Illumina sequencing as described previously [5], keeping treatment and control samples separate. Each library (treatment and control) was sequenced on an Illumina GAIIx sequencer for 75 cycles on two lanes per library, generating more than 40 million reads per library. The resulting 75-base, single-end sequences were aligned onto the latest Arabidopsis genome assembly (TAIR9, released in June 2009) using the TopHat spliced alignment tool [11].

126

Alyssa A. Gulledge et al.

Reads were then subdivided into two groups: reads that mapped to one location in the genome (SM, or single-mapping reads) and reads that mapped to more than one location in the genome (MM, or multi-mapping reads). The protocol and images described here refer to the SM reads only. The version of IGB used in this protocol is a prerelease version of IGB 7.1 corresponding to subversion repository revision 15,286. Alignment files are available from the main IGBQuickLoad site (igbquickload.org), and reads are available from the Short Read Archive via accession SRP022162.

3

Methods

3.1 Start IGB and Open Data Sets

1. Go to http://www.bioviz.org, and follow the links labeled “Download” to download and start IGB. IGB can run on any computer that has Java installed, but methods of launching IGB may vary depending on your system. For discussion of how to launch to IGB, see Note 1. 2. Open the latest Arabidopsis genome and reference gene models by clicking on the A. thaliana image on the start screen. Clicking the image triggers loading of the latest Arabidopsis genome assembly and associated gene models from the IGBQuickLoad.org Web site directly into IGB (Fig. 1). The data loading process may take several seconds or a few minutes depending on your network connection.

Fig. 1 Integrated Genome Browser after selecting the Arabidopsis genome. Plus and minus strand gene models from the TAIR10 mRNA gene model annotations are shown

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

127

3. Open cold stress and control RNA-Seq data sets. In the Available Data panel on the left side of the Data Access tab, open folder IGB Quickload > RNA-Seq > Loraine Lab > Mixed Cold > SM > Reads and select Control, alignments and Cold, alignments. Position the mouse over the sample labels to view a tooltip describing the data sets or click the “i” icon to open a Web page with details about the data sets. When you select a data set or open a file, the data set appears in the Data Management Table and a placeholder track appears in the Main View. See Note 2 for discussion of opening files from your desktop or Web sites. 4. Go to the LHY gene by entering “LHY” in the search box at the top left of the display. Note that when you enter this search term, IGB suggests several options, including “LHY1/LHY.” (LHY1 is a synonym of LHY according to the TAIR10 annotations.) Use the up and down arrow keys or the mouse to select option “LHY1/LHY,” and press ENTER to run the search. After the search, the Main View will zoom and pan to a closeup view of the LHY gene centered in the display. 5. Load the sequence data by clicking the Load Sequence button at the top right of the display to load the reference sequence, which appears as a gray bar in the Coordinate Axis track. Click within the gene to set the zoom focus, and drag the horizontal zoom slider at the top of the display to the right to zoom in to view the sequence. You can also zoom by clicking the + button to the left of the slider, click-dragging a region in the coordinate axis, or double-clicking an annotation feature. As you zoom in, the individual bases become visible. To re-center the LHY gene in the display, use the horizontal zoom slider or double-click on an intron. See Note 3 for more details on how to zoom in IGB. 6. Load the alignments data by clicking the button labeled Load Data at the top right of the display (see Note 4). When you click Load Data, the two previously empty tracks are filled with stacks of reads whose alignments correspond to the intron/exon structure of the LHY gene. 3.2

Adjust Main View

To facilitate visual analysis of splicing data, it is useful to configure the layout and presentation of reads and annotations in ways that reduce clutter and make biologically meaningful patterns easier to recognize. To adjust the view for splicing analysis of LHY, perform the following steps: 1. Edit track labels. When you first open a file from the IGBQuickLoad site, the full folder path to the file is displayed in the track label. To simplify the view, edit the track names. In the Data Management Table, click a track name to edit the

128

Alyssa A. Gulledge et al.

track name. After editing, press ENTER to trigger the change in the Main View. 2. Combine plus and minus gene model tracks, and adjust stack height. Click the track label to the left of the TAIR10 mRNA track to select it. Select the Annotation tab at the bottom of the IGB window, and choose the option labeled +/− under the Strand section to combine plus and minus strands into a single track. Also, enter “4” in the stack height section. This will cause IGB to use less vertical space to display the LHY gene models, which will be useful when other data sets and tracks are loaded. See Note 5 for discussion of strand options and stack height. 3. Make more space for visualization. Expand the space on your desktop available for visualization by hiding the Current Genome or other tabbed panels and enlarging the IGB window. After enlarging the IGB window, reposition the divider separating the main view and the bottom tabs so that the tabs occupying the smallest space are possible while still keeping the controls visible. 4. Lock track height. To ensure that the LHY gene models will remain the same size, no matter how much data is loaded into the main view, lock the gene model annotation track height. To lock track height to 200 pixels, select the track by clicking the TAIR10 mRNA track label, select the Annotation tab, choose Lock Track Height (Pixels), and enter 200. 5. Modify foreground, background, and track label colors. Select tracks by clicking the track labels, and use the Style section of the Annotation tab to change the color scheme for tracks. Also use the font option to increase or decrease the track label font size. To match figures shown in this chapter, use the following color scheme: white background for all tracks, black track labels for all tracks, black foreground for the reference gene model track (TAIR10 mRNA), and green and blue foreground colors for the control and cold stress alignment/read tracks, respectively. 6. Modify edge-matching color. For visual analysis of splicing, IGB provides an edge matching function designed to facilitate rapid comparisons between gene models and other items in the display. To activate edge matching, click the label of a gene model to select it and observe that the edges of all other items in the display that have matching boundaries become highlighted. To make these highlights easier to see with our chosen color scheme, choose File > Preferences > Other Options and change the Edge Matching color from white to red or black. See Note 6 for discussion of edge matching.

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

129

Fig. 2 Adjusting the main view to optimize the data analysis environment. The region surrounding the LHY gene is shown. The display has been adjusted as described in Subheading 3.2

7. Clamp to view. When viewing a single gene in detail, it is useful to restrict zooming to that one gene of interest. To restrict zooming to the LHY region, zoom to center LHY in the center of the display and select View > Clamp to View. While Clamp to View is active, the range of zooming and panning will be limited to the LHY region. After performing the preceding adjustments, the IGB window should look like the image shown in Fig. 2. 8. Save your work. After loading the data and configuring the IGB display, use the IGB session saving feature to save your work as an IGB session file. The next time you use IGB, you can load the IGB session file and begin again from this point, with colors, files, and other choices remembered and loaded. To save the session, choose File > Save Session. Note that Load Session, which you would use to open an IGB session file after re-starting IGB, is also in this menu.

130

Alyssa A. Gulledge et al.

3.3 Create and Analyze Junction Tracks

Genomic alignments of RNA-Seq reads derived from exon–exon junctions in spliced mRNAs typically contain gaps. These gaps define the former locations of introns and can be used to identify and analyze AS. Within the IGB display, gapped reads appear as alignment blocks separated by thin lines corresponding to the gaps. To facilitate analysis of splicing using gapped reads, IGB provides a visual analytics function called “FindJunctions” that you can use to quantify splicing choices and assess the prevalence of AS within an RNA-Seq data set. 1. Create FindJunctions tracks for each RNA-Seq data set. To create junction tracks for each RNA-Seq data set, select all of the sample tracks (SHIFT + click the track labels), select SingleTrack: Find Junctions in the Annotation tab, and click Apply. For each selected track, a new track will appear that contains junction features summarized from the read alignment tracks. Each junction feature has a numeric label indicating the number of gapped reads that supported the junction. By default, only single-mapping reads that align with at least five bases on either side of the intron are counted, and you can change these defaults by right-clicking a track and selecting the Track Operations > FindJunctions > Configure… menu item. See Note 7 on configuring FindJunctions. 2. Simplify the display by collapsing the alignment tracks. Click the − icon in the upper left corner of the alignment tracks to collapse the reads into a single row. 3. Reorganize the tracks. Click the track labels, and drag the junction tracks above their corresponding read tracks as shown in Fig. 3. 4. Use IGB zoom controls and edge matching to inspect the junctions and compare them to the annotated gene models. The FindJunctions track from the control sample contains 134 reads that support skipping of exon 5 (82 bases, 36,171– 36,089) and no reads that support its inclusion. However, the cold stress sample contains 90 reads in support of exon skipping and 9 and 16 reads supporting exon inclusion. Consistent with a previous examination of alternative splicing in LHY [10], cold stress increased the prevalence of the exon inclusion in LHY. Note that when you select a junction read, its name appears in the Selection Info box at the top right of the IGB display and the name indicates the location of the intron inferred from the gapped reads used to create the junction. Make note of junctions that have no corresponding intron in the LHY gene models. Some of these have only one or two supporting reads, while others have several. Among the best supported novel introns with respect to the annotated gene models is junction feature J:chr1:33858-33980, which has five

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

131

Fig. 3 Junction tracks created from RNA-Seq read tracks. Junction tracks derived from read alignment tracks are shown above the read alignment tracks used to create them. To conserve vertical space, read tracks are collapsed. Numbers above each junction feature indicate the number of gapped reads that supported it. The zoom stripe is positioned close to the novel acceptor site inferred from junction J:chr1:35963-36089

supporting reads and indicates a new splicing event affecting the three prime ends of the transcript. Another novel junction is J:chr1:35963-36089, which is immediately to the left of the optionally included exon (see Fig. 3) and thus is present in the cold stress and not the control data set. The donor site associated with this junction does not match any of the annotated gene models, but the acceptor site matches the boundary of the optionally included exon, as described in [10]. The RNASeq alignment data suggests that the acceptor site corresponding to this novel intron is the only one that is used and that the annotated acceptor is incorrect. 5. View the effects on protein sequence. In IGB, translated regions are shown as tall blocks with shorter blocks representing the five and three prime untranslated regions (UTRs). If the sequence has been loaded (which you did in an earlier step), you can zoom in on a location and observe the amino acid translation of a gene model. Zoom in on the option (sometimes skipped) exon 5, and look at the amino acid sequence. Compare the translations of the different gene

132

Alyssa A. Gulledge et al.

models, and note that including exon 5 does not introduce a premature stop codon. See Note 8 for discussion of viewing sequence data in IGB. 3.4 Use Depth Graphs to Detect Intron Retention

Depth graphs, also called coverage graphs, summarize the number of reads that overlap each base position within a genome. Depth graphs are useful for identifying intron retention (IR), a form of AS that is especially common in plant species. Depth graphs can also provide a rough estimate of the overall gene expression within a sample. 1. Create depth graphs. To create depth graphs for the RNA-Seq tracks and check for retained intron AS events, select the tracks as before (see preceding section), choose Single-Track: Depth in the Annotation tab, and click Apply. See Note 9 for discussion of coverage graphs. Use the Data Access tab as before if you wish to re-label the new depth graphs. 2. Simplify the display by hiding junction tracks. Since you will now focus on identifying introns, simplify the display by hiding tracks that are no longer needed. Right-click the track labels next to the junction tracks, and select Hide or use the track visibility icon (eye) next to the track’s listing within the Data Access tab. 3. Adjust the scale of the depth graphs. To use the depth graphs to find IR AS events, decrease the scale so that lower coverage regions are more obvious. Click the Graph tab, and click Select All button to select all graphs; in the section Y-axis Scale > Set by: Value, enter 20 in the Max text box to limit the upper boundary of the Y-axis. Observe that introns two and five have some coverage, indicating that these introns may sometimes be subject to intron retention (Fig. 4). See Note 9 for discussion of graph scaling. 4. Count retained intron reads to assess frequency of intron retention. Hide all the tracks except the gene models and cold reads. Adjust the horizontal zoom level to enlarge intron two so that it occupies the entire IGB screen, and use View > Clamp to View to restrict zooming to this region. Expand the read track using the + icon, then select the track, and click Optimize under the Annotation tab to show all the reads in the region. Use the vertical zoom controls to stretch the read track vertically so that you can see each read. Using the mouse, select all the reads that map within the intron. Use SHIFT-CLICK to select multiple reads and CNTRL-SHIFT-CLICK to remove reads from the current selection as needed. The Selection Info box (upper right) reports the number of items currently selected, which you can use to count the reads that overlap the intron (Fig. 5). Although IR does appear to occur, the number

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

133

Fig. 4 Coverage graphs created from RNA-Seq read tracks. Coverage graph tracks derived from read alignment tracks are shown above the read alignment tracks used to create them. To conserve vertical space, read tracks are collapsed. Each graph track has a scale on the left that shows the number of reads overlapping positions indicated in the coordinate track sequence axis

of gapped reads outnumbers the retained intron reads, suggesting that mRNAs with retained introns probably do not account for a large proportion of LHY transcripts.

4

Notes 1. Start the latest version of IGB by selecting a Web Start option at http://www.bioviz.org. If your computer system does not support downloading applications via Java Web Start, then download and unzip the file labeled igb.zip available from the IGB download page. Unpacking the igb.zip file will create a new folder named “igb” on your computer. Open the igb

134

Alyssa A. Gulledge et al.

Fig. 5 Counting reads to quantify intron retention. All tracks except the gene models (TAIR10 mRNA) and cold stress RNA-Seq read alignment track are hidden. The stack height for the RNA-Seq read alignment track is optimized to allow all reads to be shown, and the vertical zoom has been used to stretch the display vertically. Reads that appear to support intron retention are selected, and the number of selected reads is shown in the Selection Info box. Some of the selected reads are offscreen, above the current view

folder, and double-click a “.bat” IGB start script file (Windows) or “.command” start script (Mac) to start IGB. Note that the start scripts specify different amounts of memory that IGB will use when running, and the memory version you choose will vary depending on your computer. We recommend that if you have a Mac computer with 8 Gb or more of RAM, you can safely run a 5 Gb version. Windows users with less than 4 Gb of RAM should run the 1 Gb version.

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

135

2. In addition to loading data from the IGBQuickLoad.org site, you can also open data files stored on your local computer or Web sites using the File > Open File… or Open URL… menu options. In addition to using the File menu, you can also open a file by click-dragging it from your desktop or from a Web site into the IGB window. 3. IGB is unique among genome browsers in that it supports rapid, smooth, and flexible zooming. There are two types of zooming in IGB: animated zooming, in which the display appears to expand or contract in an animated fashion, and jump zooming, in which the display “jumps” to a new location. To zoom in IGB, use the horizontal or the vertical sliders (animated zooming) or click one of the zoom buttons next to the sliders (jump zooming). When you zoom in IGB, the zoom stripe, the vertical line that marks the location of your last mouse click within the main IGB window, remains stationary while the rest of the view expands or contracts around it; use the zoom stripe to focus zooming and also as a pointer when comparing boundaries of overlapping items. To jumpzoom to an item in the display, double-click it. To zoom to an entire gene model (versus a single exon within a gene model), double-click an intron (thin line) or double-click the gene model’s label. To jump-zoom to a region, click and drag over a region in the coordinate axis. You can also enter coordinates in the region box in the upper left of the display or search for a gene by name to jump-zoom to a new location in the genome. 4. RNA-Seq data sets are large, comprising many millions of sequence read alignments. For this reason, IGB does not automatically load an entire RNA-Seq data set directly into the viewer but instead waits until you request data by clicking the Load Data button. Note that the Data Management Table within the Data Access tab reports a Load Mode setting for each track, indicating how the data will be loaded into IGB. The mRNA track load mode is Genome, which means that the entire data set has been loaded into IGB, a reasonable setting for this data set, which contains 30,000 or fewer features. The load mode for the RNA-Seq data sets (cold and control), however, is set to Manual, meaning that data will only be loaded into IGB when you request it. This is because each RNA-Seq data set contains tens of millions of features, more data than IGB can display at once. Also note that when you click Load Data, only data for the currently visible region is loaded into IGB. 5. When IGB first loads reference gene model annotations or RNA-Seq reads, it reserves enough vertical space to display up to ten (the default) overlapping features in distinct rows within a track. If there are locations that contain more than ten items

136

Alyssa A. Gulledge et al.

that overlap along the sequence coordinate axis, the additional models will be indicated in the top row of the track, the “extra features” row. However, when viewing a gene or a region with fewer than ten gene models, you will observe some empty space above the models. You can eliminate this extra empty area by changing (or optimizing) the track’s stack height setting, which is the number of rows that can be shown per track in addition to the extra feature row at the top of the track. Also by default, IGB separates plus and minus strand gene models into separate tracks. However, when viewing a single gene, it often is better to combine the tracks for a more compact view by selecting the “+/−” combine strand setting. If you also select the “arrow” option, then gene models will be shown with arrowheads on the ends of the transcript and “greater” or “less” than symbols within introns to indicate the direction of transcription and strand of origin. 6. Edge matching is a visual analytics technique that was introduced by an earlier version of IGB called Neomorphic’s Annotation Station, which was used to annotate early assemblies of the Arabidopsis genome. Edge matching is designed to enable rapid comparisons between gene models and alternative splicing variants. To use edge matching, select a gene model or an exon within a gene model and observe that all other items with matching boundaries acquire an edge match highlight on compatible edges. However, the default edge match highlight color is white, which means that if you have changed your track background color to white, the edge matching highlight may be difficult to see. IGB allows you to change the edge match color by selecting File > Preferences > Other Options. Depending on your color scheme, red is usually a good choice for the edge match color setting. 7. The FindJunctions feature can be configured to consider single-mapping reads only or consider only reads with a threshold value of flanking bases on either side of the gap. The default is to use single-mapping reads that have at least five bases on either side of the gap. When junctions are created, the size of the blocks on either side of the gap is the same as the flanking threshold that was used in order to provide a visual cue indicating how the junction features were created. However, if you choose “TopHat Junctions,” the size of the flanking blocks will match the size of the largest flanking blocks present in the original gapped reads. 8. Right-click a gene model and select View Genomic Sequence in Sequence Viewer to open a new window that displays the sequence of the selected item. The Sequence Viewer has options to show spliced (cDNA) and unspliced (genomic) sequence and protein translations in different frames. You can

A Protocol for Visual Analysis of Alternative Splicing in RNA-Seq Data…

137

also copy and paste sequence from both the Sequence Viewer and the genomic sequence displayed below the coordinate axis. Also, zooming in on a gene model within the Main View reveals its amino acid translation. 9. You can set the Y-axis scale of a graph either by Value or by Percentage. Value is an absolute number; percentage is the percentage of all results currently loaded. If you load several genes, one of which has a very high peak, but choose to focus on a lower abundance region, IGB will still calculate the by Percentage based on all the loaded data, so your Y-axis may need to be reduced more than you expected to show the full height of the region you are viewing. Alternatively use the by Value option and set it to a usable height. References 1. English AC, Patel KS, Loraine AE (2010) BMC Plant Biol 10:102 2. Leviatan N, Alkan N, Leshkowitz D, Fluhr R (2013) PLoS One 8:e66511 3. Ner-Gaon H, Halachmi R, Savaldi-Goldstein S, Rubin E, Ophir R, Fluhr R (2004) Plant J 39:877–885 4. Iida K, Seki M, Sakurai T, Satou M, Akiyama K, Toyoda T, Konagaya A, Shinozaki K (2004) Nucleic Acids Res 32:5096–5103 5. Loraine AE, McCormick S, Estrada A, Patel K, Qin P (2013) Plant Physiol 162:1092–1109 6. Nicol JW, Helt GA, Blanchard SG Jr, Raja A, Loraine AE (2009) Bioinformatics 25:2730–2731

7. Perez-Santangelo S, Schlaen RG, Yanovsky MJ (2012) Brief Funct Genomics 12:13–24 8. Filichkin SA, Mockler TC (2012) Biol Direct 7:20 9. Filichkin SA, Priest HD, Givan SA, Shen R, Bryant DW, Fox SE, Wong WK, Mockler TC (2010) Genome Res 20:45–58 10. James AB, Syed NH, Bordage S, Marshall J, Nimmo GA, Jenkins GI, Herzyk P, Brown JW, Nimmo HG (2012) Plant Cell 24: 961–981 11. Trapnell C, Pachter L, Salzberg SL (2009) Bioinformatics 25:1105–1111

Chapter 9 AthaMap Web Tools for the Analysis of Transcriptional and Posttranscriptional Regulation of Gene Expression in Arabidopsis thaliana Reinhard Hehl and Lorenz Bülow Abstract The AthaMap database provides a map of verified and predicted transcription factor (TF) and small RNA-binding sites for the A. thaliana genome. The database can be used for bioinformatic predictions of putative regulatory sites. Several online web tools are available that address specific questions. Starting with the identification of transcription factor-binding sites (TFBS) in any gene of interest, colocalizing TFBS can be identified as well as common TFBS in a set of user-provided genes. Furthermore, genes can be identified that are potentially targeted by specific transcription factors or small inhibitory RNAs. This chapter provides detailed information on how each AthaMap web tool can be used online. Examples on how this database is used to address questions in circadian and diurnal regulation are given. Furthermore, complementary databases and databases that go beyond questions addressed with AthaMap are discussed. Key words Bioinformatics, Databases, Diurnal regulation, Gene expression, Transcriptional regulation, Posttranscriptional regulation, Web-server

1

Introduction The AthaMap database generates a genome-wide map of verified and predicted transcription factor-binding sites (TFBS) and small RNA target sites for A. thaliana. AthaMap was originally developed using matrix-based sequence searches with alignment matrices derived from several binding sites of the same TF [1]. Initially the database contained data for alignment matrices of 23 TFs from 13 different factor families that identified 2.4 × 106 predicted TFBS. Alignment matrices are derived by aligning a set of experimentally determined TFBS for one TF and were used to determine genomewide TFBS in the A. thaliana genome with the computer program Patser [1, 2]. Using TFBS for TFs that are known to interact with each other, the so-called colocalizing or combinatorial elements were also annotated [3]. Later, functionally verified single TFBS were annotated as well and were used to predict additional binding

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_9, © Springer Science+Business Media New York 2014

139

140

Reinhard Hehl and Lorenz Bülow

sites within the genome [4]. Currently, the database contains about 1 × 107 binding sites predicted with alignment matrices of 57 TFs from 21 TF families, 3.5 × 105 combinatorial elements, and 2.2 × 105 TFBS for 85 TFs from 21 TF families identified with TFs for which functionally verified single sites are known. The alignment matrices and the sequences used for genome-wide screenings, the name of the corresponding TF and its TF family, the corresponding publication, as well as the number of genome-wide matches detected with the screening sequences can be found on the Documentation page of AthaMap: http://www.athamap.de/ documentation.php under “Matrix-based screenings” and “Pattern-based screenings.” The alignment matrices and the sequences used for genome-wide screenings can be found by selecting the name of the TF. Currently, TFBS for a total number of 27 different TF families with 141 TFs are represented by AthaMap. One main difference between the TFBS detected with alignment matrices and with single sequences is the origin of the sequences. Matrix-based screenings were also done with TFs from other species than A. thaliana, while pattern-based screenings were done only with A. thaliana TFs [1, 4]. In recent years AthaMap data content was further enhanced with genome-wide target sites for small RNAs [5, 6]. For this, sequences from small RNA transcriptome datasets derived from different A. thaliana tissues such as seedlings, leaves, flowers, inflorescences, and siliques were employed [7–12]. Also, processed mature microRNA sequences from 243 microRNA genes obtained from the mirBase database were employed for genome-wide screenings [13]. The target sites for microRNAs were mapped genome-wide with the psRNATarget web server [14]. This analysis revealed that up to 16,600 Arabidopsis genes may be posttranscriptionally regulated by inhibitory RNAs [6]. The small RNA transcriptome datasets and the processed microRNAs used for genome-wide screenings can be found on the Documentation page of AthaMap: http://www.athamap.de/documentation.php under “Small RNAs” and “MicroRNAs.” With the data annotated to AthaMap, this database is a valuable resource to study transcriptional and posttranscriptional regulation of your genes of interest. The AthaMap database can be used for any questions related to gene expression regulation. Most recently it was used, for example, to study gene expression regulation in biosynthetic pathways, during plant development and in plant–pathogen interactions [15– 17]. Also questions related to diurnal/circadian regulation have been and can be addressed with the AthaMap database. Recent studies have identified novel components of diurnal/circadian transcriptional networks [18]. When promoters of cycling genes were analyzed, several cis-regulatory elements controlling time-ofday expression were identified, each containing distinct cisregulatory sequences [18, 19]: the morning cis-regulatory module

AthaMap Web Tools

141

consisting of the morning element (CCACAC) and the G-box (CACGTG), the evening cis-regulatory module consisting of the evening element (AAATATCT) and the GATA-box (GATA), and the midnight cis-regulatory module comprising telo-box (AAAA CCCT), starch synthesis box (AAGCCC), and protein box (ATGC CCC). Regulatory elements have also been discovered with the use of the AthaMap database. For example, during short-day photocycles, the expression of gibberellin (GA) receptor genes (GID) oscillates in seedlings, yielding a window of strong GA activity at the end of the night that overlaps with the period of maximum growth [20]. To investigate the link between the circadian clock and the transcriptional regulation of the GA receptors, an in silico analysis of GID1a, GID1b, and GID1c promoters with AthaMap indicated an enrichment of binding sites for DOF transcription factors in GID1a and GID1b. This is consistent with the identification of CYCLING DOF FACTOR 1 and 2 whose expression oscillates under diurnal conditions [20]. This analysis shows that the type of TFBS also indicates if members of a specific TF family can regulate the investigated genes, because members of the same TF family often have a similar DNA-binding specificity. Two other TFs that are already well established as circadian clock regulators are CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY) which are members of the MYB TF family [21]. Based on a single promoter study [22], a limited number of CCA1 and LHY TFBS were annotated to the AthaMap database. However, TFBS of MYB TFs constitute a large proportion of TFBS in AthaMap and may help to analyze target genes of MYB factors involved in circadian regulation, similar to the abovementioned study on the role of DOF TFs in oscillating GA receptors. Furthermore, in A. thaliana members of the TCP TF family are known to interact with components of the central oscillator such as LHY and CCA1 [20, 23]. TFBS for several TCP TFs [24, 25] were also annotated to AthaMap and can be used to identify target genes of TCP transcription factors using the AthaMap web tools. The following chapters will provide a step-bystep tutorial on how to use the AthaMap web tools. AthaMap can be accessed at http://www.athamap.de.

2

The AthaMap Web Tools Six different AthaMap web tools for the analysis of gene expression regulation are available online (Figs. 1 and 2). To analyze the regulation of specific genes, it will be important to learn which TFs and small RNAs regulate the gene(s) of interest. For this, AthaMap offers two tools: “Search” and “Gene Analysis” [1, 26]. To identify genes that harbor combinatorial TFBS, the “Colocalization” web tool was implemented [3]. For precalculated combinatorial

142

Reinhard Hehl and Lorenz Bülow

Fig. 1 AthaMap web tools. Partial screenshots of the AthaMap web tools: (a) “Search.” (b) “Colocalization Analysis.” (c) “Gene Analysis”

elements the “Gene Identifcation” web tool can be used [27]. To identify all possible target genes of a specific TF, one can also use the web tool “Gene Identification.” To identify putative posttranscriptionally regulated genes, the web tools “Small RNA targets” and “MicroRNA targets” are available [6].

AthaMap Web Tools

143

Fig. 2 AthaMap web tools. Partial screenshots of the AthaMap web tools: (a) “Gene Identification.” (b) “Small RNA Targets.” (c) “MicroRNA Targets”

144

Reinhard Hehl and Lorenz Bülow

2.1 The “Search” Web Tool

1. This web tool permits the display of predicted transcription factor and small RNA-binding sites in the A. thaliana genome. Two options for entering data are available on the search page at http://www.athamap.de/search.php (Fig. 1a). Most commonly, the Arabidopsis Gene Identifier (AGI) will be entered into the field AGI. Alternatively, a chromosomal position can be selected by first selecting the chromosome and then entering the position on the chromosome. Another parameter that can be chosen is “% restriction to highly conserved TF binding sites.” This only applies to TFBS that have been identified with alignment matrices (see Note 1). Initially, it is recommended not to restrict the search. 2. After a gene or the chromosomal position has been entered, the respective “Search” button should be activated. This results in the display of a sequence window in which the upper strand that corresponds to the TAIR annotation is displayed. This 1,000 bp window usually shows 500 bp upstream and 500 bp downstream of the chosen chromosomal position, the transcription start site (*TSS), or the start codon of the gene of interest. The untranslated regions (UTRs) as well as exon and intron structure of the gene are shown in a color code explained within the result window. Information about the gene is given below the window, and links to TAIR and MIPS are implemented [28, 29]. Potential TFBS are indicated as arrows above the sequence (see Note 2). 3. To get more information about the binding sites displayed on the result window, the mouse pointer can be moved over the arrow and information about the particular binding site is displayed in a mouse-over window. The individual position is given, and in the case of matrix-based binding sites, the score of that particular site, the threshold, and the maximum possible score are indicated as determined by the search program (see Note 1). By selecting the name of the TF or the small RNA, more information about screening sequence and corresponding literature is displayed in a pop-up window. To obtain a table with all predicted factor-binding sites, the web tool “Gene Analysis” can be used (see below).

2.2 The “Colocalization” Web Tool

1. To identify combinatorial TFBS, the Colocalization web tool at http://www.athamap.de/search_colocalization.php can be employed (Fig. 1b). First, two TFs (“Transcription-Factor 1 and 2”) should be chosen from a list that can be displayed by the selection button “Factor-Name.” To display all TFs, select “all” under “Family.” Here one can select one of the 141 TFs for which TFBS were predicted using either alignment matrices (“Matrix-Based Screenings”) or single sequences (“PatternBased Screenings”). Also the eight combinatorial elements and

AthaMap Web Tools

145

the two TFBS screenings that were done with bioinformatically established alignment matrices for the TATA-box (TBP) and the CCAAT-box (CBF) [30] can be found under “FactorName.” The different types of TFBS screenings used for selected TF are identified with symbols before the factor’s name (see Note 3). For TFBS identified with alignment matrices, the minimal threshold and the maximum score of an alignment matrix are displayed. The minimal threshold is the default value, but a higher value can be submitted into the field “Min. Threshold” (Fig. 1b). After this, the “Size of colocalization window” has to be determined. The difference between minimum spacer length and maximum spacer length must not be larger than 100. Entering 0 (“Min. Spacer length”) and 100 (“Max. Spacer length”) will determine the colocalization of the TFBS for the selected factors genome-wide in a 100 bp window. Before the search is started, one can select the way the results are sorted and if putative target genes of small RNAs should be excluded in the search (Fig. 1b). 2. After activating the “Search” button, the result is displayed on the same page. For this one needs to scroll down. The results are shown in a table that first identifies the chromosomal position of the identified TFBS and the respective chromosome. In addition, the orientation of the TFBS of the selected factors (“Factor 1” and “Factor 2”) is indicated by a “+” or a “−.” Also the spacer length between the two TFBS (“Spacer”), the nearest gene (“Gene”), and the position of the combinatorial element relative to the start of the gene (“Distance”) is listed. If genes putatively targeted by small RNAs are among the displayed genes, they are identified (see Note 4). 2.3 The “Gene Analysis” Web Tool

1. A different way to display the TFBS of a gene of interest is provided by the “Gene Analysis” web tool at http://www. athamap.de/search_gene.php (Fig. 1c). Here the gene identifier is submitted in a table (“Genes”). After this, the user can choose a specific region of the gene to be analyzed for TFBS (see Note 5). The “% restriction” applies again only to matrixbased TFBS (see Note 1). There are some options to sort the results and to exclude genes targeted by small RNAs. Finally, one needs to select specific TF families, small RNAs, and microRNAs to be displayed in the result table. By selecting “ALL,” all factors are included in the result. 2. After parameters have been selected and a “Search” is started, a result table shows up on the same page. In this table the position of TFBS or small RNA target sites, depending on the factors selected, are listed for the selected region. For TFBS, the corresponding TF, the TF family, the absolute chromosomal position, the orientation relative to the gene (“Relative orientation”),

146

Reinhard Hehl and Lorenz Bülow

and the position of the TFBS relative to the transcription or the translation start site are shown (“Relative distance”). Furthermore, for TFBS detected by matrix-based screening, the maximum score, the threshold score, and the score of the detected TFBS are shown. 3. The result table can now be downloaded into an excel table or any other word processing program simply by “copy” and “paste.” Additional online options allow the graphic display of the result (“Show Gene Analysis graphical display”) or to display a table that summarizes how many binding sites have been detected for each factor (“Show overview”). The third option “Show factors that are common in genes” applies to a “Gene Analysis” with more than one gene (see next paragraph). 4. The “Gene Analysis” web tool also permits the identification of common TFBS in a set of genes. This web tool is useful if, for example, coregulated genes are being studied. The web tool, data input, and parameter selection have been described above; the only difference is that a list of gene identifiers is being submitted into the online form. These should be separated by carriage returns. Not more than 200 genes can be submitted. Data display and content of the result tables have been described above. By selecting “Show factors that are common in genes,” one obtains a table that shows the binding sites detected in all genes or in a subset of the genes submitted for analysis. Also information about over- or underrepresentation of the binding site is provided. This can be obtained by comparing the “Sum of TFBS in total” with the “Theoretical Number of TFBS” (expected) that gives a “Ratio occurrence/ theoretical.” Numbers larger than 1 indicate an overrepresentation of a TFBS and numbers lower than 1 an underrepresentation compared to the expectation. 2.4 The “Gene Identification” Web Tool

1. To identify targets of specific TFs, this web tool can be accessed at http://www.athamap.de/gene_ident.php (Fig. 2a). First, one of the TFs should be chosen from a list that can be displayed by the selection button “Factor-Name.” To restrict the display to members of specific TF families, simply choose one of the 27 TF families listed under “Family.” Since the performance of the web tool has certain limitations, the number of TFBS had to be restricted to about 200,000 genome-wide TFBS for each factor. This applies only for TFBS identified with alignment matrices. For this, TFBS with a higher than minimum score were preselected by using an increased threshold score of each matrix to yield approximately 200,000 TFBS. This new score and the TFs for which this applies can be seen by selecting “Table of restriction scores” (Fig. 2a). The user can further increase this threshold by entering a number higher than

AthaMap Web Tools

147

the displayed “Min. Threshold” and lower than the maximum score. Also, the user needs to select a region for TFBS detection in the genes (see Note 5). Before the search is started, one can select the way the results are sorted and if putative target genes of small RNAs should be excluded in the search. 2. After starting the search, the results are shown in a table on the same page. The genes identified, the TF, and the TF family for which TFBS were identified (“Factor” and “Family”) are listed. Furthermore, the orientation of the TFBS relative to the gene and the position of the TFBS relative to the transcription or the translation start site are shown (“Relative Orientation” and “Relative distance”). In case TFBS is detected with alignment matrices, the maximum score, threshold score, and individual score of the identified TFBS are shown. The table can be downloaded for further use. 2.5 The “Small RNA Targets” Web Tool

1. To identify predicted small RNA target sites, the tool (Fig. 2b) available at http://www.athamap.de/smallRNA_targets.php provides the possibility to select from nine datasets (“Library”) representing small RNAs present in different tissues or developmental stages like seedlings, leaves, inflorescence, flowers, and siliques (see Note 6). The search parameters “upstream and downstream region” will define the query window relative to each gene start, either the TSS or the start codon of the coding sequence. For example upstream region = 0 and downstream region = 1,000 will search within the first 1,000 nucleotides of all transcripts. Prior to submitting such a query to AthaMap via the Search button, one can choose the results to be sorted by gene AGIs (“Gene”) or by relative distance to each gene start site. 2. Small RNA target sites and target genes are displayed in a table giving the corresponding AGI of the gene, the small RNA library, the absolute chromosomal position of the target site, the relative orientation, and the relative distance in respect to the corresponding gene. Gene AGIs and absolute positions can be selected in order to explore the genomic context of the gene or the target site within a newly opened sequence window. In order to display the small RNA target genes the total number of gene IDs detected can be selected. Small RNA target genes that are also targets of known microRNAs will be printed in bold within the result table. These will be displayed in an own table by selecting the number of predicted microRNA-regulated genes. For further data processing, tables can easily be exported to Excel when using Microsoft Internet Explorer or the table content can be copied to any spreadsheet software. In addition, the entire data for a selected small RNA dataset can be downloaded as a text file within a query window of −2,000– 4,000 relative to the gene starts.

148

Reinhard Hehl and Lorenz Bülow

2.6 The “MicroRNA Targets” Web Tool

1. Using the psRNATarget web server [14], microRNA target sites are predicted from a genome-wide screening using sequences from 190 different microRNAs (see Note 7). Target sites and genes can be identified by selecting a specific microRNA by family and name at http://www.athamap.de/ microRNA_ident.php (Fig. 2c). The query window relative to each gene start has to be defined as described above. By default, only trans targets are shown by suppressing the display of microRNA genes that encode the selected microRNA itself. A decreased psRNATarget score will increase the screening stringency and restrict the search to microRNA target sites displaying fewer possible mismatches. Both sense- and antisenseoriented microRNA target sites relative to the transcript will be displayed by default (“+/−”), but this can be changed to display target sites only in antisense (“−”) or in sense (“+”) orientation. Results can be sorted by gene AGIs (“Gene”), by relative distance to the TSS or start codon, or by psRNATarget score prior to submission of the query via the Search button. 2. The “results” table displays all microRNA target sites and genes matching the search criteria. Besides the target gene, microRNA name and family, the corresponding chromosomal position, orientation and distance of the target site relative to the target gene, and the individual psRNATarget score of the site are given. This tool also enables to explore the genomic context of a target gene or a target site within the sequence window by selecting a gene AGI or absolute position. A list with all unique target genes will be given when selecting the total number of gene IDs detected. In case a microRNA target gene is also being targeted by a small RNA, it will be shown in italics. These genes can also be displayed in a separate list of predicted small RNA-regulated genes.

3

Alternative and Complementary Databases and Web Tools In addition to AthaMap, there are many web-queryable resources for the analysis of gene expression regulation [31–34]. If possible, it is recommended to use more than one resource for database and web server-assisted analysis. First, each database has a different level of topicality and may also contain different data. For A. thaliana, many Internet resources are available. A selection of Internet resources discussed in this chapter is listed in Table 1. For example, A. thaliana-specific databases such as AGRIS, ATHENA, ATTED-II, PlantPAN, and AtPAN can be employed to predict cis-regulatory sequences [35–38]. Also sequences can be submitted to the PLACE, PlantCARE, and TRANSFAC databases to display either TFBS or cis-regulatory sequences that have been experimentally described before or for which no binding

AthaMap Web Tools

149

Table 1 Alphabetical list of Internet resources and their URLs Name

Link

aGFP

http://agfp.ueb.cas.cz

AGRIS

http://arabidopsis.med.ohio-state.edu/

Arabidopsis eFP

http://www.bar.utoronto.ca/

ARAMEMNON

http://aramemnon.botanik.uni-koeln.de/

ASRP

http://asrp.cgrb.oregonstate.edu/

AthaMap

http://www.athamap.de/

ATHENA

http://www.bioinformatics2.wsu.edu/cgi-bin/Athena/cgi/home.pl

AtProteome

http://fgcz-atproteome.unizh.ch/

AtPAN

http://AtPAN.itps.ncku.edu.tw/

ATTED-II

http://atted.jp/

BAR

http://bbc.botany.utoronto.ca

DIURNAL

http://diurnal.mocklerlab.org/

IAIC

http://www.arabidopsisinformatics.org/

miRBase

http://www.mirbase.org/

PatMatch

http://www.arabidopsis.org/cgi-bin/patmatch/nph-patmatch.pl

PLACE

http://www.dna.affrc.go.jp/PLACE/

PlantCARE

http://bioinformatics.psb.ugent.be/webtools/plantcare/html/

PlantPAN

http://plantpan.mbc.nctu.edu.tw/index.php

psRNATarget

http://plantgrn.noble.org/psRNATarget/

RSA tools

http://rsat.ulb.ac.be/rsat

STAMP

http://www.benoslab.pitt.edu/stamp/

SUBA

http://suba.plantenergy.uwa.edu.au/index.php

TAIR

http://arabidopsis.org

Target-align

http://www.leonxie.com/targetAlign.php

TRANSFAC

http://www.gene-regulation.com/

TF has been predicted [39–41]. Precalculated combinatorial elements were determined for A. thaliana with cis-sequences from the PLACE database [42]. Not all TFBS of predicted TFs are known. For example the Plant Transcription Factor Database (PlnTFDB) currently contains 2,657 protein models and 2,451 distinct protein sequences of A. thaliana, arranged in 81 gene families [43]. Currently, AthaMap contains TFBS for a total number of 27 different TFs or gene fami-

150

Reinhard Hehl and Lorenz Bülow

lies. Because AthaMap is actively curated, this number may be close to the number of gene families for which TFBS are known. Regulatory sequences for which no binding TF is known may be identified by detecting patterns of conserved sequence motifs in a set of coregulated genes [44]. These may represent regulatory sequences for the coregulation of these genes and may interact with TF for which binding sites have not been yet determined [45]. Such an approach has also been used to identify novel cis-regulatory sequences involved in diurnal regulation [18]. The Regulatory Analysis Tools (RSA tools) offer the possibility to submit a set of promoter sequences to identify common motifs [46, 47]. Similarly, Promomer at the Bio Array Resource (BAR) is a web tool to discover over-represented sequence motifs in regulatory regions from sets of A. thaliana genes [48]. In addition to online tools, also programs can be downloaded and can be implemented locally. This involves pattern mining programs such as MEME, AlignACE, CONSENSUS, Co-Bind, BioProspector, or MITRA [49–53]. To further refine the output of motif discovery programs, BioOptimizer [54] was integrated with four motiffinding programs within the Binding-site Estimation Suite of Tools (BEST) [55]. BEST includes four commonly used motif-finding programs, AlignACE, BioProspector, CONSENSUS, and MEME, and the optimization program BioOptimizer [55]. These programs detect sequence motifs that are conserved in a set of sequences. To find out if these motifs have been identified previously as regulatory sequences or TFBS, STAMP may be used to query databases of known motifs to compare these with motifs derived from coregulated genes [56]. STAMP contains motifs from many plant-specific databases such as PLACE, PlantCARE, AGRIS, and AthaMap. Therefore, it may be established if an identified motif has similarities to a known cis-regulatory sequence or TFBS. The positions of such putative cis-regulatory sequences within the A. thaliana genome can be determined with PatMatch, available at TAIR [57]. For identification of posttranscriptionally regulated genes targeted by microRNAs, Target-align or the psRNATarget web servers can be used. In this case, the small RNA sequence of interest needs to be submitted or selected and also the sequence to be analyzed for target sites needs to be provided [14, 58]. To learn which Arabidopsis genes are targets for small RNA-mediated mRNA degradation in vivo, the Arabidopsis Small RNA Project (ASRP) database can be employed [59]. With the application of next-generation sequencing, the analysis of gene expression is undergoing a revolutionary development [60]. We are not far from learning which genes are transcribed at a cellular level in time and space. The integration of such information in databases like Arabidopsis eFP Browser [61] and Arabidopsis

AthaMap Web Tools

151

Gene Family Profiler (aGFP) [62] will visualize the tissue- and state-specific expression of each gene during plant development. Ideally, not only transcriptome data will be integrated into databases but also proteome data [63]. Protein synthesis is under translational control, and, most importantly, proteins may move between cells [64, 65]. There are already several protein databases. For example SUBA, a SUBcellular location database for Arabidopsis proteins, comprises ten distinct subcellular locations and >6,743 non-redundant proteins and represents the proteins encoded in the transcripts responsible for 51 % of Arabidopsis expressed sequence tags [66]. Another example is ARAMEMNON, a database for Arabidopsis integral membrane proteins [67]. The AtProteome database comprises data of the high-density, organspecific proteome map for Arabidopsis thaliana [68]. A database of particular interest for scientists studying diurnal and circadian regulation is DIURNAL (Table 1). This database provides a graphical interface for mining and viewing diurnal and circadian microarray data for A. thaliana, poplar, and rice [69]. It is widely consulted to check for rhythmic transcript patterns, reflecting the growing awareness of the importance of temporal programs in gene expression [70]. A major challenge for the scientific community is to integrate the numerous yet disconnected databases into a dynamic and modular consortium of databases with a single point of access for users [71]. This integration is the goal of the International Arabidopsis Informatics Consortium (IAIC, Table 1) that will build the Arabidopsis Information Portal (AIP) which will facilitate the coordination and integration of data and scientists from around the world [71].

4

Notes 1. The program used for genome-wide screenings with matrices defines a maximum score and a threshold score for each predicted TFBS [2]. All TFBS that have a score between threshold and maximum score are annotated in AthaMap as putative TFBS. For example, if a matrix for a TF has a max. score of 10 and a threshold of 5 and the user defines the restriction to 20 % by entering a “20” in the text field “restriction,” only TFBS for that factor with a score of 6 or greater will be displayed since 5 + (5 × 0.2) = 6. If a “50” was entered in that text field, only TFBS with a score greater than or equal to 5 + (5 × 0.5) = 7.5 will be displayed in AthaMap. With this restriction it is possible to limit the results to TFBS that have a higher score and therefore a higher similarity to the matrix used for screening. 2. On the sequence display window, predicted target sites of TFs and small RNAs are shown above the sequence. The arrow indicates

152

Reinhard Hehl and Lorenz Bülow

the extension and orientation of the binding site. The arrow type depends on the type of binding site: Arrows (“----->”) symbolize matrix based, arrowheads (“>>>>>>”) symbolize pattern-based binding sites, and double lines (“======”) symbolize combinatorial elements. Small RNA and microRNA target sites are symbolized as “xxxxx>” and “XXXXX>,” respectively. 3. When selecting TFs in the “Colocalization” and “Gene Identification” web tools, the mode of TFBS detection is indicated by a symbol before the name of the TF. “−” means that TFBS for this factor were detected with matrix-based screenings (alignment matrices). “>” means that TFBS for this factor were detected with pattern-based screenings (single sequences). “=” means sites for a designated combinatorial element. 4. The gene lists usually obtained in result tables of several AthaMap web tools also identify genes that are putatively targeted by small RNAs. These genes are shown with a gene identifier in italics (small RNA), in bold (microRNA), or in italics and bold. These are usually listed when putatively posttranscriptionally regulated genes have not been omitted from the analysis by checking a box designated “exclude genes putatively regulated by small RNA” and/or “exclude genes putatively regulated by microRNAs” when using the respective web tool. 5. The upstream and downstream region selectable in the “Gene Analysis” web tool is restricted to −2,000 upstream and 3,000 downstream. Furthermore, when entering a list of genes (AGIs), the number is restricted to 200 gene identifiers. In the web tool “Gene Identification,” the upstream and downstream region is restricted to −2,000 upstream and 4,000 downstream. 6. Proveniences, references, and number of genomic hits of the nine small RNA datasets representing sequence signatures from different tissues or developmental stages, i.e., seedlings (s, s2, se3), leaves (l2, le3), inflorescence (i, i2), flowers (fl3), and siliques (si3), are given at AthaMap’s small RNA documentation site at http://www.athamap.de/documentation_ smallRNA.php. Lists with all predicted target genes can be downloaded as well. The functionality of other AthaMap tools that exclude or tag predicted small RNA-regulated genes within their respective analyses are based on these gene lists. 7. 190 distinct microRNA sequences collected by the miRBase database (http://www.mirbase.org/) were used for genomic screenings. AthaMap’s microRNA documentation site at http://www.athamap.de/documentation_microRNA.php displays the complete list of all 190 microRNAs used. Name, sequence and number of detected target sites are given. Text files harboring predicted microRNA target genes are provided for download. Other AthaMap tools use these gene lists to exclude or tag predicted microRNA-regulated genes.

AthaMap Web Tools

153

Acknowledgements The work on the AthaMap database was supported by grants from the German Federal Ministry for Education and Research (BMBF). We would like to thank Martin Schindler, Nils Ole Steffens, Claudia Galuschka, Yuri Brill, Julio C. Bolívar, Jonas Ruhe, and Stefan Engelmann for their work on the AthaMap database and web tools. Currently, AthaMap is being curated by Artyom Romanov. We also want to thank Edgar Wingender for stimulating our interest in bioinformatics. References 1. Steffens NO, Galuschka C, Schindler M, Bülow L, Hehl R (2004) AthaMap: an online resource for in silico transcription factor binding sites in the Arabidopsis thaliana genome. Nucleic Acids Res 32(1):D368–D372 2. Hertz GZ, Stormo GD (1999) Identifying DNA and protein patterns with statistically significant alignments of multiple sequences. Bioinformatics 15(7–8):563–577 3. Steffens NO, Galuschka C, Schindler M, Bülow L, Hehl R (2005) AthaMap web tools for database-assisted identification of combinatorial cis-regulatory elements and the display of highly conserved transcription factor binding sites in Arabidopsis thaliana. Nucleic Acids Res 33:W397–W402 4. Bülow L, Steffens NO, Galuschka C, Schindler M, Hehl R (2006) AthaMap: from in silico data to real transcription factor binding sites. Silico Biol 6(3):243–252 5. Bülow L, Engelmann S, Schindler M, Hehl R (2009) AthaMap, integrating transcriptional and post-transcriptional data. Nucleic Acids Res 37(Database issue):D983–D986 6. Bülow L, Bolívar JC, Ruhe J, Brill Y, Hehl R (2012) ‘MicroRNA Targets’, A new AthaMap web-tool for genome-wide identification of miRNA targets in Arabidopsis thaliana. BioData Min 5:7 7. Lu C, Tej SS, Luo S, Haudenschild CD, Meyers BC, Green PJ (2005) Elucidation of the small RNA component of the transcriptome. Science 309(5740):1567–1569 8. Axtell MJ, Jan C, Rajagopalan R, Bartel DP (2006) A two-hit trigger for siRNA biogenesis in plants. Cell 127(3):565–577 9. Rajagopalan R, Vaucheret H, Trejo J, Bartel DP (2006) A diverse and evolutionarily fluid set of microRNAs in Arabidopsis thaliana. Genes Dev 20(24):3407–3425 10. Kasschau KD, Fahlgren N, Chapman EJ, Sullivan CM, Cumbie JS, Givan SA, Carrington

11.

12.

13.

14. 15.

16.

17.

18.

JC (2007) Genome-wide profiling and analysis of Arabidopsis siRNAs. PLoS Biol 5(3):e57 Fahlgren N, Howell MD, Kasschau KD, Chapman EJ, Sullivan CM, Cumbie JS, Givan SA, Law TF, Grant SR, Dangl JL, Carrington JC (2007) High-throughput sequencing of Arabidopsis microRNAs: evidence for frequent birth and death of MIRNA genes. PLoS One 2(2):e219 Howell MD, Fahlgren N, Chapman EJ, Cumbie JS, Sullivan CM, Givan SA, Kasschau KD, Carrington JC (2007) Genome-wide analysis of the RNA-DEPENDENT RNA POLYMERASE6/DICER-LIKE4 pathway in Arabidopsis reveals dependency on miRNAand tasiRNA-directed targeting. Plant Cell 19(3):926–942 Kozomara A, Griffiths-Jones S (2011) miRBase: integrating microRNA annotation and deep-sequencing data. Nucleic Acids Res 39(Database issue):D152–D157 Dai X, Zhao PX (2011) psRNATarget: a plant small RNA target analysis server. Nucleic Acids Res 39(Suppl 2):W155–W159 Quadrana L, Almeida J, Otaiza SN, Duffy T, Correa da Silva JV, de Godoy F, Asis R, Bermudez L, Fernie AR, Carrari F, Rossi M (2013) Transcriptional regulation of tocopherol biosynthesis in tomato. Plant Mol Biol 81(3):309–325 Donner TJ, Scarpella E (2013) Transcriptional control of early vein expression of CYCA2; 1 and CYCA2;4 in Arabidopsis leaves. Mech Dev 130(1):14–24 Barah P, Winge P, Kusnierczyk A, Tran DH, Bones AM (2013) Molecular signatures in Arabidopsis thaliana in response to insect attack and bacterial infection. PLoS One 8(3):e58987 Michael TP, Mockler TC, Breton G, McEntee C, Byer A, Trout JD, Hazen SP, Shen R, Priest HD, Sullivan CM, Givan SA, Yanovsky M, Hong F, Kay SA, Chory J (2008) Network discovery

154

19. 20.

