Plant Cell Morphogenesis: Methods and Protocols [2nd ed.] 978-1-4939-9468-7;978-1-4939-9469-4

Table of contents :
Front Matter ....Pages i-xi
Essential Methods of Plant Sample Preparation for Light Microscopy (Aleš Soukup, Edita Tylová)....Pages 1-26
Selected Simple Methods of Plant Cell Wall Histochemistry and Staining for Light Microscopy (Aleš Soukup)....Pages 27-42
Chemical Fixation, Immunofluorescence, and Immunogold Labeling of Electron Microscopical Sections (Ilse Foissner, Margit Hoeftberger)....Pages 43-62
Essential Methods of Plant Sample Preparation for High-Resolution Scanning Electron Microscopy at Room Temperature (Jana Nebesářová)....Pages 63-76
Fluorescence Lifetime Imaging of Plant Cell Walls (Christine Terryn, Gabriel Paës)....Pages 77-82
Raman Spectroscopy in Nonwoody Plants (Dorota Borowska-Wykręt, Mateusz Dulski)....Pages 83-107
Image Analysis: Basic Procedures for Description of Plant Structures (Jana Albrechtová, Zuzana Kubínová, Aleš Soukup, Jiří Janáček)....Pages 109-119
From Data to Illustrations: Common (Free) Tools for Proper Image Data Handling and Processing (Fatima Cvrčková)....Pages 121-133
Visualizing and Quantifying In Vivo Cortical Cytoskeleton Structure and Dynamics (Amparo Rosero, Denisa Oulehlová, Viktor Žárský, Fatima Cvrčková)....Pages 135-149
Quantitative and Comparative Analysis of Global Patterns of (Microtubule) Cytoskeleton Organization with CytoskeletonAnalyzer2D (Birgit Möller, Luise Zergiebel, Katharina Bürstenbinder)....Pages 151-171
Using FM Dyes to Study Endomembranes and Their Dynamics in Plants and Cell Suspensions (Adriana Jelínková, Kateřina Malínská, Jan Petrášek)....Pages 173-187
Transient Gene Expression as a Tool to Monitor and Manipulate the Levels of Acidic Phospholipids in Plant Cells (Lise C. Noack, Přemysl Pejchar, Juraj Sekereš, Yvon Jaillais, Martin Potocký)....Pages 189-199
The Photoconvertible Fluorescent Protein Dendra2 Tag as a Tool to Investigate Intracellular Protein Dynamics (Alexandra Lešková, Zuzana Kusá, Mária Labajová, Miroslav Krausko, Ján Jásik)....Pages 201-214
Cellular Force Microscopy to Measure Mechanical Forces in Plant Cells (Mateusz Majda, Aleksandra Sapala, Anne-Lise Routier-Kierzkowska, Richard S. Smith)....Pages 215-230
Optical Trapping in Plant Cells (Tijs Ketelaar, Norbert de Ruijter, Stefan Niehren)....Pages 231-238
Sequential Replicas: Method for In Vivo Imaging of Plant Organ Surfaces that Undergo Deformation (Dorota Kwiatkowska, Sandra Natonik-Białoń, Agata Burian)....Pages 239-255
Time-Lapse Imaging of Developing Shoot Meristems Using A Confocal Laser Scanning Microscope (Olivier Hamant, Pradeep Das, Agata Burian)....Pages 257-268
Quantifying Plant Growth and Cell Proliferation with MorphoGraphX (Soeren Strauss, Aleksandra Sapala, Daniel Kierzkowski, Richard S. Smith)....Pages 269-290
Kinematic Characterization of Root Growth by Means of Stripflow (Tobias I. Baskin, Ellen Zelinsky)....Pages 291-305
Automated Image Acquisition and Morphological Analysis of Cell Growth Mutants in Physcomitrella patens (Giulia Galotto, Jeffrey P. Bibeau, Luis Vidali)....Pages 307-322
Live Cell Imaging of Arabidopsis Root Hairs (Tijs Ketelaar)....Pages 323-327
Morphological Analysis of Leaf Epidermis Pavement Cells with PaCeQuant (Birgit Möller, Yvonne Poeschl, Sandra Klemm, Katharina Bürstenbinder)....Pages 329-349
Imaging of Developing Metaxylem Vessel Elements in Cultured Hypocotyls (Takema Sasaki, Yoshihisa Oda)....Pages 351-358
Antisense Oligodeoxynucleotide-Mediated Gene Knockdown in Pollen Tubes (Martin Potocký, Radek Bezvoda, Přemysl Pejchar)....Pages 359-365
Plant Cell Lines in Cell Morphogenesis Research: From Phenotyping to -Omics (Petr Klíma, Vojtěch Čermák, Miroslav Srba, Karel Müller, Jan Petrášek, Josef Šonka et al.)....Pages 367-376
Back Matter ....Pages 377-380

Citation preview

Methods in Molecular Biology 1992

Fatima Cvrčková Viktor Žárský Editors

Plant Cell Morphogenesis Methods and Protocols Second Edition

METHODS

IN

MOLECULAR BIOLOGY

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

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

Plant Cell Morphogenesis Methods and Protocols Second Edition

Edited by

Fatima Cvrčková and Viktor Žárský Department of Experimental Plant Biology, Charles University, Prague, Czech Republic

Editors Fatima Cvrcˇkova´ Department of Experimental Plant Biology Charles University Prague, Czech Republic

ˇ a´rsky´ Viktor Z Department of Experimental Plant Biology Charles University Prague, Czech Republic

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-4939-9468-7 ISBN 978-1-4939-9469-4 (eBook) https://doi.org/10.1007/978-1-4939-9469-4 © Springer Science+Business Media, LLC, part of Springer Nature 2019 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. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover Illustration Caption: Growing epidermal cells color-coded according to several growth parameters (from Chapter 18) This Humana imprint is published by the registered company Springer Science+Business Media, LLC part of Springer Nature The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.

Preface Five years after the publication of the first edition of this volume, the focus of post-genomic land plant biology is definitely leaving the confines of a handful of major models, such as Arabidopsis thaliana or Physcomitrella patens, thus opening new spaces for defining and solving major questions of basic plant biology. Collecting wisdom and skills accumulated mostly from work on the founding molecular biology models can undoubtedly guide applications in other species, including crop plants. As editors of this volume, we strive to reflect this development in order to inspire future research in cellular aspects of land plant morphogenesis. Studying the dynamics of plant shapes, starting from the cellular level and advancing via tissues to organs and onward to the whole plant, is a truly fascinating perspective that we share with the nineteenth- and twentieth-century founders of our field. Here in Prague, we acknowledge continuous inspiration by Jan Evangelista Purkyneˇ and, in our field especially, his disciple Julius Sachs, who started his career in the German-speaking part of the Charles University in Prague and became the father of modern plant physiology (including pioneering studies of the processes of plant morphogenesis). We have worked on this volume in a building constructed in 1898 for the German Plant Physiology department, directed in those years by Professor Hans Molisch, author of Mikrochemie der Pflanzen (published after his move to Vienna in 1909). Several of the Czech contributors to this volume consider themselves “academic grandchildren” of Professor Bohumil Neˇmec, one of the fathers of experimental plant cell biology. When Neˇmec discovered the decisive role of starch statoliths in root columella for root gravitropism (1900), he immediately understood that, to function in graviperception, columellar cells need to be not only internally dynamically polarized but also connected in a communicative (i.e., signaling) network with other root cells. This indicated an intricate internal cellular structure and intercellular communication, beyond the imagination of scientists of those times. Neˇmec taught us, via his students (our teachers) and his impressive published volumes on plant biology, to understand tissues and cells as products, not mere constituents or “bricks,” of a living plant body as a whole. As in the first edition of this book, the first eight chapters of this volume (Chapters 1–8) focus on the visualization of plant cell structures, since seeing the objects of interest is an obvious prerequisite of understanding the processes that brought them into being. Chapters 1 and 2 present a contemporary take on light microscopy, the classical approach that was instrumental in establishing the plant cell biology field, These chapters are directly linked to a classic plant histochemistry methods book published by Bohumil Neˇmec—Botanical microtechnique (“Botanicka´ mikrotechnika” in Czech, Prague 1962). While light microscopy remains a central visualization method in plant cell biology, electron microscopy provides exciting insights into cellular ultrastructure. Chapters 3 and 4 describe a collection of useful electron microscopy techniques, including immunogold localization procedures. Chapters 5 and 6 then present techniques for in situ qualitative analysis of plant cell wall composition. Modern cell biology is dominated by digital data. This is reflected by the recurrent inclusion of digital image analysis protocols in several of the following chapters. Chapters 7 and 8 provide generally applicable techniques for quantitative image analysis, as well as for the only seemingly mundane task of presenting image data lege artis.

v

vi

Preface

The second part (Chapters 9–13) is devoted to qualitative and quantitative detection of the organization and dynamics of individual intracellular structures responsible for the generation of cell shape, in particular the two cytoskeletal systems, as well as endomembranes, whose structure and behavior can be studied using in vivo fluorescent markers. The third section (Chapters 14–19) is devoted to exciting new possibilities of manipulating intracellular structures by means of optical tweezers, and probing their mechanical features by Cellular Force Microscopy. It also covers detailed monitoring and quantifying structural dynamics of meristems and developing organs on the cellular level. The choice of the experimental model is, as a rule, tightly linked with the choice of questions that can be studied. It is hard to find a field where this would be more obvious than the study of cell morphogenesis. In the final six chapters (Chapters 20–25), we present specific techniques for studying model cell types such as filaments of the moss Physcomitrella patens, Arabidopsis root hairs, pavement cells, developing xylem vessels, pollen tubes, or plant cell lines. We decided to leave out several topics addressed in the first edition of this book. Several previously included methods received excellent coverage in recent volumes of the Methods in Molecular Biology series. This is, in particular, the case for additional electron microscopy techniques, such as cryofixation with freze substitution and subsequent immunodetection [1], electron microscopy tomography and 3D reconstruction [2, 3], or analyses of nuclear morphology [4]. Omission of other topics, such as automated microscopy application in forward genetics screens, use of laser microdissection to study gene expression, microfluidics applications, general techniques for transient gene expression in plant systems, or heterologous expression in yeast, was in effort to keep the present volume focused and its size manageable. We approached our task in editing this collection of protocols with the hope that this volume may become a source of inspiration for further research into the morphogenesis of plant cells, tissues, and organs. We are especially grateful to many colleagues—the best experts in their fields from all over the world—who accepted our invitation and contributed chapters to this volume, for making it more likely that our hope may be fulfilled. Prague, Czech Republic

Fatima Cvrcˇkova´ Viktor Zˇa´rsky´

References 1. Takeuchi M, Takabe K, Mineyuki Y (2016) Immunoelectron microscopy of cryofixed and freezesubstituted plant tissues. Methods Mol Biol 1474:233–242 2. Mai KKK, Kang BH (2017) Semiautomatic segmentation of plant Golgi stacks in electron tomograms using 3dmod. Methods Mol Biol 1662:97–104 3. Mai KKK, Kang MJ, Kang BH (2017) 3D printing of plant Golgi stacks from their electron tomographic models. Methods Mol Biol 1662:105–113 4. Desset S, Poulet A, Tatout C (2018) Quantitative 3D analysis of nuclear morphology and heterochromatin organization from whole-mount plant tissue using NucleusJ. Methods Mol Biol 1675:615–632

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

v ix

1 Essential Methods of Plant Sample Preparation for Light Microscopy . . . . . . . . . Alesˇ Soukup and Edita Tylova´ 2 Selected Simple Methods of Plant Cell Wall Histochemistry and Staining for Light Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alesˇ Soukup 3 Chemical Fixation, Immunofluorescence, and Immunogold Labeling of Electron Microscopical Sections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ilse Foissner and Margit Hoeftberger 4 Essential Methods of Plant Sample Preparation for High-Resolution Scanning Electron Microscopy at Room Temperature . . . . . . . . . . . . . . . . . . . . . . . Jana Nebesa´rˇova´

1

27

43

63

5 Fluorescence Lifetime Imaging of Plant Cell Walls . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Christine Terryn and Gabriel Pae¨s 6 Raman Spectroscopy in Nonwoody Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Dorota Borowska-Wykre˛t and Mateusz Dulski 7 Image Analysis: Basic Procedures for Description of Plant Structures . . . . . . . . . . 109 Jana Albrechtova´, Zuzana Kubı´nova´, Alesˇ Soukup, and Jirˇı´ Jana´cˇek 8 From Data to Illustrations: Common (Free) Tools for Proper Image Data Handling and Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Fatima Cvrcˇkova´ 9 Visualizing and Quantifying In Vivo Cortical Cytoskeleton Structure and Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Amparo Rosero, Denisa Oulehlova´, Viktor Zˇa´rsky´, and Fatima Cvrcˇkova´ 10

11

12

13

Quantitative and Comparative Analysis of Global Patterns of (Microtubule) Cytoskeleton Organization with CytoskeletonAnalyzer2D . . . . . . . . . . . . . . . . . . . 151 ¨ rstenbinder Birgit Mo¨ller, Luise Zergiebel, and Katharina Bu Using FM Dyes to Study Endomembranes and Their Dynamics in Plants and Cell Suspensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Adriana Jelı´nkova´, Katerˇina Malı´nska´, and Jan Petra´ˇsek Transient Gene Expression as a Tool to Monitor and Manipulate the Levels of Acidic Phospholipids in Plant Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Lise C. Noack, Prˇemysl Pejchar, Juraj Sekeresˇ, Yvon Jaillais, and Martin Potocky´ The Photoconvertible Fluorescent Protein Dendra2 Tag as a Tool to Investigate Intracellular Protein Dynamics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Alexandra Lesˇkova´, Zuzana Kusa´, Ma´ria Labajova´, Miroslav Krausko, and Ja´n Ja´sik

vii

viii

14

15 16

17

18

19 20

21 22

23

24

25

Contents

Cellular Force Microscopy to Measure Mechanical Forces in Plant Cells . . . . . . . Mateusz Majda, Aleksandra Sapala, Anne-Lise Routier-Kierzkowska, and Richard S. Smith Optical Trapping in Plant Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tijs Ketelaar, Norbert de Ruijter, and Stefan Niehren Sequential Replicas: Method for In Vivo Imaging of Plant Organ Surfaces that Undergo Deformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dorota Kwiatkowska, Sandra Natonik-Białon´, and Agata Burian Time-Lapse Imaging of Developing Shoot Meristems Using A Confocal Laser Scanning Microscope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Olivier Hamant, Pradeep Das, and Agata Burian Quantifying Plant Growth and Cell Proliferation with MorphoGraphX . . . . . . . . Soeren Strauss, Aleksandra Sapala, Daniel Kierzkowski, and Richard S. Smith Kinematic Characterization of Root Growth by Means of Stripflow . . . . . . . . . . . Tobias I. Baskin and Ellen Zelinsky Automated Image Acquisition and Morphological Analysis of Cell Growth Mutants in Physcomitrella patens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Giulia Galotto, Jeffrey P. Bibeau, and Luis Vidali Live Cell Imaging of Arabidopsis Root Hairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tijs Ketelaar Morphological Analysis of Leaf Epidermis Pavement Cells with PaCeQuant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Birgit Mo¨ller, Yvonne Poeschl, Sandra Klemm, ¨ rstenbinder and Katharina Bu Imaging of Developing Metaxylem Vessel Elements in Cultured Hypocotyls. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Takema Sasaki and Yoshihisa Oda Antisense Oligodeoxynucleotide-Mediated Gene Knockdown in Pollen Tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martin Potocky´, Radek Bezvoda, and Prˇemysl Pejchar Plant Cell Lines in Cell Morphogenesis Research: From Phenotyping to -Omics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ˇ erma´k, Miroslav Srba, Karel Mu ¨ ller, Jan Petra´ˇsek, Petr Klı´ma, Vojteˇch C ˇ Josef Sonka, Luka´ˇs Fischer, and Zdeneˇk Opatrny´

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

215

231

239

257 269

291

307 323

329

351

359

367

377

Contributors JANA ALBRECHTOVA´  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic TOBIAS I. BASKIN  Biology Department, University of Massachusetts, Amherst, MA, USA RADEK BEZVODA  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic JEFFREY P. BIBEAU  Department of Biology and Biotechnology, Worcester Polytechnic Institute, Worcester, MA, USA DOROTA BOROWSKA-WYKRE˛T  Department of Biophysics and Morphogenesis of Plants, University of Silesia in Katowice, Katowice, Poland KATHARINA BU¨RSTENBINDER  Department of Molecular Signal Processing, Leibniz Institute of Plant Biochemistry (IPB), Halle (Saale), Germany AGATA BURIAN  Department of Biophysics and Morphogenesis of Plants, University of Silesia in Katowice, Katowice, Poland ˇ ERMA´K  Department of Experimental Plant Biology, Faculty of Science, Charles VOJTEˇCH C University, Prague, Czech Republic FATIMA CVRCˇKOVA´  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic PRADEEP DAS  Laboratoire de Reproduction et De´veloppement des Plantes, Universite´ de Lyon, UCB Lyon 1, ENS de Lyon, INRA, CNRS, Lyon, France NORBERT DE RUIJTER  Laboratory of Cell Biology, Wageningen University, Wageningen, The Netherlands MATEUSZ DULSKI  Institute of Material Science, University of Silesia in Katowice, Chorzow, Poland; Silesian Center for Education and Interdisciplinary Research, Chorzow, Poland LUKA´Sˇ FISCHER  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic ILSE FOISSNER  Department of Biosciences, University of Salzburg, Salzburg, Austria GIULIA GALOTTO  Department of Biology and Biotechnology, Worcester Polytechnic Institute, Worcester, MA, USA OLIVIER HAMANT  Laboratoire de Reproduction et De´veloppement des Plantes, Universite´ de Lyon, UCB Lyon 1, ENS de Lyon, INRA, CNRS, Lyon, France MARGIT HOEFTBERGER  Department of Biosciences, University of Salzburg, Salzburg, Austria JA´N JA´SIK  Plant Science and Biodiversity Center, Institute of Botany, Slovak Academy of Sciences, Bratislava, Slovakia YVON JAILLAIS  Laboratoire Reproduction et De´veloppement des Plantes, Universite´ de Lyon, ENS de Lyon, CNRS, INRA, Lyon, France JIRˇI´ JANA´CˇEK  Institute of Physiology of the Czech Academy of Sciences, Prague, Czech Republic ADRIANA JELI´NKOVA´  Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic TIJS KETELAAR  Laboratory of Cell Biology, Wageningen University, Wageningen, The Netherlands

ix

x

Contributors

DANIEL KIERZKOWSKI  Max Planck Institute for Plant Breeding Research, Cologne, Germany; Institut de Recherche en Biologie Ve´ge´tale, University of Montre´al, Montreal, QB, Canada SANDRA KLEMM  Department of Molecular Signal Processing, Leibniz Institute of Plant Biochemistry (IPB), Halle (Saale), Germany PETR KLI´MA  Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic MIROSLAV KRAUSKO  Plant Science and Biodiversity Center, Institute of Botany of the Czech Academy of Sciences, Slovak Academy of Sciences, Bratislava, Slovakia ZUZANA KUBI´NOVA´  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic ZUZANA KUSA´  Plant Science and Biodiversity Center, Institute of Botany, Slovak Academy of Sciences, Bratislava, Slovakia DOROTA KWIATKOWSKA  Department of Biophysics and Morphogenesis of Plants, University of Silesia in Katowice, Katowice, Poland ´ MARIA LABAJOVA´  Plant Science and Biodiversity Center, Institute of Botany, Slovak Academy of Sciences, Bratislava, Slovakia ALEXANDRA LESˇKOVA´  Plant Science and Biodiversity Center, Institute of Botany, Slovak Academy of Sciences, Bratislava, Slovakia MATEUSZ MAJDA  Max Planck Institute for Plant Breeding Research, Cologne, Germany KATERˇINA MALI´NSKA´  Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic BIRGIT MO¨LLER  Institute of Computer Science, Martin Luther University HalleWittenberg, Halle (Saale), Germany KAREL MU¨LLER  Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic SANDRA NATONIK-BIAŁON´  Department of Biophysics and Morphogenesis of Plants, University of Silesia in Katowice, Katowice, Poland ˇ eske´ Budeˇjovice, Czech JANA NEBESA´RˇOVA´  Biology Centre of CAS, Institute of Parasitology, C Republic; Faculty of Science, Charles University, Prague, Czech Republic STEFAN NIEHREN  Molecular Machines & Industries GmbH, Eching, Germany LISE C. NOACK  Laboratoire Reproduction et De´veloppement des Plantes, Universite´ de Lyon, ENS de Lyon, CNRS, INRA, Lyon, France YOSHIHISA ODA  Center for Frontier Research, National Institute of Genetics, Mishima, Shizuoka, Japan; Department of Genetics, SOKENDAI (Graduate University for Advanced Studies), Mishima, Shizuoka, Japan ZDENEˇK OPATRNY´  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic DENISA OULEHLOVA´  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic; Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic GABRIEL PAE¨S  Fractionation of AgroResources and Environment (FARE) Laboratory, INRA, University of Reims Champagne-Ardenne, Reims, France PRˇEMYSL PEJCHAR  Institute of Experimental Botany, Czech Academy of Sciences, Prague, Czech Republic; Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic

Contributors

xi

JAN PETRA´SˇEK  Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic; Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic YVONNE POESCHL  Institute of Computer Science, Martin Luther University HalleWittenberg, Halle (Saale), Germany; iDiv, German Integrative Research Center for Biodiversity, Leipzig, Germany MARTIN POTOCKY´  Institute of Experimental Botany, Czech Academy of Sciences, Prague, Czech Republic; Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic AMPARO ROSERO  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic; Coordinacion de Innovacion Regional, C.I. Turipana´, Monterı´a, Cordoba, Colombia ANNE-LISE ROUTIER-KIERZKOWSKA  Max Planck Institute for Plant Breeding Research, Cologne, Germany; Institut de Recherche en Biologie Ve´ge´tale, University of Montre´al, Montreal, QC, Canada ALEKSANDRA SAPALA  Department of Biosystems Science and Engineering, ETH Zurich, Basel, Switzerland; Max Planck Institute for Plant Breeding Research, Cologne, Germany TAKEMA SASAKI  Center for Frontier Research, National Institute of Genetics, Mishima, Shizuoka, Japan JURAJ SEKERESˇ  Institute of Experimental Botany, Czech Academy of Sciences, Prague, Czech Republic; Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic RICHARD S. SMITH  Max Planck Institute for Plant Breeding Research, Cologne, Germany JOSEF SˇONKA  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic ALESˇ SOUKUP  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic MIROSLAV SRBA  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic SOEREN STRAUSS  Max Planck Institute for Plant Breeding Research, Cologne, Germany CHRISTINE TERRYN  PICT Platform, University of Reims Champagne-Ardenne, Reims, France EDITA TYLOVA´  Department of Experimental Plant Biology, Faculty of Science, Charles University, Prague, Czech Republic LUIS VIDALI  Department of Biology and Biotechnology, Worcester Polytechnic Institute, Worcester, MA, USA ˇ A´RSKY´  Department of Experimental Plant Biology, Faculty of Science, Charles VIKTOR Z University, Prague, Czech Republic; Institute of Experimental Botany of the Czech Academy of Sciences, Prague, Czech Republic ELLEN ZELINSKY  Biology Department, University of Massachusetts, Amherst, MA, USA LUISE ZERGIEBEL  Department of Molecular Signal Processing, Leibniz Institute of Plant Biochemistry (IPB), Halle (Saale), Germany

Chapter 1 Essential Methods of Plant Sample Preparation for Light Microscopy Alesˇ Soukup and Edita Tylova´ Abstract There are various preparatory techniques for light microscopy permitting access to the inner structure of plant body and its development. Minute objects might be processed as whole-mount preparations, while voluminous ones should be separated into smaller pieces. Here we summarize some of the “classical” techniques to cut more voluminous objects into slices and access their inner structure either for simple anatomical analysis or for further processing (e.g., histochemistry, immunohistochemistry, in situ hybridization, enzyme histochemistry). Key words Paraffin, Sections, Freehand sectioning, Fixation, Whole mount, Serial sections, Cryotome, Hand microtome

1

Introduction There are various ways of preparation of plant objects for investigation with light microscopy. Correct selection of an appropriate technique largely depends on the equipment available, but nature, optical character, complexity, and size of the object and purpose of the preparation have a major role. Here we present a set of simple techniques which might provide vast, however not exhaustive, information on structural and cytological features of cells, tissues, and organs. Tissues, organs, or explants, which are not voluminous and optically dense, might be processed as a cleared whole-mount preparations. Such a way of preparation became very popular with advent of confocal microscopy and Arabidopsis as a model plant. However, there are many objects where cuttings or macerations are necessary to gain adequate information on internal structure. Available sectioning techniques allow for preparation of sections of variable thicknesses according to the intended application. Tissue preservation (fixation) and embedding into a supporting matrix are common initial steps involved in most sectioning methods, which determine the quality and final application of microscopic sections.

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_1, © Springer Science+Business Media, LLC, part of Springer Nature 2019

1

2

Alesˇ Soukup and Edita Tylova´

Sectioning of fresh (not fixed) and/or nonembedded samples is a valuable alternative to avoid systemic artifacts of sample processing. 1.1

Fixation

1.2 Cleared WholeMount Preparations

Fixation is commonly the initial step of the sequence. The choice of proper fixation has strong consequences for later use of the preparation and its subsequent processing. We include here only the two common basic procedures using FAA and buffered formaldehyde. FAA (formalin–acetic acid–alcohol) penetrates rapidly and is suitable for general anatomical or morphological work. However, preservation of cytological details, antigens or enzyme activity is far less satisfactory comparing to formaldehyde. Fixation of samples with Clark’s and Carnoy’s fluids, alcohols, glutaraldehyde, acrolein, carbodiimides, chilled methanol or acetone, and others should be considered an alternative depending on the goal of the preparation [1–4]. The process of fixation includes both penetration of the fixative into the tissue and its action within the tissue. While alcoholelicited coagulation is a rapid process, saturation of the chemical linkages within the tissue by formaldehyde takes 1–2 days [5]. Diffusibility of fixative (distance that the fixative diffuses in an hour within the object) varies strongly among tissues and fixatives, being about 25 higher for ethanol than for formaldehyde solutions [6]. This fact should be considered during sampling as size and character of the object strongly influence penetration of the fixative. Most experimental tests of fixative penetration use animal tissues (see for example refs. 5, 7), and only little data is available from plant tissues [8, 9]. The reasonable expectation of formaldehyde penetration rate does not exceed more than few millimeters per hour in compact plant tissues. Therefore, it is difficult to get a coherent rule for estimation of fixation time regarding the size and character of the object. Low-pressure (“vacuum”) infiltration of the tissue might be required to facilitate penetration of an aqueous fixative with considerable surface tension into air-filled intercellular spaces. On the other hand, once filled with a fixative, such intercellular spaces might serve as an important entryway for more voluminous samples. Cleared whole-mount preparations allow for focusing on minute objects (usually not more than few hundreds of micrometers deep) and gaining information on their inner arrangement. In fact, there are several attitudes as to clearing an object. Removal of pigments, inclusions, and most cellular content decreases optical density of the object and improves the transparency of the tissue and thus enables access to its inner structure [3, 10]. Treatment using sodium chloride [11], hydrogen peroxide [12], strong alkali or acids [13, 14], phenol [15], lactic acid [16–18], chloral hydrate [19–21], and their combinations are commonly used. Alternative saturation of the object with compounds of high refractive index

Plant Preparations for Light Microscopy

3

decreases light dispersion and increases transparency of the tissue [22–24]. Various procedures combine these two attitudes. Here we present a simple protocol of gentle tissue clearing with high refractive index solution, which preserves most of the cellular content. We have introduced using sodium iodide solution [25] as a high refractive index nontoxic alternative to chloral hydrate, which is a regulated narcotic in most countries. The procedure is not selfreliant for highly pigmented and highly optically dense (e.g., secondary xylem) tissues and should be combined with pigmentation removal in such a case. An alternative protocol of ClearSee solution [26] and its combination with other fluorochromes and fluorescent proteins should be also recommended [27]. 1.3

Hand Sectioning

1.4 Paraffin Embedding and Sectioning

Hand sectioning is fast and easy method of fresh/fixed specimen sectioning. While it might seem old fashioned in a modern equipment-loaded laboratory, if done skillfully, it helps us gain quickly substantial information on structure and in combination with various detection techniques also on composition and other parameters of tissues. Freehand sectioning with a razor blade is the simplest option and should be reckoned as a basic level laboratory skill. Hand microtome and the straight razor blade (Fig. 1a) can push the sectioning further to achieve series of sections of standardized thickness (50 μm is realistic for most tissues). Hand sectioning has no necessity for infiltration and embedding. For smaller objects, additional external mechanical support might be required to facilitate manipulation or fixation into the clamp of hand microtome. We commonly use elder pith (soft dead parenchymatous tissue), but other material (carrot, styrofoam, potato, roll of Parafilm, etc.) or encasing into paraffin or agarose block surrounding the object during sectioning [28, 29] might be used. In fact the hand microtome sectioning can provide sections similar to that of vibratome. Quality of the cutting edge is the most limiting factor, and high-quality disposable razor blades (not the single-sided technical ones) or well-maintained straight razor (requires proper honing and stropping) is crucial for sectioning. The other procedures presented in this selection will involve specimen infiltration and embedding with a supporting matrix to form blocks suitable for sectioning. Such embedding allows for thinner (less than 10 μm) and routine serial sections. Paraffin is the very classical embedding medium introduced into microtechnique by Klebs [30]. Paraffin melts at rather high temperatures (54–60  C), is strongly hydrophobic, and does not allow for routine sectioning below approx. 3 μm. In spite of these disadvantages, it is still the most common embedding medium. Easy cutting and joining of sections into ribbons allow for straightforward routine of serial sections. Its high hydrophobicity requires strict dehydration of the object and use of intermedium

4

Alesˇ Soukup and Edita Tylova´

Fig. 1 Hand sectioning can be done either with bare hand or with hand microtome (a), a simple device with a micrometric screw, which allows for stepwise adjustment of section thickness. Scale on the microtome is normally graduated in 10 μm increments. Object fixed into the central clamp is cut with a straight razor or other suitable blade. To slide the blade along the glass plate smoothly, the blade should have a flat grind on the glass touching side. Press the straight blade slightly on the glass plate of hand microtome with thumb and smoothly slide along to cut the sections (b). Notice the position of hands during freehand sectioning (c). Section should be kept permanently moistened with a drop of water or buffer. To strop the straight blade (d), place the blade flat on the strop and draw it spine first along the strop so that the whole length of edge is treated. Rotate the blade over its spine, so the edge moves away from the strop and draw the blade back. Repeat as long as necessary. (e) Simple sections staining holders made of Eppendorf vials, tubing ring, and fine mesh

(intermediate anhydrous paraffin solvent) to completely saturate tissue with paraffin before embedding. Butanol is the most commonly used intermedium, which substitutes the originally more common and toxic xylene. In our laboratory we generally use nbutanol. t-Butanol is a more efficient solvent of paraffin and more potent to be used for infiltration. However, high melting point (Tm 25  C, frequently solid at lab temperature) and higher price of tbutanol make n-butanol usually the better option. There are various protocols for the paraffin infiltration and embedding, which might differ in tissue damage and time consumption. The phase of dehydration and paraffin infiltration are steps which usually induce most of the tissue shrinkage. To minimize volume changes caused by intense solvent exchange, a gradual series of solutions with decreasing water content is commonly employed. Various solvents were proposed for use in dehydration (isopropanol, acetone, methyl cellosolve, etc.) and paraffin

Plant Preparations for Light Microscopy

5

infiltration (chloroform, xylene, n-propanol, n- or t-butanol; reviewed in ref. 3). Ethanol–butanol dehydration series [31, 32] has become a method of choice in our lab. Combination of gradual dehydration and concurrent introduction of intermedia minimizes distortion and saves time. Damage and/or hardening attributed to longer action of higher temperature during infiltration with paraffin reported in literature was not observed to be that significant in our hands, but damage caused by overly fast infiltration progress or improper elimination of intermedium was recorded rather frequently. Protocol of paraffin oil-regulated rate of infiltration should be mentioned [31] in this context. This protocol uses a mixture of butanol and paraffin oil (1:1) instead of pure butanol to saturate the objects before paraffin infiltration. Paraffin oil has higher viscosity and, if slowly replaced with melted paraffin in later steps, reduces shrinkage of tissue induced during paraffin infiltration. The whole procedure of sample processing toward paraffin embedding is a chain of events. Each of them might introduce artifacts in the preparation, which cannot be corrected later on and accumulation of errors therefore commonly takes place. An alternative to paraffin embedding is embedding using low-temperature melting Steedman’s wax (see ref. 33), which is suitable for objects sensitive to higher temperature (e.g., sections for immunodetection), and infiltration protocol is significantly shorter. On the other hand, sectioning, flattening of sections on slides, and storage of samples are slightly more complicated owing to the hygroscopic nature of the polyethylene glycol distearatebased wax. There are various types of resins used for sample embedding and sectioning (Technovit, LR White, Lowicryl, GMA, Spurr, etc.) that differ in hydrophobicity, hardness, and sectioning properties. Resin-embedded objects can be sectioned to thinner slices (less than 1 μm) to achieve higher degree of cytological details. Embedding into and sectioning of resins is beyond the scope of this chapter. There is an excellent manual [3] for such protocols. The sectioning of a paraffin-embedded object is carried out on various types of microtomes as described later. 1.5 Sectioning of Frozen Material

Sectioning of frozen material does not require extensive sample dehydration and embedding medium infiltration. Cryosections are suitable for a wide range of light-microscopy applications (immunohistochemistry, in situ hybridization, enzyme histochemistry, etc.), but it should be mentioned that it might not be so straightforward to gain good quality sections for plant tissues. Fixed or fresh (not fixed) samples might be processed according to intended application. Standard thickness of sections is 8–20 μm, but thickness down to 3 μm is attainable for some samples using standard cryomicrotome. Objects are encased into cryoembedding media, which acts as an object-surrounding matrix for sectioning.

6

Alesˇ Soukup and Edita Tylova´

Freezing is the critical step of procedure, which strongly determines quality of sections. Freezing procedure should prevent formation of large ice crystals inside the sample, which causes structural damage, and related sectioning problems. Highly vacuolated plant tissues are therefore rather complicated objects from this point of view. There are two principal approaches to minimize freezing distortion—flash freezing or cryoprotection pretreatment. Flash freezing approach prevents formation of large hexagonal ice crystals due to high freezing rate and small cubic crystals or even vitreous ice should form. Efficiency of the procedure can be further increased under high-pressure conditions [34]. Isopentane supercooled with liquid nitrogen or with solid carbon dioxide is frequently used to ensure proper heat transfer from object. However, supercooling is suitable only for small specimens and even in such a case low thermal conductivity of biological samples is a limitation for freezing rate [35]. The other approach restricts formation of large ice crystals due to the presence of rather high concentrations of cryoprotective solutes. Sucrose is a common cryoprotectant used in a wide range of concentrations from 10% to over 75% [36–38]. Infiltration with 8–15% glycerol [39], 10% dimethyl sulfoxide [40] or polyvinyl alcohol, and polyethylene glycol mixtures [41] can be used. Freezing rate is far less critical for the cryoprotected specimen, and freezing directly in the cryostat chamber (freezing shelf) is possible. The process of antifreeze treatment takes several hours and therefore requires foregoing fixation of samples to minimize processing-related artifacts. Besides freezing of the object, a proper setup of cryotome (temperature, antiroll plate, blade settings) is crucial for successful sectioning.

2 2.1

Materials Fixation

1. FAA (formalin–acetic acid–alcohol): Mix together 50% (v/v) of ethanol, 5% (v/v) of acetic acid, 5% (v/v) of formalin, and 40% (v/v) of distilled water. (Volumes should be adjusted according to stock ethanol and acetic acid concentration; for variations see Note 1.) 2. 4% formaldehyde in 50 mM phosphate buffer (pH 7.2): Dissolve 8% (w/v) of paraformaldehyde (PFA) in distilled water; to facilitate dissolution, add minimal volume of 1 M KOH solution (approx. 200 μl per 100 ml) and warm the solution up to ~60  C in a fume hood. When PFA is dissolved (the solution comes clear), add equal volume of 100 mM phosphate buffer of proper pH (selected according to the purpose at hand). Check pH and titrate to required pH with 1 M HCl if necessary (see Note 2).

Plant Preparations for Light Microscopy

7

Table 1 Phosphate buffer composition x

y

pH

x

y

pH

93.5

6.5

5.7

45.0

55.0

6.9

92.0

8.0

5.8

39.0

61.0

7.0

90.0

10.0

5.9

33.0

67.0

7.1

87.7

12.3

6.0

28.0

72.0

7.2

85.0

15.0

6.l

23.0

77.0

7.3

81.5

18.5

6.2

19.0

81.0

7.4

77.5

22.5

6.3

16.0

84.0

7.5

73.5

26.5

6.4

13.0

87.0

7.6

68.5

31.5

6.5

10.5

89.5

7.7

62.5

37.5

6.6

8.5

91.5

7.8

56.5

43.5

6.7

7.0

93.0

7.9

51.0

49.0

6.8

5.3

94.7

8.0

3. Phosphate buffer: Based on the desired pH (Table 1) mix together x ml of 0.2 M acid sodium phosphate (27.8 g NaH2PO4 in 1 l) + y ml of 0.2 M middle sodium phosphate (53.65 g Na2HPO4·7H2O in 1 l) fill up to 200 ml with distilled water to prepare 100 mM buffer. 2.2 Whole-Mount Preparation

1. NaI-based clearing solution for whole mounts: Dissolve 0.04 g of sodium thiosulfate (Na2S2O3) in 20 ml of 65% (aq. v/v) glycerol. Add and dissolve 17 g of sodium iodate (NaI). Add 2% (v/v) of DMSO to the final solution. The final solution should be clear and colorless with refractive index close to 1.5. 2. 15%, 30%, and 65% (aq. v/v) glycerol containing 2% (v/v) of dimethyl sulfoxide (DMSO).

2.3 Dehydration, Paraffin Infiltration, and Embedding

1. Ethanol–butanol dehydration series: Composition of individual steps is specified in Table 2. 2. Anhydrous ethanol and butanol: To efficiently remove water from the standard stock butanol or 96% ethanol (see ref. 42), pour the solvent into a flask and introduce enough of desiccant (approx. 1/5 of volume). Desiccant can be either solventdrying molecular sieve (3 A˚ for both butanol and ethanol) or anhydrous salt (e.g., K2CO3, CaSO4, or CuSO4), which binds the water but does not dissolve in alcohol. Let the capped flask stand overnight. Filtrate or decant water-free solvent and keep

8

Alesˇ Soukup and Edita Tylova´

Table 2 Ethanol–butanol dehydration series Step no.

Distilled water (%)

Ethanol (%)

Butanol (%)

1

70

20

10

2

60

25

15

3

45

30

25

4

30

30

40

5

20

25

55

6

10

20

70

7

–

15

85

8

–

–

100

it in a tightly closed flask to prevent air humidity entrance. To regenerate the molecular sieves as well as hydrated salt, place them into drying oven at 250  C in a thin layer for about 2 h. 3. Paraffin: paraffin is a mixture of long-chain alkanes. There are various types of paraffin suitable for embedding and sectioning. The classical method is based on recycling of suitable paraffin and alchemy of preparation of such paraffin [31]. Most laboratories use commercially available and easily accessible histological paraffin these days. Various brands are on the market (e.g., Paraffin, Tissue-Tek, Paramat, Paraplast, Histoplast, see Note 3). 4. Gelatine-subbing coated microscopic slides (alum gelatin adhesive, chrome alum; [43]): Place 0.5 g of pure gelatin in 100 ml of distilled water and heat to approx. 45  C to dissolve it completely. Add 0.05 g of KCr(SO4)2·12H2O (the usage of other alums is also possible) and dissolve and filter the solution. Immerse a set of clean slides in staining rack into the solution for 10 s, blot excess of solution, and let the slides dry (48 h at room temperature or 12 h at 50  C); protect slides from dust. The slides can be submerged several times (2–5 times) to strengthen coating layer if necessary (see Note 4). 5. Poly-L-lysine-coated slides: Dilute 10 poly-L-lysine stock solution (0.1% w/v) to prepare working solution. Immerse clean slides into the solution for 10 min to 1 h. Dry and store coated slides in dust-free dry place; 4  C is recommended for longer storage (see Note 4). 6. Glycerol albumen: Mix carefully egg white with an equal volume of pure glycerol. Filter the mixture over glass wool or few layers of gauze. Add 1% of sodium salicylate or thymol as a preservative (causes background autofluorescence!).

Plant Preparations for Light Microscopy

9

Alternatively use 0.5–1 g NaN3 (Be careful, toxic!). Smear a tiny amount (pinhead volume) evenly over a clean grease-free slide with your finger to make a very fine (not wet) coating. Protein precipitates form on the slides if a high amount of albumen adhesive is used (see Note 4). 2.4

Cryosectioning

1. High-viscosity cryoembedding medium (see ref. 41, Note 5): Dissolve 65–75 g of polyvinyl alcohol (PVA) 56–98 in 1 l of distilled water or phosphate buffer (pH 7.4; 50 mM). Warm up to 100  C to completely dissolve it. Add 10 ml of Tween 20, 0.5–1 g NaN3 (preservative for long-term storage), and 40 ml polyethylene glycol 400. Optionally supplement carboxymethylcellulose (CMC) to increase medium thickness up to a semisolid gel. Add 7–10 g CMC powder on the surface of the medium, leave to rehydrate overnight, and mix well. Centrifuge to eliminate air bubbles. 2. Sucrose solutions: 3%, 10%, and 20% w/v solutions of sucrose in 0.1 M phosphate buffer (pH 7.4).

2.5 Staining of Paraffin Sections

1. Safranin O staining solution: Dissolve 3 g of Safranin O and 4 g of sodium acetate in 100 ml of 96% ethanol. Add 8 ml of formalin (40% formaldehyde). Dilute 1:1 with 50% ethanol before use. 2. Picric acid solution: Saturated solution of picric acid in 96% ethanol. 3. Fast Green FCF staining solution: Dissolve 0.5 g of Fast Green FCF in 100 ml of 100% ethanol. Add 100 ml of clove oil. 4. Hematoxylin (Ehrlich’s) staining solution: Dissolve 1 g of hematoxylin in 50 ml of 100% ethanol. Add 5 ml of acetic acid, 50 ml of glycerol, and 50 ml of saturated solution of KAl(SO4)2. Staining solution should be left to mature (oxidize) either on air and light for several days, or add 0.1 g of KIO3 to decrease maturation time to several hours.

3

Methods

3.1 Fixation of Samples

1. Cut samples of adequate size to allow rapid access of fixative to innermost tissues. In general the smaller the better. On the other hand, the size of the structure of interest and/or cell size and desire of the investigator should be considered during sampling. Use sharp razor blade to minimize damage in the vicinity of cut edge. 2. Submerge samples into adequate volume (see Note 6) of fixative solution immediately after excision. Optimally, cut tissues under suitable buffer, water, or cultivation solution to avoid drying. We found 20 ml scintillation vials to be convenient vessels for fixation.

10

Alesˇ Soukup and Edita Tylova´

3. Alcoholic solutions easily fill in intercellular spaces due to low surface tension. If aqueous solutions are used, application of lower pressure (“vacuum infiltration”) might be necessary to substitute air with the fixative solution. A vacuum pump connected to a plastic desiccator via a vacuum regulator allows for controlled gradual decrease of pressure within the chamber. The rate of pressure drop depends on the nature of object and fixative used and should be adjusted accordingly. In general decrease should not cause boiling of the solution, but only slowly escaping stream of bubbles should be stimulated. Bring the samples slowly down to minimum pressure of the pump (approx. 5 mBar), turn off the vacuum line, and let the samples to equilibrate within the chamber for 10–20 min. Then let air slowly in to fill the chamber up again. The reintroduction of pressure should be gradual and as gentle as possible to fill “vacuum” within the sample intercellular spaces with solution during this period. A quick release of pressure difference might cause collapse of the intercellular spaces. 4. Let the samples to be fixed for a selected period of time (see Notes 7 and 8). 3.2 Simple Protocol for Whole Mounts or Clearing of Thick Sections

The procedure is optimized for Arabidopsis seedlings and might need minor readjustment for other samples. Multiwell culture plates are convenient to process larger sets of samples (see Note 9). 1. Fix samples in 4% formaldehyde buffered to pH 7.2 (25 mM) overnight. 2. Wash out fixative with 15% (aq. v/v) glycerol containing 2% (v/v) of dimethyl sulfoxide (DMSO) and leave for 30 min. 3. Replace the solution with 30% glycerol containing 2% of DMSO and leave for 30 min. 4. Transfer into 50% glycerol with 0.1% Triton and leave for 30 min. 5. Replace solution with 65% glycerol containing 2% of DMSO and leave for 30 min. 6. Mount the objects into NaI clearing solution and apply coverslip. Let the objects to clear up. In most cases 24 h is sufficient, for more voluminous objects time should be prolonged. 7. Preparations can be saved for weeks in the dark at 4  C.

3.3 Free-Hand Sectioning

Good quality double-sided razor blade is indispensable to successfully cut objects in bare hands. Quality of the blade makes strong limitation to the quality and attainable thickness of the sections. Break the blade longitudinal into halves is rather convenient practice. Besides better handling it is easier to control which side is still fresh and having good edge.

Plant Preparations for Light Microscopy

11

1. Grip the sample as indicated in Fig. 1c. For larger axially symmetric objects, it does not make much sense to strive to cut complete sections. More convenient is to get partial but thinner sections. If the object is too thin to be griped, it should be supported with some moist material and cut within such a material. We prefer soft elder pith soaked with appropriate buffer or water. 2. Wet the blade and cut the section holding the object more/less vertically (Fig. 1c). It is handy to use fine brush to collect and manipulate sections. Always keep the sections in solution as drying of the tissue is destructive and rapid at lab temperature. 3. Holders made of Eppendorf vials (with conical part cut off), a ring of tubing (inside diameter 10 mm), and fine mesh (Fig. 1e) might be used for convenient handling of sections (see also ref. 44). 3.4 Hand-Microtome Sectioning

Straight razor blade, which is used with a hand microtome, should be kept very sharp during the sectioning (for maintenance see Note 10 and Fig. 1d). 1. Preparation of the sample is identical to freehand sectioning (see Subheading 3.3, step 1). 2. Clamp the specimen into the central cylinder of the microtome so that it extends over the flat glass plate. If necessary, use supporting material (e.g., water-soaked elder pith or carrot sticks) to fix small specimens in appropriate position similarly to freehand sectioning (Fig. 1b). 3. Carefully place the flat side of straight blade on the glass plate and cut the object to align it with the plate (sectioning plane). 4. The straight razor blade should be laid down completely and slide smoothly when drawn along the glass plate. Be careful not to touch the glass plate or “cut it” with blade edge directly as it can be easily damaged this way. 5. Keep the specimen moist all the time. Wet it with small drops of water from brush to prevent drying and allow sections to float effortlessly up onto the razor blade (Fig. 1b). 6. Add a drop of water on blade and collect floating sections with fine brush (or dropper in the case of very small specimens) for further processing.

3.5 Paraffin Embedding and Sectioning

1. Fix samples as indicated above. Label samples with a pencil on a slip of cardboard, as graphite lead is stable in any solvent. The cardboard will pass together with samples through the dehydration and infiltration series and will be finally embedded into paraffin.

12

Alesˇ Soukup and Edita Tylova´

2. Wash the fixative out of your samples for 2 15 min. For wash use the same water content of solution as it has been used in the fixative. In case of formaldehyde, use the buffer included in the fixative, for FAA use ethanol of approximately the same concentration as in the fixative. 3. Gradually dehydrate objects and exchange dehydrating solvent for paraffin-dissolving intermedium. Thorough dehydration is indispensable for later successful paraffin infiltration. Starting point of the dehydration series should be selected according to fixative as in previous step (e.g., third step of ethanol–butanol series for 50% FAA). Pass samples through higher steps of ethanol–butanol dehydration series with adequate time in each dehydration step (see Note 11). The sample MUST NOT dry out during any dehydration step. Use an adequate volume of dehydration solution in relation to sample volume to keep its dehydration capacity (see Note 12). At least 100 sample volume might be a good thumb rule. 4. Repeat the anhydrous butanol bath (2 in total) to completely remove remaining ethanol from samples before starting paraffin infiltration. 5. Gradually introduce paraffin to fully infiltrate the objects and exhaustively eliminate butanol (or any other intermedium) from the samples in the end. Too rash infiltration is the most common reason of object shrinkage. That is because butanol escapes faster from the object than paraffin is able to replace it and compensate for volume changes. Timing of individual steps presented below is informative and should be adjusted according to the object. Place samples in 100% (water-free) butanol in suitable vessels (we use 50 ml vessels with cap for infiltration). Add chips of paraffin (approx. 1/5 of butanol volume) and let them stand for 1 day at laboratory temperature. 6. Place covered dishes at 40  C oven and let paraffin dissolve. Add enough paraffin to keep a few undissolved chips on the bottom and let stay during the day. 7. Open the dishes in the end of the day and let overnight. If the paraffin is completely dissolved add more. Butanol will evaporate slowly; be careful not to let the samples dry. 8. Increase temperature to 58  C, add paraffin, and let stand for 1 day (samples can stand even over weekend in this or later steps). 9. Pour off approx. one-third of the paraffin–butanol mixture and bring to original volume with melted paraffin. Let samples infiltrate for 3 h or longer.

Plant Preparations for Light Microscopy

13

10. Replace half of the mixture with melted paraffin (3–12 h). Repeat this step once more. 11. Let open vessels stand overnight in oven to completely evaporate residual butanol. It is convenient to apply low pressure in vacuum oven (paraffin must not solidify in this step) to facilitate complete butanol removal. Butanol should not be smelled from the samples at the end of this step. 12. Replace melted paraffin for a pure one and let stand for half a day (paraffin can be recycled for this step) to eliminate rest of butanol from samples. 13. Replace melted paraffin for a pure one and proceed to embedding. 14. Pour the last paraffin change with objects into the paper origami dish (Fig. 2a) or suitable mold (see Note 13). Top up with pure melted paraffin (do not exceed 65  C) and arrange the samples on the bottom so that it is easy to separate them later (Fig. 2b). 15. Individual samples or their groups (e.g., sets of root segments) might be organized into rows and columns. Arrangement can be achieved with hot needle or tweezers, which are also convenient to remove potential air bubbles from hot paraffin surface. Arrangement can be done on hot plate to extend period of time for manipulation, but fast work in ceramic dishes is usually sufficient. Suitable orientation of objects for sectioning during embedding makes later steps easier. Cardboard tag should be placed into the block together with samples. 16. Cool paraffin blocks. Rapid cooling rate is recommended (e.g., cool water bath, frozen iron plate). Paraffin blocks can be stored for long time (many years at room temperature). 17. Cut individual samples from the paraffin block and fix them to wooden, metal, or plastic chucks suitable for your microtome clamp. Heat the block and chuck at the site of contact to melt surface layer of paraffin and press them together. 18. Trim paraffin blocks into desirable size and shape them for easy sectioning. Proper trimming allows easier achievement of ribbons during sectioning (Fig. 2b). 19. Fix the chuck into the microtome clamp. Use microtome clamp to adjust object into optimal position according to blade edge and cutting plane. Reassure that the clamping mechanism is tightened securely before trimming and sectioning. 20. Adjust proper microtome knife angle (see Note 14; Fig. 2c). 21. To clean the knife/blade, use petrol; this is not as dry and noxious as xylene or toluene. Do not touch the fine edge of the knife/blade as it can be very easily damaged.

14

Alesˇ Soukup and Edita Tylova´

Fig. 2 Cardboard origami for paraffin embedding. (a) Bend along the dashed lines to get a boat suitable for paraffin embedding. Grey indicates upper sided of the cardboard. (b) Arrange the objects into paraffin block so that they might be easily separated after embedding. (c) To adjust correct blade clearance angle, the shape of the blade should be respected. Incorrect angle of the blade can either crush (II) or scrape (III) the block instead of cutting (I). (e) Sectioning of paraffin-embedded object on the microtome and ribbon formation, (d) subsequent flattening of sections on slides

22. Cut sections into a ribbon (Fig. 2d). As it is always more difficult to cut the first section, it is easier to cut ribbon than individual sections. It is also easier to arrange pieces of ribbon on slides. Speed of the cutting stroke should be adjusted according to paraffin and temperature. Slow and steady strokes usually result in best sections with least compression.