21. 22.

23.

24.

25.

26.

27.

28.

29.

Reinhard Hehl and Lorenz Bülow pipeline elucidates conserved time-of-dayspecific cis-regulatory modules. PLoS Genet 4(2):e14 Priest HD, Filichkin SA, Mockler TC (2009) cis-Regulatory elements in plant cell signaling. Curr Opin Plant Biol 12(5):643–649 Marin-de la Rosa N, Alabadi D, Blazquez MA, Arana MV (2011) Integrating circadian and gibberellin signaling in Arabidopsis: possible links between the circadian clock and the AtGID1 transcription. Plant Signal Behav 6(9): 1411–1413 Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22(16):R648–R657 Hazen SP, Schultz TF, Pruneda-Paz JL, Borevitz JO, Ecker JR, Kay SA (2005) LUX ARRHYTHMO encodes a Myb domain protein essential for circadian rhythms. Proc Natl Acad Sci U S A 102(29):10387–10392 Giraud E, Ng S, Carrie C, Duncan O, Low J, Lee CP, Van Aken O, Millar AH, Murcha M, Whelan J (2010) TCP transcription factors link the regulation of genes encoding mitochondrial proteins with the circadian clock in Arabidopsis thaliana. Plant Cell 22(12):3921–3934 Kosugi S, Ohashi Y (2002) DNA binding and dimerization specificity and potential targets for the TCP protein family. Plant J 30(3): 337–348 Viola IL, Reinheimer R, Ripoll R, Manassero NG, Gonzalez DH (2012) Determinants of the DNA binding specificity of class I and class II TCP transcription factors. J Biol Chem 287(1): 347–356 Galuschka C, Schindler M, Bülow L, Hehl R (2007) AthaMap web-tools for the analysis and identification of co-regulated genes. Nucleic Acids Res 35:D857–D862 Bülow L, Brill Y, Hehl R (2010) AthaMapassisted transcription factor target gene identification in Arabidopsis thaliana. Database (Oxford) 2010:baq034 Swarbreck D, Wilks C, Lamesch P, Berardini TZ, Garcia-Hernandez M, Foerster H, Li D, Meyer T, Muller R, Ploetz L, Radenbaugh A, Singh S, Swing V, Tissier C, Zhang P, Huala E (2008) The Arabidopsis Information Resource (TAIR): gene structure and function annotation. Nucleic Acids Res 36(Database issue):D1009–D1014 Mewes HW, Ruepp A, Theis F, Rattei T, Walter M, Frishman D, Suhre K, Spannagl M, Mayer KF, Stumpflen V, Antonov A (2010) MIPS: curated databases and comprehensive secondary data resources in 2010. Nucleic Acids Res 39(Database issue):D220–D224

30. Shahmuradov IA, Gammerman AJ, Hancock JM, Bramley PM, Solovyev VV (2003) PlantProm: a database of plant promoter sequences. Nucleic Acids Res 31(1):114–117 31. Hehl R, Bülow L (2008) Internet resources for gene expression analysis in Arabidopsis thaliana. Curr Genomics 9:375–380 32. Usadel B, Obayashi T, Mutwil M, Giorgi FM, Bassel GW, Tanimoto M, Chow A, Steinhauser D, Persson S, Provart NJ (2009) Co-expression tools for plant biology: opportunities for hypothesis generation and caveats. Plant Cell Environ 32(12):1633–1651 33. Brady SM, Provart NJ (2009) Web-queryable large-scale data sets for hypothesis generation in plant biology. Plant Cell 21(4):1034–1051 34. Mejia-Guerra MK, Pomeranz M, Morohashi K, Grotewold E (2012) From plant gene regulatory grids to network dynamics. Biochim Biophys Acta 1819(5):454–465 35. Davuluri RV, Sun H, Palaniswamy SK, Matthews N, Molina C, Kurtz M, Grotewold E (2003) AGRIS: Arabidopsis Gene Regulatory Information Server, an information resource of Arabidopsis cis-regulatory elements and transcription factors. BMC Bioinforma 4(1):25 36. O’Connor TR, Dyreson C, Wyrick JJ (2005) Athena: a resource for rapid visualization and systematic analysis of Arabidopsis promoter sequences. Bioinformatics 21(24):4411–4413 37. Obayashi T, Kinoshita K, Nakai K, Shibaoka M, Hayashi S, Saeki M, Shibata D, Saito K, Ohta H (2007) ATTED-II: a database of coexpressed genes and cis elements for identifying co-regulated gene groups in Arabidopsis. Nucleic Acids Res 35(Database issue): D863–D869 38. Chen Y-A, Wen Y-C, Chang W-C (2012) AtPAN: an integrated system for reconstructing transcriptional regulatory networks in Arabidopsis thaliana. BMC Genomics 13:85 39. Higo K, Ugawa Y, Iwamoto M, Korenaga T (1999) Plant cis-acting regulatory DNA elements (PLACE) database: 1999. Nucleic Acids Res 27(1):297–300 40. Rombauts S, Dehais P, Van Montagu M, Rouze P (1999) PlantCARE, a plant cis-acting regulatory element database. Nucleic Acids Res 27(1): 295–296 41. Matys V, Fricke E, Geffers R, Gossling E, Haubrock M, Hehl R, Hornischer K, Karas D, Kel AE, Kel-Margoulis OV, Kloos DU, Land S, Lewicki-Potapov B, Michael H, Münch R, Reuter I, Rotert S, Saxel H, Scheer M, Thiele S, Wingender E (2003) TRANSFAC: transcriptional regulation, from patterns to profiles. Nucleic Acids Res 31(1):374–378

AthaMap Web Tools 42. Ding J, Hu H, Li X (2012) Thousands of cisregulatory sequence combinations are shared by Arabidopsis and poplar. Plant Physiol 158(1):145–155 43. Perez-Rodriguez P, Riano-Pachon DM, Correa LG, Rensing SA, Kersten B, Mueller-Roeber B (2010) PlnTFDB: updated content and new features of the plant transcription factor database. Nucleic Acids Res 38(Database issue): D822–D827 44. Bailey TL (2008) Discovering sequence motifs. Methods Mol Biol 452:231–251 45. Koschmann J, Machens F, Becker M, Niemeyer J, Schulze J, Bülow L, Stahl DJ, Hehl R (2012) Integration of bioinformatics and synthetic promoters leads to the discovery of novel elicitor-responsive cis-regulatory sequences in Arabidopsis. Plant Physiol 160:178–191 46. van Helden J (2003) Regulatory sequence analysis tools. Nucleic Acids Res 31(13): 3593–3596 47. Thomas-Chollier M, Defrance M, MedinaRivera A, Sand O, Herrmann C, Thieffry D, van Helden J (2011) RSAT 2011: regulatory sequence analysis tools. Nucleic Acids Res 39(Web Server issue):W86–W91 48. Toufighi K, Brady SM, Austin R, Ly E, Provart NJ (2005) The Botany Array Resource: e-northerns, expression angling, and promoter analyses. Plant J 43(1):153–163 49. Bailey TL, Elkan C (1995) The value of prior knowledge in discovering motifs with MEME. Proc Int Conf Intell Syst Mol Biol 3:21–29 50. Roth FP, Hughes JD, Estep PW, Church GM (1998) Finding DNA regulatory motifs within unaligned noncoding sequences clustered by whole-genome mRNA quantitation. Nat Biotechnol 16(10):939–945 51. GuhaThakurta D, Stormo GD (2001) Identifying target sites for cooperatively binding factors. Bioinformatics 17(7):608–621 52. Liu X, Brutlag DL, Liu JS (2001) BioProspector: discovering conserved DNA motifs in upstream regulatory regions of co-expressed genes. Pac Symp Biocomput, pp 127–138 53. Eskin E, Pevzner PA (2002) Finding composite regulatory patterns in DNA sequences. Bioinformatics 18(Suppl 1):S354–S363 54. Jensen ST, Liu JS (2004) BioOptimizer: a Bayesian scoring function approach to motif discovery. Bioinformatics 20(10):1557–1564 55. Che D, Jensen S, Cai L, Liu JS (2005) BEST: binding-site estimation suite of tools. Bioinformatics 21(12):2909–2911 56. Mahony S, Benos PV (2007) STAMP: a web tool for exploring DNA-binding motif similarities. Nucleic Acids Res 35(Web Server issue): W253–W258

155

57. Yan T, Yoo D, Berardini TZ, Mueller LA, Weems DC, Weng S, Cherry JM, Rhee SY (2005) PatMatch: a program for finding patterns in peptide and nucleotide sequences. Nucleic Acids Res 33(Web Server issue): W262–W266 58. Xie F, Zhang B (2010) Target-align: a tool for plant microRNA target identification. Bioinformatics 26(23):3002–3003 59. Backman TW, Sullivan CM, Cumbie JS, Miller ZA, Chapman EJ, Fahlgren N, Givan SA, Carrington JC, Kasschau KD (2008) Update of ASRP: the Arabidopsis small RNA project database. Nucleic Acids Res 36(Database issue): D982–D985 60. Meyers BC, Vu TH, Tej SS, Ghazal H, Matvienko M, Agrawal V, Ning J, Haudenschild CD (2004) Analysis of the transcriptional complexity of Arabidopsis thaliana by massively parallel signature sequencing. Nat Biotechnol 22(8):1006–1011 61. Winter D, Vinegar B, Nahal H, Ammar R, Wilson GV, Provart NJ (2007) An “electronic fluorescent pictograph” browser for exploring and analyzing large-scale biological data sets. PLoS One 2(1):e718 62. Dupl’akova N, Renak D, Hovanec P, Honysova B, Twell D, Honys D (2007) Arabidopsis Gene Family Profiler (aGFP) – user-oriented transcriptomic database with easy-to-use graphic interface. BMC Plant Biol 7(39):39 63. Baginsky S, Gruissem W (2006) Arabidopsis thaliana proteomics: from proteome to genome. J Exp Bot 57(7):1485–1491 64. Kawaguchi R, Bailey-Serres J (2005) mRNA sequence features that contribute to translational regulation in Arabidopsis. Nucleic Acids Res 33(3):955–965 65. Dolan L (2001) Root patterning: SHORT ROOT on the move. Curr Biol 11(23): R983–R985 66. Heazlewood JL, Verboom RE, Tonti-Filippini J, Small I, Millar AH (2007) SUBA: the Arabidopsis subcellular database. Nucleic Acids Res 35(Database issue):D213–D218 67. Schwacke R, Schneider A, van der Graaff E, Fischer K, Catoni E, Desimone M, Frommer WB, Flugge UI, Kunze R (2003) ARAMEMNON, a novel database for Arabidopsis integral membrane proteins. Plant Physiol 131(1):16–26 68. Baerenfaller K, Grossmann J, Grobei MA, Hull R, Hirsch-Hoffmann M, Yalovsky S, Zimmermann P, Grossniklaus U, Gruissem W, Baginsky S (2008) Genome-scale proteomics reveals Arabidopsis thaliana gene models and proteome dynamics. Science 320(5878): 938–941

156

Reinhard Hehl and Lorenz Bülow

69. Mockler TC, Michael TP, Priest HD, Shen R, Sullivan CM, Givan SA, McEntee C, Kay SA, Chory J (2007) The DIURNAL project: DIURNAL and circadian expression profiling, model-based pattern matching, and promoter analysis. Cold Spring Harbor Symp Quantitat Biol 72:353–363

70. Staiger D, Shin J, Johansson M, Davis SJ (2013) The circadian clock goes genomic. Genome Biol 14(6):208 71. The International Arabidopsis Informatics Consortium (2012) Taking the next step: building an Arabidopsis information portal. Plant Cell 24(6):2248–2256

Chapter 10 Analysis of mRNA Translation States in Arabidopsis Over the Diurnal Cycle by Polysome Microarray Anamika Missra and Albrecht G. von Arnim Abstract Gene regulation at the level of translation occurs in response to environmental perturbation and is increasingly recognized as a factor affecting plant development. Despite extensive knowledge of transcriptional control, very little is known about translational regulation of genes in response to the daily light/ dark cycles. Here we describe the experimental layout designed to address how the translation states of genes change at various times during a diurnal cycle in Arabidopsis thaliana seedlings. We have adopted a strategy combining sucrose-gradient profiling of ribosomes and high-throughput microarray analysis of the ribosome-associated mRNA to investigate the translational landscape of the Arabidopsis genome. This is a powerful technique that can be easily extended to study translation regulation in different genetic backgrounds and under various environmental conditions. Key words Translational regulation, Polysome profile, Sucrose density gradient, Microarray, Quantitative PCR, RNA extraction

1

Introduction Nearly all organisms anticipate and adapt to the daily environmental changes with an endogenous timekeeper known as the circadian clock. The clock’s core is composed of a network of molecular oscillators that produce a self-sustained rhythm of approximately 24 h. However, the clock is entrained and reset by the daily periods of light and darkness, known as a diurnal cycle [1, 2]. Adaptability in sessile organisms such as plants is critically dependent on the environmental cues that maintain their circadian clock. In the model plant Arabidopsis thaliana, about one-third of the transcriptome responds to diurnal cycles, and DNA microarray analysis has been employed extensively to show that transcript levels of genes belonging to various metabolic pathways show rhythmic oscillations at different times of the day [3–7]. Yet, there is very limited data available on the translatome of Arabidopsis over the diurnal cycle; whether translation of genes also shows oscillating patterns

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_10, © Springer Science+Business Media New York 2014

157

158

Anamika Missra and Albrecht G. von Arnim

remains largely unknown. Interestingly, several studies point only to a weak correlation between transcript and protein levels in different organisms [8, 9]. Additionally, recent experiments on Arabidopsis rosette leaves suggest that protein levels are generally non-cycling, in spite of robust mRNA oscillations [10, 11]. Taken together, there is strong evidence that translational regulation in response to the diurnal cycle may help to modulate protein levels in Arabidopsis. Polysome profiling using sucrose density gradient centrifugation is a widely used technique to analyze the translational status of genes. Sucrose gradient centrifugation separates free mRNA (nontranslating) from polysome-associated mRNA (translating). It is known that highly translated mRNAs are associated with more ribosomes (polysomes) compared to poorly translated mRNAs. Hence, for a given gene, the ratio of polysomal to non-polysomal mRNA gives the translation state of the gene. A high ratio indicates active translation and vice versa. We took advantage of this polysome profiling technique and combined it with microarray analysis [12] to obtain the translational landscape of 10-day-old Arabidopsis seedlings over a diurnal cycle (24 h). This approach allows us to assign translation states to more than 22,000 Arabidopsis genes at different times of the day. Additionally, we have used quantitative PCR to validate the microarray data on a few genes.

2

Materials Prepare all solutions using autoclaved ultrapure water (prepared by purifying deionized water to attain a resistance of 18.3 MΩ at 25 °C; see Note 1). Store all solutions at room temperature unless otherwise indicated.

2.1 Arabidopsis Germination, Growth, and Harvesting

1. 100 % v/v ethanol. 2. Sterilization solution: 30 % bleach (commercial bleach like Clorox), 0.1 % Triton X 100.

2.1.1 Seed Sterilization 2.1.2 Seed Germination

Germination medium (GM, 1 L): 4.3 g Murashige and Skoog basal salt mixture, 10 g sucrose, 0.5 g 2-(N-morpholino) ethanesulfonic acid (MES), 900 mL water. Mix and adjust the pH to 5.7 with 1 N potassium hydroxide (KOH) solution, and add water to make 1 L of medium. Add 8 g (0.8 %) Bacto agar. Autoclave the medium in a 2 L conical flask for 20 min. Place in a 55 °C water bath to cool. Pour about 25 mL into petri plates (BD Falcon, 100 × 15 mm). Store plates at 4 °C.

Polysome Microarray Analysis of Arabidopsis Translation 2.1.3 Seedling Harvesting

1. Liquid nitrogen.

2.2 Sucrose Gradient Preparation

1. 1 M Tris–HCl, pH 8.4: Sterilize by autoclaving.

2.2.1 Stock Solutions

3. 1 M MgCl2: Sterilize by autoclaving.

159

2. Aluminum foil.

2. 3 M KCl: Sterilize by autoclaving. 4. 10 % deoxycholic acid (Sigma): Sterilize by autoclaving. 5. 1 M DTT in 10 mM sodium acetate, pH 5.2: Sterilize by filtering, and store at −20 °C. 6. 50 mg/mL cycloheximide in 100 % v/v ethanol: Store at 4 °C.

2.2.2 Gradient Preparation

1. 15 or 50 % sucrose solution: 15 or 50 % (w/v) sucrose, 200 mM Tris–HCl, pH 8.4, 50 mM KCl, 25 mM MgCl2, 0.1 mM DTT, 50 μg/mL cycloheximide. Prepare by using the stock solutions described in Subheading 2.2.1. Sterilize by autoclaving, and store at 4 °C until use. Add DTT and cycloheximide after autoclaving and immediately before use. 2. A commercial gradient maker, with 10 mL capacity in each chamber (we use CBS Scientific, Model # GM-20). 3. A small magnetic stir bar (2 × 5 mm) and a stir plate. 4. Beckman polyallomer centrifuge tubes (14 × 89 mm).

2.3 Sample Preparation and Gradient Centrifugation

1. Clean, autoclaved mortars (10 cm diameter) and pestles. 2. Polysome isolation buffer: 200 mM Tris–HCl, pH 8.4, 50 mM KCl, 1 % deoxycholic acid, 25 mM MgCl2, 2 % Polyoxyethylene 10 tridecyl ether (Sigma), 400 U/mL recombinant RNasin (RNase inhibitor, Promega), 50 μg/mL cycloheximide. Prepare using the sterilized stock solutions described in Subheading 2.2.1, and store at 4 °C until use, but not more than a month. Add RNasin and cycloheximide immediately before use. 3. Beckman L8-M ultracentrifuge and a Beckman SW41Ti swinging bucket rotor with six buckets.

2.4 Fraction Collection, RNA Extraction, and Gel Electrophoresis

1. Gradient fractionator (we use ISCO, model #640). 2. 60 % sucrose solution in water with 0.1 % (w/v) bromophenol blue added: Sterilize by autoclaving. 3. Phenol (pH 4.3):chloroform:isoamyl alcohol (25:24:1). 4. Isopropanol. 5. 10 % SDS. 6. 3 M sodium acetate, pH 5.2. 7. 75 % v/v ethanol.

160

Anamika Missra and Albrecht G. von Arnim

8. 50× Tris–acetate–EDTA (TAE) buffer (1 L): 242 g Tris–HCl, 100 mL of 0.5 M EDTA, pH 8.0, and 57.1 mL of glacial acetic acid. Mix, and make up the volume to 1 L with water. 9. Agarose. 10. Ethidium bromide (1 % solution). 11. 5× RNA loading dye: 12.5 % Ficoll 400, 0.25 % bromophenol blue. 2.5 RNA Purification and Quality Control

1. Qiagen RNeasy Plant Mini Kit with RNase-free DNase enzyme and buffer set. 2. Spectrophotometer (we use Nanodrop, Thermo Scientific) for measuring RNA concentration. 3. Bioanalyzer (Agilent) and RNA 6000 Nano Kit (Agilent) for determining RNA quality.

2.6 Candidate Transcript Analysis by qPCR

1. M-MLV Reverse transcriptase with buffer (Promega).

2.6.1 cDNA Synthesis

4. Oligo (dT)16.

2.6.2

1. 2× SsoFast™ EvaGreen® Supermix (BioRad).

qPCR

2. 20 mM dNTP (dATP, dCTP, dTTP, dGTP) mix. 3. Recombinant RNasin (RNase inhibitor, 400 U/mL, Promega).

2. PCR primers for the genes you intend to study. 2.7

Global Analysis

2.7.1 cDNA Synthesis

1. Nugen Applause 3′-Amp System with Encore Biotin Module. 2. Affymetrix Eukaryotic Poly A Control kit. 3. Qiagen min elute cleanup kit.

2.7.2 Microarray Hybridization

3

1. Affymetrix ATH1 microarray chips. 2. Affymetrix hybridization, wash, and stain kit.

Methods

3.1 Arabidopsis Germination, Growth, and Harvesting

1. Seed sterilization: Work in the laminar flow hood, and use sterile pipette tips and tubes. Transfer about 25 mg of dry wild-type Arabidopsis seeds (Columbia ecotype) into a microcentrifuge tube. These many seeds are sufficient for two plates containing about 500 seedlings each. Prepare as many such tubes as needed for the experiment (see Note 2). To sterilize the seeds, add 1 mL of 100 % v/v ethanol to the tube, resuspend the seeds by vortexing, spin down briefly, and remove the ethanol by aspiration. Then add 1 mL of the sterilization solution, vortex briefly to resuspend the seeds, and place the tube in the hood for 10 min. Spin the tube briefly, and remove

Polysome Microarray Analysis of Arabidopsis Translation

161

the solution by aspiration. Add 1 mL of water, resuspend the seeds by vortexing, spin down briefly, and remove the water by aspiration. Repeat the wash steps three more times. Finally resuspend the seeds in 1 mL of water. They are now ready to be plated. 2. Plating seeds: Work in the laminar flow hood. Cut off the very end of a 10–200 μL sterile pipette tip such that the opening is just enough for one or two seeds to pass. Resuspend the seeds in water, and pipette 100 μL of the suspension. Gently touch the end of the tip to the surface of the agar in a GM plate to release a single seed. Repeat this until all the seeds have been plated. Work in a circular or a spiral fashion starting from the center of the plate, and maintain even spacing of 3–4 mm between two seeds (see Note 3). Plate about 500 seeds on a single plate. Leave the lid of the plate slightly open until the water around each seed evaporates, and then seal the plate with parafilm. 3. Germination and growth of seedlings: Place plates with seeds at 4 °C in the dark for 48 h, and then transfer them to a growth chamber under white light (80 μmol/s/m2). Program the chamber to provide cycles of 16 h of continuous light, followed by 8 h of darkness (lights off), and set the temperature to 22 °C. Remove the parafilm seal before placing the plates in the chamber. The seeds should germinate within 1 or 2 days. Allow the seedlings to grow for 10 days. At this stage, the cotyledons are fully emerged and the primary leaves are about 2 mm long. 4. Harvesting seedlings: Harvest seedlings on the tenth day after placing them in the growth chamber. Collect seedlings from the first plate immediately before the lights turn on and then one plate every 6 h for 24 h. Snap-freeze the seedlings by pouring liquid nitrogen directly on the petri plate. When most of the liquid nitrogen has evaporated, scrape off the shoots with a spatula that has been precooled by dipping in liquid nitrogen. Allow the frozen shoots to fall into a small bath of liquid nitrogen made by draping a pre-labeled aluminum foil over a shallow container. Once the liquid nitrogen evaporates, quickly fold the aluminum foil to wrap the seedlings and immediately store in a −80 °C freezer. 3.2 Sucrose Gradient Preparation

1. The gradient maker is fitted with an outlet tubing with a 0.5–10 μL pipette tip attached to the end. Prepare each 10 mL gradient in a Beckman polyallomer centrifuge tube (14 × 89 mm). 2. Place the gradient maker on a stir plate, and add a magnetic stir bar to the chamber proximal to the outlet tubing (mixing

162

Anamika Missra and Albrecht G. von Arnim

chamber). Place a centrifuge tube near the gradient maker, and at a lower level with respect to it, such that the gradient can flow into the tube by gravity. Ensure that the valve controlling flow between the two chambers is in the closed position. 3. Clamp the outlet tube, and add 5 mL of 50 % sucrose to the mixing chamber. In the other chamber (distal to the outlet tube), add 5 mL of 15 % sucrose (see Note 4). Open the clamp on the tubing to allow a little bit (about 1 cm) of the 50 % sucrose to flow into the tubing, and then clamp the tubing again. 4. With the stir bar running at high speed, quickly open the valve between the chambers to allow mixing of the sucrose solutions and at the same time release the clamp from the tubing to allow the sucrose to flow into the centrifuge tube by gravity. Sometimes air bubbles stuck between the two chambers need to be removed by applying gentle pressure with a gloved finger on top of the chamber containing 15 % sucrose. 5. While the gradient flows into the tube, ensure that the tip at the end of the tubing touches the surface of the gradient to minimize mixing between layers. Tilt the gradient maker when the level of 15 % sucrose becomes low to allow all the solution to flow into the tube. Cover the tube with aluminum foil and place on ice till all gradients are prepared (see Note 5). Usually, we discard the first gradient and use the rest for centrifugation, so that the chambers and tubing get rinsed with the sucrose solutions. After making each gradient, rinse the mixing chamber and tubing with 50 % sucrose solution. 3.3 Sample Preparation and Gradient Centrifugation

1. Use clean and autoclaved mortars and pestles to grind snapfrozen seedlings. Precool the mortar and pestle by pouring some liquid nitrogen into the mortar. Quickly remove the foilwrapped frozen seedlings from the freezer, and transfer them to a Dewar flask containing liquid nitrogen till they are ready to be ground. For grinding, transfer seedlings from the aluminum foil to the mortar containing liquid nitrogen. Allow the liquid nitrogen to evaporate, and then start grinding with the pestle. Grind the tissue into a fine powder for at least 5 min (see Note 6). Cool the sample after every minute by adding liquid nitrogen and scraping the sides of the mortar with a precooled spatula to collect the powder at the bottom of the mortar. 2. Immediately after grinding, transfer a little bit of the ground tissue (about 50 mg; loosely packed to the 100 μL mark of a microcentrifuge tube) to a 2 mL microcentrifuge tube and immerse it in liquid nitrogen immediately. This will be used to prepare total RNA later, using the Qiagen RNeasy Plant Mini Kit (follow the manufacturer’s protocol for extraction of total

Polysome Microarray Analysis of Arabidopsis Translation

163

RNA from plant tissue). Transfer the remaining powder into an ice-cooled 5 mL plastic beaker, and immediately add 1 mL of polysome isolation buffer to the beaker. Swirl to mix until the powder dissolves in the buffer; it may take 2–3 min. Pipette the homogenate to a microfuge tube and immediately place on ice. It is acceptable for the tubes to sit on ice for the time it takes to prepare all the samples, which is usually within half an hour. 3. After all samples have been prepared, spin the microfuge tubes in a precooled fixed-angle bench-top centrifuge for 5 min at 12,000 rpm (10,000 × g) at 4 °C. Then carefully pipette up the supernatant (extract) from each tube, and gently layer it on top of the respective gradient. Balance pairs of centrifuge tubes as precisely as possible before centrifugation, and adjust the volumes if necessary by adding polysome isolation buffer. 4. Place the tubes into the buckets for Beckman SW41Ti swinging bucket rotor, and screw the lids on with a screwdriver. Gently hang the buckets on the precooled rotor. This rotor must be run with all six buckets in place, even if they are empty. Centrifuge the samples for 3.5 h at 35,000 rpm (210,000 × g) at 4 °C in a Beckman L8-M ultracentrifuge. Set the deceleration to zero to ensure that brakes are not applied to stop the rotor at the end of the run. It usually takes about 10 min after extraction to load the samples, balance the tubes, and start centrifugation. 3.4 Fraction Collection, RNA Extraction, and Gel Electrophoresis

1. Each gradient is divided into 12 fractions. While the centrifuge is running, prepare 2 mL microfuge tubes for fraction collection. Pipette 600 μL of phenol–chloroform–isoamyl alcohol into each collection tube. Then add 15 μL of 10 % SDS, 12 μL of 0.5 M EDTA, pH 8.0, and 3 μL of 1 M DTT to each tube and mix. With a marker, mark the 1.5 mL volume level on each tube. Place the tubes on ice until use. Also prepare 1.5 mL microfuge tubes for each fraction for RNA precipitation by adding 70 μL of 3 M sodium acetate, pH 5.2, and 700 μL isopropanol and place on ice until use. 2. The fractionator (ISCO, model #640, Fig. 1) consists of a tube holder (A), a UV (254 nm) absorbance flow cell (Type 6 optical unit, B), a pump (C), and a UA-5 absorbance monitor and chart recorder (D). Additional attachments include a tube piercer (E), a glass syringe (F) with a 2 mm diameter clear tubing attached (G), and a piston (H). Turn on the UV lamp 30 min before collecting fractions. Set the chart recorder controls as follows: speed at 60 cm/h, current at 340 mA, and sensitivity at 1 for concentrated polysome samples (more tissue) or 0.5 for dilute samples (less tissue, see Note 7).

164

Anamika Missra and Albrecht G. von Arnim

Fig. 1 Density gradient fractionator setup for collecting fractions. The components marked are A tube holder with a tube attached, B UV absorbance flow cell, C pump, D UA-5 absorbance monitor and chart recorder, E tube piercer, F glass syringe, G 2 mm diameter clear tubing, and H piston. The pump (C ) pushes the piston within the glass syringe, such that 60 % sucrose flows through the tubing (G ) and up the tube piercer (E ) into the tube containing the sucrose gradient. This causes the gradient to be pushed through the UV absorbance flow cell and tubing into collection tubes. The UV absorbance monitor and chart recorder (D ) continuously record the absorbance profile of the gradient

3. After centrifugation, carefully remove the centrifuge tubes from the buckets, cover the tops with small pieces of aluminum foil, and immediately place them on ice. After centrifugation the samples are divided into two distinct layers—a top green band containing chlorophyll (about half an inch) and a colorless layer that occupies the rest of the tube. 4. Fill the glass syringe with 30 mL of the blue 60 % sucrose solution, and prime the tubing using the piston, removing all air bubbles. Attach the tube piercer to the tube holder such that the centrifuge tube can fit snugly in between the two. Connect the tubing from the syringe to the tube piercer, and place the syringe firmly into the syringe holder attached to the pump. Start the pump so that the piston moves forward and pushes out any remaining air bubbles in the tubing and needle, and then stop the pump. Attach a tube containing the gradient to the tube holder. Pierce the bottom of the tube with the tube piercer, and start the pump to push 60 % sucrose solution into the tube. Set the flow rate to 0.75 mL/min, and turn on the chart recorder. Collect 900 μL of sample in the first collection

Polysome Microarray Analysis of Arabidopsis Translation 100

60 S

Absorbance 254 nm

80

Monosome

40 S

165

Polysomes

60 40

1

2

3

4

5 6 7 Fraction number

8

9

10

11

12

Fig. 2 An example of a typical polysome profile. The trace represents RNA absorbance at 254 nm. Free ribosomal subunit peaks and polysomal peaks are labeled

tube (till the 1.5 mL mark on the tube is reached), and then proceed to the next tube. While the second fraction is being collected, vortex the first tube for 5 s to mix the organic and aqueous layers. Place vortexed tubes on ice until all 12 fractions have been collected for a gradient. A typical absorbance profile (Fig. 2) will first show high absorbance (for the top chlorophyll-containing layer), then a trough, and then peaks for the 40S ribosomal subunit, 60S subunit, and 80S (monosomes), followed by a trough and then polysomal peaks. 5. After all fractions from a gradient have been collected, centrifuge the phenol extraction tubes using a bench-top centrifuge at room temperature for 5 min at maximum speed (≥14,000 × g, see Note 8). Meanwhile, repeat step 4 to prepare a new gradient for fractionation. During the time it takes for the first fraction of the new gradient to be ready for collection, transfer the upper aqueous phase (about 700 μL) from the phenolextracted tubes into new 1.5 mL microcentrifuge tubes containing sodium acetate and isopropanol. Mix the tubes well by inversion and incubate overnight at −20 °C. 6. To precipitate RNA, centrifuge the tubes in a precooled benchtop centrifuge for 20 min at maximum speed (≥14,000 × g) at 4 °C. Wash the pellets with 300 μL of chilled 75 % ethanol and centrifuge for 2 min at room temperature in a bench-top centrifuge at maximum speed (≥14,000 × g). Decant the supernatant, briefly spin the tubes, and remove the residual ethanol by pipetting. Allow the pellets to air-dry for not more than 5 min. Dissolve each pellet in 20 μL of RNase-free water (we use the water that is included in the Qiagen RNeasy plant mini kit)

166

Anamika Missra and Albrecht G. von Arnim

by gently flicking the sides of the tube with a finger. Store the RNA at −80 °C until further use or proceed directly to analyzing the RNA by gel electrophoresis, followed by DNase treatment and purification (Subheading 3.5). 7. Analyze the RNA from each fraction of a gradient on an agarose gel. Dissolve an appropriate weight of agarose to obtain a 1 % gel in 1× TAE buffer by heating in a microwave oven. Cool the agarose to approximately 50 °C, add 0.02 % ethidium bromide, mix well, and cast a gel. Mix 1 μL of the RNA with 7 μL of RNase-free water and 2 μL of 5× loading dye, and load and run the gel in an electrophoresis tank with 1× TAE buffer at 120 V till the bromophenol blue dye front almost reaches the bottom of the gel. Observe the bright 28S and 18S rRNA bands under UV light. The upper and lower edge of each band should appear equally sharp, indicating little or no degradation. The first two fractions do not contain any ribosomal subunits, so the 18S and 28S rRNA bands are absent. The third fraction contains the 40S ribosomal subunit; therefore, only the 18S rRNA band is observed. The fourth fraction contains a strong 18S rRNA band and a weaker 28S rRNA band since it is a mix of 40S and 60S subunits. From the fifth fraction onwards, the intensity of both the 18S and 28S rRNA bands become equal. Additionally, you may observe faint bands for chloroplast rRNA (23S, 16S, 4.5S, and 5S). 3.5 RNA Purification and Quality Control

1. For our experiments, we pool RNA from gradient fractions into the non-polysomal, small-polysomal, and large-polysomal samples. Of the 12 fractions, the first 4 fractions contain free mRNA, small RNAs such as tRNAs, and rRNA from 40S and 60S ribosomal subunits. These are pooled together to get the non-polysomal RNA (NP). The next four fractions contain RNA associated with monosomes (80S ribosomes), disomes, and trisomes. These are pooled together to get the small polysomal RNA (SP). The last four fractions contain RNA associated with polysomes (≥4 ribosomes), and they are pooled to get the large-polysomal RNA (LP). 2. DNase treat and column-purify the NP, SP, and LP RNA pools using the Qiagen RNeasy Plant Mini Kit with RNase-free DNase set, according to the instructions provided in the manual. In the final step, elute the RNA in 30 μL of RNase-free water, add the eluate back to the same column, and re-elute to concentrate the RNA. 3. Measure the concentration of RNA spectrophotometrically (we use a Nanodrop spectrophotometer, Thermo Scientific), and analyze the quality with the help of a Bioanalyzer (Agilent). The Bioanalyzer separates RNAs by size, and the software generates an electropherogram in which the shape

Polysome Microarray Analysis of Arabidopsis Translation

167

of the peaks and the level of background between the peaks yield evidence for the integrity of the abundant ribosomal RNAs. RNA with sharp peaks of 28S and 18S RNA, and low troughs in between, and with a 28S:18S ratio of ~2 is of good quality (i.e., has not undergone significant degradation). The software also generates an RNA Integrity Number (RIN) for each RNA sample, which classifies the RNA quality on a scale of 1 (most degraded) to 10 (undegraded). We use RNA having a RIN of 6.5 or higher for microarray analysis and quantitative PCR (see Note 9). 3.6 Candidate Transcript Analysis by qPCR 3.6.1 cDNA Synthesis

1. We prepare oligo(dT)-primed cDNA from 1 μg RNA for subsequent qPCR analyses. First, 0.5 μg of oligo d(T)16 is mixed with 1 μg of RNA and the volume is made up to 10 μL with RNase-free water. This mix is incubated at 70 °C for 10 min and then chilled on ice immediately for at least 2 min to anneal the primer. 2. Meanwhile, 10 μL of the reverse transcription (RT) mix is prepared by combining 4 μL of 5× M-MLV reverse transcriptase buffer (Promega), 1 μL RNasin (40 U/μL, Promega), 1 μL 20 mM dNTP, 1 μL M-MLV reverse transcriptase (200 U/μL, Promega), and 3 μL RNase-free water. 3. The oligo (dT)16-primed RNA is added to the RT mix to make a final volume of 20 μL, and the reaction is incubated at 41 °C for 50 min, followed by 70 °C for 15 min. The cDNA is diluted tenfold with water and stored at −20 °C until use in qPCR reactions.

3.6.2

qPCR

1. Primers are designed according to the target region near the 3′ end of the mRNA of interest, such that their Tm is ~60 °C and the amplified product size is within a range of 90–120 base pairs. The primers are quality-tested before use by setting up a standard curve with a dilution series of any template to confirm that the primers allow log-linear amplification at a particular Tm. 2. We perform all qPCR experiments using the BioRad iQ5 thermal cycler. Each reaction is carried out in triplicate to ensure technical consistency. Reactions are set up in optical film-sealed 96-well plates in volumes of 20 μL, with 5 μL of the tenfold diluted cDNA, 10 μL of 2× SsoFast™ EvaGreen® Supermix (BioRad), 1 μL each of gene-specific forward and reverse primers (10 μM stock concentration), and 3 μL of water (see Note 10). The PCR program consists of the steps outlined in Table 1. 3. Using the raw Ct numbers (cycle number at the threshold level of log-based fluorescence) we calculate the expression level of each gene, adjusted for the reference gene eEF1α (a subunit of

168

Anamika Missra and Albrecht G. von Arnim

Table 1 Cycling conditions for qPCR Cycling step

Temperature (°C)

Time

Number of cycles

Enzyme activation

95

1 min

Denaturation

95

10 s

Annealing/extension

60

20 s

Melt curve

65–95

20 s/step

1 45

1

Translation State (TL)

5

4

GI TL array

3

TL qPCR

2

1

Zt 0

Zt 6

Zt 12

Zt 18

Zt 24

Fig. 3 Comparison of TL states of the gene GIGANTEA (At1g22770) calculated from microarray and qPCR data. qPCR was carried out on seedlings harvested over five consecutive time points, but microarray was performed only with seedlings from the first four time points. An extrapolation of the TL graph to the fifth time point is shown with the dashed line for the array data. Error bars show standard deviations

the eukaryotic translation elongation factor 1). The transcript level of this gene is known to remain constant across the diurnal cycle, making it an ideal reference gene for studying oscillating transcripts of other genes. We have found EF1α to be a uniformly translated gene in our experiments; that is, the amount of mRNA in the NP, SP, or LP pools remains constant over all time points. This makes it a good reference gene for evaluating translation states also. From the adjusted expression level for each gene, the translation state (TL) is calculated exactly as described in Subheading 3.7. As a representative example, Fig. 3 compares the translation states of one gene, GI (At1g22770), derived from microarray and qPCR analysis. Both graphs show similar trends, so the qPCR results corroborate the microarray data.

Polysome Microarray Analysis of Arabidopsis Translation

3.7

Global Analysis

3.7.1 cDNA Synthesis

169

For microarray hybridization, we prepare biotin-labeled and fragmented cDNA using the Nugen Applause 3′-Amp System with Encore Biotin Module, according to the manufacturer’s protocol. Steps for hybridization, washing, and scanning of the microarray chips are carried out exactly as described in the Affymetrix manual. Only a brief outline is provided below. 1. cDNA is prepared from 200 ng of NP, SP, LP, and total RNA (see Note 11), and spike-in RNA (Affymetrix Eukaryotic Poly A Control kit) is added to each sample before the first-strand synthesis. The spike-in RNAs serve as internal standards to check for labeling efficiency as well as linearity of amplification and hybridization (see Note 12). The kit contains a mixture of four independent polyadenylated prokaryotic (B. subtilis) RNA from the genes lys, phe, thr, and dap at staggered concentrations. The stock RNA mix is diluted 125,000 fold (20 × 50 × 125) in the provided dilution buffer, and 1 μL of this diluted RNA is added to the primer-annealing step of the first-strand cDNA synthesis (oligo (dT) primed). 2. After the second-strand synthesis and Single Primer Isothermal Amplification® (SPIA), the reaction product is purified using Qiagen min elute cleanup kit according to the manufacturer’s protocol. The cDNA is then quantified using a Nanodrop, and 3.75 μg of cDNA is used for the subsequent fragmentation and biotin-labeling reactions.

3.7.2 Microarray Hybridization

The final volume after labeling is 50 μL, and the entire volume is used for hybridization to the Affymetrix ATH1 array. Array hybridization, wash, and scan protocols are carried out according to Affymetrix’s instructions.

3.7.3 Microarray Data Analysis

The Affymetrix software represents data from each array in “CEL” files, which contain the signal intensities from each probe on the chip. The CEL files from all arrays are first quality-checked using the “affy” package in the programming software R (Bioconductor). Several graphical tools are used to analyze the distribution of probe intensities across all arrays, for example, boxplots, histograms, and scatterplots. Ideally, these tools should indicate near-equal probe intensities across all arrays and also identify any outlier arrays that should not be included for further processing. The extent of RNA degradation is also assessed for each array using the affy package to ensure that there is no significant level of RNA degradation in any particular array. After the arrays clear the quality checks, we use the Bioconductor suite in the R statistical programming language to extract normalized values for intensities of each gene on the chip. Background adjustment is carried out using one of the two methods—gc

170

Anamika Missra and Albrecht G. von Arnim

Robust Multiarray Average (GCRMA) or Robust Multiarray Average (RMA). RMA uses the perfectly matched probes for background adjustment and does not use information from the mismatched probes. The mismatched probes sometimes detect specific signal, so background correction methods (e.g., MAS 5.0) that subtract mismatched probe intensities from perfectly matched probe intensities are likely to overadjust. In contrast the background intensity might be underestimated by RMA since it ignores the possibility of probes to undergo nonspecific binding. GCRMA uses the probe sequence information from the perfectly matched and mismatched probes to calculate probe affinities adjusted for nonspecific binding. Overall, we have found RMA and GCRMA to give very similar results with our datasets. The GCRMA and RMA functions in Bioconductor also include a quantile normalization step and a summarization step using the median polish algorithm. These procedures finally yield log-transformed expression values for each gene on the array. After the normalized expression values are obtained, the next step is to calculate the translation state (TL) of each gene. The expression values from RMA or GCRMA are first unlogged, and the SP and LP values are multiplied by the mean number of ribosomes in those fractions (2 for SP and 7 for LP), so that the expression values are weighted by the ribosome density on the mRNA. The sum of the weighted SP and LP values is then divided by the sum of expression values from NP, SP, and LP to get the TL value. Using the TL value for each gene at every time point, the genes are clustered based on the time of their peak TL states (Fig. 4). For clustering, we use a MATLAB-generated code that uses a threshold value to determine whether the TL states of a gene change significantly (cycles) over time or not. The threshold value is chosen such that it eliminates most of the noise generated by the standard deviation in the TL values over biological replicates. The average standard deviation was 0.2, and the threshold chosen was 0.3. The cluster analysis gives us 12 groups of genes that show the same patterns of translation across a diurnal cycle and a 13th group that does not cycle.

4

Notes 1. We do not use DEPC-treated water while making solutions, since the autoclaved ultrapure water in our laboratory has been tested to be free of RNases. However, it is advisable to use DEPC-treated or other sources of RNase-free water, if RNase contamination in water might be suspected. We also follow other routine procedures of working with RNA, like using RNase-free disposable plasticware (pipette tips and microcentrifuge tubes), wiping the working area, soaking glassware in

Polysome Microarray Analysis of Arabidopsis Translation

171

Fig. 4 Heatmap showing the 12 clusters of genes based on the time of their peak TL states. The genes in each cluster are sorted in the descending order of their peak TL state values. The number of genes in each cluster is shown in parentheses

RNase-removing solutions (e.g., RNase-AWAY, Molecular Bioproducts), and changing gloves frequently. The gradient maker with the attached tubing is cleaned with detergent, rinsed thoroughly with water, soaked in RNase-AWAY, and finally rinsed thoroughly in ultrapure water before each use. 2. It is better to sterilize seeds in small batches of 25 mg each in microcentrifuge tubes, instead of a single large batch. This ensures that the seeds are better exposed to the stertilization solution, and it is also convenient to spin down seeds for changing solutions and rinsing. 3. Since this is a light-dependent experiment, it is important to plate seeds evenly so that the seedlings do not shade each other. Growth conditions must be followed strictly and reproducibly. Harvesting of seedlings should be done in a timely manner. Each set of biological replicates is grown and harvested on different days. 4. It is important to maintain reproducibility between sucrose gradients. Make large stocks of 15 and 50 % sucrose to suffice

172

Anamika Missra and Albrecht G. von Arnim

for all the gradients and all biological replicates. While making the gradients, accurately measure the volumes of the solutions, so that the final volume of each gradient is the same. 5. After all gradients for one experiment are prepared, they are picked up at random and labeled. This randomization ensures that the gradients for each time point are not prepared in the same order for all the biological replicates. 6. Grind the seedlings thoroughly for at least 5 min. Although the appearance of the ground tissue will not change once it turns to a powder form, thorough grinding is essential for efficient extraction of polysomes. Gently add liquid nitrogen from a small beaker to the mortar at regular intervals during grinding to maintain the tissue in a frozen state. 7. Before starting the gradient collection, it is advisable to mockrun the fraction collector with a tube containing water to ensure that the pump and fraction collector are running smoothly. If using water, set the baseline of the chart recorder to about 1/10th of the chart length to ensure that during the actual gradient collection, all the polysomal peaks are recorded within the chart limits. The initial high absorbance peak (due to chlorophyll) usually exceeds the chart dimensions, but the peak height can be extrapolated from the chart. Sometimes, this may also be the case for the 60S and monosomal peak. Another way to make all the peaks fit inside the graph paper limits is to lower the sensitivity of the chart recorder, but this is not advisable since the polysomal peaks tend to get very small and it becomes difficult to analyze differences in polysome profiles between time points. 8. Fraction collection, if performed manually, as in our experiments, should be carried out precisely to ensure reproducibility between experiments. The phenol extraction step should be performed as soon as the fractions from a single gradient are collected; do not wait till all the gradients have been processed. This could lead to RNA degradation. Run as few gradients as your experiment allows, so that the wait time for the gradients during fractionation is reduced. Usually, for each experiment we run five gradients and it takes about half an hour to collect and process fractions from each gradient. So the last gradient has to wait on ice for approximately 2 h before it is fractionated. After RNA precipitation, ensure that all the ethanol has been removed from the tube. Do not use a pipette to resuspend the RNA pellet in water, gently flick the side of the tube about ten times, let the tube sit on ice for about 10 min, and then flick again. Spin the tubes briefly before storing the tubes at −80 °C or proceeding with the subsequent steps. 9. On a number of occasions, we have found that the nonpolysomal fraction (NP) is more prone to RNase degradation.

Polysome Microarray Analysis of Arabidopsis Translation

173

If the Bioanalyzer-predicted RIN for any sample is below 6, it is advisable not to use the RNA for microarray hybridization. Unfortunately, this means that the entire time series has to be repeated until all the RNA samples meet the required quality standard. 10. When setting up qPCR reactions, we make a mastermix for each primer pair, composed of water, primers, and the polymerase mix. The template is added individually to the wells in the plate, and then an appropriate volume of mastermix is mixed with the template by pipetting five times. We keep wells around the edges of the plate empty since samples in those wells are prone to evaporation. We also run duplicate sets of no-template controls for each primer pair to ensure that there is no spurious DNA in our working solutions. 11. While making cDNA for microarray hybridization, it is advisable to initially dilute each RNA sample to a working stock of 100 ng/μL. This eliminates the possibility of having to pipette up very small volumes of RNA ( 130 kDa > 100 kDa > 70 kDa > 55 kDa > 40 kDa >

Ponceau staining

α-GFP

Fig. 1 Co-IP of CBP20 and ABH1 and IP of RACK1A-GFP. Co-IPs were performed as described here. Protein extracts prepared from 10-day-old seedlings grown on ½ MS were separated on a 10 % SDS-PAGE and transferred to a nitrocellulose membrane. (a) Co-IP of CBP20 or ABH1, respectively, in WT and mutants (cbp20, abh1) using protein-specific antibodies. Proteins were detected by immunodetection using CBP20/ABH1 antibodies and a conformation-specific secondary antibody coupled to HRP. * Note that CBP20 is not detectable in abh1 mutants. (b) Co-IP of RACK1A-GFP using GFP-TRAP. Successful IP is shown by Ponceau S staining and immunodetection using a GFP-specific antibody

2

Materials

2.1 Coimmunoprecipitation

1. IP buffer: 50 mM Tris–HCl pH 7.5, 100 mM NaCl, 10 % (v/v) glycerin, 100 μM MG132, protease inhibitor tablets Complete (Roche) according to the instruction of the supplier (see Note 1). 2. IP wash buffer: 50 mM Tris–HCl pH 7.5, 100 mM NaCl, 10 % (v/v) glycerin, 0.05 % (v/v) Triton X-100 (see Note 1). 3. Protein A or protein G-coupled agarose beads (e.g., Roche) or GFP-TRAP coupled to agarose beads (Chromotek) (see Note 2). 4. Antibody for Co-IP (e.g., CBP20/ABH1, Agrisera).

178

Corinna Speth et al.

5. Syringe-driven filter unit (45 μm, non-sterile) (e.g., Millipore). 6. Rotating wheel. 7. Bench-top centrifuge with refrigeration. 8. 1.5 and 2 mL reaction tubes. 2.2

Western Blot

2.2.1 Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoreses (SDS-PAGE)

1. Casting device for polyacrylamide gels (e.g., BioRAD). 2. Electrophoresis system for protein analysis (e.g., BioRAD). 3. 30 % acrylamide: 30 % (w/v) acrylamide/0.8 % (w/v) bisacrylamide (37.5:1). 4. Resolving gel buffer: 1 M Tris–HCl, pH 8.8. 5. Stacking gel buffer: 1 M Tris–HCl pH 6.8. 6. 10 % (w/v) sodium dodecyl sulfate (SDS). 7. 10 % (w/v) ammonium persulfate (APS); store in aliquots at −20 °C. 8. N,N,N′,N′-Tetramethylethylenediamine (TEMED). 9. 10× SDS running buffer: 250 mM Tris base, 1.92 M glycine, 1 % (w/v) SDS. 10. 1× SDS running buffer: 100 mL 10× SDS running buffer and 900 mL (Milli-Q) water. 11. 5× loading dye: 250 mM Tris–HCl pH 7.5, 50 % (v/v) glycerin, 10 % (w/v) SDS, 1 mg/mL bromophenol blue. Always pre-warm the buffer to 50 °C to resolve the SDS. Before use add 500 mM DTT to the loading dye. 12. Nitrile gloves.

2.2.2 Protein Transfer

1. Tank-blotting system for protein transfer (e.g., GE Healthcare) (see Note 3). 2. Nitrocellulose membrane (0.45 μm pore size) (e.g., GE Healthcare) (see Note 4). 3. Transfer buffer: 150 mM glycine, 20 mM Tris base, 20 % (v/v) ethanol. 4. Ponceau S staining solution: 0.5 % (w/v) Ponceau S, 2 % (v/v) acetic acid. 5. Whatman paper (0.34 mm). 6. Forceps. 7. Plastic box filled with H2O for the assembly of the transfer cassette. 8. Gloves.

2.2.3 Immunodetection

1. 10× phosphate-buffered saline (PBS): 1.37 M NaCl, 81 mM Na2HPO4, 27 mM KCl, 14.7 mM KH2HPO4. 2. 1× PBS: 100 mL 10× PBS and 900 mL (Milli-Q) water.

Protein-Protein Interaction Analysis

179

3. PAGE wash buffer: 1×PBS + 0.1 % (v/v) Tween-20. 4. Blocking solution: Blocking solution on polymer basis (e.g., Roti-Block (ROTH)) or 5 % (w/v) milk powder in 1×PBS. 5. Primary antibody (used in Fig. 1: CBP20/ABH1 (Agrisera), GFP (Roche)). 6. Secondary antibody coupled to horseradish peroxidase (HRP) (in Fig. 1: HRP-coupled anti-mouse IgG (Agrisera), antirabbit conformation-specific IgG coupled to HRP: recognizes only native rabbit IgGs (Cell signaling)) (see Note 5). 7. Plastic container for membrane incubation. 8. Enhanced chemiluminescence (ECL) reagents (e.g., GE Healthcare). 9. CCD camera (e.g., Peqlab) (see Note 6).