Plant Preparations for Light Microscopy

15

Use moistened brush to pick up and manipulate ribbons, as it is easier than forceps and less probable to cause blade damage. For troubleshooting of the most common problems see Table 3. 23. Transfer ribbon on black cardboard and cut it into equal pieces to be placed on the glass slides. Their length should be less then length of available coverslips. If series of sections is required, take care to maintain their proper order. A small nick in the paraffin block, which can be seen in the ribbon, can make proper orientation easier. There are two sides of the ribbon. The glossy one should be placed toward the slide (down), while the matt site is the upper one. 24. Use precoated slides to ensure adhesion of sections for further manipulation (see Note 4). Cover the glass slide with distilled water so that only small part stays without water and can be used to handle the slide. The surface tension of water helps to flatten sections, and enough free space surrounding ribbons should be available. 25. Float the ribbons on the water surface, arrange it, and heat it on hot plate to stretch and flatten the sections (Fig. 2e). The temperature of the plate should be approx. 5  C lower than paraffin melting temperature. Let the slides on the plate for 5–10 min, as stretching the ribbon should be slow and gradual to be efficient. Temperature can be adjusted also experimentally so that it is gradually increased till the paraffin of sections start to melt, then the temperature should lowered for 3–5  C. If the temperature is too high, the ribbons will melt (objects are lost); if too low, flattening does not get complete (lines and wrinkles are still discernible on ribbons). Stretching of sections in water bath is more convenient for large individual sections. If small bubbles form under the ribbon, use preboiled distilled water to eliminate dissolved gasses. 26. Remove the slide from hot plate, let it cool down, and rearrange the ribbons if necessary. 27. Gently remove most of the water and let the slides dry to attach sections to the slides on warm plate (40  C overnight). Protect slides from dust. 28. When dry, sections can proceed to staining or store the slides in box before further processing. 3.6 Staining of Paraffin Sections

Two examples of staining protocols (Safranin O + Fast Green; Hematoxylin) are presented with staining sequence and timing of individual steps (Table 4). There are several common steps: 1. To remove paraffin from sections, place slides into slide holder and immerse them into toluene for 3 min. Repeat this step two

16

Alesˇ Soukup and Edita Tylova´

Table 3 Troubleshooting for the most common problems encountered in paraffin sectioning Problem

Cause

Remedy

Separate sections curl up, cracks parallel to blade edge may appear

The block is too cold Sections are too thick for used temperature Wrong clearance angle of the knife resulting in irregular section thickness

Straighten the first section using a soft brush; subsequent sections within the ribbon usually do not roll Warm up the block by breathing on it, touching it with your finger or placing incandescent bulb into its vicinity Modify section thickness. Change the angle of the blade

Individual sections do not Incorrectly prepared block (opposite Realign block edges and position according to the knife ribbon sites are not parallel, side of the Warm up the block and knife block is not parallel to the blade Cut faster edge) Cold block or knife Cutting is too slow (sections are glued together with heat as blade hits the block) Individual sections are strongly compressed, folded, and may stick on the knife

Temperature is too high Dull blade Blade is dirty with paraffin Too thin sections for the type of paraffin Too low clearing angle

Cool down the block Resharpen knife or change blade Clean the knife Increase section thickness Increase knife angle Decrease speed of sectioning

Objects are separating from the section

Improperly embedded object (improper dehydration, incomplete infiltration, incomplete elimination of ethanol or intermedium) Object is too hard for used paraffin

Re-embed the object (if possible) Use harder paraffin (higher melting temperature) Cool down the block to make it harder Soften the object

Sections catch on the block Improper knife angle when travelling back Dirty or dull knife

Modify knife angle Carefully clean the knife from both sides

Ribbon is not straight but Sides of the block are not parallel turns (mutually or to the knife edge) Object is heterogeneously hard

Realign and trim the block

Longitudinal lines on the ribbon

Use other part of the edge, change blade, resharpen knife Clean the edge Decrease the knife angle Soften the object Re-embed into clean/harder paraffin

Nicks on the blade edge. Dirty edge Hard particles in object (sclerenchyma) Dust in paraffin

Plant Preparations for Light Microscopy

17

Table 4 Staining paraffin sections Safranin O + Fast Green FCF Safranin O stains in red lignified, suberinized, or cutinized cell walls, and nucleoli and chromosomes after FAA fixation. Cellulosic cell walls and cytoplasm stain green Paraffin removal

Toluene I Toluene II Toluene III

3 min 3 min 3 min

Rehydration

100% ethanol 96% ethanol 70% ethanol

3 min 3 min 3 min

Dyeing

Safranin O 70% ethanol 70% ethanol 96% ethanol Picric acid 96% ethanol + NH3 (3–5 drops/100 ml) 100% ethanol

2 h or more (test for particular materials) Rinse briefly 15 s 15 s Rinse briefly (might be omitted if staining is weak) 2 min

Counterstain and dehydration

Fast Green FCF 100% ethanol 100% ethanol

2 min Rinse to wash out excessive dye Rinse briefly

Mounting

Toluene IV Toluene V Toluene VI Resin

2 min 2 min 2 min Apply resin and mount the coverslip

Rinse briefly

Hematoxylin Hematoxylin staining provides overall, almost monochromatic contrast of structures Paraffin removal

Toluene I Toluene II Toluene III

3 min 3 min 3 min

Rehydration

100% ethanol 96% ethanol 70% ethanol 50% ethanol

3 min 3 min 3 min 3 min

Dyeing

Hematoxylin Wash in water

5–30 min (test for particular materials) 10 min

Dehydration

50% ethanol 70% ethanol 96% ethanol 100% ethanol 100% ethanol

2 min 2 min 2 min 2 min 2 min

Mounting

Toluene IV Toluene V Toluene VI Resin

2 min 2 min 2 min Apply resin and mount the coverslip

18

Alesˇ Soukup and Edita Tylova´

times with fresh toluene bath to completely eliminate paraffin, which might otherwise interfere with staining. 2. Wash out toluene with absolute ethanol and gradually rehydrate sections through reverse ethanol dehydration series to achieve hydration adequate for selected staining method. For water-soluble dyes go through the complete ethanol series (100%, 96%, 70%, 50%, 30% ethanol) down to distilled water, for ethanol-soluble dyes finish with ethanol concentration equal to selected staining solution. In most cases 3 min for each step is enough for standard sections. 3. Dye sections passing them through the series indicated in Table 4. 4. Gradually dehydrate sections through independent ethanol series (up to 100% ethanol; 2 min in each step). 5. Rinse three times in toluene for 2 min. 6. Cover sections with resin mounting medium (e.g., Pertex or other toluene or xylene soluble resin) and apply coverslip. Do not allow the sections to dry as bubbles will be trapped below the coverslip. Let the mounting medium to harden (2 days at room temperature). 3.7

Cryosectioning

1. Fix the specimen in an appropriate fixative (e.g., 4% formaldehyde in phosphate buffer). Use lowered pressure (“vacuum infiltration”) to substitute fixative for the air within the tissue if necessary. 2. Wash out fixative for 15 min with phosphate buffer used to prepare fixative. 3. Infiltrate samples gradually with 3%, 10%, and 20% sucrose solutions. Agitate gently and apply 0.1% of surfactant (Triton or Tween 20) with 3% sucrose solution to facilitate infiltration. Each step takes at least 30 min at room temperature. Individual steps should be prolonged to infiltrate properly compact and more voluminous samples (see Note 15). 4. Prepare all needed equipment (brushes, forceps, etc.) into cryotome chamber to freeze it down to the right working temperature before use. 5. Let the cryostat cool down to working temperature and turn on cryobar (freezing shelf) boost to minimize the bar temperature. Add small amount of semisolid cryoembedding medium (e.g., OCT or high-viscosity cryoembedding medium) on the specimen chuck and use heat extractor to make flat base (the extractor frequently stick to the medium if not frozen enough; apply Teflon coating spray to the extractor to minimize this problem). Add more medium on the top of frozen platform and transfer the sample into this medium. Quickly arrange the

Plant Preparations for Light Microscopy

19

sample into the desired position and freeze the block on cryobar. The sample should be covered with a thin layer of cryoembedding medium (see Note 15). The frozen samples can be stored at 80  C in a closed container if necessary. Do not store them in the cryotome chamber as the samples dry out rather quickly. 6. Trim frozen medium encasing object to adjust the specimen block size for easier cutting. Leave enough medium in surrounding of the sample. However, it is more difficult to cut thin sections from larger block. 7. Mount the chuck with object on the microtome head and let its temperature to equilibrate. Working temperature should be selected according to desired thickness, character of the sample, and composition of embedding medium beside others. Independent temperature setup of chamber (knife) and sample is of advantage; it might be convenient to use 2–3  C lower temperature of knife than sample (e.g., ref. 45). Commonly we use specimen temperature between 8 and 20  C for standard section of 8–20 μm. Thinner sections might require lower temperature. It is reasonable to start with 15  C and adjust the temperature according to appearance of sections. If the sections wrinkle and smear on knife, the working temperature is too high. If sections crumble, temperature is too low. For troubleshooting of the most common problems see Table 5. 8. Fix the chuck into holder and adjust its axial position with microtome head according to the blade. Approach slowly with objects toward the blade. Do not trim too quickly as damage may be done to blade edge and object can break off. 9. Set the position of antiroll plate parallel with blade edge. The edge of the antiroll plate should be close enough to the edge to allow emerging sections to slide underneath (up to 0.5 mm) but not too close to crush the block. It is not easy to adjust position visually, so while very gently touching the top of the plate, one should feel the cutting edge. Be careful not to hurt your fingers! Cut few sections (it is reasonable to start with section of 15 μm) and further correct the plate position. If the section rolls in front of the plate, move it slightly up. If the plate is too high, the plate crushes the object (you can feel it with finger lightly touching the plate holder during sectioning), lower the plate. 10. Cut several sections to make smooth block surface. 11. Cut the sections; they run individually or in a row of few under the plate (Fig. 3a). 12. Pick up sections with brush chilled in cryotome chamber or collect them with subbed (adhesive coated) and marked slide of room temperature. The slide should be approached very closely

20

Alesˇ Soukup and Edita Tylova´

Table 5 Troubleshooting for cryotome sectioning Problem

Cause

Remedy

Sections smear or crumple on the blade edge

Temperature of specimen is Select a lower temperature too high Wait to equilibrate the object temperature The space under antiroll and try sectioning again plate is too low

Sections shatter at the tip of Specimen is too cold the blade The antiroll plate is not correctly adjusted The blade might be dull or its clearance angle is too steep The specimen surface is large

Select a higher temperature and let the object equilibrate; if from the cryobar, let it to adopt temperature of the clamp Adjust the antiroll plate correctly Use another area of the blade or a new blade Trim the specimen parallel and increase the section thickness Knife clearance angle of 2–5 is recommended for disposable blades

Sections curl up when the antiroll plate is raised up

The antiroll plate is too warm

Lay down the antiroll plate on knife and let its temperature stabilize Minimize air exchange within the chamber

Sections do not run flat under the antiroll plate

Dirty antiroll plate and/or knife Dull blade

Clean with dry cloth or brush; it is convenient to have some frozen in the chamber If necessary, use ethanol to clean blade and antiroll plate Change blade or use another area of the edge

Sections curl in front of the The antiroll plate is too far below the edge of blade antiroll plate and do not go underneath

Readjust the antiroll plate

Section smear on the top of Antiroll plate goes beyond the antiroll plate the blade edge and crushes the object

Readjust the antiroll plate

Chatter on sections

The chuck or blade is not secured correctly Specimen is too hard, too cold or too big The clearance angle is incorrect Cutting speed is too high

Check and fix the stabilization of block holder and blade Modify temperature; let the object equilibrate with specimen head Trim the object to decrease its size Reset clearance angle of the blade Decrease the speed of cutting

Variable thickness of sections

The chuck or blade is not secured correctly. The clearance angle is incorrect

Recheck and fasten the microtome head and blade holder Reset the blade angle

Sections are torn perpendicularly to the blade edge

Dust or nick on the blade Clean front and back side of the blade Leading edge of the antiroll Replace the blade or move to another part of plate is dirty or damaged the blade Clean or replace antiroll plate

Plant Preparations for Light Microscopy

21

Fig. 3 Detail of cryotome head with chuck and object. In the lower part of the picture is the antiroll plate with section underneath (a). To pick up the section, rise up the antiroll plate and place the slide very close to the section (resting the corner of slide on the blade holder helps). The section “jumps” on the slide and melts on its surface (b–d). Do not press the slide on section, as it would freeze onto the blade holder. (e) Slide with two collected sections

to sections but not directly touching them. Sections will “jump” and melt onto the slide within second (Fig. 3b–e). If you press the slide on section, the section melts on the blade holder and it is not easy to clean it off. 13. Brush away the condensed ice from blade holder before sectioning.

4

Notes 1. There are two commonly used options of FAA based on the final ethanol concentration: 70% and more delicate 50% (v/v). Content of acetic acid can be also modified between 2% and 6%. Materials can be stored in solution for a considerable period of time. Use formalin (commercial ~40% formaldehyde solution) for preparation. 2. Buffers used with aldehyde fixatives must not react with them (e.g., TRIS, EDTA amino groups will react with aldehydes). Phosphate, HEPES, PIPES-based buffers, or other Good’s buffers are recommended. Phosphate might precipitate some divalent cations (Mg2+, Ca2+). Osmolarity of the buffer should be selected according to particular objects. For most plant samples, we use 25–100 mM buffers. Be aware that 4% formaldehyde itself is a 1.33 M solution.

22

Alesˇ Soukup and Edita Tylova´

3. Melting temperature is closely related to hardiness of the paraffin. It normally stays between 56 and 58  C, but 54 and 60  C mixtures are available too. Composition of the embedding paraffin blends differs mainly in content and composition of plasticizers (plastic polymers) improving sectioning properties and “hardness.” We prefer the use of paraffin with minimum or no additives and stabilizers as we have experienced easier infiltration and no separation/precipitation of plasticizers from paraffin at higher temperatures. Some high polymer-content mixtures seem to be rather sensitive to higher temperatures, and it is better not to exceed 65  C. Paraffin without additives can be cut down to approximately 5 μm. It is claimed that with additives sections down to 2–3 μm are accessible. This might be valid only for soft plant tissues, and we prefer resin embedding for semithin sections. We do not recycle paraffin with additives. 4. Adhesives are compounds used to glue sections on the glass slides. Unlike animal tissues, plant tissues have relatively lower protein content, and presence of cell wall and vacuole makes them less adhesive. That is why they float away from the slides easily during staining or other processing. Selection of the right adhesive depends on intended use. Glycerol albumen is the easiest to use, alum gelatin is standard subbing that holds well the sections, and poly-L-lysine is good solution for immuno and other more sensitive applications. To prepare adhesive (subbing)-coated slides, cleaning and degreasing of slides is of high importance. Even new slides should be washed with detergent followed with 96% ethanol and distilled water. Alternatively the slides can be washed in dishwasher and carefully rinsed with distilled water before use. 5. Cryoembedding media are commercially available or can be prepared in the laboratory. We have positive experience with both options. OTC (optimum cutting temperature compound) is a commercially available cryoembedding medium (e.g., Tissue-Tek OCT) based on polyvinyl alcohol (PVA) and polyethylene glycol (PEG). Cryo-Gel™, Cryomatrix™, and PolyFreeze™ are further commercial options differing in viscosity. 6. It might be accepted as a common rule of thumb that volume of fixative should not be less than 50 volume of the fixed tissue. Otherwise, the buffering capacity of the solution (pH, concentration of fixative, molarity) might not be sufficient. 7. There are very few experimental data to estimate the time needed for aldehyde penetration and fixation. In animal tissues (e.g., liver), the penetration rate normally does not exceed 1 mm per h. In the work of Mersey and McCully [8], the acrolein fixation passed about 140 μm per min along the root

Plant Preparations for Light Microscopy

23

hair. The formation of linkages (incorporation of formaldehyde) within the tissue might be also a rather slow process, taking hours to be saturated (e.g., ref. 5). 8. Rotary vane vacuum pump with pressure regulator and plastic desiccator allow for controlled gradual drop of pressure within the desiccator chamber. Evaporating fixative (or any other solution) accumulates within the oil and can cause corrosion (damage) of the pump chamber. It is necessary to let the pump run long enough to warm up the oil and evaporate the condensate from pump. During vacuum infiltration, make sure to vent pump exhaust into the fume hood and not into laboratory. Fixatives are toxic! 9. Tissues of high density or pigmentation might require removing the cytoplasm content with 2% NaOH in 30% EtOH or dimethyl sulfoxide. The latter is more efficient and can remove complete protoplasts. Extraction of chlorophyll and lipidic compounds might be done in a methanol–chloroform (1:1) mixture. To clear colored phenolic depositions, alkalized hydrogen peroxide or sodium hypochlorite-based protocols mentioned in introduction should be used. 10. Sharpening with fine-grained stone (e.g., Japanese whetstone, grit 8000) is part of the regular maintenance but does not replace stropping. The finest cutting edge is affected during sectioning, and frequently it is enough to straighten and polish the blade by stropping (Fig. 1d) to recover its sharpness. While sharpening requires edge-forward movement of the blade on the stone, it should be carefully drawn spine first to avoid cutting of the strop during stropping. The hard steel of blade is highly sensitive to corrosion. It should be therefore kept away from acids and stored clean and dry. 11. Timing of individual dehydration steps depends on the size of the objects and their texture [3]. For easy objects (e.g., roots samples with diameter up to 2 mm, leaves 5  5 mm segments, etc.), 3 h in the step should be safe. Small objects up to 1 mm might require only 15–30 min, while larger objects (minimal dimension 10–15 mm) might require days per step. Objects of high tissue density or covered with low permeability cuticle have significantly impeded exchange of solutions. Timing of changes should be prolonged accordingly. 12. Dehydration series solutions lose their properties with use, and contamination with compounds extracted from samples takes place. It is recommended to keep record on the number and type of processed samples and replace solutions regularly. Precaution is more important for anhydrous (later) steps of the series.

24

Alesˇ Soukup and Edita Tylova´

13. The embedding can be done using commercially available molds, cassettes, or origami made of smoothen cardboard which does not soak up much of the paraffin (Fig. 2a). Ceramic dishes with flat bottom are also good option. Spray/smear the ceramic dishes with 50% glycerol or commercial detergent before embedding to facilitate later separation of paraffin block from the ceramic surface. 14. Because the knife (as well as disposable blades) is not a simple wedge but has facets on the edge (Fig. 2c), the setup of knife clearance angle should respect the shape. Too low angle will crush sample; too steep adjustment might cause rolling of sections and chatter over hard objects. Clearance angle is usually between 3 and 5 or more acute approx. 10 for thinner sections and harder objects. Scale for adjustment is normally marked on the blade holder. 15. Transfer samples into a mixture of 20% sucrose solution and cryoembedding medium (1:1) and let them infiltrate overnight within refrigerator. This step is optional, but introduction of cryoembedding medium (e.g., OCT) into the object might further improve the quality of sections [46]. Alternatively, place samples into a small aluminum foil mold filled with cryoembedding medium. To prepare the mold, fold a rectangular sheet of thicker Al foil around a small coverslip box. It is easier to handle small specimens and arrange them into desirable position using forceps or needle. Freeze the mold on cryobar inside the cryostat chamber or immerse the mold into isopentane supercooled with liquid nitrogen. Peel away the Al foil and fix frozen block on the chuck with a drop of cryoembedding medium.

Acknowledgments The current update of this chapter, originally published as ref. 47, has been supported by the NPUI-LO1417 project. References 1. Pearse AG (1980) Histochemistry (theoretical and applied): preparative and optical technology. Churchill Livingstone, Edinburgh 2. Pearse AG (1985) Histochemistry (theoretical and applied): analytical technology. Churchill Livingstone, Edinburgh 3. O’Brien TP, Mccully ME (1981) The study of plant structure: principles and selected methods. Termarcarphi Pty LTD, Melbourne 4. Ruzin SE (1999) Plant microtechnique and microscopy. Oxford University Press, Oxford

5. Fox CH, Johnson FB, Whiting J, Roller PP (1985) Formaldehyde fixation. J Histochem Cytochem 33:845–853 6. Medawar PB (1941) The rate of penetration of fixatives. J Royal Micro Soc 61:46–57 7. Bancroft JD, Gamble M (2008) Theory and practice of histological techniques. Churchill Livingstone, London 8. Mersey B, Mccully ME (1978) Monitoring of the course of fixation of plant cells. J Microsc 114:49–76

Plant Preparations for Light Microscopy 9. Coetzee J, van der Merwe CF (1985) Penetration rate of glutaraldehyde in various buffers into plant tissue and gelatin gels. J Microsc 137:129–136 10. Gardner RO (1975) An overview of botanical clearing technique. Biotech Histochem 50:99–105 11. Bybd DW Jr, Kirkpatrick T, Barker KR (1983) An improved technique for clearing and staining plant tissues for detection of nematodes. J Nematol 15:142–143 12. Stebbins GL Jr (1938) A bleaching and clearing method for plant tissues. Science 87:21–22 13. Malamy JE, Benfey PN (1997) Organization and cell differentiation in lateral roots of Arabidopsis thaliana. Development 124:33–44 14. Shobe WR, Lersten NR (1967) A technique for clearing and staining gymnosperm leaves. Bot Gaz 128:150–152 15. Sporne KR (1948) A note on a rapid clearing technique of wide application. New Phytol 47:290–291 16. Simpson JLS (1929) A short method of clearing plant tissues for anatomical studies. Biotech Histochem 4:131–132 17. Lux A, Morita S, Abe J, Ito K (2005) An improved method for clearing and staining free-hand sections and whole-mount samples. Ann Bot 96:989–996 18. Peterson CA, Fletcher RA (1973) Lactic acid clearing and fluorescent staining for demonstration of sieve tubes. Biotech Histochem 48:23–27 19. Lersten NR (1986) Modified clearing method to show sieve tubes in minor veins of leaves. Biotech Histochem 61:231–234 20. Herr JM Jr (1971) A new clearing-squash technique for the study of ovule development in angiosperms. Am J Bot 58:785–790 21. Beeckman T, Engler G (1994) An easy technique for the clearing of histochemically stained plant tissue. Plant Mol Biol Rep 12:37–42 22. Bougourd S, Marrison J, Haseloff J (2000) An aniline blue staining procedure for confocal microscopy and 3D imaging of normal and perturbed cellular phenotypes in mature Arabidopsis embryos. Plant J 24:543–550 23. Cunningham JL (1972) A miracle mounting fluid for permanent whole-mounts of microfungi. Mycologia 64:906–911 24. Truernit E, Bauby H, Dubreucq B, Grandjean O, Runions J et al (2008) Highresolution whole-mount imaging of threedimensional tissue organization and gene expression enables the study of phloem development and structure in Arabidopsis. Plant Cell 20:1494–1503

25

25. Dubrovsky JG, Soukup A, NapsucialyMendivil S, Jeknic Z, Ivanchenko MG (2009) The lateral root initiation index: an integrative measure of primordium formation. Ann Bot 103:807–817 26. Kurihara D, Mizuta Y, Sato Y, Higashiyama T (2015) ClearSee: a rapid optical clearing reagent for whole-plant fluorescence imaging. Development 142(23):4168–4179 27. Ursache R, Andersen TG, Marhavy´ P, Geldner N (2018) A protocol for combining fluorescent proteins with histological stains for diverse cell wall components. Plant J 93(2):399–412 28. Zelko I, Lux A, Sterckeman T, Martinka M, Kolla´rova´ K et al (2012) An easy method for cutting and fluorescent staining of thin roots. Ann Bot 110:475–478 29. de Almeida Engler J, Van Montagu M, Engler G (1994) Hybridization in situ of wholemount messenger RNA in plants. Plant Mol Biol Rep 12:321–331 30. Klebs E (1869) Die Einschmelzungs Methode, ein Beitrag zur mikroskopischen Technik. Arch Mikrosk Anat Entwicklungsmech 5:164–166 31. Johansen DA (1940) Plant microtechnique. McGraw-Hill Book Co. Inc, New York 32. Sass JE (1940) Elements of botanical microtechnique. McGraw-Hill Book Co Inc., New York, London 33. Vitha S, Baluska F, Jasik J, Volkmann D, Barlow PW (2000) Steedman’s wax for F-actin visualization. Dev Plant Soil Sci 89:619–636 34. Sartori N, Richter K, Dubochet J (1993) Vitrification depth can be increased more than 10-fold by high-pressure freezing. J Microsc 172:55–61 35. Quintana C (1994) Cryofixation, cryosubstitution, cryoembedding for ultrastructural, immunocytochemical and microanalytical studies. Micron 25:63–99 36. Benesˇ K (1973) On the media improving freeze-sectioning of plant material. Biol Plant 15:50–56 37. Tirichine L, Andrey P, Biot E, Maurin Y, Gaudin V (2009) 3D fluorescent in situ hybridization using Arabidopsis leaf cryosections and isolated nuclei. Plant Methods 5:11–18 38. Knapp E, Flores R, Scheiblin D, Scheiblin D, Modla S et al (2012) A cryohistological protocol for preparation of large plant tissue sections for screening intracellular fluorescent protein expression. Biotechniques 52:31–37 39. Zhang Z, Niu L, Chen X, Xu X, Ru Z (2012) Improvement of plant cryosection. Front Biol 7:374–377 40. Knox RB (1970) Freeze-sectioning of plant tissues. Biotech Histochem 45:265–272

26

Alesˇ Soukup and Edita Tylova´

41. Cocco C, Melis GV, Ferri GL (2003) Embedding media for cryomicrotomy: an applicativereappraisal. Appl Immunohistochem Mol Morphol 11:274–280 42. Williams DBG, Lawton M (2010) Drying of organic solvents: quantitative evaluation of the efficiency of several desiccants. J Org Chem 75:8351–8354 43. Pappas PW (1971) The use of a chrome alumgelatin (subbing) solution as a general adhesive for paraffin sections. Biotech Histochem 46:121–124 44. Brundrett MC, Enstone DE, Peterson CA (1988) A berberine–aniline blue fluorescent

staining procedure for suberin, lignin, and callose in plant tissue. Protoplasma 146:133–142 45. Ferri GL, Cocco C, Melis GV, Aste L (2002) Equipment testing and tuning: the cold-knife cryomicrotome microm HM-560. Appl Immunohistochem Mol Morphol 10:381–386 46. Barthel LK, Raymond PA (1990) Improved method for obtaining 3-microns cryosections for immunocytochemistry. J Histochem Cytochem 38:1383–1388 47. Soukup A, Tylova´ E (2014) Essential methods of plant sample preparation for light microscopy. Methods Mol Biol 1080:1–23

Chapter 2 Selected Simple Methods of Plant Cell Wall Histochemistry and Staining for Light Microscopy Alesˇ Soukup Abstract Histochemical methods allow for identification and localization of various components within the tissue. Such information on the spatial heterogeneity is not available with biochemical methods. However, there is limitation of the specificity of such detection in context of complex tissue, which is important to consider, and interpretations of the results should regard suitable control treatments if possible. Such methods are valuable extension to specific optical and spectroscopic analytical methods. Here we present a set of selected simple methods of staining and histochemical tests with comments based on our laboratory experience. Key words Cell wall, Histochemistry, Lignin, Suberin, Pectin, Cellulose, Callose, Antibody, Staining

1

Introduction The rigid plant cell wall is a prominent structure tightly related to cell shape, function, and interactions in the context of a multicellular body and in communication with surrounding environment. In fact, plant cell walls are structures most frequently followed studying tissue and organ anatomical organization. A combination of simple methods of cell wall staining and histochemistry might provide substantial and easily accessible information on cell wall composition, modifications, and changes related to development and tissue differentiation. However, unlike the biochemical detection, it does not allow for specific separation of cross-linked, complex mixture of components, which significantly increase probability of nonspecific results and interactions during detection. Therefore, higher probability of incorrect interpretation should be compensated with use of proper controls and independent parallel reactions if possible. Histochemical detection should not be confused with procedures of “anatomical staining” because the affinity of pigment to target structure (e.g., safranin staining of lignified cell walls) depends highly on particular conditions (pH, polarity of

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_2, © Springer Science+Business Media, LLC, part of Springer Nature 2019

27

28

Alesˇ Soukup

solvent, temperature, time of dyeing, etc.) and is far less specific than colored product gained during specific histochemical reaction with substrate. 1.1 General Cell Wall Staining Methods

Toluidine blue O polychromatic staining is a simple and very useful oversight staining procedure disseminated in plant microtechnique by O’Brien et al. [1]. Besides good overall contrast of most structures, it also renders information on properties of the stained material. That is because the pigment interacts with stained material and shifts its absorption spectra toward longer wavelengths according to density of surface polyanions and subsequent dye aggregation [2]. Such coloration is referred to as metachromatic [3]. Therefore, cell walls with low pectin content will stain blue (orthochromatic color), while pectin rich material will stain purple to pink (metachromatic color). Because lignified/phenolics containing cell walls present lower concentration of acidic groups, their staining is usually greenish. Cell wall material can be stained metachromatically above pH 3. Besides, cell wall tannin-containing vacuoles might stain green to bright blue (usually on dark background), DNA-containing nucleus green. Even such “anatomical staining” can gain very useful information, but interpretation should be cautious.

1.2 Staining of Pectins, Callose, and Hemicelluloses

PAS (periodic acid—Schiff’s reagents) reaction is an example of nonselective polysaccharide detection procedure [4]. Periodic acid is a strong oxidizing agent cleaving vicinal diol linkages of polysaccharides producing thus dialdehydes. These are subsequently detected with Schiff’s reagent or its fluorescence alternatives [5, 6]. However, there is often background signal of some aldehydes present in the tissue (e.g., lignin), and some others might be introduced during treatment with aldehyde fixative, which should be therefore used thoughtfully. That is why control sections without previous periodic acid treatment should be always included to ascertain about the origin of aldehydes. Optionally the autochthonous aldehydes might be eliminated before periodic acid treatment with borohydride reduction [4]. The PAS reaction scheme has been used recently also for staining of whole mount objects in combination of fluorescent leukobases of propidium iodide [7]. Calcofluor staining might be considered another general procedure of fluorescent cell wall accentuation. Calcofluor (synonyms are Tinopal, Fluorescent brightener) is nonspecific UV-excited fluorochrome with high affinity to plant and fungi cell walls [8]. Its selectivity is considered to be related to (1 ! 3) and (1 ! 4), -β-D- glucan chains of polysaccharides similarly to Congo red [9]. Besides procedures of general cell wall detection, there are methods aimed for specific components of cell wall.

Selected Methods of Cell Wall Histochemistry

29

Alcian blue is a basic dye, which can be used to rather specifically stain dissociated acidic carboxyl groups of pectins [10]. Acidic environment used for staining further narrows spectrum of potentially dissociated (stainable) acidic groups. In fact, there are not many other compounds that might react with the dye in plant cell walls under such conditions. Staining mechanism of Alcian blue is rather similar to more generally used ruthenium red [11]. Ruthenium red is a hexavalent cation, which binds to variety of polyanions; that is why this classical reaction with pectin should be considered typical rather than specific. Ruthenium red also has its traditional use in electron microscopy (e.g., ref. 12). Specificity of the staining can be further verified after pointed carboxyl blockage via methylation [4]. Such blockage of acidic carboxyl should also abolish most of toluidine blue metachromasy discussed above. Pectins can be often (depending on the linkages within the cell wall context) extracted with hot aqueous solutions, Ca2+ chelating agents, and weak alkali solutions. Therefore, such treatment should be avoided prior to pectin staining with any of the methods described. On the other hand, extracting agents and their sequences might be used in connection with detection methods to further specify or confirm composition of extracellular material according to specific extractability (e.g., refs. 13, 14). Callose is highly dynamic polymer (e.g., ref. 15). Its presence in tissues might be easily induced, for example, with chemical fixation of samples. Callose deposition is one of rather fast responses to stress or plant cell injury. Aldehyde fixation, which is in fact it is a kind of chemical injury, also induces rapid deposition of callose into the plasmodesmata containing pit fields in order of minutes. That is why usage of cold methanol fixation or callose synthase inhibitors proved to be convenient to approach in vivo presence of callose. There are two most common ways of callose detection in tissues. The most frequently used is staining with aniline blue [16, 17], respective its common impurity—the UV excited fluorochrome Sirofluor. Because the content of Sirofluor in the raw dye is variable according to brand and batch, it is reasonable to test your dye stock with known material first (positive control) or use purified (and far more expensive) fluorochrome, which forms highly fluorescent complexes with (1–3) β-D-glucans [18]. Advantage of purified fluorochrome might be seen also in the extended staining pH range from 3 to 10, while aniline blue staining should be efficient at higher pH [18]. There are several reports on compromised specificity of the reaction and possible interaction with other polymers [19]. Unstained control sections are crucial to ensure about nature of the fluorescence emission. Besides the fluorescent staining, bright field visualization of callose with reasonable specificity might be gained with Resorcin blue [17]. Since the antibodies are commercially available (e.g., Biosupplies, Australia), callose

30

Alesˇ Soukup

immunolocalization provides easily accessible highly consistent, sensitive, and specific detection alternative. As far as we are aware there is no reliable and specific histochemical test for identification of hemicelluloses. This gap can be efficiently filled with the use of specific antibodies, which allow for precise distinction of various cell wall components (including pectins, hemicelluloses, and proteins). There are several sources of the antibodies currently available (e.g., http://www.plantprobes.net; http://cell.ccrc.uga.edu/~mao/wallmab/Antibodies/antib.htm). General information of available antibodies can be searched from Plant Cell Wall Monoclonal Antibody Database (http://glycomics. ccrc.uga.edu/wall2/antibodies/antibodyHome.html). Some of them were used in our lab in protocol similar to protocol presented bellow for callose. 1.3 Staining of Cell Wall Lipids and Lignin

Presence of lipidic compounds in cell wall is frequently connected with formation of cell layers with modified permeability of apoplast. There are two principal cell wall lipids insoluble in polar solvents historically distinguished by their position. Suberin is located in internal and secondary dermal tissues, while cutin constitutes cuticular part of epidermis on the surface of plant organs [20, 21]. Considerable variations in monomeric composition of suberin and proportion of aromatic and lipidic monomers were reported among species and during development [22, 23]. There are several staining procedures used for detection of lipidic compounds in cell walls. Lipidic Sudan dyes (Sudan III, Sudan IV, Sudan Black B) are traditionally utilized in alcoholic solutions. Polyethylene glycol/ glycerol-based staining solution of Sudan red 7B introduced by Brundrett [24] proved to be far more efficient and is the method of choice. Lipidic dyes partition from the slightly polar dyeing solution into the lipidic compartments of the tissue. It should be emphasized that intensity of staining depends highly on lipidic nature of cell wall material (quantity as well as molecular context of derivatives of fatty acid within the cell wall). Therefore, sensitivity of the detection should be considered during interpretation, and nonspecific precipitation should be avoided. Improvement of sensitivity was reported due to use of lipidic fluorochrome Fluorol yellow [24]. However, background staining and autofluorescence can be sometimes difficult to distinguish from specific Fluorol yellow signal. That is why for some objects (e.g., maize roots) Sudan red 7B is preferred in our hands. Commonly, we use fresh sections after aldehyde or no fixation. Several pitfalls are known (see method description below). Modified method of fluorol yellow staining combined with lactic acid clearing of object was published by Lux [25]. Finally, a very old technique of concentrated sulfuric acid digestion of cell wall material might also provide valuable information as only suberin and cutin impregnated material should resist it [26, 27].

Selected Methods of Cell Wall Histochemistry

31

Berberine–toluidine blue staining procedure was introduced by Brundrett et al. [28] to detect material of Casparian bands, suberin lamellae, and lignified tissue. It is a very frequently used staining based on acidophilic nature of berberine, which stains aromatic domains of lignified and suberized cell walls. The staining of berberine is combined with counter stain of toluidine blue (alternatively aniline blue, Evans blue, or Crystal violet) to quench background fluorescence. Selectivity of such a quenching is most likely related to physical properties of the cell wall (decreased accessibility of the material, for example, due to suberinization). In fact, counterstaining itself provides valuable information and combination with other acidophilic fluorochromes (e.g., acridine orange, basic fuchsin) or observation of autofluorescence of suberized cell walls is possible. It is still far from clear to which extend the aromatic domain of suberin is identical or similar to lignin and how to distinguish those. Above mentioned staining can be considered an indication but not as a proof of lignifications. There are several others historically established methods of detection of lignification. The most frequently used is Wiesner’s reaction [29] using phloroglucinol condensation with cinnamic aldehydes (coniferyl aldehyde) in acidic environment and formation of cherry red product [30, 31]. There is potential cross-reactivity with other aliphatic and aromatic aldehydes [32], but in standard conditions, the specificity can be considered rather high. Alternatively, aniline sulfate [29, 33] is proposed. The output of the reaction and localization seems to be very similar to phloroglucinol reaction, but with lower contrast of resulting yellow coloration. Another often-used traditional lignin test is M€aule’s reaction [34]. Syringyl moieties of lignin are considered the reaction target [30, 31]. The lignin composition related difference in detection, comparing to phloroglucinol can be strongly pronounced during development [31, 35] as well as in between taxonomic groups [36]. Schiff’s reagent staining might be also used for detection of aldehydes of lignin [26]. There is wide spectrum of acidophilic dyes that have some affinity to lignified cell walls (PI, DAPI, Hoechst, basic Fuchsine, Auramine O, etc.). However, because of dependence on staining conditions and low specificity of such staining, it should be considered as informative only and further confirmation of lignification is recommended. Autofluorescence of aromatic compounds is another very useful approach to follow phenolic compounds within the cell walls [37, 38]. It should be mentioned that combination of detection of lipid compounds and lignification with visualization of fluorescent proteins is restricted to usage of rather mild aqueous solutions of dyes in order to protect the protein fluorescence [39]. Such staining is highly informative, but with lower specificity than classical histochemical tests.

32

Alesˇ Soukup

1.4 Detection of Enzyme Activities

2

There is a wide range of enzymes in apoplastic space. Apoplastic plant peroxidases play a key role in various metabolic processes (e.g., lignin and suberin formation, cross-linking of cell wall components, auxin metabolism, and metabolism of reactive oxygen species) [40]. Peroxidase enzymatic activity might be probed with various cosubstrates in presence of H2O2. The most common is diaminobenzidine (DAB), which yields upon oxidation brownish polymer [41, 42] and tetramethylbenzidine (TMB)—chromogen which yields a blue reaction product upon oxidation [43, 44]. Substrate does not have specific selectivity for particular heme protein; therefore, distinction of catalase and peroxidase is based on their different pH optima. Peroxidase has its optimum at neutral range (pH ~6.5) while it is above pH 10 for catalase (see ref. 33). To optimize the reaction progress, higher temperature (37  C) is recommended which increases the enzyme activity, and adequately reduced exposure time decreases spontaneous precipitation of DAB in presence of H2O2. Precipitation is further decreased if the reaction proceeds in dark as light induces spontaneous decomposition of H2O2. It is very important to include suitable controls (e.g., reaction mixture without H2O2 and sections where peroxidase activity was inhibited; ref. 4).

Materials

2.1 Various Single Component Dyes

All solutions should be prepared in distilled water, unless stated otherwise. 1. Toluidine blue O: 0.01–0.025% (w/v) toluidine blue O in water. Store at 4  C. It can last for rather long time (months). Check periodically for mold. 2. Calcofluor white stock solution: 1% solution of Calcofluor in distilled water. Gently heat the solution and add minimum of 1 M sodium hydroxide (final pH 10–11) to dissolve the dye completely. Aliquots of stock solution can be stored at 20  C for a long period of time. 3. Alcian blue: 0.1% (w/v) Alcian blue in 3% acetic acid (alternatively citrate buffer of pH 3.5, 100 mM can be used, but is less selective). 4. Ruthenium red: 0.05% (w/v) aqueous solution. Do not use phosphate and some other anionic buffers as those might precipitate the dye. 5. Aniline blue fluorochrome: 0.005–0.01% solution of watersoluble aniline blue buffered to pH above 8.5 (e.g., 100 mM K2HPO4 with pH 9). Stock solution of purified aniline blue fluorochrome Sirofluor (1 mg/ml) in distilled water can be stored in aliquots at 20  C.

Selected Methods of Cell Wall Histochemistry

33

6. Resorcin blue: Dissolve 3 g of resorcinol (p.a.) in 200 ml of distilled water. Add 3 ml of concentrated ammonia. Heat up in steam bath for 10 min (do not boil!). Let the red–brown solution cool down to lab temperature. The solution will gradually gain blue color (after approx. 6 h). Heat again in the steam bath for about 30 min until no more ammonia escapes (test with wet pH indicator paper). Dilute prepared solution 1:50 with distilled water for staining. 7. Sudan red 7B and fluorol yellow: Dissolve Sudan Red 7B (0.1–0.2% w/v) or fluorol yellow 088 (0.01% w/v) in PEG-400 heating the solution up to 90  C. Do not exceed 100  C as overheating changes staining properties. Add equal volume of 90% aq. glycerol. Filter solution through coarse filter paper or let stand overnight, and decant supernatant or centrifuge to sediment crystals of undissolved dye (if present surface of sections will be contaminated). 8. Aniline sulfate solution: Dissolve 1 g of aniline sulfate (toxic and dangerous to environment) in 10 ml of 0.05 M H2SO4 and 90 ml of 70% EtOH. 2.2

PAS Reaction

1. Periodic acid solution: 1% (w/v) H5IO6. Significantly lower concentration (0.2%) proved to be also efficient. 2. Schiff’s reagent according to de Tomasi: Dissolve 1 g of basic fuchsin in 200 ml of boiling distilled water. Stir the solution for 5 min and let it cool down to 50  C. Add 20 ml of 1 M HCl and let it cool down to 25  C. Add and dissolve 1 g of Na2S2O5 (potassium metabisulfite). Leave in dark for 14 to 24 h to gain a pale yellowish-orange clear solution. Add enough active coal (approx. 2 g) and shake for few minutes. Remove the coal with filtration on paper to gain clear colorless solution. Store in a dark tightly stoppered bottle at 4  C. The solution deteriorates with time. Discard it when it turns colored. 3. SO2 water: Mix 5 ml of 1 M HCl with 5 ml of 10% K2S2O5 and 100 ml H2O before use. Solution remains efficient for a few days in a closed bottle. 4. Reducing solution: Dissolve 1 g of KI and 1 g of Na2S2O3·5H2O in 50 ml of H2O, add 0.5 ml of 2 M HCl. Prepare fresh before use.

2.3 Callose Immunodetection

1. Primary antibody solution: Dilute monoclonal antibody toward (1–3)-β-glucan (Biosupplies Australia PTY Ltd) 1:100 in 1 PBS with addition of 10 μl of BSA stock per 1 ml of final solution. 2. Secondary antibody solution: Select anti-mouse or anti-rabbit IgG antibody of your choice (we use Invitrogen anti-mouse IgG Alexa Fluor 488; 1:1000) and dilute accordingly in 1

34

Alesˇ Soukup

PBS with addition of 10 μl of BSA stock per 1 ml of final solution. 3. 10 PBS (phosphate-buffered saline) stock solution: Weight 80.1 g NaCl, 2 g KCl, 14.7 g Na2HPO4. 2H2O, and 2.38 g KH2PO4 to prepare 1 l of solution. 10 PBS stock has pH 6.8, after dilution to 1 PBS should be pH 7.3. It is recommended to check with pH meter before use. 4. 10% BSA (bovine serum albumin) stock solution: Dissolve 1 g of powdered BSA (Fraction V) in 10 ml of distilled H2O. Store in 1 ml aliquots at 20  C. 5. Casein 3 stock solution: Add 3.33% (w/v) of casein into distilled water and titrate to pH 10 with minimal amount of 2 M KOH. Let casein to dissolve at 40  C with constant stirring (approx. 2 h). Once completely dissolved, titrate to pH 7 with minimum of 2 M HCl. Add 10 PBS stock (11% of volume of solution prepared previously) to gain 3% casein solution in PBS. Aliquots can be stored at 20  C. Before usage dilute (1:2) with PBS. 6. Buffered glycerol with n-propyl gallate: Add 3% (w/v) of npropyl gallate (antifade reagent) into glycerol and stir overnight at room temperature (it is not readily soluble in aqueous solutions). Mix 8:2 with TRIS buffer (0.1 M, pH 9.0). Centrifuge to remove undissolved propyl gallate. Solution can be stored in dark at 4  C for about a year. 7. TRIS buffer (0.1 M, pH 9.0): Dissolve 12.1 g TRIS base in approx. 750 ml of distilled water. Adjust with 1 M HCl to pH 9.0, and fill with distilled water to final volume of 1 l. 8. High-humidity chamber is used to prevent evaporation of low volumes of antibodies from slides. Simple chamber can be made of large petri dish with water soaked tissue or filter paper on the bottom. Glass rods are used to separate slides from the soaked tissue and prevent their contact. 2.4 Berberine– Toluidine Blue Staining

1. Berberine dye solution: 0.2% Berberine hemisulfate in water. The solution is close to the saturation and crystals will form when stored at 4  C, which should be redissolved before use. 0.1% solution is used in most publications but higher concentration does not cause overstaining. 2. Toluidine blue O dye solution: 0.05% w/v of toluidine blue O in water. 3. Crystal violet solution: 0.05% w/v crystal violet in water.

2.5 HCl: Phloroglucinol (Wiesner’s Reagent)

1. Acidified phloroglucinol solution: Phloroglucinol (1% w/v, saturated) solution in 18% aq. HCl. The solution oxidizes with time, turns deep yellow–brown and the intensity of reaction decreases. That is time to change it for fresh one.

Selected Methods of Cell Wall Histochemistry

35

2. Acidic glycerol to mount sections: Mix 75% of glycerol with 15% of H2O and 10% of concentrated H2SO4. The percentage is related to final volume of solution. 2.6

Ma€ule Reaction

1. KMnO4 solution: Prepare fresh 1% w/v solution of KMnO4 in distilled water. 2. Alkalized glycerol: 15% (w/v) solution of Na2CO3 in 50% aqueous glycerol (alternatively 15% ammonium hydroxide in 75% aq. glycerol can be used).

2.7 Peroxidase Activity Detection

Two optional cosubstrate mixtures can be used (items 1 and 2). 1. DAB reaction mixture: Prepare fresh solution just before incubation, containing 500 μl DAB stock (1 mg in 1 ml of distilled water, Note 1), 499 μl acetate buffer (pH 5; 0.1 M), 50 μl NiCl2 (8% w/v in distilled water). Add 1 μl of H2O2 (30% in distilled water) just before usage. 2. TMB reaction mixture: Prepare fresh solution just before incubation, containing 10 μl TMB stock (10 mg in 1 ml 96% ethanol) and 989 μl acetate buffer (pH 5; 0.1 M). Add 1 μl of H2O2 (30% in distilled water) just before usage. 3. Acetate buffer (pH 5; 0.1 M): Mix 14.8 ml of 0.2 M acetic acid with 35.2 ml 0.2 M sodium acetate and make up to 200 ml with distilled water. 4. Solutions for peroxidase inhibition: Use either (1) fresh solution of 3% H2O2 in methanol, (2) acetate buffer containing 0.1% sodium azide and 0.5% H2O2, or (3) acetate buffer containing 0.1% phenylhydrazine.

3

Methods

3.1 Toluidine Blue Staining

1. Stain fresh sections in toluidine blue O solution for 1–5 min. 2. Wash carefully in water. 3. Mount into water or low percentage glycerol (less than 25% aqueous solution) to maintain metachromatic staining (see Note 2). If resin sections are used, let them air-dry and mount them with nonaqueous media. 4. Observe in bright-field optics.

3.2 PAS Reaction for Detection of Cell Wall Polysaccharides

We use most commonly fresh sections. Other types of sections should be fully hydrated before treatment. 1. Select parallel control sections and skip H5IO6 oxidation step for these.

36

Alesˇ Soukup

2. Oxidize sections in 1% w/v H5IO6 for 1 min in laboratory temperature. Time should be adjusted properly if bigger objects (wholemounts) are treated. 3. Wash sections 3 in distilled water. 4. Optionally apply for 3 min the reducing solution and wash again with distilled water. Solution can be applied to eliminate remaining of periodic acid, not necessary for sections but can be useful for bigger objects as wholemounts. 5. Stain in Schiff’s reagent for 10 min. 6. Wash very carefully in SO2 water 3  10 min to prevent oxidation of reduced colorless fuchsin and unspecific background staining. 7. Mount into 50% v/v glycerol in SO2 water. 8. Purple coloration of the tissue is tightly bound to cell wall material, so it is also possible to dehydrate objects and use permanent mounting. The background staining strongly depends upon efficient Schiff reagent wash out. 9. Observe in bright field. Presence of polysaccharides should be indicated with purple coloration, compare with control sections. If indigenous aldehydes are present before periodic acid treatment (in control sections), their reduction might be performed in the beginning of procedure (see Note 3). 3.3 Calcofluor Staining

1. Dilute the stock solution 1:100 with water and stain objects for 0.5–5 min. We use fresh sections but other types of sections should work if fully rehydrated. 2. Wash in water. 3. Mount in water, 50% glycerol or other aqueous mounting media. Objects might be observed also directly in staining solution as unbound Calcofluor emits very low fluorescence. 4. Observe under UV excitation. The cell wall material should yield a pale blue signal.

3.4 Alcian Blue Staining

1. Rinse the sections in acetic acid (3% aq. solution). 2. Control sections might be methylated to block free carboxyl groups in acidified methanol (1 M HCl in MetOH) for 4 h at 60  C [4]. Methylation should mask free carboxyl and therefore inhibit polyanionic staining. Methylation of the carboxyl can be reverted with alkalized ethanol (1% KOH in 70% EtOH, 10 min at laboratory temperature). 3. Stain in Alcian blue for 30 min at laboratory temperature. 4. Thoroughly wash in 3% acetic acid (at least 10 min).

Selected Methods of Cell Wall Histochemistry

37

5. Mount into 75% glycerol (sections could be also dehydrated and mount permanently). 6. Observe in bright field optics. Polyanionic compounds stain cyan. 3.5 Ruthenium Red Staining

1. Use fresh or fully rehydrated sections. 2. Control sections might be methylated to block free carboxyl groups in acidified methanol (1 M HCl in MetOH) for 4 h at 60  C [4]. Methylation should mask free carboxyl and therefore inhibit polyanionic staining. Methylation of the carboxyl can be reverted with alkalized ethanol (1% KOH in 70% EtOH, 10 min at laboratory temperature). 3. Stain the sections until the walls are red (normally within 5 min). 4. Rinse the sections with water. 5. Mount in water or 50% glycerol. 6. Observe in bright field optics. Polyanionic compounds stain intensely red. Limited penetration of the dye into the tissues was reported, which should be considered evaluating the results on thicker sections or wholemounts.

3.6 Aniline Blue Fluorochrome (Sirofluor) Staining

1. Stain the sections for 5–10 min in solution of aniline blue or flood sections with solution of Sirofluor (stock diluted 1:50 in distilled water or suitable buffer). Aqueous solutions low in ionic solutes decrease background staining of cellulose [45]. 2. Rinse carefully with water or suitable buffer. 3. Mount into water or 50% glycerol (alternatively it is possible to observe directly in staining solution due to low fluorescence of unbound fluorochrome in water solutions). 4. Observe in fluorescence setup. Fluorochrome yields yellow–green fluorescence with blue excitation and pale-yellow fluorescence with UV excitation. Control sections should be included to make sure of the fluorescence source.

3.7 Resorcin Blue Staining

1. Dilute prepared solution 1:50 with distilled water and stain sections for 1–2 min. 2. Carefully wash 3 in water. 3. Mount into citrate buffer pH 3.2 (or 50% buffered glycerol). 4. Observe in bright field optics. Callose is stained blue while lignified structures change the color to red in low pH.

3.8 Callose Immunodetection

1. Fix pieces of tissue in chilled (20  C) methanol for 5–10 min. 2. Wash 2  5 min v PBS.

38

Alesˇ Soukup

3. Prepare sections, preferentially fresh hand sections. Cryosections dried to slides should be rehydrated for 30 min in PBS. If paraffin sections are used, paraffin should be removed from sections in toluene, rehydrated via alcohol series down to water, and PBS. 4. Block the nonspecific protein binding with 1% casein solution in PBS, 15 min. 5. Wash in PBS for 2 min and blot excess of solution from the edge of slide. 6. Apply primary antibody in high-humidity chamber at laboratory temperature for 2 h (time of application might be prolonged). 7. Wash 2  10 min in PBS, blot excess of solution. 8. Apply secondary antibody in humidity chamber at laboratory temperature for 2 h. 9. Wash 2  5 min in PBS. 10. Counter stain with toluidine blue for 3 min. Overstaining might suppress detection (see Note 2). 11. Mount into glycerol with propyl gallate. 3.9 Sudan Red 7B or Fluorol Yellow Staining Procedure

1. Fresh hand sections or cryotome sections are best suitable for the staining. 2. Blot sections to minimize transfer of water into staining solution, optionally 75% glycerol wash might be included before staining to minimize precipitation of dye. 3. Stain in the solution for minimum of 1.5 h at laboratory temperature (the time can be extended significantly without a risk of overstaining; warming the staining solution up to 60  C might accelerate the staining process). 4. Quickly rinse excess of dye solution with detergent solution (e.g., 0.5% aq. solution of SDS). 5. Carefully wash with water. 6. Mount into 75% aq. glycerol. 7. Sudan red gives intense red coloration of lipidic compounds, while Fluorol yellow yields green/yellowish fluorescence with UV excitation (for possible pitfalls see Note 4).

3.10 Berberine: Toluidine Blue Staining

1. Stain sections (we preferentially use fresh ones after aldehyde fixation) for at least 1 h in Berberine solution. 2. Wash twice with water. 3. Counter stain in 0.05% toluidine blue O in water for 5–10 min. Alternatively Crystal violet (syn. Gentian violet) can be used to efficiently quench background fluorescence.

Selected Methods of Cell Wall Histochemistry

39

4. Wash carefully with water. 5. Mount into water or 25–50% glycerol. 6. Observe under UV excitation as a yellowish fluorescence or under blue excitation as a green emission. 3.11 Wiesner’s (HCl: Phloroglucinol) Reaction

1. Stain the sections (we preferentially use fresh ones, but paraffinembedded sections might be used after rehydration) at laboratory temperature with acid phloroglucinol solution till the sections turn red (within few minutes). 2. Mount the sections into glycerol acidified with sulfuric acid to maintain the reaction product, which last for several days or even weeks. Hydrochloric acid (escaping hydrogen chloride) is highly aggressive to metallic and optical parts of the microscope. That is why it is strongly recommended not to use it in close vicinity of the microscope. 3. Observe with bright field optics. Lignin modified cell walls stain cherry red.

3.12 Aniline Sulfate Procedure

1. Treat the section with aniline sulfate solution for 5 min at lab temperature. 2. Mount into acidified glycerol described for Wiesner’s reaction. 3. Observe in bright field. Lignins are stained bright yellow.

3.13 Ma€ule Reaction for Lignin

1. Oxidize sections in solution of KMnO4 for 10–20 min. 2. Wash 3 with distilled water. 3. Flood with 1 M HCl and wait until the dark precipitate disappears (normally it takes about 30–60 s). 4. Wash gently with water. 5. Mount into alkalized glycerol. 6. Lignin is colorized red or brown red.

3.14 Peroxidase Activity Detection

1. Fix the object in 4% formaldehyde in phosphate buffer (25 mM, pH 6.8) for 2–4 h at room temperature. 2. Carefully wash fixative out of sections with phosphate buffer (25 mM, pH 6.8) 2 for 15–20 min. 3. Prepare sections (we normally use hand sections) and select parallel sections for controls. 4. Recommended controls are as follows: (1) sections treated with reaction mixture without H2O2; (2) sections with peroxidase inhibited with H2O2 in methanol, 10 min at laboratory temperature; and (3) sections with peroxidase inhibited with phenylhydrazine, 10 min at laboratory temperature. 5. Wash sections carefully with acetate buffer 2  5 min.