3

Methods

3.1 Coimmunoprecipitation

In this chapter, we describe how proteins are selectively immunoprecipitated using protein-specific antibodies (CBP20/ABH1) or a GFP affinity resin (GPF-TRAP) which binds exclusively to GFP and GFP-tagged proteins (Fig. 1). One should consider proper negative controls. (1) For epitope-tagged transgenic lines, WT or a transgenic line expressing only the epitope can function as a negative control. (2) When using protein-specific antibodies, mutants lacking the protein or a WT sample treated without antibody should serve as negative control. 1. Harvest plant material, and grind it into a fine powder using liquid nitrogen, mortar, and pistil. 2. Transfer approximately 0.5 mg (1 mL) of ground plant tissue into a 2 mL reaction tube. It is important to prevent thawing of the powder. 3. Add 1.5–2 mL extraction buffer, and resuspend the plant material by vigorous vortexing. 4. Subsequently collect the plant debris by centrifugation (16,000 × g, 5 min, 4 °C). 5. Transfer the supernatant to a new 1.5 mL reaction tube, and collect the residual cell debris by a second centrifugation step (16,000 × g, 10 min, 4 °C). 6. Transfer supernatant to a new reaction tube. 7. In order to get rid of the remaining debris filter the plant extract through a syringe-driven filter unit. 8. Determine the protein concentration via standard Bradford method. This protein extraction method yields usually a protein concentration of 0.5–2 μg/μL dependent on the plant

180

Corinna Speth et al.

tissue used (protein amount: rosette leaves < seedlings < inflorescences). 9. Adjust the protein concentration of all sample to the same concentration using extraction buffer. 10. For Co-IPs, use 1–1.5 mL protein extract with a concentration of 0.5–1 μg/μL. Keep 40 μL of the protein extract as input and store on ice until the end of the experiment. 11. Pre-clear the lysate by adding 50 μL protein A agarose beads and incubate on a rotating wheel for 30 min at 4 °C. Preclearing is important to reduce background signals, which could be due to unspecific binding of proteins to protein A agarose beads (see Note 7). 12. Subsequently collect the beads by centrifugation (2,000 × g, 2 min, 4 °C). 13. Carefully transfer the supernatant to a new 1.5 mL reaction tube without disturbing the beads. 14. Afterwards add antibody to the protein extract (α-CBP20/ α-ABH1 dilution: 1:250). The amount of antibody used for Co-IP has to be determined for each antibody experimentally (dilutions usually used for Co-IPs: 1:50–1:1,000). When using the GFP-Trap skip steps 14–16 and incubate sample with 25 μL GFP-Trap for 3 h. Go on with step 17. 15. Incubate the samples for 1.5–2 h at 4 °C on a rotating wheel. 16. Afterwards add 50 μL protein A agarose beads to the sample, and incubate it for 1 h at 4 °C on a rotating wheel. 17. Collect the beads by centrifugation (2,000 × g, 2 min, 4 °C), and carefully remove the supernatant without disturbing the beads. 18. Subsequently add 1 mL Co-IP washing buffer, and incubate the sample for 5 min on a rotating wheel. 19. Repeat steps 16 and 17 twice (see Note 8). 20. Finally collect the beads by centrifugation (2,000 × g, 2 min, 4 °C), and carefully remove the supernatant. 21. Resuspend beads in 30 μL 2× Laemmli loading dye. 22. Add 10 μL 5× loading dye to the input. 23. Proceed to Western analysis or store the samples at −20 °C. 3.2

Western Blot

Western blot analyses are used to detected specific proteins in a given sample. First, proteins are separated depending on their molecular weight using SDS-PAGE (Subheading 3.2.1). Proteins are then transferred on a membrane (Subheading 3.2.2) followed by immunodetection using specific antibodies (Subheading 3.2.3).

Protein-Protein Interaction Analysis 3.2.1 Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoreses (SDS-PAGE)

181

Using denaturing conditions (SDS-PAGE) proteins are separated only by their molecular weight. The resolution of the polyacrylamide gel should be considered dependent on the expected size of the proteins to be detected (5 % gel: 60–200 kDa, 10 % gel: 16–70 kDa, 15 % gel: 12–45 kDa). Always wear nitrile gloves and goggles when handling acrylamide solutions because unpolymerized acrylamide is very toxic. 1. Prepare the casting stand and tray according to the manufacturer’s instructions. 2. Prepare the resolving gel by mixing the corresponding stock solution (listed in Table 1). It is important to add APS and TEMED at last as APS starts the polymerization reaction. Mix gently without producing air bubbles. The volume is calculated for one gel (each 8.6 × 6.7 × 0.15 cm2). Pour 7.5 mL of the resolving gel between the glass plates of the aperture. Immediately overlay the gel solution carefully with 0.1 % (w/v) SDS. The gel polymerizes within 30–45 min (see Note 9). 3. Discard the 0.1 % (w/v) SDS. 4. Prepare the stacking gel (see Table 1) as described for the resolving gel, and carefully pour it on top of the polymerized resolving gel. Immediately insert the comb. Wait for 30–45 min until the gel is completely polymerized (see Note 9). 5. Assemble the electrophoresis aperture, and add 1× SDS running buffer to the tank. 6. Remove the comb, and rinse the pocket with running buffer by pipetting up and down. Table 1 Pipetting scheme for polyacrylamide gels (table adapted from Dechert et al. [13]) Resolving gel

Stacking gel

Final concentration (%)

15

12.5

10

30 % Acrylamide (mL)

3.75

3.125 2.5

7.5

5

1.875 0.5

1 M Resolving gel buffer/ 2.813 stacking gel buffer (mL)

2.813 2.813 2.813 0.375

H2O (mL)

0.796

1.421 2.046 2.671 2.077

10 % SDS (μL)

75

75

75

75

30

10 % APS (μL)

60

60

60

60

15

TEMED (μL)

6

6

6

6

3

Total (mL)

7.5

7.5

7.5

7.5

3

182

Corinna Speth et al.

7. Denature proteins of input and immunoprecipitation (IP) (Subheading 3.1, step 21/22) by incubation at 80 °C for 10 min (see Note 10). 8. Load 20–50 μL input (15–30 μg total protein) and total IP (30 μL) on the gel. 9. Run the gel at 15 mA/per gel until the dye reaches the resolving gel. Afterwards set the currency to 25 mA/per gel, and run it until the dye reaches the bottom of the gel. 10. Use the gel immediately for protein transfer. 3.2.2

Protein Transfer

Try to handle the nitrocellulose membrane only with forceps and wear gloves. 1. Prepare the proper amount of transfer buffer, and fill the blotting tank. 2. Cut four pieces of Whatman paper and one piece of nitrocellulose which have the size of the gel. 3. Remove the gel from the electrophoresis aperture. Discard the stacking gel. 4. Equilibrate gel, Whatman paper, and membrane for 5 min in transfer buffer. 5. Assemble the protein transfer cassette in a plastic container filled with water as follows: two Whatman paper–membrane– gel–two Whatman paper. Remove the remaining air bubbles between the layers with a clean reaction tube, and place the cassette in the blotting tank. 6. Transfer proteins at 0.4–1.5 mA/cm2 for 2 h or overnight at room temperature (RT) or 4 °C. Transfer conditions have to be determined experimentally for each transfer device and protein of interest. Transfer conditions used for experiments shown in Fig. 1 are as follows: 1.5 mA/cm2, 2 h, and RT. 7. Subsequently stain the membrane with Ponceau S for 5 min to visualize highly abundant proteins on the membrane. 8. Remove residual staining by washing the membrane twice with H2O. 9. Staining might be captured by scanning. 10. Remove the Ponceau staining by washing the membrane twice with 1×PBS. Blocking and further downstream applications are not affected if some residual Ponceau staining is visible on the membrane. 11. The membrane can also be stored for 1–2 days in 1× PBS at 4 °C. Alternatively, after drying on a Whatman paper the membrane can be stored between two clean Whatman papers at 4 °C for a longer term.

Protein-Protein Interaction Analysis 3.2.3

Immunodetection

183

Try to handle the nitrocellulose membrane only with forceps and wear gloves. The optimal working dilution of the antibodies has to be determined experimentally. 1. Block the membrane using blocking solution on polymer basis (e.g., Roti-Block) or 5 % (w/v) milk powder in 1× PBS for 1 h at RT. 2. Incubate the membrane with primary antibody (CBP20/ ABH1: 1:1,000, GFP: 1:500) in 5 % (w/v) milk powder in 1× PBS for 1.5–3 h at RT. 3. Wash the membrane three times with 1× PBS supplemented with 0.1 % (v/v) Tween-20. 4. Incubate the membrane with secondary antibody (conformation-specific antibody: 1:2,000, HRP-mouse: 1:10,000) in 5 % (w/v) milk powder in 1× PBS for 1 h at RT. 5. Wash the membrane three times with 1×PBS supplemented with 0.1 % (v/v) Tween-20. 6. Detect the signal using a commercial ECL reagent kit and a CCD camera (see Note 6).

4

Notes 1. IP buffer and washing buffer might be modified by increasing the concentration of salt (NaCl; concentration can range from 0.1 to 1 M) or adding mild nonionic detergents (NP-40, Triton X-100; concentration can range from 0.05 to 2 % (v/v)). Increasing concentrations of detergents could reduce unspecific background (wash buffer) and might be necessary to solubilize some proteins (IP buffer). Increasing salt concentrations further impact the stringency of wash buffers and therefore can be used to reduce unspecific background. 2. Protein A and protein G have different binding affinities to the Fc and Fab regions of different IgG subclasses. Check which is the appropriate protein for your antibody. Protein A/G and GFP-Trap are also available coupled to magnetic beads. Magnetic beads are collected by a magnet instead of centrifugation. 3. It is also possible to use semidry blotting apertures for protein transfer. 4. It is also possible to use PVDF membranes for protein transfer. Usually PVDF membranes have a higher protein binding capacity compared to nitrocellulose membranes but produce higher background.

184

Corinna Speth et al.

5. Secondary antibodies can be coupled to alkaline phosphatases (APs), HRPs, and fluorescent dyes. Detection methods and sensitivity vary depending on the label and the substrates used for detection. For example BCIP/NBT, which is a substrate of APs and is converted into a colored precipitate, is less sensitive than detection systems producing light (e.g., when HRPs are visualized using ECL). 6. If a CCD camera is not available, it is also possible to detect chemiluminescent signals using X-ray films. 7. If the nonspecific protein binding is too high pre-clearing might be extended up to overnight. Check whether your protein of interest remains stable for that time. 8. Additional washing steps might be necessary to reduce unspecific protein binding. 9. Polymerization can be extended up to overnight to improve the polymerization reaction. When incubating gels overnight make sure to wrap them in a wet paper towel and store them in a plastic bag at 4 °C. 10. Denaturation temperature and time might be adjusted depending on the protein of interest.

Acknowledgements This work was supported by the DFG (LA2633-1/2) and the Max Planck Society (MPG)—Chemical Genomics Centre (CGC) through its supporting companies AstraZeneca, Bayer CropScience, Bayer Healthcare, Boehringer-Ingelheim, and Merck-Serono. References 1. McClung CR, Gutierrez RA (2010) Network news: prime time for systems biology of the plant circadian clock. Curr Opin Genet Dev 20(6): 588–598. doi:10.1016/j.gde.2010.08.010 2. Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22(16):R648–R657. doi:10.1016/j.cub.2012.07.025 3. Schoning JC, Staiger D (2005) At the pulse of time: protein interactions determine the pace of circadian clocks. FEBS Lett 579(15):3246– 3252. doi:10.1016/j.febslet.2005.03.028 4. Braun P, Aubourg S, Van Leene J, De Jaeger G, Lurin C (2013) Plant protein interactomes. Annu Rev Plant Biol 64:161–187. doi:10.1146/ annurev-arplant-050312-120140 5. Fields S, Song O (1989) A novel genetic system to detect protein-protein interactions.

6.

7.

8. 9.

Nature 340(6230):245–246. doi:10.1038/ 340245a0 Johnsson N, Varshavsky A (1994) Split ubiquitin as a sensor of protein interactions in vivo. Proc Natl Acad Sci U S A 91(22): 10340–10344 Kerppola TK (2008) Bimolecular fluorescence complementation (BiFC) analysis as a probe of protein interactions in living cells. Annu Rev Biophys 37:465–487. doi:10.1146/annurev. biophys.37.032807.125842 Gadella TW Jr, van der Krogt GN, Bisseling T (1999) GFP-based FRET microscopy in living plant cells. Trends Plant Sci 4(7):287–291 Isono E, Schwechheimer C (2010) Co-immunoprecipitation and protein blots. Methods Mol Biol 655:377–387. doi:10.1007/ 978-1-60761-765-5_25

Protein-Protein Interaction Analysis 10. Kuhn JM, Hugouvieux V, Schroeder JI (2008) mRNA cap binding proteins: effects on abscisic acid signal transduction, mRNA processing, and microarray analyses. Curr Top Microbiol Immunol 326:139–150 11. Adams DR, Ron D, Kiely PA (2011) RACK1, A multifaceted scaffolding protein: structure and function. Cell Commun Signal 9:22. doi:10.1186/1478-811X-9-22

185

12. Kocher T, Superti-Furga G (2007) Mass spectrometry-based functional proteomics: from molecular machines to protein networks. Nat Methods 4(10):807–815. doi:10.1038/ nmeth1093 13. Dechert U (1999) In Gelelektrophoresen. 2 (ed.), Vol. pp. 19–72, Spektrum Akademischer Verlag GmbH, Heidelberg

Chapter 12 Comparative Phosphoproteomics to Identify Targets of the Clock-Relevant Casein Kinase 1 in C. reinhardtii Flagella Jens Boesger, Volker Wagner, Wolfram Weisheit, and Maria Mittag Abstract In the green biflagellate alga Chlamydomonas reinhardtii different clock-relevant components have been identified that are involved in maintaining phase, period, and amplitude of circadian rhythms. It became evident that several of them are interconnected to flagellar function such as CASEIN KINASE1 (CK1). CK1 is involved in keeping the period. But it is also relevant for the formation of flagella, where it is physically located, and it controls the swimming velocity. In this chapter, we describe (1) how the flagellar subproteome is purified, (2) how phosphopeptides from this organelle are enriched, (3) how in vivo phosphorylation sites are determined, and (4) how direct and indirect flagellar targets of CK1 can be found using a specific inhibitor. Such a procedure can also be employed with other clock-relevant kinases if specific inhibitors or mutants are available. Key words Chlamydomonas reinhardtii, Circadian rhythms, Phosphoproteomics, Casein kinase 1, Flagella

1

Introduction In the last decades, the unicellular biflagellate green alga Chlamydomonas reinhardtii has developed as a model organism to study specific biological processes such as the composition and function of flagella, chloroplast biogenesis, nutrient, light and stress signaling pathways, or the mechanism of the circadian clock [1]. In C. reinhardtii several physiological processes are controlled by an endogenous clock (reviewed in [2] and [3]). Interestingly, some of them are closely related to flagellar function, including photoaccumulation (often described as phototaxis), chemotaxis, or stickiness to glass. Several molecular components of the circadian system of C. reinhardtii have been identified. The two subunits C1 and C3 of the RNA-binding protein CHLAMY1 are relevant for controlling circadian output, but they also influence phase (C3 subunit) and period (C1 subunit) of the circadian clock, indicating that they

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_12, © Springer Science+Business Media New York 2014

187

188

Jens Boesger et al.

are connected to the oscillatory system [4, 5]. Moreover, it was found that their expression is directly influenced by the Zeitgebers light and temperature [6, 7]. Another clock-relevant component of the circadian system of C. reinhardtii is CASEIN KINASE1 (CK1). Its silencing results in period shortening within a few days before arrhythmicity occurs [8]. An insertional mutagenesis approach revealed additional key players of the clock machinery of C. reinhardtii named RHYTHM OF CHLOROPLAST (ROC). They are relevant for maintaining period, phase, and amplitude of circadian rhythmicity of a chloroplast bioluminescence reporter [9, 10]. Also, a homologue of CONSTANS that controls the photoperiod in Arabidopsis thaliana was found in C. reinhardtii (CrCO) where it regulates processes that are under the control of the photoperiod or the circadian clock [11]. Notably, alterations in the expression of approximately 25 % of the ROCs as well as of CrCO and CK1 not only influence the circadian clock but also have in parallel severe effects on hatching (release of daughter cells from the mother cell wall during the cell cycle), flagella formation, and/or movement, underlining that these processes are interconnected in C. reinhardtii [8, 9, 11, 12]. To gain further insight into the clock network of C. reinhardtii, we investigated targets of CK1 in flagella, where this kinase is present and enriched compared to cell bodies [12–14]. For this purpose, we determined the flagellar phosphoproteome of wildtype cells [14] and compared it to the phosphoproteome of cells that had been treated with the specific CK1 inhibitor CKI-7 for 29 h [12] that was used before for studying CK1 in C. reinhardtii [15]. We did not use RNAi knockdown lines of CK1 for this approach because strong permanent reduction of the CK1 level results in shortening of flagella [8]. Phosphorylation is one of the most important posttranslational modifications and influences a wide range of cellular processes like protein function, its intracellular localization, its activity, and its affinity to interaction partners [16]. The detection of phosphorylation sites along with quantitative studies on the dynamics of phosphorylation events is important to understand cell signaling pathways. Therefore, one can use a combination of metabolic labeling and phosphopeptide enrichment via immobilized metal affinity chromatography (IMAC) [17, 18]. To find differences in the phosphorylation pattern of proteins that derive from the activity of specific kinases, another method of comparative phosphoproteome analysis can be used that is based on kinase mutants or on the application of kinase-specific inhibitors and the subsequent comparison of the phosphoproteomes. The latter method is described in this chapter. For the analysis, we took wild-type cells that were grown in a light–dark cycle and then transferred to minimal medium, where they were kept for 29 h in constant dim light before harvesting.

Comparative Phosphoproteomics to Find CK1 Targets

189

Thus, the proteome was determined from cells of the day phase. In parallel, another aliquot of the cells was treated the same way, but CKI-7 inhibitor was added when the cells were put into minimal medium and placed under dim light. The phosphoproteomes of both aliquots were determined and compared [12, 14]. At first, flagella were removed from the cell bodies of harvested cells by dibucaine treatment [14, 19]. The isolated flagella were pelleted by centrifugation, and flagellar proteins were extracted using detergent. The proteins were digested with trypsin and afterwards desalted using reversed-phase chromatography. For enrichment of phosphopeptides by IMAC, the tryptic peptides were applied to a column that was self-packed with a metal-chelating resin and charged with Ga3+ ions [20]. The bound phosphopeptides were eluted with a phosphate solution and desalted by using reversedphase chromatography. For identification of in vivo phosphorylation sites from flagellar phosphopeptides, we used tandem mass spectrometry (MS-MS) along with the acquisition of datadependent neutral loss scans (MS-MS-MS spectra). Data analysis was done using Proteome Discoverer 1.0 software including the SEQUEST algorithm [21] based on the available data from the flagellar proteome and the C. reinhardtii genome sequences [1, 13]. The spectra of the phosphopeptides along with their phosphorylation sites were in addition manually verified. The identified phosphopeptides from CKI-7-treated and non-treated C. reinhardtii cells were classified regarding their functions and compared concerning their presence or absence in the respective sample. In wild type, 32 phosphoproteins along with 126 phosphopeptides were found in the flagellar sub-proteome (see all nonlabeled and light grey-labeled proteins in Fig. 1 [14]). In the CKI-7-treated cells, several phosphoproteins were missing (see light grey-labeled proteins in Fig. 1) or were identified with a reduced number of phosphorylation sites (see proteins labeled with index a in Fig. 1), compared to untreated wild-type cells [12]. Interestingly, also novel phosphopeptides (see dark grey-labeled proteins in Fig. 1) or additional phosphorylation sites of known phosphopeptides (see proteins labeled with index b in Fig. 1) were identified in the CKI-7-treated cells. These findings were in agreement with immunodetections done with specific antibodies against phosphoSer in the CKI-7-treated and non-treated cells, respectively [12]. They clearly showed the disappearance of phosphoprotein bands as well as the appearance of novel bands after CKI-7 treatment. These data suggest that CK1 is part of a signaling network in the flagellum together with other kinases. One identified kinase is GLYCOGEN SYNTHASE KINASE3 (GSK3) that was only found in the CKI-7-treated cells as a phosphorylated protein. It was already shown that GSK3 is associated with the axoneme in a phosphorylation-dependent manner and the level of active GSK3 correlates with flagellar length [22]. GSK3 plays an important role

Fig. 1 Functional categorization of phosphoproteins of CKI-7-treated and non-treated cells [12, 14]. Phosphoproteins from wild type that were found in the flagellar sub-proteome are either non-labeled or labeled in light grey. Light grey labeling indicates at the same time that these phosphoproteins were missing

Comparative Phosphoproteomics to Find CK1 Targets

191

in the circadian system of other organisms such as Neurospora, Drosophila, or mammals [23–25]. It represents a candidate for further functional analysis in C. reinhardtii with regard to circadian control and flagellar function. The comparative phosphoproteome approach described in this chapter is also suited to investigate targets of other clock-relevant kinases if mutants or specific inhibitors are available. Thereby, the use of sub-proteomes, which have been successfully established in C. reinhardtii for several subcellular compartments, e.g., flagella, basal bodies, the eyespot, or the chloroplast (summarized in [26]), is of advantage due to the lower complexity of such data sets.

2

Materials Prepare and store all reagents at room temperature unless otherwise specified. All given percentages are % by volume (v/v) unless otherwise indicated.

2.1 Preparation of Cell Culture

1. C. reinhardtii strain 137c (CC-125 wild-type mt+, nit1, nit2) was used (for details see Chlamy Resource Center, University of Minnesota; http://chlamycollection.org/strains/). 2. Growth medium for C. reinhardtii: Tris-acetate-phosphate (TAP) [27]: Dilute 9.68 g Tris in about 3 L distilled water, and add consequently under stirring 4 mL of the phosphate buffer for TAP (a), 100 mL of the salt solution for TAP (b), 4 mL of the trace element solution (c), and about 4 mL acetic acid to adjust pH 7. Fill up the solution to 4 L with distilled water, transfer 2 L in a 5 L Erlenmeyer flask containing a magnetic stirring bar, and autoclave the flasks sealed with aluminum foil for 80 min at 120 °C. (a) Phosphate buffer: Add approximately 150 mL 1 M KH2PO4 to 250 mL 1 M K2HPO4 mL and adjust to pH 7. (b) Salt solution: 300 mM NH4Cl, 13.6 mM CaCl2, 4 mM MgSO4. (c) Trace element solution: These amounts are required for 1 L solution: 22 g ZnSO4 · 7 H2O, 11.4 g H3BO4, 5.06 g MnCl2 · 4 H2O, 5 g FeSO4 · 7 H2O, 1.61 g CoCl2 · 6 H2O, 1.57 g

Fig. 1 (continued) in CKI-7-treated cells. Novel phosphoproteins that were identified only in CKI-7-treated cells are dark grey labeled. FAP is the abbreviation of flagellar-associated protein. Accession numbers of all proteins can be found in Supplementary Tables 1 in [12, 14]. aPhosphoproteins were identified with a reduced number of phosphorylation sites in CKI-7-treated cells. bKnown phosphopeptides with additional phosphorylation sites in CKI-7-treated cells

192

Jens Boesger et al.

CuSO4 · 5 H2O, 1.1 g (NH4)6Mo7O24 · 4 H2O, 50 g Na2EDTA (see Note 1). 3. Minimal medium (MiMe) ([27], with modifications): Fill 1,798 mL of distilled water in a 5 L Erlenmeyer flask equipped with a magnetic stirring bar, and supplement it with 100 mL of the salt solution for MiMe (b) and 2 mL of the trace element solution (c). Seal the flasks with a piece of aluminum foil, and autoclave them for 80 min at 120 °C. In parallel, autoclave 100 mL of the phosphate buffer for MiMe (a) separately in a bottle. Combine 100 mL of the phosphate buffer for MiMe (a) with the solution in the 5 L Erlenmeyer flask directly before you resuspend the cells in this medium. (a) 1 M phosphate buffer: The following amounts are required for 1 L solution: 14.34 g K2HPO4, 7.26 g KH2PO4. (b) Salt solution: 150 mM NH4Cl, 13.6 mM CaCl2, 4 mM MgSO4. (c) Trace element solution (see above). 4. Casein kinase 1 inhibitor (CKI-7) stock solution: Dilute 50 mg CKI-7 dihydrochloride (Sigma-Aldrich Chemie GmbH) in 2.78 mL dimethyl sulfoxide (DMSO) immediately before use (see Note 2). 2.2 Preparation of Flagella

1. Wash buffer: 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH 7.5. Adjust pH with NaOH. 2. HMD buffer: 10 mM HEPES–NaOH pH 7.5, 5 mM MgSO4, 1 mM dithiothreitol (DTT). 3. HMD4%S buffer: 10 mM HEPES–NaOH pH 7.5, 5 mM MgSO4, 1 mM DTT, 4 % (w/v) sucrose. 4. HMD25%S buffer: 10 mM HEPES–NaOH pH 7.5, 5 mM MgSO4, 1 mM DTT, 25 % (w/v) sucrose. 5. 25 mM dibucaine hydrochloride (Sigma-Aldrich Chemie GmbH) solution in wash buffer. 6. 10 % (v/v) octylphenoxypolyethoxyethanol (IGEPAL CA-630 or Nonidet P-40, Sigma-Aldrich Chemie GmbH, see Note 3) in HMD buffer. 7. Complete protease inhibitor cocktail (PIC, Roche) (25× concentrated): Dissolve four pills of PIC in 8 mL HMD4%S buffer. 8. 100 mM ethylene glycol tetra acetic acid (EGTA) (see Note 4). 9. Phosphatase inhibitor cocktail 1 and 2 (Sigma-Aldrich Chemie GmbH, see Note 5).

Comparative Phosphoproteomics to Find CK1 Targets

2.3 Reduction, Carbamidomethylation, and Tryptic Digestion of the Flagellar Membrane–Matrix– Axoneme Fraction 2.4 Desalting of Peptides by Fast Protein Liquid Chromatography

193

1. Denaturation solution: 6 M guanidinium hydrochloride. 2. Carbamidomethylation solution: 0.5 M iodoacetamide. 3. Reduction solution: 1 M DTT. 4. Trypsin buffer: 100 mM NH4HCO3. 5. Trypsin stock solution: 500 ng trypsin/μL dissolving buffer, both provided by Promega. Acetonitrile and formic acid (Fluka) that were used in all the following experiments were of MS grade. 1. Reversed-phase column Source 15 RPC (GE Healthcare). 2. Fast protein liquid chromatography (FPLC) solution A: 2 % acetonitrile, 0.1 % formic acid. 3. FPLC solution B: 80 % acetonitrile, 0.1 % formic acid.

2.5 Enrichment of Phosphopeptides by IMAC

All solutions for IMAC, ZipTip®, and MS analysis are prepared with MS-grade water (Merck). 1. Nitrogen gas cylinder with a pressure-reducing regulator connected to a silicon tube. 2. Gel loader tips (10 μL; Eppendorf). 3. Glass fiber filter (Whatman). 4. POROS® 20 MC Metal Chelating Resin (66 % (w/w)) slurry in distilled water containing 0.02 % sodium azide (Applied Biosystems, Life Technologies Corporation). 5. 100 mM GaCl3 (Sigma-Aldrich Chemie GmbH) (see Note 6). 6. IMAC solution A: 0.1 % acetic acid. 7. IMAC solution B: 50 % acetonitrile, 0.1 % acetic acid. 8. IMAC solution C: 50 % acetonitrile, 0.1 % acetic acid, 100 mM NaCl. 9. IMAC elution buffer: 200 mM Na2HPO4.

2.6 Purification of Phosphopeptides by ZipTip®

1. Methanol (Roth), HPLC grade. 2. ZipTip® Pipette Tips (size P10; Millipore). 3. POROS® 10 R2 Reversed Phase resin (Applied Biosystems, Life Technologies Corporation). 4. ZipTip® equilibration solution: 60 % methanol. 5. ZipTip® wash solution: 5 % methanol, 5 % formic acid. 6. ZipTip® elution solution: 60 % methanol, 5 % formic acid.

2.7 Capillary Liquid ChromatographyElectrospray Ionization (LC-ESI)-MS Analysis

1. Reversed-phase column: Analytical Acclaim® PepMap C18 column (3 μm particle size and a pore size of 100 Å, Thermo Scientific, formerly Dionex).

194

Jens Boesger et al.

2. High pressure nano liquid chromatography, Ultimate 3000 HPLC (Thermo Scientific, formerly Dionex). 3. Linear ion trap mass spectrometer, Finnigan LTQ (Thermo Scientific, formerly Thermo Electron Corporation). 4. MS sample solution: 5 % DMSO, 5 % formic acid. 5. MS eluent A: 95:5 water–acetonitrile, 0.1 % formic acid. 6. MS eluent B: 20:80 water–acetonitrile, 0.1 % formic acid.

3

Methods Carry out all procedures at room temperature unless otherwise indicated.

3.1

Cell Culture

1. Inoculate 4 L TAP in two Erlenmeyer flasks, each with a 50 mL pre-culture of C. reinhardtii strain 137c. Cultivate the cells by constant stirring in a 12-h light/12-h dark cycle with a light intensity of 71 μE/m2/s (1 E = 1 mol of photons) at 24 °C. 2. Pellet the cells by centrifugation (1,100 × g, 5 min, 4 °C) when the culture has reached a cell density of 2 × 106–3 × 106 cells/ mL, and resuspend the pellet in half the volume of MiMe. Then, put the culture under constant conditions of dim light (15 μE/m2/s) for 29 h corresponding to subjective day. 3. CK1 inhibitor (CKI-7) treatment: Add 2 mL of the CK1 inhibitor solution to 2 L MiMe prior to the inoculation with the C. reinhardtii cells and the release into constant dim light.

3.2 Preparation of the Flagellar Membrane–Matrix– Axoneme Fraction

(According to Witman [19] and Boesger et al. [12, 14]) The procedure is explained based on one flagella purification using either CKI-7-treated or non-treated cells. 1. Harvest the cells by centrifugation at 700 × g and 4 °C for 15 min. Discard the supernatant, and keep the cell pellet on ice. 2. Resuspend the cell pellet in 50 mL wash buffer by gently shaking, then split the cells into two fractions, and put each in a 50 mL Falcon tube. Centrifuge the resuspended cells for 5 min at 1,300 × g and 4 °C. Discard the supernatants carefully, and repeat the washing step two times with 30 mL wash buffer. 3. Resuspend each pellet in 10 mL HMD4%S. Make sure to take a 100 μL sample for later (within the next 5 min) microscopic analysis and store it at RT. 4. For starting the deflagellation, add 2.5 mL of the 25 mM dibucaine hydrochloride solution per resuspended cell fraction and incubate them for 2 min by gently shaking on ice. Take 100 μL of the cells, which should be now deflagellated, and analyze the efficiency of deflagellation in comparison to

Comparative Phosphoproteomics to Find CK1 Targets

195

the sample without dibucaine treatment by microscopy (see Subheading 3.2, step 3). 5. If more than 95 % of the cells have lost their flagella, add 28 mL HMD4%S containing 100 μL 100 mM EGTA to each cell fraction to stop the deflagellation. 6. To separate the remaining cell bodies from the flagella, centrifuge the samples for 5 min at 1,300 × g and 4 °C and transfer the supernatants containing the flagella to new 50 mL Falcon tubes. Thereby, determine the volume of the supernatants and add PIC accordingly (finally 1× concentrated). 7. Pellet the flagella by centrifugation at 12,300 × g and 4 °C for 10 min, and discard the supernatants. Resuspend the pellet in a total volume of 10 mL HMD4%S buffer containing 400 μL PIC. Split the resuspended flagella suspension in two equal parts, and put them into two precooled 15 mL Falcon tubes. 8. Underlay each flagella suspension with 0.6 volume of the HMD25%S buffer using a long-tip pasteur pipette (see Note 7). 9. Centrifuge at 1,100 × g and 4 °C for 10 min (see Note 8). By this way, the remaining cell particles aggregate on the bottom of the 15 mL Falcon tube. Purified flagella are situated in the upper and inter phase of the sucrose cushion. 10. Transfer the purified flagella suspension from these phases into two new 15 mL Falcon tubes. At this point, the flagella fraction should be checked again for any remaining cell debris by microscopy. If the purity of the flagella fraction is not sufficient, repeat steps 8 and 9 (see Note 9). 11. Pellet the flagella suspension by centrifugation at 12,300 × g and 4 °C for 10 min, and discard the supernatant. 12. Resuspend the flagella pellets in a total volume of 450 μL HMD buffer containing 5 μL phosphatase inhibitor cocktail 1 and 5 μL phosphatase inhibitor cocktail 2. Transfer the flagella suspension in a new Eppendorf tube, add 50 μL 10 % (v/v) Nonidet P-40, and mix the solution gently to solubilize membrane proteins. Incubate the solution containing the membrane–matrix–axoneme (MMA) fraction for 10 min on ice. 13. Determine the protein concentration of the MMA fraction using the Neuhoff method [28]. 3.3 Reduction, Carbamidomethylation, and Tryptic Digestion of the MMA Fraction

Reduction and carbamidomethylation were carried out according to Zhang et al. [29] in order to cleave disulfide bridges and prevent their regeneration. 1. Proceed with 1 mg protein of the MMA fraction, and add 1 volume of 6 M guanidinium hydrochloride (final concentration 3 M) and one-hundredth volume of 1 M DTT (final concentration 10 mM). Incubate the solution for 1 h at 56 °C.

196

Jens Boesger et al.

2. Add one-tenth volume of 0.5 M iodoacetamide (final concentration 50 mM). Incubate the solution for 1 h at room temperature. 3. Add four volumes 100 mM NH4HCO3 and 20 μL of the trypsin solution to the sample, and incubate it for 16 h at 37 °C. 3.4 Desalting of Tryptic Peptides

1. Add acetonitrile to a final concentration of 2 % and formic acid to a final concentration of 0.1 % to 1 mg of trypsin-digested proteins to enhance binding of the peptides to the column resin. 2. Apply the sample to an FPLC system (GE Healthcare, formerly Amersham Biosciences) equipped with a reversed-phase column (SOURCE® 15 RPC) at a flow rate of 1 mL/min. 3. After sample application, wash the column with four column volumes of FPLC solution A. 4. Elute the peptides with 4 mL FPLC solution B and a fraction volume of 0.5 mL.

3.5 Enrichment of Phosphopeptides by IMAC

The IMAC procedure was performed according to Wagner et al. [30] using Ga3+-charged metal-chelating resin according to Shu et al. [20]. For IMAC, we used self-made micro columns prepared according to Erdjument-Bromage et al. [31]. 1. Cork out a small piece of a glass fiber filter (Whatman) with a 20 μL pipette tip, and push it down until it becomes stuck into the lower end of an Eppendorf gel-loader tip. 2. Ensure the correct placement of the filter piece by passing 20 μL 100 % ethanol through the column using 20 psi nitrogen pressure (see Note 10). 3. Pass two times 50 μL IMAC solution A through the gel-loader tip. 4. Pipette 50 μL of metal-chelating resin into the gel-loader tip, and equilibrate the column two times with 50 μL IMAC solution A. 5. Charge the column by applying 3× 50 μL of 100 mM GaCl3 and wash once with 50 μL IMAC solution A. 6. Load the eluted peptides of the desalting step (see Subheading 3.4, step 4) onto the activated IMAC column in 50 μL steps. 7. Wash the IMAC column with 50 μL IMAC solution A. 8. Wash the IMAC column with 50 μL IMAC solution B. 9. Wash the IMAC column with 50 μL IMAC solution C. 10. Wash the IMAC column with 50 μL IMAC solution A. 11. Elute the phosphopeptides using 3× 20 μL 200 mM Na2HPO4.

Comparative Phosphoproteomics to Find CK1 Targets

3.6 Desalting of Phosphopeptides by ZipTip®

197

(According to Wagner et al. [30]) 1. Make a hole in the bottom of a 0.5 mL Eppendorf PCR tube with a dissecting needle, and put a ZipTip® in front through this hole. Transfer this unit into a 1.5 mL Eppendorf tube. 2. Resuspend 10 μL of the POROS® 10 R2 reversed-phase resin in 50 μL equilibration solution, and transfer it into the ZipTip®. 3. Centrifuge the ZipTip® unit for approximately 15 s after reaching 1,180 × g at room temperature. Check if almost the entire liquid has passed through the column (see Note 11). 4. Transfer the ZipTip® unit in a new Eppendorf tube, add 50 μL ZipTip® equilibration solution, and centrifuge the ZipTip® unit like in step 3. Repeat this equilibration step once. 5. Transfer the ZipTip® unit in a new Eppendorf tube, add 50 μL ZipTip® wash solution, and centrifuge the ZipTip® unit like in step 3. Repeat this washing step once. 6. To improve binding to the reversed-phase column, add formic acid to the phosphopeptide sample to an end concentration of 5 % and then apply it to the ZipTip® unit. 7. Again, transfer the ZipTip® unit in a new Eppendorf tube, and centrifuge it like in step 3. 8. Transfer the ZipTip® unit in a new Eppendorf tube, wash the ZipTip® unit with 50 μL wash solution, and centrifuge like in step 3. Repeat this washing step once. 9. Transfer the ZipTip® unit in a new Eppendorf tube, and elute the phosphopeptides twice with 50 μL ZipTip® elution solution by centrifugation like in step 3. 10. Dry the eluate in a speed-vac for 2–3 h. You can freeze the dried sample at −80 °C. 11. For MS analysis, dissolve the dried peptides in 5 μL MS sample solution and transfer them in an HPLC sample tube.

3.7 Capillary Liquid ChromatographyElectrospray Ionization (LC-ESI)-MS Analysis

To separate the tryptic peptides prior to the MS measurements, a reversed-phase high pressure nano liquid chromatography (Ultimate 3000) is used. The eluting capillary is directly coupled to the nano-electrospray ionization (nESI) source that ionizes and transfers the peptides to the MS with its linear ion trap (Finnigan LTQ) for the measurement. For the analysis of phosphopeptides along with their in vivo phosphorylation sites, the routinely realized acquisition of one full MS and MS-MS scans for peptide identification was supplemented by an MS-MS-MS scan when a neutral loss of phosphoric acid was detected in the MS-MS scan. The following protocol is optimized for the stated system. 1. The MS and the online coupled HPLC are controlled by the Xcalibur 2.0 software developed by Thermo Scientific (formerly Thermo Electron Corp, San Jose, CA).

198

Jens Boesger et al.

2. To separate the peptide mixture on the analytical Acclaim® PepMap C18 column, use MS eluent A and MS eluent B. Set the HPLC micro pump to realize the following flow ramp: 0–5 min 4 % MS eluent B, 5–35 min increase MS eluent B to 50 %, 35–36 min up to 90 % MS eluent B, 36–46 min hold 90 % MS eluent B, 46–47 min decrease MS eluent B to 4 %, and 47–62 min hold 4 % MS eluent B (see Note 12). 3. The analytical column elutes to a fused silica needle attached to the nESI source (Thermo Scientific). Adjust the flow rate to 300 nL/min by a flowsplitter cassette with a splitting ratio of 939:1. 4. To ionize the eluting peptides at atmospheric pressure, a voltage difference of +1.7 kV is applied to the fused silica needle. The tip of the needle is positioned in 1–2 mm distance to the ion transfer capillary of the MS instrument to accomplish an efficient transfer from the charged vaporized peptide ions into the vacuum and towards the linear ion trap. 5. After the measurement of one initial full MS, the four most abundant ions are taken for a fragmentation step by collisioninduced dissociation using helium atoms and the resulting fragments are again analyzed (MS-MS). After each of these cycles, the peptide masses are excluded from the analysis for 10 s. An additional data-dependent fragmentation and measurement step (MS-MS-MS) is done when a mass difference of 98 Da (neutral loss of phosphate) in comparison to the precursor mass is detected in one of the MS-MS. 3.8 MS Data Analysis

For data interpretation, we use the Proteome Discoverer 1.0 (Thermo Scientific) comprising the SEQUEST algorithm [21] and the suitable databases (fasta protein sequences). These are the flagellar proteome [13] database (http://labs.umassmed.edu/ chlamyfp/index.php) and the Joint Genome Institute C. reinhardtii databases (Version 2 and Version 3; http://genome.jgi-psf. org/Chlre3/Chlre3.home.html) [1]. 1. As standard settings for the peptide identification, we allow two missed cleavages of trypsin and a static modification of cysteine (+57 Da) due to carbamidomethylation with iodoacetamide as well as a dynamic modification of +16 Da on methionine due to oxidation. To identify the phosphopeptides, the algorithm is adjusted to detect a dynamic modification on serine, threonine, and tyrosine of +79.97 Da in the MS-MS and MS-MS-MS spectra corresponding to the attached phosphate and additional −18 Da in the MS-MS-MS on serine and threonine due to the loss of the complete phosphate group during collision-induced dissociation. For the peptide identification, the peptide mass tolerance is set to 1.5 Da in MS mode. In MS-MS and MS-MS-MS modes, fragment ion tolerance is set up to 1 Da.

Comparative Phosphoproteomics to Find CK1 Targets

199

Only peptides with high Xcorr scores (see Note 13) of more than or equal to 2.0 for 1× positively charged, 2.5 for doubly charged, and 3.0 for triply charged peptides and a false discovery rate (FDR) of less than or equal to 1 % (see Note 14) are considered to be reliable results. Furthermore, the MS-MS and MS-MS-MS spectra of all identified phosphopeptides should be manually validated for the correct assignment of the b- and y-ion series with regard to the phosphorylation sites. 2. Repeat the analyses two to three times, take only identical peptides that are found in at least two independent experiments as trustworthy, and combine them to a final result file. Compare the result files of the protein samples from the CKI-7-treated and the non-treated C. reinhardtii cells manually. Analyze corresponding proteins also with the NCBI protein database using BLAST [32]. Examine flagellar associated proteins with no functional annotation additionally for the presence of conserved domains (Conserved domain BLAST). For positive identification of both protein and functional domain prediction, use an internal cutoff E value of 1e-05.

4

Notes 1. The trace element solution should be prepared as follows [27, 33]: Dissolve all salts except disodium EDTA in 75 mL distilled water. For the H3BO3 and the FeSO4 solutions it is necessary to heat them up. Add the disodium EDTA to 250 mL water, and heat the solution until the EDTA is dissolved. Mix the hot FeSO4 solution with the hot EDTA solution. Combine all other salt solutions in a 1 L beaker, heat them up to 70 °C, add the hot FeSO4–EDTA solution, and adjust the pH carefully to 6.5–6.8 using a 20 % KOH solution at this high temperature. Fill it up to 1 L, and let it cool down. It should be a clear green-colored solution. Stir the covered beaker at RT for several days until the solution shows a purple color, and then filtrate it through a sterile filter. The ready-to-use trace element solution should be stored in 1 mL aliquots at −20 °C. 2. CKI-7 is a highly toxic substance, and great care should be taken when handling it. Any waste of CKI-7 (e.g., the used TAP media) should be collected and disposed of as hazardous waste. 3. Do not confuse Nonidet P-40 with the NP40 type of Tergitol. IGEPAL CA-630 is offered from Sigma-Aldrich as a replacement for Nonidet P-40 and is suggested to be chemically indistinguishable. 4. EGTA is not easily soluble in water. Add 0.38 g EGTA to approximately 9 mL distilled water, and add 50–100 μL 3 N

200

Jens Boesger et al.

NaOH to dissolve the substance. Then fill up to 10 mL with distilled water. 5. Sigma-Aldrich provides phosphatase inhibitor cocktail 1 and 2 that cover the widest range of phosphatases and that were used in the present study. Recently, Sigma-Aldrich also offers a phosphatase inhibitor cocktail 3 that is a mixture of phosphatase inhibitors directed towards Ser/Thr phosphatases. This can be applied in addition to the cocktails 1 and 2. 6. GaCl3 is usually delivered in a sealed glass ampoule. Great care should be taken when opening it as GaCl3 is very hygroscopic and its reaction with water is very exothermic. To make the 100 mM solution, add 5 g of GaCl3 in small pieces to a beaker filled with 200 mL distilled water under constant stirring to prevent extensive heating of the solution. Then fill up to 284 mL with distilled water. 7. Use a battery-powered pipette filler with a long-tip pasteur pipette. Place the tip of the pipette on the bottom of the 15 mL Falcon tube, and underlay the flagella suspension with 0.6 volume of the HMD25%S by releasing the liquid out of the pipette tip at the lowest speed. Prevent mixing of the two solutions. 8. Centrifugation should be carried out without applying the brakes of the centrifuge so that the deceleration is very slow and does not influence the two phases of the sucrose cushions. 9. To examine the purity of the flagella fraction, you can further analyze proteins from the fraction by immunodetection using antibodies that are directed against proteins found in specific subcellular compartments [14]. This is an additional way besides microscopy to check for potential contaminants originating from other cellular compartments. 10. For this purpose, the N2 pressure-aided elution will be necessary during the whole procedure. 11. Please be aware to always keep a little amount of the liquid phase over the stationary phase until the last elution step. Otherwise the ZipTip® might dry out and will no longer bind the phosphopeptides. 12. Phosphopeptides have a high affinity to glass and other surfaces and bind to parts of an HPLC system, especially to the fused silica capillaries and the needle. To prevent a loss of phosphopeptides during analysis, we use a monophosphopeptide standard (Eurogentec) to block all free binding sites prior to the analyses of the samples. 13. The cross-correlation factor Xcorr of the SEQUEST analysis of the MS spectra describes the cross-correlation between the experimentally measured MS-MS spectrum and an in

Comparative Phosphoproteomics to Find CK1 Targets

201

silico-generated MS-MS spectrum of candidate peptides in the databases. 14. The FDR is a software-implemented filter criteria that analyzes the raw data against the reversed database and adjusts the Xcorr to reach a ratio of identified peptides with the reverse database (false positives) compared to the original one of less than or equal to 0.01.

Acknowledgments We appreciate the free delivery of information by the US (Department of Energy) and Japanese genome projects of C. reinhardtii very much. Our work was supported by grants of the Deutsche Forschungsgemeinschaft (DFG) and the Bundesministerium für Bildung und Forschung (BMBF). References 1. Merchant SS, Prochnik SE, Vallon O et al (2007) The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science 318:245–250 2. Schulze T, Prager K, Dathe H et al (2010) How the green alga Chlamydomonas reinhardtii keeps time. Protoplasma 244:3–14 3. Matsuo T, Ishiura M (2010) New insights into the circadian clock in Chlamydomonas. Int Rev Cell Mol Biol 280:281–314 4. Waltenberger H, Schneid C, Grosch JO et al (2001) Identification of target mRNAs for the clock-controlled RNA-binding protein Chlamy 1 from Chlamydomonas reinhardtii. Mol Genet Genomics 265:180–188 5. Iliev D, Voytsekh O, Schmidt EM et al (2006) A heteromeric RNA-binding protein is involved in maintaining acrophase and period of the circadian clock. Plant Physiol 142:797–806 6. Beel B, Prager K, Spexard M et al (2012) A flavin binding cryptochrome photoreceptor responds to both blue and red light in Chlamydomonas reinhardtii. Plant Cell 24: 2992–3008 7. Voytsekh O, Seitz SB, Iliev D et al (2008) Both subunits of the circadian RNA-binding protein CHLAMY1 can integrate temperature information. Plant Physiol 147:2179–2193 8. Schmidt M, Gessner G, Luff M et al (2006) Proteomic analysis of the eyespot of Chlamydomonas reinhardtii provides novel insights into its components and tactic movements. Plant Cell 18:1908–1930

9. Matsuo T, Okamoto K, Onai K et al (2008) A systematic forward genetic analysis identified components of the Chlamydomonas circadian system. Genes Dev 22:918–930 10. Matsuo T, Ishiura M (2011) Chlamydomonas reinhardtii as a new model system for studying the molecular basis of the circadian clock. FEBS Lett 585:1495–1502 11. Serrano G, Herrera-Palau R, Romero JM et al (2009) Chlamydomonas CONSTANS and the evolution of plant photoperiodic signaling. Curr Biol 19:359–368 12. Boesger J, Wagner V, Weisheit W et al (2012) Application of phosphoproteomics to find targets of casein kinase 1 in the flagellum of Chlamydomonas. Int J Plant Genom 2012:9 pp. doi:10.1155/2012/581460 13. Pazour GJ, Agrin N, Leszyk J et al (2005) Proteomic analysis of a eukaryotic cilium. J Cell Biol 170:103–113 14. Boesger J, Wagner V, Weisheit W et al (2009) Analysis of flagellar phosphoproteins from Chlamydomonas reinhardtii. Eukaryot Cell 8:922–932 15. Gokhale A, Wirshell M, Sale WS (2009) Regulation of dynein-driven microtubule sliding by the axonemal protein kinase CK1 in Chlamydomonas flagella. J Cell Biol 186: 817–824 16. Reinders J, Sickmann A (2007) Modificomics: posttranslational modifications beyond protein phosphorylation and glycosylation. Biomol Eng 24:169–177

202

Jens Boesger et al.

17. Hinsby AM, Olsen JV, Mann M (2004) Tyrosine phosphoproteomics of fibroblast growth factor signaling: a role for insulin receptor substrate-4. J Biol Chem 279: 46438–46447 18. Gruhler A, Olsen JV, Mohammed S et al (2005) Quantitative phosphoproteomics applied to the yeast pheromone signaling pathway. Mol Cell Proteomics 4:310–327 19. Witman GB (1986) Isolation of Chlamydomonas flagella and flagellar axonemes. Meth Enzym 134:280–290 20. Shu H, Chen S, Bi Q et al (2004) Identification of phosphoproteins and their phosphorylation sites in the WEHI-231 B lymphoma cell line. Mol Cell Proteomics 3:279–286 21. Link AJ, Eng J, Schieltz DM et al (1999) Direct analysis of protein complexes using mass spectrometry. Nat Biotechnol 17: 676–682 22. Wilson NF, Lefebvre PA (2004) Regulation of flagellar assembly by glycogen synthase kinase 3 in Chlamydomonas reinhardtii. Eukaryot Cell 3:1307–1319 23. Panda S, Hogenesch JB, Kay SB (2002) Circadian rhythms from flies to human. Nature 417:329–335 24. Tataroğlu Ö, Lauinger L, Sancar G et al (2012) Glycogen synthase kinase is a regulator of the circadian clock of Neurospora crassa. J Biol Chem 287:36936–36943 25. Spengler ML, Kuropatwinski KK, Schumer M et al (2009) A serine cluster mediates BMAL1dependent CLOCK phosphorylation and degradation. Cell Cycle 8:4138–4146

26. Wagner V, Boesger J, Mittag M (2009) Subproteome analysis in the green flagellate alga Chlamydomonas reinhardtii. J Basic Microbiol 49:32–41 27. Harris EH (2009) The Chlamydomonas sourcebook, vol 3. Academic, San Diego, CA 28. Neuhoff V, Philipp K, Zimmer HG et al (1979) A simple, versatile, sensitive and volumeindependent method for quantitative protein determination which is independent of other external influences. Hoppe Seylers Z Physiol Chem 360:1657–1670 29. Zhang Y, Wolf-Yadlin A, Ross PL et al (2005) Time-resolved mass spectrometry of tyrosine phosphorylation sites in the epidermal growth factor receptor signaling network reveals dynamic modules. Mol Cell Proteomics 4:1240–1250 30. Wagner V, Geßner G, Heiland I et al (2006) Analysis of the phosphoproteome of Chlamydomonas reinhardtii provides new insights into various cellular pathways. Eukaryot Cell 5:457–468 31. Erdjument-Bromage H, Lui M, Lacomis L et al (1998) Examination of micro-tip reversed phase liquid chromatographic extraction of peptide pools for mass spectrometric analysis. J Chromatogr A 826:167–181 32. Altschul SF, Madden TL, Schäffer AA et al (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25:3389–3402 33. Wagner V, Mittag M (2009) Probing circadian rhythms in Chlamydomonas reinhardtii by functional proteomics. Methods Mol Biol 479: 173–188

Chapter 13 Pulsed Induction of Circadian Clock Genes in Arabidopsis Seedlings Stephen M. Knowles, Sheen X. Lu, and Elaine M. Tobin Abstract The Alc-inducible system is a simple, yet effective, “gene switch” that can be used to transiently induce gene expression in Arabidopsis. Here we provide a protocol for using the Alc-inducible system to give a pulse in expression of a circadian clock gene in transgenic seedlings. The line we use as an example harbors an Alc-inducible copy of the CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) gene (Alc∷CCA1). Alc∷CCA1 seedlings are grown on solid MS medium and subsequently treated with ethanol vapor. Because the ethanol is quickly absorbed into the medium upon exposure, the seedlings are moved to fresh plates following treatment to avoid continuous induction. After the induction, the seedlings are harvested over a time-course for future total RNA and/or protein extraction that can be used for subsequent gene expression analyses. Key words Ethanol switch, Alcohol-inducible system, Circadian, Pulse, AlcR, Induction, Clock gene, Transient, Arabidopsis thaliana

1

Introduction The Alc-inducible system, also known as the ethanol (EtOH)-switch, allows for rapid and transient expression of transgenes in plants [1–3]. The system is relatively simple in design as it works with just two components: a regulator and an effector (Fig. 1). The transcription factor AlcR functions as the regulator. It is constitutively expressed in virtually all plant tissues by the constitutive 35S Cauliflower Mosaic Virus promoter. In the presence of EtOH, AlcR binds to the inducible AlcA promoter and promotes the expression of the downstream gene, i.e., the effector. This promotion is reversible—once the EtOH is removed, AlcR no longer promotes the expression of the effector gene. The induction can be performed in a number of different ways, leading to the versatility of the system. EtOH can be applied by spraying or root-drenching with a dilute solution of EtOH in water or by vapor [3–5]. Induction via EtOH vapor is key to its implementation in circadian

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_13, © Springer Science+Business Media New York 2014

203

204

Stephen M. Knowles et al.

Regulator

35S

AlcR

AlcR

EtOH

EtOH

AlcR

EtOH AlcR Effector

pAlcA

GENE

Fig. 1 Model for the Alc-inducible system

research, as the vapor source can be quickly introduced to, and subsequently removed from, the plants, thereby generating a “pulse” of effector gene expression. In our lab such pulse experiments have helped to confirm the participation of several clockassociated proteins within the central oscillator, namely CIRCADIAN CLOCK-ASSOCIATED1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY) [6], and have also helped to identify and potential downstream targets of CCA1 [7, 8]. In this chapter, we outline the protocol for pulsing clock gene expression in transgenic Arabidopsis seedlings that contain an Alcinducible construct. Here, as an example, we use the Alc::CCA1 line [6]. This line harbors both the regulator construct (35S::AlcR) and an effector construct (pAlcA::CCA1). A parental line containing only the regulator construct (and no effector), AlcR-OX is also included in the experiment as a control. After performing the induction, seedlings are harvested over a time-course for future total RNA and/or protein extraction that can be used for subsequent gene expression analyses.

2

Materials

2.1 Growth of Alc-Inducible Lines in Tissue Culture

1. Arabidopsis seeds (see Note 1): (a) Alc-inducible CCA1 line (Alc::CCA1). (b) Parental line AlcR-OX, to serve as control. 2. Chemical hood and laminar flow hood. 3. 1.7-ml Microcentrifuge tubes, 50-ml conical tubes (sterile), commercial bleach, concentrated HCl, and glass bell jar. 4. 0.1 % (w/v) electrophoresis-grade agarose in H2O, sterile. 5. Sixteen 90 mm × 25 mm plastic petri plates containing solid Murashige and Skoog (MS) medium [9].

Pulsed Induction of Clock Genes

205

6. 3 M Micropore surgical paper tape (1-in. wide; Cat. 1530-1). 7. Nylon net filter (20 μm pore size (Cat. NY20, Millipore Corp.)), cut into sixteen 80 cm diameter circles, sterile (see Note 2). 8. Two pairs of forceps. 9. Plant growth chamber. 2.2 Pulse of CCA1 Expression

1. Thick chromatography paper (Fisher Scientific, Inc.), cut into 80–85 cm diameter circles (see Note 3). 2. 100 ml of 1 % (v/v) EtOH in H2O. 3. Eight unused 90 mm × 25 mm plastic petri plates containing solid MS. 4. 150 mm Plastic petri dishes. 5. Paper towels. 6. 1.7-ml Microcentrifuge tubes. 7. Liquid nitrogen.

3

Methods

3.1 Growth of Alc-Inducible Lines in Tissue Culture

1. Place approximately 100 μl of AlcR-OX and Alc::CCA1 seeds into 1.7 ml microcentrifuge tubes. “Gas-sterilize” the seeds: place the tubes in a bell jar along with a 100 ml glass beaker containing 50 ml commercial bleach. Move the bell jar into a chemical fume hood and carefully add 1.5 ml concentrated HCl to the bleach and swirl gently to mix. Replace the lid on the bell jar and incubate for 4 h. When finished, air out the seeds in a laminar flow hood for 15 min. Use sterile technique for the remainder of this section. 2. Add 32 ml of 0.1 % agarose (in H2O) to two sterile 50 ml conical tubes. Pour the sterilized seeds from the previous step into the agarose and mix well to disperse the seeds. Cover with foil and place at 4 °C to stratify for at least 2 days. 3. Prepare 1 l MS medium with 15 g agarose (1.5 % (w/v) final concentration). Autoclave on the liquid cycle for 15 min. Once cooled, pour the medium into 90 mm × 25 mm petri dishes about halfway full and place the lids on. Once the medium begins to harder, dry thoroughly for at least 15 min with the lids off (see Note 4). 4. Using two pairs of forceps, place one 80 mm diameter nylon net filter onto each MS plate. Be sure to get rid of any trapped air bubbles underneath. 5. Sow 4 ml of Alc::CCA1 and AlcR-OX seeds from step 2 onto eight MS plates (16 plates total). Remove the lids of the plates

206

Stephen M. Knowles et al.

and let dry completely in the hood. Once dry, close the lids on the plates and seal with Millipore tape (see Note 5). 6. Place the plates in a growth chamber and let seedlings grow for 7 days under light-dark entrainment conditions (12 h light:12 h dark) (see Note 6). 3.2 Pulse of CCA1 Expression

1. After 7 days of growth, place the seedlings under constant light (LL) for at least one full cycle (24 h). 2. For both Alc::CCA1 and AlcR-OX, designate four plates that will serve as a noninduced control. The remaining four plates for both lines are to be induced. 3. Choose a time during day 2 in LL to perform the induction. Harvest a 0-h time-point for AlcR-OX and Alc::CCA1. Using forceps, take about one-eight of the seedlings from each of the four plates and shove them into the bottom of a 1.7 ml microcentrifuge tube. Snap-freeze the four tubes in liquid nitrogen and store at −80 °C (see Note 7). Return the noninduced plates to the growth chamber and proceed with the induction for the remaining plates in (next step) (see Note 8). 4. Place a chromatography paper circle on the inner side of each plate lid. Pour just enough 1 % (v/v) EtOH onto the circles until saturated. Invert the lids and tap out any excess liquid. Put the lids back onto each plate and let incubate for 10 min at room temperature (see Note 9). 5. After 10 min, using forceps, grab the nylon filter + seedlings from one side and lift off of the plate (see Note 10). Using a second pair of forceps grab the other side of the filter and bring the filter to a horizontal. Gently float on H2O in a 150 mm petri dish for 10 s. Move the filter onto a paper towel to dry for 2 min. Finally, place the filter onto a fresh MS plate and return the plates to the growth chamber (see Note 11). 6. Harvest seedlings for the seven remaining time points: 1, 2, 4, 8, 12, 16, and 24 h. 7. mRNA and/or protein can be extracted from the tissue for analysis of the CCA1 pulse and/or expression of clockcontrolled genes (quantitative RT-PCR, western blotting, etc.) [6]. The results from such an experiment are given in Fig. 2.

4

Notes 1. Our lab has generated four relevant Alc-lines that can be ordered from the Arabidopsis Biological Resource Center (ABRC; http://abrc.osu.edu): Alc::CCA1 (CS67790), Alc::LHY (CS67791), Alc::TOC1 (CS67792), and AlcR-OX (CS67789). The binary vector for creating your own effector construct,

Pulsed Induction of Clock Genes

Relative CCA1 mRNA level

40

207

Control +EtOH

30

20

10

0 24

36

48

60

72

Time in LL (h)

Fig. 2 A pulse of CCA1 mRNA using the Alc-inducible system. Nine-day-old Alc::CCA1 seedlings were exposed to EtOH vapor for 10 min (“+EtOH”). A control group remained untreated (“Control”). The induction was performed at 38 h in continuous light (LL) (arrowhead), a time when CCA1 levels are normally low. RNA levels were measured by quantitative RT-PCR. CCA1 levels are relative to that of a non-cycling control RNA, RNA HELICASE8 (RH8) [6]. White and hatched boxes denote subjective day and subjective night, respectively

pEffector [6], is available from UCLA (contact Dr. Chentao Lin, [email protected]). 2. Cutting nylon filters requires a sharp pair of scissors and a little practice. A circle template can be made out of a crisp (i.e., unused) manila folder. Draw the circle with a compass. Using the template as a guide, gently draw a circle on the nylon filter with a dull pencil. When cutting the filter, keep the scissors stationary, and move only the filter as you cut. Sterilize the filters by wrapping in aluminum foil and autoclaving on the dry cycle for 15 min (they will shrink a little). The filters are reusable and can be used for multiple induction experiments; we routinely reautoclave and use them up to three times. 3. The paper circles need to fit within the underside of the lid of the petri dish when wet. There is no need to sterilize them. 4. This growth medium is harder than typical solid MS. The reasoning behind this is twofold. Firstly, it ensures a firm platform on which to rest the nylon net filter. Secondly, it serves to reduce the level of “auto-induction” (effector gene expression in the absence of the inducer) of the system [3, 10]. 5. Before sowing, let the seeds-in-agarose come to room temperature for at least 15 min. This will reduce the viscosity and make it easier to spread on the filter. When sowing, pipette the liquid into the center of the filter and spread evenly by tilting the plate in a circular motion. At this point it is necessary for

208

Stephen M. Knowles et al.

the plates to be thoroughly dried as a wet plate may increase the level of auto-induction. Perform the drying step until no standing liquid remains. 6. Regarding entrainment, it is useful to set the timing of lightson and lights-off appropriately so that you can perform the induction at a time-of-day that is convenient for you. 7. There is no need to use sterile technique for this 24-h experiment. The induction and tissue harvesting can be performed on a lab bench. 8. Keep in mind that the Alc-inducible system is very sensitive to EtOH vapor. Hence you must keep the noninduced seedlings well clear of any source of EtOH. 9. A 10-min induction is sufficient to pulse CCA1 mRNA in Alc∷CCA1 seedlings to about the peak level observed in wildtype plants [6]. In general, the induction conditions will have to be determined empirically to achieve the desired level of induction. The percentage of EtOH in water can be altered, as well as the timing of vapor exposure to the seedlings. Indeed, different Alc-inducible lines will likely require different induction conditions to achieve the same fold-induction. 10. If you desire continuous induction of the effector gene, skip this step and leave the seedlings in their original plates. In this case it is also wise keep these plates well-clear of the noninduced group. 11. Briefly floating the nylon filter on water helps to wash away any residual EtOH. When introducing the nylon filter to a fresh MS plate, ensure that it is devoid of any standing liquid. If there is some liquid, the lids can be left off to help dry the plates. References 1. Borghi L (2010) Inducible gene expression systems for plants. Methods Mol Biol 655: 65–75 2. Caddick MX, Greenland AJ, Jepson I et al (1998) An ethanol inducible gene switch for plants used to manipulate carbon metabolism. Nat Biotechnol 16:177–180 3. Roslan HA, Salter MG, Wood CD et al (2001) Characterization of the ethanol-inducible alc gene-expression system in Arabidopsis thaliana. Plant J 28:225–235 4. Salter MG, Paine JA, Riddell KV et al (2002) Characterisation of the ethanol-inducible alc gene expression system for transgenic plants. Plant J 16:127–132 5. Sweetman JP, Chu C, Qu N et al (2002) Ethanol vapor is an efficient inducer of the alc gene expression system in model and crop plant species. Plant Physiol 129:943–948

6. Knowles SM, Lu SX, Tobin EM (2008) Testing time: can ethanol-induced pulses of proposed oscillator components phase shift rhythms in Arabidopsis? J Biol Rhythms 23:463–471 7. Lu SX, Knowles SM, Webb CJ et al (2011) The Jumonji C domain-containing protein JMJ30 regulates period length in the Arabidopsis circadian clock. Plant Physiol 155:906–915 8. Lu SX, Webb CJ, Knowles SM et al (2012) CCA1 and ELF3 interact in the control of hypocotyl length and flowering time in Arabidopsis. Plant Physiol 158:1079–1088 9. Murashige T, Skoog F (1962) A revised medium for rapid growth and bio-assays with tobacco tissue cultures. Physiol Plant 15:473–497 10. Deveaux Y, Peaucelle A, Roberts GR et al (2003) The ethanol switch: a tool for tissuespecific gene induction during plant development. Plant J 36:918–930

Chapter 14 The Use of Fluorescent Proteins to Analyze Circadian Rhythms Ekaterina Shor, Miriam Hassidim, and Rachel M. Green Abstract Compared with luciferase which is widely used as a reporter for circadian rhythms in Arabidopsis thaliana, available fluorescent markers are generally too stable to allow circadian oscillations to be measured. However, we have developed a technique to use the nuclear localization of circadian-controlled transcription factors fused to a fluorescent reporter as a means of measuring circadian rhythms. This technique has the advantage of being suitable for analyzing rhythms at the level of individual cells and in living plants. Key words Circadian, Arabidopsis, Fluorescent, Confocal, Protein, Nuclear

1

Introduction To date almost all the studies on the Arabidopsis circadian system have been carried out at the level of the whole plant and the canonical model of the circadian system that has been built [1] represents an average either of the circadian systems in all the cells of the plant or of all the cells in the above-ground parts of the plant. However, as a few recent papers have demonstrated, there are organ/cellspecific differences in oscillator mechanisms. In one report, it was shown that roots have a simplified version of the aerial oscillator consisting of a subset of the canonical oscillator components. Another report identified a vascular tissue-specific mechanism for regulating the stability of a key oscillator component [2, 3]. Given these findings, it is highly likely that there are other organ/cellspecific differences, but that such differences are hard to monitor using existing techniques for analyzing circadian rhythms. We have developed a technique to examine circadian rhythms in individual cells in intact, living plants. Our technique is based on the fact that a number of the key proteins in the oscillator are transcription factors that show circadian oscillations in nuclear accumulation. These oscillating transcription factors (OTFs) can be

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_14, © Springer Science+Business Media New York 2014

209

210

Ekaterina Shor et al.

fused to a fluorescent protein reporter (FP) such as YFP (YELLOW FLUORESCENT PROTEIN) and the rhythmic nuclear accumulation of the resulting fusion protein (OTF:FP) monitored by confocal microscopy. The great advantage of this system is that rhythms may be measured in any cell type at the resolution level of a single cell. We have successfully used the single MYB transcription factor, CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) to examine rhythms in stomatal, epidermal, and mesophyll cells. The coding region of CCA1 was fused with YFP under the control of the CCA1 upstream region (966 bp promoter + 208 bp untranslated region [UTR]). The resulting construct CCA1pro∷CCA1-YFP was cloned into the pMLBART binary vector [4] and used to transform CCA1null (cca1-1) Arabidopsis plants by the floral dip procedure [5]. Seeds were harvested and selected for the presence of the construct.

2 2.1

Materials Plant Growth

1. Plants: Homozygous plants with the OTF:FP fusion protein. 2. MS (Murashige and Skoog medium): 4.4 g/l, sucrose 3 % in double-distilled water (DDW). Adjust to pH 5.7 with KOH and autoclave. Pour 25 ml into 90 mm × 15 mm Petri dishes. Allow medium to cool and then store plates well-wrapped in plastic wrap at room temperature. 3. Sterilization solution: 50 % bleach (in DDW), 0.15 % SDS (sodium dodecyl sulfate).

2.2 Confocal Microscopy

1. DAPI stain solution: 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI) 50 μg/ml, pluronic acid 0.02 %, DDW. Prepare this solution directly before the microscopy session (see Note 1). 2. Small plastic Petri dish with a glass insert in the base: Remove a 20 mm × 25 mm rectangle from the bottom of a small (50 mm) Petri dish. Stick a glass cover slip with a colorless nail polish externally to the base of the Petri dish so that it covers the hole (Fig. 1; see Note 2). 3. (Optional) Polymer sealant to fix the upper coverslip in place and reduce water loss. We used Oranwash L (Zhermack, Badia Polesine, Italy) condensation silicone and activated its polymerization with Indurent Gel (Zhermack, Badia Polesine, Italy) by mixing at a ratio of 10:1 Oranwash L: Indurent Gel. The sealant should be used immediately after mixing. 4. Confocal microscope equipped with a linear encoded motorized stage and a ×40/1.3 oil immersion objective. 5. ImageJ program (http://rsbweb.nih.gov/ij/index.html).

The Use of Fluorescent Proteins to Analyze Circadian Rhythms

211

Fig. 1 Small Petri dish with a glass inserted in the base

3

Methods

3.1 Plant Growth for Microscopy Analysis

1. Sterilize seeds of plants expressing the OTF:FP fusion protein in sterilization solution for 10 min, then wash four times with sterile DDW. Imbibe the seeds at 4 °C for 4 days in sterile DDW to optimize germination. 2. Sow on Petri dishes with MS +3 % sucrose medium at low densities (~15 seeds/90 mm Petri dish). 3. Grow for 2 weeks in LD (long days; 14 h light 100–130 μE/ m2/s per 10 h dark) at 23 °C before either using the plants to observe FP accumulation in the nucleus at a single time-point (Subheading 3.2) or transferring plants to LL (continuous light) for 3–7 days to analyze rhythms (Subheading 3.3).

3.2 FP accumulation in Nucleus by Confocal Microscopy

Staining with DAPI stain to ensure that the OTF:FP protein accumulates in nucleus (see Note 3). 1. Cut a leaf from the plant at a time when the levels of the OTF protein are expected to be maximal (see Note 4) and put it into DAPI stain solution for 20 min. 2. Remove excess DAPI by rinsing the leaf into DDW three times. 3. Put the leaf on the microscope slide into a drop of DDW and cover with a cover slip. 4. Put the slide on the stage of confocal microscope. Use ×40/1.3 oil immersion objective. The leaves should be examined at a number of different excitation and emission wavelengths to distinguish the DAPI-stained nuclei, OPF:FP, and chloroplasts. The DAPI excitation peak is 405 nm and its emission peak is 370–430 nm. The excitation and emission wavelengths for FP fluorescence depend on the FP that is used; the information is available from the supplier. For YFP we used 515 nm excitation and measured emission at 535–565 nm.

212

Ekaterina Shor et al.

Fig. 2 Chlorophyll autofluorescence in the chloroplasts (red) and fluorescence from CCA1:YFP in the nucleus (green)

Fig. 3 Leaf ready to be imaged

The excitation and emission for chloroplast autofluorescence are 633 and 655–755 nm, respectively (Fig. 2). 5. Compare localization of DAPI staining and FP fluorescence; the FP fluorescence should clearly correlate with the DAPI nuclear staining. 3.3 Analysis of Circadian Rhythms

1. Carefully cut off a block of agar around the plant to be imaged, avoiding damage to the roots. Transfer the plant together with the block of agar to the small Petri dish with the glass coverslip inserted in the base. Keep the plant upright. 2. Choose a leaf to be imaged. The fifth or sixth leaf is usually most convenient. Bend the leaf down gently so that it is resting on the glass cover slip on the base of Petri dish. The block of agar with the plant may have to be tilted. Add a drop of DDW and cover with an upper cover slip. If desired, the upper coverslip may be gently fixed in place on three sides with polymer sealant to keep the leaf in position for imaging and prevent excess water loss (Fig. 3). 3. Put the Petri dish on the stage of confocal microscope. Use a ×40/1.3 oil immersion objective. The wavelength parameters for FP and chloroplast autofluorescence are the same as described above (step 4 in Subheading 3.2). 4. Chose three fields on the leaf using a linear encoded motorized stage and scan each field every hour. In between imaging sessions keep the plants in LL at 100–130 μE/m2/s.

The Use of Fluorescent Proteins to Analyze Circadian Rhythms

213

5. Calculate levels of FP fluorescence at each time point with ImageJ using the Plugins/Stacks - T Function/Intensity v Time Monitor option.

4

Notes 1. It is convenient to keep 100× DAPI as a stock solution and add pluronic acid at −20 °C without the pluronic acid and to dilute it as needed with DDW. 1 ml of DAPI stain solution should be sufficient for each microscopy session. 2. The glass between the sample and the objective of the microscope should be as thin as possible. 3. Nuclear staining with DAPI may cause problems with the subsequent visualization of fluorescence from the rhythmic accumulation of the OTF:FP protein. Therefore, once it has been established that the OTF:FP protein accumulates in nucleus, a separate set of leaves should be used to measure the rhythmic nuclear accumulation of the OTF:FP protein. 4. If the kinetics of translation do not match those of nuclear accumulation and the OTF:FP may not be observed in the nucleus at times when levels of the OTF protein in the cell are expected to be maximal. In this case, other times should be tried.

Acknowledgement This work was supported by US Friends of Hebrew University grant 0367445, ISF grant 0398636 and DFG grant 0308300. References 1. Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22:648–657 2. James AB, Monreal JA, Nimmo GA, Kelly CL, Herzyk P, Jenkins GI, Nimmo HG (2008) The circadian clock in Arabidopsis roots is a simplified slave version of the clock in shoots. Science 322:1832–1835 3. Para A, Farre EM, Imaizumi T, Pruneda-Paz JL, Harmon FG, Kay SA (2007) PRR3 is a vascular regulator of TOC1 stability in the

Arabidopsis circadian clock. Plant Cell 19:3462–3473 4. Yakir E, Hilman D, Kron I, Hassidim M, Melamed-Book N, Green RM (2009) Posttranslational regulation of CIRCADIAN CLOCK ASSOCIATED1 in the circadian oscillator of Arabidopsis. Plant Physiol 150:844–857 5. Davis AM, Hall A, Millar AJ, Darrah C, Davis SJ (2009) Protocol: streamlined sub-protocols for floral-dip transformation and selection of transformants in Arabidopsis thaliana. Plant Methods 5:3

Chapter 15 Measuring Circadian Oscillations of Cytosolic-Free Calcium in Arabidopsis thaliana Timothy J. Hearn and Alex A.R. Webb Abstract Circadian oscillations of cytosolic-free calcium can be measured in plants by observing luminescence of the bioluminescent calcium binding protein aequorin. Here we describe the use of intensified photon-counting CCD cameras to measure circadian oscillations in aequorin bioluminescence from Arabidopsis thaliana. Key words Circadian, Calcium, Aequorin, Photon-counting imaging, CCD

1

Introduction: Cytosolic-Free Calcium in the Circadian System The plant circadian clock is a molecular oscillator with an endogenous free-running period of approximately 24 h in constant light that is thought to maximize fitness-related traits by synchronizing the timing of key physiological processes with the phase of the light and dark cycles of the environment [1]. The circadian clock consists of a set of feedback loops between both genetic and cytosolic components (Fig. 1). Oscillations of cytosolic-free calcium ([Ca2+]cyt) with a period of 24 h that persisted in constant conditions were first reported in tobacco and Arabidopsis thaliana [2]. These circadian oscillations of [Ca2+]cyt peaked during the subjective day and troughed during the subjective night [2]. The [Ca2+]cyt rhythms were absent in constant dark, or in seedlings grown on agar containing added sucrose [2]. The dynamics of the circadian rhythms of [Ca2+]cyt are a consequence of co-regulation between circadian signalling and light signalling mediated by the cryptochromes (blue) and phytochrome A (red) [3]. Mutations in core circadian clock genes affect circadian rhythms of [Ca2+]cyt consistent with [Ca2+]cyt being regulated by a transcriptional translational feedback loop [4]. Circadian [Ca2+]cyt oscillations are downstream of CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) with a delay of about 5 h [3]. CCA1 is thought to be repressive to [Ca2+]cyt because

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_15, © Springer Science+Business Media New York 2014

215

216

Timothy J. Hearn and Alex A.R. Webb

Fig. 1 [Ca2+]cyt and the Arabidopsis thaliana circadian clock. (a) Interlocked morning and evening feedback loops in the nucleus provide robust oscillations of gene expression. Circadian oscillations of cytosolic-free calcium [Ca2+]cyt are driven by oscillations of cADPR and are downstream of CCA1. (b) Raw aequorin luminescence at ZT 0 measured from 35 samples using the equipment and protocols described in the text. (c) Aequorin luminescence measured from Col-0 plants using the Photek ICCD225 and the experimental protocols described in the text. Data was collected for two days in 12:12 light and dark cycles, represented by white and black bars, before plants were transferred to constant light for 4 days. Images are taken every 2 h with an integration of 1,500 s

both CCA1-ox gain-of-function and cca1-1 loss-of-function are arrhythmic for [Ca2+]cyt and in the CCA1-ox, [Ca2+]cyt is arrested at relatively low levels of [Ca2+]cyt, whereas in cca1-1, [Ca2+]cyt is arrested at relatively high levels [4]. It has been proposed that circadian oscillations of [Ca2+]cyt are driven by circadian oscillations in the cytosolic abundance of the Ca2+ agonist, cyclic ADP ribose [5]. In the timing of cab1-1 (toc1-1) mutant [Ca2+]cyt is uncoupled from other circadian rhythms, persisting with a period of approximately 24 h despite all other circadian rhythms tending to a short period of 21 h in this background [4]. The ability of circadian

Circadian Oscillations of Cytosolic-Free Calcium

217

oscillations of [Ca2+]cyt to uncouple from other circadian rhythms, such as Lhcb gene expression, might suggest the presence of multiple circadian oscillators differentially regulating different outputs [6], possibly in a cell type-specific manner [4]. [Ca2+]cyt is an important second messenger, usually associated with rapid signalling [7]. Therefore, we study circadian oscillations of [Ca2+]cyt to understand why this key regulator of signalling is under circadian control and the mechanism by which the circadian clock regulates physiology. A variety of techniques have been used to measure [Ca2+]cyt in plants [8], most of which are invasive (e.g., microelectrodes or microinjection of fluorescent indicators) or require fluorescent detection in non-photosynthetic tissues using high energy light excitation (e.g., recombinant FRET-based reporters) and are therefore currently not suited to the prolonged in vivo measurements from leaf tissues required for circadian analyses. Circadian rhythms of [Ca2+]cyt have so far only been recorded from plant tissues using recombinant aequroin from Aequorea victoria. Aequorin consists of the protein apoaequroin and the luminophore substrate coelentrazine. The functional aequorin protein has three EF binding sites for Ca2+ which when bound cause the protein to undergo a conformational change and the release of light with peak emission at 470 nm. Thus the level of luminescence from reconstituted aequorin is proportional to the free Ca2+ in the compartment to which the aequorin protein has been targeted. To measure [Ca2+]cyt in plants, the apoaequorin cDNA is expressed under control of the 35S promoter from Cauliflower Mosaic Virus (35S:AEQ) to give high constitutive expression in all tissues [9]. Coelentrazine is added to reconstitute the functional aequorin protein. The aequorin bioluminescence is detected using low background, high sensitivity photon-counting detectors (Fig. 2). Here we present the standard protocol used in our laboratory to measure circadian oscillations of [Ca2+]cyt from A. thaliana expressing the 35S:AEQ construct.

2

Materials

2.1 Preparation of Coelentrazine Working Solution

1. Coelentrazine stock solution: Coelentrazine is light sensitive until reconstituted in to the aequorin protein, therefore work under shaded conditions. Add 591 μl 100 % (w/v) methanol to 1 mg powdered coelentrazine (Nanolight, http://www. nanolight.com) to give a concentration of 4 nmol/μl. Aliquot into 25 μl microcentrifuge tubes. Dry samples in a speed dryer until methanol has evaporated giving 100 nmol coelentrazine pellets (in a Speedvac with medium heat this takes 25 min). Wrap aliquots in silver foil and store at −80 °C.

218

Timothy J. Hearn and Alex A.R. Webb

Fig. 2 Cartoon of the apparatus for measuring circadian oscillations of [Ca2+]cyt. The intensified CCD camera is positioned in a light-tight camera chamber vertically above a 25-well plate. Illumination for seedling growth and to control the photoperiod is provided by two panels comprising an equal mix of red (660 nm) and blue (470 nm) light emitting diodes. The camera Photek ICCD225 camera system is controlled from a computer running IFS32

2. Coelentrazine working solution: work under shaded conditions. Dissolve 100 nmol aliquots in 100 μl 100 % (w/v) methanol. Make up to 5 ml with autoclaved dH2O in a larger tube to give 20 μM coelentrazine. Wrap working solution in aluminum foil and store at −4 °C, use within 72 h. 2.2 PhotonCounting Camera

1. CCD intensified photon-counting camera (Photek ICCD225), mounted in a light-tight camera chamber containing LED arrays comprising both blue (660 nm) and red (470 nm) LEDs (see Note 1) and a computer running IFS32 control software (see Note 2). Camera should be stored in a darkroom with temperature controlled to 19 °C (see Note 3).

2.3 Photon Multiplier Tubes

1. A low-noise 30-channel photomultiplier tube (Hamamatsu R2693P) connected to a photon counter (Hamamatsu C3866) stored in a light-tight chamber (see Note 4). Apparatus should be stored in a darkroom with temperature controlled to 19 °C.

2.4

1. A. thaliana transformed with APOAEQUORIN cDNA under the control of the 35S Cauliflower Mosaic Virus promoter (35S) promoter (35S::AEQ). This can be achieved using Agrobacterium tuberfaciens-mediated transformation.

Plant Materials

Circadian Oscillations of Cytosolic-Free Calcium

219

There are two aequorin constructs in main use, pMAQ2.4 [2] and pJAAAEQ [4]. As an alternative to A. tuberfaciens-mediated transformation the 35S:AEQ reporter can be introduced into the genotype of interest by crossing. Transformed plants are selected for the aequorin-containing T-DNA insertion with either kanamycin (pMAQ2.4) or dl-phosphinothricin (PPT, pJAAAEQ). Resistant individuals expressing high levels of AEQUORIN are selected for in the T1/T2 or F1/F2 generation. A plate reader (FluostarOptima; www.bmglabtech. com/) is used to measure luminescence from individual leaf discs from 21-day-old plants left to incubate overnight in 20 μM coelentazine. Luminescence is recorded every 1 s for a total of 60 s following addition of 100 μl 2 M CaCl2 20 % (w/v) ethanol to fully discharge the AEQUORIN. Only the lines with highest levels of integrated AEQUORIN luminescence following discharge are selected for further use. The above constructs have already been introduced into a range of ecotypes and circadian mutants, e.g., C24; toc1-1, toc1-2, ztl-1, Col-0; CCA1-ox, elf3-1, cry2-1, WS; cca1-1, Ler; lhy-1, phyA-201, phyA-201 phyB-5, cry1, cry1 cry2, phyB-9 [4]. 2.5 Media and Assay Plate

1. 0.5 MS 0.8 % (w/v) agar media; 2.15 g MS salts and 8 g bacto agar per liter sterile dH2O (see Note 5). 2. Assay plate; Cover the wells of a 25 well plate with black insulation tape in a sterile flow hood and rinse in 70 % ethanol to sterilize. Sterilize 1 cm tall pieces of 7 mm surgical tubing (rings). Add 2 ml 0.5 MS 0.8 % (w/v) agar media to each well. Add four rings per well on top of the media.

3

Methods

3.1 Measuring Circadian Oscillations of Cytosolic-Free Calcium ([Ca2+]cyt) in A. thaliana

1. Prepare a 25-well plate for the assay. 2. Surface sterilize and sow clusters of 20 seeds of 35S::AEQ plants into each ring (see Note 6). Stratify for 96 h at 4 °C. Transfer to growth cabinet with 12:12 light/dark cycles. 3. Dose 9-day-old plants with 40 μl coelenterazine under shaded conditions. Dose plants at the onset of the dark cycle. This gives the plants 12 h to absorb the coelenterazine before it is broken down by light. Repeat this on day 10. 4. Place 11-day-old plants in the camera chamber at dawn (ZT0) and focus the camera, do not expose the plants to any form of illumination except green light before dawn. Start the automated schedule for the experiment immediately after focussing.

220

Timothy J. Hearn and Alex A.R. Webb

5. Schedule. To record a high-quality data set it is desirable to measure AEQ luminescence from two light dark cycles before releasing the plants into free-run in constant light or constant darkness. Data should be collected for a minimum of 4 days in constant light as the first cycle in constant light is not a true free-running oscillation. To take a measurement the LED arrays must be turned off, and a pause inserted for 200 s to let the delayed fluorescence from the chloroplasts disperse [10]. Luminescence is usually integrated for 1,500 s to increase the signal to noise ratio. As this process takes 1,700 s in total, measurements should only be taken every 2 h, so that the plant does not interpret the measurements as dark breaks in constant light. To automate the imaging process a LRN (learn) file can be created using the Learn function in the Photek Software by going to Learn → Learn and run the program by going to Learn → Recite. Below is an example of a Learn file instruction to turn the LED arrays on for 5,500 s, turn the LED arrays off, wait for 200 s, and then integrate a photon count for 1,500 s. The file is then saved as Sample 0001.sxy. zap ND0 100 camera_off lamp on message Lamp on for 5500s pause 5500.00 lamp off message Lamp off waiting 200 seconds pause 200.00 camera_photoncount clear_integration start_integration delete A colour_log display_image A window A pause 1500 stop_integration display_image A camera_off save_tri 'C:\My Documents\Automated\Sample0001. sxy' A Forty-eight cycles of this instruction set would be sufficient for photon-counting imaging every 2 h in “constant” light for 96 h. Light dark cycles can be programmed using the lamp on and lamp off command. Likewise constant darkness can be maintained by changing to: lamp off for all cycles.

Circadian Oscillations of Cytosolic-Free Calcium

221

It is important to change the file name sequentially for each integration, i.e., Sample 0001.sxy, Sample 0002.sxy, Sample 0003. sxy. Otherwise each save file will overwrite the previous (Fig. 3). 3.2 Measuring Circadian Oscillations of Cytosolic-Free Calcium ([Ca2+]cyt) in A. thaliana Using Photon Multiplier Tubes

1. Surface sterilize and sow individual seeds of 35S:AEQ plants onto 0.5 MS 0.8 % (w/v) agar. Stratify for 96 h at 4 °C. Transfer to a growth cabinet with 12 h light, 12 h dark cycles. 2. Dose 6-day-old plants with 40 μl coelenterazine. Dose plants at the onset of the dark cycle. This gives the plants 12 h to absorb the coelenterazine before it is broken down by light. 3. Transfer ten to twelve 7-day-old plants to vials containing 2 ml 0.5 MS 0.8 % (w/v) agar. Place tubes in the photomultiplier tube apparatus. To make a measurement, turn the light off for 1 min to allow light from delayed fluorescence to scatter and then measure aequorin bioluminescence for 1 min. Repeat every hour for the duration of the experiment.

3.3

Data Analysis

Using the LRN file code from Subheading 3.1 integrations are saved as .sxy files in the specified file location, in this case a folder in the C:/ drive named automated. To open .sxy files in ISF32 go to TRI → select TRI file. Save the opened image as a .pct file (File → save image (ctrl + s). Repeat to save all the images in a sequence as .pct. To analyze the images in the sequence, load the entire sequence (Sequence > load frames), select the first image in the sequence and assign a memory buffer letter (A,B,C, etc.) for the first image, the stored images in the folder will now open as a sequence. Regions of interest for image analysis are defined using the overlay tool (measure > OAS > OAP > circle, etc). Regions of interest can be moved, copied, or resized by selecting the relevant option from the OAP menu. Background luminescence also should be measured from equal sized regions of interest placed in an area of the image devoid of plant material. The OAS layout can be saved for repeated use (OAS > save OAS). Data are extracted from the regions using the OAS menu (measure OAS > extract OAS data) and choose the memory buffer (A,B,C, etc.) in which the image sequence is stored. A text file (.txt) containing the extracted values opens automatically. Saving the .txt allows subsequent import into a spreadsheet (e.g., LibreOffice Calc, Microsoft Excel) for data analyses. Simple analyses include calculation and subtraction of the mean background signal. Advanced analyses might include estimates of circadian period and other parameters. A widely used tool in the plant community for estimates of circadian period is fast Fourier transformed nonlinear least-square analysis (FFT NLLS) [11] implemented in a BRASS interface for MS Excel (http://amillar.org) (see Note 7).

222

Timothy J. Hearn and Alex A.R. Webb

Fig. 3 Learn file editing page in ISF32. Note the sequential increase in file name number in the save_tri command line

Circadian Oscillations of Cytosolic-Free Calcium

4

223

Notes 1. Light: The circadian amplitude of [Ca2+]cyt oscillations is strongly influenced by light intensity [12]. Under 60 μmol/m2/s, a widely used light intensity for other circadian experiments [10], the amplitude of oscillations of [Ca2+]cyt are very low [12]. We recommend using light intensities of 70–80 μmol/ m2/s equal red (660 nm) and blue (470 nm; RB) light. At the time of writing, plant growth chambers containing LED illumination are only now gaining adoption, therefore consideration should be given for the effect of transferring plants from growth chambers using, for example, white fluorescent illumination, to the imaging chamber containing LEDs emitting in a very narrow wavelength range and that for similar illumination intensity by the different light sources, the activation of the red and blue photoreceptors might be very different. 2. Choice of photon-counting camera: A number of photon-counting cameras are available to measure bioluminescence, but few are sufficiently sensitive to measure AEQUORIN in vivo. Cooled CCD and EM-CCD cameras routinely used for luciferase assays are not sufficiently sensitive. At the time of writing circadian oscillations of [Ca2+]cyt have been imaged in Arabidopsis using only CCD cameras coupled to amplifying image intensifiers manufactured by Photek (Hastings, UK). We currently use the fourth generation high resolution photon-counting package (HRPCS4), which incorporates the ICCD225 photon-counting camera. The camera is mounted in the roof of a light-tight chamber and images down on plant material placed within. A high numerical aperture lens is required to maximize signal acquisition. The ICCD225 fits 1 in. cine (C) mount camera lenses. We currently fit 0.8 mm Schneider lenses. To increase the depth of field, spacer rings can be fitted onto the female thread of the camera C mount. 3. Temperature: Although circadian rhythms are not affected by temperature in the same manner as other biological processes and have low Q10 values, temperature is an entrainment cue and anthropomorphic-induced temperature cycles, such as air conditioning, is enough to entrain plants. Therefore during experimentation the temperature should be carefully controlled. In our laboratory the imaging experiments are performed at 20 ± 1 °C.

224

Timothy J. Hearn and Alex A.R. Webb

4. Measuring circadian [Ca2+]cyt oscillations using photon-counting multiplier tubes: Circadian [Ca2+]cyt oscillations were first reported in Nicotiana plumbaginiolla (tobacco) [2]. For these initial experiments photon multiplier tubes were used (Hamamatsu R2693P) and manually assayed every hour. Two tobacco seedlings per cuvette provide adequate luminescence. Low background count photomultiplier tubes can be an order of magnitude more sensitive than even the most sensitive of photon-counting cameras. This can be an advantage, particularly considering the lower cost of multiplier tubes compared to photon-counting camera systems. However, the low cost and high sensitivity is counteracted by the relatively small circular detector area (c. 19.6 cm2) that allows only one or two seedlings at a time to be measured, necessitating the use of mechanical or manual devices to allow movement of seedlings in and out of the plane of the detector. This coupled with a light-tight housing for measurement requires a relatively specialized engineering solution and computerized electronics to provide control. This contrasts with photon-counting imaging which allows an imaging area of 10 cm2 using the Photek HSPRC4 package and ICCD225 camera with a Schneider 0.8 mm lens. 5. Sucrose: Sucrose should not be added to media prepared for an experiment measuring circadian [Ca2+]cyt oscillations. Sucrose (1–3 % (w/v)) abolishes circadian [Ca2+]cyt oscillations, through an unknown mechanism [2, 13]. 6. Sensitivity: To increase the signal to noise ratio we recommend using 20 seedlings grown in clusters to increase the photon density at each pixel of the CCD array. Due to the low intensity aequorin luminescence it is required to have the plant material relatively close to the detector. In our laboratory this is achieved using a jack to raise the level of the 25-well plate to a distance of roughly 10–15 cm from the camera lens. 7. Calibrating levels of [Ca2+]cyt using luminometry: The assays presented here will return the level of [Ca2+]cyt in terms of aequorin bioluminescence. Absolute [Ca2+]cyt can be calculated from aequorin bioluminescence in short-term experiments using a calibration curve [14]. Calibration of aequorin biolumin\escence using the pMAQ2 construct is achieved using the following equation derived from empirical studies.

Circadian Oscillations of Cytosolic-Free Calcium

225

Photon counts are converted to [Ca2+]cyt by the following equation: − log Ca 2+ 

cyt

 L = a ×  − log  c   Lt

  + b  

where Lc = luminescence counts per second; Lt = total luminescence counts over the entire experiment, including discharge of all functional aequorin in permeabilized cells using excess CaCl2 (see Subheading 2.3); a = 0.3326; b = 5.559 [15]. However, these approaches have not proved suitable for the long-term [Ca2+]cyt measurements associated with circadian studies in plants because the dynamic range of intensified CCD detectors means it is not possible to discharge all the functional aequorin pool with excess [Ca2+]cyt without saturating the detector and also risking long-term damage to the P43 phosphor screen which converts the electrons back to photons. Therefore it is not possible to obtain a measurement of Lt. As an alternative [12], compared the fold changes in signal intensity of aequorin due to circadian rhythms of [Ca2+]cyt with the fold changes in aequorin signal induced by a range of [NaCl] measured using the same CCD apparatus. These values were compared with previously obtained data estimating of the magnitude of NaCl-induced increases in [Ca2+]cyt in aequorin containing seedlings [16]. The use of photon-counting photomultiplier tubes permitted the estimation of Lt and thus calibration of NaCl-induced changes in aequorin luminescence to [Ca2+]cyt. Love et al. [12] estimated that the changes in bioluminescence during circadian oscillation of [Ca2+]cyt were equivalent to transitions from c. 150 nm to a peak of c. 350 nM. References 1. Dodd AN, Salathia N, Hall A et al (2005) Plant circadian clocks increase photosynthesis, growth, survival, and competitive advantage. Science 309:630–633 2. Johnson CH, Knight MR, Kondo T et al (1995) Circadian oscillations of cytosolic and chloroplastic free calcium in plants. Science 269:1863–1865 3. Dalchau N, Hubbard KE, Robertson FC et al (2010) Correct biological timing in Arabidopsis requires multiple light-signaling pathways. Proc Natl Acad Sci U S A 107:13171–13176 4. Xu X, Hotta CT, Dodd AN et al (2007) Distinct light and clock modulation of cytosolic free Ca2+

oscillations and rhythmic CHLOROPHYLL A/B BINDING PROTEIN2 promoter activity in Arabidopsis. Plant cell 19:3474–3490 5. Dodd AN, Gardner MJ, Hotta CT et al (2007) The Arabidopsis circadian clock incorporates a cADPR-based feedback loop. Science 318:1789–1792 6. Sai J, Johnson CH (1999) Different circadian oscillators control Ca2+ fluxes and lhcb gene expression. Proc Natl Acad Sci U S A 96: 11659–11663 7. Dodd AN, Kudla J, Sanders D (2010) The language of calcium signaling. Annu Rev Plant Biol 61:593–620

226

Timothy J. Hearn and Alex A.R. Webb

8. McAinsh M, Ng CKY (2005) Measurement of cytosolic-free Ca2+ in plant tissue. In: Lambert DG (ed) Calcium signalling protocols, methods in molecular biology. Humana, New York, pp 289–302 9. Knight MR, Campbell AK, Smith SM, Trewavas AJ (1991) Transgenic plant aequorin reports the effects of touch and cold-shock and elicitors on cytoplasmic calcium. Nature 352:524–526 10. Gould PD, Diaz P, Hogben C et al (2009) Delayed fluorescence as a universal tool for the measurement of circadian rhythms in higher plants. Pant J 58:893–901 11. Plautz JD, Straume M, Stanewsky R et al (1997) Quantitative analysis of Drosophila period gene transcription in living animals. J Biol Rhythms 12:204–217 12. Love J, Dodd AN, Webb AAR (2004) Circadian and diurnal calcium oscillations encode photoperiodic information in Arabidopsis. Plant Cell 16:956–966

13. Haydon MJ, Bell LJ, Webb AAR (2011) Interactions between plant circadian clocks and solute transport. J Exp Bot 62:2333–2348 14. Cobbold PH, Rink TJ (1987) Fluorescence and bioluminescence measurement of cytoplasmic free calcium. Biochem J 248:313–328 15. Fricker MD, Plieth C, Knight H et al (1999) Fluorescence and luminescence techniques to probe ion activities in living plant cells. In: Mason WT (ed) Fluorescent and luminescent probes for biological activity, 2nd edn. Academic Press, San Diego, CA, pp 569–596 16. Tracy FE, Gilliham M, Dodd AN et al (2008) NaCl-induced changes in cytosolic free Ca2+ in Arabidopsis thaliana are heterogeneous and modified by external ionic composition. Plant Cell Environ 31:1063–1073

Chapter 16 Circadian Life Without Micronutrients: Effects of Altered Micronutrient Supply on Clock Function in Arabidopsis Patrice A. Salomé, Maria Bernal, and Ute Krämer Abstract The plant circadian clock is formed by a number of interlocked feedback loops that control the expression of thousands of genes. Genetic and pharmacological approaches towards the study of the plant clock are routinely carried out on Murashige and Skoog growth medium, which is both Fe-replete and Cu-deficient. However, it has recently become clear that the plant clock responds to available iron (Fe) supply: circadian pace slows down under conditions of Fe deficiency; circadian period progressively shortens with increasing Fe supply. Here, we describe several growth media that may be used to study the effects of varying micronutrient supply on the circadian clock, in which deficiency in a given micronutrient are imposed by the addition of a specific chelator or, alternatively, by using EDTA-washed agar as gelling agent, thus minimizing micronutrient contamination. Key words Micronutrient, MS medium, Minimal medium, Hoagland medium, FerroZine, Agar, Circadian clock

1

Introduction Because of their sessile nature, plants must cope with various levels of micronutrient availability from the soil, from severe deficiency to toxic excess [1–3]. The transition metals iron (Fe), copper (Cu), manganese (Mn), and zinc (Zn) are all considered to be essential for plant growth and survival, and their deficiency negatively impacts plant development and productivity. Fe deficiency symptoms are perhaps the most common and are characterized by reduced biomass and much lower chlorophyll levels, as one of the enzymes involved in chlorophyll biosynthesis requires Fe for function [4]. An intact circadian clock is critical to reap maximal fitness advantage by anticipating daily environmental transitions (dawn, dusk, and the associated changes in light quantity and quality and temperature, predatory pressure from herbivores and pathogens), as well as coordinating daily events to the proper time of day and proper rate (e.g., starch degradation during the night) [5–8].

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_16, © Springer Science+Business Media New York 2014

227

228

Patrice A. Salomé et al.

Although the circadian clock controls the expression of many genes encoding proteins containing Fe, Zn, Cu, and Mn, little is known about what consequences on clock function, if any, are associated with varying micronutrient supply. Several recent publications have highlighted the importance of the micronutrient Fe in the control of circadian pace [9–11]. One additional publication also suggested that altered Cu uptake affected clock gene expression, although the associated circadian phenotypes were not strong [12]. All three Fe studies reported period lengthening under conditions of low Fe availability, demonstrating the robustness of the phenotype, but also differed in some of their conclusions. Interestingly, the growth medium and growth conditions were not common across publications, raising the possibility that experimental considerations might, at least in part, contribute to the observed differences [13]. In this chapter, we present all three published methods to study the effects of micronutrient deficiency with a focus on Fe. We also offer some notes about the strengths and weaknesses of each approach. Seedlings grown on any of the growth media described here may be used for such downstream analyses as seedling morphology and root elongation under micronutrient deficiency conditions, gene expression profiling with real-time quantitative PCR, or recording of luciferase circadian reporter lines. Many transgenic lines carrying luciferase under the control of a clock-controlled promoter are available on request from our laboratory or the Arabidopsis stock centers NASC (arabidopsis.info) and ABRC (arabidopsis.org). For a complete description of setting up a circadian assay on the Perkin Elmer TopCount Luminometer, the reader is invited to consult Kim et al. [14].

2

Materials

2.1 Components for MS Medium

1. Macronutrients (for full-strength MS): 20.6 mM NH4NO3, 3 mM CaCl2, 18.79 mM KNO3, 1.5 mM MgSO4, 1.25 mM KH2PO4, 100 μM Na2H2EDTA. 2. Micronutrients (for full-strength MS): 1.03 μM (NH4)6Mo7O24, 100 μM H3BO3, 100 μM FeIISO4, 100 nM CuSO4, 100 μM MnSO4, 29.91 μM ZnSO4, 110 nM CoCl2, 5 μM KI, (see Note 1). 3. Vitamins and organic components: 560 nM myo-inositol, 26.64 μM glycine, 4.06 μM nicotinic acid, 2.43 μM pyridoxine HCl, 300 nM thiamine (see Note 2) [15].

2.2 Components for Fe-Limited Minimal Medium

1. Macronutrients: 2 mM Ca(NO3)2, 750 μM K2SO4, 650 μM MgSO4, 100 μM KH2PO4 [16] (see Note 3). 2. Micronutrients: 10 μM H3BO3, 100 nM MnSO4, 50 nM CuSO4, 50 nM ZnSO4, 5 nM (NH4)6Mo7O24 [16].

Circadian Life Without Micronutrients…

229

2.3 Components for Hoagland Medium Lacking Fe and/or Other Micronutrients

1. Macronutrients: 1.5 mM Ca(NO3)2, 1.25 mM KNO3, 750 μM MgSO4, 280 μM KH2PO4 [17] (see Note 3).

2.4

1. Agar (type M from Sigma or similar).

Other Reagents

2. Micronutrients (drop one of Fe, Cu, Mn, or Zn, as needed): 50 μM KCl, 25 μM H3BO3, 5 μM FeIIIHBED, 0.5 μM CuSO4, 5 μM MnSO4, 5 μM ZnSO4, 100 nM (NH4)6Mo7O24 [17].

2. 0.5 M Ethylenediaminetetraacetic acid (EDTA), pH 8.0. 3. Sucrose (see Note 4). 4. 0.2 M HCl. 5. 1 M 2-(N-morpholino) ethanesulfonate (MES)–KOH, pH 5.8. 6. Chelators: 3-(2-Pyridyl)-5,6-bis(4-sulfophenyl)-1,2,4-triazine (ferroZine) for Fe, N,N,N9,N9-tetrakis[2-pyridylmethyl] ethylenediamine (TPEN) for Zn, bathocuproine sulfonate (BCS) for Cu (see Note 5). 7. D-Luciferin, potassium salt. A stock solution of 2.5 mM should be prepared by resuspending the yellow powder into deionized H2O, and pH to 5.8 with 1 M KOH. 2.5

Consumables

1. Plastic petri dishes. 2. Glass petri dishes (for Cu deficiency), 150 mm diameter.

3

Methods All media should be prepared with ultrapure H2O, and seedlings handled in a sterile environment. Table 1 provides a comparison of media composition for quick reference.

3.1 Fe-Deficient MS Medium

Based on Chen et al. [11], seedlings are first grown on half-strength MS medium (thus containing 50 μM FeIIINaEDTA) for 8 days. Micronutrient deficiency conditions are imposed by transferring seedlings to new plates with half-strength MS medium lacking one micronutrient (and 50 μM EDTA in the case of Fe, see Notes 1 and 2) and additionally supplemented with a specific chelator to chelate any contaminant micronutrients contained in the agar. Control conditions are provided by maintaining seedlings on halfstrength MS medium (see Note 6). 1. Mix all macro- and micro-nutrient solutions and organic components (see Notes 2 and 7). Add 0.05 % (w/v) MES and adjust pH to 5.8 with 1 M KOH. 2. Add 0.5–1 % (w/v) sucrose to provide a carbon source, and pH medium to 5.8 with 1 M KOH, and 0.5–1 % (w/v) agar (see Note 8).