40

Alesˇ Soukup

6. Treat the section with the incubation medium at 37  C for 1 h (or longer if peroxidase activity is weak). 7. Wash section carefully with acetate buffer 2  5 min. 8. Mount into 50% glycerol. 9. Observe with bright field optics.

4

Notes 1. DAB is commonly used as hydrochloride, which is more soluble. If DAB is not in the form of hydrochloride it should be dissolved first in a drop of dimethylformamide and then add to buffer. Low concentration of DMF (up to ~0.5%) should not affect peroxidase activity. DAB is carcinogenic. Waste should be oxidized (commercial bleach, KMnO4) before being discarded. NiCl2 catalyzes precipitation of the reaction product and decreases its run from reaction site, thus improving the accuracy of localization. 2. Toluidine blue staining is very convenient for fresh sections. Metachromasy is stable only in aqueous (highly polar) solutions and disappears in organic solvents [46]. It fades even if mount in stronger (we normally do not exceed 25%) glycerol solutions. The intensity of the staining (concentration of dye solution) should be adjusted according to type and thickness of the section. As thicker freehand sections might be overstained with presented dye concentration, it is reasonable to dilute staining solution 5–10. Other types of sections (e.g., paraffin [47] or hydrophilic resin sections [48]) work well if fully hydrated before staining and air-dried afterward before mounting into nonaqueous mounts. Acetate buffer pH 4.4 can be used instead of water to prepare dye solution for more consistent results. 3. To reduce aldehydes on sections, dissolve 5 mg of NaBH4 in 10 ml of borate buffer (pH 7.6) and treat section for 1 h in lab temperature [4]. 4. There are several pitfalls of the Sudan red staining procedure. First, there might be problem with unspecific precipitation of Sudan red pigment on sections. The primary reason might be in water contact with the dyeing solution, which produces crystals of hydrophobic Sudan. The dyeing solutions remains stable for considerable period of time (months), but is sensitive to water absorption and deteriorates if let open for a long time. We have also experienced staining problems dues to long-term storage (several years) of PEG 400 used for preparation of the solution. Be also careful with microscope setup to localize well cell wall response as plasma membrane staining may in some

Selected Methods of Cell Wall Histochemistry

41

cases cause coloration of cell walls due to refraction. The possibility of nonspecific staining of strongly acid structures (chromosomes) was indicated by Lillie [49].

Acknowledgments The current update of this chapter, first published as ref. 50, has been supported by the NPUI-LO1417 project. References 1. O’Brien TP, Feder N, McCully ME (1964) Polychromatic staining of plant cell walls by toluidine blue O. Protoplasma 59:368–373 2. Sylve´n B (1954) Metachromatic dye- substrate interactions. Q J Microsc Sci 93–95:327–358 3. Bergeron JA, Singer M (1958) Metachromasy: an experimental and theoretical reevaluation. J Biophys Biochem Cytol 4:433–457 4. Pearse AG (1985) Histochemistry (theoretical and applied). Churchill Livingstone, Edinburgh 5. Rost FWD (1995) Fluorescence microscopy. Cambridge University Press, Cambridge 6. Kasten FH, Burton V, Glover P (1959) Fluorescent Schiff-type reagents for cytochemical detection of polyaldehyde moieties in sections and smears. Nature 184:1797–1798 7. Truernit E, Bauby H, Dubreucq B, Grandjean O, Runions J et al (2008) Highresolution whole-mount imaging of threedimensional tissue organization and gene expression enables the study of phloem development and structure in Arabidopsis. Plant Cell 20:1494–1503 8. Herth W, Schnepf E (1980) The fluorochrome, calcofluor white, binds oriented to structural polysaccharide fibrils. Protoplasma 105:129–133 9. Wood PJ, Fulcher RG, Stone BA (1983) Studies on the specificity of interaction of cereal cell wall components with Congo red and Calcofluor. Specific detection and histochemistry of (1–3), (1–4), β-D-glucan. J Cereal Sci 1:95–110 10. Benesˇ K (1968) On the stainability of plant cell walls with alcian blue. Biol Plant 10:334–346 11. Luft JH (1971) Ruthenium red and violet. I. Chemistry, purification, methods of use for electron microscopy and mechanism of action. Anat Rec 171:347–368 12. Muhlethaler K (1950) Electron microscopy of developing plant cell walls. Biochim Biophys Acta 5:1–9

13. Soukup A, Votrubova´ O (2005) Woundinduced vascular occlusions in tissues of the reed Phragmites australis: their development and chemical nature. New Phytol 167:415–424 14. Redgwell RJ, Selvendran RR (1986) Structural features of cell-wall polysaccharides of onion Allium cepa. Carbohydr Res 157:183–199 15. Mullendore ML, Windt CW, Van As H, Knoblauch M (2010) Sieve tube geometry in relation to phloem flow. Plant Cell 22:579–593 16. Currier HB, Strugger S (1956) Aniline blue and fluorescence microscopy of callose in bulb scales of Allium cepa L. Protoplasma 45:552–559 17. Eschrich W, Currier HB (1964) Identification of callose by its diachrome and fluorochrome reactions. Stain Technol 39:303–307 18. Evans NA, Hoyne PA, Stone BA (1984) Characteristics and specificity of the interaction of a fluorochrome from aniline blue (sirofluor) with polysaccharides. Carbohydr Polym 4:215–230 19. Smith MM, McCully ME (1978) A critical evaluation of the specificity of aniline blue induced fluorescence. Protoplasma 95:229–254 20. Priestley JH (1921) Suberin and cutin. New Phytol 20:17–29 21. Soukup A, Tylova´ E (2018) Apoplastic barriers: their structure and function from a historical perspective. In: Sahi VP, Balusˇka F (eds) Concepts in cell biology-history and evolution. Springer, pp 155–183 22. Soukup A, Armstrong W, Schreiber L, Franke R, Votrubova´ O (2007) Apoplastic barriers to radial oxygen loss and solute penetration: a chemical and functional comparison of the exodermis of two wetland species, Phragmites australis and Glyceria maxima. New Phytol 173:264–278 23. Pollard M, Beisson F, Li Y, Ohlrogge JB (2008) Building lipid barriers: biosynthesis of cutin and suberin. Trends Plant Sci 13:236–246

42

Alesˇ Soukup

24. Brundrett MC, Kendrick B, Peterson CA (1988) Efficient lipid staining in plant material with Sudan red 7B or fluorol yellow 088 in polyethylene glycol. Biotech Histochem 66:133–142 25. Lux A, Morita S, Abe J, Ito K (2005) An improved method for clearing and staining free-hand sections and whole-mount samples. Ann Bot 96:989–996 26. Jensen WA (1962) Botanical histochemistry. Freeman, San Francisco 27. Zimmermann A (1892) Die botanische mikrotechnik: Ein handbuch der mikroskopischen pr€aparations-reaktions- und tinktions- methoden. Verlag der H.Laupp’schen Buchhandlung, Tu¨bingen 28. Brundrett MC, Enstone DE, Peterson CA (1988) A berberine–aniline blue fluorescent staining procedure for suberin, lignin, and callose in plant tissue. Protoplasma 146:133–142 ˝ber das Verhalten des 29. Wiesner J (1878) Note u Phloroglucins und einiger verwandter Ko¨rper zur verholzten Zellmembrane. SitzungsberAkad Wiss Math-naturw Kl 77:60–66 30. Akin DE (1989) Light microscopy and histology of lignocellulose related to biodegradation. In: Chesson A, Ørskov ER (eds) Physicochemical characterization of plant residues for industrial and feed use. Elsevier Applied Science, London, NY, pp 58–64 31. Pomar F, Merino F, Barcelo AR (2002) O-4linked coniferyl and sinapyl aldehydes in lignifying cell walls are the main targets of the Wiesner (phloroglucinol-HCl) reaction. Protoplasma 220:17–28 32. Clifford MN (1974) Specificity of acidic phloroglucinol reagents. J Chromatogr 94:321–324 33. Gahan PA (1984) Plant histochemistry and cytochemistry—an introduction. Academic, London 34. Johansen DA (1940) Plant microtechnique. McGraw-Hill Book Co, New York 35. Stafford HA (1962) Histochemical and biochemical differences between lignin-like materials in Phleum pratense L. Plant Physiol 37:643–649 36. Ibrahim RK, Towers GHN, Gibbs RD (1962) Syringic and sinapic acids as indicators of differences between major groups of vascular plants. J Linn Soc Lond Bot 58:223–230 37. Harris PJ, Hartley RD (1976) Detection of bound ferulic acid in cell walls of the Gramineae by ultraviolet fluorescence microscopy. Nature 259:508–510

38. Harris PJ, Hartley RD (1980) Phenolic constituents of the cell walls of monocotyledons. Biochem System Ecol 8:153–160 39. Ursache R, Andersen TG, Marhavy´ P, Geldner N (2018) A protocol for combining fluorescent proteins with histological stains for diverse cell wall components. Plant J 93:399–412 40. Almagro L, Go´mez Ros LV, Belchi-Navarro S, ˜ o MA (2009) Class III Ros Barcelo´ A, Pedren peroxidases in plant defence reactions. J Exp Bot 60:377–390 41. Frederick SE (1987) DAB procedures. In: Vaughn KC (ed) CRC handbook of plant cytochemistry–cytochemical localization of enzymes. CRC, Boca Raton, FL, pp 3–23 42. Graham RC, Karnovsky MJ (1966) The fine structural localization of peroxidase activity. J Histochem Cytochem 14:291–302 43. Brand JA, Tsang VC, Zhou W, Shukla SB (1990) Comparison of particulate 3, 30 , 5, 50 -tetra-methylbenzidine and 3, 30 -diaminobenzidine as chromogenic substrates for immunoblot. BioTechniques 8:58–60 44. Mesulam MM (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferents. J Histochem Cytochem 26:106–117 45. Evans NA, Hoyne PA (1982) A fluorochrome from aniline blue: structure, synthesis and fluorescence properties. Aust J Chem 35:2571–2575 46. Pal MK (1965) Effects of differently hydrophobic solvents on the aggregation of cationic dyes as measured by quenching of fluorescence and/or metachromasia of the dyes. Histochem Cell Biol 5:24–31 47. Sakai WS (1973) Simple method for differential staining of paraffin embedded plant material using toluidine blue O. Biotech Histochem 48:247–249 48. O’Brien TP, McCully ME (1981) The study of plant structure: principles and selected methods. Termarcarphi Pty LTD, Melbourne 49. Lillie RD (1977) HJ Conn’s biological stains: a handbook on the nature and uses of the dyes employed in the biological laboratory. Sigma Chemical Company, St. Louis 50. Soukup A (2014) Selected simple methods of plant cell wall histochemistry and staining for light microscopy. Methods Mol Biol 1080:25–40

Chapter 3 Chemical Fixation, Immunofluorescence, and Immunogold Labeling of Electron Microscopical Sections Ilse Foissner and Margit Hoeftberger Abstract Knowledge about the spatiotemporal distribution patterns of proteins and other molecules of the cell is essential for understanding their function. A widely used technique is immunolabeling which uses specific antibodies to reveal the distribution of molecular components at various structural levels. Immunofluorescence gives an overview about the distribution of molecules at the level of the fluorescence or confocal laser scanning microscope. Electron microscopy offers the highest resolution of morphological techniques and is thus an indispensable tool for the analysis of molecule distribution patterns at the subcellular level. In this chapter we describe selected routine methods for immunofluorescence and for labeling ultrathin sections of resin-embedded material with antibodies conjugated to colloidal gold, including protocols for chemical fixation, embedding, and sectioning. Key words Chemical fixation, Colloidal gold, Embedding, Whole-mount immunofluorescence, Immunolabeling, Sectioning, Transmission electron microscopy

1

Introduction

1.1 General and on Preparation of Samples

Immuno techniques use antibodies to detect, characterize, and localize specific antigens in cell extracts (e.g., dot and Western blots, immunoprecipitation) and in situ inside tissues and cells (immunolabeling or immunocytochemistry; ref. 1; Fig. 1a–f). A prerequisite for successful immunolabeling/immunocytochemistry is the preservation of the antigen, especially on electron microscopy (EM) sections where only few binding sites are available for recognition by the antibody. For immunofluorescence chemical fixation is the method of choice (Fig. 1a). The fixation solutions mostly contain a mixture of formaldehyde, which better preserves the antigen, and glutaraldehyde, which better preserves the cytoarchitecture. Often a compromise between preservation of cytoarchitecture and preservation of antigen has to be made. Prior to treatment with antibodies the cell wall has to be permeabilized chemically (by cell wall-degrading enzymes; e.g., ref. 2) or mechanically

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_3, © Springer Science+Business Media, LLC, part of Springer Nature 2019

43

44

Ilse Foissner and Margit Hoeftberger

Fig. 1 Immunofluorescence confocal images and immunogold labeled EM sections. (a) Immunofluorescence confocal image of a cotyledon epidermal cell of a mutant of Arabidopsis thaliana [30] expressing GFP-HDEL, a marker of the endoplasmic reticulum, and labeled with anti-tubulin and a red fluorescent secondary antibody. The primary antibody recognizes microtubules (red strands) and an unrelated protein which accumulates in perinuclear tubules of the endoplasmic reticulum (green; yellow color indicates colocalization at the nucleus N). (b) Transmission electron micrograph of a glutaraldehyde and OsO4-fixed, Epon-embedded cotyledon epidermal cell of the same Arabidopsis mutant with perinuclear tubules (arrow) near the nucleus (N). (c) EM section of the aldehyde-fixed and LR Gold-embedded cotyledon epidermal cell after immunolabeling with a primary antibody against tubulin and a secondary antibody conjugated to 15 nm gold particles (arrows). M mitochondrion, N nucleus. (d) Confocal laser scanning microscope image of a root of Arabidopsis thaliana expressing a vacuolar protein fused to GFP. The sample was counterstained with the red fluorescent endocytic marker FM4-64. (e) Similar root tip as in (d) which was high pressure-frozen, cryosubstituted, and embedded in LR Gold. The EM image shows a section through a rhizodermal cell after immunolabeling with anti-GFP. Gold particles (arrows) coupled to the secondary antibody are visible in the vacuole (V). G Golgi stack. (f) EM section of high pressure frozen, cryosubstituted, and LR Gold-embedded internodal cell of the green alga Chara braunii after immunolabeling with a primary antibody against a cell wall (CW) polysaccharide and a secondary antibody conjugated to 10 nm gold particles (arrows). M mitochondrion, V vacuole. Scale bars are 25 μm (d), 5 μm (a), and 500 nm (b, c, e, f)

Immunofluorescence and Immunogold Labeling

45

(sectioning or freeze-shattering; ref. 3). For immunogold labeling of EM sections, high pressure-frozen and cryosubstituted materials are frequently used because ultrarapid freeze fixation is considered to preserve structure and antigen binding sites more efficiently than chemical fixation (refs. 4–7, Fig. 1e, f). For antigen-rich material, however, chemical fixation with aldehydes is also sufficient with the advantage that sophisticated instruments are not required (ref. 6, Fig. 1b, c). Before sectioning, cells or tissues are embedded in polar acrylic resins [1, 8–12]. Epoxy resins because of their hydrophobicity and the high temperature necessary for polymerization are not suited for immunogold labeling. To study their distribution, antibodies must be conjugated to markers which differ according to the method used for observation. Enzymes which produce colored products with chromogenic substances are suitable for bright field microscopy. Fluorescent markers can be detected by using conventional epifluorescence microscopy (FM) or confocal laser scanning microscopy and related techniques (Fig. 1a). For EM, electron-dense gold particles are commonly used (ref. 13, Fig. 1c, e, f). After increasing the size of gold particles by silver enhancement, the distribution of gold conjugated probes can also be visualized in the bright field microscope. Immunolabeling is complemented by genetic and molecular biological techniques. Proteins can be visualized in the confocal laser scanning microscope by expressing them together with the amino acid sequence of a fluorescent marker (e.g., green fluorescent protein (GFP)). Coexpression with known organelle markers (proteins or signaling peptides) tagged with a different fluorescent protein gives a first hint for the localization of the protein under study. If a suitable antibody is available and if the fluorescent markers are fixable, this method can be combined with immunofluorescence (Fig. 1a). The distribution of the fluorescent probes can also be studied at much higher, even “suborganellar” resolution in the EM using specific commercially available gold-conjugated antibodies which recognize the fluorescence markers (Fig. 1d, e). Immunogold labeling thus bridges the gap between molecular biology and cellular architecture at the fine structural level [5–7, 14]. 1.2 Direct and Indirect Immunolabeling

Markers for immunolabeling can either be conjugated to the primary antibody which binds to the epitope of a protein or other molecule under study, or to a secondary antibody which specifically recognizes the primary unlabeled antibody. Accordingly, we distinguish between direct and indirect immunolabeling. Direct immunolabeling is used, for example, for lectins which recognize specific sugar residues and for enzymes to localize their substrates [8]. However, the most widely applied method is indirect immunolabeling. Although it requires additional incubation and washing steps, indirect immunolabeling—in general—is time and cost saving because there is no need to label every primary antibody and

46

Ilse Foissner and Margit Hoeftberger

one secondary antibody can be used for a variety of primary antibodies. Even more importantly, indirect immunolabeling amplifies the signal because several secondary antibodies can bind to the primary antibody (see Subheading 1.4). 1.3 Primary Antibodies Used for Immunolabeling

Antibodies are either polyclonal or monoclonal [15]. Polyclonal antibodies are collected from the serum of an immunized animal (usually an herbivore) and recognize different epitopes, thus increasing the likelihood of antibody binding. Unfortunately, the serum will not only contain antibodies against the protein, peptide, or carbohydrate used for immunization but also antibodies against contaminant molecules and native antibodies of the immune system (e.g., antibodies raised against the food which is especially disturbing when working with plant cells). Therefore, preimmune serum should be collected prior to immunization and should be used for control experiments. Monoclonal antibodies are produced by cell cultures originating from fusion of an antibody producing cell with a myeloma cell and are specific for one epitope thus reducing unspecific binding. However, there is no guarantee that this epitope is exposed and available for antibody binding. Different classes of immunoglobulins are present in the serum of immunized animals. Most antibodies used for immunolabeling are IgGs which are purified in order to reduce background staining. The antigen binding sites are localized at the light chains located at the ends of the arms of the Y-shaped immunoglobulin.

1.4 Secondary Antibodies and Markers

The secondary antibody must specifically recognize the first antibody, that is, it must be directed against the host species. For example, if the primary antibody is mouse IgG an anti-mouse immunoglobulin raised in another animal (e.g., goat) is required. The most frequently used secondary antibodies are affinity-purified polyclonal IgGs (whole molecules) generated by immunizing an animal with whole IgG. These secondary antibodies recognize different epitopes so that several molecules with their attached markers will bind to one primary antibody which results in an amplification of the signal as compared with direct immunolabelling. Polyclonal anti-IgGs will also react with the light chains of IgAs, IgMs, IgDs, and IgEs when used as primary antibodies. Monoclonal secondary antibodies, fragments of secondary antibodies and antibodies raised against fragments are likewise commercially available but not so often used for immunolabeling of plant cells. The secondary antibodies are either coupled with fluorescent molecules for immunofluorescence or with colloidal gold for immunolabeling at the EM level. Secondary antibodies with different markers can be used for double immunostaining. This is usually done by using primary antibodies raised in different animals (e.g., in mouse and in goat).

Immunofluorescence and Immunogold Labeling

47

Besides of antibodies, a large variety of fluorescent markers which differ in excitation and emission wavelengths are commercially available. The decision for a specific marker depends on the available microscope, that is, on the wavelengths available for excitation and the filters available for detection of the emission fluorescence (a spectral detector allows arbitrary selection of wavelengths). For double immunofluorescence or for the comparison between fluorescently tagged protein and an antibody-labeled molecule, markers should either differ in excitation or in emission, preferably in both. Prior to buying expensive antibodies, plant cells and tissues must always be inspected for autofluorescence of chloroplasts, cell walls, and vacuoles. The colloidal gold method is applicable to most of the electron microscopical systems and techniques (transmission EM, scanning EM; ultrathin sections and freeze fracture). Gold particles can be absorbed to antibodies by complex electrochemical interactions and are available in distinct sizes between 1.4 (nanoparticles) and 30 nm. The larger gold particles can be seen at low magnifications of the EM. They have the disadvantage, however, that because of steric hindrance, a lower number of secondary antibody molecules can bind to the primary antibody which decreases the extent of signal amplification. Therefore, it is better to use antibodies conjugated to small particles and to increase their size by silver amplification, if necessary. Instead of secondary antibodies, Protein A-gold or Protein G-gold can be used. Both bind to a single site at the Fc region of antibodies of various host species (but not to all!) which means that signal amplification is not possible. Streptavidin-gold can be used in combination with a biotinylated antibody. An alternative to conventional colloidal gold is fluorescent nanogold particles [16] and quantum dots [17, 18]. These probes allow for studying the same specimen by fluorescence microscopy and by EM (correlative microscopy). Visualization of quantum dots in the EM requires the use of an energy-filtered transmission electron microscope. In the following, we describe selected protocols for immunofluorescence and for immunogold labeling of resin embedded EM sections of chemically fixed plant material. Further information, including paraffin sections, can be found in refs. 1, 5, 6, 9–17, 19–22, as well as in protocols provided on the home pages of various laboratories (e.g., https://www.liverpool.ac.uk/emunit/ working-with-us/protocols/, https://synapseweb.clm.utexas. edu/protocols) or companies (e.g., https://www.leica-micro systems.com/products/sample-preparation-for-electron-micros copy/). Detailed procedures for high-pressure freezing and cryosubstitution are described in refs. 19, 23, 24. Protocols for preembedding immunogold labeling are to be found elsewhere [25] as well as immunolabeling of cryosections [4, 26]. The latter method offers the advantage that epitope antigens are

48

Ilse Foissner and Margit Hoeftberger

much better preserved and accessible for antibody binding but requires the use of a cryostat microtome. Methods to quantify immunogold localization are described in ref. 27.

2

Materials All solutions should be prepared with double distilled water if available (see Note 1). Reagents should be of analytical grade and can be stored at room temperature, except bovine serum albumin (BSA, 4
C) and antibodies (4
C or frozen, see Subheading 2.4). Buffers can be prepared as 10 stock solutions and supplemented by 0.02% NaN3 for longer storage. All other solutions should be freshly prepared. Fixatives and their vapors, as well as Formvar and chloroform used for coating of EM grids, are harmful. Therefore, always use gloves and work under a fume hood when using these chemicals.

2.1 Chemical Fixation and Permeabilization for Whole-Mount Immunofluorescence

Fixation solutions should be prepared immediately before use and the osmolarity of the fixation solution should correspond to that of the material to be fixed. 1. Fixation solution for immunofluorescence: 1.5% (w/v) paraformaldehyde (see Note 2) and 0.5% (v/v) glutaraldehyde (diluted from 25% commercially available stock solution) in PEMT buffer: 100 mM PIPES (piperazine-1,4-bis(2-ethanesulfonic acid)), 5 mM EGTA (ethylene glycol-bis(2-aminoethylether)-N,N,N0 ,N0 -tetraacetic acid), 2 mM Mg2+SO4, 0.05% (v/v) Triton X-100; pH 6.9 [3]. 2. Water jet pump and accessories for vacuum infiltration, if necessary. 3. Cell wall digestion: 0.05% (w/v) Pectolyase Y-23 in PEM buffer (50 mM PIPES, 2 mM EGTA, 2 mM MgSO4) [28] or cell wall digesting enzyme mix: Cellulysin 1%, pectinase 1%, driselase 2% in PEM buffer [3]; enzymes are available from, for example, Sigma-Aldrich. 4. Instead of chemical cell wall digestion, samples can be plungefrozen and shattered in liquid nitrogen [3]. Necessary equipment: microscope slides or metal plates, clothes-peg or clamp and chilled aluminum blocks, liquid nitrogen (see Note 3, Fig. 2a–c). 5. Washing solution: PEMT and PEM buffer, respectively. 6. Blocking buffer for immunofluorescence (see Note 4): 1% BSA (w/v) and 50 mM glycine (w/v; Note 5) in PBS (phosphate buffered saline: 140 mM NaCl, 3 mM KCl, 2.38 mM KH2PO4, 76.1 mM Na2HPO4; pH 7). 7. Mounting solution: 50% glycerol in PBS (see Note 6).

Immunofluorescence and Immunogold Labeling

49

Fig. 2 Equipment for freeze shattering and for preparing the LR Gold–benzil embedding solution. (a) and (b) Two metal plates connected at one side by a clamp viewed from above (a) and from side (b). The red dot indicates the position of the chemically fixed sample. (c) Aluminum block and aluminum plate for breaking the sample after plunge freezing in liquid nitrogen. (d) The UV initiator benzil is conveniently dissolved by passing nitrogen through the LR Gold. The nitrogen expels oxygen whose presence prevents polymerization of LR Gold. A flexible tube connects the pipette tip to a nitrogen flask 2.2 Chemical Fixation, Dehydration, Embedding, and Sectioning for EM

1. Fixation solution for EM: 3% (w/v) paraformaldehyde (see Note 2), 1% (v/v) glutaraldehyde, 50 mM PIPES, pH 7.0 + 2 small droplets of 0.1% CaCl2 per 50 ml (see Note 7 and Subheading 2.1, item 1). 2. Washing solution: 50 mM PIPES. 3. Dehydration: 10%, 20%, 30%, 40%, 50%, 70%, 80%, 90%, and 100% ethanol (see Note 8). 4. Embedding in LR Gold: ethanol LR Gold 3:1, 1:1 1:3; LR Gold; LR Gold +0.1% benzil (w/v). Mix LR Gold and benzil (e.g., Agar Scientific, Stansted, UK; see Note 9) thoroughly but without introducing air bubbles because LR Gold does not polymerize in the presence of oxygen. If available, let N2 flow through the solution for several minutes in order to get rid of oxygen (Fig. 2d). 5. Gelatine capsules (size 000), plastic molds suitable for LR Gold embedding or aluminum dishes (diameter ca. 4 cm), UV lamp, freezer (temperature between 15
C and 20
C). 6. Nickel or gold grids (see Note 10). 7. 0.3% Formvar solution for grid coating: under a fume hood dissolve Formvar (Agar Scientific, Stansted, UK) in chloroform using a sonicating water bath, or use commercially available Formvar solutions. Store bottle well sealed with Parafilm at 4
C. 8. Sectioning: Trimmer (e.g., Leica EM TRIM, Leica Microsystems Vienna), Ultramicrotome (e.g., Leica UC7), Diamond

50

Ilse Foissner and Margit Hoeftberger

knife (e.g., Diatome, Hatfield, USA), distilled water, eyelash manipulator, fine antimagnetic tweezers with curved tips (e.g., Dumont no 7; Note 11), grid holder pad, grid storage box. 2.3 Immunogold Labeling of EM Sections

1. pH buffers: we use either PBS (see above) or one of the following pH buffers as a component of the blocking buffer. Buffer components are sequentially dissolved in water and adjusted to pH 7.4 with HCl or NaOH. Tris–HCl (tris(hydroxymethyl) aminomethane; 50 mM): Tris base or Trizma preset crystals (Sigma-Aldrich, St. Louis, USA; T7693; Note 12). TBS (Tris-buffered saline): 10 mM Tris base or Trizma preset crystals in 150 mM NaCl. 2. Blocking buffer for immunogold staining: Use a magnetic stirrer at moderate speed (see Note 13) to mix: Tris–HCl (see Note 14), TBS or PBS, 1% (w/v) BSA (see Note 15), 0.1% (v/v) Tween 20 (polyoxyethylene-sorbitan monolaurate; Note 16), 50 mM glycine (for aldehyde fixed cells only; Note 5).

2.4 Antibody Solutions

Antibodies are diluted in blocking buffer for immunofluorescence or for labeling EM sections (see above). The working solutions should be prepared immediately before use and centrifuged (maximum speed of microfuge or up to 10,000  g for 5 min) in order to remove larger aggregates. 1. Stock solutions of most primary antibodies should be stored at 20
C and frozen in aliquots in order to prevent repeated freezing and thawing; some primary antibodies require storage at 4
C or at room temperature. For each primary antibody the optimum concentration must be tested (see Note 17). 2. Stock solutions of secondary antibodies are usually stored at 4
C and often can be used for months and even years. 3. For working solutions, start with information given in the manufacturer’s data sheets (see Note 18). Further information on antibodies (problems, testing) can be found elsewhere [29].

2.5

Miscellaneous

1. Microfuge, magnetic stirrer, low-power dissection microscope, shaker, pH meter, sonicator, lamp. 2. Dust mask, protective glasses, and lint-free gloves. 3. Disposable Pasteur pipettes and pipettes with adjustable volumes between 0.1 and 10 μl, 10 and 50 μl, 100 and 1000 μl. 4. Petri dishes of various diameters, beakers. 5. Wet chamber (a plastic box with a rack to support petri dish or a plastic box lined with wet filter paper). 6. Scalpel, scissors, and new razor blades.

Immunofluorescence and Immunogold Labeling

51

7. Parafilm, Whatman filter paper, silicon-free lens paper (e.g., Pelco Optical Lens Tissue; Ted Pella, Redding, USA). 8. Acetone.

3

Methods The aim of all immunolabeling procedures is to obtain a maximum specific signal (reflecting the distribution of the antigen) with a minimum of nonspecific background. The here described methods must eventually be adapted to the sample investigated. In case of a low signal-to-noise ratio first try to find the optimum concentrations of primary and secondary antibodies (see Notes 17 and 18). Then the concentration and composition of blocking buffer should be varied in order to reduce nonspecific binding of antibodies [29] or of colloidal gold to the sample (see Notes 14–16). On EM sections nonspecific labeling is mostly a problem with primary antibodies. They often stick to cell walls, chloroplasts, and starch granules even in well fixed and embedded cells or tissues as well as to the Formvar film which supports the EM sections on grids.

3.1 Immunofluorescence: Fixation, Permeabilization, Labeling, and Mounting

Before immunolabeling the sample of interest a number of positive controls should be made (see Note 19). The necessary negative controls should be run in the same experiment (see Note 20). 1. Place sample (whole plant or organ) into a small petri dish filled with immunofluorescence fixation solution and fix for 30 min at room temperature. Fixation should be as fast as possible and the volume of the fixative should be 5–10 times higher than that of the sample. If the sample does not sink, keep it immersed with forceps and make some incisions with a sharp razor blade or excise small parts (see Note 21). Alternatively or in addition, samples can be fixed under vacuum with the aid of a water vacuum pump in order to replace the air within the intercellular spaces by the fixative (see Note 22). 2. Transfer samples into Eppendorf tubes and wash with buffer (washing solution) at least three times for 15 min each (see Notes 23 and 24). 3. Permeabilize cell walls with enzymes or use freeze-shatter method [3] to obtain fragments of tissues and cells. For freeze shattering, plants or plant fragments are loosely sandwiched between two microscope slides or metal plates, connected at one side by clothes-peg or clamp, plunge-frozen in liquid nitrogen, and broken between chilled aluminum blocks (Fig. 2a–c). The sample fragments are scratched off the metal or thawed and washed away with buffer and collected in Eppendorf tubes.

52

Ilse Foissner and Margit Hoeftberger

4. For immunofluorescence of detergent insoluble actin filaments or microtubules further permeabilization for up to 30 min with 1% Triton X-100 dissolved in fixation buffer or PBS is recommended (see Note 25). After this step wash again three times each with buffer. 5. Replace buffer with blocking buffer and incubate for at least 30 min. 6. Replace blocking buffer with primary antibody solution and incubate for at least 1 h at room temperature or overnight at 4
C (see Note 26). 7. Gradually bring sample and solutions (buffers, antibody solutions) back to room temperature (if necessary) and replace primary antibody solution with blocking buffer. 8. Wash with blocking buffer three times for 15 min each (see Note 24). 9. After removing excess buffer, add secondary antibody solution and incubate for at least 1 h at room temperature or overnight at 4
C (see Note 27). 10. Remove excess liquid and wash with buffer or with PBS (3  15 min each). 11. Replace buffer or PBS by mounting solution and prepare samples for fluorescence or confocal laser scanning microscopy (see Note 28). 3.2 Chemical Fixation and Embedding for Electron Microscopy

1. Fix material as described above (see Subheading 3.1) with EM fixative for 2 h at room temperature (see Note 29). 2. Transfer fixed plant fragments into Eppendorf tubes (see Note 23). 3. Wash several times with 50 mM buffer for 15 min each. 4. Dehydrate in a graded water–ethanol series for at least 15 min each: 10%, 20%, 30%, 40%, 50%, 70%, 80%, 90%, followed by 3  100% ethanol at 4
C on ice or in a fridge. 5. Replace 100% ethanol by ethanol–LR Gold 3:1, 1:1, 1:3 and 100% LR Gold; each step 6–24 h at 4
C. For the last step place the Eppendorf tubes in an exsiccator which is stored in a fridge or in a cold room. Leave Eppendorf tubes open so that any residual ethanol or water can evaporate. 6. Finally transfer samples into gelatine capsules filled with LR Gold–benzil and cover with lid. If plastic molds or aluminum dishes are used shield the embedding solution with a transparent plastic foil (cling film) and omit air bubbles (see Note 30). Place capsules or molds in an exsiccator stored at 4
C for 12 h.

Immunofluorescence and Immunogold Labeling

53

7. Polymerize in freezer (between 15 and 20
C) with a UV lamp (386 nm) for at least 12 h. The samples should be placed several cm beneath the UV lamp. 3.3 Preparation of Coated Grids

1. Choose a day of low humidity and make sure that all used materials are free of water and dust. 2. Wash grids several minutes with acetone (100%), place them on lint free filter paper and cover them with a lid of petri dish. 3. Fill a large glass petri dish (about 20 cm in diameter and 4.5 cm in height) with double distilled water and cover with the lid. 4. Fill a 25 ml glass beaker with the Formvar solution and cover it to protect it from dust and air moisture. Cool Formvar solution should be brought to room temperature before opening the glass bottle to avoid water condensation. 5. Clean a slide thoroughly with lint- and silicon-free lens paper. 6. Place the slide into the Formvar solution and after a few seconds lift it evenly. The slower the lift, the thinner the film will be, since chloroform vapor dissolves the film. 7. Lean the slide transversally on lint free filter paper and scrape around the film edge (about one to 2 mm from the rim) with a metal needle or a clean razor blade. 8. Dunk the slide (with the scored film side up) very slowly into the water at ~45
angle to float off film in the water. At first the film peels away from the front edge of the slide. The deeper the slide immerses, the more film area will detach from the slide till the whole film floats on the water surface. If the film does not detach, breath on the slide (water vapor facilitates the separation of the film) immediately before you immerse it. 9. Illuminate the film with a lamp and check the thickness. The film should be uniformly grey which corresponds to a thickness of about 50 nm. Gold or colored areas are too thick. 10. Take a grid with a fine curved tweezers and put it carefully (with the dull side down) on the film. Proceed with other grids till all “good” areas of the film are occupied. 11. Cut a strip of Parafilm and lay it with the clean side down on the Formvar film. As soon as the whole Formvar film adheres to the Parafilm, remove it with a tweezers and place it with the grids up on lint free filter paper in a petri dish. Close with a lid and let the film dry.

3.4

Sectioning

Please be aware that the following instructions do not replace the personal introduction and training at the available technical equipment!

54

Ilse Foissner and Margit Hoeftberger

1. Material embedded in aluminum dishes must be excised with a heated scalpel and mounted on the tip of Beem capsules using two component glue or super glue. Material embedded in adequate molds or in gelatine capsules can be used immediately. 2. Remove surrounding embedding material with a trimmer. The upper side of the block should get a pyramidal shape with the sample on its top and the surface of the top should have a trapezoid shape. Remove as much embedding material as possible, since the smaller the cutting face is trimmed, the easier the sectioning will be. Make the top and the bottom edge of the trapezoid as parallel as possible in order to get a ribbon of sections. 3. Clamp the trimming block on your ultramicrotome (e.g., Leica EM UC7; user manual), put the segment arc in its opening, fasten the specimen holder, and insert your block in a way that the bottom edge of the trapezoid is facing you. 4. Clean a razor blade very carefully with lint and silicon free lens paper to remove oil and dust, then hold it with thumbs and forefingers at an approximately 45
angle and cut small slices top down the pyramid to sharpen the edges. 5. Move the block forward so that the next edge of the pyramid is facing you and sharpen it with the razor blade. Proceed in the same way with the two other edges. Be sure that top and bottom edges of the trapezoid remain parallel. 6. Insert the segment arc in the specimen arm. The parallel edges of the trapezoid should be oriented horizontally. 7. Place the knife stage on the support. 8. Rinse the diamond knife carefully with a jet of double distilled water from a spray bottle. Dry it with lint-free filter paper but neither touch the cutting edge of the diamond nor the inside of the boat. 9. Place the diamond knife in the knife holder, fix it carefully and set the clearance angle (usually 6
). Fill the boat with double distilled water and be sure that the knife edge is fully coated. Draw water off with the refiller syringe, until the surface of the water in front of the cutting edge reflects silvery. 10. Move the knife stage close to the block in the specimen arm. Use the back-light for alignment of the diamond knife with the sample. If the knife’s reflection on the block face is not uniform, the cutting edge and sample edge are not oriented parallel to each other. Correct the position of the knife by the precision drive for pivoting. 11. Control by up and down movements of the block over the cutting edge to ensure the block face and cutting edge are

Immunofluorescence and Immunogold Labeling

55

oriented vertically parallel to each other. If this is not the case adjust it by tilting the sample via the segment arc. 12. Set cutting window, speed (usually 1 mm/s), and thickness (usually 40–60 nm) and start cutting. 13. Look through the binoculars and control sectioning. Sections will float on the surface of the water and reflect grey (up to about 50 nm) to silver (about 60 nm). Thicker sections from golden to colored are less suitable for immune EM. 14. Stop the block when it is below the level of the knife, for collecting the sections. 15. Increase the water level with the refiller syringe and collect the sections with the eyelash manipulator. 16. Take a Formvar coated nickel or gold grid with fine, curved tweezers and control the intactness of the Formvar film under the binoculars. 17. Bring the grid parallel to the water level with the Formvar coated side facing the sections. Then touch the sections with caution. The sections and a drop of water will adhere to the grid. 18. Carefully touch the rim of the grid with lint free filter paper to draw off excess water (Fig. 3c). 19. Control whether the sections adhere to the grid and put the grid with the film/sections side up on a grid holder pad. After drying overnight, grids may be stored in a grid storage box. 20. To continue sectioning, decrease the water level and start again as described above. 21. When sectioning is finished, take the diamond knife out of the knife block and rinse it carefully with a spray of double distilled water. Dry it with lint free filter paper, but neither touch the cutting edge nor the inner side of the boat. Put the knife in its box and cover it with lint-free filter paper so that remaining water will be absorbed. Close the box and store the knife until next use. 3.5 Immunogold Labeling of EM Sections

Work in a dust-free clean environment and avoid contamination of solutions, sections and material. Wash forceps in distilled water between each step. Cover petri dish between handling of grids in order to reduce evaporation. Never let grids completely dry out before the end of the incubation procedure! Perform positive and negative controls (see Notes 19 and 20). 1. Place a sheet of Parafilm on the bottom of a petri dish (see Note 31; Fig. 3a). Be careful not to touch the surface of the Parafilm with ungloved hands.

56

Ilse Foissner and Margit Hoeftberger

Fig. 3 Immunogold labeling of EM sections. (a) Petri dish with Parafilm onto which a mesh-like pattern was engraved using a blunt needle. (b) Series of incubation steps for immunogold labeling and two important negative controls (see Note 20). The sections on the grid in the upper row (now in banding buffer BB) will be incubated in primary antibody (pAB). The sections on the grid in the middle row will be incubated in preimmune serum (PIS) to check specificity of primary, polyclonal antibody. The sections on the grid in the lower row will be incubated in blocking buffer to check specific binding of the secondary antibody (sAB). All other steps are the same. Ad ¼ distilled water. (c) Removal of excess liquid by vertically touching filter paper with grid. (d) A wedge of filter paper is used to remove liquid between prongs of tweezers

2. Place 20–50 μl droplets of blocking buffer onto the Parafilm (see Note 32; Fig. 3b). 3. Put grids on top of droplets with section side facing downward and incubate for 30 min at room temperature (see Note 33; Fig. 3b). 4. Place droplets of the primary antibody solution onto the Parafilm (Fig. 3b). 5. Remove grids from blocking buffer; remove excess liquid by touching the edge of the grid with filter paper (Fig. 3c) and place on top of droplets containing the primary antibody. Incubation time should be 1–2 h at room temperature or up to 24 h in fridge (see Note 34). 6. Remove grids from droplet with primary antibody solution, remove excess liquid and pass grids through a series of at least

Immunofluorescence and Immunogold Labeling

57

four droplets (50 μl) of blocking buffer with at least 10 min incubation each (see Note 24). 7. After removing excess liquid, place grids on top of droplets containing secondary, gold-conjugated antibody diluted in blocking buffer and incubate for at least 1 h at room temperature (or up to 12 h at about 4
C). 8. Remove excess liquid and pass grids through a series of at least four droplets of blocking buffer (10 min each) and two droplets of distilled water. 9. Remove grids from droplets and rinse both sides by drop-like addition of about 5  1 ml of distilled water from a Pasteur pipette. Remove excess liquid as described above and use a small wedge of filter paper to absorb liquid trapped between prongs of forceps that hold grid (Fig. 3d). 10. Allow grids to dry on a grid holder pad with sections facing upward. 11. Option 1: After step 9 use silver enhancement to enlarge the size of small (2–5 nm) gold particles (see Note 35). 12. Option 2: After step 9 stain sections by placing grids on droplets of 2% aqueous uranyl acetate for 10 min (see Note 36).

4

Notes 1. The conductivity of the water should be 1 μS.m1. It is advisable to use freshly distilled water not stored for a longer time in plastic containers. 2. Paraformaldehyde should be dissolved freshly or diluted from stock solutions which can be stored at 20
C. A 10% stock solution is prepared by dissolving 2 g of paraformaldehyde in 20 ml water heated to 60
C under a fume hood. While stirring, add several droplets of 1 N KOH until the solution is clear (some particles will eventually remain). 3. We prefer unbreakable metal plates although they have the disadvantage that the exact position of the sample is not visible. Protect your fingers by wearing cold insulating gloves! 4. The blocking buffer provides the necessary pH for antibody binding and contains various agents which help to reduce unspecific staining. 5. Glycine masks free aldehyde groups introduced by fixation; the concentration should vary between 10 and 50 mM. Free aldehyde groups can also be quenched by washing with 10–100 mM NH4Cl, pH 7, or by 0.1% NaBH4.

58

Ilse Foissner and Margit Hoeftberger

6. The addition of antifade reagents helps to reduce bleaching of sensitive “old fashioned” fluorochromes like FITC (fluorescein isothiocyanate) and/or when samples are investigated with conventional fluorescence microscopes where irradiation exposure is higher than with the laser scanning microscopes. 7. This fixation solution yields good results for cotyledons and leaves but is less suited for roots. For fixation of roots see for example ref. 3. 8. Prepare solutions 1 day prior to embedding and store at 4
C. Anhydrous ethanol is not required for embedding in acrylic resins. 9. Benzil is a UV catalyst, and LR Gold–benzil solutions should therefore be protected from intense light. The concentration of benzil is dependent on the power and emission spectrum of the UV lamp used for polymerization. The cross-link density of the final resin is important. If resins are not penetrating sufficiently quickly reduce the benzil concentration rather than the light intensity or exposure time. 10. Copper grids may form precipitates upon exposure to buffers. 11. Tweezers’ points should always be protected with a cap or fixed by wire wrapped around the levers. Damaged tweezers can be deburred by pulling a fingernail file or double-sided sandpaper between the closed forceps prongs, but often they are no longer suited for picking up EM grids. 12. We prefer Trizma preset crystals because they need no pH adjustment. Notice that certain silver chloride pH electrodes do not give accurate results with Tris buffers. In that case better use pH sticks. 13. Fast stirring will produce excessive BSA foam. 14. If sample contains positively charged components which unspecifically attract negatively charged gold particles the ionic concentration of the Tris–HCl buffer might be too low. In that case the use of TBS or PBS is suggested. 15. BSA saturates nonspecific binding sites and neutralizes positively charged samples. The BSA should be free of IgGs. Acetylated BSA (available from Aurion, Wageningen, The Netherlands) has a much higher binding affinity than nonacetylated BSA and more effectively blocks polycationic sites. The concentration of BSA should vary between 0.1% and 2.5%. Instead of or additionally to BSA preimmune or nonimmune serum from the host species of the secondary antibody can be used up to 5%. 16. Tween facilitates wetting of the sections and prevents hydrophobic interactions between gold particles and sample; the concentration should vary between 0.1% and 1%.

Immunofluorescence and Immunogold Labeling

59

17. The appropriate concentration of the antibodies is crucial for successful immunolabeling. When a new primary antibody is applied, a dilution series between 1:1 up to 1:5000 should be tested. In general, EM immunolabeling requires an up to 10 higher concentration of primary antibody than used for Western blotting because of the lower number of binding sites, whereas the concentration for whole-mount immunofluorescence often is in between. Ideally, the primary antibody should be diluted so that nonspecific binding is reduced below the level of detection. 18. In the absence of the primary antibody and at the optimum concentration of the second antibody no fluorescence should be visible in immunofluorescence samples and no gold conjugates should be present on sections (compare Note 20). 19. Dot blots can be used to prove the presence of the antigen in cell extracts and to detect inhibitory effects of fixatives (e.g., glutaraldehyde or OsO4) on antibody binding. Western blots are required to show that the antibody recognizes a (single) protein with the appropriate mass. It is also a good idea to first test a new antibody for immunofluorescence where more antigens are available for binding. When immunofluorescence fails labeling of EM sections is unlikely to be successful. It must also be kept in mind that antibodies suited for Western blots are not necessarily suited for immunostaining. In Western blots antibodies recognize the denatured antigen whereas in samples for immunolabeling antigens are supposed to be in a more natural conformation. But it is also possible that the antigen has become inaccessible, destroyed, or extracted during preparation. 20. Several negative controls are needed in order to confirm the specificity of staining, especially when working with polyclonal primary antibodies. But also monoclonal and secondary antibodies may bind nonspecifically to the sample. The most important negative controls are the following: (1) In case of polyclonal antibody replace primary antibody with preimmune serum collected from the animal before immunization (see Subheading 1.3). If preimmune serum is not available, use at least nonimmune (normal) serum of the same species but this is considered not to be an adequate substitute. (2) Replace the primary antibody with a control antibody from the same species raised against an antigen known to be absent in the tissue. (3) If available, use mutants lacking the antigen or, when the distribution of transgenic proteins is studied, use nontransformed material as controls. (4) Absorb the primary antibody with its antigen by addition of antigen which should abolish binding. (5) Use blocking buffer instead of primary antibody

60

Ilse Foissner and Margit Hoeftberger

to get information about the background signal caused by unspecific binding of the secondary antibody. 21. Work under a fume hood in order to avoid breathing of the aldehyde vapor. 22. During operating the vacuum pump open and close cap of the vacuum glass bottle several times until the sample sinks. 23. Use a steel pin for engraving the Eppendorf tubes; felt pen markings are likely to be dissolved during handling. 24. All washing steps given in this protocol are minimums! In case of high background staining or dirt increase number of washing steps and incubation times. Carefully remove solutions and disperse “pellet” gently while adding solutions. 25. Other proteins (cytosolic or membrane-bound) may be dissolved or strongly diluted by excess permeabilization with Triton X-100. 26. Incubation times can be reduced at higher temperatures up to 37
C but care must be taken that the samples do not dry out! 27. Protect secondary fluorescent antibody solutions from intense light. Short time exposure to room light is not harmful. 28. When applying new protocols check success of washing before this step! 29. Unlike as with conventional EM postfixation with OsO4 is omitted because of its detrimental effect on epitope structure. 30. LR Gold does not polymerize in the presence of oxygen. 31. A small droplet of water between Parafilm and petri dish helps to fix the sheet if required. 32. Droplets placed on the intersections of a mesh like pattern scratched into the Parafilm with the tip of blunt tweezers or needles are more likely to remain in place. But be careful not to perforate the Parafilm! 33. Do not immerse grids into droplet. If that happens deliberately, either thoroughly wash the grid in distilled water and proceed again or continue by immersing the grid with section side facing upward into the droplets until the end of the procedure. If sections are collected on uncoated nickel or gold grids both surfaces can be stained by fully submerging grids in reagents. Such grids can also be used for double labeling when each side is exposed to different primary and secondary antibodies (but not immersed!). Uncoated grids must have a higher number of meshes than Formvar-coated grids in order to support the EM sections. 34. For prolonged incubation place covered petri dish in a wet chamber. Vibration of the samples during incubation helps to shorten incubation time and may also reduce nonspecific

Immunofluorescence and Immunogold Labeling

61

background labeling. The dish of shakers may become hot after prolonged use. A thin sheet of polystyrene beneath the petri dish or the moist chamber prevents heating of samples. 35. For easier detection at low magnification the size of the gold particles can be increased with silver enhancement kits available from different companies. 36. This step is only required for sections with very low contrast and when using conventional EMs without energy-filtering. Too much contrast makes recognition of small gold particles difficult. Thoroughly wash grids with water after staining.

Acknowledgments This work was supported by the Austrian Science Fund (project no. P 27536-B16 to IF). References 1. Hawes CR, Satiat-Jeunemaitre B (2001) Plant cell biology: a practical approach. Oxford University Press, Oxford 2. Shimamura M (2015) Whole-mount immunofluorescence staining of plant cells and tissues. In: Plant microtechniques and protocols. Springer International Publishing, pp 181–196 3. Wasteneys GO, Willingale-Theune J, Menzel D (1997) Freeze shattering: a simple and effective method for permeabilizing higher plant cell walls. J Microsc 188:51–61 4. van Donselaar E, Posthuma G, Zeuschner D, Humbel BM, Slot JW (2007) Immunogold labeling of cryosections from high-pressure frozen cells. Traffic 8:471–485 5. Oliver C, Jamur MC (2010) Immunocytochemical methods and protocols. In: Methods in molecular biology. Humana Press, Totowa, NJ 6. Schwartzbach SD, Osafune T (2010) Immunoelectron microscopy: methods and protocols. Humana Press, Totowa, NJ 7. Stierhof YD, El Kasmi F (2010) Strategies to improve the antigenicity, ultrastructure preservation and visibility of trafficking compartments in Arabidopsis tissue. Eur J Cell Biol 89:285–297 8. Hall JL, Hawes CR (1991) Electron microscopy of plant cells. Academic Press, London 9. Bozzola JJ, Russell LD (1999) Electron microscopy: principles and techniques for biologists. Jones & Bartlett Publishers International, London

10. Griffiths G (1993) Fixation for fine structure preservation and immunocytochemistry. In: Fine structure immunocytochemistry. Springer, Berlin, Heidelberg, pp 26–89 11. Newman GR, Hobot JA (2012) Resin microscopy and on-section immunocytochemistry. Springer Science & Business Media, Berlin, Heidelberg 12. Verkleij AJ (1989) Immuno-gold-labeling in cell biology. CRC Press, Boca Raton 13. Hayat MA (1991) Colloidal gold. Principles, methods and applications. Academic Press, London 14. Koster AJ, Klumperman J (2003) Electron microscopy in cell biology: integrating structure and function. Nat Rev Mol Cell Biol (Suppl):SS6–SS10 15. Javois LC (1999) Immunocytochemical methods and protocols. Humana Press, Totowa, NJ 16. Takizawa T, Robinson JM (2000) FluoroNanogold is a bifunctional immunoprobe for correlative fluorescence and electron microscopy. J Histochem Cytochem 48:481–486 17. Nisman R, Dellaire G, Ran Y, Li R, BazettJones DP (2004) Application of quantum dots as probes for correlative fluorescence, conventional, and energy-filtered transmission electron microscopy. J Histochem Cytochem 52:13–18 18. Alivisatos AP, Gu W, Larabell C (2005) Quantum dots as cellular probes. Annu Rev Biomed Eng 7:55–76

62

Ilse Foissner and Margit Hoeftberger

19. McDonald KL, Webb RI (2011) Freeze substitution in 3 hours or less. J Microsc 243 (3):227–233 20. Maunsbach A (1998) Immunolabeling and staining of ultrathin sections in biological electron microscopy. In: Celis JE (ed) Cell biology: a laboratory handbook, vol 3. Academic, New York, pp 268–276 21. Sarraf CE (2000) Immunolabeling for electron microscopy. Methods Mol Med 40:439–452 22. VandenBosch K (1992) Localization of proteins and carbohydrates using immunogold labelling in light and electron microscopy. Mol Plant Pathol 2:31–44 23. Austin JR 2nd (2014) High-pressure freezing and freeze substitution of Arabidopsis for electron microscopy. Methods Mol Biol 1062:473–486 24. Karahara I, Kang B-H (2014) High-pressure freezing and low-temperature processing of plant tissue samples for electron microscopy. ˇ a´rsky´ V, Cvrcˇkova´ F (eds) Plant cell morIn: Z phogenesis: methods and protocols. Humana Press, Totowa, NJ, pp 147–157 25. Humbel BM, de Jong MD, Muller WH, Verkleij AJ (1998) Pre-embedding immunolabeling

for electron microscopy: an evaluation of permeabilization methods and markers. Microsc Res Tech 42:43–58 26. Tokuyasu K (1980) Immunochemistry on ultrathin frozen sections. Histochem J 12:381–403 27. Mayhew TM (2011) Quantifying immunogold localization on electron microscopic thin sections: a compendium of new approaches for plant cell biologists. J Exp Bot 62:4101–4113 28. Kawamura E, Himmelspach R, Rashbrooke MC, Whittington AT, Gale KR et al (2006) MICROTUBULE ORGANIZATION 1 regulates structure and function of microtubule arrays during mitosis and cytokinesis in the Arabidopsis root. Plant Physiol 140:102–114 29. Griffiths G, Lucocq JM (2014) Antibodies for immunolabeling by light and electron microscopy: not for the faint hearted. Histochem Cell Biol 142:347–360 30. Jancowski S, Catching A, Pighin J, Kudo T, Foissner I et al (2014) Trafficking of the myrosinase-associated protein GLL23 requires NUC/MVP1/GOLD36/ERMO3 and the p24 protein CYB. Plant J 77:497–510

Chapter 4 Essential Methods of Plant Sample Preparation for HighResolution Scanning Electron Microscopy at Room Temperature Jana Nebesa´rˇova´ Abstract In this chapter, conventional techniques are described for the preparation of plant samples at room temperature before examination in the high resolution scanning electron microscopy. Protocols are given on how to collect, to fix, to dehydrate, and to dry plant samples. Subsequently, it is described how to stick them to stubs and cover with a thin conductive layer. These methods are suitable for a wide variety of plant specimens, ranging from microalgae to higher plants. Key words High-resolution scanning electron microscope, Plant sample, Chemical fixation, Dehydration, Critical point drying, t-Butanol freeze drying, Metal coating, Sample storage

1

Introduction High-resolution scanning electron microscopy (HRSEM) is an excellent tool for visualization of plant surfaces at high magnification. However, there are many restraints on the specimen to be examined. Firstly, it has to withstand the vacuum. Secondly, it must be capable of dissipating the energy of the focused electron beam without building up a local surface electric charge. And third condition is that the specimen has to be a good source of secondary or backscattered electrons which are detected and used for image forming. Therefore, the basic prerequisite for successful surface imaging in most HR SEMs operated under high vacuum is the dehydration of plant specimens and then covering by a conductive layer. A number of processing techniques and protocols have been developed for preserving plant specimens as close as possible to their original structure and form, however none of them guarantees that the prepared samples will be artifact-free. The aim of this chapter is to describe the common protocol which can be used as

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_4, © Springer Science+Business Media, LLC, part of Springer Nature 2019

63

64

Jana Nebesa´rˇova´

a starting protocol for the preparation of a wide range of plant specimens, from microalgae to higher plants (see Note 1). The whole procedure involves several steps [1–6]: fixation in buffered aldehyde solution, postfixation in osmium tetroxide solution, and dehydration in an organic solvent. The fixed and dehydrated specimens are transferred to a drying apparatus, then mounted on a specimen stub, coated with a conductive layer and examined in the HR SEM. All the steps are carried out at room temperature unless stated otherwise (see Note 2). It should be noted that this protocol represents only the initial approach. It is clear that different types of plant specimens (e.g., bulk tissues such as leaf and root tissue, single cells such as algae or cyanobacteria in suspensions, or isolated cell components) may require a slightly different approach [7–10], which needs often to be determined empirically. Many of these modifications for a particular type of samples can be found in the literature.