230

Patrice A. Salomé et al.

Table 1 Comparison of MS-, minimal-, and Hoagland-based growth media for macro- and micronutrients (in mM) Component Macronutrient NH4NO3 CaCl2 Ca(NO3)2 Na2EDTA MgSO4 KNO3 KH2PO4 K2SO4 Micronutrient FeSO4 FeNaEDTA FeHBED MnSO4 CuSO4 H3BO3 ZnSO4 KCl KI CoCl2 (NH4)6Mo7O24

MS medium

1/2 MS

20.6 3

10.3 1.5

0.1 1.5 18.79 1.25

50 × 10−3 0.75 9.4 0.625

0.1

0.05

Minimal

Hoagland

2

1.5

0.65

0.75 1.25 0.28

0.1 0.75

0.05 0.1 0.1 × 10−3 0.1 29.91 × 10−3

50 × 10 0.05 × 10−3 50 × 10−3 14.95 × 10−3

0.1 × 10 0.05 × 10−3 10 × 10−3 0.05 × 10−3

5 × 10−3 5 × 10−3 0.5 × 10−3 25 × 10−3 5 × 10−3 50 × 10−3

5 × 10−3 0.11 × 10−3 1.03 × 10−3

2.5 × 10−3 0.055 × 10−3 0.5 × 10−3

5 × 10−6

0.1 × 10−3

Organic compounds Myo-inositol Glycine Nicotinic acid Pyridoxine HCl Thiamine

0.56 × 10−3 26.64 × 10−3 4.06 × 10−3 2.43 × 10−3 0.3 × 10−3

0.28 × 10−3 13.32 × 10−3 2.03 × 10−3 1.21 × 10−3 0.15 × 10−3

Selected components Ca2+ NH4+ NO3− SO42− K+a Na2 EDTA

3 21.6 39.4 1.73 20.1 0.1 0.1

1.5 10.8 19.7 0.865 10.05 0.05 0.05

2 5 × 10−6 2 1.4 0.85 0.05 0.05

1.5 0.1 × 10−3 2.75 2.01 1.58

−3

−3

Luciferase substrate is provided as a potassium salt. We routinely add 30 μl of 2.5 mM D-luciferin, K salt, to each well. With 200 μl of growth medium, this will bring in an additional 0.33 mM K

a

3. After autoclaving, add 100 μM ferroZine to the medium to chelate Fe for Fe-deficient medium (see Notes 9 and 10). 4. For medium deficient in Cu, add 10 μM BCS instead of ferro Zine; for medium deficient in Zn, add 50 μM TPEN instead of ferroZine (see Note 5).

Circadian Life Without Micronutrients…

3.2 Fe-Deficient Minimal Medium

231

Based on Hong et al. [10], Fe-deficient conditions are achieved by preparing a minimal growth medium lacking Fe, very much as for Fe-deficient MS medium above, although the composition of this minimal medium is much simpler [16, 18]. Strong Fe deficiency is imposed by further adding ferroZine to the growth medium, while Fe sufficiency conditions are obtained by adding 50 μM FeIIINaEDTA to the minimal medium. Seedlings are therefore exposed to Fe deficiency from the start of germination. 1. Mix all macro- and micro-nutrient solutions. 2. Add 0.5–1 % (w/v) sucrose to provide a carbon source. Add 0.05 % (w/v) MES and adjust pH to 5.8 with 1 M NaOH, and 0.5–1 % (w/v) agar (see Notes 4 and 8). 3. After autoclaving, add 100 μM ferroZine to the medium to chelate Fe (see Notes 9 and 10), or add 50 μM FeIIINaEDTA to reach Fe sufficiency (see Note 11).

3.3 Varying Micronutrient Supply in Hoagland-Based Medium Using EDTA-Washed Agar

In hydroponic test experiments, we observed that the addition of the chelators HEDTA or BCS generated symptoms in plants that were distinct from those induced by the mere omission of Zn or Cu, respectively, from hydroponic solutions (Ute Krämer, unpublished, ref. 19). The washed-agar approach for the cultivation of Arabidopsis under micronutrient transition metal deficiency conditions was thus adopted from prior work on Chlamydomonas, described in Quinn and Merchant [20]. In contrast to the previous two Fe-deficient growth media (Subheadings 3.1 and 3.2), our Hoagland-based medium recipe does not rely on chelation to render contaminant Fe unavailable to plant roots, but instead on the combination of a micronutrient solution lacking an Fe source with the use of EDTA-washed agar [20]. In addition, the Fe source here is FeIIIHBED (a complex of FeIII and N,N′-di(2-hydroxybenzyl) ethylenediamine-N,N′-diacetic acid) rather than FeIIINaEDTA [9, 21, 22] (see Note 12). The time investment for this medium is higher than for MS or minimal medium-based Fe deficiency studies, but offers a tighter control over Fe supply. 1. Weigh out 40 g of agar, and place in a 2-l conical glass flask filled with 1,000 ml of 5 mM EDTA, pH 8.0, using deionized water. 2. Stir agar suspension overnight at room temperature. 3. Remove EDTA solution by filtering agar over one layer of Miracloth inside a large funnel. Squeeze excess liquid gently, so as not to break the Miracloth. 4. Most agar will peel off easily from Miracloth and can be collected into the large funnel, placed over the 2-l glass flask. Remove remaining agar from Miracloth with deionized H2O, and resuspend agar in a final volume of 1,000 ml.

232

Patrice A. Salomé et al.

5. Perform two more EDTA washes for a total of three EDTA washes. 6. Resuspend agar in 1,000 ml deionized H2O, and stir overnight at room temperature. 7. Drain H2O by filtering agar suspension over one layer of Miracloth inside a large funnel. Squeeze excess liquid gently, so as not to break the Miracloth. 8. Repeat steps 6 and 7 two more times for a total of three H2O washes (see Note 13). 9. Drain as much H2O by filtering agar suspension over one layer of Miracloth inside a large funnel, taking care to squeeze as much H2O as possible. 10. Weigh final wet agar, and split among sixteen 50-ml Falcon tubes (when using 0.5 % (w:v) agar for solid medium, and assuming negligible loss of agar during the washing process). The consistency of the wet agar will be that of a soft solid. 3.4 Varying Fe Supply in HoaglandBased Medium

Our medium recipe relies on FeHBED as Fe source and is prepared according to Chaney [23], as described below. 1. Dissolve HBED with KOH, by adding 223 mg (or 0.525 mmol) HBED to 80 ml 15 mM KOH in deionized H2O. Stir solution to partly dissolve HBED (see Note 14). 2. Add 202 mg (or 0.5 mmol) ferric nitrate (Fe(NO3)3) and stir for 2–3 h. The solution should turn dark red. 3. Add drops of 1 M KOH to the solution to help HBED solubilization. Adjust final pH to 5.7 with KOH or diluted HNO3. 4. Make up to final volume of 100 ml. 5. As the stock solution is not sterile, FeHBED should either be added to Hoagland medium before autoclaving, or filtersterilized using a 0.2-μm filter. 6. Mix all macro- and micro-nutrient solutions (lacking Fe). 7. Add 0.5–1 % (w/v) sucrose to provide a carbon source, and adjust pH to 5.8 with the addition of 3 ml 1 M MES–KOH, pH 5.8 for a final concentration of 3 mM. Add one aliquot of washed agar (see Note 4). 8. Autoclave. 9. When pouring plates, calculate how many plates for each Fe concentration will be needed, and add FeHBED to Hoagland medium to reach desired Fe concentration. We routinely use 0.25 and 1 μM for Fe-limited conditions, 5 μM for Fe-deficient conditions, and 20 μM (sometimes up to 100 μM) for Fe-sufficient up to Fe-luxury conditions (see Note 15).

Circadian Life Without Micronutrients…

3.5 Varying Cu Supply in HoaglandBased Medium

233

To generate controlled Cu deficiency, we recommend pouring Cu-free Hoagland-based medium into acid-washed glass plates (see Notes 16 and 17). 1. Let glass dishes soak in 0.2 N HCl overnight. 2. The next day, rinse once with ultrapure H2O. 3. Rinse once more, with ultrapure H2O. 4. Autoclave dishes, wrapped individually, and use rapidly, as Cu trace contamination will reappear on the glass surface. 5. To prepare Cu-free medium, mix macro- and micro-nutrient (lacking Fe and Cu, or only Cu when adding 5 μM FeHBED) solutions. 6. Add 0.5–1 % (w/v) sucrose to provide a carbon source, and adjust pH to 5.8 with the addition of 3 ml 1 M MES–KOH, pH 5.8 for a final concentration of 3 mM. Add one aliquot of washed agar (see Note 4). 7. Autoclave. 8. When pouring plates, Cu sufficiency is obtained by adding CuSO4 to a final concentration of 0.5 μM. Cu toxicity is reached with concentrations as low as 1.5 μM CuSO4 (see Note 14).

3.6 Varying Zn Supply in Hoagland Medium

1. Mix all macro- and micro-nutrient (lacking Fe and Zn, or only Zn when adding 5 μM FeHBED) solutions. 2. Add 0.5–1 % (w/v) sucrose to provide a carbon source, and adjust pH to 5.8 with the addition of 3 ml 1 M MES–KOH, pH 5.8 for a final concentration of 3 mM. Add one aliquot of washed agar (see Note 4). 3. Autoclave. 4. When pouring plates, Zn sufficiency is obtained by adding ZnSO4 to a final concentration of between 1 and 5 μM. Slight Zn toxicity is reached with 30 μM ZnSO4.

3.7 Varying Mn Supply in Hoagland Medium

1. Mix all macro- and micro-nutrient (lacking Fe and Mn or only Mn when adding 5 μM FeHBED) solutions. 2. Add 0.5–1 % (w/v) sucrose to provide a carbon source, and adjust pH to 5.8 with the addition of 3 ml 1 M MES–KOH, pH 5.8 for a final concentration of 3 mM. Add one aliquot of washed agar (see Note 4). 3. Autoclave. 4. When pouring plates, Mn sufficiency is obtained by adding MnSO4 to a final concentration of 5 μM. Mn toxicity is reached with 100 μM MnSO4.

234

4

Patrice A. Salomé et al.

Notes 1. Fe can have two alternative oxidation states in biology: +2 (denoted here as FeII) and +3 (denoted here as FeIII). Ambient oxygen will result in the oxidation of FeII to FeIII. FeIII is far less soluble than FeII, but FeIII also forms complexes of higher stability than FeII in the presence of chelators. 2. In the absence of an added chelator, FeIII precipitates from nutrient solutions in the presence of oxygen. When preparing MS medium without added Fe, EDTA should therefore also be omitted because it will chelate other divalent or trivalent cations in the absence of equimolar concentrations of FeIII. 3. Media containing NO3− as sole N source will be alkalinized by the plant in the course of N assimilation, thus gradually increasing the pH (especially when not buffered). At pH > 6, FeIII hydroxides and phosphates will begin to precipitate (a kinetically slow process), even in the presence of the chelator EDTA. 4. Sucrose in the growth medium can affect the timing of processes under circadian control, such as growth [24]. Period lengthening under Fe-deficient conditions is maintained even when sucrose is omitted from the medium (Patrice A. Salomé, unpublished). The addition of sucrose supports uniform germination and growth across individuals, but should be kept at moderate concentrations. 5. TPEN is known to be highly membrane-permeable, a property that is commonly used in cell biology studies of non-plant systems. A non-membrane permeable option for Zn chelation is HEDTA, but it unspecifically chelates a range of divalent cations [25]. 6. Seedlings first grow for 8 days under micronutrient-sufficient conditions. The addition of a chelator is necessary to prevent uptake of any micronutrient following transfer to deficient conditions, but the underlying circadian phenotype will only become apparent after seedlings have exhausted their internal stores. A delay in the appearance of circadian phenotypes (such as the long period seen under low Fe conditions) is therefore to be expected, and was indeed observed by Chen et al. [11]. 7. One of the organic constituents of MS medium is glycine. In non-photosynthetic organisms, the tetrapyrrole precursor aminolaevulinic acid (ALA) is formed by a single-step condensation reaction between glycine and succinyl-CoA. In plants and algae, ALA is synthesized from glutamate in a multi-step process and therefore does not rely on glycine [26]. Nonetheless, glycine has been dropped from the growth medium commonly used by researchers working on the plastidto nucleus retrograde signaling. Because a retrograde signal from the plastid to the nucleus has been postulated [9, 11],

Circadian Life Without Micronutrients…

235

glycine may be omitted from the growth medium, as in Linsmaier and Skoog medium [27]. 8. Fe-deficient MS medium uses regular plant tissue culture-tested agar, and the total concentration of contaminant Fe in the agar will range from 2 to 12 μM depending on the agar supplier [28]. 9. Upon the addition of ferroZine to a medium containing Fe3+, the medium will turn magenta, the intensity of the color depending on the extent of Fe traces present in the agar. In contrast, the addition of ferroZine to Hoagland medium with washed agar does not lead to the formation of the characteristic magenta complex, consistent with the removal of trace Fe from the agar by the EDTA washing steps. 10. FerroZine is generally accepted to be membrane-impermeable, and its use in the quantification and in the scavenging of FeII is widespread. FerroZine can also bind FeIII, yet with far lower binding affinity, forming a similar magenta-colored complex of lower extinction coefficient [29]. FerroZine addition to growth media has often been employed to generate Fe deficiency in yeast and Arabidopsis, both of which reduce FeIII to FeII extracellularly prior to the binding of Fe2+ by the plasma membrane uptake systems. We found that ferroZine does not remain in the growth medium but can accumulate in plant tissues, when the growth medium contains some Fe3+ [9]. Regular agar will carry enough Fe contaminants to allow ferroZine to find its way into the plant. Interestingly, seedlings grown on medium containing ferroZine have phenotypes reminiscent of mutants affected in chlorophyll biosynthesis (such as gun4 or glk1glk2): low chlorophyll accumulation, similar pale yellow appearance of cotyledons and leaves [30, 31]. When grown on plates, irt1 seedlings (lacking high-affinity Fe transport) have more pronounced chlorosis with concomitant accumulation of anthocyanins. 11. EDTA is used as a FeIII chelator to prevent its precipitation, but will also form chelates with other metals, including Cu2+, Zn2+, and Ca2+, as well as Mg2+. The addition of a large excess of EDTA can therefore restrict the bioavailability of micronutrients, such as Cu2+ and Zn2+, and result in an effective deficiency of these metals. In fact, standard MS medium containing FeNaEDTA is commonly used to induce Cu deficiency [12]. 12. HBED (N,N′-di(2-hydroxybenzyl)ethylenediamine-N,N′diacetic acid) differs from both EDTA and EDDHA, which has also been used for Fe chelation in plant media, by showing only a limited affinity for Cu and by remaining chelated to FeIII over a wider pH range [23]. It can be purchased from Strem Chemicals, catalog number 07-0422). 13. Any EDTA remaining in the agar as a result of insufficient washing will reduce the availability of multiple divalent cations,

236

Patrice A. Salomé et al.

most prominently of Cu2+. Millimolar concentrations of free EDTA have been observed to cause toxicity in plants [32]. 14. HBED is provided in a 5 % excess over the final Fe concentration to avoid large changes in (initially free) Fe3+ concentrations from small weighing errors (following Sébastien Thomine, Institut des Sciences du Végétal, CNRS, Gif-sur-Yvette, personal communication). 15. Rather than preparing several bottles of Hoagland medium and adding FeHBED to reach the desired Fe concentration, we make up our Hoagland medium omitting Fe, and add FeHBED to 50 ml Falcon tubes immediately before pouring plates. 16. Cu deficiency is more clearly seen when seedlings are grown on acid-washed glass dishes when compared to plastic dishes, based on the phenotypes of Col-0 and spl7 mutant [33]. Copper or other transition metal cations are often used to support polymerization in plastic production, and will subsequently contaminate solutions in contact with the surface. Depending on the manufacturing process employed, this can cause problems or variability in generating Zn or Cu deficiency. Copper ions will also be present on glass surfaces, but can be removed with an acid wash. 17. We and others [9, 11] did not observe changes in circadian parameters when Col-0 seedlings were grown under Cu-deficient conditions. We have not tested circadian phenotypes of lines overexpressing the copper transporters COPT1 or COPT3 [12].

Acknowledgments We thank Sabeeha Merchant (UCLA), Sébastien Thomine (Institut des Sciences du Végétal, CNRS, Gif-sur-Yvette), and Stephan Clemens (Universität Bayreuth) for stimulating discussions about methods to generate metal deficiencies without resorting to the use of chelators. This work was supported by an EMBO long-term fellowship (PAS), German Research Foundation Heisenberg Fellowship DFG Kr1967/4-1 and grant DFG Kr1967/5-1 (M.B. and U.K.). P.A.S. also thanks Detlef Weigel for support from the Max Planck Society. References 1. Kramer U (2010) Metal hyperaccumulation in plants. Annu Rev Plant Biol 61:517–534 2. Sinclair SA, Kramer U (2012) The zinc homeostasis network of land plants. Biochim Biophys Acta 1823(9):1553–1567

3. Palmer CM, Guerinot ML (2009) Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nat Chem Biol 5(5):333–340 4. Tottey S, Block MA, Allen M, Westergren T, Albrieux C, Scheller HV, Merchant S, Jensen PE

Circadian Life Without Micronutrients…

5.

6.

7.

8.

9.

10.

11.

12.

13. 14.

15. 16.

17.

(2003) Arabidopsis CHL27, located in both envelope and thylakoid membranes, is required for the synthesis of protochlorophyllide. Proc Natl Acad Sci U S A 100(26):16119–16124 Dodd AN, Salathia N, Hall A, Kevei E, Toth R, Nagy F, Hibberd JM, Millar AJ, Webb AAR (2005) Plant circadian clocks increase photosynthesis, growth, survival, and competitive advantage. Science 309:630–633 Graf A, Schlereth A, Stitt M, Smith AM (2010) Circadian control of carbohydrate availability for growth in Arabidopsis plants at night. Proc Natl Acad Sci U S A 107(20):9458–9463 Wang W, Barnaby JY, Tada Y, Li H, Tor M, Caldelari D, Lee DU, Fu XD, Dong X (2011) Timing of plant immune responses by a central circadian regulator. Nature 470(7332):110–114 Goodspeed D, Chehab EW, Min-Venditti A, Braam J, Covington MF (2012) Arabidopsis synchronizes jasmonate-mediated defense with insect circadian behavior. Proc Natl Acad Sci U S A 109(12):4674–4677 Salomé PA, Oliva M, Weigel D, Kramer U (2013) Circadian clock adjustment to plant iron status depends on chloroplast and phytochrome function. EMBO J 32(4):511–523 Hong S, Kim SA, Guerinot ML, McClung CR (2013) Reciprocal interaction of the circadian clock with the iron homeostasis network in Arabidopsis. Plant Physiol 161(2):893–903 Chen YY, Wang Y, Shin LJ, Wu JF, Shanmugam V, Tsednee M, Lo JC, Chen CC, Wu SH, Yeh KC (2013) Iron is involved in the maintenance of circadian period length in Arabidopsis. Plant Physiol 161(3):1409–1420 Andres-Colas N, Perea-Garcia A, Puig S, Penarrubia L (2012) Deregulated copper transport affects Arabidopsis development especially in the absence of environmental cycles. Plant Physiol 153(1):170–184 Wilson GT, Connolly EL (2013) Running a little late: chloroplast Fe status and the circadian clock. EMBO J 32(4):490–492 Kim WY, Salome PA, Fujiwara S, Somers DE, McClung CR (2010) Characterization of pseudo-response regulators in plants. Methods Enzymol 471:357–378 Murashige T, Skoog F (1962) A revised medium for rapid growth and bio-assays with tobacco tissue cultures. Physiol Plant 15(3):473–497 Yi Y, Guerinot ML (1996) Genetic evidence that induction of root Fe(III) chelate reductase activity is necessary for iron uptake under iron deficiency. Plant J 10(5):835–844 Becher M, Talke IN, Krall L, Kramer U (2004) Cross-species microarray transcript profiling reveals high constitutive expression of metal

18.

19.

20. 21.

22.

23.

24.

25.

26.

27. 28.

237

homeostasis genes in shoots of the zinc hyperaccumulator Arabidopsis halleri. Plant J 37(2): 251–268 Marschner H, Römheld V, OssenbergNeuhaus H (1982) Rapid method for measuring changes in pH and reducing processes along roots of intact plants. Z Pflanzenphysiol 105:407–416 Talke IN, Hanikenne M, Kramer U (2006) Zinc-dependent global transcriptional control, transcriptional deregulation, and higher gene copy number for genes in metal homeostasis of the hyperaccumulator Arabidopsis halleri. Plant Physiol 142(1):148–167 Quinn JM, Merchant S (1998) Copper-responsive gene expression during adaptation to copper deficiency. Methods Enzymol 297:263–279 Tennstedt P, Peisker D, Bottcher C, Trampczynska A, Clemens S (2009) Phytochelatin synthesis is essential for the detoxification of excess zinc and contributes significantly to the accumulation of zinc. Plant Physiol 149(2):938–948 Lanquar V, Lelievre F, Bolte S, Hames C, Alcon C, Neumann D, Vansuyt G, Curie C, Schroder A, Kramer U, Barbier-Brygoo H, Thomine S (2005) Mobilization of vacuolar iron by AtNRAMP3 and AtNRAMP4 is essential for seed germination on low iron. EMBO J 24(23):4041–4051 Chaney RL (1988) Plants can utilize iron from Fe-N, N′-di-(2-hydroxybenzoyl)-ethylenediamine-N, N′-diacetic acid, a ferric chelate with 106 greater formation constant than FE-EDDHA. J Plant Nutr 11:1033–1050 Stewart JL, Maloof JN, Nemhauser JL (2011) PIF genes mediate the effect of sucrose on seedling growth dynamics. PLoS One 6(5): e19894 Wintz H, Fox T, Wu YY, Feng V, Chen W, Chang HS, Zhu T, Vulpe C (2003) Expression profiles of Arabidopsis thaliana in mineral deficiencies reveal novel transporters involved in metal homeostasis. J Biol Chem 278(48): 47644–47653 Tanaka R, Kobayashi K, Masuda T (2011) Tetrapyrrole metabolism in Arabidopsis thaliana. The Arabidopsis book. Am Soc Plant Biol 9:e0145 Linsmaier EM, Skoog F (1965) Organic growth factors requirements of tobacco tissue culture. Physiol Plant 18:100–127 Jain A, Poling MD, Smith AP, Nagarajan VK, Lahner B, Meagher RB, Raghothama KG (2009) Variations in the composition of gelling agents affect morphophysiological and molecular responses to deficiencies of phosphate

238

Patrice A. Salomé et al.

and other nutrients. Plant Physiol 150(2): 1033–1049 29. Viollier E, Inglett PW, Hunter K, Roychoudhury AN, van Cappellen P (2000) The ferrozine method revisited: Fe(II)/ Fe(III) determination in natural waters. Appl Geochem 15(6):785–790(6) 30. Larkin RM, Alonso JM, Ecker JR, Chory J (2003) GUN4, a regulator of chlorophyll synthesis and intracellular signaling. Science 299(5608):902–906 31. Waters MT, Wang P, Korkaric M, Capper RG, Saunders NJ, Langdale JA (2009) GLK transcription factors coordinate expression of the

photosynthetic apparatus in Arabidopsis. Plant Cell 21(4):1109–1128 32. Vassil AD, Kapulnik Y, Raskin II, Salt DE (1998) The role of EDTA in lead transport and accumulation by Indian mustard. Plant Physiol 117(2):447–453 33. Bernal M, Casero D, Singh V, Wilson GT, Grande A, Yang H, Dodani SC, Pellegrini M, Huijser P, Connolly EL, Merchant SS, Kramer U (2012) Transcriptome sequencing identifies SPL7-regulated copper acquisition genes FRO4/FRO5 and the copper dependence of iron homeostasis in Arabidopsis. Plant Cell 24(2):738–761

Chapter 17 Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity Katharina König, Helena Galliardt, Marten Moore, Patrick Treffon, Thorsten Seidel, and Karl-Josef Dietz Abstract Redox homeostasis is an important parameter of cell function and cell signaling. Spatial and temporal alterations of redox state control metabolism, developmental processes, as well as acute responses to environmental stresses and stress acclimation. Redox homeostasis is also linked to the circadian clock. This chapter introduces methods to assess important redox parameters such as the low molecular weight redox metabolites glutathione and ascorbate, their amount and redox state, and H2O2 as reactive oxygen species. In vivo redox cell imaging is described by use of the reduction–oxidation sensitive green fluorescent protein (roGFP). Finally, on the level of posttranslational redox modifications of proteins, methods are shown to assess hyperoxidation of 2-cysteine peroxiredoxin and glutathionylation of peroxiredoxin IIE. The redox state of 2-cysteine peroxiredoxin has been identified as a transcription-independent marker of circadian rhythmicity. Key words Ascorbate, Glutathione, Glutathionylation, Hydrogen peroxide, Peroxiredoxin, Redox signaling, roGFP

1

Introduction The circadian clock controls the redox state of the cell. This intimate relationship between the time of day and the cellular redox state is well established for insects and vertebrates and increasingly for plants [1–3]. Work with Arabidopsis thaliana mutants defective in elements of the circadian clock such as CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) revealed that the circadian clock regulates the daytime-specific adjustment of the cell redox state [4]. Several hallmarks define the redox state of the cell. Among these are the amount and the redox state of the ascorbate and glutathione systems, the H2O2 levels and the oxidation state of the 2-cysteine peroxiredoxin (2-CysPrx). Plant 2-CysPrx adopts five different conformation states in dependence on the redox state of its catalytic cysteinyl residues [5]. The occurrence of the

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_17, © Springer Science+Business Media New York 2014

239

240

Katharina König et al.

Circadian rhythm

H2O2

Asc/DHA Redox

GSH/GSSG

ROS

roGFP

Peroxiredoxin

Glutathionylation

Fig. 1 Schematics depicting the relationship between the redox/ROS state of the cell and the measured parameters in this method chapter. On the left hand site, ascorbate and glutathione define part of the cellular redox milieu which can also be assessed in vivo by in vivo redox imaging. 2-CysPrx hyperoxidation and PrxIIE glutathionylation link redox and ROS. H2O2 is considered to be involved in ROS stress and redox signaling. The circadian clock appears to affect many of these parameters. Asc—ascorbate, DHA—dehydroascorbate, GSH—reduced glutathione, GSSG—oxidized glutathione

hyperoxidized sulfinic acid form of 2-CysPrx recently was described as a transcription- and translation-independent readout of circadian rhythmicity [6]. These biochemical indicators of the redox and reactive oxygen species (ROS) state of plant cells can be assessed ex vivo by biochemical and in situ by cell biological analyses. To this end this chapter introduces into methods that allow to quantify (a) H2O2 levels, the amount and redox state of (b) glutathione and (c) ascorbate in extracts, the (d) redox state of the glutathione system in vivo and two methods to test for the redox state of target proteins ex vivo, (e) the redox state of the chloroplast 2-cysteine peroxiredoxin, and (f) glutathionylation of proteins using the example of peroxiredoxin IIE. Figure 1 gives a schematic overview of the interdependency of the six parameters described in this chapter.

2

Materials All solutions are prepared with high-quality ultrapure water and all chemicals are of reagent grade and used without further purification

2.1 Quantification of H2O2 in Extracts by Luminol

1. 0.1 M Sodium carbonate (Na2CO3) solution in water. 2. 0.1 M Sodium bicarbonate (NaHCO3) solution in water. 3. 0.1 M Sodium carbonate buffer solution pH 10.2 (see Note 1).

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

241

4. Luminol stock solution is prepared by adding 11.5 mg of luminol to 10 mL of 0.1 M sodium carbonate buffer solution pH 10.2. 5. Co (II) stock solution is prepared by dissolving 7.14 mg cobalt (II) chloride in 0.1 M sodium carbonate buffer solution pH 10.2. 6. Mixed reagent stock solution is prepared by diluting 10 mL luminol stock solution and 2 mL Co (II) stock solution (see Note 2) to 100 mL with 0.1 M sodium carbonate buffer solution. 7. Working solution was prepared by diluting the mixed reagent stock solution tenfold (see Note 3). 8. Hydrogen peroxide (H2O2) standard solutions: 1, 2, 5, 10, 15 μM in sodium carbonate buffer solution pH 10.2. 9. Trichloroacetic acid (TCA) solution 5 % (v/v) in 0.1 M sodium carbonate. The solution should be stored in a dark bottle at 4 °C (see Note 4). 10. Catalase (bovine liver, Sigma, USA). 11. Incubator for microfuge tubes at 30 °C. 12. Luminometer (e.g., Sirius L Tube (Berthold Detection Systems, Germany)). 13. Plant material was harvested prior to experiments in liquid N2 and afterwards stored at −80 °C until further use. 2.2 Measurement of Reduced and Oxidized Ascorbate

1. 0.1 M Perchloric acid (HClO4). 2. 0.1 M HEPES–KOH pH 7.0. 3. 5 M Potassium carbonate. 4. 0.1 M Sodium phosphate buffer (Na-Pi) pH 5.6 (see Note 9). 5. 0.1 M Sodium phosphate buffer (Na-Pi) pH 7.5 (see Note 9). 6. 1 M Dithiothreitol (DTT), freshly prepared. 7. Ascorbate oxidase 1 U/μL (AppliChem ascorbate oxidase 1,000 U; A3377) (see Note 10). 8. Standard Solution: 5 mM ascorbic acid in 0.1 M HClO4, 5 mM dehydroascorbate (DHA) in 0.1 M HClO4. 9. Quartz cuvettes. 10. UV–Vis spectrophotometer capable of performing kinetic measurements, e.g., Varian Cary 300 (Agilent Technologies, Böblingen, Germany).

2.3 Enzymatic Analysis of Reduced and Oxidized Glutathione

1. Sample buffer: 0.1 N HCl, 1 mM EDTA. 2. Test buffer: 120 mM sodium phosphate buffer (Na-Pi) pH 7.8, 6 mM EDTA (see Note 13). 3. 6 mM DTNB (2,2′-Dinitro-5′5-dithiodibenzoic acid) dissolved in test buffer (see Note 14).

242

Katharina König et al.

4. Freshly prepared 10 mM NADPH dissolved in test buffer. 5. Glutathione-reductase (sigma G3664) 2.4 mg/mL (2.5 kU); dilute 1:50 in test buffer (see Note 15). 6. Freshly prepared glutathione standard solutions: 100 mM GSH in sample buffer and 50 mM GSSG in sample buffer (see Note 16). 7. 2-Vinylpyridine (see Note 17). 8. 5 M K2CO3 or 1 M NaOH for neutralization. 9. Half micro cuvettes. 10. UV–Vis spectrophotometer capable of performing kinetic measurements, e.g., Varian Cary 300 (Agilent Technologies, Böblingen, Germany). 2.4 Detection of Glutathione In Vivo by Redox-Sensitive Green Fluorescent Protein (roGFP)

1. 1 M Dithiothreitol (DTT) stock solution, prepared fresh. 2. 1 M Hydrogen peroxide stock solution (H2O2). 3. W5-medium for single cell analysis in protoplasts: 2 mM MES pH 5.7, 154 mM NaCl, 12 mM CaCl2, 5 mM glucose, 5 mM KCl. 4. MS-medium for analysis in plants. 5. Plants expressing roGFPs: Plants expressing the following roGFP constructs were published: cyt-roGFP1 (cytosolic), mitroGFP1 (mitochondrial), per-roGFP1 (peroxisomal) [7], plaroGFP2 (plastids [8]), and cyt-Grx1-roGFP2 (cytosolic [9]). 6. Technical requirements for conventional fluorescence microscopy: filter sets for excitation of EGFP and T-Sapphire. Both should allow detection of roGFP2 in the range of 500–530 nm, such as band pass filter BP515/30. For excitation, one allows for excitation of the neutral form of wild-type GFP (e.g., BP 395/30) and the other for excitation of the anionic form of wild-type GFP (e.g., BP475/30). Finally, a dichroic mirror should separate excitation and emission at 500 nm. 7. Prerequisites for analysis by confocal laser scanning microscopy: Laser lines 405 and 488 nm and one detector, e.g., a photomultiplier (PMT). Main beam splitter/dichroic mirror for excitation at 488 and 405 nm.

2.5 Oxidation State of 2-Cysteine Peroxiredoxin 2.5.1 Protein Extraction from A. thaliana Plant Material

1. Extraction buffer: 0.5 M Tris–HCl, pH 6.8. Store at room temperature. 2. Extraction buffer for oxidized conditions: Add 50 μM H2O2 to extraction buffer. 3. Hydrogen peroxide (H2O2) stock solution: 100 mM in water. Prepare fresh.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity 2.5.2 Protein Quantification Assay According to Bradford

243

1. Coomassie Brilliant Blue Dye-G250 solution, e.g., Roti-Quant reagent (Roth, Karlsruhe, Germany). 2. Bovine serum albumin (BSA) stock solution for standardization: 1 mg in 1 mL water. Store at −20 °C. 3. Microplate autoreader, e.g., PowerWave™ 200 (BioTek Instruments, Bad Friedrichshall, Germany) and software for processing, e.g., KC4™ software.

2.5.3 Discontinuous Reducing and Non-reducing SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

1. Separating gel buffer: 1.5 M Tris–HCl, pH 8.8. Store at 4 °C. 2. Stacking gel buffer: 0.5 M Tris–HCl, pH 6.8. Store at 4 °C. 3. Acrylamide solution: 30 % (w/v) Acrylamide, 0.8 % (w/v) bisacrylamide (Roth, Karlsruhe, Germany) (see Note 24). 4. N,N,N′,N′-Tetramethylethylenediamine (TEMED) (SigmaAldrich, Seelze, Germany), store at 4 °C. 5. 10 % Ammonium persulfate (APS) solution: dissolve 100 mg in 1 mL water. Prepare fresh and store at –20 °C. 6. 100 % Isopropanol – to overlay the separating gel to ensure a flat surface and to exclude oxygen during polymerization. 7. 10× Running buffer: 125 mM Tris, 960 mM glycine, 0.5 % (w/v) SDS. Store at room temperature. For 1× running buffer, add 100 mL of 10× running buffer to 900 mL ultrapure H2O. 8. Prestained molecular weight marker: e.g., PageRuler™ Prestained Protein Ladder (Fermentas, St. Leon-Rot, Germany). 9. SDS electrophoresis system: chamber, comb, set of glass plates, spacer, rubber gum, clamps, power supply, cables. A Hamilton syringe to apply the samples and to wash the wells after removing the comb. 10. 5× Loading buffer (without reducing agent): 225 mM Tris– HCl pH 6.8, 5 % (w/v) SDS, 50 % glycerol, 0.05 % bromophenol blue. Store at room temperature. 11. 1 M Dithiothreitol (DTT) should be prepared fresh in aqueous buffer of pH 7.0 (see Note 25). 12. 5× Loading buffer + DTT: 225 mM Tris–HCl pH 6.8, 5 % (w/v) SDS, 50 % glycerol, 0.05 % bromophenol blue, 50 mM DTT. Store at room temperature.

2.5.4 Western Blot Analysis: Semidry Western Transfer

1. Nitrocellulose membrane (e.g., Protran® nitrocellulose membrane, GE Healthcare, Munich, Germany). 2. Blotting paper 550 g/m2 (e.g., Hartenstein, Würzburg, Germany). 3. 10× Anode buffer: 300 mM Tris–NaOH pH 10.4, 20 % (v/v) methanol. Store at 4 °C. For 1× anode buffer, add 100 mL 10× anode buffer to 900 mL ultrapure H2O.

244

Katharina König et al.

4. Cathode buffer: 25 mM Tris–HCl (pH 9.4), 40 mM ε-aminocapronic acid, 20 % methanol. Store at 4 °C. 5. Semidry blotter, e.g., Fastblot B44 (Biometra, Göttingen, Germany). 2.5.5

Ponceau S Staining

1. Ponceau S solution: 0.2 % (w/v) Ponceau S (3-hydroxy-4-[2sulfo-4-(sulfophenylazo) phenylazo]-2, 7-naphtalene disulfonic acid) (AppliChem, Darmstadt, Germany), 5 % (v/v) acetic acid, make up to 100 mL with water. 2. 10× TBS buffer: 0.5 M Tris–HCl (pH 7.5), 1.5 M NaCl. For 1× TBST buffer, add 100 mL 10× TBST buffer to 900 mL ultrapure H2O and 0.05 % (v/v) Tween 20. Store at room temperature.

2.5.6 Immunodetection of Immobilized Proteins with Specific Antibodies and Chemiluminescence

1. Blocking solution: 1 % (w/v) milk powder in 1× TBS, freshly prepared (see Note 26). 2. Primary antibody: 2-Cys Prx specific antibody from rabbit diluted 1:2,000 in blocking solution. Store at −20 °C (see Note 27). 3. Secondary antibody: goat anti-rabbit antibody with conjugated horseradish peroxidase (Sigma Aldrich, St. Louis, USA) diluted 1:3,000 in blocking solution. Store at −20 °C (see Note 28). 4. Lumi-light Luminol substrate, e.g., Supersignal® West Pico Chemiluminescence Substrate (Thermo Scientific, USA). The Lumi-light luminol substrate should consist of one bottle Lumi-Light Luminol/Enhancer solution and of one bottle Lumi-Light Stable Peroxide solution. 5. X-Ray films, e.g., X-Ray XBA (Photochemische Werke, Germany). 6. Developer solution and fixation solution (Kodak GBX). 7. Glass plate. 8. Commercial plastic wrap/wrapping film. 9. Software for analyzing and quantification of bands, e.g., GelScan (BioSciTec, Frankfurt/Main, Germany).

2.6 S-Glutathionylation of Proteins by ESI-MS

1. Dithiothreitol (DTT) stock solution: 1 M in 0.1 M Tris–HCl, pH 8.0 (see Note 32).

2.6.1 Chemical Reduction of Proteins

3. Incubator for microfuge tubes set at 25 °C.

2.6.2 Removal of Excess DTT

1. PD-10 desalting columns (GE Healthcare, Munich, Germany).

2. Sample buffer for glutathionylation: 0.1 M Tris–HCl, pH 8.0.

2. Coomassie Brilliant Blue Dye-G250 solution, e.g., Roti-Quant (Roth, Karlsruhe, Germany).

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

245

3. 5,5'-Dithiobis-(2-nitrobenzoic acid) DTNB: 6 mM prepared fresh in sample buffer (see Note 33). 4. 96-Well microtiter plates. 2.6.3 Reaction of Reduced Thiol Proteins with the Disulfide Form of Glutathione 2.6.4 Mass Spectrometric Analysis (ESI-MS)

1. Oxidized glutathione (GSSG): 50 mM stock prepared directly in buffer just prior to use. Working solutions are prepared by further dilutions in sample buffer (see Note 34). 2. 80 % Acetone (v/v). 1. 30 % Ethanol (EtOH) (v/v). 2. 100 % Formic acid. 3. ESI-MS mass spectrometer: e.g., Esquire 3000 plus quadrupole ion trap mass spectrometer (Bruker Daltonics, Bremen, Germany). 4. Software that uses the MS data to obtain the molecular weights of the proteins via deconvolution, e.g., DataAnalysis Version 3.2 or higher (Bruker Daltonics, Bremen, Germany).

3

Methods

3.1 H2O2Quantification in Extracts by Luminol

Hydrogen peroxide (H2O2) functions as oxidant that may cause oxidative damage to cell constituents, but it also acts as key signaling molecule generated by plants in response to abiotic and biotic stresses [10]. Especially in green tissues H2O2 is generated at high rates in chloroplasts [11] and mitochondria [12] via electron transport during photosynthesis and respiration. Compared to other ROS H2O2 is characterized as moderately reactive and hence rather long-lived molecule which can diffuse across membranes and react with proteins by oxidizing their thiol groups [13]. Therefore H2O2 plays an important role in signal transmission and cellular regulation. Thus, precise and efficient methods to determine H2O2 levels in different plant tissues are of great importance. There are various methods commonly used for H2O2 level determination like fluorimetry [14], spectrophotometry [15], and chemiluminescence [16]. This chapter introduces the advanced chemiluminescence (CL) method suggested by Perez and Rubio [17]. This detection method uses luminol as reaction partner of H2O2 according to the following non-stochiometric equation: Luminol + H 2O2 → 3-aminophthalate* → 3-aminophthalate + light

H2O2 reduces luminol to 3-aminophthalate1 resulting in the excited singlet state S1. The transition of 3-aminophthalate* to the lower energetic level S0 occurs by emitting light which can be 1

*

Molecule in the excited singlet state S1.

246

Katharina König et al.

detected at 425 nm. This method is highly sensitive and allows detection of H2O2 levels on the physiological reasonable scale. 3.1.1 Sample Preparation

1. Plant material is ground to fine powder in liquid N2. Transfer approximately 100 mg powder to 1.5 mL reaction tubes. Note down the exact mass of the material that you weighed in for use in the calculation (see Note 5). 2. Homogenize the material with 0.5 mL of 5 % TCA solution at 4 °C using micropistils. 3. Invert the samples for 40 min in the dark at 4 °C (see Note 6). 4. Centrifuge the samples for 15 min at 10,000 × g and 4 °C to sediment all proteins (see Note 7). 5. Transfer 50 μL of the supernatant into a new reaction tube. 6. Dilute samples 1,000-fold for measurement in sodium carbonate buffer. 7. Prepare two 1.5 mL reaction tubes per sample and add 20 μL of the dilution (see step 6) to each tube. (a) Add 5 μL (50 U) catalase to the first reaction tube. (b) Add 5 μL H2O to the second reaction tube. 8. Incubate in the dark for 15 min at 30 °C.

3.1.2 Measurement of Samples

1. As blank use 1 mL sodium carbonate buffer solution pH 10.2 in a 1.5 mL reaction tube. 2. Prepare ten reaction tubes per sample with 998 μL working solution. 3. Add 2 μL of catalase-treated or untreated sample, then homogenize and immediately quantify the emitted chemiluminescence (CL). 4. The emitted photons of the samples are measured with a luminometer in suitable tubes (e.g., 1.5 mL reaction tubes) at 425 nm for 5–8 s (see Note 8). 5. The difference between catalase-treated and untreated samples (ΔCL) is considered as H2O2-specific CL. 6. For standardization 2 μL of the H2O2 standards containing 0, 1, 2, 5, 10, and 15 μM H2O2 are measured in 998 μL working solution in 1.5 mL reaction tubes. Calculate with the resulting standard curve (see Fig. 2) the H2O2 content of the samples. Especially under changing environmental conditions or during the circadian rhythm strong changes in signal transduction occur. The afore described method allows a very precise measurement of the H2O2 content in various plant tissues and therefore can be used for a characterization of H2O2-based signal transduction.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

247

30x104

RCLU

20x104

10x104

0 0

10

20

30

40

H2O2 [nmol/µl]

Fig. 2 Calibration curve for luminol-based H2O2 determination. See text for details

3.2 Measurement of Reduced and Oxidized Ascorbate

3.2.1 Preparation of the Standard Samples

Ascorbate (ASC, vitamin C) represents a multifunctional metabolite that plays an important role in redox balancing [18–21]. Abiotic stresses generally result in an imbalance of redox status, leading to an increase in ROS such as hydrogen peroxide (H2O2), superoxide anion (O2·−), and hydroxyl radical (OH·). Facilitated by their high intracellular concentrations two molecules ASC detoxify H2O2 to water, either spontaneously or catalyzed by ascorbate peroxidase, thereby generating monodehydroascorbate (MDHA). Two molecules of MDHA disproportionate to ascorbate and DHA. The method described here enables the measurement of reduced ASC and the oxidized form DHA of ascorbate, allowing the determination of the redox ratio between these forms after, e.g., stress treatment or in context of circadian rhythm. Here the method of Foyer et al. [22] is described with slight modifications. 1. Mix 400 μL 5 mM ascorbic acid or 5 mM DHA with 200 μL pre-cooled HEPES buffer. 2. Adjust the solution to pH 5–6 with potassium carbonate. Be very careful when adjusting the pH and start out with small volumes of K2CO3 solution (e.g., 2 μL). Generally keep the added volume as small as possible to minimize dilution effects. 3. Centrifuge for 5 min at 16,000 × g, 4 °C.

3.2.2 Preparation of Plant Extract

1. Grind 100 mg of leaf material (fresh weight: FW) in liquid nitrogen (see Note 11). 2. Add 600 μL ice cold perchloric acid, grind further. 3. Transfer the solution into a 1.5 mL tube. 4. Centrifuge for 5 min at 16,000 × g, 4 °C.

248

Katharina König et al.

5. Transfer 400 μL of the supernatant to a new tube. 6. Add 200 μL HEPES–KOH-buffer. 7. Adjust the solution to pH 5–6 with potassium carbonate. Be very careful when adjusting the pH. Typically a small volume (3 μL) is sufficient. Generally keep the added volume as small as possible to minimize dilution effects. Note the volume added as VK. 8. For the removal of the precipitated potassium perchlorate centrifuge for 5 min at 16,000 × g, 4 °C and transfer the supernatant (neutralized plant extract) into a new tube. 3.2.3 Sample Preparation for Reduced Ascorbate Content

1. Mix 200 μL of the neutralized plant extract (or neutralized ascorbic acid standard) with 200 μL pre-cooled 0.1 M Na-Pi, pH 7.5. 2. Add 31.8 μL H2O. 3. Incubate on ice for 30 min.

3.2.4 Sample Preparation for Total Ascorbate Content

3.2.5 Photometric Measurement

1. Mix 200 μL of the neutralized plant extract (or neutralized DHA standard) with 200 μL pre-cooled 0.1 M Na-Pi, pH 7.5. 2. Add 31.8 μL 1 M DTT (see Note 12). 3. Incubate on ice for 30 min. 1. The absorption is taken at 265 nm. 2. Mix 150 μL sample with 850 μL 0.1 M Na-Pi, pH 5.6 in a quartz cuvette. 3. Record absorption for 2 min with measurements taken at least every 0.5 min to get a baseline. 4. Add 5 U of ascorbate oxidase. 5. Measure the decrease of absorption until no change at Abs265nm is detectable. 6. It is recommended to make at least triplicate determinations.

3.2.6 Calculation of the Content of Reduced and Oxidized Ascorbate

1. Calculate the final ASC or DHA concentrations in your standard samples. 2. Determine the Δ absorbance (ΔAbs) of your standard measurements out of the initial absorbance and the final absorbance (Fig. 3). 3. Plot ΔAbs against ASC/DHA amount in nmol to get the calibration lines. 4. Calculate the regression line for ASC and DHA, respectively (Fig. 4). 5. Estimate ΔAbs for total ASC (ΔAbsASCt) and reduced ASC (ΔAbsASCr) as denoted in step 2.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

249

+ Ascorbate oxidase

Absorbance [265nm]

1

∆AbsASCt

0.5 ∆AbsASCr

0 0

1

2

3

4

5

6

7

8

Time [min]

Fig. 3 Time-dependent kinetics of ΔAbs in a typical assay for ASC determination. Upon addition of ascorbate oxidase, H2O2 is rapidly consumed and the absorbance drops to a new steady state. The difference between absorbance prior to ascorbate oxidase addition and at the end of the reaction is equivalent to the H2O2 in the test

∆Abs265 nm

2

Ascorbate, reduced y=0.0111 x r2=0.999 1 Dehydroascorbate y=0.0065 x r2=0.982 0 0

40

80

120

160

Ascorbate or dehydroascorbate [nmol]

Fig. 4 Calibration curve for reduced ascorbate and dehydroascorbate represented as absorbance change at 265 nm. The calibration curve was linear up to a concentration of 160 nmol/L. r2 – correlation coefficient

6. Obtain the ΔAbs for DHA (ΔAbsDHA) according to the following equation: ∆ AbsDHA = ∆ Abs ASCt − ∆ Abs ASCr .

250

Katharina König et al.

7. Derive the amounts of reduced ASC and DHA, respectively, in the cuvette sample from the calibration equation from step 4. 8. To calculate the concentration of ASC or DHA per g fresh weight (FW) you also have to consider the steps that were made during the extraction and sample preparation according to the following equation exemplified here. nmol ASC / DHA nmol ASCS / DHA S 600 µ l + 90 µ l 400 µ l + 200 µ l + V K µ l = x x gFW 0.1g FW 400 µ l 200 µ l x =

200 µ l + 200 µ l + 31.8µ l 150 µ l

nmol ASCS / DHA S x (14.90 + 0.025xV K ) . 0.1g FW

Dehydroascorbate concentration is determined indirectly by this method. Therefore one has to obtain the total ascorbate content after non-enzymatic reduction using DTT. Subtraction of reduced ascorbate from total ASC finally gives the DHA content. The proposed method for the enzymatic determination of ascorbate as well as dehydroascorbate described here is simple, sensitive, and precise. Thus, the results obtained by this technique can be applied to different topics, e.g., after stress treatment, during development of different cell types or in context of circadian rhythm. 3.3 Enzymatic Analysis of Reduced and Oxidized Glutathione

Glutathione is the predominant non-enzymatic thiol antioxidant in plants, and participates not only in the detoxification of ROS but also in many other reactions including glutathionylation of proteins as discussed below. It is present in either reduced (GSH) or oxidized form (GSSG), but only GSH can donate electrons. The oxidized form GSSG functions in signal transduction, e.g., in posttranslational modification of proteins [23]. In order to estimate the cellular thiol redox state, one has to measure not only the total glutathione content (2[GSH] + [GSSG]) but also the ratio of reduced to oxidized glutathione. It is important to note that the glutathione content of an extract only gives the average over all cell types and compartments. However, the concentrations of [GSH] and [GSSG] and redox states of the [GSH][GSH]/[GSSG] system vary considerably among cells and their cellular compartments. The enzymatic method for measuring GSH and GSSG was developed by Griffith in 1980 [24] and was later on adapted for the use in plant extracts [25–27]. To measure the content of glutathione the thiol-reagent 2,2′-dinitro-5′5-dithiodibenzoic acid (DTNB) is used. DTNB reacts with GSH to a colored product that can be measured photometrically. In this reaction GSH is oxidized to GSSG. By adding glutathione reductase and NADPH a re-reduction cycle of GSSG

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

a

b

Total glutathione content

251

GSSG content GS-vinylpyridin GSH

GSH GSSG

+ DTNB

GSSG

+ red. DTNB

+ Vinylpyridine

GSSG

+ Vinylpyridine

GSH

GR

GSSG

+ DTNB

GR

GSH

+ Vinylpyridine

DTNB

GR

GSSG

DTNB

GR

GSSG

GSSG

GSSG

Fig. 5 Schematic overview of the enzymatic analysis of glutathione. (a) Assay for the measurement of total GSH content, i.e., reduced and oxidized form. (b) Assay for the measurement of GSSG content. GR—glutathionereductase, DTNB—2,2′-dinitro-5′5-dithiodibenzoic acid, GSH—reduced glutathione, GSSG—oxidized dimer of glutathione

is established. The use of a standard curve of known concentrations allows calculation of the content of glutathione in a sample. In order to distinguish between the total glutathione and the GSSG content of a sample, GSH is selectively derivatized by 2-vinylpyridine. The remaining GSSG is then measured after reduction to GSH. 2-Vinylpyridine is more suitable for the assay than the previously used N-ethylmaleimide (NEM) which inhibits the glutathione reductase that is essential in the assay (Fig. 5). 3.3.1 Preparation of Plant Extract

1. Grind 200 mg of leaf material in liquid N2 (see Note 18). 2. Add 1 mL pre-cooled sample buffer; grind further. The sample buffer prevents degradation and further oxidation of GSH. 3. Centrifuge for 5 min at 16,000 × g, 4 °C. 4. Transfer 900 μL of the solution to a 1.5 mL tube. 5. Keep the sample on ice. 6. Check the pH value and adjust to pH 6–7 with 5 M K2CO3 or 1 M NaOH, if necessary. Be very careful when adjusting the pH and start with a very small volume (e.g., 2 μL). Generally keep the added volume as small as possible to minimize dilution effects. Note the volume of K2CO3 solution added for each sample (VK).