2

Materials

2.1 Laboratory Equipment and Supplies

1. Tubes (e.g., Eppendorf tubes), dishes, containers. 2. Disposable plastic or Pasteur glass pipettes, automatic pipettes. 3. Small tools such as scalpels or razor blades (single edge) fine tweezers, needles, wood sticks, etc. 4. Glass bottles or metal tins with tight lid for storing of unused tetroxide osmium solutions. 5. Balances for measuring buffer salts, pH meter, magnetic stirrer, and stirring bar. 6. Laboratory refrigerator designated for the storage of samples and solutions. 7. Stubs and materials for gluing of dry specimens, like carbon or double tapes, silver or carbon paints (categories of mounting tabs and adhesives from the offer of EMS, SPI, Agar Scientific, and others). 8. Specimen stub storage box (e.g., Agar Scientific, Cat. No. AG16709, Ted Pella, Inc., Cat. No. 16770). 9. Stereo microscope, critical point dryer, sputtering coater. 10. Fume hood, desiccator to store unmounted dry specimens. 11. Personal protection: lab coat, goggles, latex gloves.

2.2

Reagents

1. Buffers: A number of buffer systems are available, however the most common used buffers for SEM preparation are the phosphate buffer of Sorensen, which is inexpensive and nontoxic, and cacodylate buffer suitable for a long-term storage [1–4]. Phosphate buffer is prepared by mixing of two

Sample Preparation for Scanning Electron Microscopy

65

Table 1 Phosphate buffer composition pH

A (ml)

B (ml)

pH

A (ml)

B (ml)

6.8

62.5

37.5

7.0

39.0

61.0

7.0

51.0

49.0

7.2

28.0

72.0

solutions: Solution A (2.76 g of NaH2PO4 · H2O or 3.12 g of NaH2PO4 · 2H2O in 100 ml of distilled water), and Solution B (2.84 g of NaHPO4 or 5.36 g of NaHPO4 · 7H2O or 7.17 g of NaHPO4 · 12H2O in 100 ml of distilled water). The pH of the buffer is adjusted by the ratio of both components as shown in Table 1. Cacodylate buffer: 4.28 g of sodium cacodylate ((CH3)2AsO2Na, CAS#124-65-2, EMS, Cat. No. 12200) is dissolved in 100 ml distilled water and the pH is adjusted by adding the appropriate volume of 0.2 M hydrochloric acid (HCl, CAS#7647-01-0) (1.7 ml concentrated HCl/100 ml ddH2O). The stock buffer solution with the concentration 0.2 M is possible to store at 4
C for several months. Caution: contains arsenic, poisonous, carcinogenic. Absorbed through skin, wear lab coat and gloves, work in a fume hood. Used solution of sodium cacodylate is collected in the labeled waste bottle at back of fume hood for picking up by waste service company. 2. Glutaraldehyde fixative solution: Glutaraldehyde is usually used as the primary fixative. It penetrates the specimen rapidly, crosslinks proteins, and stops the dynamics of living cell. In case of higher plants, their surface is usually covered by waxy substances in the cuticle that makes it hydrophobic. The penetration of GA solution can be speed up by the microwave irradiation [11–15]. Glutaraldehyde (GA, electron microscopy grade, CAS#111-30-8) is commercially available as a solution with the concentration 25% in sealed ampoules. The fixative solution with the concentration of 2.5% GA is easily prepared by making a 1:10 dilution in the cacodylate or phosphate buffer, that is, by mixing 50 ml of 0.2 M buffer stock solution at proper pH, 10 ml of 25% glutaraldehyde (EM grade) and 40 ml of distilled water. Caution: poisonous, irritant, work in a fume hood, wear gloves and lab coat. Used solutions of GA are collected in the labelled waste container placed at back of fume hood for picking up by waste service company. 3. Osmium tetroxide fixative: Osmium tetroxide is used as a secondary fixative following aldehyde fixation improving electrical conductivity. OsO4 (electron microscopy grade, CAS#20816-12-0) is commercially available as a solution with the concentration 4% in sealed glass ampoules (EMS,

66

Jana Nebesa´rˇova´

Cat. No. 19140). Use an ampoule cracker to open the vial with the fixative. Discard if the solution looks blackish (oxidized). The unused stock solution can be stored in a glass tube (e.g., Schott Duran type with Teflon-lined caps) at 4
C inside of a secondary glass jar with tight lid or metal tins. It is common practice to seal these bottle lids with Parafilm. Caution: OsO4 is toxic and extremely volatile; fumes are very corrosive, especially to mucous membrane and eyes. Work in a well-ventilated fume hood and wear gloves, goggles and lab coat. Collect used solutions and solid osmium tetroxide in a labeled leakproof waste container placed in a fume hood for picking up by a waste service company. 4. Dehydration solutions: Acetone or ethanol solutions with the concentration 25%, 50%, 70%, 90% are prepared mixing ddH2O and pure acetone/ethanol. We produce 100% (absolute) acetone/ethanol by adding molecular sieve or other desiccant (anhydrous CaCl2, CuSO4) into a bottle with pure acetone/ethanol. Anhydrous tert-butanol is commercially available (Sigma-Aldrich, Cat. No. 471712). Caution: Acetone/Ethanol is highly flammable. The solvents are used in a fume hood to avoid breathing of vapor. Used solutions of acetone–ethanol are collected in a labeled waste container at back of a fume hood for picking up by a waste service company.

3

Methods

3.1 Specimen Collection

3.2

Fixation

The method of collection depends on the type of the specimen. In the case of cell suspension, it is possible to use either repeated centrifugations or cells adhere to poly-L-lysine coated slides or microporous filters. Small plants may be examined in their entirety after surface cleaning using mechanical means or chemicals (see Note 3). If the specimen is larger than several centimeters, it is necessary to excise pieces having suitable dimensions with a scalpel or fresh razor blade (Fig. 1) and clean up their surface as in the previous case to remove any debris, since they can obscure surface details. 1. Specimens are immersed in fixative solution containing 2.5% of GA in 0.1 M buffer (recommended pH 7.0–7.2; see Note 4) and left here at room temperature in the case of single cells minimally 1–2 h, in the case of tissue minimally 4 h. The fixation solution is dispensed in small stoppered glass vials or Eppendorf tubes and the volume of fixative should exceed the specimen volume minimally ten times (e.g., 1 mm3 of sample into 1 ml of the fixative solution). If necessary, specimens may be stored in this fixative up to several weeks in a refrigerator, as long as evaporation is prevented.

Sample Preparation for Scanning Electron Microscopy

67

Fig. 1 Specimen fixation. The ivy leaf (Hedera hibernica) is cut into small pieces, which are washed in buffer and immersed in Eppendorf tubes with fixation solution containing 2.5% of GA. The specimen is usually placed on a sheet of soft material, like rubber or dental wax, and the cutting is performed by a sharp, clean razor blade to minimize the specimen compression

2. The fixative solution is removed with a Pasteur pipette and replaced with 0.1 M buffer. Then the specimen is rinsed three times in buffer, 5–15 min each. 3. 2% osmium tetroxide solution is added to the specimen in the appropriate amount so that it is diluted 1:1 by the buffer residue. The specimen is postfixed for 1–3 h depending on its size (see Note 5). 4. The postfixative solution is removed and the specimen is rinsed in buffer or distilled water three times, 5–15 min each. 3.3

Dehydration

1. Dehydration is performed in graded acetone series (20%, 50%, 70%, 90%), 10–15 min at each concentration (see Note 6). Practically, the exchange of solutions is performed using a glass Pasteur or plastic disposable pipette. First the solution with lower concentration of acetone is removed and immediately replaced with a solution of higher concentration of acetone. It is important to work quickly so that samples do not dry out. 2. In the case of necessity, the specimen can be stored in 70% acetone for no more than 1 or 2 h without damage.

68

Jana Nebesa´rˇova´

3. The dehydration is finished by two changes of 100% acetone, for about 10–20 min each change. 4. Transfer specimens, still submerged in pure acetone, to the drying device. 3.4 Drying Using the Critical Point Method

The method of critical point drying (CPD) is the most commonly used drying method for biological specimens including plants [16, 17]. It is based on the removal of surface tension effects which cause drying artifacts. At the critical point (critical temperature and pressure) the transition from liquid to gas takes place without an interface because the densities of liquid and gas are equal at this point. Carbon dioxide (CO2) is the most used transitional medium, which has the critical point at 31
C and 7.4 MPa. The procedure requires a special device supplied by a number of manufacturers (Leica, Quorum, Agar, etc.). 1. The pressure chamber of CPD apparatus is cooled to the temperature between 5 and 12
C. 2. The dehydrated specimens in pure acetone are loaded into the specimen holders of the critical point dryer, like special capsules or baskets made from stainless steel, polyethylene or microporous plastic (see Note 7). 3. The specimen holders are quickly inserted into the pressure chamber and the door is sealed. 4. The chamber is filled with liquid CO2. 5. The liquid CO2 is changed five times over a period about 10 min in order to flush out all residual acetone. 6. The specimens are left in CO2 for 20–45 min, depending on their size and hardness. Then the pressure chamber is gently warmed at the temperature 38
C and the pressure 9–9.7 MPa over a period of about 20 min. Now the liquid is converted to gas. 7. Once the critical point has been reached, the pressure is released by gentle opening of the vent valve while the temperature is still maintaining high to prevent recondensation of the liquid CO2. As far as possible, the gas should flow from the chamber at a constant rate so the chamber is depressurized in 10–15 min. 8. The specimen holders are removed from the pressure chamber of CPD and store in a dry and dust-free environment (desiccator).

3.5 Alternative Drying Procedure: Freeze Drying from tert-Butanol (TBA)

This method is quite old, but simple [18, 19], allowing to eliminate sample damage in the CPD procedure due to high pressure (shrinkage or wrinkling of the surface of the sample) and removing of residual dehydration solvent by washing (breakage of fine surface

Sample Preparation for Scanning Electron Microscopy

69

structures). TBA has a melting point between 25 and 26
C and is therefore solid at room temperature. The specimen is infiltrated with liquid TBA, which is subsequently frozen and then sublimed away. Since TBA is fully miscible with ethanol, it is better in this case to dehydrate the sample with graded ethanol series. 1. Dehydration is performed in graded ethanol series (20%, 50%, 70%, 90%, 100%, 100%), 10–15 min each step at RT. 2. In the meantime, TBA is warmed to 26
C or higher to liquefy it before mixing with the pure ethanol. The mixture of TBA and 100% ethanol consisting of 1:1 parts is prepared. 3. The 100% ethanol is removed from the vial with the sample and replaced by 1–2 ml of the TBA–ethanol mixture and left for 10 min. 4. The mixture is replaced by two changes of pure TBA, for about 10–15 min each change. 5. Before putting the vial with the specimen into the fridge, the volume of TBA is reduced as low level as possible, but be careful, samples must not dry out! 6. The specimen is placed in refrigerator with temperature of about 4
C for at least 2 h. 7. The sample is transferred in a vacuum chamber connected with a rotary pump. At the beginning it is necessary to slowly pump the chamber to prevent the rapid escape of a large amount of solidified TBA. 1 ml of solidified TBA is sublimated for approximately 1 h. To ensure complete removal of the TBA, it is necessary to pump at minimum for 30 min longer. 8. Dry specimens are removed from the vacuum chamber and stored in a dry and dust-free environment (desiccator). 3.6 Alternative Drying Procedure: In Osmium Tetroxide Vapors

This method is very useful in special cases when the manipulation with a specimen can cause artifacts (brakeage of fine surface structures or their displacement). On the contrary, it is not suitable for samples containing a lot of water or dispersed in water. 1. Pieces of the fresh unfixed specimen are placed in a suitable container which can be perfectly sealed (weighing bottle or a small bottle with a sealing cap). 2. A small crystal of osmium tetroxide is added into the weighing bottle with the specimen. The container is closed and the end with the cap is wrapped in Parafilm. 3. The sealed container is transferred into a freezer with temperature around 20
C and left there for 1–3 weeks depending on the amount of water in the sample and its size. The surface of the specimen is immediately fixed and hardened by vapors of osmium tetroxide. This is followed by a slow sublimation of the ice from the sample leading to its complete drying.

70

Jana Nebesa´rˇova´

4. The container is removed from the freezer and left unopened until its temperature will be equal to room temperature and the moisture on the container surface will be dried. The dry sample is stored in a dry and dust free environment (desiccator). 3.7 Specimen Mounting

Before the examination in HR SEM, dried specimens must be mounted on special stubs appropriate to a particular microscope (Fig. 2). In general, the sample should be mounted in the optimum direction with respect to microscope detectors and sufficiently firmly so it does not drift during the examination. The connection between the sample and the stub must be electrically conductive. Therefore, liquid adhesives, such as conductive silver- or carbondoped paints and glues, cyanoacrylates, or special cements available through suppliers of consumables for electron microscopy, are recommended for mounting larger and harder samples such as seeds, hard stems or leaves. Smaller and finer samples such as pollen grains, pieces of excised plant tissue, and algae are mounted using double-sided tacky tapes (see Note 8). The most often used tapes are carbon or copper double tapes, which are electrically conductive. If the tape is used to mount a bigger specimen or suspension glued to a cover glass, it is recommended to connect the top of specimen surface with the stub or tape by creating a conductive “bridge” by means of silver or carbon paint [1–3].

3.8 Specimen Coating

In order to increase image contrast of plant specimens and to eliminate charging during their examination in HR SEM, it is usually necessary to coat the specimen with a thin layer with a

Fig. 2 Specimen mounting. After drying, pieces of the ivy leaf are mounted to special metal stubs using conductive tapes containing carbon. The sample identification number may be written on the bottom of the stub

Sample Preparation for Scanning Electron Microscopy

71

high electrical and thermal conductivity, usually made of heavy metals such as gold, platinum, palladium, or their alloys (gold–palladium, platinum–palladium; refs. 1–3). The most frequently used method of the deposition is sputter coating since it is fast and reliable and is possible to spread a thin and homogeneous film with good electrical conductivity even on a very irregular specimen surface (Figs. 3 and 4).

Fig. 3 Specimen coating. The ivy leaf sample, which was fixed with 2.5% GA, dehydrated through a graded acetone series, dried in CPD and coated with gold, is ready for the examination in HRSEM

Fig. 4 The surface structure of the Hedera hibernica leaf. Specimen was prepared according to the described protocol without postfixation with osmium tetroxide, using acetone as dehydration agent and dried by the CPD method

72

Jana Nebesa´rˇova´

1. Dry samples mounted on stubs are transferred into the chamber of the sputter coater. Normally, 5–10 standard sized specimen stubs can be placed in the chamber. 2. The chamber is closed and the rotary pump is activated to evacuate it down to the required pressure. The lower the pressure, the less contaminated coating layer will be. Pumping may take 5–10 min to the recommended vacuum level. 3. Then the inert gas, generally dry argon, is introduced in the chamber by opening the relevant valve. The gas should flow at the rate that keeps the pressure in the chamber around 10 Pa. 4. After the recommended setting of the sputter coater is reached and stabilized, the high voltage can be switched on and specimen coating will start. The film thickness is monitored most readily by means of a quartz crystal thickness monitor or similar device, or by means of simple timed runs at given parameters (see Note 9). 5. When the film has reached the desired thickness, the gun and pumps are switched off and the chamber is vented. The coated sample is taken out and stored in a dry, dust-free environment before its examination in HR SEM. 3.9 Specimen Storing

4

Dried and coated specimens must be stored in such a way to protect them from mechanical damage, dust, moisture, oxidation, temperature changes, and other unfavorable effects. Microscopy suppliers offer a lot of special plastic containers that firmly hold specimen stubs. Moisture-sensitive samples may be protected by desiccants (silica gel, or calcium sulfate) added into the sealed storage container. Any contact of the specimen with the desiccant must be avoided since it can transfer dust onto the specimen surface. Samples with extreme moisture sensitivity can be stored under an inert atmosphere (dry nitrogen or argon) or in vacuum.

Notes 1. Plants represent a large spectrum of samples with a different content of water. The choice of technique will depend on the type of sample, the equipment available and the surface structure that should be visualized. At present different types of scanning electron microscopes are at disposal: an environmental scanning electron microscope (ESEM) allows to observe fresh hydrated samples but with lower resolution [20–22]. A low vacuum scanning electron microscope [23] can visualize mechanical stable plant samples with lower content of water like seeds, roots or stems without any treatments. The

Sample Preparation for Scanning Electron Microscopy

73

possibility to visualize hydrated samples close to their native state with high resolution brings low temperature scanning electron microscopy (LTSEM) allowing for the examination of cryo-fixed samples directly at the temperature of 140 to 120
C [24, 25]. LTSEM is also very often used to visualize internal structures of plant samples by freeze-fracture. 2. Every step of specimen preparation may induce some artifacts [26] which can complicate the interpretation of images obtained by HRSEM. We can observe most often a sample compression caused by the dehydration and high pressure during CPD drying. Also charging can complicate the HRSEM examination. The low electrical conductivity of dry plant samples, insufficient or irregular thickness of conductive layer, or the presence of wax layers on the observed surface contribute to this phenomenon. Endless efforts to observe plant sample specimens as close as possible to their native state and to minimize artifacts have led, in recent years, to a significant expansion of freezing methods in specimen preparation [24, 25] and using of HRSEM working at low voltage [27, 28]. 3. A clean sample surface is essential for obtaining a sharp and clear image. It means that all debris should be removed. Sample surface cleaning can be accomplished using air blowing in combination with sweeping with a soft brush. The surface of wet specimens can be gently washed with a suitable buffer solution. To remove stubborn debris, it may be necessary to try a solution of middle detergents, surfactants, or containing enzymes. 4. GA is considered to be the best fixation agent and is therefore most commonly used. However, there are many other fixation protocols using other fixative agents like formaldehyde (sometimes in the combination with GA), methanol, acetone, and ethanol [29–31]. Due to the presence of waxy substances in the plant cuticle, the penetration of aqueous fixation solution is slow. Therefore, a microwave irradiation [14], agitation or addition of a detergent into the fixation solution is frequently used to accelerate the fixation process. 5. Plant cells are characterized by considerable mechanical stability due to the presence of cellulose in the cell wall. Therefore, the postfixation with osmium tetroxide can be omitted in a number of cases. 6. The dehydration is carried out by gradual replacing the specimen water with increasingly concentrated solutions of organic solvents like ethanol or acetone. I prefer using of acetone which can contribute to hardening of the specimen surface and cause less shrinkage. It is also completely miscible with liquid carbon dioxide and does not require the use of a transition solvent before specimen drying by the method of critical point.

74

Jana Nebesa´rˇova´

7. The specimen transfer into baskets can be done at any point after osmium tetroxide fixation. Then dehydration can be done more quickly if the baskets are placed in a small beaker or bottle, in which we can exchange the dehydrating solution for all samples at one time. 8. Dry plant specimens are very often fragile, therefore special tools, like needles, wooden sticks, tweezers, and micropipettes are used for their transfer onto stubs. For very small objects, a single eyelashes glued to a wooden stick may be used. Under the stereomicroscope the fine tip of the eyelashes is soaked by an adhesive of the double sticky tape and then the sample is glued to it and transferred to the stub. Soft entomological tweezers can be used for transferring of larger specimens without mechanical damaging. 9. The most important parameters of specimen coatings are film thickness, its homogeneity and granularity. The film thickness is influenced by deposition rates that are affected by the material being sputtered, the efficiency of the sputtering system, the power that the system is being run and the distance from the metal target to the specimen position. Various methods for film thickness measurements have been reviewed by Flood [32]. At present, majority of sputtering systems is equipped with devices like quartz crystal thickness monitor allowing directly measure the thickness of a deposited film. As concerning film contamination, the greatest source is usually material remaining on the chamber wall from previous sputtering using a different metal. The vacuum chamber must be kept clean; it means that it should be cleaned regularly. The granularity can be influenced by the selection of coating materials. Gold, platinum, palladium, and their alloys are used standardly for coating, materials like tungsten, iridium, or chromium are recommended for ultrahigh-resolution imaging, because they produce films with finer grain sizes.

Acknowledgments I acknowledge the Czech-BioImaging large RI project (LM2015062 funded by MEYS CR) and by European Regional Development Fund-Project “National infrastructure for biological and medical imaging” (No. CZ.02.1.01/0.0/0.0/16_013/ 0001775) for their support.

Sample Preparation for Scanning Electron Microscopy

75

References 1. Bozzola JJ (2014) Conventional specimen preparation techniques for scanning electron microscopy of biological specimens. In: Kuo J (ed) Methods in molecular biology: electron microscopy – methods and protocols, 3rd edn. Springer ScienceþBusiness Media, New York 2. Robards AW, Wilson AJ (1993) Procedures in electron microscopy. John Wiley and Sons, Chichester 3. Bozzola JJ, Russel LD (1999) Specimen preparation for scanning electron microscopy. In: Electron microscopy principles and techniques for biologists, 2nd edn. Jones and Bartlett, Sudbury, MA, pp 48–71 4. Dykstra MJ, Ruess LE (2003) Biological electron microscopy: theory, techniques, and troubleshooting. Kluwer, New York 5. Fisher ER, Hansen BT, Nair V, Hoyt FH, Dorward DW (2012) Scanning electron microscopy. Curr Protoc Microbiol 25:2B.2.1–2B.2.47 6. Echlin P (2009) Handbook of sample preparation for scanning electron microscopy and X-ray microanalysis. Springer, New York 7. Pathan AK, Bond J, Gaskin RE (2008) Sample preparation for scanning electron microscopy of plant surfaces–horses for courses. Micron 39:1049–1061 8. Allen TD, Goldberg MW (1993) Highresolution SEM in cell biology. Trends Cell Biol 3:205–208 9. Bomblies K, Shukla V, Grajam C (2008) Scanning electron microscopy (SEM) of plant tissues. Cold Spring Harb Protoc 3:1–3 10. Vesk M, Dibbayawan TP, Vesk PA, Egan EA (2000) Field emission scanning electron microscopy of plant cells. Protoplasma 210:138–155 11. Zechmann B, Zellnig G (2008) Microwaveassisted rapid plant sample preparation for transmission electron microscopy. J Microsc 233:258–268 12. Kok LP, Boon ME (1992) Microwave cookbook for microscopists, art and science of visualisation, 3rd Revised edn. Coulomb Press, Leyden 13. Giberson RT, Demaree RS (2001) Microwave techniques and protocols. Humana Press, Totowa, NJ 14. Heumann HG (1992) Microwave-stimulated glutaraldehyde and osmium tetroxide fixation of plant tissue: ultrastructural preservation in seconds. Histochemistry 97:341–347

15. Login GR, Dvorak AM (1993) A review of rapid microwave fixation technology: its expanding niche in morphologic studies. Scanning 15:58–66 16. Bray DF (2008) Critical point drying of biological specimens for scanning electron microscopy. Methods Biotechnol 13:235–243 17. Bray DF, Bagu J, Koegler P (1993) Comparison of hexamethydisilazane (HMDS), Peldri II, and critical-point drying methods for scanning electron microscopy of biological specimens. Microsc Res Tech 26:489–495 18. Inoue´ T, Osatake H (1988) A new drying method of biological specimens for scanning electron microscopy: the t-butyl alcohol freeze drying method. Arch Histol Cytol 51:53–59 19. Kaneko Y, Matsushima H, Sekine M, Matsumoto K (1990) Preparation of plant protoplasts for SEM observation by t-butanol freeze drying method. J Electron Microscopy 39:426–428 20. McGregor JE, Donald AM (2010) The application of ESEM to biological study. J Physics: Conference Series 241:012021 21. Cheng YT, Rodak DE, Angelopoulos A, Gacek T (2005) Microscopic observation of condensation of water on lotus leaves. Appl Phys Lett 87:194112 22. Kirk SE, Skepper JN, Donald AM (2009) Application of environmental scanning electron microscopy (ESEM) to determine biological surface structure. J Microsc 233:205–224 23. Talbot MJ, White RG (2013) Cell surface and cell outline imaging in plant tissue using backscattered electron detector in a variable pressure scanning electron microscope. Plant Methods 9:40 24. Read ND, Jeffree CE (1991) Low-temperature scanning electron microscopy in biology. J Microsc 161:59–72 25. Sargent JA (1983) The preparation of leaf surfaces for scanning electron microscopy: a comparative study. J Microsc 129:103–110 26. Crang RFE (1988) Artifacts in specimen preparation for scanning electron microscopy. In: RFE C, Klomparens KL (eds) Artifacts in biological electron microscopy. Plenum Press, New York, pp 107–129 27. Cox G, Vesk P, Dibayawan T, Baskin TI, Vesk M (2008) High-resolution and low-voltage SEM of plant cells. In: Schatten H, Pawley JB (eds) Biological low-voltage scanning electron microscopy. Springer, New York

76

Jana Nebesa´rˇova´

28. Schatten H (2011) Low voltage highresolution SEM (LVHRSEM) for biological structural and molecular analysis. Micron 42:175–185 29. Talbot MJ, White RG (2013) Methanol fixation of plant tissue for scanning electron microscopy improves preservation of tissue morphology and dimensions. Plant Methods 9:36

30. Schwab B, Hulskamp M (2010) Quick and easy fixation of plant tissues for scanning electron microscopy. Cold Spring Harb Protoc 8:1 31. Neinhuis C, Edelmann HG (1996) Methanol as a rapid fixative for the investigation of plant surfaces by SEM. J Microsc 184:14–16 32. Flood PR (1980) Thin film thickness measurements. Scanning Electron microscopy I. Scanning Electron Microscopy Inc, Chicago, pp 183–200

Chapter 5 Fluorescence Lifetime Imaging of Plant Cell Walls Christine Terryn and Gabriel Pae¨s Abstract Fluorescence is a versatile property of many molecules called fluorophores. In plant cell walls, fluorescence is generally attributed to aromatic molecules such as lignin. In contrary to fluorescence intensity, fluorescence lifetime is independent from fluorophore concentration. So mapping fluorescence lifetime of plant cell walls represents a complementary approach to acquire chemical and structural information of cell wall components and interactions. Key words Fluorescence, Lifetime, Microscopy, Plant, Lignocellulose, Lignin

1

Introduction Imaging of plant cell walls has benefited from important instrumental advances during the last decades, so that multiscale imaging using different microscopy modalities has become a reality. Among electron, photon, and atomic force microscopy techniques, fluorescence microscopy is widely used as a routine method. Indeed, the use of fluorescent probes such as antibodies or small organic molecules, having a more or less specific affinity for plant cell wall components, has revolutionized our understanding of plant cell wall structure and evolution [1]. Nonetheless, native autofluorescence of plant cell wall, which mainly originates from aromatic molecules contained in lignin [2, 3], is not as commonly used, for at least two reasons: lignin maximum excitation wavelength is around 400 nm or below and thus requires UV light sources which are quite expensive; interpretation of fluorescence intensity imaging is subtle and most of the time only qualitative. More recently, it has become possible to image not only fluorescence intensity of cell wall samples but also their fluorescence lifetime. The fluorescence lifetime of a fluorescent molecule (a fluorophore) is defined as the average duration in which it remains in its excited state [4]. When it returns to the ground

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_5, © Springer Science+Business Media, LLC, part of Springer Nature 2019

77

78

Christine Terryn and Gabriel Pae¨s

state, it follows an exponential decay, in the order of the nanosecond. Like fluorescence intensity, fluorescence lifetime is highly dependent on the fluorophore environment and its interactions, thus it can be a versatile marker. But fluorescence lifetime is also independent from fluorophore concentration [5], which facilitates comparison of sample lifetime properties. By acquiring fluorescence lifetime not only for a whole sample but for each image pixel, fluorescence lifetime imaging (FLIM) is definitely a relevant method to investigate distribution of lignin in plant cell walls [6–9] and evaluate its chemical state [10], depending on biomass species, biological, chemical, physical, or enzymatic modifications. Below is thus presented the protocol to acquire and to analyze fluorescence lifetime data and images of plant cell walls.

2 2.1

Materials Plant Samples

1. Plant samples: should be flat sections of 30 to 60 μm thickness prepared with a microtome, commonly obtained from plant straws or wood blocks. 2. Mounting buffer: aqueous buffered solution using ultrapure water (see Note 1). 3. Microscope slides: standard glass slides. 4. Microscope glass coverslips: should be 0.17 mm thickness or less.

2.2 Microscope and Detectors

1. Microscope: wide field or laser scanning microscope with pulsed source in UV (monophoton) or in infrared (multiphoton). In this protocol we use the following configuration: laser scanning microscope LSM 710 NLO (Zeiss) and infrared pulsed laser CHAMELEON Ti Sa (Coherent) accordable 80 MHz infrared pulsed laser which allows high resolution and sensitivity. 2. Fluorescence Lifetime detector: lifetime measurements can be made with phase-modulation system or Time-Correlated Single Photon Counting (TCSPC; ref. 5). Here, we focus on TCSPC acquisition setting. The acquisition is done using a MW-FLIM detector system along the SPC 150 photocounting card from Becker & Hickl. This system allows simultaneous acquisition of lifetime decay over 1024 temporal channels for each of the 256  256 pixels of the sample area. The acquisition card is synchronized with the laser then one laser pulse triggers the photon counting until the next pulse. Lifetime decay curve is drawn by accumulation during acquisition time (Fig. 1).

Fluorescence Lifetime Imaging

79

Fig. 1 FLIM acquisition setup. Incident pulse induces fluorophore switch from ground state to excited state. At the end of the pulse, fluorophore returns to ground state by emitting photons which are collected by TCSPC until the next pulse. This system requires that pulsed laser, microscope and photon counting are synchronized 2.3

3

Software

Acquisition is done with SPCm software and analysis with SPCImage 5.0 software, both from Becker & Hickl.

Methods

3.1 FLIM Measurements Acquisition Protocol

1. Put the slide on the microscope stage. 2. Adjust microscope settings to use a 750 nm excitation wavelength with biphoton laser and select corresponding filter set for emission (see Note 2). 3. Choose suitable sample area under a conventional imaging mode and select best Z focus (see Note 3). 4. Activate counting mode on the FLIM system and select acquisition duration of 30 s (see Notes 3 and 4). 5. Switch optical pathway toward FLIM detector. 6. Activate scanning mode of laser scanning microscope and trigger photon counting with the FLIM software. 7. When collecting time is over, stop scanning on the microscope and save fluorescence lifetime image (see Note 5).

3.2

FLIM Analysis

1. Open SPCimage software from Becker & Hickl and load fluorescence lifetime image (Fig. 2A). 2. Check that collected photon number is sufficient to do efficient fit (see Note 4). If no, adjust acquisition parameters. You can

80

Christine Terryn and Gabriel Pae¨s

Fig. 2 Lifetime imaging analysis of a plant sample section. Acquired intensity matrix (A) shows lifetime decay on each pixel (B). Optimize fit parameters to have less residuals and lower χ 2 (B). Saved parameters are used to calculate lifetime on each pixel. Global result shows lifetime distribution (in ps) for the whole image as a histogram (C). Lifetime image displays lifetime values as color-coded pixels (blue-shift: long lifetime, red-shift: short lifetime) on the sample (D)

also change “Bin” option to have more photons by binning pixels but consequently spatial resolution is reduced (Fig. 2B). 3. Select the fit model with “Incomplete Multiexponentials” with “Repetition Time” of 12.5 ns (see Note 6). 4. Move cursor to empty area of the sample to determine the noise and up the threshold parameter until this background is no more taken into account. 5. Move cursor to area with high intensity and check “Multiexponential Decay” box with two components (see Note 7). 6. Select “IRF auto” and change time cursor T1 or T2 to have minimum residuals of fit (see Note 8). 7. Save all fit parameters and apply them to all the samples of the experiment (see Notes 9 and 10). 8. Do “Calculate Decay matrix” (Fig. 2C). 9. Results show lifetime distribution for the whole image (Fig. 2C) and the lifetime image (Fig. 2D).

Fluorescence Lifetime Imaging

81

10. To perform a lifetime analysis on different samples, determine the maximum lifetime in the histogram and do for example statistical analysis. 11. Specific areas such as cell corners or middle lamella can be selected on the lifetime image so that specific local lifetimes can be measured.

4

Notes 1. Lifetime is highly sensitive to pH variations, so plant section samples must be placed in a buffered solution of known pH to avoid any variations when comparing measurements. 2. Laser power should be set at a value as low as possible to avoid photodamage of the sample and FLIM detector saturation. 3. High laser power and long acquisition time can cause local heating which can lead to sample defocusing. 4. To determine the best acquisition duration, make a compromise between the collected photon number per pixel (at least 1000, ref. 11 and the laser power (see Note 3). 5. Check yearly instrumental response function of your FLIM detector by acquiring FLIM measurement on hydroxyl-urea crystals on glass coverslip and by measuring Full Width at Half Maximum which depends on the system (here, it is 170 ps). 6. Laser frequency limits the temporal range of measured lifetime. Here, since an 80 MHz pulsed laser is used, only lifetime under 12.5 ns can be measured. 7. Autofluorescence of plant cell is complex signal so it cannot be fitted to a monoexponential decay function [3, 12]. 8. To determine the best fit of the lifetime decay with 2 or 3 exponentials, choose each of them and compare the adjustment parameter χ 2. If they are similar, choose 2 components, if not, choose the number of components with χ 2 closer to 1. 9. The fit model depends on the used FLIM system. The acquisition card here has specific counting mode with refolding in the first temporal channels, then “Incomplete Multiexponentials” is used to take into account the repetition time of the laser used to get a good fit. 10. For a set of studied samples, it is better to have the same time interval T1 and T2 to fit lifetime decay.

82

Christine Terryn and Gabriel Pae¨s

Acknowledgments This work was made possible through the funding from the French National Research Agency (LIGNOPROG project ANR-14CE05-0026). References 1. Pae¨s G (2014) Fluorescent probes for exploring plant cell wall deconstruction: a review. Molecules 19:9380–9402 2. Ralph J (2010) Hydroxycinnamates in lignification. Phytochem Rev 9:65–83 3. Donaldson L, Radotic´ K, Kalauzi A, Djikanovic´ D, Jeremic´ M (2010) Quantification of compression wood severity in tracheids of Pinus radiata D. Don using confocal fluorescence imaging and spectral deconvolution. J Struct Biol 169:106–115 4. Ishikawa-Ankerhold HC, Ankerhold R, Drummen GPC (2012) Advanced fluorescence microscopy techniques-FRAP, FLIP, FLAP, FRET and FLIM. Molecules 17:4047–4132 5. Becker W (2012) Fluorescence lifetime imaging - techniques and applications. J Microsc 247:119–136 6. Hafre´n J, Oosterveld-Hut HMJ (2009) Fluorescence lifetime imaging microscopy study of wood fibers. J Wood Sci 55:236–239 7. Coletta VC, Rezende CA, da Conceic¸˜aao FR, Polikarpov I, Guimara˜es FEG (2013) Mapping

the lignin distribution in pretreated sugarcane bagasse by confocal and fluorescence lifetime imaging microscopy. Biotechnol Biofuels 6:43 8. Donaldson LA, Radotic K (2013) Fluorescence lifetime imaging of lignin autofluorescence in normal and compression wood. J Microsc 251:178–187 9. Zeng Y, Zhao S, Wei H, Tucker MP, Himmel ME et al (2015) In situ micro-spectroscopic investigation of lignin in poplar cell walls pretreated by maleic acid. Biotechnol Biofuels 8:126 10. Auxenfans T, Terryn C, Pae¨s G (2017) Seeing biomass recalcitrance through fluorescence. Sci Rep 7:8838 11. Spriet C, Trinel D, Riquet F, Vandenbunder B, Usson Y et al (2008) Enhanced FRET contrast in lifetime imaging. Cytom Part A 73A:745–753 12. Harter K, Meixner AJ, Schleifenbaum F (2012) Spectro-microscopy of living plant cells. Mol Plant 5:14–26

Chapter 6 Raman Spectroscopy in Nonwoody Plants Dorota Borowska-Wykre˛t and Mateusz Dulski Abstract Confocal Raman spectroscopy (RS) enables obtaining molecular information from the nondestructive analysis of plant material in situ. It can thereby be a useful method to investigate spatial distribution and heterogeneity of cell-wall polymers. The authors’ intention is to present some examples of RS application and its capabilities for investigations of nonwoody plants. In this context, we present protocols for qualitative analysis of main polymers of plant wall and application of RS in a semiquantitative study of the arrangement of selected polymers in the wall in its native state. Key words Raman spectroscopy, Cell wall polymers, Raman imaging, Polarization

1

Introduction Plant cell walls are composed of interconnected networks of polymers. Some of them are polysaccharides, that is, cellulose, hemicellulose, or pectin, and one is a polymer of phenolic compounds— lignin. The structural and functional changes of plant cell walls are usually investigated by light microscopy (LM), transmission/scanning electron microscopy (TEM/SEM), confocal laser scanning microscopy (CLSM), fluorescence microscopy (FM), or atomic force microscopy (AFM) [1, 2]. Employment of classic histochemical methods or immunocytochemical analysis, as well as more advanced electron microscopy for the analysis of spatial arrangement of the cell-wall polymers, requires a special chemical treatment of samples (extraction, fixation, hydrolysis, staining) that could be more or less destructive. Furthermore, such approaches usually provide information about a selected wall component. Thus, a simultaneous analysis of chemical distribution, quantity or structural arrangement of the main wall building compounds is very difficult. A new approach developed in the last decade is Raman spectroscopy (RS) providing an opportunity to follow all of the polymers building the plant cell walls at the same time [3–5]. More

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_6, © Springer Science+Business Media, LLC, part of Springer Nature 2019

83

84

Dorota Borowska-Wykre˛t and Mateusz Dulski

specifically, RS acquires information about the molecular structure of various plant cell-wall polymers such as crystallinity, spatial arrangement or even tensile deformation or mechanical processing in the nearly native state of the wall [6–15]. The incident laser light used in the RS is usually polarized, creating a possibility to obtain information about the preferential orientation of the functional groups of cell-wall polymers. One of the most representative examples of the polarization-sensitive compound is cellulose that forms microfibrils. The intensity of Raman bands assigned to vibration of its functional groups (O-glycosidic linkage between glucopyranose rings, methine CH, or methylene CH2 bonds) strongly depends on the spatial orientation of the cellulose microfibril with respect to the polarization direction of the incident light [4, 16]. To sum up, Raman approach is helpful in the analysis of cell-wall polymers’ distribution, and orientation of some of the wall polymers. The latter requires data calibration relying on the single spectra measurements [17, 18]. 1.1

Basic Theory

In order to understand what is the origin of the Raman signal, it is crucial to understand the basic theory of the Raman effect. In the Raman approach, vibrational transitions that take place in the ground electronic state of molecules are considered. More precisely, photons of the laser light (of a monochromatic nature) interact with the sample molecule and distort (polarize) the cloud of electrons around the atom nuclei. Involvement of the electron cloud distortion in the process results in photon scattering, that is, the Raman effect. The effect comprises the elastic scattering (the so-called Rayleigh scattering) and two types of inelastic scattering (Stokes and anti-Stokes). In the Rayleigh scattering, which is the most intense, there is a lack or only very small change of the original photon excitation (the laser frequency; Fig. 1). This type of scattering is, however, not useful in the RS analysis, unlike the inelastic scattering, that is, changes of frequency (energy) of photons from the incident light upon interaction with a sample. In the inelastic scattering, the periodical deformation of a molecule results in a shift, up (Stokes effect) or down (anti-Stokes effect), of the frequency of scattered photons in comparison with the laser frequency (Fig. 1). The relative intensity of the two effects depends on the contribution of molecules in different states: the ground vibrational state or the excited vibrational state. For example, in the Stokes effect, which is the most useful in RS approach, the transition from the ground vibrational state leads to absorption of energy by the molecule and its promotion to the so-called virtual state of higher energy. The scattering from the virtual to excited state involves the energy transfer to the scattered photon. Because at a room temperature the contribution of molecules in the excited state is marginal, the strongest signal in RS originates from the Stokes effect [19].

Raman Spectroscopy in Nonwoody Plants

85

Fig. 1 Raman scattering processes. Line g—ground vibrational state, e—excited vibrational state, vs—virtual state of molecule created under laser interaction

Fig. 2 Water vibration modes. (a and b) stretching (ν), (c) bending (δ)

The number of possible vibrations in the molecule is 3N-6 where N is the number of atoms in a nonlinear molecule (3N-5 for a linear molecule), or the number of vibrational degrees of freedom. For example, in the case of water, only three vibrations are distinguished (Fig. 2). Some bands in the Raman spectrum are a manifestation of vibrations of individual functional groups. These vibrations can be classified, according their shape, as (Fig. 2): symmetrical stretching (νsym), asymmetrical stretching (νasym), and deformation (δ). This chapter presents the methods to create chemical images representing the distribution of various polysaccharides within the plant cell walls, as well as to obtain the calibration equations, necessary for calculations performed in the analysis of polymer orientation. A scarious involucral bract of Helichrysum bracteatum

86

Dorota Borowska-Wykre˛t and Mateusz Dulski

is used as a model of a nonwoody plant organ, which illustrates how to solve various ambiguities associated with the spatial arrangement of polysaccharides in native plant cell walls.

2

Materials

2.1 Samples, Reagents, Tools, and Consumables

1. Plant samples. We recommend sections, several dozen micrometers thick. 2. Distilled water, filter paper. 3. Microscope slides—CaF2—calcium fluoride recommended (only one Raman band, at 321 cm
1, is observed). 4. Glass coverslips (0.13–0.16 mm thick). Caution: collect the Raman spectra of coverslips at the beginning of the experiment. Special attention should be paid on Raman spectrum of some coverslips which may overlie the cellulose spectrum. Namely, regions: 200–750 cm
1 and 1050–1150 cm
1 are strongly affected by the bending and stretching of O–Si–O and Si–O–Si vibration of the silica glass. These regions, especially the latter one, are correlated with the C–O–C vibration of Oglycosidic linkages (two bands centered around 1095 and 1125 cm
1) in the cellulose-based systems. Our Raman measurements show that sodium silica glass is not recommended because of overlapping of silica and cellulose bands while a borosilicate glass produces an acceptable signal in the range of the cellulose bands due to the silica band shift toward lower wavenumber (Fig. 3). 5. Nail polish (any brand). 6. Razor blades, tweezers. 7. Scotch tape. 8. Dissecting needle—with holder equipped with an adjustable chuck for tightening or replacing the needles or microblades (Fig. 4). 9. Microblades—made of a fragment of a razor blade cut off from the blade. The fragment should form a triangle curved blade (Fig. 4).

2.2 Instrumentation and Software

We use the WITec confocal Raman microscope CRM alpha 300M with the equipment listed below. Before starting a work with a spectrometer, it is extremely important to optimize the system for a tested plant sample in order to achieve the highest transmission and efficiency of the spectrometer [20]. The optimization of the system concerns the optical fiber and objective selection as well as the CCD and the grating adjustment (see Notes 1–4).

Raman Spectroscopy in Nonwoody Plants

87

Fig. 3 Raman spectra of various glass coverslips (red and blue) and bacterial cellulose (black). Due to the overlap of silica band (blue) with cellulose signal (black) the soda–lime glass is not recommended for a Raman experiment

Fig. 4 Ensiform sample (scarious involucral bracts of Helichrysum bracteatum) immobilization and orientation with respect to coverslip for cross (a and c) and paradermal (b and d) sections. (c) A cross section of the bract in region o, (d) paradermal section of the periclinal epidermis (cuticle side faces the microscopic slide). A—long axis of the sample, B—orientation of the first coarse cutting, C—fine cutting orientation, AD—adaxial epidermis, AB—abaxial epidermis, P—parenchyma,  —the so-called hinge part, * outer periclinal wall of the adaxial epidermis (a single wall)

88

Dorota Borowska-Wykre˛t and Mateusz Dulski

1. An air-cooled solid-state laser (λ ¼ 532 nm—green line) coupled to spectrometer by a single-mode optical fiber. Caution: the use of a green line in some kind of samples may induce the fluorescence phenomena due to the presence of conjugated bonds within chlorophyll or provide sample decomposition by the usage of the inappropriate, that is, too high, laser power. The fluorescence effect may be reduced by long-time illumination of an interesting point, alteration of the objective to smaller magnification or modification of the diameter of the optical fiber. In turn, to avoid thermal decomposition we advise immersing the material, for example, in water (H2O, D2O). An alternative is to work at the power below 10 mW (for transparent samples) and 30 mW (for nontransparent samples) but the power should always be adapted to the individual sample. 2. Back-illuminated charge-coupled device (CCD) detector linked to the multimode optical fibers. 3. Dichroic filters, objectives (10/NA, 50/NA, 100/NA, etc.). 4. Calibration silicon plate oriented according to (111) plane. 5. WITecProjectFour (Plus) Software (Fig. 5), GRAMS, ORIGIN or FitTic.

3

Methods Raman spectroscopy (RS) involves sampling of an area of interest by directing a laser excitation beam and analyzing the collected molecularly specific scattered light. However, the first step in the experimental design is the sample preparation. The sampling method in RS can be somewhat modified in order to deal with the special properties of plants. Note that the procedure details of the next steps (including data acquisition, data analysis) of the Raman imaging depend on the type of microscope equipment and software. Additionally, some differences in data acquisition (and analysis) step occur depending on whether qualitative or quantitative studies are performed. Therefore, these steps are presented separately.

3.1 Sample Preparation

In some cases, embedding of samples prior to sectioning is necessary. PEG or resin is used as an embedding medium for tissue structure supporting and enables to cut appropriately thin and properly oriented sections [16, 21–23]. Proper sample geometry and adjustment with respect to laser polarization direction is necessary for investigation of cellulose arrangement. One of the advantages of PEG as an embedding agent is that after the sections are ready, the PEG can be completely removed by water rinsing.

Raman Spectroscopy in Nonwoody Plants

89

Fig. 5 WITecProjectFour (Plus) Software—right panel; toolbars are displayed on the left (more details in the text)

90

Dorota Borowska-Wykre˛t and Mateusz Dulski

1. When preparing samples by sectioning, the appropriate thickness of the section should be selected so that the region of interest can be recognized in a bright field or in the light reflected microscope. Considerations that should be taken into account are listed below (steps 2–6), the actual procedure of sample preparation is outlined in steps 7–17. 2. Thickness of specimen sections should not be lesser than the focal depth of the laser beam (i.e., 1 (so water, oil immersion is necessary). Remember that new objective has to be calibrated in the software (Add new objective ! Introduce a type of the objective—air, water, etc., its magnification and numerical aperture ! Calibrate). 4. The grating disperses the signal onto the CCD detector. Usually, the choice of the grating of 600 line/mm ensures a compromise between spectral resolution (i.e., the number of points reproducing the spectrum), experiment duration and the number of information obtained from the investigations. Moreover, it gives an opportunity to collect a full spectrum range from 0 to 3800 cm
1 at one time. In turn, application of grating with a higher number of lines/mm leads to increase of the dispersion and thus higher spectral resolution (because of distribution of the signal over a large number of pixels in the CCD detector). Unfortunately, this type of gratings reduces spectral range collected during the experiment. 5. Cuticle forms a film covering the epidermis of numerous aboveground plant organs. Because of its composition (insoluble cuticular membrane impregnated by and covered with soluble waxes), the cuticle is a water permeability barrier.

106

Dorota Borowska-Wykre˛t and Mateusz Dulski

Therefore, it is appropriate to cut out a piece from the intact organ to enable water to penetrate the whole sample as quickly as possible.

Acknowledgments This work was financially supported by the National Science Centre, Poland (ERA-CAPS CALL 2016; project No. 2017/24/Z/ NZ3/00548 and 2017/26/D/ST8/01117). We thank Dorota Kwiatkowska and Roman Wrzalik for discussion and comments. References 1. Ng JK, Schro¨der R, Brummell DA, Sutherland PW, Hallett IC et al (2015) Lower cell wall pectin solubilization and galactose loss during early fruit development in apple (Malus x domestica) cultivar ‘Scifresh’ are associated with the slower softening rate. J Plant Physiol 176:129–137 2. Zdunek A, Kozioł A, Pieczywek PM, Cybulska J (2014) Evaluation of the nanostructure of pectin, hemicellulose, and cellulose in the cell walls of pears of different texture and firmness. Food Bioproc Tech 7:3525–3535 3. Atalla RH, Ranua J, Malcolm EW (1984) Raman-spectroscopic studies of the structure of cellulose: a comparison of Kraft and sulfite pulps. TAPPI J 67:96–99 4. Atalla RH, Whitmore RE, Heimbach CJ (1980) Raman spectral evidence for molecular-orientation in native cellulosic fibres. Macromolecules 13:1717–1719 5. Wiley JH, Atalla RH (1987) Band assignments in the Raman-spectra of celluloses. Carbohydr Res 160:113–129 6. Edwards HGM, Farwell DW, Webster D (1997) FT Raman microscopy of untreated natural plant fibres. Spectrochim Acta A Mol Biomol Spectrosc 53:2383–2392 7. Himmelsbach DS, Khahili S, Akin DE (1999) Near-infrared–Fourier transform–Raman microspectroscopic imaging of flax stems. Vib Spectrosc 19:361–367 8. Eichhorn SJ, Sirichaisit J, Young RJ (2001) Deformation mechanisms in cellulose fibres, paper and wood. J Mater Sci 36:3129–3135 9. J€ahn A, Schro¨der MW, Fu¨ting M, Schenzel K, Diepenbrock W (2002) Characterization of alkali-treated flax fibres by means of FT Raman spectroscopy and environmental scanning electron microscopy. Spectrochim Acta A Mol Biomol Spectrosc 58:2271–2279

10. Morrison WH, Himmelsbach DS, Akin DE, Evans JD (2003) Chemical and spectroscopic analysis of lignin in isolated flax fibres. J Agric Food Chem 51:2565–2568 11. Fischer S, Schenzel K, Fischer K, Diepenbrock W (2005) Applications of FT Raman spectroscopy and micro spectroscopy characterizing cellulose and cellulosic biomaterials. Macromol Symp 223:41–56 12. Gierlinger N, Schwanninger M, Reinecke A, Burgert I (2006) Molecular changes during tensile deformation of single wood fibres followed by Raman microscopy. Biomacromolecules 7:2077–2081 13. Peetla P, Schenzel KC, Diepenbrock W (2006) Determination of mechanical strength properties of hemp fibres using near-infrared Fourier transform Raman microspectroscopy. Appl Spectrosc 60:682–691 14. Schenzel K, Almlof H, Germgard U (2009) Quantitative analysis of the transformation process of cellulose I to cellulose II using NIR FT Raman spectroscopy and chemometric methods. Cellulose 16:407–415 15. Atalla RH, Agarwal UP (1985) Raman microprobe evidence for lignin orientation in the cell walls of native woody tissue. Science 227:636–638 16. Gierlinger N, Luss S, Ko¨nig C, Konnerth J, Eder M et al (2010) Cellulose microfibril orientation of Picea abies and its variability at the micron-level determined by Raman imaging. J Exp Bot 61:587–595 17. Gierlinger N, Schwanninger M (2006) Chemical imaging of poplar wood cell walls by confocal Raman microscopy. Plant Physiol 140:1246–1254 18. Gierlinger N, Schwanninger M (2007) The potential of Raman microscopy and Raman

Raman Spectroscopy in Nonwoody Plants imaging in plant research: a review. Spectroscopy 21:69–89 19. Smith E, Dent G (2005) Modern Raman spectroscopy—a practical approach. John Wiley & Sons Ltd, Chichester 20. Hollricher O (2010) Raman instrumentation for confocal Raman microscopy. In: Dieing T, Hollricher O, Toporski J (eds) Confocal Raman microscopy, vol 2011. Springer, New York, pp 43–60 21. Wolosewick JJ (1980) The application of polyethylene-glycol (PEG) to electron microscopy. J Cell Biol 86:675–681 22. Schreiber N, Gierlinger N, Pu¨tz N, Fratzl P, Neinhuis C et al (2010) G-fibres in storage roots of Trifolium pratense (Fabaceae): tensile stress generators for contraction. Plant J 61:854–861 23. Gierlinger N, Reisecker C, Hild S, Gamsjaeger S (2013) Raman microscopy: insights into chemistry and structure of biological materials. In: Fratzl P, Dunlop JWC, Weinkamer R (eds) Materials design inspired by nature: function through inner architecture. Royal Society of Chemistry, London, pp 151–179 24. Griffith PR (2009) Infrared and Raman instrumentation for mapping and imaging. In: Salzer R, Siesler HW (eds) Infrared and Raman spectroscopic imaging. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, pp 3–64, 2014 25. Zeise I, Heiner Z, Holz S, Joester M, Bu¨ttner C et al (2018) Raman imaging of plant cell walls in sections of Cucumis sativus. Plants (Basel) 7:7 26. Schmidt M, Schwartzberg AM, Carroll A, Chaibang A, Adams PD et al (2010) Raman imaging of cell wall polymers in Arabidopsis thaliana. Biochem Biophys Res Commun 395:521–523 27. Piot O, Autran J-C, Manfait M (2001) Investigation by confocal Raman microspectroscopy of the molecular factors responsible for grain cohesion in the Triticum aestivum bread wheat. Role of the cell walls in the starchy endosperm. J Cereal Sci 34:191–205 28. Chylin´ska M, Szyman´ska-Chargot M, Zdunek A (2014) Imaging of polysaccharides in the tomato cell wall with Raman microspectroscopy. Plant Methods 10:14 29. Atalla RH, Agarwal UP (1986) Recording Raman-spectra from plant cell walls. J Raman Spectrosc 17:229–231 30. Gierlinger N, Keplinger T, Harrington M (2012) Imaging of plant cell walls by confocal Raman microscopy. Nat Protoc 7:1694–1708

107

31. Agarwal UP, Ralph SA (1997) FT-Raman spectroscopy of wood: identifying contributions of lignin and carbohydrate polymers in the spectrum of black spruce (Picea mariana). Appl Spectrosc 51:1648–1655 32. Agarwal UP (1999) An overview of Raman spectroscopy as applied to lignocellulosic materials. In: Argyropoulos DS (ed) Advances in lignocellulosics characterization. TAPPI Press, Atlanta, Georgia, pp 209–225 33. Borowska-Wykre˛t D, Rypien´ A, Dulski M, Grelowski M, Wrzalik R et al (2017) Gradient of structural traits drives hygroscopic movements of scarious bracts surrounding Helichrysum bracteatum capitulum. Ann Bot 119:1365–1383 34. Dieing T, Ibach W (2010) Software requirements and data analysis in confocal Raman microscopy. In: Dieing T, Hollricher O, Toporski J (eds) Confocal Raman microscopy, vol 2011. Springer, New York, pp 61–89 35. Savitzky A, Golay MJE (1964) Smoothing and differentiation of data by simplified least squares procedures. Anal Chem 36:1627–1639 36. Liland KH, Rukke EO, Olsen EF, Isaksson T (2011) Customized baseline correction. Chemom Intell Lab Syst 109:51–56 37. Schulze G, Jirasek A, Marcia ML, Lim A, Turner RF et al (2005) Investigation of selected baseline removal techniques as candidates for automated implementation. Appl Spectrosc 59:545–574 38. Prakash BD, Wei YC (2011) A fully automated iterative moving averaging (AIMA) technique for baseline correction. Analyst 136:3130–3135 39. de Juan A, Maeder M, Hancewicz T, Duponchel L, Tauler R (2009) Chemometric tools for image analysis. In: Salzer R, Siesler HW (eds) Infrared and Raman spectroscopic imaging. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, pp 65–108 40. Snyder R (1971) Raman scattering activities for partially oriented molecules. J Mol Spectrosc 37:353–365 41. Bower D (1972) Investigation of molecular orientation distributions by polarized Raman scattering and polarized fluorescence. J Polym Sci 10:2135–2153 42. Bremard C, Dhamelincourt P, Laureyns J, Turrell G (1985) The effect of high-numericalaperture objectives on polarization measurements in micro-Raman spectrometry. Appl Spectrosc 39:1036–1039

Chapter 7 Image Analysis: Basic Procedures for Description of Plant Structures Jana Albrechtova´, Zuzana Kubı´nova´, Alesˇ Soukup, and Jirˇı´ Jana´cˇek Abstract This chapter gives examples of basic procedures of quantification of plant structures with use of image analysis, which are commonly employed to describe differences among experimental treatments or phenotypes of plant material. Tasks are demonstrated with the use of ImageJ, a widely used public domain Java image processing program. Principles of sampling design based on systematic uniform random sampling for quantitative studies of anatomical parameters are given to obtain their unbiased estimations and simplified “rules of thumb” are presented. The basic procedures mentioned in the text are: (1) sampling, (2) calibration, (3) manual length measurement, (4) leaf surface area measurement, (5) estimation of particle density demonstrated on an example of stomatal density, and (6) analysis of epidermal cell shape. Key words Counting frame, Experimental design, Image analysis, Image calibration, Length estimation, Number estimation, Stereology, Stomatal density, Systematic uniform random sampling, Unbiased estimation

1

Introduction Image analysis (IA) has become a powerful tool for quantification of plant structures—macroscopic or microscopic ones. The quality of captured digital signal (macroscopic or microscopic image) is essential for further processing and quantitative analysis. Image spatial resolution is a number of picture elements (pixels, pxl in 2D) per length or area unit, which determines distinction of structures within the image. Resolution in color or gray scale (bit depth) is another important parameter. Standard 8-bit images allow for 256 (28) levels, while 16-bit images provide a more detailed scale of 65,536 levels to choose from. Colors of the image are most commonly encoded by three principal color channels—R (red), G (green), and B (blue) of the additive color mixing model (RGB) or HSB (or HSI or HSL) model, which uses channels hue, saturation, and brightness (or intensity or lightness) to specify colors. During color image processing it is, thus, possible to work with

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_7, © Springer Science+Business Media, LLC, part of Springer Nature 2019

109

110

Jana Albrechtova´ et al.

individual channels of RGB or HSB containing different information. Objects in the 2D plane perpendicular to the objective axis could be acquired without distortion (see Note 1) and their physical properties might be analyzed, if the size of the image pixels is known. Dimensions of pixels are determined during calibration. Selection of a proper image format affects quality of the image data. Some formats (e.g., JPG) use lossy compression and might cause some data degradation. That is why lossless formats (e.g., TIFF) are far more suitable for primary data storage. The images intended for image analysis should have sufficient both image resolution and bit depth. The size of the pixel shall be at least one half of the least requested detail in the image as defined by the Nyquist sampling theorem. Higher bit depth (12 bits and more) enables more accurate image processing and segmentation into objects for image analysis. Image preprocessing can be used to correct some deficiencies (background gradients, dust, etc.) and to improve visual image quality. Unlike human the computer does not identify the object of interest, so image segmentation should be used to separate those from background. The most common and simplest method of segmentation is thresholding, which results in a binary image (binary values per pixel; 0/1; black/white) definitions of the objects. Subsequent adjustments of a binary image and borders of the segmented objects based on mathematic morphology—operations of dilation, erosion, closing, and opening available in menu of any IA software—are used to improve object representation. Objects, once defined, can be measured and classified using various geometrical parameters (see Note 2). Subset of objects might be selected in specified region of interest (ROI) with superimposed mask or a counting frame [1–3]. The majority of plant structures exhibits gradients [4] in quantitative anatomical parameters, reflecting polarity of plant structures established during early plant ontogenetic development and organogenesis. A known example is the dependence of some anatomical parameters of a leaf on its insertion—that is, distance from the root system. Because of the spatial heterogeneity of values of anatomical parameters in biological structures, it might be difficult to specify average values of particular anatomical parameters within an organ or tissue. Proper sampling designs based on sound theoretical principles (e.g., ref. 2) are necessary to be applied to gain reproducible values. Unbiased estimation of a measured parameter can be achieved, for example, by systematic uniform random (SUR) sampling, which ensures the prerequisite of unbiased estimation, that is, that each particle or object has the same probability to be selected by the sampling system or rule. This is a very important issue, often neglected in formation of a sampling design of biological studies. Quite often, the sampling design and the

Basics of 2D Image Analysis

111

method of quantification are not fully described, and thus the results can be difficult to compare and discuss with other results as a consequence of biological treatment or processes. Stereological methods might be also employed to analyze an image. Stereological methods provide 3D characteristics of objects based on measurement of parameters in 2D or 3D image, which is done by application of some test system (composed of points, lines, quarter-circles, cycloids, etc.) and counting the intersections of the structure under study and parts of a test system (e.g., refs. 2, 3). Stereological methods along with uniform random sampling can be very efficiently used to get unbiased estimations of quantitative anatomical parameters of plant organs. For more details on principles of stereological method application see the source studies [5–11]. In this chapter, several examples of basic procedures for quantification of plant structures are presented. We focused on parameters, which can be quantified from 2D images: length, surface area in 2D, particle density of particles laying in one plain, and analysis of a shape of planar objects. The very important rules for obtaining reproducible IA results are the following: (1) object sampling should be done on principles to achieve unbiased estimation, for example on the base of uniform random sampling, (2) right calibration has to be set to obtain results in right units and not only in pixels, (3) proper statistical rules and methods have to be applied during result processing and interpretation, and (4) all used methods including sampling have to be thoroughly described with enough details to enable reproduction or comparison of obtained results by other researchers.

2

Materials There are many commercial image analysis programs, which can be successfully used for the following tasks. All examples in this chapter are described on the public domain program ImageJ [12]. The ImageJ program can be downloaded from http://imagej.nih.gov/ij/, which also provides a detailed installation manual, documentation, user manual and many plug-in modules for various tasks. The program is based on Java and runs on Windows, Mac OS, Mac OS X, and Linux. Its source code is freely available and anyone can program further plug-ins to be added into the program (see Notes 3 and 4). The program window consists of three lines: Menu Commands (“Menu” in the subsequent text), Toolbar (“Tools” in the subsequent text), and Status and Progress Bar (Fig. 1).

112

Jana Albrechtova´ et al.

Fig. 1 ImageJ window with menu bars: upper bar Menu Commands (“Menu” in the text), middle Toolbar (“Tools” in the text), and lower Status and Progress Bar

3 3.1

Methods Sampling

To get unbiased estimation of quantitative parameters of studied objects, it is necessary to sample the specimens in an unbiased way, so that each item has the same probability to be sampled. The more heterogeneous the structure is, the more segments, samples and sampling windows are needed [13]. The aim of a good experimental design is to quantify natural variability of data due to biological reasons (treatment, or biological process). To characterize natural variability of data, the other, artificial sources of the error have to be minimalized—such as an error due to wrong sampling and a design experimental error (observer, devices, etc.) or sample processing (see Note 1). The unbiased sampling principles such as SUR applied in a design lead to unbiased estimations of studied structural parameters. Answer the following questions and modify the design based on your answers: 1. What and where you have to sample? Are there known gradients of the parameter you are going to measure? Sample the same/comparable parts of plants to exclude natural variation in anatomical gradients (e.g., a leaf of the same insertion). 2. Do you need the average value for the whole plant/organ/ tissues? In that case you have to apply unbiased sampling (e.g., the SUR principle for sampling). Then sampled segments of an organ have to be further processed to get sections, to which any measurement procedure is applied. For example, for study on leaves you have to sample several segments in SUR way—see steps 4 and 5 below. 3. Are you interested just in comparison of some specific parts of the organ/plant? For example, for more extensive study on leaves you can compromise to comparison of a well defined part within the leaf blade (e.g., the middle of the coniferous needle). 4. Principle of SUR sampling for elongated plant structures (e.g., coniferous needles): Sample the positions of segments along

Basics of 2D Image Analysis

113

longitudinal organ axis. Determine the interval (distance T) between the segments, in which segments will be sampled along the organ. The value of T can be derived from the average length of the axis, divided by 5–10 (you want to sample between 5 and 10 segments). Get a random number from the set {0, 1, 2, . . ., T
1} from a table of random numbers or from some generator (e.g., http://www.random.org) to determine the position of first segment, that is, z (for a needle in mm). Cut the initial segment at distance z, next segment at z + T, then at z + 2 T etc. (for details see refs. 7–10). 5. Principle of SUR sampling for planar structures, such as a leaf blade of broadleaved trees or many dicot herbs. Sample the positions in a flat (2D) organ in a SUR way. Get random numbers for coordinates x, y from a table of random numbers or from some generator (see above)—for the horizontal coordinate x from the set {0, 1, 2, . . ., a
1} and for the vertical coordinate y from the set {0, 1, 2, . . ., b
1}, where a is a distance in x axis direction and b is a distance in y axis direction to determine the position of the initial point of the superimposed point grid (square or rectangular), the points of which denote the same corner of a sampled segment (for details see refs. 5, 7–9). 6. Apply the principle of SUR sampling design on all hierarchical levels of your design, that is, segments, sections, sampling windows, or measured objects (e.g., refs. 2, 7, 8). 7. Calculate variability from the results obtained in a pilot study. The convenient measure of how precise the estimate is called the coefficient of error, or CE. The coefficient of error is a statistical value used extensively in the stereological literature (defined as the standard deviation divided by a mean of the sample). Coefficient of error depends on the number of samples and corresponds to proportional variability of the estimate, its value should be lower than 0.05. If CE is higher, modify the sampling design by adding more segments, sections, sampling windows, or measured objects (e.g., refs. 2, 13). 8. In the majority of cases it is possible to apply just few basic rules, so called “rules of thumb”: (1) Five individuals per group are usually enough for the parameter estimation [14]; (2) when sampling sections along the longitudinal axis, cover the whole object systematically, in such a manner, that 5–10 segments or sections per organ are sampled; (3) when using sampling windows, 5–10 of them should be superimposed on each section; and (4) in most cases it is not necessary to count more than 200 points or intersections with stereological probe per organ in each compartment of interest [13].

114

Jana Albrechtova´ et al.

3.2 Calibration: Manual Setting of Microscopic Image Calibration

Correct calibration is a must to measure absolute values of a measured parameter in real units not jus in pixels. The calibration is specific for particular acquisition settings. Objective magnification, tubus factor, camera model (size and resolution of a chip), and settings of pixel binning on camera are particularly important while acquiring the image with microscope. Calibrate the pixel size with a stage micrometer or reticle of known dimensions (commercially available) as follows: 1. Acquire the image of a stage micrometer or reticle with a particular optical setup (objective and other components affecting magnification). 2. Open the image of a stage micrometer acquired with the same objective (and other acquisition settings such as an image size in pixels and zoom) as the image (or images) you would like to analyze. For the instructions to manual setting of known calibration (see Note 5). 3. Draw a line (Tools/Straight line) along the side of a line whose size is known. Consider the thickness of the lines of the stage micrometer—include it only once (Fig. 2). 4. Fill in the known length and units of length of the drawn line in the Set Scale window (Menu/Analyze/Set Scale; see Note 5).

Fig. 2 Calibrating square of known length of a square side. Yellow line corresponds to a measurement of a square length to determine a square size

Basics of 2D Image Analysis

115

If you tick “Global,” the calibration is saved and used for all subsequently opened images until the calibration is changed again. 5. Check the calibration by measuring the same line (Menu/ Analyze/Measure). 3.3 Manual Length Measurement

1. Open your image and set right calibration (see Subheading 3.2; Note 5). 2. For objects under study draw curves using the Tools, most often the Straight line tool (Tools/Straight line), possibly other convenient tools (Tools/Oval or Rectangular), trace the boundaries of the object (Tools/Polygon selection or Freehand selection) and measure their lengths. The length of the line appears in the dialog box. 3. To record your measurements, use the command Measure (Menu/Analyze/Measure) after every measurement. The results will appear in the Results window. 4. Save the acquired results (File/Save as in the Results window).

3.4 Measurement of Leaf Surface Area and Other Geometric Characteristics of the Leaf

1. Open your image of scanned leaves (or other objects, whose area you want to determine) and set right calibration (see Note 5). 2. Threshold the image (Menu/Image/Adjust/Threshold; Ctrl + Shift + T) to specify area of leaves within the image. Depending on the color of the background tick or uncheck “Dark background” in the dialog box. The threshold defines brightness values by the ruler under the histogram. Automatic segmentation is offered. Several values can be sampled from the image and used for segmentation. At the end, the leaves shall be masked with color, while the background shall be unchanged. When you have set the threshold, press “Select” and then “Sample” to get the binary image. 3. Specify which geometrical characteristics you intend to measure (Analyze/Set Measurements). For their explanation, see the manual parameter description. This applies particularly, for example, to the parameter of circularity, formula of which can vary in different image analysis software. 4. To exclude thresholded tiny objects in the background, set a minimal area of a measured object (Analyze/Set Measurements). 5. Measure the objects by command Analyze Particles (Menu/ Analyze/Analyze Particles) what will generate more geometrical parameters of measured objects, such as a perimeter. 6. Save the acquired results (File/Save as in the Results window).

116

Jana Albrechtova´ et al.

3.5 Estimation of Stomatal Density

1. Install the Sampling_Window plug-in to ImageJ: Download the file “Sampling_Window.class” from http://rsbweb.nih. gov/ij/plugins/sampling-window/index.html and copy it to the folder (Program files\...) ImageJ\plugins. Restart ImageJ. 2. Open your image of an epidermal peel, which was sampled in a uniform random way, and set right calibration. 3. Superimpose the counting frame on the image (Menu/Plugins/Sampling_Window). Set the size, color, line width, and position of the frame in the dialog box. Note the size of the frame and calculate its area. 4. Count the stomata (see Note 6), which are selected by the counting frame according to the following rules: stomata fully inside, stomata lying partly inside and intersecting the dashed line of the sampling frame. Do not count stomata which lie partly inside and simultaneously intersect full exclusion lines of the frame (Fig. 3). You can use the Multipoint selection tool (Tools/Multipoint selection tool) to mark the counted stomata by a point, which are then numbered. You can undo the selection by Alt + left click. 5. Relate the recorded number of stomata to the area of the counting frame and get the estimation of stomatal density.

3.6 Analysis of Epidermal Cell Shape

1. Install the Sampling_Window plug in to ImageJ (see Subheading 3.5, step 1). 2. Open your image of an epidermal peel, which was sampled in a uniform random way and acquired in such way to get a high contrast of cell walls (e.g., using cell wall polyphenolic fluorescence or stained cell walls with toluidine blue). Set right calibration. 3. If your image is a multichannel image (e.g., color image), split the channels (Image/Color/Split Channels) and for further analysis choose the most contrasting channel. 4. Smooth the image (Process/Smooth). 5. Enhance the contrast (Process/Enhance contrast). 6. Set the threshold (Image/Adjust/Threshold). 7. Dilate the image (Menu/Process/Binary/Dilate). 8. If needed, edit the image manually with the pencil or paint brush tool (More Tools/Drawing Tools). 9. Superimpose the sampling window on the image (Menu/Plugins/Sampling_Window) at random position. Set the size, color, line width, and position of the frame in the dialog box. 10. Measurement: Set the parameters for analysis in Menu/Analyze/Set measurements (e.g., Area, Shape descriptors, Perimeter, Feret’s diameter). Specify the minimal size of analyzed

Basics of 2D Image Analysis

117

Fig. 3 Analysis of epidermis—estimation of stomatal density and analysis of epidermal cell shape: (a) Original image of epidermis acquired by fluorescence microscope. (b) Red channel separated from the original image (Menu/Image/Color/Split Channels). (c) Sampling window superimposed on the image. Three stomata are manually selected, because they are lying at least partly in the frame and are not intersected by the (fulldrawn) exclusion line. (d) Sampling window superimposed on the adjusted image. Cells number 5, 12, 16–20, 22, 24, 26–28, 30, and 32 are selected, because they do not intersect the frame rectangle (or are inside the rectangle) and simultaneously do not intersect the exclusion lines. Results for each cell appear in the Results table

particles in Menu/Analyze/Analyze Particles (e.g., 5-Infinity), tick Display results, Clear results, and Add to ROI manager. 11. Save results for the cells selected by the counting frame—You can browse from a cell to cell in the ROI manager window to see which cell corresponds to which number. Other possibility

118

Jana Albrechtova´ et al.

is to measure the cells selected by the counting frame by the Wand (tracing) tool and command Ctrl + M (or Analyze/ Measure). The selected cells are those which intersect the frame rectangle (or are inside the rectangle) and simultaneously do not intersect the exclusion lines (Fig. 3). 12. Calculate the shape complexity of the object under study (ratio of its perimeter to square root of its area: b/√a, where b is perimeter and a is area of the object, [13]. b/√a is minimal for circle: 2π/√π ¼ 3.54, for highly structured objects, such as endoplasmic reticulum, it can exceed the value of 30.

4

Notes 1. The real object size could change in a consequence of specimen processing. Therefore, possible size artifacts (e.g., in structure shrinking, collapsing, and distortion) should be identified to get proper information about real dimensions (see for example ref. 5). 2. Some parameters, such as circularity, can be defined in different ways—thus, it is recommended to check the formula in the program manual. 3. Sequence of commands can be recorded as macro, saved to text file, and repeatedly run. This can speed up routine analyses of many images substantially. Saving the image analysis steps in macro is also useful for documentation purposes. This technique is accessible also to nonexperts as a sequence can be in most programs “recorded” during work. 4. The new plug-ins can be written in JAVA using ImageJ application interface. Extensive documentation as well as a lot of source files of plug-ins published on ImageJ website can be used for study of programming the image analysis modules. Some programming experience and knowledge of JAVA and ImageJ API is necessary. 5. Open the image you would like to analyze. Fill in the known information in the Set Scale window (Menu/Analyze/Set Scale). If the known scanner resolution is in pixels per inch and you need your results in mm, consider the unit transfer (1 in. ¼ 25.4 mm). The pixels are commonly square shaped but some chips have rectangular sensor elements. In case of nonsquare rectangular pixels, fill in proper value of the pixel aspect ratio (width/height). The ImageJ works with decimal point, not with decimal comma. If you tick “Global,” the calibration is saved and used for all subsequently opened images until the calibration is changed again. You can also fill the calibration in

Basics of 2D Image Analysis

119

the Image Properties (Menu/Image/Properties). Check the calibration by measuring a line length (Menu/Analyze/ Measure). 6. The same method can be used for estimation of a density of any object in 2D, such as epidermal cells. The particles must lie in one plane, otherwise it is not possible to estimate their density properly from 2D image. Detailed methodology of estimation of the number and sizes of stomata is described in the literature [8].

Acknowledgments The authors wish to acknowledge the funding by the projects GACR P501/10/0340, 18-23702S, by the Charles University in Prague, project SVV 265203, and by the Ministry of Education, Youth and Sports of the Czech Republic, project NPUI LO1417 and institutional support RVO:67985823. We thank Miroslav Barta´k for his technical help with graphical image processing. References 1. Sterio DC (1984) The unbiased estimation of number and sizes of arbitrary particles using the disector. J Microsc 134:127–136 2. Howard CV, Reed MG (1998) Unbiased stereology. BIOS, Oxford 3. Weibel ER (1979) Stereological methods, Vol 1: practical methods for biological morphometry. Academic, London 4. Pazourek J (1966) Anatomical gradients. Acta Univ Carol Biol Suppl 1/2:19 5. Lhota´kova´ Z, Albrechtova´ J, Jana´cˇek J, Kubı´nova´ L (2008) Advantages and pitfalls of using free-hand sections of frozen needles for threedimensional analysis of mesophyll by stereology and confocal microscopy. J Microsc 232:56–63 6. Albrechtova´ J, Kubı´nova´ L (1991) Quantitative analysis of the structure of etiolated barley leaf using stereological methods. J Exp Bot 42:1311–1314 7. Kubı´nova´ L (1993) Recent stereological methods for the measurement of leaf anatomical characteristics: estimation of volume density, volume and surface area. J Exp Bot 44:165–173 8. Kubı´nova´ L (1994) Recent stereological methods for measuring leaf anatomical characteristics: estimation of the number and sizes of stomata and mesophyll cells. J Exp Bot 45:119–127

9. Albrechtova´ J, Jana´cˇek J, Lhota´kova´ Z, Radochova´ B, Kubı´nova´ L (2007) Novel efficient methods for measuring mesophyll anatomical characteristics from fresh thick sections using stereology and confocal microscopy: application on acid rain-treated Norway spruce needles. J Exp Bot 58:1451–1461 10. Lhota´kova´ Z, Urban O, Duba´nkova´ M, Cvikrova´ M, Toma´sˇkova´ I et al (2012) The impact of long-term CO2 enrichment on sun and shade needles of Norway spruce (Picea abies): photosynthetic performance, needle anatomy and phenolics accumulation. Plant Sci 188:60–70 11. Kubı´nova´ Z, Jana´cˇek J, Lhota´kova´ Z, Kubı´nova´ L, Albrechtova´ J (2013) Unbiased estimation of chloroplast number in mesophyll cells: advantage of a genuine three-dimensional approach. J Exp Bot 65:609–620 12. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH image to ImageJ: 25 years of image analysis. Nat Methods 9:671–675 13. Gundersen HJG, Jensen EB (1987) The efficiency of systematic sampling in stereology and its prediction. J Microsc 147:229–263 14. Cruz-Orive LM, Weibel ER (1990) Recent stereological methods for cell biology: a brief survey. Am J Phys 258:L148–L156

Chapter 8 From Data to Illustrations: Common (Free) Tools for Proper Image Data Handling and Processing Fatima Cvrcˇkova´ Abstract Studying morphogenesis is unthinkable without visualizing shapes, and sharing the results of such studies critically depends on communicating image data. Despite a wealth of literature dealing with acquisition and analysis of image data, visualizing them for publication or presentation purposes remains a craft learned mainly by experience. This chapter provides a practical guide to producing publication-grade illustrations out of raw microscopic (or other) digital images, using mostly or exclusively free software, and points out some common problems and their solutions. Key words Computer graphics, Data visualization, Microscopy, Bitmap, Raster, Vector, Resolution, Bit depth, Inkscape, ImageJ

1

Introduction Anyone studying plant cell morphogenesis will, sooner or later, wish to share their research results in a publication. Long gone are times when institutions provided professional photography and graphics services, and a contemporary scientist has to produce publication-ready (i.e., professional quality) illustrations while being everything but a professional illustrator. Luckily, digital technologies provide aid with this task. Moreover, as a cell biologist interested in morphology and morphogenesis, our model scientist undoubtedly has above-average experience with computers (as well as a rather terrifying volume of digital photo data to produce illustrations from)—and only needs to choose appropriate software tools and learn to use them. This means deciding between either using commercial software or employing (in part or exclusively) freeware tools. Besides of valid ideological considerations [1], the freeware approach has at least two distinct practical advantages, even taking into account the highly evolved capabilities of commercial programs, which,

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_8, © Springer Science+Business Media, LLC, part of Springer Nature 2019

121

122

Fatima Cvrcˇkova´

however, are being rapidly matched also by free, often open source, software [2]. First, commercial software is costly, and thus often available to a particular investigator only through an institutional license on an institutionally owned computer. This not only might restrict the researchers’ possibilities of working from home but also poses problems when engaging students (in practice, the first experience with preparing scientific illustrations is often acquired while writing an undergraduate essay or a M.Sc. thesis). Since even special “education discount” licenses are usually administered through an institution with the abovementioned restrictions, students may face an unpleasant choice between either working long hours on a shared office computer or acquiring software by ethically dubious means. This problem simply ceases to exist when moving to free software. Second, commercial software is usually purchased through institutional multilicenses, leading to institution-wide preferences. On the other hand, science itself is collaborative, and publications (including their illustrations) may be coauthored by scientists from institutions using incompatible programs. While switching to freeware does not remove the need to learn unfamiliar procedures, it at least puts everyone on equal footing and eliminates the cost barrier. This chapter focuses on generating scientific illustrations, which typically combine microscope- or camera-generated photos and line graphics, using free tools, in particular the powerful Inkscape graphics editor. This program, undoubtedly also due to its open-source nature, gains currently on popularity in the scientific community, and specialized Inkscape-based or Inkscapecompatible solutions are springing up in areas ranging from molecular graphics [3] to utilities for continuous updating of graphs and diagrams when source data are modified [4]. Some of the tips and tricks provided below can also be, with some modifications, applied in nonfree software environments. Basic experience with image editing (either in Inkscape or in a commercial program such as Corel Draw or Adobe Illustrator) is assumed. Readers without such experience may benefit from Inkscape tutorials available at https://inkscape.org/cs/learn/tutorials/. Out of the scope of the present chapter, but also highly relevant, are issues concerning standards of good practice in data analysis and presentation. Good guidelines as to what is, and what is not, considered appropriate with regard to photo editing, as well as to primary data handling and archiving, can be found in refs. 5, 6. 1.1 Image Types and Formats

Digital images can, in principle, be stored in either bitmap or vector formats [7]. Bitmap, or raster, images are described by parameters of individual pixels arranged in a grid of finite resolution, with each pixel defined by its coordinates within the image and by a “bit depth” parameter characterizing its shade of color on a predefined scale

Image Data Handling and Processing

123

Fig. 1 The difference between bitmap and vector images becomes obvious at increased magnification

that may contain as few as two values (for a black and white image) or as many as 4,294,967,296 colors (for a 32-bit image). Bitmap file size at a given resolution (in terms of both pixel density, expressed usually in dots per inch, or dpi, and bit depth) is proportional to the image area. Scaling a bitmap image without adding or removing pixels leads to changes in resolution. A telltale sign that an image is a bitmap is the apparent loss of resolution if the image is sufficiently enlarged (Fig. 1). Photos are always stored in bitmap formats. Unlike bitmaps, vector images do not lose resolution when enlarged, because all image elements are precisely defined by parameters characterizing their outlines, as well as fills within these outlines (which may be nonuniform). Thus, the definition of any element in a vector image can be viewed as infinitely precise. Upon scaling such an image, file size remains constant. Vector formats are typically used for storing drawings, graphs, or diagrams. Specific file formats are commonly used for each type of images (Table 1). Bitmap image files tend to be large. To limit file size and memory usage, various compression methods are commonly applied. Certain compression algorithms always lead to some irreversible degradation of image resolution and quality (the so-called lossy compression, exemplified by the commonly used *.jpg image format), while others maintain image quality (lossless compression, as in the LZW compression algorithm). Avoid lossy compression when working with image data. Bitmap size can be altered also by resampling, that is, changing the number of pixels the image consists of. Like lossy compression, also resampling is irreversible, causes some degradation of image quality and should only be used carefully when downsizing very large files (Fig. 2). Scientific illustrations usually contain both annotated bitmap images (such as microscope-generated images or gel photos) and graphics (e.g., graphs, diagrams, or schematic drawings). Such a figure can, in principle, be produced either by drawing the graphics

124

Fatima Cvrcˇkova´

Table 1 Examples of common file formats for storing bitmap- and vector-type image data File types

Characteristics

Bitmap bmp, gif, jpg, jp2, png, tiff cpt (Corel Photo Paint), psd (Adobe Photoshop), xcf (Gimp)

cr2 (Canon), nef (Nikon) czi (Zeiss), lei (Leica), lif (Leica), lim (Laboratory Imaging/Nikon), nd2 (Nikon)

General-purpose bitmap image formats. Proprietary or internal bitmap formats used by image-editing programs (in parentheses). Files usually contain additional information and may, for example, consist of several layers. Formats produced by camera software, containing bitmaps and additional information. Formats produced by advanced microscopy software, containing bitmaps and additional information.

Vector emf, eps, pdf, svg, wmf

General-purpose vector formats which may differ in availability of some functions such as font embedding or object rotation. ai (Adobe Illustrator), cdr (Corel Draw), Proprietary vector formats used by image-editing programs.

Fig. 2 Image degradation by lossy compression (top) and by excessive resampling (bottom). Inkscape logo is used as an example

elements into a bitmap image and storing them subsequently in a bitmap format, or by embedding bitmaps into a vector image. The later approach has the advantage of preserving the infinite resolution of vector drawing for the vector parts of the illustration, and should thus be preferred. The method shown below aims toward producing such vector images with embedded bitmaps.

Image Data Handling and Processing

2

125

Materials

2.1 Hardware and Operation System

Publication quality images can be generated on any reasonably upto-date desktop or even notebook computer that is fit for routine office work. Given the rapid development of technology, any specific advice regarding processor type, disk size, or memory capacity would probably become obsolete even before this book reaches print. 1. If purchasing new equipment, consider a desktop computer optimized for gaming, rather than office use, because such machines are typically equipped with better graphics adaptor and larger memory capacity than typical office ones. These features also will come handy for purposes such as image data analysis or bioinformatic studies (see Note 1). 2. While not necessary, a second screen set in the “extended desktop” mode is helpful when keeping multiple files and programs open at the same time, especially if using a notebook with a small screen. 3. Make sure to have enough disk space. Keep a dedicated backup space on another device (e.g., a server connected through a network, or at least an external disk, preferentially connected via an USB3.0 or faster interface), and back up your data frequently. 4. While most personal computers now run the Microsoft Windows operating system, the choice of OS is very much a matter of personal (or institutional) preferences. Unless stated otherwise, all programs discussed in this chapter are free (though not always open source) and exist in versions compatible with recent versions of Windows, Mac, and Linux, and thus the instructions below can be considered OS-independent. Some of the procedures described below may require administrator rights on your computer.

2.2

Software

1. A table calculator, such as MS Excel (Microsoft, nonfree, only available for Windows and Mac) or the OS-independent Libre Office Calc (https://www.libreoffice.org/) that will be used to generate graphs and diagrams from quantitative data (see Note 2). This is usually installed as a part of an office software package that also contains a text editor. 2. The Fiji distribution of the free image analysis package ImageJ ([8, 9]; http://fiji.sc) for visualizing and analyzing raw microscope data and performing tasks such as format conversion and inserting scale bars. 3. A simple, fast image viewer such as IrfanView (https://www. irfanview.com/) for quick inspection of images.

126

Fatima Cvrcˇkova´

4. An up-to-date version of Inkscape (https://inkscape.org/)— this will be the program used for most of the work. 5. An up-to-date version of Gimp (https://www.gimp.org/), a graphics editor that will mainly be used for format conversion purposes (see Note 3). 6. A free PDF generator, such as the PrimoPDF virtual printer (http://www.primopdf.com/), while not essential, may be useful in some situations. 2.3

Source Data

1. For bitmaps, always start with an excess of source material (be it raw data files generated by a camera or microscope, or images processed in the course of previous analysis such as Z projections, color overlays, 3D reconstructions, and similar) to be able to choose figures that are both representative and of sufficient quality (see Note 4). The data may be placed in any location accessible to your OS. 2. If generating graphs or diagrams from data processed by a table calculator of your choice, perform the necessary calculations and generate source diagrams within the calculator. Perform any necessary adjustments such as adding error bars or setting an appropriate axis range (see Note 5), but do not optimize colors, data point markers, line thickness, etc., since this can be done more efficiently later on. If working in MS Excel, export the graphs from your calculator in the *.pdf format, if using LibreOffice Calc, save the *.ods file containing the graph. 3. Graphs and diagrams also may be generated by other programs or by online utilities. In such a case, save or export the graphics of interest in a vector format, if possible. Sometimes this might be achieved by printing to PrimoPDF If you have no other choice, export the graphics in a bitmap format—at the very worst, a screenshot will have to suffice. 4. Preexisting illustrations generated using commercial software such as Corel Draw or Adobe Illustrator can be incorporated either by direct import into Inkscape, or after conversion into an Inkscape-compatible format in the source program.

3

Methods

3.1 Processing Bitmap-Type Data

1. Inspect the source data using a suitable viewer and select which files contain the representative, or typical, images you want to show (see ref. 10 for in-depth discussion what this may mean). For standard bitmap formats such as *.jpg or *tiff, IrfanView is usually sufficient. If the files are in another format (especially generated by a microscope or camera), or do not open in IrfanView, try Fiji. Bitmap files with greater bit depth (16 bit

Image Data Handling and Processing

127

or more) sometimes open in Irfan View but look black or nearblack; in this case also use Fiji (see Note 6). 2. Based on results of this inspection, copy all files you are considering for further use to a local working directory on your computer. Original files should never be altered, all subsequent editing will be done on the working copy. 3. For microscope-generated files, use Fiji to perform any required processing such as artificial coloring (by choosing one of the preset look-up tables), multichannel overlay, Zprojection, or particle tracing. You may also adjust contrast, brightness, etc. Such modifications should always be applied to the whole image and need to be documented in figure legend or methods of your publication (see Note 7). Do not use ImageJ to annotate the photo—that will be done later during final image assembly. When happy with the appearance of the photo, convert it to 8-bit or RGB using the Image—Type command and save the image as *.tif (File—Save As). 4. If the source file does not contain internal calibration data ((e.g., in the case of photos taken using a general-purpose digital camera attached to a microscope via an aftermarket adaptor), set scale size manually based on an object of known size (e.g., an objective micrometer scale) photographed under the same conditions. Open the calibration object-containing file, draw a line along a known distance and set scale using the Analyze—Set Scale tool. Check the “Global” box to apply the scale to all subsequently opened files (until restart of Fiji or new scale setting). If the source file contains embedded calibration data (as usual in newer microscope-generated files), skip this step. 5. Insert a scale bar as an overlay using the Analyze—Tools—Scale Bar command (make sure that the “Overlay” field is checked before clicking OK). You can adjust the length of the scale bar and position it at one of the four corners of the image. Save the image as *.tif (File—Save As); you might replace the file generated in step 3 to limit cluttering your disk. Then go back to Analyze—Tools—Scale Bar, uncheck the “Overlay” field, click OK and save a second version of the file with an embedded scale bar under a new name. At this point, you will have two versions of the file—one without a scale, the other with. 6. Camera-produced photos do not carry information on scale. Include a ruler or other object of defined size already when taking the picture (plant pots, Petri dishes or gel comb imprints of known dimensions can serve as an internal scale). You can process these photos in Fiji as described in steps 4 and 5, using manual calibration. Alternatively, uniform well-documented adjustments of contrast, brightness, etc. (see step 3) can be

128

Fatima Cvrcˇkova´

performed using Gimp. If adding a scale bar in Gimp, calculate its size using internal standard information, and draw the bar of desired length somewhere at the very margin of the photo. This bar is part of the bitmap image and will scale with the photo, but can be removed by subsequent cropping and replaced by a vector scale bar during final figure assembly. When done, export the image as .tiff (LZW compression may be used to save space). Do not use the*.jpg format, which applies lossy compression. 7. You can also employ Gimp to roughly crop an image (either Fiji- or Gimp-processed). It is recommended to leave a generous margin for final cropping during composite figure assembly. 8. Gimp or IrfanView can also be used to scale and resample bitmaps, which may be useful in the case of very large, highresolution photos (such as composites covering multiple microscope fields) that will be scaled down for display. To remain safe, do not resample to less than twice the expected final resolution (taking into account expected image size). 3.2 Preparing Graphs and Diagrams for Presentation

1. Create a new file in Inkscape and import preexisting graphics prepared in another application (see Subheading 2.3, items 2 and 3) and saved in a vector format. If the source graph was prepared in LibreOffice Calc, open the *.ods file, copy the graphics and paste into an open Inkscape document (see Note 8). Save the inkscape file as *.svg (see Note 9). 2. Resize the graphics by dragging its corners as close to the desired size, as possible. Ungroup all objects and remove any unwanted objects (frames, background grid, etc.). 3. Adjust the thickness and color of lines and modify data point marker shape or fill colors and patterns as desired. Adjust font and size of any lettering (avoid using lines thinner than 0.5 pt. and letters smaller than 8 pt). Save the file. 4. If your graphics is only available in a bitmap format, paste it into an Inkscape file and use it as a template to manually redraw a vector version (see Note 10). 5. The final editing of graphs can be done either separately, with one graph per file, or combined with composite figure assembly (Subheading 3.3). Save Inkscape files in the default *.svg format.

3.3 Making Composite Figures

1. Check Instructions to Authors of your destination journal for maximum image size and for resolution requirements for bitmaps. Open Inkscape, go to File—Inkscape Settings, and set “Bitmaps—Resolution to create bitmap copy” to between once and twice the desired final value (more is better but do not

Image Data Handling and Processing

129

exceed 600 dpi, if possible). Make sure that the “When scaling objects. . .scale stroke width” setting in the top menu is off (it is on by default in freshly installed Inkscape). These settings will stay until you change them next time. 2. For each figure, start by setting canvas size to the maximum allowed by the journal for a 1- or 2-column figure, as desired (see Note 11) by setting Custom Page size in File—Document properties. Save the file as *.svg. 3. If the figure comprises multiple composite panels consisting of several photos, start by assembling the individual panels and then reshuffle them and make final adjustments. It is recommended to save your work at least at the end of each of the following steps. 4. Import all the photos that will form an individual composite panel. If you expect to crop the photos and remove the embedded scale bar, use only the scale bar-containing files from Subheading 3.1, step 5. If the scale bar overlaps a portion of the image you want to show, import also the scale bar-free version of the file, overlay the scale-bar containing version over it and group the two. 5. Draw a thick line or rectangle whose length corresponds to the scale bar(s) you want to use in the final figure (remember or note its length). 6. Roughly arrange the photos and scale bars to the desired layout by dragging (use of guides or the Align and Distribute tool, as well as grouping objects at intermediary stages of the process, may help) and group them together. The scale bars may be located either outside the photos, or “floating” above them. Resize the whole group approximately to the target size, or a bit larger, while holding Ctrl (to ensure proportional resizing), Ungroup. You will now have multiple photos awaiting final cropping, some of them duplicate (if there is both a scalecontaining and scale-free layer), and some free-floating scale bars positioned close to their intended final positions. 7. Remove the scale-bar containing overlays from the duplicate photos. 8. Resample each photo to the resolution set in step 1 by the following procedure. Select the photo and overlay it by its bitmap copy at the desired resolution (Edit—Create a bitmap copy). Select the bitmap copy (image on top) and send it behind the original photo (Layer—Lower layer to bottom). Delete the original photo, which is now on the top. Save after processing each panel. If the source photos were big, the*.svg file will shrink in size considerably (see Note 12).

130

Fatima Cvrcˇkova´

9. Crop each photo to its desired dimensions. To do so, draw an unfilled or semitransparent template rectangle and make a copy of it for each photo to be cropped to the same size. For each photo, select a template, position it over the image as required (avoiding the embedded scale bar), and then select the photo while holding Shift (resulting in both the template and the photo being selected). Run the Object—Clip—Set command, which will results in the template disappearing and the photo appearing cropped (see Note 13). 10. When all photos have been processed, arrange them and the scale bars to the final layout. Use of Align and Distribute tools is recommended. Annotate the scale bars and any features you want to highlight (e.g., by arrows). Type in captions as required, and group all components of the panel together. 11. Process any additional photo-containing panels as in steps 4–10. 12. To insert a panel containing graphics, open the corresponding file from Subheading 3.2, copy its contents and paste into your nascent composite figure. Group all parts of a graphics together and position or resize it as required (see Note 14). 13. When all panels have been arranged to your satisfaction, which might necessitate some more resizing, check once more for font size consistency in annotation and adjust fonts as needed (ungrouping will be necessary for this). Group all parts of the figure and adjust canvas size to get rid of excess free space on the page. Save the *.svg file. 3.4 Exporting Images for Publication and Presentation

1. Check which formats are acceptable for submission. Some publishers may accept *.svg files. If this is not an option, check for the *.pdf or *.eps possibility or bitmap formats (see Note 15). 2. Inkscape can generate files in several vector formats with embedded bitmaps, including *.pdf. However, it can only export the whole image as bitmap in a *.png format (which few if any publishers want), albeit with a rich selection of available settings. If you need to submit a bitmap for publication, generate a *.png file of the desired final resolution in Inkscape and convert it to the bitmap format of your choice in Gimp. 3. If you need to generate a multipage *.pdf document containing both text and images to be used as Supplementary information, the document may be assembled in a text editor of your choice, such as MS Word or LibreOffice Writer, with embedded images and printed to PrimoPDF. Multipage documents may be generated by repeated printing from PrimoPDF to the same file using the “Append” option. Alternatively, generate the

Image Data Handling and Processing

131

individual pages of the document in Inkscape, save them as *. pdf and merge using PDF Merge (https://www.pdfmerge. com) or a similar online tool (see Note 16).

4

Notes 1. Memory may become limiting when working with multiple large images. The less memory your computer has, the more it is important to keep the size of your project under control (see Subheading 3.1, steps 7 and 8 and Subheading 3.3, step 8). 2. In theory, MS Excel can be run on Linux in a virtual Windows environment or using the Wine application. However, for the purposes described here this does not bring any obvious benefits, especially since existing Excel files can be opened in LibreOffice Calc (albeit possibly with some limitations). 3. Gimp, as any other bitmap editor, is an extremely powerful tool that might tempt you to use it for other purposes, including those generally considered as unethical (see Subheading 1). Never use bitmap editors to locally alter your image! 4. The importance of choosing from excess of primary data cannot be overstressed. If you have only one photo which happens to contain a hair, a bubble, a mold colony, or an aphid in addition to your object of interest, repeat the experiment. After all, only reproducible data should be published. 5. MS Excel provides rather misleading possibilities of error bar insertion. Avoid the preset “standard deviation” option, and use only custom error bars whose length you have calculated yourself (see ref. 11 for general considerations regarding the use of error bars). 6. If the files are at a remote destination, copy them to your local machine or a directly attached storage device. Opening a large file over a network may lead to a crash. 7. See refs. 5, 6 for what is and what is not considered acceptable especially in the case of nonlinear brightness/contrast adjustments. 8. Content copied from MS Excel can be pasted as extended metafile, that is, in a vector format, into nonfree vector editors such as Corel Draw or Adobe Illustrator. Unfortunately, this does not work in Inkscape, and graphs directly pasted from MS Excel to Inkscape end up in a bitmap format. 9. Keep saving your work repeatedly to prevent data loss in case there is a crash.

132

Fatima Cvrcˇkova´

10. Although Inkscape provides a bitmap-tracing utility, its results usually require extensive postprocessing that tends to take longer than redrawing, unless the graphics is very large and complex. 11. If only maximum image width is given, but not image height, assume that the image should fit on an A4 or Letter paper with a reasonably sized white margin. If anticipating possible resubmission to another journal, try to find image size compatible with both. 12. Some files will be generated on your disk. You can safely delete them. 13. The cutoff portions of the photo can be recovered using Object—Clip—Release. If you want to get really rid of a part of the photo, use Gimp to crop it, or replace the original photo with a bitmap copy after cropping (see Subheading 3.3, step 8). 14. You may use nonproportional resizing for graphs, but be prepared that you will end up rewriting annotations and possibly replacing data point markers deformed during resizing. 15. A disadvantage of *.eps files is the impossibility to preview file content directly in a vector mode. These files carry an embedded bitmap preview that increases file size but cannot accurately reproduce the vector elements (which only show properly after import into a graphics editor). 16. Inkscape usually produces smaller *.pdf files with better quality than the text editor + PrimoPDF approach, but sometimes has problems with files that contain a large amount of text.

Acknowledgments I thank all the students, colleagues, and manuscript authors whom I witnessed to struggle with the issues discussed here and whose questions brought me to the idea of writing this chapter, Edita Jankova´ Drdova´ and Lucie Brejsˇkova´ for critical reading of the manuscript, and the Ministry of Education of the Czech Republic for financial support from the NPUI LO1417 project. References 1. Stallman R (2005) Free community science and the free development of science. PLoS Med 2:e47 2. Tchantchaleishvili V, Schmitto JD (2011) Preparing a scientific manuscript in Linux: today’s possibilities and limitations. BMC Res Notes 4:434

3. Yuan S, Chan HCS, Filipek S, Vogel H (2016) PyMOL and Inkscape bridge the data and the data visualization. Structure 24:2041–2042 4. Bigelow A, Drucker S, Fisher D, Meyer M (2017) Iterating between tools to create and edit visualizations. IEEE Trans Vis Comput Graph 23:481–490

Image Data Handling and Processing 5. Rossner M, Yamada KM (2004) What’s in a picture? The temptation of image manipulation. J Cell Biol 166:11–15 6. Martin C, Blatt M (2013) Manipulation and misconduct in the handling of image data. Plant Cell 25:3147–3148 7. Tan L (2006) Image file formats. Biomed Imaging Interv J 2:e6 8. Schindelin J, Rueden CT, Hiner MC, Eliceiri KW (2015) The ImageJ ecosystem: an open platform for biomedical image analysis. Mol Reprod Dev 82:518–529

133

9. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682 10. North AJ (2006) Seeing is believing? A beginners’ guide to practical pitfalls in image acquisition. J Cell Biol 172:9–18 11. Cumming G, Fidler F, Vaux DL (2007) Error bars in experimental biology. J Cell Biol 177:7–11

Chapter 9 Visualizing and Quantifying In Vivo Cortical Cytoskeleton Structure and Dynamics Amparo Rosero, Denisa Oulehlova´, Viktor Zˇa´rsky´, and Fatima Cvrcˇkova´ Abstract The cortical microtubule and actin meshworks play a central role in the shaping of plant cells. Transgenic plants expressing fluorescent protein markers specifically tagging the two main cytoskeletal systems are available, allowing noninvasive in vivo studies. Advanced microscopy techniques, in particular confocal laser scanning microscopy (CLSM), spinning disk confocal microscopy (SDCM), and variable angle epifluorescence microscopy (VAEM), can be nowadays used for imaging the cortical cytoskeleton of living cells with unprecedented spatial and temporal resolution. With the aid of free computing tools based on the publicly available ImageJ software package, quantitative information can be extracted from microscopic images and video sequences, providing insight into both architecture and dynamics of the cortical cytoskeleton. Key words Actin, Microtubules, Fluorescent proteins, CLSM, SDCM, VAEM, Image analysis, ImageJ

1

Introduction Cortical microtubules are long known to play a major part in the morphogenesis of plant cells, in particular due to their intimate relationship with the biosynthesis of the cellulosic cell wall microfibrils (see for example ref. 1). However, the actin cytoskeleton, which undergoes constant dynamic remodeling [2], is crucial for processes such as trichome morphogenesis [3], tip growth in root hairs [4], or development of epidermal cell lobes [5], and apparently contributes to the localization of exocytosis, affecting also the positioning of cellulose synthase complexes [1, 6]. Detailed characterization of the spatial structure and temporal behavior of the two main cytoskeletal systems in vivo may thus substantially contribute to our understanding of plant cell shaping. For such studies, specific and nondisruptive fluorescent cytoskeletal markers must be introduced into the tissues of interest, and suitable high-resolution imaging technology must be available. If quantitative information

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_9, © Springer Science+Business Media, LLC, part of Springer Nature 2019

135

136

Amparo Rosero et al.

is to be extracted from image data, appropriate protocols and software are needed. A variety of fluorescent protein-based markers has been used to trace cytoskeletal structures in living cells, including those of plants. Plant microtubules have been successfully visualized using both GFP–tubulin fusions [7, 8] and GFP-tagged autologous or heterologous microtubule-associated proteins (MAPs) such as mammalian MAP4 [9] or several isoforms of Arabidopsis MAP65 [10]. In addition to labeling microtubules along their whole length, microtubule ends can be specifically marked by tags based on end-binding proteins such as EB1 preferring minus ends [11] or mammalian CLIP170 for plus ends [12]. For actin visualization, fluorescent protein-tagged mammalian talin [13], constructs based on the C-terminal actin-binding domain of Arabidopsis fimbrin (FABD, refs. 14, 15,) or the 17-amino acid actin-binding LifeAct peptide [16] have been used to target fluorescent proteins to actin filaments in plant cells. It has to be stressed that any experiments including (over) expression of tagged (i.e., modified) and possibly heterologous proteins have to be interpreted with caution, as (1) only a subset of the relevant cytoskeletal structures may be labeled, as shown for example for the various MAP65 isoforms [10] and (2) the tag itself may affect cytoskeletal structure and dynamics. Both talin and MAP4-based markers cause visible phenotypic alteration on the whole plant level [17], and in particular GFP-tagged talin was shown to interfere with actin dynamics and aggravate the effects of some treatments and mutations affecting the actin cytoskeleton [18, 19]. Even LifeAct-based constructs may interfere with normal plant development if overexpressed [20]. A suitable high resolution fluorescence microscopy and microphotography equipment is required to make full advantage of in vivo cytoskeletal labeling. Conventional fluorescence microscopy, although useful, is limited by spatial resolution, interfering background (auto)fluorescence and usually also by long exposure times. However, advanced microscopy techniques, such as confocal laser scanning microscopy (CLSM), can be used to improve spatial resolution, while spinning disk confocal microscopy (SDCM) provides excellent possibilities to image rapid processes [21]. Very thin samples can be observed with supreme spatial and temporal resolution using the total internal reflection microscopy (TIRFM) technique. While TIRFM can only reach up to some 200 nanometers from the coverslip, TIRFM hardware can also be used in variable angle epifluorescence microscopy (VAEM) mode with a reasonable trade-off between lateral resolution and imaging depth, thus allowing for visualization of a thin cortical layer of the cytoplasm through the cell wall [22–28]. In plant cells, the evanescent wave might be initiated not only on the coverslip surface but also between the cell wall and plasmalemma, the cell wall thus being a

Quantitative Cortical Cytoskeleton Analysis In Vivo

137

part of the optical system, and even true TIRFM may thus work [24]. A variety of computational techniques can be used to analyze high-resolution images of the plant cortical cytoskeleton and quantify their biologically relevant parameters. With a bit of exaggeration, there may be as many, or even more, image analysis methods as there are publications devoted to the topic, which often hampers comparison of data from different laboratories. Here we are presenting the protocols currently used in our laboratory to study in vivo cytoskeletal structure and dynamics in Arabidopsis thaliana [20, 27–29], but based to a large extent on previous work published by others [30–33], with the hope to contribute to the standardization of basic approaches. Some of the quantification methods presented here can be used also for evaluation of images obtained from fixed material for example after antibody staining.

2

Materials Besides specialized equipment and materials listed below, standard equipment, tools and consumables for plant in vitro culture will be required.