3.3.2 Sample Preparation for Total GSH Content

1. Mix 200 μL of the neutralized plant extract (or 200 μL GSH standard) with 100 μL DTNB (see Note 19). 2. Incubate for 5 min. 3. Add 5 μL 2-vinylpyridine. Here, GSH was already oxidized in step 1 by DTNB. Thus, 2-vinylpyridine cannot react with GSH.

252

Katharina König et al.

It is added to establish equal chemical conditions as in the assay for GSSG described below. 4. Incubate for 15 min. 5. Centrifuge for 5 min at 16,000 × g, 4 °C. 3.3.3 Sample Preparation for GSSG Content

1. Mix 200 μL of the neutralized plant extract (or 200 μL GSSG standard) with 5 μL 2-vinylpyridine (see Note 19). Here, 2-vinylpyridine is added to derivatize GSH in the sample. 2. Incubate for 15 min. 3. Add 100 μL DTNB. 4. Incubate for 5 min. 5. Centrifuge for 5 min at 16,000 × g, 4 °C.

3.3.4 Photometric Measurement

The test is conducted for the standard as well as for the samples to measure both the total GSH and the GSSG content, respectively. It is recommended to make at least triplicate determinations. 1. The measurement is performed at 412 nm. 2. Pipet 1 mL test buffer into the cuvette. 3. Add 20 μL NADPH; mix thoroughly. 4. Use the “autozero” function of the photometer to set the background value to zero before starting the measurement. 5. Add 200 μL sample either for total GSH or GSSG or from standard solutions (after performing the sample preparation step—see Note 19). 6. Record the absorbance for at least 2 min with measurements taken at least every 0.5 min to get the baseline. 7. Start the reaction by adding 5 μL glutathione reductase and swift mixing of the contents inside the cuvette. 8. Measure the time-dependent increase Abs412nm.

3.3.5 Calculation of Total GSH and GSSG Contents

1. Calculate the amount of either GSH or GSSG that is present in the cuvette during the measurement of the standards. 2. Calculate the slope for each standard concentration over 1 min (ΔAbs/min). This value corresponds to the calculated amount (see Fig. 6). 3. Plot ΔAbs/min versus nmol glutathione and calculate the regression line. 4. Calculate ΔAbs/min for your samples. 5. Calculate the amount of glutathione in your measured sample applying the regression equation.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

253

1.2

Absorbance [412 nm]

+ Glutathione reductase

0.8

0.4 ∆Abs/min

0 0

2

4

6

8

10

Time [min]

Fig. 6 Time-dependent absorbance change in a typical test for glutathione determination. The absorbance was stable prior to the addition of glutathione reductase which caused a linear increase in absorbance due to DTNB reduction

6. To calculate the concentration of GSH or GSSG per mg or g fresh weight (FW) you also have to consider the dilutions that were made during the extraction and sample preparation according to the following equation exemplified here for total GSH. nmol GSH / GSSG nmol GSH S / GSSG S 1000 µ l + 180µ l 900 µ l + V K µ l = x x 0.2g FW 900 µ l g FW 200 µ l x =

200 µ l + 100 µ l + 5 µ l 200µ l

nmol GSH S / GSSG S 99 + 0.09xV K ) x ( 8.9 0.2g FW

7. Reduced GSH amount is obtained by subtracting the amount of oxidized glutathione (GSSG) from the total amount of glutathione. The method described here for the measurement of GSH and GSSG in plant extracts relies on equipment available in most laboratories. It is fast and easily carried out and, therefore, provides the researcher with easy access to data reflecting the amount and the redox state of GSH and GSSG in dependency of applied stress conditions as well as circadian regulation.

254

Katharina König et al.

3.4 Detection of Glutathione In Vivo by Redox-Sensitive Green Fluorescent Protein (roGFP)

Fluorescent proteins are powerful tools to image expression, subcellular localization, or particular chemical properties with subcellular resolution. Thus the development of thiol roGFP was a significant progress in addressing the thiol state of specific subcellular compartments. The redox state of roGFP2 is in equilibrium with the cellular glutathione redox state mediated by glutaredoxins, whereas the redox state of thioredoxins does not interfere with roGFP [28, 29]. Since roGFP1 and roGFP2 share the same structure and amino acid sequence except of the single S65T mutation buried inside the β-can structure, glutaredoxins mediate between glutathione and roGFP1 at least in vitro as well [30]. In vitro in the absence of glutaredoxin the response of the proteins to alterations of the environmental redox state is comparatively slow. The fusion of glutaredoxin 1 (Grx1) to the N-terminus of roGFP2 was reported to increase sensitivity and accelerate the response of roGFP2 [29].

3.4.1 Midpoint-Redox Potential of roGFP1 and roGFP2

The midpoint-redox potentials and dynamic ranges differ between roGFP1 and roGFP2. Whereas roGFP1 has a midpoint-redox potential of −288 mV, the potential of roGFP2 is shifted towards oxidizing conditions (−272 mV [31]. roGFP2 has been reported to have a larger dynamic range than roGFP1 and photoisomerization was observed for roGFP1 [30]. Hence, roGFP2 has been chosen for further analysis in plant cells [28].

3.4.2 Image Acquisition by Confocal Laser Scanning Microscopy

A 405 nm diode laser and a 488 nm laser line (see Note 20), e.g., from an argon ion laser are required for excitation (see Note 21). Both fluorescent proteins can be excited at 405 and 488 nm, but one excitation maximum is less pronounced in comparison to the other. In roGFP1 the neutral state (maximal Abs at 395 nm) and in roGFP2 the anionic state (AbsMax at 475 nm) are prominent. Due to the unequal efficiency of excitation at 405 and 488 nm, the intensity of laser lines has to be adjusted with respect to each other. A recommendable pre-adjustment is to aim at similar emission intensities in both channels. The emission of reduced and oxidized roGFP is detected by a single detector to avoid deviations due to individual sensitivity and linearity of signal amplification. This enables image-specific adjustment of the detector gain. Offset values shall be chosen that result in low level of background noise. However, complete lack of background noise has to be avoided in order to maintain the maximum dynamic range. To achieve the maximum dynamic range, the gain has to be adjusted to a voltage that allows for acquisition of images close to saturation of the detector, but being aware of the linear range of the detector. The pixel depth corresponds to the resolution of the dynamic range. A pixel depth of 8 bit per pixel results in 256 increments of intensity, whereas a pixel depth of 12 bit results in 4,096. The latter is recommended.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity 3.4.3 Calibration and Data Evaluation

255

The redox potential of glutathione can be expressed by the Nernst equation that involves the gas constant R (8.315 J/K/mol), the absolute temperature T, the number of transferred electrons z, the Faraday constant F (96,485 Cmol−1), and the midpoint-redox 0′ potential of glutathione E GSH at pH 7 (−240 mV): 2

E GSH = E 3.4.4 pH Dependency of Measurement

0′ GSH

RT [GSH ] − ln . zF GSSG

Since the midpoint potential depends on the pH as well and, e.g., proton pumps that are known to adjust the cytosolic pH were shown to be activated in a blue-light-dependent manner [32] (see Note 22) and hence show a diurnal circadian rhythm, the response of roGFPs relies on two components: (a) the (sub)cellular redox state and (b) the (sub)cellular pH, although the cytosolic pH is independent on the proton transport by the photosynthetic electron transport chain [33]. In all cases described here the pH had less impact on the oxidized form of roGFPs but affected the reduced form to different extents. Fluorescence quenching of the reduced roGFP2 was pronounced under acidic conditions (Fig. 7). The emission ratio increases with increasing pH, the slope is steeper for roGFP2 and Grx1-roGFP2 than for roGFP1 (Fig. 8). roGFP1 appears relatively insensitive to changes in the pH as reported before (see Note 23). The pH of subcellular compartments differs generally and ranges from acidic conditions at pH 5.2 in the vacuole to basic conditions at pH 8.4 in peroxisomes [34]. The obtained redox potentials have to be corrected with respect to the pH. Reference values for subcellular pH can be found in Table 1. The redox potential (volt) can be corrected for the environmental pH by the following equation: pH E roGFP = E roGFP − 0.0615 ( pH − 7 ) .

3.5 Oxidation State of 2-Cysteine Peroxiredoxin

Redox-dependent posttranslational modifications cause changes in the regulation of protein activity. During oxidation and reduction of proteins the involved cysteinyl thiols undergo posttranslational modifications, where the thiol groups of the cysteinyl residues form, e.g., inter- or intramolecular disulfide bonds and sulfenic, sulfinic, or sulfonic acid derivates. Thus the redox state of cysteine plays an important regulatory role in protein functions besides their involvement in structural stability. Plant 2-CysPrx adopts different conformations and functions depending on its redox state [35]. Thus the reduced form functions as thiol peroxidase, while the hyperoxidized form acts as chaperone. The hyperoxidized form was recognized as one of the first transcription/translation-independent molecular marker of circadian rhythmicity [6, 36]. In order to

256

Katharina König et al.

Ratio emission

1.2

roGFP1,

reduced

pH 8 pH 7 pH 6 pH 5

0.8 0.4

0.4 0 300

1.6

Ratio emission

oxidized

0.8

0 400

300

500 0.8

roGFP2, reduced

1.2

0.6

0.8

0.4

0.4

0.2

400

500

roGFP2, oxidized

0

0 300

1.2

Ratio emission

roGFP1,

1.2

400

500

300

Grx1-roGFP, reduced

400

500

Grx1-roGFP, oxidized

0.6

0.8 0.4 0.4

0.2

0

0 300

400

500

300

Wavelength [nm]

400

500

Wavelength [nm]

Fig. 7 pH Dependency of roGFP1, roGFP2, and Grx1-roGFP2. Recombinant proteins were incubated in buffers of pH 5–8 for 30 min and proteins were either reduced or oxidized with DTT. Excitation spectra were recorded using a fluorescence spectrometer (Kontron SFM25)

Ratio emission

roGFP1 0.12

roGFP2 16

reduced

0.08

12

reduced

8 0.04

oxidized

0 6

pH 8

12

reduced

8 oxidized

4 0

4

Grx1-roGFP2 16

4 oxidized

0 4

6

pH 8

4

6

pH 8

Fig. 8 pH Dependency of the relative emission ratios of roGFP1, roGFP2, and Grx1-roGFP2. The ratios were calculated based on the recorded spectra of Fig. 7. The ratios of the reduced form are shown in black and the ratios of the oxidized form in grey

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

257

Table 1 pH of subcellular compartments according to Shen et al. [34] Compartment

pH

Cytosol

7.3

Nucleus

7.2

Mitochondria

8.1

Chloroplast (stroma), dark-cultured protoplasts

7.2

Peroxisomes

8.4

ER

7.1

Cis-Golgi

6.8

TGN/EE

6.3

MVB/PVC

6.2

Vacuole

5.2

ER endoplasmatic reticulum, TGN trans Golgi network, EE early endosomes, MVB multivesicular bodies, PVC prevacuolar compartment

analyze the oxidation state of 2-cysteine peroxiredoxin (2-Cys Prx), Western blot and subsequent immunodetection was performed. These are efficient techniques for detection of specific proteins ex vivo. By extracting 2-Cys Prx under oxidized conditions it is possible to confirm the different oxidation states of this protein and to draw conclusions on the changes of their conformational state during the circadian rhythmicity. 3.5.1 Protein Extraction from A. thaliana Plant Material

1. For extraction of A. thaliana 2-Cys Prx, harvest leaves at different time points of the day–night cycle. Homogenize 50 mg leaf material with a micropistil in a 1.5 mL reaction tube. 2. 50 mg Homogenized leaf material are supplemented with 300 μL extraction buffer or with extraction buffer for oxidized conditions. 3. Centrifuge the extract at 16,000 × g at 4 °C for 10 min. 4. Transfer the protein containing supernatant to a fresh reaction tube. Store the protein samples on ice up to further usage.

3.5.2 Protein Quantification Assay According to Bradford

1. Mix 40 μL of Roti-Quant reagent and 150 μL sterile water in a microtiter plate (see Note 29). 2. Add 10 μL of protein extract and mix briefly. Air bubbles should be avoided. For standardization, use BSA standards of 0.1–1 mg instead of protein extracts. 3. Incubate the reaction at room temperature for 5 min.

258

Katharina König et al.

4. The protein concentration is determined by measuring the absorbance at 595 nm with a microplate autoreader. 5. The protein concentration is calculated on the basis of the standard curve generated by using BSA of known concentrations. 6. The proteins are analyzed using reducing and non-reducing SDS-PAGE and Western blot. Immunodetection after Western blotting allows for the identification of the oxidation state of specific proteins (see below). 3.5.3 Reducing and Non-reducing SDS-PAGE

1. These instructions assume the use of an SDS-PAGE system with gels of 18 cm (width) × 10 cm (height) × 1 mm (thickness). The volume can be easily adapted to other formats. 2. Thoroughly clean all components of the gel electrophoresis system (e.g., glass plates, spacers, and comb) with 70 % ethanol and a lint-free tissue (e.g., Kimtech Kimwipes, Kimberley-Clark). 3. Assemble the system according to the manufacturer’s recommendations. 4. Prepare 15 mL of 12 % separating gel by mixing 3.7 mL separating gel buffer with 6 mL acrylamide solution, 5.2 mL water, 100 μL APS solution, and 14 μL TEMED. The polymerization process will start as soon as APS and TEMED have been added. Avoid formation of air bubbles. 5. Pour the separating gel between the plates carefully, overlay with 1 mL 100 % isopropanol, and wait until the gel is polymerized (takes approximately 30 min). 6. Pour off the isopropanol. 7. Prepare 5 mL 6 % stacking gel by mixing 1.25 mL stacking gel buffer with 1 mL acrylamide solution, 2.65 mL water, 50 μL APS solution, and 5 μL TEMED. 8. Pour the stacking gel and carefully insert the comb. Avoid formation of air bubbles. 9. Once the stacking gel is polymerized (~20 min), assemble the gel electrophoresis system by attaching the gel to the chamber and fill the reservoir chambers with 1× running buffer. After removing the comb, wash the wells with 1× running buffer using a Hamilton syringe. 10. Prepare the protein samples for the gel: 20 μg protein of each sample is filled up with extraction buffer to an adequate equal volume. 11. In order to check the oxidation state of the proteins use SDSPAGE under reducing and non-reducing conditions. For reducing conditions add 10 mM DTT to 5× loading buffer, sufficient for all samples.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

259

12. Add 5× loading buffer to each sample in a volume of 1:5 to the protein sample. Ensure that only the samples with the Extraction buffer without H2O2 (extraction buffer) are supplemented with the Loading buffer containing DTT (reducing conditions). The samples with the Extraction buffer with H2O2 (extraction buffer for oxidized conditions) are supplemented with loading buffer without DTT (non-reducing condition). 13. Heat the samples supplemented with loading buffer at 95 °C for 10 min. This causes denaturation of tertiary and quaternary structures. 14. Centrifuge the samples supplemented with loading buffer at 16,000 × g at room temperature for 5 min to pellet any insoluble matter. 15. Carefully apply 10 μL of sample per well using the Hamilton syringe. 16. Apply 3 μL of prestained molecular weight marker in one well. 17. Run the gel at constant 60 mA until the running front reaches the bottom. 3.5.4 Western Blot Analysis: Semidry Western Transfer

1. Cut the nitrocellulose membrane and six pieces of blotting paper to the same size as the separation gel. 2. Incubate the separation gel and three pieces of blotting paper in cathode buffer for 5 min. 3. Incubate the nitrocellulose membrane and one piece of blotting paper in 1× anode buffer. 4. Incubate two pieces of blotting paper in 10× anode buffer. 5. Using the Semidry blotter Fastblot B44 the blot consists of: (a) Two pieces of blotting paper soaked in 10× anode buffer. (b) One piece of blotting paper soaked in 1× anode buffer. (c) Pre-soaked membrane. (d) Gel equilibrated in cathode buffer. (e) Three pieces of blotting paper soaked in cathode buffer. 6. Apply a constant current of 2 mA per cm2 of the membrane surface for 30 min. The appropriate transfer time can be calculated based on the gel size (see Note 30).

3.5.5

Ponceau S Staining

1. To control the efficiency of protein transfer after blotting, stain the membrane with Ponceau S for 2 min. 2. Subsequently wash the membrane in deionized water until protein bands are clearly visible. 3. Remove the residual dye by washing the membrane with 1× TBST buffer.

260

Katharina König et al.

3.5.6 Immunodetection of Immobilized Proteins with Specific Antibodies and Chemiluminescence

1. Incubate the membrane in blocking solution with constant shaking for 1 h at room temperature to block the unoccupied membrane sites. 2. Incubate the membrane with the specific primary antibody (here: 2-CysPrx specific antibody from rabbit diluted 1:2,000 in blocking solution) either for 3 h at room temperature with constant shaking or alternatively over night at 4 °C. 3. Wash the membrane with 1× TBST for 3 × 5 min. 4. Incubate the membrane with secondary goat anti-rabbit antibody with conjugated horseradish peroxidase (diluted 1:3,000 in blocking solution) for 1 h at room temperature with constant shaking. 5. Repeat step 3 of this section. 6. During the wash step wrap the glass plate with the commercial wrapping film. Remove air bubbles by gently wiping the surface with a tissue to press the bubbles towards the edge of the glass plate. 7. Place the membrane protein side up on the wrapped glass plate. 8. Mix equal amounts of Lumi-Light Enhancer and Lumi-Light Stable Peroxid solution (see Note 31). 9. Add the Lumi-Light substrate solution onto the membrane. 10. Immediately cover the membrane with commercial wrapping film, remove air bubbles, and incubate for 5 min. (For increased sensitivity the substrate incubation time can be increased up to 30 min.) 11. Gently squeeze out excess liquid onto a filter paper. 12. All further steps have to be done in the dark to avoid an overexposure of the X-ray film. 13. Place X-ray film on the membrane and expose for 1 min. 14. Incubate the film in the developer till bands are visible. 15. Rinse the film sufficiently with water. 16. Incubate the film with fixation solution and water it for at least 3 min. 17. Repeat the steps 13–16. Adjust the exposure time between 10 s and up to 1 h according to the result with the first film. 18. The membrane can be scanned and analyzed with software for quantification of bands, e.g., GelScan (BioSciTec, Frankfurt/ Main, Germany) (Fig. 9).

3.6 Determination of S-Glutathionylation of Proteins by ESI-MS

Glutathione (GSH) is highly abundant (1–10 mM) in plasmatic compartments of most cells, with the reduced form being predominant often with estimated 99 %. Plant chloroplasts exhibit a GSH concentration between 1 and 4 mM [18]. GSH is the major redox

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

261

Fig. 9 Detection of oxidation state of 2-Cys Prx at different time points of the day–night cycle. Proteins extracted from plant material under oxidizing conditions were subjected to non-reducing SDS-PAGE analysis. Subsequent Western blotting and detection of the proteins with chemiluminescence was performed. Under the test conditions, 2-Cys Prx exhibits three different redox states, namely oxidized dimer (ox) with the size of about 44 kDa with two disulfide bridges, the single hyperoxidized form with one disulfide bridge and the double hyperoxidized form (overox) running as dimer under oxidizing conditions since the disulfide bridge cannot be built any more. The double-overoxidized form of 2-Cys Prx appears as monomer with the size of about 25 kDa in non-reducing SDS-PAGE. (a) Plant material was harvested at different time points. 0 h corresponds to the beginning of the light phase (0 h = timepoint 0 = lights on), 10 h equates the end of the light phase. At the beginning of the light phase reduced amounts of overoxidized 2-CysPrx are present. Overoxidized 2-CysPrx increased with advanced light treatment. (b) Plant material was harvested after 36 h of continuous dark (DD) treatment which started at dawn (time point 0). It should be noted that this method only detects those overoxidized subunits where both peroxidatic cysteinyl residues within the dimer were overoxidized. For more sensitive detection, anti sulfenic acid peroxiredoxin antibody needs to be used

buffer that keeps the cellular environment reduced. The small molecular weight thiol is also involved in a posttranslational modification termed S-glutathionylation [37–39]. Glutathionylation has two proposed functions, namely to protect protein thiols from oxidation to sulfinic and sulfonic derivatives and to regulate protein function. The basic mechanism of glutathionylation is a spontaneous thiol/disulfide exchange between the small molecular weight disulfide and a protein cysteine thiol (Prot-SH) as described by the following reaction: Prot – SH + GSSG ↔ Prot – SSG + GSH.

262

Katharina König et al.

Glutathionylation occurs predominantly under oxidative stress conditions, leading to an alteration in steric properties, charge distribution, protein conformation, and thereby function [40–43]. To analyze the S-glutathionylation of a protein, mass spectrometry is a powerful method due to its ability to provide highly accurate molecular weight information on intact molecules [44, 45]. 3.6.1 Adjustment of the Redox State of the Protein

1. Protein concentration: Purified protein is adjusted to a concentration of 20–300 μM in sample buffer for glutathionylation. 2. Complete reduction of proteins is achieved by incubating 20–300 μM (198 μL) protein with 2 μL of 1 M DTT (final concentration 10 mM) overnight at 4 °C or at room temperature for 4 h.

3.6.2 Removal of Excess DTT

Remove excess DTT after the reaction by size exclusion chromatography (PD-10 columns) with gravity flow. 1. Column preparation: Remove the top cap, cut the sealed end of the column at notch, and pour off the column storage solution. 2. Equilibration: Fill up the column with sample buffer and allow the buffer to enter the packed bed completely. Discard the flow through. Repeat this step four times, so that in total 25 mL sample buffer was used for equilibration. 3. Sample application: Add the sample slowly to the middle of the column. After the sample has entered the packed bed completely, add sample buffer. 4. Elution: Elute with 6 mL sample buffer and collect the eluate in 0.5 mL fractions. 5. Analysis of protein-enriched fractions: Prepare 160 μL H2O with 40 μL of Roti-Quant per fraction and aliquot it into a microtiter plate. Add 20 μL of fractions, mix well by pipetting and incubate at room temperature for 5 min. Protein-enriched samples appear blue. To test whether these fractions contain DTT, mix 80 μL H2O with 20 μL DTNB and 20 μL of the fractions. A yellow color indicates the presence of DTT, which occurs usually at fractions 10–11.

3.6.3 Reaction of Reduced Thiol Proteins with the Disulfide Form of Glutathione

1. Combine the protein-enriched, desalted samples and adjust the concentration to 50 μM reduced protein in sample buffer and add 0.25–5 mM GSSG (final concentration). Incubate at 4 °C over night or for 4 h at room temperature. For controls skip the addition of GSSG and proceed further with step 2. This step can be refined by adjusting the thiol redox potential with defined [GSH]/[GSSG] ratios in order to titrate the glutathionylation reaction.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

263

2. Stop the reaction by adding 800 μL of ice cold (−20 °C) acetone to the samples, mix by vortexing and incubate on ice for 1 h. 3. The precipitated proteins are obtained by centrifugation at 20,000 × g at 4 °C for 15 min. 4. Wash the pellet in 1 mL of cold 80 % acetone (see Note 35). 5. Repeat steps 3 and 4. 6. Collect final protein precipitate by centrifugation at 20,000 × g at 4 °C for 15 min. 7. Carefully remove the acetone solution completely without disturbing the pellet and dry samples under dry air to eliminate any acetone residue (see Note 36). 8. Resuspend final pellet in 50 μL ultra-pure water (see Note 37). 3.6.4 Mass Spectrometric Analysis (ESI-MS)

The mass of intact modified and unmodified proteins is determined by ESI-MS. 1. Instrumental parameters for Esquire 3000 plus Quadrupole Ion Trap Mass Spectrometer, which gave the best ion abundances for glutathionylation measurements, are explained in the following. Check the parameters with your instrument. Capillary voltage = 4,000 V. Nebulizer gas pressure = 15 psi. Drying gas flow = 4.0 L/min. Drying gas temperature = 300 °C. The mass-to-charge (m/z) values depend on the molecular weight of the analyzed protein (see Note 38). 2. Sample preparation: Mix 10–25 μL of the protein solution with 489–474 μL 30 % ethanol and 1 μL formic acid and vortex (see Note 39). 3. Inject the sample with a flow rate of 180 μL/h. Check the parameter with your instrument. 4. Start the acquisition and process the resulting mass spectra with any deconvolution software, e.g., DataAnalysis (Bruker Daltonics, Bremen, Germany) according to the manufacturer’s instructions (Figs. 10 and 11). Here, a method is described to detect S-glutathionylation in vitro. This method can be combined with immunoprecipitation and subsequent LC-mass spectrometric analysis. Alternatively, S-glutathionylation can be monitored by immunoblot analysis with an anti-glutathione antibody as primary antibody. In the future, it will be interesting to monitor the glutathionylated proteome along with the circadian clock.

Katharina König et al.

a 21+ 926.6

100 22+ 884.5

+MS Prx II E reduced

20+ 972.8

relative intensity (%)

80 24+ 810.9

19+ 1024.1

23+ 846.0

60

40

20

25+ 778.5 26+ 748.7

0 800

1000

1200

m/z

b 21+ 22+ 941.2 898.4

100

+MS Prx II E + GSSG

23+ 859.4

80

relative intensity (%)

264

20+ 988.1

24+ 823.7

19+ 1040.1

60

40

20

25+ 790.7 18+ 1097.8

26+ 760.3 27+ 29+ 28+732.2 681.8 706.1

17+ 1162.3

0 800

1000

1200

m/z

Fig. 10 Representative MS spectra of reduced and glutathionylated PrxIIE. ESI-MS spectra of the heterologously expressed and purified Arabidopsis thaliana peroxiredoxin IIE (PrxIIE; At3g52960). (a) Ion signals of the reduced form of the protein with their corresponding charge. The seven charge states from +19 to +26 with a maximum at +21 are plotted. ESI-MS generates highly charged ion species. Therefore an ESI mass spectrum is characterized by multiple peaks that differ by one charge. The distribution of charge states occur after multiple protonation of the protein (in positive ion mode). (b) Glutathionylation was carried out by disulfide exchange with 100 μM GSSG for 4 h at room temperature. The glutathionylated protein species is presented in 12 charge states, from +17 to +29

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

265

a 1+ 19438.0

100

deconvoluted +MS Prx II E reduced

relative intensity (%)

80

60

40

20

0 19000

19200

19400

19600

19800

20000

Mass (Da)

b 100

1+ 19743.8

deconvoluted +MS Px II E + GSSG

relative intensity (%)

80

60

40

20

0 19000

19200

19400

19600

19800

20000

Mass (Da)

Fig. 11 Deconvoluted ESI-MS spectra of Arabidopsis thaliana PrxIIE treated with GSSG. The untreated PrxIIE showed a molecular mass of 19,438.0 Da (a). The reduced form of the protein has a theoretical mass of 19,438.7 Da. The observed mass of PrxIIE after incubation with 100 μM GSSG resulted in a peak of 19,743.8 Da (b), which could be assigned to a posttranslational modification with one bound glutathione molecule (307 Da)

4

Notes

4.1 Quantification of H2O2 in Extracts by Luminol

1. The sodium carbonate buffer is prepared by adding 60 mL of 0.1 M sodium carbonate solution to 40 mL of 0.1 M sodium bicarbonate solution. The resulting pH is 10.2 but should be checked with pH indicator strips.

266

Katharina König et al.

2. The mixed reagent stock solution can be used undiluted but the method sensitivity is reduced doing so. Solution should be kept in a dark bottle at RT and stored in the dark for at least 1 h before usage. 3. The working solution should be prepared in a dark bottle and stored in the dark for 12 h before usage. 4. Working with solid TCA or solutions should be done with care since TCA is a strong acid. 5. While grinding plant material always add sufficient amount of liquid N2 to prevent thawing. 6. To ensure darkness and temperature invert the samples underneath some cover in a cooling room at 4 °C. 7. For every step afterwards use a dark room. 8. Use the tubes that are recommended for your luminometer by the supplier and which exhibit the least background. 4.2 Measurement of Reduced and Oxidized Ascorbate

9. Sodium phosphate buffers (Na-Pi) can be prepared by combining 1 M solutions of Disodium phosphate (Na2HPO4) and Monosodium phosphate (NaH2PO4). This way buffers with a range between pH 5 and 8.0 are achieved. After combining both solutions to get the desired pH value the stock solution can be diluted to the concentration needed. 10. Dissolve 1,000 U of ascorbate oxidase in 100 μL 0.1 M NaH2PO4, pH 5.6. This 10 U/μL stock solution can be stored at −80 °C. For fresh working solution mix 10 μL of stock together with 90 μL 0.1 M NaH2PO4, pH 5.6 (= 1 U/μL). 11. Plant material should be stored at −80 °C after harvesting, but it should not be stored longer than 2 weeks prior to ascorbate measurements. 12. Dithiothreitol causes the complete reduction of the ASC pool. Together with the data obtained from the reduced samples (see Subheading 3.2) the amount of oxidized ascorbate can be determined.

4.3 Enzymatic Analysis of Reduced and Oxidized Glutathione

13. Sodium phosphate buffers (Na-Pi) can be prepared by combining 1 M solutions of disodium phosphate (Na2HPO4) and monosodium phosphate (NaH2PO4). This way buffers with a range between pH 5 and 8.0 are achieved. After combining both solutions to get the desired pH value the stock solution can be diluted to the concentration needed. 14. DTNB is light sensitive and therefore should be kept in the dark. Only prepare solution for use within 2 days. 15. The age of the glutathione reductase used in the assay and the storage conditions affect its activity. Therefore the ideal dilution has to be tested before starting the sample measurements.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

267

If the activity of glutathione reductase is too high in the assay, then the reaction will proceed too fast to measure. 16. Make several dilutions from the standard concentrations and make several measurements before starting to measure the samples. If you get the same results (ΔAbs/min) for a given standard concentration in several independent measurements then it is sufficient to use only one standard concentration in triplicate when measuring the samples. 17. 2-Vinylpyridine is a solution with high viscosity and great care should be applied when pipetting it. It could be helpful to cut off a part of the tip to get a wider opening of the tip. 18. Plant material should be stored at −80 °C after harvesting, but it should not be stored longer than 2 weeks prior to GSH/ GSSG measurements. 19. For the standards of GSH and GSSG the same preparation steps have to be performed exactly as for the samples. Two hundred microliters of the different dilutions are used instead of plant extracts. 4.4 Detection of Glutathione In Vivo by Redox-Sensitive Green Fluorescent Protein (roGFP)

20. Stability of laser intensity has to be ensured over time. Running the lasers for approximately 1 h before starting the measurements usually overcomes larger fluctuations. 21. If an argon ion multi-line laser is applied, it should be considered that the 488 nm line is one of the strongest lines, so that bleaching effects and cell damage can occur at high irradiation intensities. 22. Excitation of roGFP might induce signaling cascades originated from blue-light receptors in plants. Furthermore, near UV excitation causes some greenish autofluorescence in particular in the vacuole. Therefore, analysis of regions of interest spanning the whole cell should be avoided. Blue-light excitation can be avoided by pulsed excitation of high photon density in the far red/infrared region at approximately doubled wavelength so that roGFP is excited by simultaneous absorption of two photons in the focal plane. Thus, 2-photon excitation apparently is of advantage to avoid induction of blue-lightdependent signaling and suppresses autofluorescence since cellular absorption is widely limited to the focal plane. 23. Since roGFP1 shows photoconversion during the acquisition of time series roGFP2 is recommended for time-dependent measurements. On the other hand, roGFP1 is superior to roGFP2, if changes of environmental pH have to be considered. The application of both sensors allows correcting for fluorescence fluctuations and pH-dependent responses and hence represents the best experimental performance.

268

Katharina König et al.

4.5 Oxidation State of 2-Cysteine Peroxiredoxin

24. Acrylamide is a potent lethal neurotoxin and is absorbed through the skin. Take appropriate safety measures (gloves, fume hood, goggles, and mask), particularly if weighing solid acrylamide/bisacrylamide and while working with the solutions and gels. 25. DTT (1 M) in water is acidic with pH value around 5.0. The strong reducing agent DTT is highly prone to oxidation by air, a process promoted also by multiple freeze–thaw cycles. Therefore, aliquots for single use can be prepared and stored at −20 °C, but it is strongly recommended to prepare DTT solutions fresh prior to usage. 26. Blocking solution can be used once only. Blocking solution can be prepared and stored at 4 °C for 1 week at most. But it is strongly recommended to prepare blocking solution fresh prior to usage. 27. Use a primary antibody which is specific for the protein of interest. Regard in which host species this antibody has been generated. Application of the primary antibody should be done according to the manufacturer’s recommendations. Consider in the manufacturer’s recommendations how often the primary antibody can be used. 28. Use a secondary antibody which is reactive towards the primary antibody. Dependent on the detection method, e.g., chemiluminescence or nitro blue tetrazolium chloride/5Bromo-4-chloro-3-indolyl phosphate (NBT/BCIP) staining, the secondary antibody should either be conjugated to peroxidase or to alkaline phosphatase. Application of the secondary antibody should be done according to the manufacturer’s recommendations. The secondary antibody diluted in blocking solution can be used up to two times. But it is recommended to prepare secondary antibody diluted in blocking solution fresh prior to usage. 29. It is recommended to prepare a master mix, sufficient for all samples, with 10 % additional volume in order to compensate for reagent loss. 30. The blotting time is highly dependent on the molecular weight of proteins. Higher molecular mass proteins (>100 kDa) require longer blotting times than smaller proteins. Consider that at low current the mobility of proteins will be decreased and proteins will not be completely transferred from the gel. But if the current is too high proteins might pass through the membrane without binding. Regarding the power supply approaches (maximum of current and voltage) consider the manufacturer’s recommendations. 31. The amount of the Lumi-Light substrate solution depends on the size of the membrane. Mix 5 μL of each solution per centimeter square blot surface.

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity

4.6 S-Glutathionylation of Proteins by ESI-MS

269

32. Preparing aliquots of DTT and/or storage at −20 °C for single use is recommended, because DTT is highly prone to oxidation by air. 33. DTNB is soluble at neutral pH values. At higher concentrations neutralization with NaOH is required. Too much NaOH causes a sudden deepening of the yellow color so that a positive reaction is masked; too little NaOH and the reagent might not be alkaline enough to react quickly with thiol groups. 34. Because GSSG is insoluble at higher concentrations, avoid preparation of stock solutions >100 mM. 35. Be sure to thoroughly resuspend the pellet by pipetting or vortexing. 36. Allow the acetone to evaporate from the open tube at room temperature. Do not over-dry pellet, otherwise it may not dissolve properly. Instead of air drying, the pellet can be dried by inverting the tube and placing it on a Kimwipe for 15 min at 37 °C. 37. Samples can be shock-frozen in liquid N2 and stored at −80 °C prior to mass spectrometric analysis. 38. For the ~19 kDa peroxiredoxin II E (At3g52960) a range between 650 and 1,200 m/z achieved good results. 39. Ionization depends on the protein. Instead of the buffer mentioned above a (80 %) acetonitrile, (0.1 %) trifluoroacetic acid solution can be used.

References 1. Beaver LM, Klichko VI, Chow ES, KotwicaRolinska J, Williamson M, Orr WC, Radyuk SN, Giebultowicz JM (2012) Circadian regulation of glutathione levels and biosynthesis in Drosophila melanogaster. PLoS One 7:e50454 2. Hardeland R, Coto-Montes A, Poeggeler B (2002) Circadian rhythms, oxidative stress, and antioxidative defense mechanisms. Chronobiol Int 20:921–962 3. White BP, Davies MH, Schnell RC (1987) Circadian variations in hepatic glutathione content, gamma-glutamylcysteine synthetase and gamma-glutamyl transferase activities in mice. Toxicol Lett 35:217–223 4. Lai AG, Doherty CJ, Mueller-Roeber B, Kay SA, Schippers JH, Dijkwel PP (2012) CIRCADIAN CLOCK-ASSOCIATED 1 regulates ROS homeostasis and oxidative stress responses. Proc Natl Acad Sci U S A 109: 17129–17134 5. Muthuramalingam M, Seidel T, Laxa M, Nunes de Miranda SM, Gärtner F, Ströher E, Kandlbinder A, Dietz KJ (2009) Multiple

redox and non-redox interactions define 2-Cys peroxiredoxin as a regulatory hub in the chloroplast. Mol Plant 2:1273–1288 6. Edgar RS, Green EW, Zhao Y, van Ooijen G, Olmedo M, Qin X, Xu Y, Pan M, Valekunja UK, Feeney KA, Maywood ES, Hastings MH, Baliga NS, Merrow M, Millar AJ, Johnson CH, Kyriacou CP, O'Neill JS, Reddy AB (2012) Peroxiredoxins are conserved markers of circadian rhythms. Nature 485:459–464 7. Rosenwasser S, Rot I, Meyer AJ, Feldman L, Jiang K, Friedman H (2009) A fluorometerbased method for monitoring oxidation of redox-sensitive GFP (roGFP) during development and extended dark stress. Physiol Plant 138:493–502 8. Jiang K, Schwarzer C, Lally E, Zhang S, Ruzin S, Machen T, Remington SJ, Feldman L (2006) Expression and characterization of a redox-sensing green fluorescent protein (reduction–oxidation-sensitive green fluorescent protein) in Arabidopsis. Plant Physiol 141:397–403

270

Katharina König et al.

9. Marty L, Siala W, Schwarzländer M, Fricker MD, Wirtz M, Sweetlove LJ, Meyer Y, Meyer AJ, Reichheld JP, Hell R (2009) The NADPHdependent thioredoxin system constitutes a functional backup for cytosolic glutathione reductase in Arabidopsis. Proc Natl Acad Sci U S A 106:9109–9114 10. Møller IM, Jensen PE, Hansson A (2007) Oxidative modifications to cellular components in plants. Annu Rev Plant Physiol Plant Mol Biol 58:459–481 11. Asada K et al (1999) The water-water cycle in chloroplast: scavenging of active oxygens and dissipation of excess photons. Annu Rev Plant Physiol Plant Mol Biol 50:601–639 12. Møller IM et al (2001) Plant mitochondria and oxidative stress: electron transport, NADPH turnover, and metabolism of reactive species. Annu Rev Plant Physiol Plant Mol Biol 52:561–591 13. Takeda T, Yokota A, Shigeoka S (1995) Resistance of photosynthesis to hydrogen peroxide in algae. Plant Cell Physiol 36: 1089–1095 14. Genfa Z, Dasgupta PK (1992) Hematin as a peroxidase substitute in hydrogen peroxide determinations. Anal Chem 64:517–522 15. Ngo TT, Lenhof HM (1980) A sensitive and versatile chromogenic assay for peroxidase and peroxidase-couples reactions. Anal Biochem 105:389–397 16. Warm E, Latjes GG (1982) Quantification of hydrogen peroxide in plant extracts by the chemiluminiscence reaction with luminol. Phytochemistry 21:827–831 17. Perez FJ, Rubio S (2006) An improved chemiluminescence method for hydrogen peroxide determination in plant tissues. Plant Growth Reg 48:89–95 18. Noctor G, Foyer CH (1998) ASCORBATE AND GLUTATHIONE: keeping active oxygen under control. Annu Rev Plant Physiol Plant Mol Biol 49:249–279 19. Pignocchi C, Foyer CH (2003) Apoplastic ascorbate metabolism and its role in the regulation of cell signalling. Curr Opin Plant Biol 6:379–389 20. Smirnoff N (2000) Ascorbate biosynthesis and function in photoprotection. Phil Trans R Soc Lond [Biol] 355:1455–1464 21. Smirnoff N (2000) Ascorbic acid: metabolism and functions of a multi-facetted molecule. Curr Opin Plant Biol 3:229–235 22. Foyer C, Rowell J, Walker D (1983) Measurement of the ascorbate content of spinach leaf protoplasts and chloroplasts during illumination. Planta 57:239–244

23. Oelze M-L, Kandlbinder A, Dietz K-J (2008) Redox regulation and overreduction control in the photosynthesizing cell: complexity in redox regulatory networks. Biochim Biophys Acta 11:1261–1272 24. Griffith OW (1980) Determination of glutathione and glutathione disulfide using glutathione reductase and 2-vinylpyridine. Anal Biochem 1:207–212 25. Foyer C, Lelandais M, Galap C, Kunert KJ (1991) Effects of elevated cytosolic glutathione reductase activity on the cellular glutathione pool and photosynthesis in leaves under normal and stress conditions. Plant Physiol 3:863–872 26. Arisi AC, Noctor G, Foyer CH, Jouanin L (1997) Modification of thiol contents in poplars (Populus tremula x P. alba) overexpressing enzymes involved in glutathione synthesis. Planta 3:362–372 27. Noctor G, Foyer CH (1998) Simultaneous measurement of foliar glutathione, γ-glutamylcysteine, and amino acids by highperformance liquid chromatography: comparison with two other assay methods for glutathione. Anal Biochem 1:98–110 28. Meyer AJ, Brach T, Marty L, Kreye S, Rouhier N, Jacquot JP, Hell R (2007) Redox-sensitive GFP in Arabidopsis thaliana is a quantitative biosensor for the redox potential of the cellular glutathione redox buffer. Plant J 52:973–986 29. Gutscher M, Pauleau AL, Marty L, Brach T, Wabnitz GH, Samstag Y, Meyer AJ, Dick TP (2008) Real-time imaging of the cellular glutathione redox potential. Nat Methods 5:553–559 30. Schwarzländer M, Fricker MD, Müller CA, Marty L, Brach T, Novak J, Sweetlove LJ, Hell R, Meyer AJ (2008) Confocal imaging of glutathione redox potential in living plant cells. J Microsc 231:299–316 31. Hanson GT, Aggeler R, Oglesbee D, Cannon M, Capaldi RA, Tsien RY, Remington SJ (2004) Investigating mitochondrial redox potential with redox-sensitive green fluorescent protein indicators. J Biol Chem 279: 13044–13053 32. Klychnikov OI, Li KW, Lill H, de Boer AH (2007) The V-ATPase from etiolated barley (Hordeum vulgare L.) shoots is activated by blue light and interacts with 14-3-3 proteins. J Exp Bot 58:1013–1023 33. Netting AG (2002) pH, abscicic acid and the integration of metabolism in plants under stressed and non-stressed conditions. II. Modifications in modes of metabolism induced by variation in the tension of the water column and by stress. J Exp Bot 63:151–173

Assessing Redox State and Reactive Oxygen Species in Circadian Rhythmicity 34. Shen J, Zeng Y, Zhuang X, Sun L, Yao X, Pimpl P, Jiang L (2013) Organelle pH in the Arabidopsis endomembrane system. Mol Plant 6:1419–1437 35. Dietz KJ (2011) Peroxiredoxins in plants and cyanobacteria. Antioxid Redox Signal 15:1129–1159 36. O'Neill JS, van Ooijen G, Dixon LE, Troein C, Corellou F, Bouget FY, Reddy AB, Millar AJ (2011) Circadian rhythms persist without transcription in a eukaryote. Nature 469: 554–558 37. Chae HZ, Oubrahim H, Park JW, Rhee SG, Boon Chock P (2012) Protein glutathionylation in the regulation of peroxiredoxins: a family of thiol-specific peroxidases that function as antioxidants, molecular chaperones, and signal modulators. Antioxid Redox Signal 16:505–523 38. Mieyal JJ, Boonchock P (2012) Posttranslational modification of cysteine in redox signaling and oxidative stress: focus on S-glutathionylation antioxid. Redox Signal 16:471–475 39. Ghezzi P (2005) Regulation of protein function by glutathionylation. Free Rad Res 39(6): 573–580

271

40. Park JW, Mieyal JJ, Rhee SG, Boon Chock P (2009) Deglutathionylation of 2-Cys peroxiredoxin is specifically catalyzed by sulfiredoxin. J Biol Chem 284:23364–23374 41. Melchers J, Dirdjaja N, Ruppert T, KrauthSiegel RL (2007) Glutathionylation of trypanosomal thiol redox proteins. J Biol Chem 282:8678–8694 42. Manevich Y, Feinstein SI, Fisher AB (2003) Activation of the antioxidant enzyme 1-CYS peroxiredoxin requires glutathionylation mediated by heterodimerization with πGST. Proc Natl Acad Sci U S A 101(11):3780–3785 43. Noguera-Mazon V, Lemoine L, Walker O, Rouhier N, Salvador A, Jacquot J-P, Lancelin J-M, Krimm I (2006) Glutathionylation induces the dissociation of 1-Cys D-peroxiredoxin non-covalent homodimer. J Biol Chem 281:31736–31742 44. Trauger SA, Webb W, Siuzdak G (2002) Peptide and protein analysis with mass spectrometry. Spectroscopy 16:15–28 45. Neubauer G, Mann M (1999) Mapping of phosphorylation sites of gel-isolated proteins by nanoelectrospray tandem mass spectrometry: potentials and limitations. Anal Chem 71:235–242

Chapter 18 Circadian Regulation of Plant Immunity to Pathogens Robert A. Ingle and Laura C. Roden Abstract The plant circadian clock primes the immune response of Arabidopsis thaliana to infection with the bacterial pathogen Pseudomonas syringae pv tomato DC3000 (Pst DC3000) such that there is a more robust response during the subjective day than subjective night. Thus Pst DC3000 growth in plants infected with the same initial titre of bacteria varies depending on the time of day of infection (Bhardwaj et al., PLoS One 6: e26968, 2011; Zhang et al., PLoS Pathog 9:e1003370, 2013). We describe here a protocol for assaying bacterial leaf titres following pressure infiltration or spray inoculation of Arabidopsis thaliana with Pst DC3000. We also describe a method for assaying plant susceptibility to infection with the necrotrophic fungal pathogen Botrytis cinerea. These methods can be used in studying the circadian clock regulation of signal transduction pathways controlling plant defense responses. Key words Arabidopsis thaliana, Pseudomonas syringae, Botrytis cinerea, Plant–pathogen interaction, Circadian, Diurnal, Bacterial titre, Necrotrophic pathogen, Disease symptoms

1

Introduction In developing an understanding of plant defenses and pathogenesis, plant pathologists have examined interactions between Arabidopsis thaliana and bacteria, oomycetes, and fungi. In studying each interaction, assays to assess the outcome in terms of disease symptom measurement or pathogen multiplication have been developed [3–6]. The variables measured usually correlate the extent of symptom development with the level of pathogen multiplication, allowing conclusions to be drawn regarding the relative strength of the defense response by the plant versus the virulence of the pathogen. We describe protocols for measuring the outcome of infections of Arabidopsis leaves with a biotrophic bacterium, Pseudomonas syringae pv tomato DC3000 (Pst DC3000), and a necrotrophic fungus, Botrytis cinerea. For Pst DC3000, we outline both the pressure infiltration protocol that bypasses stomatal mediated defense and the spray inoculation protocol which does not. Diurnal experiments may be carried out under light–dark and/or

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_18, © Springer Science+Business Media New York 2014

273

274

Robert A. Ingle and Laura C. Roden

temperature cycles, where the time of infection is measured according to a particular time cue or zeitgeber [5] such as light, where dawn is considered to be zeitgeber time 0 (ZT0). To analyze the role of the endogenous clock in the absence of external time cues, experiments in conditions of constant light (or darkness) and temperature where the plant circadian rhythms are “free-running” should be carried out. Entraining plants to diurnal conditions synchronizes the circadian clock to external time before transfer to constant conditions and allows one to experiment in subjective or circadian time (CT), where subjective dawn on the first day is CT0 [7]. In these assays, equivalent doses of pathogen are applied to leaves of the same age and developmental stage at different times of the day by pressure infiltration and spray or drop inoculation. When studying the temporal regulation of plant defenses, it is important that the “strength” or the magnitude of the pathogenic stimulus that is applied at each time point is equivalent. That is, it is important that not only is the titre the same each time, but that the pathogen is in the same state or phase of growth each time too [1, 2]. Because of the number of time points involved in these experiments and the need to inoculate cultures at defined times prior to infection, it is beneficial to have more than one controlled environment chamber for plant growth, each set to different light– dark schedules for differential entrainment. This allows one access to plants at different circadian/diurnal phases at a given time point, reducing the need for round-the-clock experimenting. After inoculation, the plant and pathogen are incubated together under controlled conditions for a defined period of time before symptoms or pathogen titres are measured. In the case of the Pst DC3000 infection, success of leaf colonization is determined by enumerating at regular intervals the number of colony-forming units (cfu) that can be extracted per area of infected leaf [3]. The size of the lesion caused by B. cinerea infection is measured at a defined time interval after infection to determine the disease susceptibility at each infection time [6]. It is also useful to harvest tissues at defined time points for molecular analyses of plant defenses elicited at each infection time.

2

Materials

2.1 Arabidopsis Plants

1. Araflats, with Aratrays and Arabaskets (www.arasystem.com), are the simplest way to grow the plants for these assays. Alternatively, plants may be grown individually in 7 cm diameter pots with a drip tray. It is more convenient to stand multiple pots in a single, large drip tray than in individual ones. 2. Plastic wrap (cling film) or transparent food wrap, or propagator dome.

Clock Regulation of Innate Immunity

275

3. Jiffy-7 Peat Soil 42 mm Pellets (www.jiffygroup.com). 4. Vermiculite. 5. Controlled environment rooms or chambers. 2.2 Pseudomonas syringae DC3000 Infection of Arabidopsis

Use ultrapure water for all solutions. 1. Magnesium sulphate (1 M stock): Dissolve 24.65 g of MgSO4 · 7H2O in water, adjust final volume to 100 mL, and filter sterilize. Store at room temperature. 2. King’s Media B (KB): Dissolve 10 g tryptone, 10 g peptone, 1.5 g anhydrous K2HPO4, 10 g glycerol in 900 mL of water. Adjust pH to 7 with HCl (see Note 1), make up to a final volume of 1 L with water, and autoclave. After autoclaving add 6 mL of sterile 1 M MgSO4 (see Note 2). 3. KB agar: Make KB as above, and then add 12 g of agar per L prior to autoclaving. After autoclaving add 6 mL of sterile 1 M MgSO4. 4. Magnesium chloride (10 mM stock): Dissolve 2.03 g MgCl2 · 6H2O in water, adjust final volume to 1 L, and autoclave. Store at room temperature. 5. Rifampicin (20 mg/mL stock): Dissolve 100 mg of rifampicin in 5 mL of methanol. Aliquot the solution into foil-wrapped microfuge tubes and store at −20 °C. Rifampicin should only be added to KB or KB agar once the medium has cooled to the point at which it is possible to hold the glass bottle comfortably without gloves (see Note 3). 6. Rotary shaker in controlled environment. 7. 70 % (v/v) ethanol. 8. Silwet L77: Only required if plants are to be sprayed with Pst. Can be obtained from Lehle Seeds (www.arabidopsis.com). 9. Plastic wrap (cling film) or transparent food wrap. 10. Spray bottle. 11. 0.5 cm diameter cork borer. 12. Micropestle for tissue grinding in 1.5 or 2.0 mL tubes.

2.3 Botrytis cinerea Infection of Arabidopsis

1. 1 % agar: Add 10 g of agar to 1 L of water and autoclave. 2. Tinned apricot halves in fruit juice (low sugar brand). 3. Grape juice (low sugar brand): Dilute to half-strength by adding an equal volume of sterile water. 4. Colorless transparent plastic boxes (sandwich boxes are ideal). 5. Petri dishes. 6. Hemocytometer. 7. Light microscope.

276

Robert A. Ingle and Laura C. Roden

8. Sterile 1 L beaker. 9. Glass rod. 10. Digital camera.

3

Methods

3.1 Arabidopsis Plant Growth

1. Mix jiffy pellets with vermiculite in a 1:1 ratio and drench with water. 2. Fill Araflat pot cavities or plant pots with the moist peat– vermiculite mixture. 3. Sow Arabidopsis seed on the surface of peat–vermiculite mixture and cover with plastic wrap or clingfilm, or plastic domes, to maintain high humidity for germination (see Note 4). 4. Grow plants in controlled environment of 22 °C with 16-h photoperiod of cool white fluorescent light (approximately 80–120 μmol/m2/s). 5. When after about a week the seeds germinate, make holes in the plastic wrap or open vents on the plastic dome. After another few days the covering may be removed completely. 6. When the cover has been removed, remove extra seedlings from each pot so that each contains one individual plant. 7. Irrigate the plants from below, i.e., into the Aratray or drip tray. Do not let the soil dry completely before watering, but do not overwater the plants.