2.1

Plants

1. The fluorescent markers listed in the Introduction are likely to be available upon request from the authors who published them, either in the form of a plasmid suitable for transformation (which may be useful for introducing the marker into mutants), or in the form of seeds of stable transgenic lines. Transgenic A. thaliana lines carrying GFP-tagged tubulin markers GFP-TUB6 and GFP-TUA6 can be obtained also from the public Arabidopsis stock collections—NASC (http://ara bidopsis.info) and ABRC (http://www.arabidopsis.org)— under stock codes N6550 and N6551 (NASC) or CS6550 and CS6551 (ABRC), respectively. Stable transgenic plants carrying the marker of interest can be then used to introduce the markers into different genetic backgrounds (e.g., various mutants) by crossing (see Note 1). While, in principle, any fluorescent cytoskeletal marker can be used, our experience is based mainly on observations in plant lines carrying two marker constructs expressed under the viral 35S promoter—GFPMAP4 [9] and GFP-FABD [34], as well as GFP–, YFP–, and mRFP–LifeAct fusion proteins expressed from the weaker ubiquitin (UBQ10) promoter [20]. 2. We usually observe roots and cotyledons of young A. thaliana seedlings (5–8 days after germination) grown on vertical MS plates at 22
C with a 16 h-light/8 h-dark cycle (see Note 2). Pharmacological treatments may be included during cultivation, and seedlings may be alternatively grown in the dark to

138

Amparo Rosero et al.

achieve etiolation, as etiolated hypocotyls provide another interesting model especially for TIRF observation (ref. 22; see Note 3). Appropriate controls (e.g., wild type for mutants, or nontreated plants for pharmacological studies) have to be included at the same time, since all measurements can be interpreted only in comparison with data from simultaneously grown control plants (see Note 4). 2.2 Microscopy and Image Processing

For CLSM, SDCM and VAEM, we provide information on instrument configuration we are using, as well as basic settings (in Subheading 3.2) as a guide, albeit modifications and some experimenting will be necessary with different hardware (see Note 5). 1. CLSM: Leica TCS SP2 or Zeiss LSM880 confocal laserscanning microscope equipped with a 63/1.2 waterimmersion objective and 488-nm argon laser for excitation. 2. SDCM: inverted spinning disk confocal microscope (Yokogawa CSU-X1 on a Nikon Ti-E platform, laser box Agilent MLC400, camera Andor Ixon) with plan apochromat  100 oil (NA ¼ 1.45) lens, laser lines 488 and 561 nm. 3. VAEM: Leica AF6000 LX fluorescence platform with integrated TIRF module, HCX PL APO 100/1.46 oil immersion objective, equipped with the a Leica DFC350FXR2 digital camera for recording. 4. Microscopy slides, coverslips (preferentially larger size to accommodate the whole length of a stretched seedling), chambered slides (Nunc Lab-Tek II, 1 well, catalog number 154453), sterile water, immersion oil, tweezers, sterile toothpick, paper tissues, (optional) nail polish. 5. Personal computer (see Note 6) with an up-to-date version of the Fiji distribution of ImageJ ([35, 36]; http://fiji.sc) installed, together with the Lpixel set of plugins ([37]; https://lpixel.net/products/lpixel-imagej-plugins/). The macros hig_skewness.txt and hig_255counts.txt [32] that were formerly available at the now-defunct Higaki Laboratory web (see Note 7) can be downloaded from the “Methods and software” section of our laboratory website (follow the Cell Morphogenesis link at https://www.natur.cuni.cz/biology/ plant-biology/science-and-research). Three macros, MovieProcess.ijm, MovieProcessFineGrid.ijm, and LineToSkeleton. ijm, from the QuACK package can be downloaded as Supplementary data of ref. 20. A table calculator (such as Microsoft Excel or Libre Office Calc) will be also required.

Quantitative Cortical Cytoskeleton Analysis In Vivo

3

139

Methods

3.1 Preparing Plant Materials for Visualization

1. For CLSM, place a seedling (collected off the agar plate using sterile toothpick) into a drop of water or cultivation medium on a microscope slide, and cover with a coverslip, avoiding bubbles as far as possible. Remove excess water at the slide edges with a torn bit of paper tissue (see Notes 8 and 9). 2. For SDCM and VAEM, cut a piece of agar containing the seedlings (width 1.5 cm, length according to the seedling size). We usually observe two or three 5 days old plants per agar piece. Put a drop of water on the chambered slide and put the piece of agar placing the seedlings in contact with the glass, avoiding bubbles (the piece of agar helps to press the seedling tissues in contact with the slide). Remove excess water by gently touching the edges of agar with paper tissue.

3.2 Image Acquisition

Follow the microscope.

recommended

standard

procedures

for

your

1. In case of CLSM, we record single slices (capturing an area of the cell cortex adjacent to the coverslip) and Z series using the following settings: excitation laser (488-nm argon) intensity 25 mW, detector window using the GFP preset values (see Note 10), XY field size 1024  1024 pixels, line averaging of 4–8 times, Z series interval 0.7–1 μm, color depth of 12 bits. 2. In case of SDCM, we record single slices using excitation lines 488 and 561 nm for GFP and mRFP, respectively, with frame interval 1 s. 3. In case of VAEM, we use 400 nm peak excitation for GFP constructs,150–210 ms exposure time, frames taken in 0.5 s intervals over the course of 2 min, color depth of 8 bits. 3.3 Measuring Cytoskeletal Network Density on CLSM Stacks

1. Obtain serial optical sections (XYZ, i.e., Z-stack) of the cortical cytoplasm of a cell expressing a suitable marker by CLSM. In general, we aim toward imaging about 7–10 plants per sample, with 5–10 cells per plant evaluated (see Note 11). 2. Open the stack by dragging the microscope-generated file onto the Fiji icon or window; use the “open as hyperstack” option in the dialog box. Skeletonize the original serial optical sections (Fig. 1a) using Plugins > LPX > Lpx_Filter2d (set filter: lineFilters and parms linemode: thinLine; Fig. 1b). Generate a Z projection (Image > Stacks > Zproject) using the maximum intensity option and save the resulting image as a new 8-bit *.tif file (Image > Type > 8bits; Fig. 1c; see Note 12). 3. Select the area to be analyzed (a whole cell, or a well-focused region of the image) manually by ROI selection

140

Amparo Rosero et al.

Fig. 1 Stages of image processing prior to determining actin density and bundling using ImageJ. (a) Serial optical sections from CSLM. (b) Skeletonization of serial optical sections. (c) Single image from maximum intensity projections

(Analyze > Tools > ROI Manager). Add the selected ROI using the Add button. To specify multiple ROIs of the same size and shape within an image, you may duplicate the selected ROI (right mouse click > duplicate). 4. Evaluate the filament density within the ROI by estimation of the GFP signal occupancy, that is, the fraction of pixels constituting the skeletonized filaments relative to the total pixel number of the ROI. Install the macro hig_255counts.txt using Plugins > Macro > Install, browse to the location of the macro, select it and click Open. The macro will now appear in the Plugins menu. Count pixel number of selected ROI using the macro: hig_255counts.txt (Plugins > hig_255counts). The occupancy value is proportional to the overall filament density

Quantitative Cortical Cytoskeleton Analysis In Vivo

141

in the cell region of interest, and was shown to serve as an useful indicator to evaluate, for example, the changes in the microfilament organization induced by physiological processes, treatments with inhibitors, such as latrunculin B [32] or by gene mutations (refs. 27, 28, 38, 39, see Note 13). 5. The size and organization of microtubules makes their analysis easier than in the case of actin; microtubule density can be estimated in small specific area of a defined size also by direct manual counting [27]. 3.4 Evaluating Actin Bundle Thickness by Measuring the Skewness of Fluorescence Distribution

1. Record an image stack, prepare a skeletonized image of a maximum intensity projection and select ROIs as needed, as described in Subheading 3.3, steps 1–3.

3.5 Evaluating Actin Bundle Thickness Using the Histogram Method

1. From original stack of optical sections obtained by CLSM, prepareamaximumintensityprojection (Image> Stacks>Zproject) and save it as an 8 bits *.tif image (Image > type > 8bits).

2. Determine the skewness of the fluorescence intensity distribution (a measure of the degree of asymmetry of a distribution, correlated with microfilament bundling because bundles exhibit brighter fluorescence) in the microfilament-containing pixels using the macro: hig_skewness.txt. To do so, install the macro as described in Subheading 3.3, step 4 and then run it (Plugins > hig skewness).

2. Draw a line of a defined length (length shows at the bottom of the toolbar) across a representative, well-focused part of a cell; use Shift if you wish to constrain the line direction (see Note 14). 3. Generate a profile of GFP fluorescence intensity (Analyze > Plot profile) and record the brightness values of all peaks corresponding to microfilament bundles crossed by the line (values appear on mouseover or generate a list of values by pressing “List”). Record also background values in an area devoid of actin filaments. 4. Using the table calculator, subtract the average background value from the peak values and generate a histogram of the distribution of the resulting net peak values into three or four equally broad classes of grey level (in arbitrary units). The resulting plot documents microfilament bundling, as low intensity represents weakly labeled bundles or single filaments, and high intensity corresponds to brightly labeled bundles.

3.6 Quantifying Filament Dynamics from SDCM or VAEM Image Series—A Manual Method

1. Acquire temporal series of single-plane optical sections (XYT) of the cortical cytoplasm of a cell expressing a suitable marker by SDCM or VAEM. We aim toward imaging at least five plants per sample, with 15–20 movies per sample evaluated (see Note 11).

142

Amparo Rosero et al.

2. To measure microfilament dynamics, select randomly ten actin bundles per sample and measure their pause duration (for monitoring over time, use multipoint selection tool in ImageJ and register manually the time when the filament end shows a change in behavior). Values and distribution of pause duration can serve, for example, as an indicator of differences either in bundle size or in the degree of actin cross-linking (ref. 27, see Note 15). 3. To quantify microtubule turnover, select randomly 10–20 microtubule ends per sample and monitor their behavior over time (2 min); use the pen or brush tool in ImageJ to mark the already evaluated ends. Count microtubules in the four distinct phases (growing, shrinking, pausing, and alternating between growth and shrinkage). 4. To estimate of microtubule growth and shrinkage rates, select randomly 5–10 microtubule ends per cell and measure their distances from the starting position during specific time using ImageJ. 3.7 Kymogram Construction from SDCM or VAEM Image Series

Kymograms can be used to visualize aspects of microfilament and microtubule dynamics that are not easily observed in the video sequences. 1. Open the microscope-generated file by dragging it onto the Fiji icon or window; use the “open as hyperstack” option in the dialog box, and select the desired image to evaluate if the file contains multiple images. 2. Draw a line of defined length across a representative area of the image as described in Subheading 3.6, step 2. 3. Generate the kymograms using the plugin Multiple Kymograph (Plugins > MultipleKymograph with linewidth: 3). 4. The image generated shows velocity, movement, and different phases of microfilament or microtubule turnover (see Note 16).

3.8 Quantitative Analysis of Cytoskeletal Kymograms

We recently developed a new technique, QuACK (Quantitative Analysis of Cytoskeletal Kymograms) that utilizes kymograms derived from SDCM or VAEM recordings to obtain a quantitative estimate of lifetime and lateral mobility of cytoskeletal structures [20]. 1. Select the SDCM or VAEM files to be analyzed and open them in Fiji using the following Bio-Formats import option: View stack with Hyperstack, Stack order XYZT, Open files individually, Autoscale, Split channels. Open all the movies you are considering for analysis, including those where you are in

Quantitative Cortical Cytoskeleton Analysis In Vivo

143

doubt as far as the image quality is concerned; discard any obviously bad (e.g., drifting or out of focus) movies. 2. Set the parameters to be measured:(Analyze > Set Measurements, check Shape Descriptors). 3. If working with anisotropic samples, rotate all movies to align a well-defined structural landmark to the horizontal or the vertical (the same for all movies). To do so, draw a straight line segment along the landmark, press Ctrl + M and note the Angle value, go to Image > Transform > Rotate, input the Angle value (in whole degrees), click OK. This will produce a rotated movie. Process all movies and close the Results window without saving when done. 4. Install the MovieProcess.ijm (for actin) or MovieProcessFineGrid.ijm (for microtubules) as described in Subheading 3.3, step 4. Apply this macro to each of the open movies. This will convert each movie to grayscale, stretch the contrast in a defined manner, and overlay a randomly positioned 5 μm  5 μm (or 2 μm  2 μm) grid over the movie. Discard any obviously bad, especially very noisy, movies. 5. Decide on the length of transect you want to analyze. For highly dynamic structures (such as actin in pavement cells), aim to the longest line that fits into most cells in both horizontal and vertical direction. 15–20 μm is a good start for actin, a shorter line (5 or 2 μm) for microtubules. If dealing with a new type of sample, try several transect lengths. If the structures evaluated are very stable, try also varying the duration of the movie—for example, if working with 2 min movies, compare results of evaluating, for example, the first 30 s, 60 s and complete recording (you will need to generate shortened copies of movies using ImageJ tools for this purpose). The transect and recording length should be selected so that only a minority of skeletonized kymograms generated in the next step will contain traces corresponding to structures that either moved laterally across the whole transect, or persisted throughout the whole evaluated length of the recording (see Note 17). 6. Install the LineToSkeleton.ijm macro: as described in Subheading 3.3, step 4. For each movie, generate four skeletonized kymograms (“skeletons”) in the x or y dimension of the movie window (see Note 17) over the given transect length. To produce a skeleton, select the straight line tool and draw a transect of the selected length along the grid while holding Shift. Aim for equally spacing the transects across the focused part of the image. Apply the LineToSkeleton macro to each transect. A skeletonized kymogram will pop up in a separate window. Proceed until you have processed all the movies and

144

Amparo Rosero et al.

accumulated a full set of kymograms for the given parameter/ data point on the desktop. Close any unused movies but keep the used movie windows open if you intend to remeasure them at different transect length. 7. To measure structure lifetime, determine the length of the longest trace in the time (vertical) dimension of each skeleton. Select the kymogram window and adjust magnification as convenient using the Fiji magnifying glass tool. Select the straight line tool and draw a vertical line along the length of the tallest contiguous trace while holding Shift. Press Ctrl + M. A new row of values will appear in the Results window. If not sure which is the tallest trace, measure several candidates and delete those which are not tallest from Results. Do not close the kymogram yet if you plan to measure also lateral mobility; if you do not plan measuring other parameters, close the kymogram without saving. 8. Proceed until all kymograms have been measured, and copy the contents of the Results window to a spreadsheet program of your choice. Clean up and annotate the data in the spreadsheet—you will only be interested in the Length (last) column. Save the spreadsheet. 9. To measure lateral mobility, determine the width of the widest trace in the space (horizontal) dimension of each skeletonized kymogram using the same procedure as for structure lifetime (steps 7 and 8). When done, close all Fiji windows without saving and exit Fiji. 10. To evaluate between-sample differences, use statistical methods that do not require normal distribution of data, because events which are not recorded whole may lead to nonnormal distribution of measured values. To compare two groups, use the Wilcoxon–Mann–Whitney test (U-test). To compare more than two groups, use the Kruskal–Wallis test. A good collection of online tools for performing these (and other) statistical tests is available, for example, at http://vassarstats.net. 11. Because of possible nonnormal value distribution, presenting results should include median and quartile/range (e.g., in a box plot setup).rather than mean values  SE. A good boxplot utility is included in the commonly used free R software package, and several box plot generators are also available online, such as the R-based BoxPlotR (http://shiny.chemgrid.org/ boxplotr/).

Quantitative Cortical Cytoskeleton Analysis In Vivo

4

145

Notes 1. In wild-type genetic backgrounds, some transgenes tend to suffer from gene expression silencing leading to gradual loss of fluorescence. This can be mitigated by using an rdr6 mutant genetic background defective in silencing [40]. An rdr6-12 line is available from NASC and ABRC under stock code CS24286. 2. Use the culture media and protocols established in the laboratory; any medium and culture setup that allows for easy removal of intact seedlings from plates should work. Alternatively, seedlings may be grown on a medium-covered slide surface in situ to avoid disturbance of, for example, root hairs. 3. Any tissue that can be positioned flat toward the coverslip surface ought to be accessible to CLSM, SDCM and VAEM; especially for the later, tight contact with the coverslip is critical. Leaves of glabrous mutants and petals may be especially worth exploring. 4. Ideally, the measurements should be done at least in a singleblind manner to eliminate observer bias (i.e., the person performing quantitative image analysis should not know which image series belongs to which genotype or experimental treatment). 5. Our microscopes are in the inverted configuration. For an upright microscope, sample preparation may have to be modified. For each microscope, we assume that it is controlled by a personal computer running the microscope manufacturer’s software. 6. While we are using predominantly the Windows operation system, the software described in this chapter exists also in versions for other operation systems such as Linux. 7. Although tools for measuring cytoskeletal bundling are also included in the LPixel plugin set [37], it is not clear whether they produce identical results to the previously published version [32] for any given data set. We therefore prefer using the original version of the Higaki lab macros to ensure comparability with older data, at least until thorough comparison of the two versions of the method is performed. 8. Take care to treat all the seedlings equally, since mechanical stress may elicit modification of cytoskeletal organization and dynamics during the plant manipulation, media exchange or even coverslip placement. For longer observation, edges of the coverslip may be sealed with nail polish, but this is usually not necessary. Do not use too much water, as the coverslip should be held in place by capillary forces rather than move around on excess liquid (this is easier to achieve with large coverslips).

146

Amparo Rosero et al.

9. If working carefully, live seedlings can be recovered from the slide after observation and transferred ex vitro for further cultivation, but do not expect 100% survival. 10. In case of high autofluorescence background the window should be narrowed (i.e., longer wavelengths should be cut off in the Beam Path settings). 11. Cytoskeletal structure and dynamics varies dramatically with anatomical location; therefore, imaged cells should be located consistently (e.g., at the bottom of the root tip elongation zone, or in the middle of the cotyledon). 12. Excessive background noise can be removed by thresholding the image stack prior to skeletonization (Image > Adjust > Threshold; see ref. 32). Use a constant threshold value across all images included in your analysis (preliminary optimization may be necessary). 13. Maximum intensity projections from the serial optical sections can serve also for determining cytoskeletal filament orientation. Reference 32 describes a procedure employing noise reduction, conversion of the images to binary and skeletonization. The resulting skeletonized image is used to evaluate cytoskeletal architecture in guard cells of the stomata. Mean angular difference between microfilament pixel pairs and the nearest pixel pairs of a specific cell edge (the stomatal pore) is used as a measure of microfilament orientation. The procedure can be checked by obtaining synthesized images. Analogously, microtubule orientation can be determined with respect to the cell’s specific axis, for example, the longitudinal axis of the hypocotyl [41]. 14. It is recommendable to maintain a constant location/direction of the sampling line within a cell, for example, along the longitudinal axis in case of rhizodermis. 15. Additional data about actin assembly rates, filament origin and severing frequency can be obtained by the analysis of stochastic dynamics as described in ref. 22 where actin filaments are tracked manually through the time-lapse stack of images and different actin dynamic parameters are estimated by overlapping images or monitoring breaks along the filament over time. 16. Choose the length and location of the sampling line consistently (e.g., parallel to the longitudinal axis in roots and hypocotyls, see ref. 33). While 1 min is usually enough to document microfilament dynamics, in the case of the less mobile microtubules 2 min provide a more informative result. 17. In a typical experiment design, we analyze ten movies, originating from at least five biological replicates, per condition investigated, and generate four kymograms per movie at each

Quantitative Cortical Cytoskeleton Analysis In Vivo

147

transect length examined for each parameter measured, which means 40 kymograms per data point at the typical setup for actin. Double the number of transects for microtubules to enhance sensitivity. For isotropic (or randomly positioned) samples such as cotyledon epidermis, horizontal and vertical transects can be mixed; for anisotropic samples such as roots or hypocotyls, decide on the direction.

Acknowledgments The update and extension of this chapter, which was previously ˇ R 15-02610S published as ref. 29,, has been supported by the GAC project. We thank Boris Voigt, Richard Cyr and Matya´sˇ Fendrych for transgenic Arabidopsis lines, Ondrˇej Sˇebesta, Ondrˇej Horva´th, and Alesˇ Soukup for expert microscopy advice, Shiqi Zhang for ˇ adyova´ for technical assistance. helpful suggestions, and Marta C References 1. Crowell EF, Gonneau M, Vernhettes S, Ho¨fte H (2010) Regulation of anisotropic cell expansion in higher plants. C R Biol 333:320–324 2. Blanchoin L, Boujemaa-Paterski R, Henty JL, Khurana P, Staiger CJ (2010) Actin dynamics in plant cells: a team effort from multiple proteins orchestrates this very fast-paced game. Curr Opin Plant Biol 13:714–723 3. Szymanski DB (2005) Breaking the WAVE complex: the point of Arabidopsis trichomes. Curr Opin Plant Biol 8:103–112 4. Pei W, Du F, Zhang Y, He T, Ren H (2012) Control of the actin cytoskeleton in root hair development. Plant Sci 197:10–18 5. Armour W, Barton DA, Law AM, Overall RL (2015) Differential growth in periclinal and anticlinal walls during lobe formation in Arabidopsis cotyledon pavement cells. Plant Cell 27:2484–2500 ˇ a´rsky´ V, Cvrcˇkova´ F, Potocky´ M, Ha´la M 6. Z (2009) Exocytosis and cell polarity in plants–exocyst and recycling domains. New Phytol 183:255–272 7. Ueda K, Matsuyama T, Hashimoto T (1999) Visualization of microtubules in living cells of transgenic Arabidopsis thaliana. Protoplasma 206:201–206 8. Nakamura M, Naoi K, Shoji T, Hashimoto T (2004) Low concentrations of propyzamide and oryzalin alter microtubule dynamics in Arabidopsis epidermal cells. Plant Cell Physiol 45:1330–1334

9. Marc J, Granger CL, Brincat J, Fisher DD, Kao T et al (1998) A GFP-MAP4 reporter gene for visualizing cortical microtubule rearrangements in living epidermal cells. Plant Cell 10:1927–1939 10. Van Damme D, Van Poucke K, Boutant E, Ritzenthaler C, Inze´ D et al (2004) In vivo dynamics and differential microtubule-binding activities of MAP65 proteins. Plant Physiol 136:3956–3967 11. Chan J, Calder G, Fox S, Lloyd C (2005) Localization of the microtubule end binding protein EB1 reveals alternative pathways of spindle development in Arabidopsis suspension cells. Plant Cell 17:1737–1748 12. Dhonukshe P, Gadella TWJ (2003) Alteration of microtubule dynamic instability during preprophase band formation revealed by yellow fluorescent protein-CLIP170 microtubule plus-end labeling. Plant Cell 15:597–611 13. Kost B, Spielhofer P, Chua NH (1998) A GFP-mouse Talin fusion protein labels plant actin filaments in vivo and visualizes the actin cytoskeleton in growing pollen tubes. Plant J 16:393–401 14. Sheahan MB, Staiger CJ, Rose RJ, McCurdy DW (2004) A green fluorescent protein fusion to actin-binding domain 2 of Arabidopsis fimbrin highlights new features of a dynamic actin cytoskeleton in live plant cells. Plant Physiol 136:3968–3978

148

Amparo Rosero et al.

15. Voigt B, Timmers ACJ, Sˇamaj J, Mu¨ller J, Baluska F et al (2005) GFP-FABD2 fusion construct allows in vivo visualization of the dynamic actin cytoskeleton in all cells of Arabidopsis seedlings. Eur J Cell Biol 84:595–608 16. Era A, Tominaga M, Ebine K, Awai C, Saito C et al (2009) Application of Lifeact reveals F-actin dynamics in Arabidopsis thaliana and the liverwort, Marchantia polymorpha. Plant Cell Physiol 50:1041–1048 17. Hashimoto T (2002) Molecular genetic analysis of left-right handedness in plants. Philos Trans R Soc Lond Ser B Biol Sci 357:799–808 18. Ketelaar T, Anthony RG, Hussey PJ (2004) Green fluorescent protein-mTalin causes defects in actin organization and cell expansion in Arabidopsis and inhibits actin depolymerizing Factor’s actin depolymerizing activity in vitro. Plant Physiol 136:3990–3998 ˇ a´rsky´ V (2012) 19. Cvrcˇkova´ F, Grunt M, Z Expression of GFP-mTalin reveals an actinrelated role for the Arabidopsis class II formin AtFH12. Biol Plant 56:431–440 20. Cvrcˇkova´ F, Oulehlova´ D (2017) A new kymogram-based method reveals unexpected effects of marker protein expression and spatial anisotropy of cytoskeletal dynamics in plant cell cortex. Plant Methods 13:19 21. Jonkman J, Brown CM (2015) Any way you slice it-a comparison of confocal microscopy techniques. J Biomol Tech 26:54–65 22. Staiger CJ, Sheahan MB, Khurana P, Wang X, McCurdy DW et al (2009) Actin filament dynamics are dominated by rapid growth and severing activity in the Arabidopsis cortical array. J Cell Biol 184:269–280 23. Smertenko A, Deeks MJ, Hussey P (2010) Strategies of actin reorganisation in plant cells. J Cell Sci 123:3019–3028 24. Vizcay-Barrena G, Webb S, Martin-FernandezM, Wilson ZA (2011) Subcellular and singlemolecule imaging of plant fluorescent proteins using total internal reflection fluorescent microscopy (TIRFM). J Exp Bot 62:5419–5428 25. Wan Y, Ash WM, Fan L, Hao H, Kim MK et al (2011) Variable-angle total internal reflection fluorescence microscopy of intact cells of Arabidopsis thaliana. Plant Methods 7:27 26. Sparkes I, Graumann K, Martiniere A, Schoberer J, Wang P et al (2011) Bleach it, switch it, bounce it, pull it: using laser to reveal plant cell dynamics. J Exp Bot 62:1–7 ˇ a´rsky´ V, Cvrcˇkova´ F (2013) AtFH1 27. Rosero A, Z formin mutation affects actin filament and

microtubule dynamics in Arabidopsis thaliana. J Exp Bot 64:585–597 28. Rosero A, Oulehlova´ D, Stillerova´ L, Schiebertova´ P, Grunt M et al (2016) Arabidopsis FH1 formin affects cotyledon pavement cell shape by modulating cytoskeleton dynamics. Plant Cell Physiol 57:488–504 ˇ a´rsky´ V, Cvrcˇkova´ F (2014) Visua29. Rosero A, Z lizing and quantifying the in vivo structure and dynamics of the Arabidopsis cortical cytoskeleton using CLSM and VAEM. Methods Mol Biol 1080:87–97 30. van der Honing H, Kieft H, Emons A, Ketelaar T (2012) Arabidopsis VILLIN2 and VILLIN3 are required for the generation of thick actin filament bundles and for directional organ growth. Plant Physiol 58:1426–1438 31. Higaki T, Kutsuna N, Sano T, Hasezawa S (2008) Quantitative analysis of changes in actin microfilament contribution to cell plate development in plant cytokinesis. BMC Plant Biol 8:80 32. Higaki T, Kutsuna N, Sano T, Kondo N et al (2010) Quantification and cluster analysis of actin cytoskeletal structures in plant cells: role of actin bundling in stomatal movement during diurnal cycles in Arabidopsis guard cells. Plant J 61:156–165 33. Sampathkumar A, Lindeboom J, Debolt S, Gutierrez R, Ehrhardt DW et al (2011) Live cell imaging reveals structural associations between the actin and microtubule cytoskeleton in Arabidopsis. Plant Cell 23:2302–2313 34. Ketelaar T, Allwood EG, Anthony RG, Voigt B, Menzel D et al (2004) The actininteracting protein AIP is essential for actin organization and plant development. Curr Biol 14:145–149 35. Schindelin J, Rueden CT, Hiner MC, Eliceiri KW (2015) The ImageJ ecosystem: an open platform for biomedical image analysis. Mol Reprod Dev 82:518–529 36. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682 37. Higaki T (2017) Quantitative evaluation of cytoskeletal organizations by microscopic image analysis. Plant Morphology 29:15–21 38. Henty JL, Bledsoe S, Khurana P, Meagher RB, Day B et al (2011) Arabidopsis actin depolymerizing factor 4 modulates the stochastic dynamic behavior of actin filaments in the cortical array of epidermal cells. Plant Cell 23:3711–3726

Quantitative Cortical Cytoskeleton Analysis In Vivo 39. Li J, Henty JL, Huang S, Wang X, Blanchoin L et al (2012) Capping protein modulates the dynamic behavior of actin filaments in response to phosphatidic acid in Arabidopsis. Plant Cell 24:3742–3754 40. Sasse J, Simon S, Gu¨beli C, Liu GW, Cheng X et al (2015) Asymmetric localizations of the ABC transporter PaPDR1 trace paths of

149

directional strigolactone transport. Curr Biol 25:647–655 41. Yao M, Wakamatsu Y, Itoh T, Shoji T, Hashimoto T (2008) Arabidopsis SPIRAL2 promotes uninterrupted microtubule growth by suppressing the pause state of microtubule dynamics. J Cell Sci 121:2372–2381

Chapter 10 Quantitative and Comparative Analysis of Global Patterns of (Microtubule) Cytoskeleton Organization with CytoskeletonAnalyzer2D Birgit Mo¨ller, Luise Zergiebel, and Katharina Bu¨rstenbinder Abstract The microtubule cytoskeleton plays important roles in cell morphogenesis. To investigate the mechanisms of cytoskeletal organization, for example, during growth or development, in genetic studies, or in response to environmental stimuli, image analysis tools for quantitative assessment are needed. Here, we present a method for texture measure-based quantification and comparative analysis of global microtubule cytoskeleton patterns and subsequent visualization of output data. In contrast to other approaches that focus on the extraction of individual cytoskeletal fibers and analysis of their orientation relative to the growth axis, CytoskeletonAnalyzer2D quantifies cytoskeletal organization based on the analysis of local binary patterns. CytoskeletonAnalyzer2D thus is particularly well suited to study cytoskeletal organization in cells where individual fibers are difficult to extract or which lack a clearly defined growth axis, such as leaf epidermal pavement cells. The tool is available as ImageJ plugin and can be combined with publicly available software and tools, such as R and Cytoscape, to visualize similarity networks of cytoskeletal patterns. Key words Cytoskeleton, Microtubules, Actin, Pattern analysis, Texture measure, ImageJ, MiToBo, R software

1

Introduction The microtubule cytoskeleton forms a highly dynamic intracellular network with important roles in many cellular processes, including cell division, cell expansion, and intracellular and intercellular transport [1–3]. The organization and dynamics of the cytoskeleton change in response to diverse developmental and environmental stimuli, which is choreographed by a combination of selforganization and the regulation of cytoskeletal behavior by microtubule-associated proteins (MAPs) that facilitate, for example, bundling, cross-linking, or membrane tethering of microtubules [4–6]. Microtubule (re)organization, for example, during developmental transitions or in response to changing environments, can occur rapidly, within minutes, and often precedes

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_10, © Springer Science+Business Media, LLC, part of Springer Nature 2019

151

152

Birgit Mo¨ller et al.

changes in growth [7]. Quantification of morphological properties of cytoskeletal arrays is key to evaluate changes in cytoskeletal organization, for example, in mutants defective in MAP function or in seedlings upon perception of developmental or environmental stimuli. Analysis of organizational patterns of the microtubule cytoskeleton commonly relies on extraction of individual fibers within the cell type of interest, followed by extraction of feature data, such as fiber length [8], fiber orientation relative to the growth axis [9], or angular differences to the local cell region skeleton [10]. Alternatively, average fluorescence intensities are quantified as a measure for microtubule density [11]. In most cases, only a small number of manually selected regions of interest is analyzed, for example, along the anticlinal to outer periclinal cell wall [12, 13] or within the outer epidermal layer [14–16]. Individual fibers are segmented or labeled either manually [9], semiautomatically in a given subregion of a cell of interest [14], or automatically, for example, by analysis of local image gradients in combination with binarization methods and morphological operations [10, 17, 18]. These methods are best suited for quantification of well-organized, mostly parallel and clearly separable fibers but limited in their usability for quantification of wavy or curved fibers, of densely packed fibers, or of fibers in images, which vary in local image contrast. In such situations, approaches aiming to extract individual fibers for quantification of structural characteristics at best extract only a subset of striking, but not necessarily representative fibers. These limitations can be overcome by avoiding an explicit segmentation of individual fibers and by focusing on a quantitative analysis of the overall appearance of microtubule patterns, for example, by analysis of local gradient distributions or characteristics of local structure tensors [14, 19, 20]. In combination with methods for (semi)automatic segmentation of cell contours, global pattern analyses can provide quantitative measures for the complete cell area, thereby introducing less bias than quantifications from manually selected subregions. Here, we demonstrate how texture measures can be used to quantify structural characteristics of microtubule organization within the complete cell area, ranging from simple shaped cylindrical cells with a clearly defined growth axis, for example, in roots or hypocotyls, to complex-shaped leaf epidermis pavement cells. The approach combines a local structure analysis using a sliding-window approach with an analysis of global distributions of structural patterns (Fig. 1). To quantify local microtubule organization, the area of each cell is partitioned into a set of small subregions. For each subregion a feature vector characterizing the local microtubule structure is extracted. In general, all texture measures can be applied for this task, but local binary patterns (LBPs, ref. 21) have already proven to be well suited [22]. The local analysis results in an

Quantification of Cytoskeleton Organization with Texture Measures

153

Fig. 1 Workflow of cytoskeleton pattern analysis with CellBoundaryExtractor2D and CytoskeletonAnalyzer2D. (a) As input, two-channel images with cell boundaries labeled in one channel and cytoskeletal arrays labeled in the second channel are required. (b) Outlines of cells are automatically extracted from the “cell boundary” channel of an input image using the CellBoundaryExtractor2D. The resulting contours can optionally be postprocessed manually if quality is insufficient. (c) Corresponding label images or ImageJ ROI files containing information on cell contours are required as input for CytoskeletonAnalyzer2D. (d) CytoskeletonAnalyzer2D extracts texture features using a sliding-window approach within given cell boundaries and clusters them into categories representative of distinct cytoskeletal patterns. (e) CytoskeletonAnalyzer2D performs a comparative analysis of global distributions of structural patterns between individual cells and cell groups, which can be visualized, for example, in heat maps or similarity networks

LBP code distribution vector for each subregion and the set of all vectors represents the cytoskeleton properties of the complete cell. To facilitate comparative analysis of microtubules between individual cells or groups of cells, all structural feature vectors of a given data set are categorized into disjoint clusters according to pairwise similarity. Each of these clusters represents a specific pattern of microtubule organization and the fraction of subregions in a cell belonging to each of these clusters, that is, the distribution of clusters in each cell, is interpreted as a characteristic structural fingerprint of the cell. The method is implemented in the image analysis toolbox MiToBo [23] available as plugin for ImageJ/Fiji and offers (semi)automatic segmentation of cell contours [24], quantification of global (microtubule) cytoskeleton patterns and subsequent comparative analysis by generation of similarity matrices or networks using publicly available software like R [25] and Cytoscape [26]. The method has successfully been applied to analyze effects of MAPs on microtubule organization in transient expression assays in Nicotiana benthamiana, and its validity was confirmed by analysis of transgenic Arabidopsis thaliana seedlings overexpressing select candidates [22]. Moreover, the basic methodology has also been used to quantify changes in F-actin

154

Birgit Mo¨ller et al.

organization in human melanoma-derived HT144 cells in reaction to the depletion of RNA-binding proteins [27, 28]. The texture measure-based quantification approach thus can be applied to quantitatively assess structurally diverse cytoskeletal arrays in morphologically different cell types.

2

Materials

2.1 Sample Preparation and Image Acquisition

1. Plant materials may include, (1) leaves of N. benthamiana plants infiltrated with Agrobacterium tumefaciens carrying expression constructs for a plasma membrane marker, such as LTI6b/GFP29-1 [29] or AtPIP2A [30] and a microtubule marker (for example, TUA5 [31], TUB6 [32] or MBD [33]) or a MAP fused to fluorescent proteins (FP) suitable for two-channel imaging [22], or (2) transgenic Arabidopsis thaliana plants expressing FP-fused variants of a microtubule marker (as above) or a MAP of interest in combination with dyes suitable to visualize cell outlines (for example, propidium iodide (PI) or FM4-64) or with a stably transformed fluorescent protein-fused plasma membrane marker (see Notes 1–3). 2. Confocal laser scanning microscope, microscopy slides, and coverslips.

2.2 The MiToBo Software

The CytoskeletonAnalyzer2D tool and the CellBoundaryExtractor2D for cell segmentation are part of the Microscope Image Analysis Toolbox MiToBo (http://mitobo.informatik.uni-halle.de, Ref. [23]) implemented as plugin for ImageJ and Fiji [34]. ImageJ and Fiji are implemented in Java and are available for Windows, Linux, and MacOS. The current release of Fiji can be downloaded from Fiji’s website (https://fiji.sc) where you find installation instructions for all operating systems. Note that MiToBo requires Java 8 to run properly (see Note 4). To install MiToBo within Fiji, proceed as follows: 1. Run Fiji by double-clicking its executable (Windows) or running the starter (Linux/MacOS). 2. Select from the menu bar “Help” ! “Update. . .”. Fiji will run its update routine and probably show you several plugins that can be updated. 3. Click the button “Manage update sites” at the bottom and select “MiToBo” from the list of available sites. Then click the “Close” button. 4. Click “Apply Changes” in the updater window, which will install all available updates and will also install MiToBo in your Fiji distribution (see Note 5).

Quantification of Cytoskeleton Organization with Texture Measures

155

5. Once the updater is finished you need to restart Fiji. Afterward, you will find “MiToBo” in the “Plugins” menu. If you prefer to use ImageJ, the plugin needs to be installed manually. You can find information on how to accomplish this on MiToBo’s installation page (http://mitobo.informatik.uni-halle. de/index.php/Installation). 2.3 Data Analysis Tools

CytoskeletonAnalyzer2D provides distribution diagrams for all analyzed cells and groups of cells. To generate heat maps of pairwise similarities we provide a supplemental script using the R statistical computing environment [25] and its comfortable graphical user interface RStudio [35] (step 1 below). Result data can also be displayed as similarity networks using the network analysis software suite Cytoscape [26] (step 2 below). 1. Installation of R and RStudio: The R Project for Statistical Computing [25] is a free software environment for statistical computing and graphics. It is available for Windows, Linux, and MacOS operating systems. To install the software visit https://www.r-project.org where you find all information related to download and installation of the software. The R basic software suite can be supplemented by the free and opensource integrated development environment (IDE) RStudio [35]. The IDE offers many features for convenient access to R functionality (for example, an integrated code editor with autocompletion and syntax-highlighting, a debugger, project management options, and straightforward installation and removal of additional R packages). The most recent release of RStudio can be downloaded from the project website, https:// www.rstudio.com/products/rstudio/download, where you also find installation instructions. 2. Installation of Cytoscape: The Cytoscape software suite [26] is developed for network data analysis and visualization. The software platform is open-source and can be downloaded from http://www.cytoscape.org/download.php. Detailed information on how to install Cytoscape can be found in Subheading 2 of the Cytoscape user manual, http://manual. cytoscape.org/en/stable/.

3

Methods

3.1 Imaging of Cell Contours and Cytoskeletal Structures

The methods implemented in CellBoundaryExtractor2D and CytoskeletonAnalyzer2D enable (semi) automatic segmentation of cell contours from one channel and automatic quantification of microtubule patterns from a second channel. For generation of input images, choose a combination of FPs or dyes suitable for

156

Birgit Mo¨ller et al.

two-color imaging, for example, CFP and YFP, or mCherry/RFP/ PI/FM4-64 and GFP to label the plasma membrane and the cytoskeleton, respectively. 1. To generate samples for analysis of microtubule patterns in transient expression assays in N. benthamiana, transiently transform leaves of N. benthamiana by pressure infiltration with Agrobacterium tumefaciens strains harboring plasmids that facilitate the expression of FP-fusions of your MAP of interest and a FP-fused plasma membrane marker. We recommend to coexpress viral silencing suppressors, such as p19 or p23 [36]. Adjust the OD600 of each construct to 0.5, mix Agrobacterium strains harboring plasmids for expression of your MAP of interest, the plasma membrane marker and the silencing inhibitor in a ratio of 1:1:2, respectively. Image microtubule patterns at 2–3 days after infiltration (see Note 6). 2. To generate samples for analysis of microtubule patterns in transgenic Arabidopsis seedlings in response to, for example, stress treatments or for comparison between mutants, use microtubule marker lines such as pTUB6::GFP-TUB6 (see Note 7). Label cell outlines, for example, by staining with the cell wall dye PI or with the lipophilic dye FM4-64. Alternatively, a FP-fused plasma membrane marker can be introduced into the line of interest for simultaneous imaging of cell outlines and the cytoskeleton. 3. Image the samples by generating Z-stacks of the cell type of interest in two channel setting with consecutive imaging of both channels, for example, excite with a 488 nm and a 555/561 nm laser and detect emission between 490 and 555 nm and between 560 and 630 nm for detection of GFP and mCherry/RFP, respectively. Select the optimal Z distance (~0.5 μm) to cover the entire cytoskeleton without gaps in the Z direction. Image with an image resolution of 1024  1024 pixels and a pixel dwell of 1–1.7 μs. Adjust laser intensity relative to fluorescence intensities of your construct of choice. Include the upper half of the cell in the Z stack and make sure anticlinal walls are reached, but avoid to include lower half (see Notes 8 and 9). 4. Save images in meta format including all meta information, for example, lsm or similar formats. Maximum projections of Z stacks will be automatically generated during subsequent image analysis by CellBoundaryExtractor2D and CytoskeletonAnalyzer2D. 3.2 Cell Segmentation

For segmentation, the CellBoundaryExtractor2D available in MiToBo is applied, which is optimized for detection of cell outlines from 2D images with optimal contrast differences between cell

Quantification of Cytoskeleton Organization with Texture Measures

157

boundaries and background [23, 24]. For images with low quality or with complex and potentially ambiguous structures extracted boundaries are sometimes inaccurate and may require manual corrections for which functionality from ImageJ/Fiji and MiToBo can be used as outlined below (see Note 10). 1. Prerequisites: MiToBo installed in Fiji/ImageJ (see Subheading 2.2 for details); input images with two channels, one with labeled cell boundaries and the other with labeled microtubules (see Subheading 2.1 for details). 2. To perform (semi)automatic segmentation of cell contours, select “MiToBo” ! “MiToBo Runner” in the “Plugins” menu of Fiji, which will open the main window of the MiToBo operator runner. In the tree of available operators navigate to “de” ! “unihalle” ! “informatik” ! “MiToBo” ! “apps” ! “cells2D” and select the entry “CellBoundaryExtractor2D” by double click, which will open the main window of the cell boundary segmentation operator (see Fig. 2a). Configure and start the operator by performing steps 3–8. 3. Choose the mode in which the operator should run. In the “SINGLE_IMAGE” mode, a single image will be processed and all results will be shown directly in the graphical user interface. In the “BATCH” mode, images will be processed from a specified input folder (see Notes 11 and 12). Depending on the chosen operator mode, either one of the currently opened images or a folder containing all images of interest has to be defined as input for the operator. 4. Select the channel, which contains information on the labeled cell boundaries. For all analyzed images, cell boundary information must be provided in the identical channel. 5. Select the cell boundary contrast of your images, that is, specify whether the cell boundaries are darker or brighter than the image background. 6. You can provide a minimal and maximal size of valid cells measured in pixels. Extracted cell regions with an area smaller or larger than the minimal or maximal size threshold, respectively, will automatically be discarded. 7. Optionally enable the generation of images with intermediate results and supplementary visualizations by checking the box “Save/show additional results”. 8. When configuration of the operator is completed the “Run” button at the bottom of the main window will turn yellow (see Fig. 2a). Click the button to start processing. Depending on the number, size, and quality of your images and on the speed of your computer the image analysis might take several minutes.

158

Birgit Mo¨ller et al.

Fig. 2 Screenshots of CellBoundaryExtractor2D and CytoskeletonAnalyzer2D graphical user interfaces and recommended data organization. (a) CellBoundaryExtractor2D and (b) CytoskeletonAnalyzer2D both provide graphical user interfaces for configuration. See Subheadings 3.2 and 3.3 for information on configuration parameters of CellBoundaryExtractor2D and CytoskeletonAnalyzer2D, respectively. We recommend to organize input images in a folder structure shown in (c). Define a top-folder (for example, “experiment”), which contains subfolders for all individual groups, for example, different genotypes or treatments (see Note 11). Save all input images for a specific group within the respective “group” folder. The results of CellBoundaryExtractor2D and CytoskeletonAnalyzer2D will be stored in group-specific subfolders “results_segmentation“and “results_features”, respectively

9. As output, the operator generates a label image in which each extracted cell is marked in a unique gray value and the background in black, and a set of ImageJ regions of interest (ROIs) referring to the contours of the extracted cell regions for each analyzed image. In the “SINGLE_IMAGE” mode the label image is directly displayed in the GUI and the ROIs can easily be transferred from MiToBo’s result window to ImageJ’s ROI manager (https://imagej.nih.gov/ij/docs/guide/146-30. html#sub:ROI-Manager...). In the “BATCH” mode result data will be written to a new subfolder named “results_segmentation” (Fig. 2c), which is generated in each processed folder containing at least one image. The “results_segmentation” folder will contain the label images in TIF format, a “zip” file containing the ImageJ

Quantification of Cytoskeleton Organization with Texture Measures

159

ROIs, and a binary contour image where cell boundaries are marked in white and the cell areas in black (see Note 13). 10. Examine the quality of the segmentation by visual inspection of the label images in ImageJ. Alternatively, the ROI files can be opened in the ROI manager of ImageJ/Fiji and be overlaid to the input image. If you are satisfied with the result from the automatic segmentation you can directly proceed with the next step of feature extraction and clustering (see Subheading 3.3). 11. If segmentation data require manual postprocessing you can choose to either edit the label images using MiToBo (see steps 12–17), an image-editing program (step 18) or the ImageJ ROI files (step 19) as outlined below. 12. Label images may be edited with MiToBo’s “LabelImageEditor”, which is part of MiToBo and allows to easily exclude segmented regions with local inaccuracies from further analysis. The “LabelImageEditor” will display each label image in a given folder to the user and allows for removal of regions by simple mouse-click. The resulting label images will either be saved to the input folder or to a separate output folder if provided. 13. To run the label editor, select from the “Plugins” menu “MiToBo” ! “MiToBo Runner,” which will open the main window of the MiToBo operator runner. 14. Navigate to “de” ! “unihalle” ! “informatik” ! “MiToBo” ! “tools” ! “interactive” and select the entry “LabelImageEditor” by double click. This will open the main window of the label editor. 15. Select as “Input Directory” the folder in which label images are saved. All image files in the folder which are in a suitable format will be processed (see Notes 11 and 14). 16. Optionally: specify an output folder where edited label images will be saved. 17. Start the operator by pushing the “Run” button, which will open the interactive editor window displaying the first image in the given folder. To remove a region from the image just click with the left mouse button somewhere into that region. Use the “Next” button to save the current image and proceed with the next one. Once all images have been processed the operator will terminate. 18. Label images may alternatively be edited using software applications such as Gimp (https://www.gimp.org/) or Photoshop (https://www.adobe.com/products/photoshop.html) that provide a wide variety of options to manually process regions in label images. Compared to MiToBo’s “LabelImageEditor” these tools not only allow to remove labeled regions but also to

160

Birgit Mo¨ller et al.

add new ones by drawing manually defined boundaries. Refer to the documentation of your favorite software tool for details about available functionality (see Note 15). 19. ROI editing functionality of ImageJ/Fiji: To edit cell boundaries in ImageJ ROI files, open the ROI files generated by the CellBoundaryExtractor2D with ImageJ’s ROI manager. ImageJ offers a large variety of possibilities to edit polygon selections, for example, the removal of single points along the ROI (Ctrl + click), the addition of new points by splitting segments (Shift + click), the removal of complete regions, or smoothing a boundary by interpolation. More information on the various options can be found in the ImageJ documentation of the Polygon Section Tool (https://imagej.nih.gov/ij/docs/ guide/146-19.html#sub:Polygon-Selection-Tool) and the ROI Manager (see step 9). 3.3 Feature Extraction and Cluster Analysis of Texture Features

Once accurate cell boundaries are available, CytoskeletonAnalyzer2D can be applied to extract texture features from the microtubule channel of the input images for each cell. CytoskeletonAnalyzer2D is an extended version of the ActinAnalyzer2D operator [27] (see also http://mitobo.informatik.unihalle.de/index.php/Applications/ActinAnalyzer2D) and is included in MiToBo. As input, a set of images with fluorescently labeled cytoskeletal elements and the corresponding label images or ImageJ ROI files containing information on cell contours are required. CytoskeletonAnalyzer2D requires a special folder organization (Fig. 2c) with an “experiment” folder containing separate subfolders for each treatment/genotype/protein of interest, referred to as “group”. The cell boundary files, that is, the label images or ROI files, must be saved in a subfolder named “results_segmentation” within each individual “group” folder. If you use the CellBoundaryExtractor2D from MiToBo to segment cell contours as explained above, cell boundary files will automatically be saved in the correct folder. To automatically match a boundary file and a given image, both files must share the same file name prefix and label images must end with “-mask”. In the case of ROI files, the ROI file must end with “roi” or “zip” (see Note 16). To extract quantitative data, proceed as follows: 1. In Fiji, select from the “Plugins” menu “MiToBo” ! “MiToBo Runner”, which will open the MiToBo operator runner. Navigate to “de” ! “unihalle” ! “informatik” ! “MiToBo” ! “apps” ! “cytoskeleton” and select the entry “CytoskeletonAnalyzer2D” by double click, which will open the main window of the operator (see Fig. 2b). 2. Configure the operator. Details about available configuration parameters and their meaning can be found in Table 1 (see Notes 9 and 17–19).

Quantification of Cytoskeleton Organization with Texture Measures

161

Table 1 Configuration parameters of the CytoskeletonAnalyzer2D Name of parameter

Options

Image file folder

Boundary file format

Description Folder containing data for all groups of interest. All images in suitable formats (see Note 11) in all subfolders will be processed. Plots of cluster distributions will be generated for each group individually

LABEL_IMAGE Boundaries are provided as gray scale images with a unique label for each cell and a value of zero for the background. The gray values of individual cells are used as unique identifiers for the cells, that is, are applied as labels in the output plots IJ_ROIS

Boundaries are provided as ImageJ ROIs, one ROI for each cell. For analysis of a single cell within an image a file with ending “roi” is suitable, for images containing multiple cells ImageJ ROI files must be provided in “zip” format. A unique identifier per cell is derived from the order of the cell boundaries in the ROI file and a label image is saved to the output folder where cells are marked with their IDs to ease visual assessment of results

Cytoskeleton channel

Channel with the image of the labeled cytoskeleton. For all analyzed images, cytoskeleton images must be provided in the identical channel

Calculate features

Activate feature calculation (see Note 17)

Feature extractor

Feature operator to apply. For analysis of microtubule structures use the default operator “CytoskeletonFeatureExtractorLBPsRIU” using rotation invariant and uniform local binary patterns (see Note 18)

Tile size

Tile size in x and y direction for the sliding window used for feature extraction. Size should be chosen according to the resolution of the input images and the relative size of individual cells, that is, number of cells per image (see Note 9). Sizes of 16  16 or 32  32 are advisable

Tile shift

Tile shifts in x and y direction, that is, pixel distance between subsequent positions of the sliding window. If shifts are smaller than the tile size sliding windows overlap. Shifts of 8 or 16 pixels are common for tile sizes of 16  16 and 32  32

Number of feature clusters

Number of feature clusters applied in feature vector clustering. Feature vectors are clustered to automatically identify structural patterns suitable to characterize microtubule organization within the cells and each feature vector will be assigned to a cluster. The set of all feature vectors and their corresponding clusters yields a characteristic cluster distribution per cell, which forms the basis for subsequent analyses. The number of clusters should approximately match the number of structural patterns expected to occur in your images (see Note 19) (continued)

162

Birgit Mo¨ller et al.

Table 1 (continued) Name of parameter Do PCA in stage II?

Options

Description Optionally, a principal component analysis (PCA) can be applied to the extracted cluster distribution vectors and subsequent distance calculations can be restricted to the most significant principal components only. Select this option to enable PCA of the cluster distribution vectors. By default, PCA is disabled

3. If configuration of the operator is complete the “Run” mode will be activated, as indicated by a yellow color of the “Run” button (see Fig. 2b). Click the “Run” button to start processing. Depending on the number and size of your images and the computational complexity of the selected features image analysis might take a few minutes, also depending on the speed of your computer. 4. The operator generates different result data for each image and the complete data set. For each group of cells, that is, all images in a given folder, the cluster distribution per cell will be visualized as stacked bar plot. In addition, a box-whisker plot will be generated showing average distributions of all clusters and corresponding standard deviations for all groups. In the “BATCH” mode these plots are stored to disk together with additional files (Table 2, Notes 20 and 21). The text files contain the data underlying the cluster distribution plots in tab separated format, which can directly be imported in applications like R or Excel for generation of individual plots or more sophisticated data visualizations. 3.4 Data Visualization and Analysis

The interpretation and further analysis of structural cluster distributions extracted by CytoskeletonAnalyzer2D largely relies on the underlying research questions and thus cannot be generalized. Here, we will introduce how cluster distributions can be visualized, analyzed, and interpreted in CytoskeletonAnalyzer2D (step 1) or by using R (steps 2–7) and Cytoscape (steps 8–11) for generation of similarity matrices and similarity networks, respectively. 1. Visual inspection of cluster maps and distributions. CytoskeletonAnalyzer2D provides basic visualizations of extracted clusters and their distributions in terms of images in TIF and PNG format. For each processed input image, a cluster map image is generated and stored as “-clusters. tif” in the automatically created subfolder “results_features” of each group folder. A cluster map image shows the assignment of feature vectors, that is, tiles, to the structural clusters, in

Quantification of Cytoskeleton Organization with Texture Measures

163

Table 2 Files generated by CytoskeletonAnalyzer2D. “imageID” denotes the name of a single image, “groupID” names a specific group of cells located in a common folder File name

Description

-features.txt

Feature data for single image file

-features.tif

Image stack visualizing the feature data for single image

-clusterDistro.txt

Cluster distributions for each cell individually and for all cells in total of the image.

-clusters.tif

Pseudocolored image illustrating the cluster distribution per image

-distributionChart. png

Stacked bar plot of the cluster distribution for each cell of the group

AllCellsClusterStats.txt

Cluster distribution of raw data for all images and cells

AllCellsPCASubspaceStats.txt

If PCA is applied to the cluster distribution vectors this file contains the subspace feature vectors for all cells

AllCellsDistanceData.txt

Matrix of pairwise normalized Euclidean distances between cluster distribution vectors of all cells

AllGroupsDistanceData.txt

Matrix of pairwise normalized Euclidean distances between average cluster distribution vectors of all groups

AllGroupsSimilarityNetworkData. Similarity network suitable for import and visualization in Cytoscape txt

which each structural cluster is associated with a unique color and each tile is marked with the color of its corresponding cluster (Fig. 3a). The cluster map images provide information on the spatial localization of different structural patterns within a cell and can help to identify outlier cells in a group for which the cluster distribution significantly differs from all other cells. To facilitate comparison of cluster distributions between cells of different groups a “-distributionChart.png” is generated for each group of cells and saved in the “results_features” subfolder, which shows the distribution of structural clusters for each cell of the group in a stacked bar plot (Fig. 3b). In addition, a combined plot of all groups and cells is provided, named “AllCellsClusterDistributionChart.png”, which allows for direct comparison of the different groups. This plot will be saved to the given input folder. Note that this plot becomes useless if the number of cells is high and renders a proper visualization impossible. 2. Visualizing similarities and differences with heat maps. Heat maps support the visual comparison of structural properties of individual cells or groups against each other (Fig. 3c). In a heat map, pairwise differences between cluster distribution vectors of either individual cells or the average

164

Birgit Mo¨ller et al.

Fig. 3 Sample visualizations of the CytoskeletonAnalyzer2D results (a) Each subregion of all cells is assigned to a cluster, marked with unique colors for the individual clusters in the cluster image map. (b) For each cell, the distribution of cluster IDs assigned to its subregions is visualized in a bar plot where each cluster ID gets a unique color. Analysis of cluster distributions allows for direct comparison of structural properties between cells. Cells from the same group typically are characterized by group-specific cluster distributions. (c) Similarities can be visualized by pairwise distance heat maps. Each field in the map refers to the pairwise distance between feature vectors of either two cells or two groups of cells. Red colors refer to small distances and high similarity, yellow to white colors indicate large dissimilarity. (d) Visualization in similarity networks is suitable for identification of groups with comparable structural properties. Edge thickness is visualized proportional to group similarities and nodes can be colored, for example, according to individual phylogenetic groups. (c, d) modified from [22]

vectors of different groups are visualized by applying a suitable distance measure. We provide a supplemental R script, which is available from the CytoskeletonAnalyzer2D webpage http://mitobo.informatik.uni-halle.de/index.php/Applications/ CytoskeletonAnalyzer2D. It supports generation of heat maps from “AllCellsDistanceData.txt” and “AllGroupsDistanceData.txt” result data provided by CytoskeletonAnalyzer2D. In the first file a distance matrix in tab-separated format is stored where in each row for a specific cell (the identifier can be found in the first column) the Euclidean distances of its cluster distribution vector to the vectors of all other cells are

Quantification of Cytoskeleton Organization with Texture Measures

165

listed. In the second file “AllGroupsDistanceData.txt” a similar distance matrix for the distances between the average vectors of the different cell groups can be found. Note that the distances in both files are normalized to a range of [0,1] with distances of 0 representing high similarities and distances of 1 representing low similarities. To generate heat map visualizations of the data, proceed as described in steps 3–7. 3. Start RStudio (see Note 22). 4. Navigate to “File” ! “Open File. . .” in the menu bar and open the R script file “generateDistanceHeatmaps.R”. 5. Edit the script file. Change the names and locations of the files containing the distance matrices in lines 8 and 10 to the correct names and locations where they are located on your system, and likewise specify the names and paths for the heat map image files in lines 15 and 17. 6. Optional configuration of heat map plots: We use the R gplots library for generation of heat maps. The plot function “heatmap.2” of the library offers various options to configure the final outcome, for example, to perform a hierarchical clustering of data prior to heat map generation and to include the resulting cluster tree in the final plot. For common configuration options read the corresponding comments in the script file. More details can be found in the gplots library documentation (https://www.rdocumentation.org/packages/gplots/vers ions/3.0.1/topics/heatmap.2). 7. Run the R script by clicking the “Source” button in the top right corner of the editor window, which will execute the script. The two heat map plots will be saved to the locations specified in lines 15 and 17 of the script. 8. Network analysis with Cytoscape. For visualization of pairwise similarities in similarity networks (Fig. 3d) the publicly available software suite Cytoscape can be used [26]. The CytoskeletonAnalyzer2D generates a file named “AllGroupsSimilarityNetworkData.txt”, which contains a network definition that can directly be imported in Cytoscape as outlined in steps 9–11. 9. Run Cytoscape. 10. From the menu bar select “File” ! “Import” ! “Network” ! “File. . .” and select the data file. Click “Open”. 11. A new window showing the network will open in the work pane of Cytoscape. In the original network, nodes most likely overlap strongly, and the overall layout is suboptimal. The view can be adjusted by applying different layout algorithms implemented in Cytoscape. Navigate to “Layout” in the menu bar and select one of the available layouts, for example, “yFiles

166

Birgit Mo¨ller et al.

Layouts” ! “Organic” and/or move nodes manually. Cytoscape additionally offers various options to configure the layout and appearance of the network, for example, by changing the color of nodes and edges, or by defining similarity thresholds for edge weights to be included in the similarity network. Further details can be found in the Cytoscape documentation (http://manual.cytoscape.org/en/stable/).

4

Notes 1. For subsequent image analysis, cell contours and cytoskeleton patterns must be recorded in separate channels. We thus recommend combining fluorescent proteins or dyes suitable for two-channel imaging for labeling of the plasma membrane/cell wall and cytoskeleton, respectively, for example, by combining GFP and RFP/mCherry/propidium iodide/FM4-64, or CFP and YFP. 2. Prepare propidium iodide or FM4-64 stock solutions and adjust working solution and incubation time according to the specific requirements of the analyzed tissue and cell type. Use identical staining conditions for all analyzed samples. 3. Grow seedlings under controlled conditions. For comparative analysis of mutants randomize samples to rule out effects on the cytoskeleton by, for example, the position of seedlings within the growth chamber/cabinet, the daytime of imaging, or differences in handling of seedlings prior to imaging. 4. MiToBo, CellBoundaryExtractor2D, and CytoskeletonAnalyzer2D have not been extensively tested with newer Java versions so far and thus might not properly work with newer versions. 5. If you do not want to install other updates you can change their “Status/Action” to “Keep as-is”, which will leave the corresponding files unchanged. 6. Microtubule organization is sensitive to expression levels of MAPs. To analyze dose-dependent effects of a single MAP of interest on microtubule organization we recommend to compare cytoskeletal patterns between cells with low, medium, and high fluorescence intensities. To compare microtubule organization between different MAPs of interest analyze cells with similar fluorescence levels and exclude cells with the highest fluorescence intensity from your analysis as high levels of overexpression can cause localization artifacts. 7. Most available microtubule markers affect the organization and dynamics of the microtubule cytoskeleton when compared to unlabeled microtubules and their overexpression alters plant

Quantification of Cytoskeleton Organization with Texture Measures

167

growth. When using CaMV 35S or UBI10 promoter-driven microtubule markers only include lines heterozygous for the microtubule marker and without visible changes in plant morphology in your analysis. We recommend to use TUB6 or TUA5 fluorescent protein fusions expressed under their native promoters or under control of other weaker promoters. Note that the MAP4/MDB microtubule marker, which typically has a higher signal intensity at microtubules than TUA5 or TUB6, induces microtubule bundling by itself. Introduce your microtubule marker of choice into a mutant background by crossing to avoid position effects of the T-DNA cassette and recover both wild-type and the homozygous mutant from the segregating progeny for best comparability. 8. Make sure to not include the lower half of the cell as the different orientation of the cytoskeleton on the lower half may hamper optimal pattern analysis. 9. The accuracy of pattern analysis is related to the resolution of individual cells. Including too many cells in a single image will reduce the resolution of individual cells and information on cytoskeletal elements may be lost. For imaging of pavement cells in N. benthamiana, we recommend to include a maximum of three individual cells per image. Adjust laser intensity and gain to optimize signal-to-noise ratio and to facilitate comparative analysis of cytoskeletal organization. 10. You can also manually label your cell images, for example, by using the freehand tool in ImageJ combined with component labeling functionality from MiToBo. More information on how this works can be found at http://mitobo.informatik. uni-halle.de/index.php/Tips/Labeling. 11. MiToBo uses the Bioformats library (see https://www. openmicroscopy.org/bio-formats/) for reading and writing images and thus supports all image formats handled by the library. Please consult the Bioformats documentation if you encounter problems with some of your images to check if your format of choice is supported. 12. The operators for segmentation and feature extraction are optimized to operate on the same data folders. The CytoskeletonAnalyzer2D requires a defined structure of folder organization, in which image files and boundary files are deposited. We recommend to use the structure already for the CellBoundaryExtractor2D. Define a common parent folder for your experiment (Fig. 2c), which contains individual subfolders for all analyzed groups, for example, image sets for different treatments, different MAPs or different genotypes. The names of the subfolders will be used as identifiers for the different groups in the output plots. In each group folder

168

Birgit Mo¨ller et al.

CellBoundaryExtractor2D and CytoskeletonAnalyzer2D will add a subfolder “results_segmentation” and “results_features” in which all group-specific feature and cluster results will be saved. Results covering all groups, for example, the group distance matrix file, will be saved to the top-level folder. As input directory only the top folder “experiment” (Fig. 2c) needs to be selected and all subfolders will automatically be processed in a single run. 13. The label images are saved in TIF format with 32-bit integer pixel values, which can best be viewed in ImageJ/Fiji directly. Most other common image viewers are not capable of displaying the contents of such images properly. 14. The “File Filter” argument of the LabelImageEditor allows to exclude image files from processing based on their file names. If you provide a string for the “File Filter” argument only image files with names containing the specified string will be considered. For example, if the string “.tif” is provided, only image files with file ending “.tif” will be considered and other files, for example, in lsm format with ending “.lsm” will be skipped. In general, we suggest to store the label images in a folder separate from that with the original input images. 15. For label images postprocessed with external software such as Gimp, all pixels of a region must still share a unique gray value. Make sure that no interpolation is performed along the boundaries and no shading is applied during use of filling tools. Such operations will most likely result in inhomogeneous region labels and cause failure of the CytoskeletonAnalyzer2D operator. 16. The naming convention for images and corresponding files with cell boundary information is as follows: For an image named “image.tif” the operator either requires a label image named “image-mask.tif” or an ImageJ ROI file named “image. roi” or “image.zip” for a single cell or multiple cells per image, respectively, which must be saved in the specified folder. The names of the images are used as labels in the heat map visualization (see Subheading 3.4, step 2). We thus recommend to use short and meaningful file names. 17. The CytoskeletonAnalyzer2D plugin allows to skip the potentially time-consuming feature extraction step, for example, if features have already been extracted before and only the cluster analysis is to repeated with different parameters. To skip feature extraction, disable the option “Calculate features” and provide the formerly calculated feature files in a subfolder named “results_features” in each group folder. 18. Each feature operator provides individual configuration parameters that can be adjusted by the user. For details refer to the

Quantification of Cytoskeleton Organization with Texture Measures

169

MiToBo documentation of the operators describing the underlying theory of the methods. 19. If the number of clusters and expected structural patterns is too small the sensitivity of the operator will be low and only significantly different patterns with large structural variance will be identified. Vice versa, if the number of clusters is too high very specific patterns occurring only in a fraction of cells will be identified, which may result in bad generalizability. In our experience, a number of 6–10 clusters has performed well for most analyses. 20. In the “BATCH” mode results of the feature extraction stage will be saved to disk. In each processed folder a new subfolder named “results_features” will be created by the operator where all result data are stored. 21. MiToBo automatically generates a supplemental file for each result image file in TIF format with identical name and the ending “.tif.ald”. The file contains internal information on operator configurations and software versions and can usually be ignored. The information, however, might be helpful to solve problems in case you need support from the MiToBo development team. 22. We recommend to use RStudio. If you prefer to use plain R, just edit the script file with a text editor and run it from the R interpreter’s command line using the command “source generateDistanceHeatmaps.R”.

Acknowledgments This work was supported by IPB core funding (Leibniz Association) from the Federal Republic of Germany and the state of Saxony-Anhalt. References 1. Staiger CJ, Lloyd CW (1991) The plant cytoskeleton. Curr Opin Cell Biol 3:33–42 2. Wasteneys GO, Yang Z (2004) New views on the plant cytoskeleton. Plant Physiol 136:3884–3891 3. Wasteneys GO, Yang Z (2004) The cytoskeleton becomes multidisciplinary. Plant Physiol 136:3853–3854 4. Lloyd C, Hussey P (2001) Microtubuleassociated proteins in plants–why we need a map. Nat Rev Mol Cell Biol 2:40–47 5. Chakrabortty B, Blilou I, Scheres B, Mulder BM (2018) A computational framework for

cortical microtubule dynamics in realistically shaped plant cells. PLoS Comp Biol 14: e1005959 6. Wasteneys GO (2002) Microtubule organization in the green kingdom: chaos or self-order? J Cell Sci 115:1345–1354 7. Chen X, Wu S, Liu Z, Friml J (2016) Environmental and endogenous control of cortical microtubule orientation. Trends Cell Biol 26:409–419 8. Matschegewski C, Staehlke S, Birkholz H, Lange R, Beck U et al (2012) Automatic actin filament quantification of osteoblasts and their morphometric analysis on microtextured

170

Birgit Mo¨ller et al.

silicon-titanium arrays. Materials (Basel) 5:1176–1195 9. Hamada T, Nagasaki-Takeuchi N, Kato T, Fujiwara M, Sonobe S et al (2013) Purification and characterization of novel microtubuleassociated proteins from Arabidopsis cell suspension cultures. Plant Physiol 163:1804–1816 10. Akita K, Higaki T, Kutsuna N, Hasezawa S (2015) Quantitative analysis of microtubule orientation in interdigitated leaf pavement cells. Plant Signal Behav 10:e1024396 11. Sugiyama Y, Wakazaki M, Toyooka K, Fukuda H, Oda Y (2017) A novel plasma membrane-anchored protein regulates xylem cell-wall deposition through microtubuledependent lateral inhibition of Rho GTPase domains. Curr Biol 27:2522–2528 12. Belteton SA, Sawchuk MG, Donohoe BS, Scarpella E, Szymanski DB (2018) Reassessing the roles of PIN proteins and anticlinal microtubules during pavement cell morphogenesis. Plant Physiol 176:432–449 13. Armour WJ, Barton DA, Law AM, Overall RL (2015) Differential growth in periclinal and anticlinal walls during lobe formation in Arabidopsis cotyledon pavement cells. Plant Cell 27:2484–2500 14. Boudaoud A, Burian A, Borowska-Wykret D, Uyttewaal M, Wrzalik R et al (2014) FibrilTool, an ImageJ plug-in to quantify fibrillar structures in raw microscopy images. Nat Protoc 9:457–463 15. Wang XL, Zhang J, Yuan M, Ehrhardt DW, Wang ZY et al (2012) Arabidopsis microtubule destabilizing protein40 is involved in brassinosteroid regulation of hypocotyl elongation. Plant Cell 24:4012–4025 16. Liu X, Yang Q, Wang Y, Wang L, Fu Y et al (2018) Brassinosteroids regulate pavement cell growth by mediating BIN2-induced microtubule stabilization. J Exp Bot 69:1037–1049 17. Higaki T, Kutsuna N, Sano T, Kondo N, Hasezawa S (2010) Quantification and cluster analysis of actin cytoskeletal structures in plant cells: role of actin bundling in stomatal movement during diurnal cycles in Arabidopsis guard cells. Plant J 61:156–165 18. Faulkner C, Zhou J, Evrard A, Bourdais G, MacLean D et al (2017) An automated quantitative image analysis tool for the identification of microtubule patterns in plants. Traffic 18:683–693 19. Jacques E, Buytaert J, Wells DM, Lewandowski M, Bennett MJ et al (2013) MicroFilament Analyzer, an image analysis tool for quantifying fibrillar orientation, reveals

changes in microtubule organization during gravitropism. Plant J 74:1045–1058 20. Pu¨spo¨ki Z, Storath M, Sage D, Unser M (2016) Transforms and operators for directional bioimage analysis: A survey. Adv Anat Embryol Cell Biol 219:69–93 21. Ojala T, Pietikainen M, Maenpaa T (2002) Multiresolution gray-scale and rotation invariant texture classification with local binary patterns. IEEE Trans Pattern Anal Mach Learn 24:971–987 22. Bu¨rstenbinder K, Mo¨ller B, Plo¨tner R, Stamm G, Hause G et al (2017) The IQD family of calmodulin-binding proteins links calcium signaling to microtubules, membrane subdomains, and the nucleus. Plant Physiol 173:1692–1708 23. Mo¨ller B, Glaß M, Misiak D, Posch S (2016) MiToBo–A toolbox for image processing and analysis. J Open Res Software 4:e17 24. Mo¨ller B, Poeschl Y, Plo¨tner R, Bu¨rstenbinder K (2017) PaCeQuant: A tool for highthroughput quantification of pavement cell shape characteristics. Plant Physiol 175:998–1017 25. R Development Core Team (2008) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria 26. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT et al (2003) Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res 13:2498–2504 27. Mo¨ller B, Piltz E, Bley N (2014) Quantification of actin structures using unsupervised pattern analysis techniques. In: 22nd International conference on pattern recognition, IEEE, Stockholm, pp 3251–3256. https://doi.org/ 10.1109/ICPR.2014.560, Electronic ISBN: 978-1-4799-5209-0, Print ISSN: 1051-4651. http://ieeexplore.ieee.org/stamp/stamp.jsp? tp=&arnumber=6977272&isnumber= 6976709 28. Zirkel A, Lederer M, Stohr N, Pazaitis N, Hu¨ttelmaier S (2013) IGF2BP1 promotes mesenchymal cell properties and migration of tumorderived cells by enhancing the expression of LEF1 and SNAI2 (SLUG). Nucleic Acids Res 41:6618–6636 29. Cutler SR, Ehrhardt DW, Griffitts JS, Somerville CR (2000) Random GFP :: cDNA fusions enable visualization of subcellular structures in cells of Arabidopsis at a high frequency. Proc Natl Acad Sci USA 97:3718–3723 30. Nelson BK, Cai X, Nebenfu¨hr A (2007) A multicolored set of in vivo organelle markers

Quantification of Cytoskeleton Organization with Texture Measures for co-localization studies in Arabidopsis and other plants. Plant J 51:1126–1136 31. Gutierrez R, Lindeboom JJ, Paredez AR, Emons AM, Ehrhardt DW (2009) Arabidopsis cortical microtubules position cellulose synthase delivery to the plasma membrane and interact with cellulose synthase trafficking compartments. Nat Cell Biol 11:797–806 32. Abe T, Hashimoto T (2005) Altered microtubule dynamics by expression of modified alphatubulin protein causes right-handed helical growth in transgenic Arabidopsis plants. Plant J 43:191–204 33. Marc J, Granger CL, Brincat J, Fisher DD, Kao T et al (1998) A GFP-MAP4 reporter gene for

171

visualizing cortical microtubule rearrangements in living epidermal cells. Plant Cell 10:1927–1940 34. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682 35. RstudioTeam (2015) RStudio: integrated development for R. RstudioTeam, Boston, MA. http://www.rstudio.com/ 36. Shamloul M, Trusa J, Mett V, Yusibov V (2014) Optimization and utilization of Agrobacterium-mediated transient protein production in Nicotiana. J Vis Exp 86:51204

Chapter 11 Using FM Dyes to Study Endomembranes and Their Dynamics in Plants and Cell Suspensions Adriana Jelı´nkova´, Katerˇina Malı´nska´, and Jan Petra´sˇek Abstract FM (Fei-Mao) styryl dyes are commonly used for the fluorescence imaging of plasma membrane (PM) and endocytosis in vivo. Thanks to their amphiphilic character, these dyes are incorporated in the outer leaflet of the PM lipid bilayer and emit fluorescence in its hydrophobic environment. The endocytic pathway of FM dye uptake starts with rapid PM staining and continues in PM invaginations and membrane vesicles during endocytosis, followed by staining of trans-Golgi network (TGN) and ending in tonoplast (vacuolar membrane). FM dyes do not stain endoplasmic reticulum and nuclear membrane. The time-lapse fluorescence microscopy could track endocytic vesicles and characterize the rate of endocytosis in vivo. On the other hand, fixable FM dyes (FX) can be used for the visualization of particular steps in the FM dye uptake in situ. Staining with FM dyes and subsequent microscopic observations could be performed on both tissue and cellular level. Here, we describe simple procedures for the effective FM dye staining and destaining in root tip of Arabidopsis thaliana seedlings and suspension-cultured tobacco cells. Key words FM styryl dyes, Endocytosis, Plasma membrane, Endomembranes, Arabidopsis thaliana root tip, BY-2, Microscopy

1

Introduction Microscopic observation of endocytosis after labeling cells with FM dyes is one of the most frequently used methods for the tracking of endocytosis in plants [1]. FM styryl dyes were originally developed to stain synaptic vesicles in vivo [2] and later optimized for tracking of endocytosis and exocytosis in animal cells [3]. However, as documented in numerous studies, they contributed substantially also to our understanding of processes of endocytosis and plasma membrane recycling in various fungi, algae and vascular plants [1, 4]. There are in principle three important characteristics that favor some FM dyes like FM 4-64 (N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino) phenyl) hexatrienyl) pyridinium dibromide) and FM 1-43 (N-(3-triethylammoniumpropyl)-4(4-(dibutylamino) styryl) pyridinium dibromide) to be suitable

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_11, © Springer Science+Business Media, LLC, part of Springer Nature 2019

173

174

Adriana Jelı´nkova´ et al.

Fig. 1 Structures and spectral characteristics of FM 4-64 and FM 1-43 dyes [3]. (a) Comparison of head, bridge, and tail regions of FM 4-64 and FM 1-43. Positively charged head region prevents penetration of FM dyes through the PM. More double bonds in the bridge shifts fluorescence emission of FM 4-64 to the red part of the spectrum. More carbons in the tail region of FM 1-43 increase its lipophilicity in comparison with FM 4-64. (b) Excitation (dashed lines) and emission (solid lines) spectra of FM 4-64 and FM 1-43. Laser line 488 nm could be used for the excitation of both FM dyes. Data for FM dyes in water solution with CHAPS are taken from ThermoFisher Fluorescence SpectraViewer. Note that both excitation and emission spectra are pH-sensitive and might be shifted upon insertion of FM dyes into PM, as reported [1]

markers for studies of endocytosis [2]. Firstly, thanks to their positively charged head tail they cannot penetrate through the PM and could be therefore internalized exclusively by endocytotic processes. Secondly, they are intercalated efficiently into the outer leaflet of PM depending on the composition of their lipophilic tail (Fig. 1a). Moreover, in contrast to water solutions they have greatly enhanced fluorescence when intercalated into the PM, providing very good opportunity for selective observations of membranes. And thirdly, they could have various spectral characteristics

FM Dyes for Endomembrane Imaging

175

corresponding to the number of double bonds in the middle bridge region (Fig. 1a, b). Thanks to their properties, FM dyes mark all early processes of PM internalization [1], including clathrinmediated [5] and sterol endocytosis [6–8]. From early endosomes and numerous intermediate membrane vesicles, FM dyes reach several destinations that include trans-Golgi, prevacuolar compartments, vacuoles as well as exocytic vesicles that are recycled back to the PM [1]. They do not stain endoplasmic reticulum or nuclear envelope. Here we describe simple protocols for in vivo microscopic tracking of endocytosis with FM dyes in favorite models of Arabidopsis thaliana root tips and suspension-cultured cells of Nicotiana tabacum L., cv. Bright Yellow (BY-2) [9]. Under room temperature, the progression of FM dye internalization into these cells is observable already within several minutes after the addition of the dye. Therefore, preincubation of seedlings or suspension-cultured cells on ice is often needed for the effective PM labeling, synchronous onset of the FM internalization process and subsequent observation in predefined time points [7, 10]. Washing steps removing the excess of the dye applied are necessary for pulse–chase experiments [10–12], which allow to quantify the amount of internalized dye and to stain endomembranes without disturbing staining of PM. For the staining of PM in cells that are subsequently fixed, FX analogues fixable with aldehyde-based fixatives could be used (Fig. 2c, d). Using FX analogues, FM dye internalization could be studied stopped at certain time points by aldehyde fixation, that is, at the PM, early endocytic vesicles or all later steps of the endocytic machinery until its incorporation to the tonoplast in situ. It is necessary to label the cells in vivo and then fix, because fixed cells are not be labeled correctly. Unfortunately, the FX dyes cannot be used together with protocols needing solubilizations, as it is washed out of the membranes after their permeabilization by surfactant. This might be overcome by applying mCLING fluorescent probe instead of FM dyes [13].

2

Materials Prepare all water solutions using standard distilled (deionized) water; there is no need of ultrapure water. Use analytical grade chemicals and solvents. All culture media should be sterilized as well as equipment for handling with cells, seeds or seedlings (see Note 1).

2.1

Plant Material

1. 4- to 5-day-old Arabidopsis thaliana seedlings. 2. Suspension-cultured cells of Nicotiana tabacum L., cv. Bright Yellow (BY-2) cell line (see Note 2).

176

Adriana Jelı´nkova´ et al.

Fig. 2 In vivo FM 4-64 and in situ FM 4-64FX staining of Arabidopsis thaliana root tip cells. (a) Multi-well test plate with seedlings in the individual baskets incubated in 1 ml of 2 μM FM 4-64 (or FM 4-64FX) and fine tip tweezers Dumont #5. (b) 5 min FM 4-64 incubation on ice followed by internalization of the dye in root tip

FM Dyes for Endomembrane Imaging

2.2 Culture Media, Stock, and Incubation Solutions

177

1. Arabidopsis thaliana medium (AM), also known as halfstrength Murashige–Skoog (MS) medium [14]: 2.15 g/l MS basal salt mixture, 1% (w/v) sucrose, pH 5.8. For 1 l, weigh 2.15 g MS salt mixture, transfer it into glass or plastic beaker with around 800 ml of distilled water and add 10 g sucrose (see Note 3). Stir thoroughly with magnetic stirrer at room temperature, adjust pH with 3 M KOH to 5.8 and bring volume to 1 l. For solidified medium, add agar (10 g/l (w/v); final concentration 1%). Autoclave at 121
C for 20 min under 0.1 MPa in laboratory glass bottles (see Note 4). Allow the agar medium to cool to 45–50
C (until the container can be held with hands) and pour aseptically into sterile plastic square plates to cover approximately half of the depth of each plate. Leave the plates opened in the flow bench for 30 min to allow the agar to solidify. 2. Modified MS [14] medium for suspension-cultured tobacco cells: 4.3 g/l of MS basal salt mixture, 3% (w/v) sucrose, 100 mg/l myoinositol, 1 mg/l thiamine, 200 mg/l KH2PO4 and 0.2 mg/l (0.9 μM) 2,4-D, pH 5.8. For 1 l, weigh 4.3 g MS salt mixture and transfer it into glass or plastic beaker with around 800 ml of distilled water, add 30 g sucrose, 100 mg myoinositol, 200 mg KH2PO4, 1 ml stock solution of thiamine and 2 ml stock solution of 2,4-D (see Note 3). Stir thoroughly with magnetic stirrer at room temperature, adjust pH with 3 M KOH to 5.8 and bring volume to 1 l. Autoclave at 121
C for 20 min under 0.1 MPa (see Note 4). 3. Stock solution of thiamine (3.3 mM): for 100 ml dissolve 100 mg of thiamine in 100 ml of distilled water. Store in 2 ml aliquots in plastic tubes in 20
C. 4. Stock solution of 2,4-D (0.45 mM): for 100 ml dissolve 10 mg of 2,4-D in 1 ml pure ethanol. After addition of distilled water to the final volume 100 ml, heat the solution under continuous stirring for at least 4 h to dissolve 2,4-D completely. 5. Seed sterilization solution: for 10 ml mix 5 ml of 50% household bleach (containing 5% sodium hypochlorite) and 5 ml of

ä Fig. 2 (continued) epidermal cells after 5, 10, 20 and 30 min at the room temperature in AM. After rapid PM staining we are able to observe first endosomes as early as 5 min after transferring the seedlings to the room temperature. Later endosomes appear 10–20 min and after 30 min FM dye stains TGN (displayed as bigger endosomes). Scale bars, 10 μm. (c) Retention of the FM 4-64FX dye at the PM after 1 h fixation in 4% PFA on ice followed by 20 min incubation in AM medium at the room temperature. There are first endosomes attached to the PM and PM invaginations observed. In contrast to FM 4-64 dye, which is not retained at the PM and endosomes are observed. Scale bar 10 μm. (d) Longitudinal section of the Arabidopsis thaliana root tip after FM 4-64FX staining at the room temperature before and PFA fixation for 1 h. After 20 min of the FM incubation all the cell tissue layers throughout the central longitudinal section of the root tip are stained and fixation does not disturb the pattern. Scale bar 10 μm

178

Adriana Jelı´nkova´ et al.

sterile water, add 5 μl 0.05% (v/v) Tween® 20. Do not store this solution, always prepare fresh. 6. Sterile distilled (deionized) water (autoclaved at 121
C for 20 min under 0.1 MPa). 7. Stock solutions of FM 4-64, FM 4-64FX and FM 1-43: dissolve 1 mg of lyophilized powder in DMSO or water to make 1–20 mM stock solution (see Note 5). Store aliquots at 80
C, keep in darkness. Alternatively, FM dyes are available as 100 μg lyophilized aliquots (Molecular Probes,). FM dyes are light-sensitive and unstable at room temperature. Therefore, during experiments keep aliquots on ice and ideally in darkness. 8. Fixative: prepare 4% paraformaldehyde fixative solution by diluting 32% PFA (Electron Microscopy Sciences) in MTSB (50 mm PIPES, 5 mm EGTA, 5 mm MgSO4, pH 6.8 adjusted with 3 M KOH). 2.3 Equipment for In Vitro Cultivation, Staining Procedure, and Fluorescence Microscopy

1. Autoclave. 2. Laminar flow cabinet for handling cells and seedlings during the inoculation and sampling. 3. Sterilized standard laboratory equipment: glass Erlenmeyer flasks (250 or 100 ml) covered with aluminum foil for the incubation of cell suspensions, pipette for 1–5 ml for handling with cell suspensions (with cut tips), orbital incubator placed at 27
C in darkness in the culture room or any air-conditioned incubator with shaker, square sterile plastic plates, 24-well test plates, pipette for micro-volumes, medium incubation baskets (e.g., Intavis). 4. Microscope slides (76  26  1 mm) and cover glasses of thickness No. 1 (18  18 mm or 22  60 mm) or microscopy chambers in 8-well format (chambered cover glass for inverted and chamber slide for upright microscopes (see Note 6). Note that for upright microscopes are microscopy chambers suitable for long working distance objectives only. 5. Fluorescence microscope equipped with sensitive camera (usually CCD or sCMOS) for acquiring weak fluorescent signals, apochromatic objectives 40 or 60 with high numerical aperture (NA 1.2 water immersion or 1.4 oil immersion) and long-pass or band-pass filters for the fluorescence excitation at 488 nm and emission at 505–550 nm (for FM 1-43 and FM 2-10) or excitation at 514 nm (or 488 nm or 561 nm) and emission >570 nm (for FM 4-64, FM 4-64FX, and FM 5-95). Ideally, confocal microscope system (laser scanning or spinning disc) with GaAsP or PMT detectors (laser scanning microscope, e.g., Zeiss LSM 880 or Leica SP8) or sCMOS/CCD/ EM-CCD cameras (laser spinning disc microscope, e.g.,

FM Dyes for Endomembrane Imaging

179

Yokogawa spinning disc unit-equipped systems) equipped with filters or beam splitters for abovementioned spectral ranges of FM dye excitations and emissions. Software for image acquisition (supplied by microscope system producer). 6. The postprocessing of acquired data can be performed in the software of the microscope supplier. However, often open source software is used (e.g., open source Image analysis software ImageJ, National Institutes of Health, https://imagej. nih.gov/ij/) or even more suitable ImageJ distribution called Fiji (https://fiji.sc/) that has already integrated Bio-Formats plugin for import life sciences formats including images of all common microscopy producers (Zeiss.czi files; Nikon .nd2 files etc.; ref. 15).

3

Methods

3.1 Tracking Endocytosis in Root Cortex and Epidermal Cells of A. thaliana Seedlings in vivo

1. In the laminar flow cabinet put seeds of Arabidopsis thaliana into a microcentrifuge tube (small amount is enough, not more than several hundreds), add 1 ml of freshly made sterilization solution and incubate 10 min with occasional manual agitation. Remove sterilization solution with 1 ml pipette (seeds settle well at the bottom of the tube placed in the stand). Add 1 ml of sterile water and incubate for 10 min, repeat this washing step three times (3  10 min). 2. Plant individual seeds onto the agar square plates. Use sterile toothpick or plant seeds with 1 ml pipette. Upon suction of sterilized seeds in water (or 0.1% cooled top agar), individual seeds might be planted one by one in rows by pipette. Close plates with Parafilm and place them at 4
C for 48 h for stratification (in darkness). Keep vertically oriented plates at 16/8 h light/dark photoperiod at 21/18
C for 4–5 days. 3. Prepare a 24-well test plate with 1 ml aliquots of liquid AM medium with 2 μM FM dye by the addition of 0.1 μl of 20 mM FM dye stock solution (see Note 7) into 1 ml AM medium (see Note 8) and 1 ml aliquots of plain AM medium for rinsing the seedlings. Place the test plate on ice for 15 min. 4. Transfer seedlings into medium incubation baskets placed in the individual wells (5–8 seedlings/well) with fine tip tweezers (Dumont #5 or similar) and incubate them for 5 min on ice. Wash the seedling three times in liquid AM medium, still on ice (Fig. 2a). All the procedures, incubations and washing are performed with the seedlings placed in the baskets.

180

Adriana Jelı´nkova´ et al.

5. Place seedlings on the microscope slide into the drop of AM medium (room temperature), close with cover glass and start the timer (see Note 9). Shifting seedling into the room temperature triggers all processes of plasma membrane internalization, thus representing the time 0 for the FM uptake experiment. 6. Observe the fluorescence at predefined time points (e.g., 5, 10, 20, and 30 min) using epifluorescence microscope equipped with confocal laser scanning unit. To obtain detailed images of individual cells within the root epidermis and cortex, use 40 or 60 C-apochromatic water or oil immersion objectives (see Note 9). Fluorescence of all major FM dyes (FM 1-43, FM 4-64, FM 4-64FX, FM 2-10, and FM 5-95) could be stimulated with laser line 488 nm. Alternatively, FM 4-64, FM 4-64FX and FM 5-95 could be excited with longer wavelengths (typically with 514 or 561 nm), which might be useful for parallel observation of GFP (excited with 488 nm). The fluorescence emission is detected between 505 nm and 550 nm for FM 1-43 and FM 2-10 and above 575 nm for, FM 4-64, FM 4-64FX and FM 2-10 dyes. Sequential scanning, that is, using two timely isolated single excitations and emissions, can be used to avoid any cross talk of fluorescence channels in double labeling experiments with GFP for FM 4-64, FM 4-64FX, and FM 2-10. To separate FM 1-43 and GFP fluorescence linear unmixing method might be used on systems equipped with spectral detection unit (e.g., Zeiss LSM 880; see Note 10). FM dyes are quite photostable, which makes them suitable for long time experiments. Both speed and distribution patterns of FM dye internalization into root epidermal cells during selected uptake period might be used to characterize functions of individual elements of endocytic machinery. After rapid PM staining, first endosomes appear 5 min after transferring the seedlings to the room temperature, after 10 and 20 min later endosomes appear and after 30 min FM dye stains TGN (Fig. 2b). In such a time-lapse fluorescence microscopy, sequential tracking of endocytosis, trans-Golgi network (TGN) and incorporation into the vacuolar system in vivo, could be observed. 7. Acquired images could be analyzed with image analysis software (Subheading 2.3, item 6) to determine the amount of FM dye internalized during the experimental time window. Typically, mean fluorescence intensity in the area spanning whole cytoplasm is measured and expressed in relation to the average fluorescence at the PM [12]. FM uptake experiment should be performed in at least three biological repetitions to obtain enough data for statistical evaluation.

FM Dyes for Endomembrane Imaging

3.2 Staining of the PM with Fixable FM 4-64FX Dye in the Root Tip of A. thaliana Seedlings in situ

181

1. For in situ staining of the PM of the epidermal and cortex cells of the root tip follow the procedure Subheading 3.1, until step 5. 2. Place the seedlings into the baskets in the ice-cold aldehydebased fixative solution (4% PFA in MTSB) and leave to fix on ice for 1 h. 3. Wash the seedling three times in liquid AM medium. The FM 4-64FX is retained at the PM even 20 min after transferring seedlings to the room temperature (Fig. 2c, see Note 11). 4. For in situ staining of PM of the all tissue cell layers of the central longitudinal section of the root tip, follow steps 1 and 2 of Subheading 3.1, then transfer seedlings into individual baskets in the wells (5–8 seedlings/basket) with fine tip tweezers and incubate 20 min in 2 μM FM 4-64FX liquid AM medium at the room temperature to allow penetration of the FM dye into the deeper tissues of the root tip (Fig. 2d). 5. Fix the seedlings by transferring baskets into the ice-cold aldehyde based fixative solution (4% PFA in MTSB) and leave to fix for 1 h on ice. 6. Wash the seedling three times in liquid AM medium at the room temperature. 7. Place seedlings on the microscope slide into the drop of AM medium, close with cover glass and observe the fluorescence using epifluorescence microscope equipped with confocal laser scanning unit. To obtain detailed images of individual cells within the root epidermis and cortex, use 40 or 60 C-apochromatic water or oil immersion objectives (see Note 9). Fluorescence of the FM 4-64FX can be stimulated with laser line 488 nm, 514 nm or 561 nm. The fluorescence emission is above 575 nm. FM 4-64FX is photostable and can be thus used for fine Z-scanning method to farther make precise 3D reconstructions of the cells (root tissues, etc.) as the roots are fixed and do not grow out of the focus.

3.3 Tracking Endocytosis in Tobacco Cell Suspension BY-2 cells—Simple Labeling Protocol

1. Inoculate 1 ml of 7-day-old BY-2 cells into 30 ml of modified MS cultivation medium (see Note 12). Use pipette with tips with cut top to transfer cells under sterile conditions into the fresh medium. The laminar flow cabinet should be sterilized with UV for 20 min in advance. Cultivate cells in darkness under continuous shaking at 27
C on an orbital incubator at 150 rpm (orbital diameter 30 mm; see Note 13). The length of cultivation depends on the purpose of the experiment, but exponentially growing cells are the most suitable (see Note 2). 2. Pipet 1 ml aliquots of cell culture into 24-well test plate. The number of aliquots depends on the experimental setup.

182

Adriana Jelı´nkova´ et al.

3. Add 0.1 μl of 20 mM FM dye stock solution into 1 ml of cell culture (one well) to reach 2 μM final concentration (see Notes 7 and 8), place test plate on orbital shaker at 26–27
C and start the timer. 4. Observe the fluorescence at predefined time points by following instructions described in steps 6–7 of Subheading 3.1. For multi-time point experiments we preferred to use microscopy chamber in multiwell setup (e.g., 8-wells) (see Notes 6 and 9). For FM 4-64, PM of BY-2 cells is stained within 2–3 min, endocytosed endosomes appear immediately afterward followed by the gradual appearance of the dye in the transGolgi, prevacuolar compartments and some other endomembrane structures (15–45 min), but never ER and nuclear envelope. The dye might be applied for longer time and used for effective staining of tonoplast. 3.4 Tracking Endocytosis in Tobacco Cell Suspension BY-2 Cells—Cold Pretreatment Labeling Protocol

1. Prepare cell culture according to Subheading 3.3, step 1. 2. Pipette 1 ml aliquots of cell culture into 1.5 ml plastic tube. The number of aliquots (tubes) depends on the experimental setup. Place tubes on ice for 15 min. 3. Still on ice, add 0.1 μl of 20 mM FM dye stock solution into 1 ml of cell culture (i.e., into one tube) to reach 2 μM final concentration (see Notes 7 and 8). Keep tubes on ice for 15 min; there is no need for shaking during this period. 4. Place tubes horizontally on orbital shaker (150 rpm) at 26–27
C (fix them with the tape) and start the timer. 5. Observe the fluorescence at predefined time points by following instructions in Subheading 3.1, steps 6 and 7 and Subheading 3.2, step 7. Take one drop of cell culture with dropper and make the preparation using microscope slide and cover glass or chamber in multiwell setup and start the timer (see Notes 6 and 9). Initial cold pretreatment of cells reducing more than 90% of the dye uptake [10] helps to standardize the initial steps of staining by saturating PM with the dye. BY-2 cells stained by FM 4-64 on ice are shown 1 min after transfer to RT, when only PM is stained (Fig. 3a, e, i). After few minutes first endocytosed endosomes appear (Fig. 3b, f, j) and shows their development after 15 min. Endosome staining is followed by the gradual appearance of the dye in the transGolgi, prevacuolar compartments after 30 min and some other endomembrane structures, but never ER and nuclear envelope (Fig. 3c, g, k). The dye might be applied for a longer time (here 36 h) and used for effective staining of tonoplast (Fig. 2d, h, i).

FM Dyes for Endomembrane Imaging

183

Fig. 3 FM 4-64 internalization in tobacco BY-2 cells. (a, e, i) Staining of the PM after 1 min; (b, f, j) early endosomes after 15 min; (c, g, k) TGN and other endomembranes after 45 min and (d, h, i) vacuolar system after 40 h. (a–c) One confocal cross section; (d–f) detailed view of dashed frame and (i–l) maximum intensity projection of a confocal z-stack are presented. Scale bars 20 μm (a–c, i–l) and 5 μm (e–h) 3.5 Tracking Endocytosis in Tobacco Cell Suspension BY-2 Cells—Pulse–Chase Labeling

1. Prepare cell culture according to Subheading 3.3, step 1. 2. Pipet 1 ml cell culture aliquots into 1.5 ml plastic tubes. The number of aliquots (tubes) depends on the experimental setup. Place tubes on ice for 15 min. 3. Still on ice, add 0.1 μl of 20 mM FM dye stock solution into 1 ml of cell culture (i.e., into one tube) to reach 2 μM final concentration (see Notes 7 and 8). Keep tubes on ice for 15 min; there is no need for shaking during this period.

184

Adriana Jelı´nkova´ et al.

4. Spin tube with cells in ice-cold BY-2 medium (1600  g, 4
C, 5 min) to remove medium containing the dye. Repeat the washing step three times. 5. Place tubes horizontally on orbital shaker at 26–27
C (fix them with the tape) and start the timer. 6. Observe the fluorescence at predefined time points (see Subheading 3.3, step 4; Notes 6 and 9). Removal of excess dye is useful mainly for the staining of later phases of the dye uptake [16] without disturbing PM fluorescence, but it also helps to observe early phases of endocytosis.

4

Notes 1. The sterilization by autoclaving of materials and solutions for in vitro cultivation of cell cultures and Arabidopsis seedlings is necessary. Although the application of FM dyes and subsequent microscopic tracking of endocytosis is not performed aseptically, for handling and incubation of cells or seedlings it is suggested to use sterile disposable plastic (tips, multiwell plates, etc.). This approach helps to standardize the whole technique. Note that even barely detectable contamination could interfere with the growth of cell cultures and with the progression of endocytosis. 2. For the reproducibility of results it is absolutely necessary to work with high quality cell suspensions. Tobacco BY-2 cell suspensions are normally yellowish, dense cultures at the end of subculture interval that normally takes 7 days. FM-staining might be performed throughout this whole period; however, exponentially growing cells (day 2–4 old) have the fastest progression of endocytosis. Avoid using cell suspensions that do not grow optimally. This could be monitored without opening the culture flask by checking cells using an inverted microscope during the first 3 days of cultivation. In general, any other suspension-grown plant cell cultures can be used for labeling with FM dyes. 3. MS salts are hygroscopic; the bottle should not be left opened unless necessary. For the storage of MS salts at 4
C, seal the cap carefully with Parafilm. 4. When autoclaving liquids always use the program containing slow release of the pressure at the end of the run. Autoclaves not equipped with this option are not suitable for this purpose. Always keep the lid loose to allow the pressure equalization between the bottle and autoclave chamber. Avoid repetitive autoclaving of any media.

FM Dyes for Endomembrane Imaging

185

5. All FM dyes are easily soluble in DMSO. For most applications, 20 mM stock solution is appropriate. With limited solubility, they could be also dissolved in water as reported for 17 mM water stock solution of FM 4-64 [1]. Since DMSO as an organic solvent interferes with the composition of biomembranes, its concentration must be kept as low as possible, lower than 0.1% (v/v) [1]. 6. For observing BY-2 cells the usage of microscopy chambers is beneficial. Chambered cover glass is designed for inverted microscopes in 1–8-well setup. For multi-time point experiments 8-well chambered cover glass are preferentially used, each well for single time observation. The chambers can be washed in water and used repeatedly. Note that chambered slides for upright microscopes are suitable for long working distance objectives only. Therefore FM-dye tracking with 40 or 60 objective is not convenient. Alternatively, time-lapse observations might be performed in perfusion system for open cultivation (e.g., POC cell cultivation system) (http://www. pecon.biz/?page_id¼283). 7. Keep the FM dye stock solutions at least 15 min at the room temperature before the start of the experiment as they are dissolved in DMSO (higher melting point, 19
C). 8. Working concentrations of FM dyes ranges from 1 μM to 50 μM. However, it is suggested to keep the concentration as low as possible (1–2 μM). Since FM dyes are very lipophilic and easily integrated into the PM, in higher concentrations they interfere with endomembrane dynamics and induce other unwanted effects like dragging of some integral PM proteins out of the PM [11]. The most frequently used FM dyes in plant research are FM 4-64 [1] and FM 1-43 [10]. For both of them, slightly less lipophilic versions exist, such as FM 5-95 and FM 2-10 (Molecular Probes Handbook, https://www. thermofisher.com/cz/en/home/references/molecularprobes-the-handbook.html), respectively. Less lipophilic dyes are suitable in cases, when the endocytotic processes are too fast and upon the addition of the dye the uptake is too fast. Then, the less lipophilic variant allows slower uptake, and also, washing out these dyes is easier. In addition, the dynamics of FM dye internalization is much slower when using water instead of MS medium for all staining and washing steps. Therefore, it is suggested to use MS medium for all experiments. For DMSO-dissolved FM dyes, do not forget to add DMSO into mock treatments. 9. Seedlings of Arabidopsis thaliana placed at the slide might be completely covered with cover glass. However, cotyledons might also remain uncovered. For microscopy with oil

186

Adriana Jelı´nkova´ et al.

immersion objectives it is possible to attach cover glass to the slide with paper tape to avoid undesired movement of cover glass. Always prepare the microscope (finding the focal plane, setting the cameras, etc.) and image acquisition software before imaging experimental sample. The progression of endocytosis is very fast, and there is usually not enough time for various settings. 10. Since the emission spectra of FM 1-43 and GFP overlap, the separation of these fluorochromes might be performed on confocal systems with spectral detection unit [11]. Linear unmixing of emission spectra acquired using specimens with two fluorochromes must be performed after acquiring reference spectra of samples with only FM 1-43 or GFP. 11. After 1 h fixation of the seedlings on ice with 4% PFA, followed by 20 min at the room temperature in AM, the FM 4-64FX dye is still retained at the PM (note the early endocytic compartments attached to the PM) in compare to the FM 4-64 dye where at the room temperature the dye propagates into the cells and the endocytic compartments are formed (Fig. 2c). 12. The inoculation density is critical for the successful growth of suspension-cultured cells. For routine propagation of BY-2 cells with a 7-day-long subculture interval, stationary cells are diluted (at the day 7) 1:50 (2 ml/100 ml of medium in 250 ml Erlenmeyer flask) or 1:30 (1 ml/30 ml of medium in 100 ml Erlenmeyer flask). Never fill flasks with more medium than specified above. 13. The temperature during cultivation of cells should not exceed 27–28
C for BY-2 cells. Cells are also very sensitive to prolonged standing without shaking. Try to minimize this time as much as possible.

Acknowledgments We acknowledge the Imaging Facility of the Institute of Experimental Botany CAS supported by the MEYS CR LM2015062 Czech-BioImaging and CZ.02.1.01/0.0/0.0/16_013/0001775 and Operational Programme Prague Competitiveness (OPPC) CZ.2.16/3.1.00/21519. References 1. Bolte S, Talbot C, Boutte Y, Catrice O, Read ND et al (2004) FM-dyes as experimental probes for dissecting vesicle trafficking in living plant cells. J Microsc 214:159–173

2. Betz WJ, Bewick GS (1992) Optical analysis of synaptic vesicle recycling at the frog neuromuscular junction. Science 255:200–203

FM Dyes for Endomembrane Imaging 3. Betz WJ, Mao F, Smith CB (1996) Imaging exocytosis and endocytosis. Curr Opin Neurobiol 6:365–371 4. Sˇamajova´ O, Taka´cˇ T, von Wangenheim D, Stelzer E, Sˇamaj J (2012) Update on methods and techniques to study endocytosis in plants. In: Sˇamaj J (ed) Endocytosis in plants. Springer, Berlin, Heidelberg, pp 1–36 5. Dhonukshe P, Aniento F, Hwang I, Robinson DG, Mravec J et al (2007) Clathrin-mediated constitutive endocytosis of PIN auxin efflux carriers in Arabidopsis. Curr Biol 17:520–527 6. Ueda T, Yamaguchi M, Uchimiya H, Nakano A (2001) Ara6, a plant-unique novel type Rab GTPase, functions in the endocytic pathway of Arabidopsis thaliana. EMBO J 20:4730–4741 7. Grebe M, Xu J, Mobius W, Ueda T, Nakano A et al (2003) Arabidopsis sterol endocytosis involves actin-mediated trafficking via ARA6positive early endosomes. Curr Biol 13:1378–1387 8. Klima A, Foissner I (2008) FM dyes label sterol-rich plasma membrane domains and are internalized independently of the cytoskeleton in characean internodal cells. Plant Cell Physiol 49:1508–1521 9. Nagata T, Nemoto Y, Hasezawa S (1992) Tobacco BY-2 cell-line as the Hela-cell in the cell biology of higher plants. Int Rev Cytol 132:1–30

187

10. Emans N, Zimmermann S, Fischer R (2002) Uptake of a fluorescent marker in plant cells is sensitive to brefeldin A and wortmannin. Plant Cell 14:71–86 11. Jelı´nkova´ A, Malı´nska´ K, Simon S, KleineVehn J, Parezova´ M et al (2010) Probing plant membranes with FM dyes: tracking, dragging or blocking? Plant J 61:883–892 12. Robert S, Kleine-Vehn J, Barbez E, Sauer M, Paciorek T et al (2010) ABP1 mediates auxin inhibition of clathrin-dependent endocytosis in Arabidopsis. Cell 143:111–121 13. Revelo NH, Kamin D, Truckenbrodt S, Wong AB, Reuter-Jessen K et al (2014) A new probe for super-resolution imaging of membranes elucidates trafficking pathways. J Cell Biol 205:591–606 14. Murashige T, Skoog F (1962) A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol Plant 15:473–497 15. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Iongair M et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682 16. Kutsuna N, Hasezawa S (2002) Dynamic organization of vacuolar and microtubule structures during cell cycle progression in synchronized tobacco BY-2 cells. Plant Cell Physiol 43:965–973

Chapter 12 Transient Gene Expression as a Tool to Monitor and Manipulate the Levels of Acidic Phospholipids in Plant Cells Lise C. Noack, Prˇemysl Pejchar, Juraj Sekeresˇ, Yvon Jaillais, and Martin Potocky´ Abstract Anionic phospholipids represent only minor fraction of cell membranes lipids but they are critically important for many membrane-related processes, including membrane identity, charge, shape, the generation of second messengers, and the recruitment of peripheral proteins. The main anionic phospholipids of the plasma membrane are phosphoinositides phosphatidylinositol 4-phosphate (PI4P), phosphatidylinositol 4,5-bisphosphate (PI4,5P2), phosphatidylserine (PS), and phosphatidic acid (PA). Recent insights in the understanding of the nature of protein–phospholipid interactions enabled the design of genetically encoded fluorescent molecular probes that can interact with various phospholipids in a specific manner allowing their imaging in live cells. Here, we describe the use of transiently transformed plant cells to study phospholipiddependent membrane recruitment. Key words Microscopy, Nicotiana benthamiana, Nicotiana tabacum, Phosphoinositides, Phospholipid-binding domains, Pollen tube, Transient expression

1

Introduction All living cells are surrounded by membranes, which help to define their spatial identity and create the semipermeable boundary between intracellular and extracellular space [1]. Typical cellular membrane is composed of a bilayer of lipids and proteins, whose organization and interactions are crucial for its function as organizing platforms for cellular processes. Historically, cellular membrane studies were once dominated by a protein-centric view, where proteins executed majority of membrane-related functions and the membrane lipids were often regarded only as passive players whose role was to provide structural support for bilayer formation [2]. It is now generally accepted that both lipids and proteins play indispensable active roles in the various functions of cellular membranes [3].

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_12, © Springer Science+Business Media, LLC, part of Springer Nature 2019

189

190

Lise C. Noack et al.