3.2 Preparation of Pst DC3000 Inocula

1. Streak out Pst DC3000 on a KB agar plate containing 50 μg/mL rifampicin (see Note 3). Grow at 28 °C for 2 days in the dark. 2. Set up an overnight culture by inoculating 10 mL of KB (containing 50 μg/mL rifampicin) with bacteria from the streaked plate. Incubate on a shaker at 28 °C in the dark for approximately 12 h, at which point OD600 of the culture should be in the range of 0.6–1, indicative of growth to mid to late log phase (see Note 5). 3. a. If pressure infiltrating plants with Pst DC3000: Transfer 1.5 mL of the overnight culture to a 2 mL microfuge tube and centrifuge at 3,000 × g for 5 min at RT. Wash cells by resuspending the bacterial pellet in 1 mL of 10 mM MgCl2 and centrifuge again for 5 min at 3,000 × g. Resuspend the bacterial pellet in 0.5 mL of 10 mM MgCl2. b. If spraying plants with Pst DC3000: A larger volume is required, so centrifuge the entire 10 mL overnight culture at 3,000 × g for 5 min at RT. Wash by resuspending the bacterial pellet in 6 mL of 10 mM MgCl2, and centrifuge again for 5 min at 3,000 × g. Finally, resuspend the bacterial pellet in 3 mL of 10 mM MgCl2.

Clock Regulation of Innate Immunity

277

4. Dilute an aliquot of the cell suspension 1:20 in 10 mM MgCl2, and determine the OD600 in a spectrophotometer. 5. a. Using the OD reading obtained for the 1:20 dilution, calculate the volume of original cell suspension required to give an OD600 of 0.2 in a final volume of 5 mL (pressure infiltration) or 50 mL (spraying). Make the appropriate dilution in 10 mM MgCl2, and check the OD600 to confirm that it is indeed 0.2. This OD600 corresponds to 108 cfu/mL. This is the standard concentration used if plants are to be sprayed with bacteria. b. If plants are to be pressure infiltrated with Pst DC3000, dilute the 5 mL suspension 1:100 to give a final concentration of 106 cfu/mL. 3.3 Pressure Infiltration of Arabidopsis with Pst DC3000

1. Grow Arabidopsis plants for 4 weeks under a 16-h light/8-h dark cycle at 22 °C. For experiments with plants in free-running conditions, transfer plants to constant light conditions during the 16-h light period at least 24 h prior to infection. Subjective or circadian times are calculated from the last ZT0. 2. We recommend using three to five plants per time point in experiments and infecting three leaves per plant with Pst DC3000 (see Note 6). Since resistance to Pst DC3000 has been reported to vary with leaf age, it is recommended to use the three youngest fully expanded leaves on each plant to minimize variation between replicates. Identify and mark the leaves to be pressure infiltrated with a permanent marker. 3. Use a needleless 1 mL syringe to pressure infiltrate Pst DC3000 into the leaf tissue by placing the end of the syringe against the abaxial (lower) surface of the leaf and your finger on the adaxial (upper) surface (Fig. 1). Avoid the midrib to minimize damage

Fig. 1 Pressure infiltration of a Pst DC3000 suspension into an Arabidopsis leaf via the abaxial surface. The midrib is avoided to minimize damage to the vascular system

278

Robert A. Ingle and Laura C. Roden

to the leaf vascular system. Gently depress the plunger to infiltrate the leaf tissue with the bacterial cell suspension (see Note 7). Only a small volume (7, so adjust pH to 7 (or just below) using HCl if necessary. 2. KB will become cloudy if MgSO4 is added prior to autoclaving. 3. For virulent Pst DC3000 only rifampicin is required. Avirulent Pst DC3000 strains where the Avr gene is carried on a plasmid will require additional antibiotic selection. 4. The pots/Araflats may be placed at 4 °C for 3–4 days for stratification to synchronize germination and then moved to the growth room/chamber.

282

Robert A. Ingle and Laura C. Roden

5. Overnight cultures with an OD600 >1 should not be used as they will contain high proportion of dead bacteria, resulting in poor infection of the plant. It is very important that the bacterial inoculum used for each infection is made up from cultures in the same phase of growth each time. We advise that the growth characteristics for Pst DC3000 are determined empirically for your own laboratory prior to experimentation. Inocula should then be prepared each time from cultures that are in mid-log phase. 6. For example, if you want to determine bacterial titres 4 and 48 hpi in plants infected at CT24 and CT42 you would need 12 (3 per time point) to 20 plants (5 per time point). It is also recommended to infiltrate (or spray) three plants with 10 mM MgCl2 and harvest tissue 48 hpi as a negative control. 7. Wear eye protection here to avoid splashes of bacterial suspension. 8. We recommend preparing the required number of microfuge tubes containing 900 μL of 10 mM MgCl2 prior to harvesting any leaf tissue. 9. Multiplying by 100 takes into account the fact that only 10 μL of the 1 mL diluted sample is plated onto the KB agar plate. For example, using a cork borer 0.5 cm in diameter and observing 23 cfu at the 10−4 dilution cfu/cm2 = (23 cfu × 10,000 × 100)/(3 × 0.1963 cm−2) = 39,055,867 or 3.91 × 107 cfu/cm2. 10. At 4 hpi the values obtained from plants pressure infiltrated with a 106 cfu/mL suspension of Pst DC3000 should be approximately 104 cfu/cm2 as the bacterial numbers will not have increased significantly by that time. We typically find that bacterial titres of virulent Pst DC3000 are 1,000–10,000 times higher by 48 hpi and disease symptoms are evident, while avirulent Pst DC3000 strains will show minor (95 % for the hybrid aspen T89 wild type. 4. Avoid potting plants that are too small as their roots are easily damaged during transplanting. The plants to be transplanted should have a height of around 10 cm and have well-developed roots. If using in vitro-cultured plants, also avoid potting plants that are so large that they have grown into the lid of the propagation container, as this may result in disturbed growth at later stages. 5. Avoid crowding the plants as this may impair growth and make it difficult to make measurements inside the chamber. 6. Do not measure the diameter at the actual nodes, as this will skew your measurements. 7. Expected growth rate of Populus trees is around 10–20 cm or around 10 % per week under optimal growth conditions. 8. When buds reach a bud score of 0, they are hard and have a prickly and shiny appearance. 9. As some plants may refuse to enter dormancy or resume growth due to unknown internal factors, when at least six plants or preferably 90 % or more of the population has set bud, it is reasonable to proceed with further treatments. 10. This protocol describes the study of certain seasonal key events and provides ample room for additional treatments during other growth studies. It could, for example, be used when studying specific wavelengths of light, during the dark period, or when studying growth under suboptimal temperature conditions as when working with light receptor RNAi or overexpressor trees [18]. 11. Use a green light when samples are collected during the night, as other wavelengths are more likely to trigger a light response [19].

Monitoring Seasonal Bud Set, Bud Burst, and Cold Hardiness in Populus

323

Acknowledgements The authors are grateful for financial and other support from various funding bodies and institutions. M.J. is supported by a scholarship from the Alexander von Humboldt Foundation and through SPP1530 (DFG). C.I. is supported by FONDECYT grant no. 1110831 (CONICYT—Chile). M.E.E. is a VINNMER Marie Curie International Qualification Fellow funded by the Swedish Governmental Agency for Innovation Systems (VINNOVA) and the European Union, with current support from the Swedish Research Council (VR), Umeå University Career Grant, and Carl Trygger’s Foundation for Scientific Research. Further M.E.E. would like to acknowledge the Department of Plant Sciences and Churchill College at Cambridge University, Umeå Plant Science Centre, and support from the VRand VINNOVA-funded Berzelii Centre of Forest Biotechnology, FORMAS, the Kempe Foundation, and Nils och Dorti Troëdssons forskningsfond. References 1. Jansson S, Douglas CJ (2007) Populus. A model system for plant biology. Annu Rev Plant Biol 58:435–458 2. Ibáñez C, Kozarewa I, Johansson M, Ögren E, Rohde A et al (2010) Circadian clock components regulate entry and affect exit of seasonal dormancy as well as winter hardiness in populus trees. Plant Physiol 153:1823–1833 3. Nilsson O, Aldén T, Sitbon F, Anthony Little CH, Chalupa V et al (1992) Spatial pattern of cauliflower mosaic virus 35S promoter-luciferase expression in transgenic hybrid aspen trees monitored by enzymatic assay and non-destructive imaging. Transgenic Res 1:209–220 4. Eriksson ME, Israelsson M, Olsson O, Moritz T (2000) Increased gibberellin biosynthesis in transgenic trees promotes growth, biomass production and xylem fiber length. Nat Biotechnol 18:784–788 5. Dickmann DI, Isebrands JG, Eckenwalder JE, Richardson J (eds) (2002) Popular culture in North America. NRC Research Press, Ottawa, ON 6. Erickson RO, Michelini FJ (1957) The plastochron index. Am J Bot 44:297–305 7. Rohde A, Prinsen E, Rycke R, de Engler G, van Montagu M et al (2002) PtABI3 impinges on the growth and differentiation of embryonic leaves during bud set in poplar. Plant Cell 14:2975

8. Rohde A, Bastien C, Boerjan W (2011) Temperature signals contribute to the timing of photoperiodic growth cessation and bud set in poplar. Tree Physiol 31:472–482 9. Olsen JE, Junttila O, Moritz T (1997) Longday induced bud break in salix pentandra is associated with transiently elevated levels of GA1 and gradual increase in indole-3-acetic acid. Plant Cell Physiol 38:536–540 10. Rohde A, Bhalerao RP (2007) Plant dormancy in the perennial context. Trends Plant Sci 12:217–223 11. Ghelardini L, Santini A, Black-Samuelsson S, Myking T, Falusi M et al (2010) Bud dormancy release in elm (Ulmus spp.) clones – a case study of photoperiod and temperature responses. Tree Physiol 30:264–274 12. Perry TO (1971) Dormancy of trees in winter. Science 171:29–36 13. Welling A, Rinne P, Viherä‐Aarnio A, Kontunen‐Soppela S, Heino P et al (2004) Photoperiod and temperature differentially regulate the expression of two dehydrin genes during overwintering of birch (Betula pubescens Ehrh.). J Exp Bot 55:507–516 14. UPOV (1981) Guidelines for the conduct of tests for distinctness, homogeneity and stability in Poplar (Populus L), International Union for the Protection of New Varieties of Plants, Geneva, Switzerland. TG 21

324

Mikael Johansson et al.

15. Kalcsits LA, Silim S, Tanino K (2009) Warm temperature accelerates short photoperiodinduced growth cessation and dormancy induction in hybrid poplar (Populus x spp.). Trees Struct Funct 23:971–979 16. Olsen J, Junttila O, Nilsen J, Eriksson M, Martinussen I et al (1997) Ectopic expression of oat phytochrome A in hybrid aspen changes critical day length for growth and prevents cold acclimatization. Plant J 12:1339–1350 17. Tanino KK, Kalcsits L, Silim S, Kendall E, Gray GR et al (2010) Temperature-driven plasticity in growth cessation and dormancy development in deciduous woody plants: a

working hypothesis suggesting how molecular and cellular function is affected by temperature during dormancy induction. Plant Mol Biol 73:49–65 18. Kozarewa I, Ibáñez C, Johansson M, Ögren E, Mozley D et al (2010) Alteration of PHYA expression change circadian rhythms and timing of bud set in Populus. Plant Mol Biol 73:143–156 19. Arana MV, La Marin-de Rosa N, Maloof JN, Blazquez MA, Alabadi D et al (2011) Circadian oscillation of gibberellin signaling in Arabidopsis. Proc Natl Acad Sci U S A 108: 9292–9297

Chapter 22 Transformation and Measurement of Bioluminescence Rhythms in the Moss Physcomitrella patens Setsuyuki Aoki, Ryo Okada, and Santosh B. Satbhai Abstract Gene targeting is a highly effective and straightforward technique for the functional analysis of a gene of interest. However, its efficiency is not satisfactorily high in many model plants including Arabidopsis thaliana. In the moss Physcomitrella patens, a model species of basal plants, the efficiency of gene targeting is as high as in yeasts, and this moss is becoming widely recognized as an experimental model of choice in various areas of plant biology. Here we focus on the transformation of protoplast cells and on the measurement of bioluminescence rhythms from protonema tissues of luciferase reporter strains in P. patens, both of which are important for mechanistic studies of the circadian clock. Key words Physcomitrella patens, Moss, Gene targeting, Homologous recombination, Luciferase, Circadian rhythm

1

Introduction The mechanisms that generate circadian rhythmicity in land plant cells have largely been studied using Arabidopsis thaliana. Genetic and biochemical studies using this model dicot have revealed many component genes (clock genes) of the plant circadian system and their encoded products [1, 2]. Based on the structure of the regulatory network between these molecules, several mechanistic models of the A. thaliana circadian system have been published, in parallel with mathematical approaches to understand the dynamics which these hypothetical “plant circadian networks” generate [1–3]. On the other hand, homologues of clock genes and/or clock-related genes of A. thaliana have been isolated from many other species [1, 4]. In some cases, functional studies of the homologous genes have also been conducted using spontaneous as well as artificially induced mutants that showed aberrant rhythmicity [1, 4]. However, in any single species other than A. thaliana, systematic analyses of clock genes have not yet been performed, and no detailed mechanistic model based on experimental results is available. One reason for this

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_22, © Springer Science+Business Media New York 2014

325

326

Setsuyuki Aoki et al.

is the lack of an efficient transformation system and large-scale genome information in many plant species, which are important in the isolation and functional analysis of a gene of interest. Physcomitrella patens is a species of moss, which diverged from vascular plant lineages at least 450 Ma [5]. In 1997, Schaefer and Zrÿd [6] described efficient gene targeting in this moss, which occurs by homologous recombination between the moss genome and an externally introduced genomic DNA fragment. They reported that the efficiency of gene targeting in this moss exceeds 90 %, which matches those reported for yeasts [6]. P. patens has since then been used for many areas of plant biology ranging from basic cytological and developmental studies to plant phylogeny and genome evolution [7, 8], and it became the first nonangiosperm land plant to have its genome sequenced [9]. If the molecular mechanism of the circadian clock of P. patens is unraveled, it will shed light on the divergence, evolution, and origin of circadian systems in green plants. Moreover, comparisons of the circadian systems between plant species, as distantly related as A. thaliana and P. patens, would greatly contribute to an understanding of the machinery essential for the generation of circadian oscillation. Based on this reasoning, we established luciferase reporter strains in P. patens to monitor circadian gene expression as bioluminescence [10, 11]. Moreover, we performed functional analysis of some clock gene homologues by disrupting them through gene targeting based on homologous recombination in this moss ([11, 12]; SBS, unpublished data). In this manuscript, we summarize the methods for (1) transformation of moss cells, which is necessary to introduce reporter constructs to the moss genome or to disrupt genes related to clock functions, and (2) monitoring bioluminescence rhythms from moss tissues using the firefly luciferase gene as a reporter, which is highly suitable to analyze clock-controlled gene expression.

2

Materials We introduce DNA constructs into the moss nuclear genome by polyethylene glycol (PEG)-mediated transformation. Particle bombardment-mediated transformation is also applicable to P. patens [13, 14]. We use the PHYSCOmanual [15] (ver. 1.4; the latest version is 2.0) protocol kindly shared by the Hasebe lab, modifying its details according to the requirements and settings of our lab.

2.1 Transformation of Moss Cells

1. Stock A (100×): 0.5 M Ca(NO3)2, 4.5 mM FeSO4. Make up to 1 L with water in a glass bottle, sterilize by autoclaving, and store at room temperature. Wrap the glass bottle with aluminum foil to protect from light.

Transformation and Measurement of Bioluminescence Rhythms…

327

2. Stock B (100×): 0.1 mM MgSO4. Make up to 1 L with water, sterilize by autoclaving, and store at room temperature. 3. Stock C (100×): 1.84 mM KH2PO4, pH 6.5. Dissolve 25 g KH2PO4 in 900 mL water in a glass beaker. Adjust pH to 6.5 with 4 M KOH, make up to 1 L with water, sterilize by autoclaving, and store at room temperature. 4. Stock D (100×): 1 M KNO3, 4.5 mM FeSO4. Make up to 500 mL with water, and store at room temperature. Wrap the glass bottle with aluminum foil to protect from light. Use prior to coloration caused by the precipitation of iron. Even if coloration does not occur, discard the stock solution that has been kept over 3 months and make a new bottle of the solution. 5. Alternative TES (1,000×): 0.22 mM CuSO4, 10 mM H3BO3, 0.23 mM CoCl2, 0.1 mM Na2MoO4, 0.19 mM ZnSO4, 2 mM MnCl2, 0.17 mM KI. Make up to 1 L with water, sterilize by autoclaving, and store at room temperature. 6. Stock for ammonium tartrate (100×): 500 mM ammonium tartrate. Make up to 1 L with water, sterilize by autoclaving, and store at room temperature. 7. Stock for CaCl2 (50×): 50 mM CaCl2. Make up to 1 L with water, sterilize by autoclaving, and store at room temperature. 8. Germination agar medium (BCD + 10 mM Ca): 1× Stock B, 1× stock C, 1× stock D, 1× alternative TES, 5 mM ammonium tartrate, 10 mM CaCl2, 0.8 % (w/v) agar. Add 10 mL stock B, 10 mL stock C, 10 mL stock D, 1 mL alternative TES, and 10 mL ammonium tartrate stock to 900 mL deionized water in a glass media bottle. Weigh 1.5 g CaCl2 · 2H2O and 8 g agar (A6924, SIGMA-ALDRICH), and transfer them to the bottle. Make up to 1 L with water, and sterilize and melt the agar by autoclaving. Cool down to about 50 °C, mix, and pour the agar into Petri dishes (9 cm diameter). Dry the solidified agar by blowing air for 20 min with lids half open on a clean bench. Store the dishes in airtight plastic container at room temperature. 9. BCD medium (BCD + 1 mM Ca): 1× Stock B, 1× stock C, 1× stock D, 1× alternative TES, 1 mM CaCl2, 0.8 % (w/v) agar. Add 10 mL stock B, 10 mL stock C, 10 mL stock D, and 1 mL alternative TES to 900 mL water in a glass media bottle. Weigh 0.15 g CaCl2 · 2H2O and 8 g agar and transfer to the bottle. Make up to 1 L with water, and sterilize and melt the agar by autoclaving. Make and store the agar plates as in item 8. 10. BCDAT medium: 1× Stock B, 1× stock C, 1× stock D, 1× alternative TES, 5 mM ammonium tartrate, 1 mM CaCl2, 0.8 % (w/v) agar. Add 10 mL stock B, 10 mL stock C, 10 mL stock D, 1 mL alternative TES, and 10 mL ammonium tartrate stock to 900 mL water in a glass media bottle. Weigh 0.15 g CaCl2 · 2H2O and 8 g agar and transfer to the bottle.

328

Setsuyuki Aoki et al.

Make up to 1 L with water, and sterilize and melt the agar by autoclaving. Make and store the agar plates as in item 8. 11. BCDATG medium: 1× Stock B, 1× stock C, 1× stock D, 1× alternative TES, 5 mM ammonium tartrate, 1 mM CaCl2, 0.5 % (w/v) glucose, 0.8 % (w/v) agar. In addition to the stocks and reagents for making BCDAT medium, also add 5 g glucose to 900 mL water in a glass media bottle. Make up to 1 L with water, and sterilize and melt the agar by autoclaving. Make and store the agar plates as in item 8. 12. Mannitol solution: 8 % (w/v) mannitol. Sterilize by autoclaving, and store at room temperature. 13. Protoplast liquid medium: 1× Stock A, 1× stock B, 0.1× stock C, 0.005 % (w/v) ammonium tartrate, 6.6 % (w/v) mannitol, 0.5 % (w/v) glucose. Add 1 mL stock A, 1 mL stock B, 0.1 mL stock C, 1 mL 0.5 % (w/v) ammonium tartrate solution to 90 mL water in a glass media bottle. Weigh 6.6 g mannitol and 0.5 g glucose, transfer to the bottle, and mix. Make up to 100 mL with water, sterilize by autoclaving, and store at room temperature. 14. PEG/T solution: 20 % (w/v) polyethylene glycol 6000 (PEG6000), 7.2 % (w/v) mannitol, 0.1 M Ca(NO3)2, 10 mM Tris–HCl. Weigh 2 g PEG6000, transfer to a 20-mL glass vial along with a small magnetic stirrer, and melt by autoclaving. Add 9 mL mannitol solution (8 % (w/v)), 1 mL Ca(NO3)2 solution (1 M), and 100 μL Tris–HCl (1 M, pH 8.0) in a glass beaker, and mix by whirling. Sterilize this solution by filtration through a membrane filter (pore size, 0.22 μm), add to the PEG6000 solution in the 20-mL glass vial, and mix with a stirrer. 15. MMM solution: 9.1 % (w/v) mannitol, 15 mM MgCl2, 0.1 % (w/v) 2-morpholinoethanesulfonic acid (MES). Weigh 910 mg mannitol and transfer to a small glass beaker. Add 150 μL MgCl2 solution (1 M) and 1 mL MES solution (1 % (w/v); pH 5.6, adjusted with 0.1 N KOH) to the beaker, and mix by whirling. Sterilize the resulting solution by filtration through a membrane filter (pore size, 0.22 μm). Prepare this solution just before use on day 1 of transformation (step 20 of Subheading 3.1). 16. PRM/T solution: 1× Stock B, 1× stock C, 1× stock D, 1× alternative TES, 5 mM ammonium tartrate, 8 % (w/v) mannitol, 10 mM CaCl2, 0.8 % (w/v) agar. Add 2 mL stock B, 2 mL stock C, 2 mL stock D, 0.2 mL alternative TES, and 2 mL ammonium tartrate stock in a glass media bottle. Weigh 16 g mannitol, 0.29 g CaCl2 · 2H2O, and 1.6 g agar, and transfer to the bottle. Make up to 200 mL with water, sterilize and melt the agar by autoclaving, and keep at 45 °C until use.

Transformation and Measurement of Bioluminescence Rhythms…

329

17. PRM/B agar plates: 1× Stock B, 1× stock C, 1× stock D, 1× alternative TES, 5 mM ammonium tartrate, 6 % (w/v) mannitol, 10 mM CaCl2, 0.8 % (w/v) agar. Add 10 mL stock B, 10 mL stock C, 10 mL stock D, 1 mL alternative TES, and 10 mL ammonium tartrate stock in a glass media bottle. Weigh 60 g mannitol, 1.47 g CaCl2 · 2H2O, and 8 g agar and transfer to the bottle. Make up to 1 L with water, and sterilize and melt the agar by autoclaving. Make and store the agar plates as in item 8. 18. Driselase solution: Weigh 0.5 g Driselase (Kyowa Hakko, Tokyo) and transfer to a 50-mL plastic tube. Add 25 mL mannitol solution, and mix well. Centrifuge at 2,500 × g for 5 min. Filter the supernatant through a membrane filter (pore size, 0.45 μm) and transfer to a 50-mL glass centrifuge tube. 19. Homogenizer (Physcotron NS-310E, MICROTECH, Funabashi). 2.2 Monitoring Bioluminescence Rhythms from Moss Tissue

1. Bioluminescence monitoring machine [16] (see Note 1). 2. Stock solution for luciferin (100×): 100 mM D-luciferin, potassium salt. Add 785.2 μL sterilized water to a vial containing 25 mg D-luciferin and potassium salt, mix, divide into 10-μL aliquots in 1.5-mL microcentrifuge tubes, and store at −80 °C (see Note 2). 3. Luciferin solution for bioluminescence monitoring: Dilute the luciferin stock solution by adding 990 μL water to the 10-μL aliquot from item 2. 4. BCDATG agar plates for bioluminescence monitoring. Prepare BCDATG agar medium as item 12 of Subheading 2.1 except that smaller Petri dishes with 4-cm diameter are used instead of dishes 9 cm in diameter.

3

Methods Below is the protocol for PEG-mediated transformation using moss protoplast cells. Figure 1 helps to understand the protocol. Moss cells and tissues should be handled aseptically in a clean bench unless they are in sterilized agar plates sealed with surgical tape or a Parafilm strip. Sterilize all the tools for handling moss tissue, such as forceps, by autoclaving. The procedure for preparation of bioluminescent moss tissue samples and setting the samples to the bioluminescence monitoring machine is also described.

3.1 Transformation of Moss Cells

1. Overlay a sterilized cellophane sheet on the germination agar medium with forceps (see Note 3). 2. Immerse several sporangia in 1 mL of antiformin in a 1.5-mL microcentrifuge tube, and mix gently by slowly inverting the

330

Setsuyuki Aoki et al. Propagation of protonema germinated from spores Subculture cycles of protonema Day 1. Isolation of protoplasts • Introduction of DNA into • protoplasts Day 2. • Start of regeneration of protoplasts on agar medium (culture without antibiotics) Day 5. • Start of the selection period 1 (culture with antibiotics) Day ~26. • Start of the relaxation period (culture without antibiotics) Day ~33. • Start of the selection period 2 (culture with antibiotics) Day ~40. • Isolation of candidate stable transformants

Fig. 1 Flow diagram of the transformation protocol. See Subheading 3.1 of the main text for details

tube for 3 min for sterilization. After the sporangia sink to the bottom of the tube, discard the supernatant. 3. Add 1 mL sterilized water into the tube, and mix gently for 3 min. After the sporangia sink, discard the supernatant. 4. Repeat step 3 three times. 5. Add 1 mL sterilized water into the tube, and crush the sporangia with the Pipetman tip to disperse spores in water. 6. Transfer the dispersed spores onto four germination agar plates, each containing 0.25 mL of spore suspension. 7. Incubate at 25 °C under continuous light. Light intensity is ~50 μmol/m2/s throughout the entire culture of protonema tissue. Incubation for 1–2 weeks is enough to obtain sufficient protonema tissue for subculture. 8. Start subculture by collecting protonema tissue grown on germination agar plates (or BCDATG agar plates in the case of the second or the later round of subculture) with forceps and transfer into a glass tube (see Note 4). Add water (2 mL for each new agar plate) into the tube. 9. Shred the tissue with Physcotron for about 10 s (see Note 4). Use the shaft NS-4 or NS-7 at a rotation speed of 4–5.

Transformation and Measurement of Bioluminescence Rhythms…

331

Transfer the fragmented tissue onto new BCDATG agar plates, pouring 2 mL protonema suspension onto each plate with a Komagome pipette. 10. Grow protonema tissue under continuous light until when water on the BCDATG agar medium has dried (4–5 days). 11. Repeat steps 8–10 until enough protonema tissue is obtained for transformation. Usually protonema tissue collected from 2 to 3 agar plates is fragmented and spread onto 6–10 plates (see Note 5). 12. (Day 1 of transformation) Collect the protonema tissue grown on BCDATG agar plates for 3–5 days (usually ~12 plates are used for four DNA samples) and transfer to Driselase solution in a 50-mL glass centrifuge tube (see Note 4). 13. Wrap the tube with aluminum foil and incubate at 25 °C for 30 min. Mix gently by slowly rotating the tube by hand once every 5 min, and check if protoplast cells have detached from protonema tissue. 14. Transfer the protoplast suspension into a new 50-mL glass centrifuge tube, filtering through a nylon mesh sheet (pore size, 40 μm) to obtain protoplast cells finely separated from each other. 15. Centrifuge at 180 × g for 2 min (see Note 6). 16. Discard the supernatant with a Komagome pipette, leaving 2–3 mL Driselase solution at the bottom of the tube. 17. Resuspend protoplast cells in the remaining Driselase solution. 18. Add 40 mL mannitol solution (see Note 7). 19. Repeat steps 15–17 twice. 20. Count protoplast cells in 40-mL mannitol solution with a hemocytometer and resuspend in MMM solution at 1.6 × 106 cells/mL. 21. Add 30 μL plasmid DNA (see Note 8), 300 μL protoplast suspension in MMM solution, and 300 μL PEG/T solution in a 15-mL conical tube, and mix gently by swirling. 22. Incubate DNA–protoplast mixture at 45 °C for 5 min and then at 20 °C for 10 min. 23. Dilute DNA–protoplast mixture by adding 300 μL protoplast liquid medium every 3 min five times and then adding 1 mL protoplast liquid medium every 3 min five times. 24. Transfer diluted DNA–protoplast mixture into a small plastic dish, wrap the plate with aluminum foil, and incubate at 25 °C overnight in the dark. 25. (Day 2) Transfer DNA–protoplast mixture into a 15-mL conical tube with Pipetman (P-1000). Centrifuge at 180 × g for 2 min at room temperature, and discard the supernatant (see Note 6).

332

Setsuyuki Aoki et al.

26. Add 8 mL autoclaved PRM/T medium at ~45 °C into the conical tube, and gently resuspend the spun-down protoplasts (see Note 9). 27. Transfer PRM/T medium containing protoplasts onto four PRM/B agar plates overlaid with sterilized cellophane sheets, pouring 2 mL suspension onto each plate (see Note 9). 28. Incubate at 25 °C under continuous light for 3 days. 29. (Day 5) Transfer regenerated protoplasts and the underlying cellophane sheet with forceps onto a BCDAT agar plate supplemented with appropriate concentration of antibiotics (see Note 10). 30. Incubate at 25 °C under continuous light for 2 weeks. 31. Transfer protonema colonies with the underlying PRM/T medium and cellophane sheet using forceps onto a new BCDAT agar plate supplemented with appropriate concentration of antibiotics (see Note 11). 32. Incubate at 25 °C under continuous light for 1 week. 33. (Day ~26) Transfer the surviving protonema colonies, one by one, with forceps onto antibiotic-free BCDAT agar medium. 34. Incubate at 25 °C for ~1 week under continuous light. 35. (Day ~33) Transfer a small fragment of each protonema colony with forceps onto BCDAT medium supplemented with appropriate concentration of antibiotics. 36. Incubate at 25 °C for 1 week under continuous light. 37. (Day ~40) Pick up protonema colonies that have grown on selection medium as candidate stable transformant clones (see Note 12). 3.2 Monitoring Bioluminescence Rhythms from Moss Tissues

1. Obtain fresh protonema tissue of a P. patens reporter strain on BCDATG agar medium by repeating subculture cycles (steps 8–10 of Subheading 3.1) under continuous light. During the last round of subculture cycles, grow tissue under 12-h light:12-h dark cycles (12:12LD; light intensity of the 12-h light period is 40 μmol/m2/s) to synchronize the clocks in protonema cells. 2. Shred protonema tissue as described in step 9 of Subheading 3.1, and pour fragmented tissue onto BCDATG agar plates for bioluminescence monitoring. 3. Incubate the plates at 25 °C under 12:12LD for 3 days. 4. Pour 100 μL luciferin solution for bioluminescence monitoring on top of propagated protonema tissue in each plate (see Note 13). 5. Set plates on the bioluminescence monitoring machine [16] and initiate monitoring in continuous darkness.

Transformation and Measurement of Bioluminescence Rhythms…

4

333

Notes 1. A compact bioluminescence monitoring machine developed for small-scale but flexible monitoring of gene expression from various organisms is used [16] although other automatic machines that enable accurate and long-term monitoring of bioluminescence from reporter organisms exist [17, 18]. There are also commercially available systems (http://www.atto.co. http://www.churitsu.co.jp/ jp/eng/prodct_Kronos.html, products/index.html). 2. To make the correct concentration of luciferin stock solution, we avoid weighing luciferin powder but add water directly into the vial containing D-luciferin provided by Life technologies (L-2916). D-luciferin concentration strongly affects average levels of bioluminescence from P. patens reporter strains but does not affect relative bioluminescence patterns. 3. Prepare the sterilized cellophane sheets as follows. Cut out circles that are slightly smaller in size than the plastic plate from a cellophane sheet. Pile up the circular cellophane sheets in a glass Petri dish, and rinse the sheets with deionized water several times. Add ~10 mL deionized water in the Petri dish, wrap the dish with aluminum foil, and sterilize by autoclaving. 4. Take care of sterilization in these steps; otherwise, contamination can occur. When collecting protonema tissue from an agar plate with forceps, do not use the tissue growing near the rim as much as possible to avoid contamination by bacteria or fungi. Sterilize the homogenizer shaft by wrapping in aluminum foil and autoclaving before shredding protonema tissue. Sterilize the motor (and the main body), to which the shaft is attached, by wiping well with 70 % ethanol and exposing to UV light in the clean bench. Be careful not to touch the cutting edge of the shaft when attaching to the motor. Instead of a homogenizer such as Physcotron and Polytron, ordinary mortar and pestle can be used to shred protonema tissue, but the possibility of contamination may be higher. For vigorous growth, protonema tissue should be homogenously fragmented and evenly spread on the entire cellophane sheet. 5. When protonema tissue that has been grown continuously on a same agar medium for a long time or kept at a low temperature is used as starting material, it takes more subculture cycles to get enough vigorous protonema tissue than when using fresh tissue. Therefore, we recommend routine subculture so that fresh tissue is always available. In addition, do not use protonema tissue that has been subcultured more than ~8 times from the time of spore germination to avoid accumulation of somatic mutations.

334

Setsuyuki Aoki et al.

6. Centrifuge with the break function off so that quick changes in centrifugal force do not damage protoplast cells. 7. Add mannitol solution gently and slowly along the wall of the tube to avoid damaging protoplast cells. 8. In P. patens, transformed DNA containing homologous sequences at both termini integrates predominantly by homologous recombination with its target site on the genome. Therefore, plasmid DNA must be linearized with a restriction enzyme(s) at appropriate positions to conduct gene targeting. We usually cut 30 μg plasmid DNA, extracted with a kit (Qiagen, midi plasmid kit), with enough restriction enzyme overnight. To confirm whether DNA digestion is complete or not, check a small part of the reaction by agarose electrophoresis. Purify DNA by phenol/ chloroform extraction followed by ethanol precipitation. After the remaining ethanol dries up, dissolve DNA in 30 μL sterilized TE (10 mM Tris–HCl, 1 mM EDTA, pH 8.0) buffer and store at −20 °C until use for transformation. 9. To obtain protoplasts uniformly resuspended in PRM/T medium, mild pipetting (adding and aspirating PRM/T medium once) is enough. Repetitive pipetting is not required. After pouring on PRM/B plate, PRM/T agar solidifies rapidly. Therefore, rotate the plate quickly to evenly distribute the PRM/T medium across the entire surface of the bottom agar, before it solidifies: it is helpful to prewarm PRM/B plates at 37 °C. 10. Make concentrated stock solution (1,000×) of each antibiotic with sterilized water, add 1/1,000 volume into the agar medium, which was autoclaved and allowed to cool to ~50 °C, mix, and pour into plastic Petri dishes. Final concentration of antibiotics is 30 mg/L for Hygromycin B (Wako Pure Chemical Industries), Geneticin (Gibco), and zeocin (Invitrogen). 11. Step 31 can be omitted. However, if omitted, the number of transient transformant colonies at the beginning of the relaxation period (step 33) will greatly increase. 12. The selection period in the abovementioned transformation protocol eliminates transient transformants and thereby allows selective isolation of stable transformants. However, the stable transformants include not only clones in which the DNA construct was integrated to the target site on the genome by homologous recombination but also those in which the DNA construct was randomly inserted in the genome by nonhomologous recombination. Therefore, we need to check if the DNA construct was introduced to the target site as designed by genomic Southern blotting analysis and genomic polymerase chain reaction (PCR). Details are described in the caption of Fig. 2, which shows an example of the results of the Southern and PCR analyses.

335

Transformation and Measurement of Bioluminescence Rhythms…

a

P1 HindIII

HindIII

213L

213R P2

P1 P4

HindIII

PCCA1b

213L

Luc+

Tnos

E7

hpt

WIn

Tnos

P3

P2

c 50 75

50 75

4.4 2.0

213R 1kbp

b (kb) M

HindIII

1 2 (kb) 3.7

M

3

4

(kb)

(kb) 23.0 9.4 6.6 1.7 4.4 2.0

M

50

75 WT

5

6

7

(kb)

WT 50

75

(kb)

(kb)

(kb) 23.0

23.0

9.4 10.1 6.6 3.2 4.4

11.2 9.4 6.6 4.4

2.3

2.3

WT 50

75

(kb)

11.2

3.7

Fig. 2 Analyses of the genomes of the PCCA1b::LUC+ reporter strains. (a) The genomic DNA structure of the 213 loci in the wild type (WT) (upper figure) and the PCCA1b::LUC+ reporter strains (lower figure) of P. patens. We used this 213 locus (a kind gift from Jean-Pierre Zrÿd) as a target site to introduce reporter constructs into the moss genome. Thin lines of both ends represent the genomic DNAs outside of the 213 loci. Closed boxes represent the 213 region, which is divided into 213L and 213R by introduction of the reporter construct. The reporter construct, corresponding to the lower figure except for thin lines and closed boxes of both ends, is a fusion of the firefly luciferase gene (Luc+) driven by the PpCCA1b gene promoter (PCCA1b) and the hygromycin resistance gene (hpt) driven by an artificial strong promoter (E7ΩIn). Tnos indicates the terminator of the nopaline synthase gene (nos). Small arrows show primers which were used for PCR experiments. Small grey rectangles show DNA probes for Southern experiments specific to the 213L region or LUC+. (b) Results of genomic PCR analyses to confirm homologous recombination events. By this analysis, we selected from the regenerating transformants several clones that carried a single copy of the reporter construct integrated only in the 213 site. In these selected clones, DNA fragments with expected sizes were amplified with various combinations of primers derived from flanking regions of the 213 site and from the reporter construct. Shown are the results of two clones #50 and #75, indicated above lane numbers, which showed the same results. The primer combinations are lanes 1 and 2, P1 and P3; lanes 3 and 4, P2 and P4; lanes 5, 6, and 7, P1 and P2. The results of PCR on the genomic DNA from the WT strain (carrying no reporter gene) using primer set P1/P3 or P2/P4 are not displayed because they showed no DNA band. DNA size markers (M) are HindIII digests of Lambda DNA. The sizes of the amplified bands predicted from (a) are indicated on the right of the photos. (c) Results of Southern analyses performed using the LUC+ (left photo) and 213L (right photo) probes shown in (a). Genomic DNAs (0.5 μg for each lane) from the WT strain (carrying no reporter gene) #50 and #75 were digested with HindIII, separated by agarose gel electrophoresis, and used for blotting. In the Southern analysis, (1) a single band with a predicted size (~11.2 kb) was observed only for the transformant clones but not for the WT strain when using the LUC+-specific probe and (2) a predicted band shift was observed between the selected clones and the wild-type strain when using the 213 site-specific probe (reproduced from [11] with permission from John Wiley and Sons)

336

Setsuyuki Aoki et al.

13. The luciferin solution is applied to the tissue sample immediately before starting monitoring bioluminescence. The timing of application of the luciferin solution does not affect the pattern of bioluminescence from P. patens reporter strains (Aoki, unpublished data).

Acknowledgement We thank Mitsuyasu Hasebe and members of his laboratory (National Institute for Basic Biology, Okazaki) for kindly sharing the PHYSCOmanual protocol and various materials throughout our study using P. patens. We also thank Jean-Pierre Zrÿd (University of Lausanne) for kindly sharing a plasmid containing the 213 locus and Kyowa Hakko for kindly sharing Driselase. This work was supported by grants (21570005 and 24570007) from the Japan Society for the Promotion of Science to SA. References 1. McClung CR (2011) The genetics of plant clocks. Adv Genet 74:105–139 2. Nagel DH, Kay SA (2012) Complexity in the wiring and regulation of plant circadian networks. Curr Biol 22(16):R648–R657 3. Bujdoso N, Davis SJ (2013) Mathematical modeling of an oscillating gene circuit to unravel the circadian clock network of Arabidopsis thaliana. Front Plant Sci 4:3 4. Song YH, Ito S, Imaizumi T (2010) Similarities in the circadian clock and photoperiodism in plants. Curr Opin Plant Biol 13(5):594–603 5. Lang D, Zimmer AD, Rensing SA et al (2008) Exploring plant biodiversity: the Physcomitrella genome and beyond. Trends Plant Sci 13: 542–549 6. Schaefer DG, Zrÿd JP (1997) Efficient gene targeting in the moss Physcomitrella patens. Plant J 11(6):1195–1206 7. Cove D (2005) The moss Physcomitrella patens. Annu Rev Genet 39:339–358 8. Knight C, Perround P-F, Cove D (eds) (2009) The moss Physcomitrella patens. WileyBlackwell, London 9. Rensing SA, Lang D, Zimmer AD et al (2008) The Physcomitrella genome reveals evolutionary insights into the conquest of land by plants. Science 319(5859):64–69 10. Aoki S, Kato S, Ichikawa K et al (2004) Circadian expression of the PpLhcb2 gene encoding a major light-harvesting chlorophyll a/b-binding protein in the moss Physcomitrella patens. Plant Cell Physiol 45(1):68–76 11. Okada R, Kondo S, Satbhai SB et al (2009) Functional characterization of CCA1/LHY

12.

13.

14.

15.

16.

17.

18.

homolog genes, PpCCA1a and PpCCA1b, in the moss Physcomitrella patens. Plant J 60(3):551–563 Satbhai SB, Yamashino T, Okada R et al (2011) Pseudo-response regulator (PRR) homologues of the moss Physcomitrella patens: insights into the evolution of the PRR family in land plants. DNA Res 18(1):39–52 Cho SH, Chung YS, Cho SK et al (1999) Particle bombardment mediated transformation and GFP expression in the moss Physcomitrella patens. Mol Cells 9(1):14–19 Ichinose M, Sugita C, Yagi Y et al (2013) Two DYW subclass PRR proteins are involved in RNA editing of ccmFc and atp9 transcripts in the moss Physcomitrella patens: first complete set of PPR editing factors in plant mitochondria. Plant Cell Physiol 54(11): 1907–1916 Hiwatashi Y, Fujita T, Sato Y et al (2008) PHYSCOmanual, version 1.4. http://www. nibb.ac.jp/~evodevo/PHYSCOmanual_ v14/00Eindex.htm. Accessed on March 7 2013 Okamoto K, Ishiura M, Torii T et al (2007) A compact multi-channel apparatus for automated real-time monitoring of bioluminescence. J Biochem Biophys Methods 70(4):535–538 Okamoto K, Onai K, Ezaki N et al (2005) An automated apparatus for the real-time monitoring of bioluminescence in plants. Anal Biochem 340(2):187–192 Okamoto K, Onai K, Ishiura M (2005) RAP, an integrated program for monitoring bioluminescence and analyzing circadian rhythms in real time. Anal Biochem 340(2):193–200

Chapter 23 Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO Christoph Schmal, Jean-Christophe Leloup, and Didier Gonze Abstract Circadian clocks are endogenous timekeepers that produce oscillations with a period of about one day. Their rhythmicity originates from complex gene regulatory networks at the cellular level. In the last decades, computational models have been proven to be a powerful tool in order to understand the dynamics and design principles of the complex regulatory circuitries underlying the circadian clocks of different organisms. We present the process of model development using a small and simplified two-gene regulatory network of the Arabidopsis circadian clock. Subsequently, we discuss important numerical techniques to analyze such a mathematical model using XPP-AUTO. We show how to solve deterministic and stochastic ordinary differential equations and how to compute bifurcation diagrams or simulate phase-shift experiments. We finally discuss the contributions of modeling to the understanding and dissection of the Arabidopsis circadian system. Key words Circadian clocks, Feedback loops, Computational models, Ordinary and Stochastic Differential Equations, Limit-cycle oscillations, Bifurcation diagrams, XPP-AUTO

1

Introduction The Earth’s celestial mechanics faces its inhabitants with periodically changing environmental conditions such as 24h light–dark or temperature cycles. In order to anticipate these environmental changes and to optimally schedule physiological and behavioral processes, nearly all species, ranging from bacteria [1], fungi [2, 3], insects [4], and plants [5] to mammals [6], have evolved an endogenous timekeeper, the circadian clock [7]. This pacemaker is able to generate robust oscillations in gene expression that persist even under constant conditions of light and temperature with a period of approximately 24 h. Specific periodic environmental stimuli, known as zeitgeber signals, such as light–dark or temperature cycles are able to entrain the clock to the period of the Earth’s rotation. The molecular mechanisms underlying these circadian clocks are commonly described for most organisms as multiple interlocked

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7_23, © Springer Science+Business Media New York 2014

337

338

Christoph Schmal et al.

regulatory feedback loops, usually including transcriptional as well as post-transcriptional or post-translational steps [8, 9]. Astonishingly enough, circadian rhythms solely governed by posttranslational regulation were discovered in cyanobacteria recently [10] and, on top of that, could be reconstructed in vitro [11]. The dynamics of such complex regulatory networks which are species-specific and involve regulatory interactions of various kinds are hard to apprehend intuitively. Computational models helped in many cases to untangle the design principles underlying the circadian clocks and led to a better understanding of their experimentally observed complex dynamical behavior. A shining bright example, showing the power of computational modeling, is the recent progression in the dissection of the Arabidopsis thaliana circadian clockwork. As it will be discussed in more detail in Subheading 4, new experimental findings and model improvements were always closely associated. However, even in times before anything was known about the molecular springs and levers comprising the circadian clockwork, ordinary differential equation models based on generic or physical oscillators (e.g., phase-models, Poincaré, or Van der Pol oscillators) were used to explain the observed experimental data (see, e.g., [12–14]). They are still in use today in studies rather focusing on generic properties of the circadian system, such as synchronization or entrainment properties (see, e.g., [15–17]) than on breaking the clockwork down to its molecular organization. Once the molecular mechanisms underlying the circadian clocks were identified, detailed mathematical models made a decisive contribution to the understanding and dissection of the circadian networks of the fruit fly Drosophila melanogaster [18], the red bread mould Neurospora crassa [19], the cyanobacterium Synechococcus elongatus [20], mammals [21, 22], the small flowering plant Arabidopsis thaliana [23–29], and the unicellular green algae Ostreococcus tauri [30]. Most of these molecular models take the form of a set of coupled ordinary differential equations that describe the time evolution of the concentration of clock gene mRNA and proteins. These models are by definition deterministic and can be used to explore the molecular origin of the circadian oscillations, to predict the effect of various mutations, and to account for entrainment by light–dark cycles. Due to the inherent probabilistic nature of biochemical reactions and the variability of external influences (light, temperature, etc.), noise effects may play an important role in the behavior of circadian clocks and were studied by various stochastic models for several organisms [31–34]. Since more and more experimental data on circadian clocks becomes available, the corresponding clock models quickly grow in their number of dynamical variables and parameters.

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

339

Boolean Modeling, long known as a useful tool to analyze the dynamics on large gene regulatory networks [35], were successfully applied to describe circadian oscillations in Neurospora and Arabidopsis recently [36]. In this logical description of the dynamics, the expression value of a given gene is described by Boolean values (on or off) and their mutual regulation is described by Boolean functions. This reduces the dimension of the parameter space tremendously while it is, on the other hand, accompanied by a loss of biochemical detail. In the following, we will describe step-by-step how to develop, simulate, and analyze a molecular clock model based on ordinary and stochastic differential equations (ODEs and SDEs, respectively). In Subheading 2, we will give a detailed description of how to model certain reaction steps in a simple two-gene clock model. Subheading 3 will then show how one can simulate the equations, developed in Subheading 2, using the numerical tool XPP-AUTO. On top of that, we will learn how XPP-AUTO can be used to apply a bifurcation analysis and how we can simulate models incorporating noise terms. After an introduction to these basic techniques, Subheading 4 will summarize the evolution of the Arabidopsis thaliana circadian clock with an emphasis on the important contributions dynamical systems modeling made to the understanding of this clockwork.

2

Modeling the Arabidopsis Circadian Clock The first model of the Arabidopsis circadian clock, published by Locke et al. in 2005 [23], was based on a single negative feedback loop, where two partially redundant MYB transcription factors LATE ELONGATED HYPOCOTYL (LHY) and CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) inhibit the transcription of their transcriptional activator TIMING OF CAB EXPRESSION 1 (TOC1). Albeit the activating role of TOC1 has been questioned recently [28, 37] and though the real structure of the Arabidopsis clockwork is far more complex, we will use a toy model (similar to Locke’s 2005 model [23]) based on this two-gene negative feedback assumption hereinafter for the sake of simplicity and didactic reasons. Our model is illustrated in Fig. 1. In this simplified clock model we lump together LHY and CCA1 into one variable (as it is also done by all other Arabidopsis clock models published so far) and consider only the mRNA concentration Mi(t) and protein concentration Pi(t) of LHY/CCA1 (i = L) and TOC1 (i = T). In particular, we do not take into account any compartmentalization of the cell (trafficking between nucleus and cytoplasm) and no further post-transcriptional or posttranslational processes apart from translation and the degradation

340

Christoph Schmal et al.

Fig. 1 Schematic representation of our simple two-gene negative feedback loop model. Numbers indicate biochemical reactions as described in the correspondingly numbered paragraph of Subheading 2

of mRNAs and proteins. We model the temporal evolution of the dynamical variables ML,PL, MT, and PT as follows: Pa mM dM L = L (t ) + v1 a T a − 1 L g 1 + PT k1 + M L dt

(1)

dPL mP = p1M L − 2 L k2 + PL dt

(2)

mM gb dM T = v2 b 2 b − 3 T g 2 + PL k3 + M T dt

(3)

dPT mP = p2 M T − 4 T . k4 + PT dt

(4)

Concentrations are typically expressed in units of nanomolar (nM) and time in units of hours (h) [23]. The specific kinetic functions applied in this model are based on the subsequent considerations: 1. Transcriptional Regulation: The effect of transcription factors on the transcription rate of its target gene is commonly described by means of Hill functions: In the absence of LHY/ CCA1 proteins, i.e., PL = 0, TOC1 is transcribed by its maximal transcription rate v2. In presence of LHY/CCA1 protein

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

a

0.5

b

0 0

0.5

1

H + (PT ,g1 ,a)

H – (PL ,g2 ,b)

1

b=1 b=2 b= 4

∞

a

1

b

341

∞ 4 a= =2 a a=1

0.5

0 1.5

0

2

Relative Repressor Concentration PL / g2

0.5

1

1.5

2

Relative Activator Concentration PT / g1

−

Fig. 2 (a) Hill repressor function H for the Hill coefficients b = 1, 2, 4, and b → ∞. (b) Hill activator function H+ for the Hill coefficients a = 1, 2, 4, and a → ∞

(i.e., PL ≠ 0), the rate of TOC1 transcription is reduced to a fraction H − ( PL , g 2 , b ) ∈ [0, 1] of v2. The Hill repressor function H − ( PL , g 2 , b ) :=

g 2b g 2b + PLb

(5)

is a monotonically decreasing function of sigmoidal shape that converges to zero for high repressor concentrations (see Fig. 2a). g2 is called inhibition constant and marks a threshold of necessary LHY/CCA1 protein concentration PL in order to repress the transcription of TOC1 notably. The Hill coefficient b determines the steepness of the Hill function and generally accounts for cooperative binding of transcription factors to their target promoters. Similar considerations apply for the kinetics of the transcriptional activation of LHY/CCA1 transcription by TOC1 proteins PT: In this case the corresponding Hill activator function H + ( PT , g 1 , a ) := 1 − H − ( PT , g 1 , a ) =

PTa g 1a + PTa

(6)

is of sigmoidal and monotonically increasing shape (see Fig. 2b). The transcription rate of LHY/CCA1 saturates to a maximal expression value v1 in case when TOC1 protein concentrations PT are much larger than the activation constant g1. The Hill functions H+ and H− can be seen as the probability that the transcription factors are bound or unbound to the promoter of their target gene, respectively [38]. This interpretation allows us to extend the formalism to situations when multiple transcription factors affect the transcription rate of a given target gene. If, for example, an activator X and an inhibitor Y bind to a promoter independently, the target gene will be transcribed when X binds while Y does not. The probability that X binds to the promoter and Y does not is then given by the probability P(X and not Y ) = H + (X , kx , n x ) × H − (Y , ky , n y ).

(7)

342

Christoph Schmal et al.

The resulting promoter activity or transcription rate is then given by the product vP(X and not Y), where v denotes the maximal transcription rate. 2. Translation: The translation of LHY/CCA1 mRNA (ML) and TOC1 mRNA (MT) into protein (gain terms of Eqs. 2 and 4 in Fig. 1) is assumed to rely on first-order or linear kinetics. Proteins are therefore produced at a rate that depends on the translation rates p1, p2 and the actual mRNA concentrations ML and MT for LHY/CCA1 and TOC1, respectively. 3. Degradation: For all dynamical variables, degradation is assumed to follow Michaelis–Menten kinetics (last terms in Eqs. 1–4). Such an approach accounts for saturation in an enzyme-driven degradation process. In contrast to linear degradation, the Michaelian degradation rate then converges to a maximal value mj for increasing concentration values. Nonlinearity is well known as a necessity in order to generate self-sustaining limit-cycle oscillations. In models using linear degradation and translation kinetics, high Hill coefficients may be needed to reach the required nonlinearity [39, 40]. Using Michaelis–Menten kinetics for degradation processes raises the possibility of obtaining limit-cycle oscillations with lower, biologically more common, Hill coefficients but also has some “risk”: It contains the possibility of diverging solutions (i.e., the concentrations may shoot to infinity for t → ∞) if the maximal degradation rate under completely saturated conditions is not sufficient to compensate the production of a given species. 4. Light Input: Light is an important zeitgeber signal for plant circadian clocks. The effect of light is usually modeled through the modulation of one or several light-controlled parameters. A light–dark (LD) cycle can be modeled by the periodic modulation of the light-controlled parameters. In our simplified model the effect of light enters Eq. 1 via a light-input function L(t). This function will change its value in a binary on–off fashion, following a specific periodic pattern and will be described in more detail in Subheading 3.4. 5. Remark On More Sophisticated Models: Our simplified model is based on the mutual transcriptional regulation of LHY/CCA1 and TOC1, light-regulation of LHY/CCA1 transcription as well as translation and degradation processes. Other—more elaborated—models take into account important aspects like compartmentalization of the cell (distinguishing between cytoplasmatic and nuclear protein concentrations) [23], light-regulated degradation [24], and post-translational modifications such as protein complex formation [28] or phosphorylation [41]. 6. Parameter Estimation: Even our simple two-gene model contains 16 parameters and the dimension of parameter space

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

343

quickly grows with increasing model detail. In many biological models, some or even all of these parameters are not known from experiments. Finding parameter sets that account for known experimental data is a classical inverse problem and there are two conceptually different ways how to deal with this issue: (1) On the one hand, the modeler can try to find “good” parameter sets “by hand” in a trial-and-error fashion. In doing so, the modeler gets a good feeling and intuition for the system’s behavior but it may become a cumbersome task for increasingly detailed models. (2) On the other hand, many studies used an automated and computer-based sampling and evaluation method in order to find “good” solutions (see, e.g., [23]). The quality of a given, previously sampled parameter set is then determined by a test function that quantifies the overlap between the numerical solution and known experimental data. These experimental data could be time traces (χ2 function fitting, see, e.g., [42]) or qualitative features such as the phase or the period of an oscillating solution (so-called cost function fitting, see, e.g., [23]).

3

Simulating the Arabidopsis Circadian Clock

3.1 Installation of XPP-AUTO

XPP-AUTO was developed by Bart Ermentrout (University of Pittsburgh) and is widely used by modelers in biology [43]. It is a stand-alone software, free for academic users. It runs on all types of computer (under Windows you will need to install an X-server). The program as well as the documentation and some demonstration files can be found here: http://www.math.pitt.edu/~bard/xpp/xpp.html. XPP-AUTO is designed to simulate, analyze, and visualize dynamical systems. It offers a large choice of ODE solvers (and uses by default a fourth-order Runge-Kutta (RK4) method with adaptive step size). Its graphical interface allows the user to display time series or trajectories in the 2D and 3D phase space and to study the effect of changes in parameter values and initial conditions. XPP-AUTO also incorporates AUTO (developed by E. Doedel), which computes bifurcation diagrams using advanced continuation methods [44–46]: http://indy.cs.concordia.ca/auto/

3.2 Creating an ODE File

Before launching XPP-AUTO, you have to prepare a text file (with the extension.ode) that contains the equations of your model, as well as the values of the parameters. Although not mandatory, it is recommended to define the initial conditions. Otherwise, the initial values are set to zero. You may also specify in the ODE file some integration parameters, such as the total time of integration

344

Christoph Schmal et al.

(“total”), the time step (“dt”), and the integration method (“method”), as well as plotting parameters. Here is the content of the ODE file for the toy, 2-variable model, described above. In this first version, we only focus on the interaction between LHY/CCA1 and TOC1 and consider only the mRNA and protein levels. In particular, we do not consider the effect of light (tantamount to constant darkness (DD) condition with the light function L set to zero). # LHY/CCA1-TOC1 model (Arabidopsis) # Equations dmlhy/dt=L+v1*ptoc^a/(g1^a+ptoc^a)-m1*mlhy/ (k1+mlhy) dplhy/dt=p1*mlhy-m2*plhy/(k2+plhy) dmtoc/dt=v2*g2^b/(g2^b+plhy^b)-m3*mtoc/ (k3+mtoc) dptoc/dt=p2*mtoc-m4*ptoc/(k4+ptoc) # Parameters param L=0 param v1=0.3, param p1=0.5, param v2=0.6, param p2=0.3,

a=2, g1=0.5, m1=0.4, k1=1 m2=0.6, k2=0.5 b=2, g2=0.1, m3=0.6, k3=1 m4=0.3, k4=1

# Initial conditions ini mlhy=0.1, plhy=0.1 ini mtoc=0.1, ptoc=0.1 # Integration parameters @ total=120, dt=0.05 # Plotting parameters @ nplot=2, yp1=mlhy, yp2=mtoc done 3.3 Time Series and Phase Space

XPP-AUTO can be launched either by clicking on the program icon or by typing the following command in a terminal: > xppaut A window then opens, inviting you to select your ODE file and allowing you to browse your directories. Alternatively, you may also directly specify the name and location of your ODE file when launching XPP-AUTO: > xppaut arabidopsis.ode The main XPP-AUTO window then opens. On the left appear various menus. Note that the name of each menu is written with one capital letter. This letter is a keyboard shortcut. Here below they are noted with square brackets. On the top of the XPP-AUTO window, there are a few buttons allowing you to open some additional

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

345

windows. On the bottom, three windows allow you to change the value of three parameters using a sliding cursor. You can then proceed with the following instructions. These lines give you a rapid overview of the possibilities offered by XPPAUTO and are far from exhaustive. 1. Click on Initialconds [I], and, in the submenu that appears, on “(G)o” [G]. The equations are integrated and the time series are displayed on the screen: The variable mlhy is shown in black and the variable mtoc is in red. 2. To fit axes scales to the computed time series, click on Window/ zoom [W] and then on “(F)it” [F]. You then observe the circadian oscillations of mlhy and mtoc. They seem slightly damped. It is thus necessary to check if they stabilize or if they are really damped. 3. Type [E] to Erase (clear) the screen. Click then on Initialconds [I], and, in the submenu that appears, on “(L)ast” [L]. The equations are integrated using as initial conditions the last point reached in the previous simulation. Type [W][F] to fit the axes to the size of the window. You can now verify that the oscillations stabilize. 4. Click on nUmerics [U], and, then on “Total” [T]. You can then give a new value for the total integration time. You can set Total, for example, to 240. Confirm by typing [ENTER] and quit the “nUmerics” menu by typing [ESC]. 5. To add a curve on the plot, click on Graphic stuff [G], and then on “(A)dd curve” [A]. As Y-axis, enter the variable you wish to add (e.g., plhy). Colors are coded by numbers. You can select, for example, “3” (yellow). Click on OK. Type [W][F] to fit the axes limits. The added curve is now displayed in orange on the screen. An alternative way to define the variable to be plotted is to select them in the ICs window (click on the little squares in front of the variables you want to display on the screen and then on the button “xvst”). 6. To display the trajectory in the phase space, click on Viewaxes [V] and then on “2D” [ENTER]. A window opens. Enter then the variables you wish to plot on X and Y-axes (e.g., mlhy and mtoc). Click on OK. Type [W][F] to fit the axes. The limitcycle corresponding to the oscillations is now displayed. Again, this could be also realized in the ICs window (click on two little squares corresponding to the variables you want to display and then on the button “xvsy”). 7. Click then on the button Param. The “Parameters” window opens. You can then change the value of some parameters. Set, for example, m4 (which represents the maximum rate of TOC1 protein degradation) to 2. Click then on “OK” and on “Go” to run a simulation with these new parameter values. The trajectory in the phase space is a spiral: the system shows damped

346

Christoph Schmal et al.

oscillations and converges to a stable steady state. The model thus predicts that increasing the degradation of TOC1 protein leads to an arrest of the self-sustained oscillations. 8. To save the data in a file, open the Data window, click on “Write,” enter a file name (otherwise the default name is test.dat), and click on OK. This tab-delimited data file contains 5 columns: the first one is the time and the following are the variables. 9. To quit XPP-AUTO, click on File [F] and then on “Quit” [Q]. Figure 3A shows the simulated time series of our four dynamical variables under DD conditions for the parameter set as proposed in Subheading 3.2 and after neglecting the transient dynamics to the limit cycle. Figure 3B shows these limit-cycle oscillations, plotted in the MT–PT phase space. The graphs were obtained using the XPP-AUTO–Python interface XPPy for numerical simulations and the Python plotting library Matplotlib for visualization [47, 48]. 3.4 Effect of Light and Entrainment by an LD Cycle

An important factor for circadian clocks is light [49]. Light was shown to affect one or several molecular processes at the core of the circadian clock [50]. The effect of light is usually modeled through the modulation of the value of one (or several) light-controlled parameters. A light–dark (LD) cycle can be modeled by the periodic modulation of the light-controlled parameter(s)[51, 52]. Such periodic forcing can be described in XPP-AUTO by adding the following lines to the previous ODE file: tm=mod(t,per) tmtest=tm-per*(1-php) F=heav(tmtest) L=amp*F aux light=L param amp=0.02 param per=24 param php=0.5 The function mod(t,per) returns the outcome of “t modulo per” and heav is the Heaviside step function, defined as 0, x < 0, heav (x) :=  1, x ≥ 0 . The three first lines thus generate a square-wave function F, switching between F = 0 (dark phase) and F = 1 (light phase), with a period per and a “photoperiod” php, defined by the ratio of the duration of the light phase divided by the period. Other functions, like sine, semi-sine, or atan, can also be implemented.

347

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

TOC1 Protein (PT (t))

b

Concentration [nM]

a MT PT ML PL 1

0.5

0 24

c

36

48 Time [h]

60

72

1 0.5 0 1

0

0.5 TOC1 mRNA (MT (t))

d

1

Short Day

T = 20h

0.5 0

LHY/CCA1 mRNA (ML (t))

e

50

0

100 Time [h]

150

f

a

b

3 2

HB

1

0.1

0.3

0.5 v1 [nM/h]

0.7

0.9

0.5 0 Long Day

1 0.5 0

200

4

0

TOC1 mRNA (MT (t))

T = 24h

0 1

LHY/CCA1 mRNA (ML (t))

TOC1 mRNA (MT (t))

1 0.5

1

12

0

24 Time [h]

36

48

1

a

0 2

b 1 0

0

50

100

150

200

250

300

Time [h]

Fig. 3 Dynamical behavior of our simplified Arabidopsis thaliana circadian clock model. (A) Solutions of the model for the parameter set from Subheading 3.2 under DD conditions without considering any transient dynamics. One can observe limit-cycle oscillations with a period of approximately 24. 8h. (B) The same solution as in (A) drawn in the MT –PT phase plane. (C) Upper: The system can be entrained to LD cycles with a period of T = 24h. Lower: The circadian clock desynchronizes from the external LD cycles for periods that differ markedly from the system’s internal period, as shown for a period of T = 20h. (D) The model can be entrained to different photoperiods. (E) One parameter bifurcation diagram: Plotted are the maximal transcription rate of LHY/CCA1 (v1) versus the LHY/CCA1 mRNA concentration (ML ). Continuous black lines denote stable fixed points, while dashed black lines denote unstable fixed points. The system undergoes a Hopf-Bifurcation (HB) at v1 = 0.194 nM/h . Diverging dynamics of PT are observed for v1 < 0. 067 nM/h. Circle markers denote peak and trough values of oscillatory solutions for a given v1. (F) Damped (a) and self-sustained (b) oscillations under DD conditions for v1 = 0.16 nM/h and v1 = 0.5nM/h , respectively

348

Christoph Schmal et al.

The function “aux” in the code above allows XPP to save the value of “light” in the output. The time series of the function “light” can then be displayed on the screen: @ nplot=3, yp1=mlhy, yp2=mtoc, yp3=light You can then examine the effect of light by performing the following operations: 1. Launch XPP-AUTO with the amended ODE file. 2. Type [I][G][W][F]. The light modulation is now displayed in orange. If you repeat [I][L] a couple of times, you can verify that the oscillations are entrained by the LD cycle: they acquire the period of 24h and are phase locked with respect to the periodic forcing. 3. In the parameter window (button Param), set the forcing period per to 20 and click on “OK.” Run the simulation by typing [I][G] and, a few times, [E][I][L]. The oscillations do not stabilize. There is no entrainment and the phase is not locked. As a matter of fact, this quasi-periodic behavior is commonly observed when the period of the oscillator and of the periodic forcing are not sufficiently close. 4. Entrainment can be recovered by increasing the amplitude amp of the forcing (try, for example, amp = 0. 04). 5. The data can be saved as described above. Here, the data file contains an additional column, with the value of “light.” Depending on the period and amplitude of the periodic forcing by light, circadian oscillations can be entrained or undergo complex behaviors, see Fig. 3C. Typically, the oscillations are properly entrained when the period of the forcing is close to the intrinsic period of the oscillator or to a rational multiple of that period. Increasing the forcing amplitude also facilitates entrainment. The domains of entrainment in the (per,amp) space are called Arnold tongues. XPP allows to easily change the value of per and amp and to observe the resulting dynamics, but there is no automatic way to construct the Arnold tongues. It has been shown that several components of the Arabidopsis circadian clock are expressed in a photoperiod-dependent way, i.e., they depend on the length of the daily light phase [53]. You can readily test the sensitivity of the model to the applied photoperiod by changing the value of php for a given period T. Figure 3D illustrates the earlier phase of TOC1 mRNA under short day (6h of light and 18h of dark) conditions, compared to the later phase under long day (18h of light and 6h of dark) conditions. Another effect of light is to phase shift the oscillations. If a brief pulse of light is applied at a given time, the system will react by undergoing a transient perturbation. Once the perturbation is damped (as expected for limit-cycle oscillations), the period and amplitude of the oscillations are recovered, but a phase shift with respect to the unperturbed oscillations is expected.

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

349

The magnitude of that phase shift depends on the phase at which the pulse is applied. A phase response curve (PRC) gives the phase shift as a function of the time (phase) of the perturbation. Here again there is no automatic way to construct PRCs with XPPAUTO, but the phase shift for a given pulse of light can easily be computed and visualized: 1. In the previous file, replace the three first lines by: F=heav(t-t1)*heav(t2-t) L=amp*F Define parameters t1 and t2 (times of beginning and end of the light pulse): param t1=24 param t2=28 and, for more visibility, plot only one variable: @ nplot=2, yp1=mlhy, yp3=light 2. Launch XPP with this modified file, set parameter amp = 0, and run the simulation. 3. Without “erasing” the window, set parameter amp, for example, to 0.1 and run the simulation. You then see the “perturbed” time series superimposed to the “unperturbed one” and you can appreciate the phase shift between the two curves. 4. The effect of the phase of the perturbation and of its duration can be assessed by changing the values of t1 and t2. 3.5 Bifurcation (AUTO)

As seen above, it is easy to change the value of a given parameter and rerun a simulation to assess the effect of that change. A bifurcation diagram shows how the steady state of a system changes as a control parameter is changed. Such plots can be very instructive because it also shows when and how stable solutions (steady states) become unstable. They summarize the effect of a parameter on the (long-term) behavior of a dynamical system. Besides steady state solutions, the amplitude or the period of oscillations can also be plotted as a function of the control parameter. Such bifurcation diagrams could be constructed by extensive numerical simulation of the ODE (for every value of the parameter) followed by a time series analysis. An alternative is to use AUTO which computes such diagrams via continuation methods (and does not necessitate intensive integration of ODEs). The following instructions show how to run AUTO for the above model. As a control parameter we selected the maximum transcription rate of LHY/CCA1, v1. 1. Launch XPP-AUTO with the original arabidopsis.ode (not the amended model that accounts for the effect of light). 2. To start AUTO, we need to start at the steady state. We then first set parameter v1 to 0. 1 nM/h. Type [I][G] and repeat

350

Christoph Schmal et al.

Table 1 Important numerical parameters of AUTO [43]. Note that Ds is adaptive and thus only a “suggestion.” A reduction of Ds, Dsmin, and Dsmax leads to a higher precision of the numerical calculations Parameter name

Description

Ntst

Number of mesh intervals for discretization of periodic orbits

Nmax

Maximum number of steps along any branch

NPr

Complete info will be given on every “Npr” steps

Ds

Initial step size for the continuation process. The sign of Ds determines the direction of the parameter continuation

Dsmin/Dsmax

Minimum and maximum step sizes

Par Min/Par Max

Left-hand and right-hand limit for the calculation to stop

two or three times [I][L] so that the system converged sufficiently close to the steady state. 3. Click on Sing Pts [S], and then on “Go.” A window opens with the question “Print eigen values?” Click “Yes.” A new window gives the value of the variables at the steady state, as well as information about its stability (c+ and c− gives the number of complex eigen values with positive or negative real part, resp.; im gives the number of purely imaginary eigen values; and r+ and r− give the number of positive and negative eigen values, resp.). In case the system did not converge to the steady state, the window displays the message “no convergence.” 4. Click on the menu File and then on “Auto.” The AUTO window opens. 5. In the AUTO window, click on “Parameter” and define as “Par1” v1. Click on “OK.” 6. Click then on “Axes.” You can determine the axes of the plot. We will keep mlhy as the variable (displayed in the y-axis), but we want v1 as the main parameter. 7. Click on “Numerics” to set up or change the numerical parameters for the continuation process. A description of the main parameters can be found in Table 1. 8. Click on “Run” and subsequently on “Steady state.” A curve is drawn. This is the steady state value of LHY/CCA1 as a function of v1. The thick red part indicates that the steady state is stable, while the thin black part indicates that the steady state is unstable. The transition corresponds to a Hopf bifurcation. 9. From the Hopf bifurcation, we can “follow” the oscillations. To this end we need to grab the Hopf point. Click then on “Grab.” Using the tabulation key [TAB], you can “jump” to

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

351

Table 2 Marker names that can appear in the bifurcation diagrams obtained by AUTO [43] Marker name

Description

LP

Limit Point, Turning Point, Saddle Node

HB

(Andronov-)Hopf Bifurcation Point

TR

Torus Bifurcation

PD

Period Doubling Bifurcation

BP

Bifurcation or Branch Point (e.g., transcritical or pitchfork bifurcation)

EP

Endpoint of a branch. Starting-point and endpoint of the bifurcation diagram calculation

MX

Failure of convergence

the various points marked by AUTO. Stop at the point number 2 and click on [ENTER] to select this point. You can see that AUTO correctly identified this point as a Hopf bifurcation: Below the main window, this point is marked “HB” and its location is given by v1 v1 = 0.1936 nM/h . Click on “Run” and select “Periodic.” The two green curves which appear on the screen are the maxima and minima of mlhy. A list of important “special” points that can appear in the bifurcation diagrams of XPP-AUTO is given in Table 2. 10. Axes can be fitted to the window. Click on “Axes,” then on “Fit.” This model predicts that the amplitude of the LHY/ CCA1 mRNA increases with v1, and that the oscillations are lost below a critical value of this control parameter, see Fig. 3E. Simulated time series data showing solutions with damped dynamics ( v1 = 0.16 nM/h ) and solutions with a higher amplitude ( v1 = 0.5nM/h ) compared to the oscillations obtained from the original parameter set (i.e., v1 = 0.3nM/h ) is presented in Fig. 3F. 11. It is also possible to display the period. Click on “Axes,” then on “Period.” Click then again on “Axes” and on “Fit.” The diagram now shows that the period of the oscillations increases nearly linearly with v1. 12. To save data in a file, click on File and then on “Write pts.” Give a file name and click on “OK.” The data are saved in a tabdelimited format. Note that only the variable displayed on the screen (or the period) is saved. The three first columns are the controlled parameter, the maximum, and the minimum of the variable (identical at steady state). The last two columns give information about the type (“1” = steady state, “2” = oscillations) and the stability (“1” = stable, “2” = unstable) of the solution.

352

Christoph Schmal et al.

3.6 Stochastic Simulations

Noise is an omnipresent feature of biological systems that may affect the timing precision of the circadian clock [54]. In biological systems we often distinguish between internal and external sources. While external noise could be induced by fluctuating environmental influences, internal noise arises from the inherent probabilistic nature of biochemical reactions [55, 56]. Fluctuating environmental influences can be accounted by parameter variations. Computational models treating internal, molecular noise due to a limited number of molecules in the observed system are usually treated by the Gillespie algorithm, a Monte-Carlo method for the discrete and stochastic simulation of reaction events [57]. Another way to model stochasticity in biochemical systems is to add noise to the previously deterministic equations. The stochastic ordinary differential equation for a certain species xi(t) then reads as   xɺi (t ) = f i (t , x ) + g i (x (t ))ξi (t ), (8)  where f i (t , x ) are the original deterministic equations, ξi(t) are inde pendent delta-correlated white noises of zero mean, and g i (x (t )) denote the noise strengths at a given time t. In analogy with [58], we  choose g i (x (t )) = 2σ xi (t ) with a global noise parameter σ. You can simulate the stochastic description of our Eqs. 1–4, obtained as defined by Eq. 8, by applying the following protocol: 1. XPP-AUTO allows you to integrate stochastic differential equations by means of the so-called Wiener parameters that return scaled white noises. They are defined as normally distributed random numbers with zero mean and a variance of dt , where dt is the numerical time step [43]. Wiener parameters are declared in the ODE File via wiener w. 2. In our case we want to add four independent noise terms with  concentration-dependent noise strengths g i (x (t )) = 2σ xi (t ) and thus have to declare four independent Wiener parameters. These enter our original ordinary differential equations as follows: # LHY/CCA1-TOC1 model (Arabidopsis) with noise # Noise Parameters wiener w1, w2, w3, w4 param sigma=0.001 # Equations dmlhy/dt=⋯ + dplhy/dt=⋯ + dmtoc/dt=⋯ + dptoc/dt=⋯ + ⋮ @ meth=euler done

sqrt(2*sigma*mlhy)*w1 sqrt(2*sigma*plhy)*w2 sqrt(2*sigma*mtoc)*w3 sqrt(2*sigma*ptoc)*w4

b Concentration [nM]

a MT

1

ML

0.5

0

0

24

48 Time [h]

72

96

TOC1 Protein (PT (t))

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

353

1

0.5

0

0

0.5

1

TOC1 mRNA (MT (t))

Fig. 4 Stochastic dynamics of the two-gene model from Eqs. 1–4 for σ = 0. 001 under DD conditions. (a) Dashed lines denote the deterministic solution of the LHY/CCA1 mRNA as well as TOC1 mRNA as shown in Fig. 3A. Continuous lines denote the corresponding stochastic solutions. (b) Stochastic dynamics of TOC1 mRNA and protein under DD condition. See Fig. 3B for a comparison with the deterministic dynamics

3. It is important that you choose the Euler method for integration since XPP-AUTO does not have any special integrators for stochastic differential equations. 4. Launch XPP-AUTO with the modified ODE file. 5. Type [I][G][W][F] as described above in order to start the integration and to fit the window width. 6. You can observe the non-deterministic behavior of our stochastic simulations by relaunching the integrations several times ([I][G]). 7. You can choose the seed of the underlying random number generator by choosing New seed in the Stochastic submenu of the root menu Numerics, i.e., type [U][H][N]. Figure 4 shows solutions of the stochastic two-gene model with σ = 0. 001 and the corresponding deterministic solutions under constant darkness (DD) for identical initial conditions. You can now investigate the impact of different noise strengths on the systems dynamics by varying the parameter σ. Note that XPP-AUTO is not designed to further analyze statistical time series data (e.g., spectral or wavelet analysis), but a variety of specialized programs such as the statistics programming language R [59] are available for such purposes.

4

Evolution of the Arabidopsis thaliana Circadian Clock Models The first molecular model for the circadian clock of Arabidopsis was based on a single negative loop where TOC1 protein, the product of TOC1 mRNA, activates the transcription of LHY/ CCA1 mRNA, whose product, LHY/CCA1 protein, inhibits the transcription of TOC1 mRNA [23]. This model is based on seven

354

Christoph Schmal et al.

Ordinary Differential Equations (ODEs): The dynamics of both LHY/CCA1 and TOC1 are each described by three equations (mRNAs, cytosolic, and nuclear proteins) and one equation is introduced for a light-sensitive protein. Numerical analysis of this model showed that limit-cycle circadian oscillations can readily be obtained in a two-gene system and demonstrated the possibility to fit noisy experimental data using a cost function optimization technique as described above (see Subheading 2). This model, however, failed to explain the short-period phenotype of plants carrying loss-of-function mutations in either LHY or CCA1 [60] or the late phase of the oscillations under longer photoperiods [61]. In order to overcome these limitations, two extensions of the model were proposed: the extended one-loop and the two-loop models [24]. The extended one-loop model incorporates an additional unknown gene (variable X between TOC1 and LHY/ CCA1), while the two-loop model involves a second unknown element (variable Y forming a new negative feedback loop with TOC1 and introducing a new point of entry for light perception). While the extended one-loop model still lacks some essential features of the circadian clock (e.g., wrong period for the cca1 mutant and arrhythmicity under long photoperiods), the two-loop model better agrees with experimental observations. Remarkably, comparison with experimental data in the two-loop model also predicts that the unknown Y variable could be the GIGANTEA (GI) protein. In parallel with the discovery of new clock genes, several extensions of these models were proposed, integrating these genes/proteins as new variables, as well as additional positive or negative regulatory feedback loops (see Fig. 5). The PSEUDO RESPONSE REGULATORS 7 and 9 (PRR7 and PRR9, respectively) were incorporated, either through a single variable for both PRRs [25] or through two separate variables [26], adding thus one or two additional negative feedback loops between LHY/CCA1 and the PRRs and giving rise to the three-loop and four-loop models, respectively. Analysis of the three-loop network leads to the idea that the plant clock consists of morning and evening oscillators. These successively extended models account for phenotypes observed in varying environments and mutants. Additional (speculative) post-translational modifications were also taken into account in a multiple-loop model [27]: a modified TOC1 form (TOC1mod) replaces the X variable and a modified LHY form (LHYmod) activates the PRRs; GI and Y are treated as two different variables; TOC1 inhibits PRR9 and GI inhibits TOC1 through the F-Box protein ZEITLUPE (ZTL). This later model was then updated 2 years later [28] not only by adding new elements/variables (e.g., ELF3, ELF4, and LUX forming an Evening Complex, EC) and new regulatory feedback loops (e.g., between the EC and other clock elements) but also by removing some variables (e.g., TOC1mod, LHYmod, or Y). Two striking facts about this model are worth mentioning. First, TOC1, a key

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO

355

Fig. 5 Stages of development in the Arabidopsis thaliana circadian clock models

element of the circadian clock, so far assumed to act as an activator, now appears to be an inhibitor. This change of function predicted by the model was recently corroborated experimentally [37, 62]. Second, the core mechanism of the Arabidopsis circadian clock appears to be reducible to a 3-variable model, similar to the “repressilator“ (see [63]), where the EC inhibits the PRRs, that inhibit LHY/CCA1, that in turn inhibits the EC, thus closing the core negative feedback loop. Recently, even this multiple-loop model was further extended by the inhibitory effect of TOC1 on the expression of various clock genes (including LHY/CCA1, PRR9, PRR7, PRR5, LUX, ELF4, and GI) and an additional input to the clock through the abscisic acid (ABA) signaling pathway[29]. While all the models presented here above are based on ODEs, alternative approaches were recently developed for some of them such as (1) a model based on Boolean logic [36] for the two-loop and three-loop models and (2) a stochastic version of the 2010 multiple-loop model [34]. All the modifications in the models (from the one-loop model to the 2012 multiple-loop model) give rise to more and more complex/detailed models (see the evolution in Fig. 5). They are also based on experimental data and even if the models do not corroborate all the experimental facts, they often give rise to interesting predictions that are then tested experimentally to finally

356

Christoph Schmal et al.

“feed back” to the modelers that incorporate new elements or regulations into their models. They also explain robustness (to parameter values and to noise) and flexibility (e.g., adaptation to various day lengths) of the circadian network, and, thereby, reveal the design principles of the clock [23–29]. Future studies will use computational models to further untangle the complex web of output pathways controlling the physiological rhythms as well as the input pathways that process and transmit zeitgeber signals, such as light and temperature, to the clock. These processes include post-transcriptional slave oscillators [64], the regulation of stomata aperture [29], the photoperiodic control of flowering [65], the role of calcium in the circadian regulation [66], or the influence of the nutrients on the clock and clock-controlled processes [67]. The combination of modeling and experimental approaches will undoubtedly lead to a better understanding of clock-controlled processes in plants.

Acknowledgments This work was supported by the program “Actions de Recherche Concertée” (ARC2012-2017) launched by the Division of Scientific Research, Ministry of Science and Education, French community of Belgium and by DFG grants to Dorothee Staiger (STA 653 and SFB 613). C. S. is a fellow of International Graduate Program “Bioinformatics of Signaling Networks” funded by Bielefeld University. J.-C. L. is Chercheur qualifié du Fonds National de la Recherche Scientifique (F.N.R.S., Belgium). References 1. Golden SS, Canales SR (2003) Cyanobacterial circadian clocks - timing is everything. Nat Rev Microbiol 1(3):191–199 2. Bell-Pedersen D, Crosthwaite SK, LakinThomas PL, Martha M, Okland M (2001) The Neurospora circadian clock: simple or complex? Phil Trans Roy Soc Lond B Biol Sci 356 (1415):1697–1709 3. Eelderink-Chen Z, Mazzotta G, Sturre M, Bosman J, Roenneberg T, Merrow M (2010) A circadian clock in Saccharomyces cerevisiae. PNAS 107(5):2043–2047 4. Allada R, Chung BY (2010) Circadian organization of behavior and physiology in Drosophila. Ann Rev Physiol 72(1):605–624 5. Robertson McClung C (2006) Plant circadian rhythms. Plant Cell 18(4):792–803 6. Mohawk JA, Green CB, Takahashi JS (2012) Central and peripheral circadian clocks in mammals. Ann Rev Neurosci 35(1):445–462

7. Bell-Pedersen D, Cassone VM, Earnest DJ, Golden SS, Hardin PE, Thomas TL, Zoran MJ (2005) Circadian rhythms from multiple oscillators: lessons from diverse organisms. Nat Rev Genet 6(7):544–556 8. Brown SA, Kowalska E, Dallmann R (2012) (Re)inventing the circadian feedback loop. Dev Cell 22(3):477–487 9. Staiger D, Köster T (2010) Spotlight on posttranscriptional control in the circadian system. Cell Mol Life Sci 68(1):71–83 10. Tomita J, Nakajima M, Kondo T, Iwasaki H (2005) No transcription-translation feedback in circadian rhythm of KaiC phosphorylation. Science 307(5707):251–254 11. Nakajima M, Imai K, Ito H, Nishiwaki T, Murayama Y, Iwasaki H, Oyama T, Kondo T (2005) Reconstitution of circadian oscillation of cyanobacterial KaiC phosphorylation in vitro. Science 308(5720):414–415

Modeling and Simulating the Arabidopsis thaliana Circadian Clock Using XPP-AUTO 12. Pittendrigh C, Bruce V, Kaus P (1958) On the significance of transients in daily rhythms. PNAS 44(9):965–973 13. Strogatz SH (1987) Human sleep and circadian rhythms: a simple model based on two coupled oscillators. J Math Biol 25(3):327–347 14. Pittendrigh C, Kyner WT, Takamura T (1991) The amplitude of circadian oscillations: temperature dependence, latitudinal clines, and the photoperiodic time measurement. J Biol Rhythms 6(4):299–313 15. Westermark PO, Welsh DK, Okamura H, Herzel H (2009) Quantification of circadian rhythms in single cells. PLoS Comput Biol 5(11):e1000580 16. Granada AE, Cambras T, Díez-Noguera A, Herzel H (2011) Circadian desynchronization. Interface Focus 1(1):153–166 17. Oda GA, Otto Friesen W (2011) Modeling two-oscillator circadian systems entrained by two environmental cycles. PLoS ONE 6(8): e23895 18. Leloup J-C, Goldbeter A (1998) A model for circadian rhythms in Drosophila incorporating the formation of a complex between the PER and TIM proteins. J Biol Rhythms 13(1): 70–87 19. Leloup J-C, Gonze D, Goldbeter A (1999) Limit cycle models for circadian rhythms based on transcriptional regulation in Drosophila and Neurospora. J Biol Rhythms 14(6):433–448 20. Kurosawa G, Aihara K, Iwasa Y (2006) A model for the circadian rhythm of cyanobacteria that maintains oscillation without gene expression. Biophys J 91(6):2015–2023 21. Leloup J-C, Goldbeter A (2003) Toward a detailed computational model for the mammalian circadian clock. PNAS 100(12): 7051–7056 22. Forger DB, Peskin CS (2003) A detailed predictive model of the mammalian circadian clock. PNAS 100(25):14806–14811 23. Locke JCW, Millar AJ, Turner MS (2005) Modelling genetic networks with noisy and varied experimental data: the circadian clock in Arabidopsis thaliana. J Theor Biol 234(3): 383–393 24. Locke JCW, Southern MM, Kozma-Bognar L, V Hibberd, Brown PE, Turner MS, Millar AJ (2005) Extension of a genetic network model by iterative experimentation and mathematical analysis. Mol Syst Biol 1:2005.0013 25. Locke JCW, Kozma-Bognar L, Gould PD, Feher B, Kevei E, Nagy F, Turner MS, Hall A, Millar AJ (2006) Experimental validation of a predicted feedback loop in the multi-oscillator clock of Arabidopsis thaliana. Mol Syst Biol 2:59

357

26. Zeilinger MN, Farre EM, Taylor SR, Kay SA, Doyle FJ (2006) A novel computational model of the circadian clock in Arabidopsis that incorporates PRR7 and PRR9. Mol Syst Biol 2:58 27. Pokhilko A, Hodge SK, Stratford K, Knox K, Edwards KD, Thomson AW, Mizuno T, Millar AJ (2010) Data assimilation constrains new connections and components in a complex, eukaryotic circadian clock model. Mol Syst Biol 6:416 28. Pokhilko A, Piñas Fernández A, Edwards KD, Southern MM, Halliday KJ, Millar AJ (2012) The clock gene circuit in Arabidopsis includes a repressilator with additional feedback loops. Mol Syst Biol 8:574(1) 29. Pokhilko A, Mas P, Millar AJ (2013) Modelling the widespread effects of TOC1 signalling on the plant circadian clock and its outputs. BMC Syst Biol 7(1):23 30. Morant P-E, Thommen Q, Pfeuty B, Vandermoere C, Corellou F, Bouget F-Y, Lefranc M (2010) A robust two-gene oscillator at the core of Ostreococcus tauri circadian clock. Chaos 20(4):045108 31. Gonze D, Halloy J, Goldbeter A (2002) Robustness of circadian rhythms with respect to molecular noise. PNAS 99(2):673–678 32. Gonze D, Halloy J, Leloup J-C, Goldbeter A (2003) Stochastic models for circadian rhythms: effect of molecular noise on periodic and chaotic behaviour. CR Biologies 326(2):189–203 33. Forger DB, Peskin CS (2005) Stochastic simulation of the mammalian circadian clock. PNAS 102(2):321–324 34. Guerriero ML, Pokhilko A, Piñas Fernández A, Halliday KJ, Millar AJ, Hillston J (2012) Stochastic properties of the plant circadian clock. J R Soc Interface 9(69):744–756 35. Bornholdt S (2005) SYSTEMS BIOLOGY: less is more in modeling large genetic networks. Science 310(5747):449–451 36. Akman OE, Watterson S, Parton A, Binns N, Millar AJ, Ghazal P (2012) Digital clocks: simple boolean models can quantitatively describe circadian systems. J R Soc Interface 9(74): 2365–2382 37. Huang W, Pérez-García P, Pokhilko A, Millar AJ, Antoshechkin I, Riechmann JL, Mas P (2012) Mapping the core of the Arabidopsis circadian clock defines the network structure of the oscillator. Science 336(6077):75–79 38. Alon U (2007) An introduction to systems biology: design principles of biological circuits. Chapman & Hall/CRC, Mathematical and Computational Biology series, Taylor & Francis Group, Boca Raton

358

Christoph Schmal et al.

39. Goodwin BC (1965) Oscillatory behavior in enzymatic control processes. Adv Enzyme Regul 3:425–437 40. Griffith JS (1968) Mathematics of cellular control processes I. Negative feedback to one gene. J Theor Biol 20(2):202–208 41. Leloup J-C (2009) Circadian clocks and phosphorylation: insights from computational modeling. Cent Eur J Biol 4(3):290–303 42. Brown KS, Sethna JP (2003) Statistical mechanical approaches to models with many poorly known parameters. Phys Rev E 68(2): 021904 43. Ermentrout B (2002) Simulating, analyzing, and animating dynamical systems: a guide to XPPAUT for researchers and students. SIAM 2002, Philadelphia 44. Doedel EJ (1981) AUTO, a program for the automatic bifurcation analysis of autonomous systems. Cong Numer 30:265–384 45. Doedel EJ, Champneys AR, Fairgrieve TF, Kuznetsov YA, Sandstede B, Wang X (1997) AUTO97: continuation and bifurcation software for ordinary differential equations (with HomCont). Technical Report, Concordia University 46. Doedel EJ, Paffenroth RC, Champneys AR, Fairgrieve TF, Kuznetsov YA, Oldeman BE, Sandstede B, Wang X (2002) AUTO 2000: continuation and bifurcation software for ordinary differential equations (with HomCont). Technical Report, Concordia University 47. Nowacki J (2011) XPPy, URL: https:// github.com/jsnowacki/xppy 48. Hunter JD (2007) Matplotlib: A 2D Graphics Environment. Comput Sci Eng 9(3):90-95, URL: http://matplotlib.org 49. Fankhauser C, Staiger D (2002) Photoreceptors in Arabidopsis thaliana: light perception, signal transduction and entrainment of the endogenous clock. Planta 216(1):1–16 50. Millar AJ (2004) Input signals to the plant circadian clock. J Exp Bot 55:277–283 51. Gonze D (2011) Modeling circadian clocks: roles, advantages, and limitations. Cent Eur J Biol 6(5):712–729 52. Gonze D (2011) Modeling circadian clocks: from equations to oscillations. Cent Eur J Biol 6(5):699–711 53. Edwards KD, Akman OE, Knox K, Lumsden PJ, Thomson AW, Brown PE, Pokhilko A, Kozma-Bognar L, Nagy F, Rand DA, Millar AJ (2010) Quantitative analysis of regulatory flexibility under changing environmental conditions. Mol Sys Biol 6:424 54. Barkai N, Leibler S (2000) Biological rhythms: circadian clocks limited by noise. Nature 403 (6767):267–268

55. Raser JM, O’Shea EK (2005) Noise in gene expression: origins, consequences, and control. Science 309(5743):2010–2013 56. Swain PS, Longtin A (2006) Noise in genetic and neural networks. Chaos 16(2):026101 57. Gillespie DT (1977) Exact stochastic simulation of coupled chemical reactions. J Phys Chem 81(25):2340–2361 58. Steuer R (2004) Effects of stochasticity in models of the cell cycle: from quantized cycle times to noise-induced oscillations. J Theor Biol 228(3):293–301 59. R Core Team (2012) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, 2012. ISBN 3-900051-07-0 60. Mizoguchi T, Wheatley K, Hanzawa Y, Wright L, Mizoguchi M, Song H-R, Carré IA, Coupland G (2002) LHY and CCA1 are partially redundant genes required to maintain circadian rhythms in Arabidopsis. Dev Cell 2(5):629–641 61. Millar AJ, Kay SA (1996) Integration of circadian and phototransduction pathways in the network controlling CAB gene transcription in Arabidopsis. PNAS 93(26):15491–15496 62. Gendron JM, Pruneda-Paz JL, Doherty CJ, Gross AM, Earl Kang S, Kay SA (2012) Arabidopsis circadian clock protein, TOC1, is a DNA-binding transcription factor. PNAS 109(8):3167–3172 63. Elowitz MB, Leibler S (2000) A synthetic oscillatory network of transcriptional regulators. Nature 403(6767):335–338 64. Schmal C, Reimann P, Staiger D (2013) A circadian clock-regulated toggle switch explains AtGRP7 and AtGRP8 oscillations in Arabidopsis thaliana. PLoS Comput Biol 9(3): e1002986 65. Salazar JD, Saithong T, Brown PE, Foreman J, Locke JCW, Halliday KJ, Carré IA, Rand DA, Millar AJ (2009) Prediction of photoperiodic regulators from quantitative gene circuit models. Cell 139(6):1170–1179 66. Dalchau N, Hubbard KE, Robertson FC, Hotta CT, Briggs HM, Stan G-B, Gonçalves JM, Webb AAR (2010) Correct biological timing in Arabidopsis requires multiple lightsignaling pathways. PNAS 107(29): 13171–13176 67. Hermans C, Vuylsteke M, Coppens F, Craciun A, Inzé D, Verbruggen N (2010) Early transcriptomic changes induced by magnesium deficiency in Arabidopsis thaliana reveal the alteration of circadian clock gene expression in roots and the triggering of abscisic acid-responsive genes. New Phytologist 187(1): 119–131

INDEX A Aequorin....................................216, 217, 219, 221, 223–225 Alcohol-inducible system ......................... 203, 204, 207, 208 Amplitude ...................................... 4, 7, 8, 27, 28, 38–42, 54, 67, 108, 188, 223, 303, 348, 351 Antibody....................................... 58, 64, 108, 110, 115, 119, 177, 179, 180, 183, 244, 260, 261, 263, 268 Arabidopsis thaliana .......................................58, 63, 108, 124, 126, 139–153, 157, 175, 188, 215–225, 239, 265, 273, 286–288, 298, 325, 326, 337–356 Ascorbate.......................................... 239–241, 247–250, 266 AtGRP7 ................................................... 108, 110, 111, 118 AthaMap ..................................................................139–153

B Bacterial titre .................................................... 278–280, 282 Bifurcation diagram .................................. 343, 347, 349, 351 BioDare ........................................................................13–43 Bioluminescence .......................................188, 217, 221, 223, 224, 325–336 Botrytis cinerea ........................................... 273–276, 280–281 Bradford ................................................... 179, 243, 257–258 BRASS ..................................................5, 14, 15, 32–34, 221 Bud burst .................................................... 303–305, 312–323 set ........................................301–303, 305, 306, 313–323

C Calcium ions..................................................... 215–225, 336 CCD camera ............................. 3–7, 179, 183, 184, 218, 223 cDNA synthesis..............................................54, 95, 99, 101, 160, 167, 169 Chlamydomonas reinhardtii ........................................ 187–201 Chromatin chromatin immunoprecipitation (ChIP) ..................................... 57, 58, 60, 64, 66, 111 ChIP-qPCR ...........................................................57, 58 ChIP-Seq ...............................................................57–68 modification ...........................................................57, 58 Circadian Clock Associated 1 (CCA1)...............................................4–6, 58, 65, 66, 124, 141, 204–208, 210, 212, 215, 216, 219, 239, 298–300, 302, 303, 339–342, 344, 347, 349–355

Clock ............................................ 1, 13, 45, 57, 93, 107, 123, 141, 157, 175, 187, 203, 210, 215, 227, 239, 274, 286, 296, 314, 325, 337 Cold acclimation ..............................................................301 Colocalization web tool ............................................ 141, 144 Confocal microscopy ................................................210–212 CONSTANS.................................................... 188, 286, 298 Critical day length ........................................... 285, 287, 301, 314, 317 Cross-linking formaldehyde ...............................109, 112, 114, 118, 119

D Data repository .............................................................14, 15 Day length ........................................285–288, 293, 301, 304, 306, 313, 314, 316, 317, 320, 322, 356 Differential equation ordinary .............................................. 338, 339, 352, 354 stochastic .................................................... 339, 352, 353 Disease ..................................................... 274, 278, 282, 292 Diurnal ......................................... 93, 96, 140, 141, 150, 151, 157–173, 273, 274, 303, 305 Dormancy ........................................................ 301, 303–306, 317–320, 322

E Epitope ..............................................108, 110, 118, 176, 179 ESI-MS............................................193, 197, 244, 245, 260, 263, 264, 269, 625 Ethanol switch..................................................................203 Evening element ...............................................................141

F Fast Fourier transform nonlinear least squares (FFT-NLLS) ......................... 5, 14, 28, 29, 32, 39, 41 Feedback loops .................................108, 175, 215, 216, 298, 338–340, 354, 355 FerroZine ......................................................... 229–231, 235 Flagella .....................................................................187–201 Flowering ................................................. 285–294, 298, 302 Flowering Locus C ............................................. 287, 302 Flowering Locus T .....................................................302 Fluorescent protein ........................................... 209–213, 254 Freeze test.........................................................................318

Dorothee Staiger (ed.), Plant Circadian Networks: Methods and Protocols, Methods in Molecular Biology, vol. 1158, DOI 10.1007/978-1-4939-0700-7, © Springer Science+Business Media New York 2014

359

PLANT CIRCADIAN NETWORKS: METHODS AND PROTOCOLS 360 Index G

M

G-Box ..............................................................................141 Gene Analysis web tool .................................... 144–146, 152 Gene Identification web tool ............................................142 Gene targeting .......................................................... 326, 334 GIGANTEA ........................................... 168, 286, 298, 354 Glutathion glutathionylation..........................240, 244, 250, 261–264 Green Fluorescent Protein (GFP) GFP Trap Beads ..................109–112, 115, 116, 118, 177

Mapman .......................................................................75, 87 Microarrays...........................................45–55, 57, 58, 71, 75, 85, 94, 151, 157–173 Micronutrients..........................................................227–236 MicroRNAs........................................93, 140, 142, 143, 145, 147, 148, 150, 152 MicroRNA targets web tool .............................................148 Midnight module .............................................................141 Minimal medium..................................................... 188, 189, 192, 228, 231 Modeling ..................................................................337–356 Morning element..............................................................141 Moss .........................................................................325–336 Multiexperiment Viewer ..............................................75, 87

H Heading date .................................................... 287, 291, 294 Histone methylation.............................................................57, 58 modifications ..........................................................57–67 Homologous recombination ............................. 326, 334, 335 Hordeum vulgare................................................ 286, 288, 290 Hybridization ................................... 1, 45, 46, 50–52, 55, 57, 58, 111, 160, 169, 173 Hydorgen peroxide ................................... 241, 242, 245, 247

I Immunity..................................................................273–282 Immunoblot......................................................................263 Immunodetection .....................................177, 178, 180, 183, 189, 200, 244, 257, 258, 260 Immunoprecipitation ..............................57, 58, 64, 107–119, 175–184, 263 Inducible promoter pulsed induction .................................................203–208 Input ......................................... 24, 34, 39–41, 65, 66, 68, 80, 82, 83, 86, 87, 109, 111, 115–117, 119, 146, 180, 182, 302, 342, 355, 356 Integrated Genome Browser ....................................123–137 Internode .................................................................. 315, 321 Intron retention ........................................................ 132, 134 Iron ........................................................................... 227, 327

J JTK_CYCLE ............................................. 46, 47, 50, 53–55 Junction tracks ..........................................................130–132

L Late Elongated Hypocotyl (LHY) ......................... 58, 65, 66, 124, 125, 127–130, 133, 141, 204, 206, 219, 298–305, 314, 319, 339–342, 344, 347, 349–355 Limit-cycle oscillation ...................................... 342, 346–348 Luciferase Firefly (Photinus pyralis) luciferase ...........2, 3, 8, 326, 335 Renilla luciferase .............................................................2 Luciferin ........................ 2–5, 7, 229, 230, 329, 332, 333, 336 Luminescence ........................7, 216, 217, 219–221, 223–225

N Next generation sequencing................................ 94, 111, 150

O Output ...................................5, 15, 18, 34, 39, 41, 52, 54, 79, 80, 82–85, 150, 187, 217, 302, 348, 356

P Pathogen necrotrophic................................................................273 plant–pathogen interaction .........................................140 PCR quantitative PCR ....................................................65, 66 Perennial ................................................... 297–306, 313, 314 Period .............................................................4, 7, 13–43, 45, 53, 54, 108, 141, 157, 175, 187, 188, 215, 216, 218, 221, 228, 234, 274, 276–278, 280, 285, 291, 302, 303, 305, 314, 315, 317, 318, 322, 332, 334, 337, 343, 346–349, 351, 354 Peroxiredoxins ......................................... 239, 240, 242–244, 255–261, 264, 268, 269, 305 Phase ...............................4, 25, 27–29, 33, 34, 37, 39, 41, 42, 54, 73, 76–78, 94, 100, 107, 116, 165, 187–189, 193, 195–197, 200, 215, 261, 274, 276, 282, 290, 299, 300, 302, 317, 338, 343–349, 354 Phosphoproteomics ..................................................187–201 Photon-counting imaging ........................................ 220, 224 Photoperiodic pathway .............................................286–289 Physcomitrella patens ...............................................325–336 Polyacrylamide gels .......................................... 178, 181, 243 Polysome profiling ...................................................... 158, 165, 172 Poplar ...............................................................................151 Post-transcriptional ..........................107, 108, 123, 139–152, 338, 339, 356 Protein immunoprecipitation ...................................175–184 Protein–protein interaction ......................................175–184

PLANT CIRCADIAN NETWORKS: METHODS AND PROTOCOLS 361 Index Proteomics phosphoproteomics.............................................187–201 Protonema ................................................................330–333 Protoplasts ............................. 7, 242, 328, 329, 331, 332, 334 Pseudomonas syringae ......................................... 110, 273, 275 Pseudoresponse Regulator .......................... 58, 108, 287, 298

R Reactive oxygen species (ROS)................................. 239–268 Read mapping .......................................72, 74, 75, 80, 83–85 Read Quality assessment ..............................................80–82 Real time PCR .........................................65–68, 94, 98, 109, 113, 117, 118 Redox roGFP ................................................ 242, 254–256, 267 Reporter constructs................................................... 326, 335 Reverse transcription (RT) .....................94–96, 99, 101–103, 113, 117, 167, 293 RT-PCR ...................................94–97, 99, 103–105, 108, 111, 206, 207, 293 RFP Trap beads ........................................ 109–112, 115, 116 Rhythms ...............................................1–9, 14, 15, 209–212, 325–336, 338, 356 RNA RNA immunoprecipitation.................................107–119 RNA-protein interactions........................... 109, 118, 119 RNA-Seq ............................................... 71–89, 123–137 RNA-Seq read tracks ......................................... 131, 133

S Search Webtool ................................................................144 Season.......................................................285, 286, 292, 299, 303–304, 305, 313–323 seasonal regulation .......................229, 301, 302, 304, 306

Short interfering RNA .......................................................83 Simulation ...........................................13, 345, 348, 349, 352 Small RNA (sRNA) targets web tool...........................................................147 Splicing alternative ................................................... 108, 123–137 Sucrose density gradient ...................................................158

T Timing of Cab Expression 1 (TOC1) .................... 4, 6, 58, 65, 206, 216, 219, 286, 298–302, 304, 339–342, 344–346, 348, 352–355 TopCount ....................................................... 3, 5–8, 32, 228 Transcriptional ...................................93, 108, 124, 139–152, 175, 215, 298, 338–342, 356 Transcription factor binding site ......................................... 139–141, 144–152 Transcriptome ......................................... 45–55, 71–89, 107, 140, 151, 157 Transformation ....................................84, 218, 219, 325–336 Translation state ............................................... 158, 168, 170

V Vernalization ................................................... 285–287, 289, 291, 293, 294 Virtual Plant ........................................................... 75, 88–89

W Western blot .....................................178–183, 206, 243–244, 257–259, 261

X XPP-AUTO.............................................................337–356