Among plant plasma membrane lipids, negatively charged (anionic) phospholipids, phosphoinositides like phosphatidylinositol 4-phosphate (PI4P) and phosphatidylinositol 4,5-bisphosphate (PI4,5P2), together with phosphatidylserine (PS) and phosphatidic acid (PA), constitute low-abundant but essential component [4–6]. They possess many important roles, which include defining membrane identity, generation of downstream signaling molecules, generating membrane negative charge, modulating membrane curvature, and creating binding sites for the targeting of effector proteins [7–12]. The realization of the important roles of anionic phospholipids has created a need for methods that would enable their noninvasive spatiotemporal monitoring in living cells. The identification and characterization of protein modules that specifically bind to various anionic phospholipids led to the idea that these protein modules might be used to detect phospholipids in living cells. This resulted in the development of genetically encoded phospholipid sensors that consist of the specific phospholipid-binding domains (either single or in tandem) fused with various fluorescent proteins, enabling the live cell imaging of phospholipid dynamics [13–15]. Especially in the past decade, this approach has been successfully used in plants to generate sensors for a wide variety of phospholipids including PI4P [16–18], PI4,5P2 [17, 19, 20], PA [21, 22], and PS [17, 22]. Concomitantly, many enzymes involved in the production and degradation of anionic lipids were identified and the phenotypes of their knockout or overexpressing mutant lines described (for plants see for example refs. 23, 24). This also led to the development of tools allowing the manipulation of phospholipid levels in the cell upon generic or targeted overexpression of phospholipid-modifying enzymes [25]. The recruitment of phospholipid-binding peripheral proteins to cell membranes is essential for many cellular processes. The targeting of proteins to specific phospholipids or to the membranes of particular lipid compositions, mediated by lipid-binding domains, allows their recruitment to be precisely controlled in spatiotemporal fashion. Despite the manifold biological consequences associated with the targeted recruitment of peripheral protein to their target membranes, only a basic understanding of the interactions of proteins with membrane surfaces exists because these questions are inaccessible by commonly used structural techniques [26, 27]. Therefore, the selective colocalization of peripheral proteins (or individual protein domains, protein deletions, point mutations etc.) with the particular lipid marker together with its relocalization after coexpression with corresponding phospholipid-modifying enzymes may bring valuable information about the nature of protein–membrane interface. Transient gene expression approaches are particularly beneficial, since they enable quick screening of many proteins or protein variants and allow for

Exploring Phospholipid Signalling in Plant Cells Using Transient Gene Expression

191

easy manipulation of expression level. Here, we describe the protocols allowing transient coexpression of proteins of interest with genetically encoded lipid markers or phospholipid-modifying enzymes in two different plant cell types: biolistics-mediated transformation of growing tobacco pollen tubes (that show spatially separated plasma membrane domains enriched with distinct phospholipids, refs. 21, 22), and agroinfiltration of Nicotiana benthamiana leaf epidermal cells (where high transformation efficiency can be achieved).

2

Materials Prepare all solutions using ultrapure water and store at room temperature (RT), unless stated otherwise.

2.1 Particle Bombardment Solutions

1. Gold particles: resuspend 30 mg of 1.6 μm gold microcarriers (Bio-Rad, #1652264, see Note 1) in 1 ml absolute ethanol, vortex vigorously for 3 min, and spin down at table top centrifuge (1 min, maximum speed). Wash twice with H2O and resuspend in 1 ml of 50% glycerol (sterile). Store at 4
C. 2. 2.5 M CaCl2: dissolve 3.675 g CaCl2 · 2H2O (Sigma, #C7902) in 10 ml H2O. Filter-sterilize and keep 1 ml aliquots at 20
C. Working aliquot can be kept at 4
C for several months. 3. Protamine: dissolve 10 mg of protamine sulfate (Sigma, #P4505) in 10 ml H2O. Filter-sterilize and keep 0.5 ml aliquots at 20
C. Working aliquot can be kept at 4
C for several weeks.

2.2

DNA

1. For tobacco pollen particle bombardment, dilute DNA stock with H2O to 0.25–1 μg/μl working solution. In order to achieve good transformation frequency and expression levels, target construct expression must be driven by promoters active in pollen (typically LAT52p or UBQ10p). 35S promoter is not recommended. Usually, clean miniprep is enough for several transformations. Store at 20
C. 2. For N. benthamiana leaves infiltration, use DNA plasmid concentrated at 0.25–1 μg/μl to transform Agrobacterium. Expression of the construct is usually driven by UBQ10 or 35S promoters.

2.3 Biological Materials

1. Pollen: flowers of outdoor- or glasshouse-grown tobacco plants (N. tabacum cv. Samsun) are collected in warm and dry weather conditions before opening; anthers are taken out and kept in laboratory conditions on a filter paper for 1 day to let anthers open and dehydrate (anthers might be surface sterilized and dried in the laminar box to harvest sterile pollen). Dried

192

Lise C. Noack et al.

pollen grains are sifted through to remove anthers. Harvested pollen can be kept frozen at 20
C without apparent loss of the germination capacity for several years. 1 mg of pollen is usually used per transformation. 2. N. benthamiana leaves: use leaves from 2 to 3 weeks old plants. For infiltration, select leaves that are heart shaped. If the plant has already flowered, it is too late to perform the infiltration. 3. Agrobacterium: the electrocompetent A. tumefaciens C58pmp90 strain is used for tobacco leaf infiltration. 2.4

Cultivation Media

1. 2 pollen tube medium (2PTM): 10% w/v sucrose, 25% w/v PEG-6000, 2 mM CaCl2, 2 mM KCl, 1.6 mM MgSO4, 3.2 mM H3BO3, 60 μM CuSO4, 0.06% w/v casein acid-hydrolysate, 0.6% w/v MES, pH 5.9. Prepare 10 stock solution for salts (1470 mg/l CaCl2 · 2H2O, 746 mg/l KCl, 1972 mg/l MgSO4 · 7H2O, 989 mg/l H3BO3, 75 mg/l CuSO4 · 5H2O) and 50 stock for casein hydrolysate (1.5% w/v). Store the stocks at 20
C. Dissolve appropriate amounts of sucrose, MES, salt, and casein hydrolysate stocks in H2O, adjust pH to 5.9 with 4 M KOH, add PEG-6000 and make up to the final volume with H2O. Store 50 ml aliquots at 20
C. 2. Solid pollen tube medium: 1PTM solidified with 0.25% Phytagel (see Note 2). Thaw 50 ml aliquot of 2PTM and warm it in water bath to at least 60
C. Prepare 0.5% Phytagel (Sigma, #P8169) solution: weigh 0.25 g of Phytagel into 50 ml H2O and resuspend. Dissolve Phytagel by carefully heating up to the boiling point while stirring. Mix 2PTM and Phytagel solutions and keep the 1PTM/0.25% Phytagel medium in hot water bath. In laminar box, prepare plates with solidified medium by rapid pouring of 4 ml of hot PTM/Phytagel solution onto 5 cm petri dishes. Let dry and store up to 1 month at 4
C. 3. LB medium: Agrobacterium are grown on LB liquid medium (Difco™ LB Broth, Lennox, #240230, 20 g/l) and LB plate (Difco™ LB Broth, Lennox, #214010, 20 g/l, 15% w/v agar, Difco™ Bacto Agar). 4. Infiltration medium: 10 mM MES (Sigma-Aldrich, #D8250250G), 10 mM MgCl2 (Sigma-Aldrich, #M2670), 0.15 mM acetosyringone (Sigma-Aldrich, #D134406). Prepare a stock solution of 100 mM MES. Weight and dissolve the appropriate amount of MES in H2O and adjust the pH to 5.7 using KOH solution. Autoclave the MES solution for 30 min and store it at RT. Prepare a stock solution of acetosyringone at 100 mM in EtOH and store it at 20
C. Just before performing the infiltration, dilute the MES solution in H2O, add the MgCl2 and the acetosyringone to a final concentration of 10 mM and 0.15 mM, respectively.

Exploring Phospholipid Signalling in Plant Cells Using Transient Gene Expression

3

193

Methods Carry out all procedures at RT, unless otherwise specified (see Note 3).

3.1 DNA Macrocarrier Preparation for Pollen Particle Bombardment

1. Add following to the 1.5 ml microcentrifuge tube to prepare one macrocarrier (the sample should be mixing continuously, keep sequence and timing): 25 μl of gold particles suspension, vortex 1 min, add 2.5–7 μl of plasmid DNA (~1–10 μg, see Note 4), 25 μl of 2.5 M CaCl2, and 10 μl of protamine (1 mg/ml) solution. 2. Vortex vigorously for at least 3 min. Spin down in tabletop centrifuge for 30 s at max speed. Remove the supernatant carefully using yellow pipette tip or vacuum and discard it. 3. Resuspend the pellet completely (see Note 5) in 200 μl of absolute ethanol and vortex for 3 min. Spin down again and discard the supernatant. 4. Resuspend the pellet in 18 μl of absolute ethanol, vortex for 1 min and load the suspension on the macrocarrier (Bio-Rad, #1652335). Keep the suspension dispersed by constant pipetting. For future manipulations, it is better to have macrocarrier fitted in the steel macrocarrier holder (Bio-Rad, #1652322) before loading. 5. Let the macrocarrier dry in a vibration-free environment. One should obtain evenly distributed layer of gold particles without any visible clumps. Although dried macrocarriers can be stored in dry chamber at RT for several hours, it is better to perform the transformation immediately after preparation.

3.2

Pollen Plating

1. For one transformation, resuspend 1 mg of tobacco pollen per 5 ml of 1PTM. Pour the suspension on the prewetted nylon 47 mm (0.8 μm) filter disc (Whatman, #Z746282) placed on filtration apparatus (e.g., Millipore, #XX1004720) and remove medium using vacuum. 2. Transfer the pollen to the solidified PTM by placing the filter disc upside-down briefly. Repeat for the next DNA sample and/or proceed to the particle bombardment immediately (see Note 6).

3.3 Particle Bombardment

1. Set up the PDS-1000/He system (Bio-Rad, #1652257) according to the instruction manual using standard settings. 2. Put the rupture disc (1100 psi, Bio-Rad, #1652329) in place and tighten gently with the screwdriver. 3. Assemble macrocarrier and stopping screen (Bio-Rad, #1652336) into microcarrier launch assembly and insert it into the uppermost position.

194

Lise C. Noack et al.

4. Use the second free slot from above for the pollen plated on solidified medium. 5. Evacuate the chamber to 28 inHg and perform the bombardment. 6. Release the vacuum immediately, seal the sample plate with Parafilm and store it at RT. 3.4 Agrobacterium Transformation

1. For each construct, add 1 μl of DNA plasmid into 50 μl of electrocompetent Agrobacterium on ice. 2. Transfer the Agrobacterium into cold 1 mm wide electroporation chamber (Eurogentec, #CE00150). Put the electroporation chamber into the MicropulserTM (Bio-Rad, #165-2100) and give a pulse of 2 kV, 335 Ω, 15 μF, for 5 ms. 3. Add 1 ml of liquid LB medium and transfer the bacteria into a new tube and incubate them at 29
C for at least 2 h. 4. Plate the Agrobacterium onto LB plates containing the appropriate antibiotics to select the Agrobacterium strain (gentamycin 20 μg/ml and rifampicin 50 μg/ml) and the target construct. Incubate the plate at 29
C for 48 h.

3.5 Nicotiana benthamiana Infiltration

1. Scoop transformed Agrobacterium from the transformation plate with a tip and resuspend the bacteria into 2 ml infiltration medium (see Subheading 2.4) by pipetting. Measure the OD600 using a spectrophotometer (Biophotometer, Eppendorf). Adjust the OD600 to 1 by adding infiltration medium. For coinfiltration of several constructs, mix the same quantity of each transformed Agrobacterium to obtain a final OD600 of 1. 2. Using 1 ml syringe (Terumo, #125162229), press the infiltration solution with the Agrobacterium onto the abaxial side of the chosen tobacco leaf keeping your finger on the other side of the leaf. The solution must spread into the leaf (see Note 7). 3. Mark the place where the infiltration has been made with a permanent marker. Put the plant back to the growth chamber for 2–3 days.

3.6 Data Acquisition, Analysis, and Quantification

Images can be exported from the microscope-specific acquisition software and analyzed with suitable analysis software. We use Fiji for this purpose, which is a distribution of well-known software ImageJ [28, 29], bundling a lot of plugins which facilitate scientific image analysis, and which is freely available at https://fiji.sc/. A number of basic and advanced tools are available within this software, including subtraction of background, and measurements of intensities, both based on the definition of a region of interest. 1. Particle-transformed pollen tubes: for the initial evaluation of protein overexpression on pollen tube growth and polarity,

Exploring Phospholipid Signalling in Plant Cells Using Transient Gene Expression

195

Fig. 1 Elevation of PI4,5P2 levels by the overexpression of the CFP-tagged PI4P-5 kinase (PI4P5K) results in increased NtSEC3a:YFP recruitment to the plasma membrane in tobacco pollen tubes. Pollen tubes with comparably low expression level of NtSEC3a:YFP were selected. For coexpression with PI4P5K, cells expressing high levels of PI4P5K:CFP (not shown) and displaying characteristic PI4PK5-overexpression phenotype were selected. White dashed lines mark the site of intensity profiles and yellow dotted lines indicate length of the membrane fluorescence signal. Micrographs are shown using a color intensity code in order to display local enrichment of the YFP fluorescence

observe the cells 12–24 h after transformation with 5–10  lenses. Identify transformed cells based on FP fluorescence and take images using the same acquisition settings. Mean pollen tube length, pollen tube width, tip swelling and cell “curviness” (calculated as the ratio of the distance between the pollen grain/pollen tube tip and the pollen tube length, is close to 1 for straight pollen tubes) are good parameters for the initial quantitative assessment. Several simple measures can be used to monitor the binding of protein of interest to the plasma membrane (e.g., measuring of the membrane- and cytoplasmicassociated intensities from the line scan (Fig. 1, see also refs. 21, 30) and calculating the ratio as a proxy for membrane recruitment index, and/or measuring the length of membrane signal (in the case of asymmetric localizations, see Fig. 1)). For the quantitative assessment of colocalization, Pearson or Spearman rank correlation coefficients can be calculated from the data (e.g., with Coloc 2 plugin available in Fiji). 2. Confocal observation of N. benthamiana leaves: cut 5 mm2 regions of the leaf that surround the place where the infiltration has been made. Place the piece of leaf into water between slide and coverslip with the abaxial side of the leaf facing the coverslip. It may be convenient to tape the slide and coverslip together to maintain the coverslip on the slide as the leaf sample is thick. Using the appropriate wavelength, an epifluorescent microscope and the smallest objective (10), screen the

196

Lise C. Noack et al.

Fig. 2 Organelle-targeted phosphatase as a way to locally interfere with acidic phospholipids. (a) The top panel represents a schematic representation of a possible construct for organelle-specific targeting of a lipid modifying enzyme. Such system include an organelle-specific targeting anchor/protein (which may be at the N- or C-terminal end of the synthetic chimeric protein), a fluorescent protein to verify localization specificity, and the isolated catalytic domain of a lipid modifying enzyme. An important criterion for the use of the catalytic domain is that it should be free of any endogenous targeting capacity. The bottom panel represents an example of construct for the specific depletion of PI4P at the plasma membrane (i.e., MP-Sac1). (b) Schematic representation of MP-Sac1 action at the PM but not endosome and (c) effect of MP-Sac1 on PI4P accumulation. In the control condition (left, for example with expression of a catalytically dead MP-Sac1 enzyme), there is much more PI4P at the PM than in endosomes and as a result, a PI4P biosensor such as P4M is localized preferentially at the PM. Upon expression of MP-Sac1 (right), the pool of PI4P at the PM is reduced, which triggers the redistribution of the P4M PI4P sensor to both PM and endosomes

surface of the leaf to find the transformed cells. Then, switch to confocal microscope and 63 objective to look at the subcellular localization of the fluorescent protein. 3. Analysis of the effect of PM-targeted phosphatases on anionic phospholipid localization: in order to perturb anionic phospholipid production with subcellular accuracy, it is possible to target an isolated phosphatase (or kinase) domain to a specific organelle. For example, the 4-phosphatase SAC domain of the yeast Sac1p protein is targeted to the PM using a myristoylation and palmitoylation anchor (MP) (ref. 18; Fig. 2). This MP-Sac1 construct was fused to a mTurquoise2 (mTu2) protein to monitor protein localization (MP-mTu2-Sac1) in order to verify that this synthetic enzyme was indeed efficiently targeted to the PM. Cotransfection with genetically encoded anionic phospholipid sensors allowed to determine the effect of the 4-phosphatase activity on the production of a given phospholipid. Quantification of the effect of the phosphatase activity may be performed using the analyses mentioned above

Exploring Phospholipid Signalling in Plant Cells Using Transient Gene Expression

197

for pollen tubes. Typically, three behaviors may be anticipated for the biosensors following coexpression with an organelle targeted phosphatase: (1) no effect, (2) redistribution of the sensor from a membrane to a cytosolic pool, and (3) redistribution of the sensor to a different organelle. For example, MP-Sac1 expression induced the redistribution of PI4P biosensors from the PM to endosomes (Fig. 2, see also refs. 18, 22, 31). This can be quantified qualitatively, as a percentage of cells with endosomal labeling by the PI4P sensor, as compared to the total number of cells analyzed. It can also be quantified by making a ratio of membrane vs. soluble signal, but this later quantification method is difficult given the reduced cytoplasm of N. benthamiana leaf cells. Once validated, such heterologous transient assay may be used to probe the importance of a given lipid for targeting a protein. It may also be used to validate in vivo catalytic activity of a phosphoinositide phosphatase of unknown specificity.

4

Notes 1. Different sizes of particles (0.6 or 1.0 μm) may be also used; this will however affect the amount of coatable DNA. Alternatively, cheaper tungsten particles may be used, their size distribution is however much more variable, resulting in yet greater variability in expression levels. 2. Do not use agar or agarose as they would cause the precipitation of pollen tube medium. 3. For the details of PDS-1000/He assembly and operation, consult the PDS-1000/He Particle Delivery System Instruction Manual (http://www.bio-rad.com/webroot/web/pdf/ lsr/literature/M1652249.pdf) and watch the YouTube tutorial (https://www.youtube.com/watch?v¼dfD95gsEdrg&t¼90s). 4. When transforming with more than one construct, premix the DNA before coating. Use only small amount (0.5–1 μg of plasmids expressing phospholipid markers to prevent the perturbation of phospholipid signaling due to overexpression of lipid-binding domain). 5. This is crucial for obtaining good transformation frequency. The more DNA is added the longer it takes to resuspend the pellet completely. 6. We routinely transform up to 12 plates in a row. If more transformations are needed, split the plating/bombardment into batches of ten. 7. The infiltration might not work if the stomata are closed. To get around this problem, make small holes with a needle.

198

Lise C. Noack et al.

Acknowledgments Research in the Prague lab is supported by the Czech Science Foundation (grants no. 17-27477S, 18-18290J and 19-21758S) and by the Ministry of Education Youth and Sport of the Czech Republic (project no. NPUI LO1417). Y.J. is funded by ERC no. 3363360-APPL under FP/2007-2013, and L.C.N is funded by a fellowship from the French Ministry of Higher Education. References 1. Bernardino de la Serna J, Schu¨tz GJ, Eggeling C, Cebecauer M (2016) There is no simple model of the plasma membrane organization. Front Cell Dev Biol 4:106 2. Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175:720–731 3. Nicolson GL (2014) The fluid-mosaic model of membrane structure: still relevant to understanding the structure, function and dynamics of biological membranes after more than 40 years. Biochim Biophys Acta 1838:1451–1466 4. Devaiah SP, Roth MR, Baughman E, Li M, Tamura P et al (2006) Quantitative profiling of polar glycerolipid species from organs of wild-type Arabidopsis and a PHOSPHOLIPASE Dα1 knockout mutant. Phytochemistry 67:1907–1924 5. Mosblech A, Ko¨nig S, Stenzel I, Grzeganek P, Feussner I et al (2008) Phosphoinositide and inositolpolyphosphate signalling in defense responses of Arabidopsis thaliana challenged by mechanical wounding. Mol Plant 1:249–261 6. Furt F, Simon-Plas F, Mongrand S (2011) Lipids of the plant plasma membrane. In: Murphy AS, Schulz B, Peer W (eds) The plant plasma membrane. Springer, Berlin Heidelberg, pp 3–30 7. Balla T (2013) Phosphoinositides: tiny lipids with giant impact on cell regulation. Physiol Rev 93:1019–1137 8. Kay JG, Grinstein S (2013) Phosphatidylserine-mediated cellular signaling. In: Capelluto D (ed) Lipid-mediated protein signaling. Springer, Dordrecht, pp 177–193 ˇ a´rsky´ V, 9. Sekeresˇ J, Pleskot R, Pejchar P, Z Potocky´ M (2015) The song of lipids and proteins: dynamic lipid-protein interfaces in the regulation of plant cell polarity at different scales. J Exp Bot 66:1587–1598 10. Noack LC, Jaillais Y (2017) Precision targeting by phosphoinositides: how PIs direct

endomembrane trafficking in plants. Curr Opin Plant Biol 40:22–33 11. Pokotylo I, Kravets V, Martinec J, Ruelland E (2018) The phosphatidic acid paradox: too many actions for one molecule class? Lessons from plants. Prog Lipid Res 71:43–53 12. Tanguy E, Kassas N, Vitale N (2018) Protein–phospholipid interaction motifs: a focus on phosphatidic acid. Biomolecules 8:20 13. Vermeer JEM, Munnik T (2010) Imaging lipids in living plants. In: Munnik T (ed) Lipid signaling in plants. Springer, Berlin Heidelberg, pp 185–199 14. Platre MP, Jaillais Y (2016) Guidelines for the use of protein domains in acidic phospholipid imaging. In: Waugh MG (ed) Lipid signaling protocols. Springer, New York, pp 175–194 15. Va´rnai P, Gulya´s G, To´th DJ, Sohn M, Sengupta N et al (2017) Quantifying lipid changes in various membrane compartments using lipid binding protein domains. Cell Calcium 64:72–82 16. Vermeer JEM, Thole JM, Goedhart J, Nielsen E, Munnik T et al (2009) Imaging phosphatidylinositol 4-phosphate dynamics in living plant cells. Plant J 57:356–372 17. Simon MLA, Platre MP, Assil S, van Wijk R, Chen WY et al (2014) A multi-colour/multiaffinity marker set to visualize phosphoinositide dynamics in Arabidopsis. Plant J 77:322–337 18. Simon MLA, Platre MP, Marque`s-Bueno MM, Armengot L, Stanislas T et al (2016) A PtdIns (4)P-driven electrostatic field controls cell membrane identity and signalling in plants. Nature Plants 2:16089 19. Kost B, Lemichez E, Spielhofer P, Hong Y, Tolias K et al (1999) Rac homologues and compartmentalized phosphatidylinositol 4, 5-bisphosphate act in a common pathway to regulate polar pollen tube growth. J Cell Biol 145:317–330 20. van Leeuwen W, Vermeer JEM, Gadella TWJ, Munnik T (2007) Visualization of

Exploring Phospholipid Signalling in Plant Cells Using Transient Gene Expression phosphatidylinositol 4,5-bisphosphate in the plasma membrane of suspension-cultured tobacco BY-2 cells and whole Arabidopsis seedlings. Plant J 52:1014–1026 21. Potocky´ M, Pleskot R, Pejchar P, Vitale N, Kost B et al (2014) Live-cell imaging of phosphatidic acid dynamics in pollen tubes visualized by Spo20p-derived biosensor. New Phytol 203:483–494 22. Platre MP, Noack LC, Doumane M, Bayle V, Simon MLA et al (2018) A combinatorial lipid code shapes the electrostatic landscape of plant endomembranes. Dev Cell 45:465–480 23. Heilmann I (2016) Phosphoinositide signaling in plant development. Development 143:2044–2055 24. Yao HY, Xue HW (2018) Phosphatidic acid (PA) plays key roles regulating plant development and stress responses. J Integr Plant Biol 60(9):851–863 25. Idevall-Hagren O, De Camilli P (2015) Detection and manipulation of phosphoinositides. Biochim Biophys Acta 1851:736–745 26. Pu M, Orr A, Redfield AG, Roberts MF (2010) Defining specific lipid binding sites for a

199

peripheral membrane protein in situ using subtesla field-cycling NMR. J Biol Chem 285:26916–26922 ˇ a´rsky´ V, 27. Pleskot R, Cwiklik L, Jungwirth P, Z Potocky´ M (2015) Membrane targeting of the yeast exocyst complex. Biochim Biophys Acta 1848:1481–1489 28. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M et al (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682 29. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9:671–675 30. Sekeresˇ J, Pejchar P, Sˇantru˚cˇek J, ˇ a´rsky´ V et al (2017) Analysis Vukasˇinovic´ N, Z of exocyst subunit EXO70 family reveals distinct membrane polar domains in tobacco pollen tubes. Plant Physiol 173:1659–1675 31. Gronnier J, Crowet JM, Habenstein B, Nasir MN, Bayle V et al (2017) Structural basis for plant plasma membrane protein dynamics and organization into functional nanodomains. elife 6:e26404

Chapter 13 The Photoconvertible Fluorescent Protein Dendra2 Tag as a Tool to Investigate Intracellular Protein Dynamics Alexandra Lesˇkova´, Zuzana Kusa´, Ma´ria Labajova´, Miroslav Krausko, and Ja´n Ja´sik Abstract Fluorescence proteins changing spectral properties after exposure to light with a specific wavelength have recently become outstanding aids in the study of intracellular protein dynamics. Herein we show using Arabidopsis SYNAPTOTAGMIN 1 as a model protein that the Dendra2 green to red photoconvertible protein tag in combination with confocal scanning laser microscopy is a useful tool to study membrane protein intracellular dynamics. Key words Photoconvertible fluorescence proteins, Dendra2, Photoactivated localization microscopy, Arabidopsis SYT1

1

Introduction The latest innovations in fluorescent protein technology combined with superresolution microscopy and advanced imaging techniques enable successful study of protein intracellular dynamics. Since discovery of the green fluorescent protein (GFP), over a hundred naturally occurring DNA sequences encoding proteins analogous to GFP have been isolated. Moreover, the original fluorescence proteins (FPs) have been genetically engineered, and there are now numerous modified versions available with improved suitability for live imaging of tagged proteins [1, 2]. An interesting new class of FPs developed recently is characterized by their ability to change reversibly or irreversibly their emission spectra (photoconvertible FP, PCFP). Other FPs can be photoactivated and photoswitched. The conversion is mostly accomplished by illuminating the samples with light at the specific wavelength. Unfortunately, naturally occurring PCFPs have a tendency to form oligomeric complexes and this significantly limits their application in protein visualization because fusion proteins

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_13, © Springer Science+Business Media, LLC, part of Springer Nature 2019

201

202

Alexandra Lesˇkova´ et al.

may be mislocalized and nonfunctional. Site-directed mutagenesis of PCFPs has solved some problems in their oligomerization, slow maturation and low fluorescence intensity [3–5]. Over the last decade also plant researchers have recognized the potential of these unique proteins and employed some in photoactivated localization microscopy [6]. Kaede was the first PCFP employed in plant research. Its fusion to mitochondrial markers first enabled study of mitochondrial fusion and fission [7], and then tracking [8]. Tobacco BY-2 cells expressing Kaede fused to vacuolar and Golgi markers were utilized to explore the plant secretory pathway [9]. Kaede was also suitable for monitoring protein homooligomerization by Fo¨rster resonance energy transfer [10]. The advantage of this technique is that it requires transformation with only a single plasmid fusion construct. Finally Kaede was used in analysis of auxin-efflux transporter ATP-BINDING CASSETTE B4 turnover in the plasma membrane [11]. Internalisation and recycling of another auxin-efflux carrier PIN-FORMED2 (PIN2) has been estimated by EosFP [12]. Several organelle markers labeled with monomeric mEosFP were prepared by Malthur et al. [13] and these were used in chloroplast [14, 15], peroxisomes [16, 17] and mitochondria investigation [18]. The same laboratory also used this tag to explore the actin cytoskeleton [19] and changes in living-cell nuclear DNA content [20]. Dendra, and especially its monomeric Dendra2 variant which was developed in Lukyanov’s laboratory [4] aided solution of several plant cell problems. The PIN-FORMED1 (PIN1) auxin transporter labeled with Dendra2 was investigated in transiently transformed Arabidopsis leaf pavement cells [21]. We prepared Dendra2 fusion with the homologous auxin transporter PIN2. Expression was driven by native promoter and stably transformed Arabidopsis lines were used to solve problems in PIN2 dynamics [22–24]. Dendra2 was also successfully used to demonstrate directional passage of some macromolecules through plasmodesmata in Physcomitrella patens protonemata [25] and to verify the mobility of the transcription factors between Arabidopsis root cells [26]. Finally nuclear influx and efflux of RSZp22 splicing factor was studied with this tool in tobacco leaf cells [27]. Other photoconvertible and photoswitchable fluorescent proteins have only occasionally been employed in plant research. Monomeric mMaple should be more photostable and matures considerably faster than mEosFP [6]. Another PCFP mKikGR was used to distinguish between two nuclei during double fertilisation [28]. Photoswitchable DRONPA-s fused to ARABIDOPSIS THALIANA GLYCINE-RICH RNA-BINDING PROTEIN 7 was employed to confirm this protein’s nucleocytoplasmic shuttling [29] and more recently to demonstrate cell-type-specific differences in total protein efflux and symplasmic connectivity between Arabidopsis roots cells [30]. An interesting possibility is

Monitoring Intracellular Protein Dynamics with Dendra2

203

provided by “fluorescent timers” which change their emission properties over time. They can be useful for monitoring the fate of the proteins in cells when conversion is sufficiently slow. However this tool has been rarely adapted in plant research. We used the DsRed-E5 timer in SYNAPTOTAGMIN 2 tracking [31]. As already mentioned our studies showed that the Dendra2 tag was fruitful in investigating the PIN2 auxin efflux carrier dynamics. We challenged the correctness of using the Brefeldin A (BFA) fungal lactone in studying PIN2 protein endocytosis [23], and we established that PIN2 endocytosis was not inhibited by auxins in stable transgenic plant roots which express the pPIN2-PIN2-Dendra2 construct [24]. Herein we summarize our experiences with this probe and provide a protocol for using Dendra2 tag in the study of membrane protein intracellular dynamics.

2

Materials 1. Standard chemicals, enzymes, and tools for recombinant DNA preparation, DNA analysis, and documentation, cultivation media and equipment for bacteria transformation and cultivation (see Note 1). 2. Round petri dishes (94
16 mm), plant growth medium (see Note 2), equipment for plant cultivation and work in sterile conditions (see Note 3). 3. For seedling manipulation and cultivation during imaging: glass slides and coverslips, square petri dishes (Greiner dishes, 120
120
17 mm), timer, microwave, fine forceps. 4. Fluorescent stereomicroscope with filters for GFP. 5. Confocal laser scanning microscope equipped with lasers, dichroic mirrors, filters and fluorescent filter cubes enabling observation and imaging of standard green and red fluorescent proteins. The microscope must be equipped with both a 405 nm laser line and lamp and filter cube allowing illumination of samples with light around 400 nm (see Note 4). 6. Image J software (https://rsbweb.nih.gov/ij).

3

Methods

3.1 Translation Fusion Construct Preparation and Creation of Transgenic Lines

1. Prepare binary plasmid encompassing plant specific expression cassette containing DNA encoding gene of interest, CDS for Dendra2, and suitable regulatory sequences (see Note 5). 2. Transform Arabidopsis plants by Agrobacterium tumefaciens (see Note 6). 3. Select homozygous plants with single T-DNA insert (see Note 7).

204

Alexandra Lesˇkova´ et al.

3.2 Seeds Sterilization and Seedling Cultivation

1. Sterilize seed surface for 30 s with 70% ethanol and then for 20 min with 1.5% sodium hypochlorite (Sigma) containing 0.05% Tween 20 detergent. Wash seeds several times with sterile Millipore water. 2. Prepare cultivation medium for Arabidopsis seedlings and pour it into petri dishes (see Note 2). 3. Sow Arabidopsis seeds on medium under sterile conditions and transfer the agar plates to the refrigerator for 2 days (see Note 8). 4. Move plates into a growth cabinet, position them vertically and germinate seedlings under a regime of 12 h light (100 μmol · m2 s1 light intensity, 22  C) and 12 h dark (19  C) for 4–5 days (see Note 9). 5. Twelve hours prior to the experiment, transfer the seedlings onto plates with fresh agar medium (see Note 10).

3.3 Setting Up Photoconversion of Whole Root Tips

For general considerations applying to this step (see Notes 11 and 12). 1. Prepare the cultivation slides for the experiment. Melt the cultivation medium and distribute 1.5 ml on microscopic slides and then wait a few minutes until the medium solidifies (see Note 13). 2. Select five seedlings from homozygous population under a fluorescence stereomicroscope equipped with filter for GFP which show similar expression of fusion protein in their roots. Place seedlings on microscope slide with cultivation medium so that roots are parallel to each other and root tips at the same level. 3. Image plant roots expressing fusion protein in the green channel with a 20
objective (Uplan FI, 0.50 NA) with 4
internal zoom and frame size of 512
800 pixels. Excite the samples with the 488 nm laser line (see Note 14) and collect the signal with a 505–525 nm band-pass filter and 405/488/543 nm beam splitter (see Note 15). Optimize the imaging conditions to achieve minimal background but wide signal dynamic range. Use as low laser power as possible to avoid bleaching the protein, and examine the images in the Hi-Lo look-up table to avoid pixel saturation. You are imaging live samples, so you must set up a reasonably fast acquisition time (3–5 s) to prevent capturing shifted images in x, y, or z directions. Acquire several frames of the same image to make sure the fluorescent protein is not bleached by the set-up. 4. Switch to 40
objective (see Note 16) and focus on the surface of the root tip which should be situated at the center of the field of view. Illuminate the roots with a 100 W mercury lamp of an epifluorescence microscope for 20 s (see Note 17) using a filter cube allowing for excitation of Dendra2 with light around 400 nm (see Note 18).

Monitoring Intracellular Protein Dynamics with Dendra2

205

5. Switch again to 20
objective and image the roots in the red channel with the same zoom and frame as for the green channel. Excite roots with the 543-nm line of laser (see Note 19) and collect the fluorescence with a 560–620 nm band-pass filter (see Note 20) and 405/488/543 beam splitter. Optimise the imaging conditions for capturing good quality images within reasonable time and without bleaching the protein (as described in Subheading 3.3, step 3). 6. Select new roots, illuminate them with epifluorescence microscope using 40
objective (see Subheading 3.3, step 4) for a few seconds and image root tips in multitrack mode with 20
objective in both green and red channels; sequentially and switching between lines. Capturing conditions should be set as they were optimized (see Subheading 3.3, steps 3 and 5). Illuminate the roots again for a few seconds with the 40
objective, and repeat several times. Image the roots in both channels after each illumination pulse using the same settings. Measure the absolute fluorescence intensities of both the green and red signal with the ImageJ program using line modus (see below). Plot the fluorescence intensities against the period of illumination and determine the optimal time required to photoconvert the protein (see Note 21). The example obtained with selected SYT1-Dendra2 line showing duration of illumination dependent decrease in signal intensity of the green variety of SYT1-Dendra2 in the plasma membrane and concomitant increase in abundance of the red variety is displayed in Fig. 1c. 3.4 Study of Protein Turnover in the TimeLapse Experiment

1. Prepare the cultivation slides for the experiment as described in Subheading 3.3, step 1. Add compounds of interest to the medium, recheck and readjust the pH if needed (see Note 22). 2. Select five seedlings with similar expression of fusion protein (see Subheading 3.3, step 2). 3. Image the root tips with the 20
objective in multitrack mode in both green and red channels. To capture the signal of green unconverted protein, excite the samples with the 488 nm laser line and collect the signal with a 505–525 nm band-pass filter. For red signal, excite the samples with the 543-nm laser and collect the fluorescence with a 560–620 nm band-pass filter. Use 405/488/543 nm beam splitter and acquire the signals in sequential mode and switching between lines. Optimize the acquisition settings by reaching a compromise between image quality and the time needed for capture. This is described in Subheading 3.3, step 3; see also Note 23 for an example of the settings.

206

Alexandra Lesˇkova´ et al.

Fig. 1 (a) Distribution of SYT1-Dendra2 in Arabidopsis root epidermis. Root tips were imaged simultaneously in the green and the red channel before and after photoconversion. Photoconversion was carried out with 100-W Hg lamp for 20 s using LD C-Apochromat 40
/1.1 W Corr M27 objective and BP 390/40 excitation filter cube. The arrow points to the plasma membrane and the arrowhead to the endomembrane system for which emission spectra shown in (b) were measured. For this we employed a Zeiss LSM 880 confocal microscope and 40
objective mentioned above. Fluorescence of nonconverted samples was excited at 458 nm, fluorescence of converted samples at 514 nm and both were detected with a Quasar multichannel spectral

Monitoring Intracellular Protein Dynamics with Dendra2

207

4. Switch to the 40
objective and illuminate the roots by 100 W mercury lamp of an epifluorescence microscope using a suitable filter cube for the optimal period you determined for your microscope. 5. Switch to 20
objective and image the roots again in both channels using identical imaging parameters as before conversion. The laser power and detector settings can be set up in a way that both the background signal of the red channel and the average fluorescence intensity of the red signal after conversion are equal to, or very close to that of green signal before conversion. This setup allows for direct visual and quantitative comparison of changes in the green and red signals over time. 6. Image the roots after conversion in both channels at different time points under identical acquisition settings. Remove the coverslip after each imaging (see Note 24). The pattern of SYT1-Dendra2 turnover in roots cells is shown in Fig. 1d. 7. To avoid the agar medium drying-out on the microscope slides between imaging, especially in longer-experiment times from several hours to 24 h, place the slides in humidity chambers (see Note 25). 8. Quantify the green and red signal intensity in the samples using ImageJ software. To measure membrane fluorescence intensities, use the straight line selection mode with line width completely covering the membrane, and extract the average intensities for each plasma membrane for both channels. Analogically, for example, you can analyze cell structures or cytoplasm areas with free hand selection mode. 9. If the roots have significant differences in signal intensities, you should express your results as relative values. In single roots, the means of the green signal intensities at each experiment time point are normalized to the mean of green signal ä Fig. 1 (continued) detector. Fluorescence intensities were quantified using Zeiss Zen software. Each point on the graph with SE bar represents the mean of 30 ROIs in the plasma membrane or the intracellular space between nucleus and the plasma membrane. (c) Fluorescence intensities of the green and red varieties of SYT1-Dendra2 in the plasma membrane after photoconversion with different light pulses. Samples were illuminated for different period with a 100-W Hg lamp using a BP 360–370 filter cube and a 40
UPLSAPO Super Apochromat (0.90) objective of an Olympus FV1000 confocal laser scanning microscope. Intensities of green and red fluorescence signals were assessed with the ImageJ software and each point on the graph with SE bar represents the mean of 50 membranes. Note that the standard Olympus BP 360–370 filter cube is not optimal for photoconversion of Dendra2 and better results are achieved using filters with a band pass around 400 nm. (d) Merged images of the green and red channel of the same root expressing SYT1-Dendra2 fusion captured at different time points in the time-lapse experiment. Notice the decrease of red signal intensity and increase of green signal intensity in the plasma membranes within 15 h after photoconversion when compared with intensities measured shortly after photoconversion and before photoconversion respectively. Bars ¼ 10 μm

208

Alexandra Lesˇkova´ et al.

intensities emitted by the membranes of the same root before conversion. Time point “0” is the relative green signal intensity recorded immediately after conversion. Analogically, in single roots the means of membrane red signal intensities at each time point in the trial are related to the mean red fluorescence intensities measured immediately after conversion (time point “0”). Values of the time points “0” are set to “1”. 3.5 Local Photoconversion of a Defined ROI

1. Select five seedlings with similar expression levels using the fluorescent stereo microscope equipped with GFP filter, and transfer them to preprepared microscopic slides with cultivation medium. 2. Set up three channel acquisition mode; where the first is set for the converting light, the second for the green and the third for the red form of Dendra2. For the first channel, set excitation with a 405 nm LD laser line and 505–525 emission range. For the second channel, select 488 nm laser line and 505–525 nm emission filter, and for the third 543 nm laser line and 560–620 nm emission range. Select 405/488/543 dichroic mirror. 3. Activate only the channels for green and red form of Dendra2. Restrict the area of the scan to a few root cells by imaging the root tips with 40
objective and 6
internal zoom. Acquire the preconversion image in sequential mode and scan by switching between lines. Scan with low laser power and fast acquisition mode (see Note 26). Save the acquisition settings. 4. Deactivate the red channel, activate only the green channel and zoom into the plant structure/membrane/organelle you wish to photoconvert in live acquisition mode. Select the ROI in the magnified region with the point-scan selection tool (see Note 27). 5. Deactivate the green channel and activate only the first 405 nm excitation channel. Set up low laser power and scan the ROI continuously for 5–20 s with a long exposure time (see Note 28). 6. Reload the acquisition settings you used to capture Dendra2 before photoconversion. Deactivate the first channel and activate the green and red channel. Make a postconversion image. Examine the extent and the size of photoconversion in the ROI and in neighboring cell structures (see Note 29). Adjust the imaging settings for the red signal, if needed, and repeat the conversion with those settings on a different cell. 7. To examine the diffusion and any other movement of the converted red variety of Dendra2 and the green Dendra2 toward the photoconverted area, image the root cells repeatedly at different time points (see Note 30). It is generally not necessary to remove the coverslip between imaging for up to 30 min. Analyze the captured images with the ImageJ program.

Monitoring Intracellular Protein Dynamics with Dendra2

4

209

Notes 1. For Escherichia coli we use LB cultivation medium and for Agrobacterium tumefaciens YEB medium. 2. We use ½ MSMO medium (half concentration of MSMO salts, Sigma, #M6899) supplemented with sucrose to 1% (w/v) (Sigma-Aldrich) and solidified with agar (Difco) to 0.7% (w/v). The pH of the medium is adjusted with KOH to 5.7 before autoclaving. 3. For plant cultivation avoid using strong light sources with full spectrum emitting light at UV or violet spectra potentially inducing unwanted photoconversion. It was recently documented that even red light can convert Dendra2 under certain conditions [32]. Set the light intensities to 100 μmol. m2 s1. 4. The protocol is suggested for Olympus FV1000 equipped with a 100 W mercury burner, U-MNV2 mirror unit with 400–410 nm excitation filter; 405, 488, and 543 nm laser, 405/488/543 nm dichroic mirror, and emission filters in range of 505–525 nm, 560–620 nm. 5. We routinely amplify DNA by PCR from the genomic DNA of Arabidopsis (Col-0) including the regulatory regions, and we clone the fragment into the pAMPAT-MSC vector (GenBank: AY436765.1). The sequence encoding Dendra2 [4] is multiplied from Gateway®Dendra2-At-N entry clone (Evrogen, Moscow, Russia) and attached to the 30 or 50 end of genomic DNA of gene or inside the genomic sequence. In the current protocol, we used translational fusion of Dendra2 with Arabidopsis thaliana SYNAPTOTAGMIN 1 protein (SYT1) encoded by At2g20990 locus which we characterized previously [33]. Expression is driven by native SYT1 promoter and DNA encoding Dendra2 is attached to the 30 end of genomic SYT1 DNA. 6. We introduce plasmid into Agrobacterium tumefaciens GV3101 (pMP90RK, [34]) by electroporation and transform Arabidopsis (Col-0) by floral dipping [35]. 7. Selection of transformants in our experiment is carried out on medium supplemented with 7.5 mg/l phosphinothricin (PPT, Duchefa, Haarlem, The Netherlands) or under fluorescent stereo microscope based on fusion tag protein expression. Further selection of single copy transformants is performed by segregation analysis. 8. It may be necessary to keep petri dishes at 4  C for 2 days to synchronize seed germination. Dry seeds can be freeze-stored or they can be kept in the refrigerator.

210

Alexandra Lesˇkova´ et al.

9. You may need modified cultivation conditions if the steadystate protein level is regulated by light or temperature; for example, this would involve permanent light and constant temperature. 10. Protein steady state can be affected after moving seedlings to fresh medium because their synthesis may be accelerated. 11. We convert whole root tips with an Hg arc lamp for protein turnover experiments and in studies where we assess the time and/or dose-dependent effects of different compounds/factors on protein dynamics. 12. Imaging conditions are set for our configuration of Olympus FV1000 and selected SYT1-Dendra2 transgenic line. 13. Shortly before the medium is completely solidified, we place a coverslip on the top of the agar medium to achieve a flat surface and uniform thickness of the medium layer. 14. Excitation maximum for Green Dendra2 is 490 nm [36]. 15. The chart in Fig. 1b demonstrates the emission spectrum of nonconverted Dendra2 fused to SYT1 located to the plasma membrane and the endomembrane system (Fig. 1a). Maximum emission in live cells is around 505 nm, and this suggests that common filters used for green fluorophores are suitable for capturing the green variety of Dendra2. 16. We use in our experiments the UPLSAPO Super Apochromat objective (40
, 0.90 NA, F.N. 26.5) which allows for illumination of the whole root tip and with sufficient number of photons to photoconvert the fusion protein within a reasonable time. Objectives with lower numerical aperture and magnification may require longer illumination time. 17. This time is generally sufficient to efficiently convert samples without undesirable bleaching of Dendra2 fluorescent proteins; but it may not be optimal. 18. Although it was stated that Dendra2 should also be convertible with 488 nm laser light [4], we could not achieve conversion in any of our constructs when expressed in Arabidopsis tissues at this wavelength; we needed illumination with light around 400 nm. 19. The excitation maximum for red Dendra2 is 553 nm [36], but the 543 laser line excites the red Dendra2 form sufficiently. 20. The chart in Fig. 1b highlights the emission pattern of the red Dendra2 variety fused to SYT1. Maximum emission of photoconverted Dendra2 in live cells is around 570 nm. This verifies that common filters used for red fluorophores are suitable for capturing the red variety of Dendra2.

Monitoring Intracellular Protein Dynamics with Dendra2

211

21. The time required for conversion can be affected by different conditions; including the light intensity provided by Hg arc lamp, the type of fluorescence cube, amount of fluorescent protein, type of tissue expressing the fusion protein and also the objective parameters. An example of dependence of photoconversion on duration of illumination pulses is demonstrated in Fig. 1c. 22. The emissions of Dendra2 variants are pH sensitive [22, 37, 38]. Some compounds added to the media in higher doses may change the pH of the medium. Since the volume of the cultivation agar medium on the slide is only about 1.5 ml, we verify the pH using bromocresol purple indicator. We prepare a pair of 2 ml Eppendorf tubes with melted agar medium and the 0.006% (w/v) indicator. We add the studied compound to the cooled but unsolidified medium in one of the tubes and compare the colors. If the colour remains unchanged we prepare new medium with the compound and use it in the experiment. If the colour changes after adding a compound, we adjust it by addition of HCl/KOH until we reach the coloration of the reference medium without the compound. We prepare new medium with the compound and the required volume of HCl/KOH as previously determined. Do not use the medium with the indicator in the actual experiment because the dye may interfere with the protein fluorescence. 23. We use the following imaging parameters in Olympus FV1000 confocal microscope: 3% power for 488 nm argon laser and 25% for 543 nm He-Ne laser, frame size of 512
800 pixels, 12-bit depth, 150 μm pinhole size, dwell time of 4 μs/pixel, and line averaging 4 for the selected SYT1-Dendra2 line. 24. Prolonged cultivation of seedlings covered by a coverslip can cause decreased protein abundance in the cells because of the anaerobic conditions [20]. 25. We use square-shaped petri dishes, filled with 50 ml ½ MS agar medium. The surface of the medium is covered with a Parafilm square to avoid cross-contamination by liquid leakage from the medium. 26. Protein diffusion is a relatively fast process, therefore the speed of the imaging and photoconversion is critical for local and short-term studies of protein movement. Both the image acquisition and photoconversion should be as fast as possible. We should not be too concerned with the quality of the images, as long as the signal-to-noise ratio is high enough and the pixels are not saturated. For the Olympus FV1000 confocal microscope, we use the following settings: 1% power for 488 nm argon laser and 25% for 543 nm He-Ne laser, frame

212

Alexandra Lesˇkova´ et al.

size of 512
512 pixels, 12-bit depth, 150 μm pinhole size, dwell time of 2 μs/pixel, and line averaging 4. 27. The efficiency of the photoconversion and the size of the photoconverted region depend on the zoom factor, so we use the high zoom factor to photoconvert the smallest possible region. We set our 488 nm argon laser to low, 0.2% laser power and zoom in 50-times. For the local point conversion, we choose the point-scan tool to select a 4
4 pixel-sized (48
48 nm) region of the original 512
512 pixel-sized frame. The converted area will be larger than expected due to light dispersion in the sample. 28. Optimize the laser power and the time required for the photoconversion of the selected ROI to your own system. Using high laser power and prolonged photoconversion can lead to bleaching of the ROI center, while photoconversion will occur at the ROI periphery. We have found that optimization of the exposure time is easier than optimization of the laser power, because in our hands the exposure time correlated better with the efficiency of the photoconversion than the laser power. For Olympus FV1000 we use 0.8% power of 405 nm LD laser, exposure time of several seconds depending on the ROI size and maximum possible dwell time of the system (200 μs/ pixel). 29. If the photoconverted area is wider than expected, repeat the conversion with smaller ROI and/or lower laser power. 30. The frequency and overall time you need to observe the protein movement depends on the mobility of your protein, and this may vary considerably.

Acknowledgments This work has been supported by the Slovak Research and Development Agency (grant no. APVV-16-0398). J.J. thanks also the Alexander von Humboldt Foundation for supporting his renewed research stay at MPIPZ, Cologne. References 1. Kremers GJ, Gilbert SG, Cranfill PJ, Davidson MW, Piston DW (2011) Fluorescent proteins at a glance. J Cell Sci 124:157–160 2. Rodriguez EA, Campbell RE, Lin JY, Lin MZ, Miyawaki A et al (2017) The growing and glowing toolbox of fluorescent and photoactive proteins. Trends Biochem Sci 42:111–129 3. Wiedenmann J, Ivanchenko S, Oswald F, Schmitt F, Ro¨cker C et al (2004) EosFP, a

fluorescent marker protein with UV-inducible green-to-red fluorescence conversion. Proc Natl Acad Sci U S A 101:15905–15910 4. Gurskaya NG, Verkhusha VV, Shcheglov AS, Staroverov DB, Chepurnykh TV et al (2006) Engineering of a monomeric green-to-red photoactivatable fluorescent protein induced by blue light. Nat Biotechnol 24:461–465

Monitoring Intracellular Protein Dynamics with Dendra2 5. Lummer M, Humpert F, Wiedenlu¨bbert M, Sauer M, Schu¨ttpelz M et al (2013) A new set of reversibly photoswitchable fluorescent proteins for use in transgenic plants. Mol Plant 6:1518–1530 6. Griffiths N, Jaipargas EA, Wozny MR, Barton KA, Mathur N et al (2016) Photo-convertible fluorescent proteins as tools for fresh insights on subcellular interactions in plants. J Microsc 263:148–157 7. Arimura SI, Yamamoto J, Aida GP, Nakazono M, Tsutsumi N (2004) Frequent fusion and fission of plant mitochondria with unequal nucleoid distribution. Proc Natl Acad Sci U S A 101:7805–7808 8. Watanabe W, Shimada T, Matsunaga S, Kurihara D, Fukui K et al (2007) Singleorganelle tracking by two-photon conversion. Opt Express 15:2490–2498 9. Brown SC, Bolte S, Gaudin M, Pereira C, Marion J et al (2010) Exploring plant endomembrane dynamics using the photoconvertible protein Kaede. Plant J 63:696–711 10. Wolf H, Barisas BG, Dietz KJ, Seidel T (2013) Kaede for detection of protein oligomerization. Mol Plant 6:1453–1462 11. Cho M, Lee ZW, Cho HT (2012) ATP-binding cassette B4, an auxin-efflux transporter, stably associates with the plasma membrane and shows distinctive intracellular trafficking from that of PIN-FORMEDs. Plant Physiol 159:642–654 12. Dhonukshe P, Aniento F, Hwang I, Robinson DG, Mravec J et al (2007) Clathrin-mediated constitutive endocytosis of PIN auxin efflux carriers in Arabidopsis. Curr Biol 17:520–527 13. Mathur J, Radhamony R, Sinclair AM, Donoso A, Dunn N et al (2010) mEosFP based green to red photoconvertible subcellular probes for plants. Plant Physiol 154:1573–1587 14. Schattat MH, Griffiths S, Mathur N, Barton K, Wozny MR et al (2012) Differential coloring reveals that plastids do not form networks for exchanging macromolecules. Plant Cell 24:1465–1477 15. Hanson MR, Sattarzadeh A (2013) Trafficking of proteins through plastid stromules. Plant Cell 25:2774–2782 16. Scott I, Sparkes IA, Logan DC (2007) The missing link: inter-organellar connections in mitochondria and peroxisomes? Trends Plant Sci 12:380–381 17. Sinclair AM, Trobacher CP, Mathur N, Greenwood JS, Mathur J (2009) Peroxule extension over ER-defined paths constitutes a rapid

213

subcellular response to hydroxyl stress. Plant J 59:231–242 18. Jaipargas EA, Barton KA, Mathur N, Mathur J (2015) Mitochondrial pleomorphy in plant cells is driven by contiguous ER dynamics. Front Plant Sci 6:783 19. Schenkel M, Sinclair AM, Johnstone D, Bewley JD, Mathur J (2008) Visualizing the actin cytoskeleton in living plant cells using a photoconvertible mEos:: FABD-mTn fluorescent fusion protein. Plant Methods 4:21 20. Wozny M, Schattat MH, Mathur N, Barton K, Mathur J (2011) Colour recovery after photoconversion of H2B: mEosFP allows detection of increased nuclear DNA content in developing plant cells. Plant Physiol 158:95–106 21. Nagawa S, Xu T, Lin D, Dhonukshe P, Zhang X et al (2012) ROP GTPase-dependent actin microfilaments promote PIN1 polarization by localized inhibition of clathrin-dependent endocytosis. PLoS Biol 10:e1001299 22. Ja´sik J, Boggetti B, Balusˇka F, Volkmann D, Gensch T et al (2013) PIN2 turnover in Arabidopsis root epidermal cells explored by the photoconvertible protein Dendra2. PLoS One 8:e61403 23. Ja´sik J, Schmelzer E (2014) Internalized and newly synthesized Arabidopsis PIN-FORMED2 pass through brefeldin A compartments: a new insight into intracellular dynamics of the protein by using the photoconvertible fluorescence protein Dendra2 as a tag. Mol Plant 7:1578–1581 24. Ja´sik J, Bokor B, Stuchlı´k S, Micˇieta K, Turnˇa J et al (2016) Effects of auxins on PIN-FORMED2 (PIN2) dynamics are not mediated by inhibiting PIN2 endocytosis. Plant Physiol 172:1019–1031 25. Kitagawa M, Fujita T (2013) Quantitative imaging of directional transport through plasmodesmata in moss protonemata via single-cell photoconversion of Dendra2. J Plant Res 126:577–585 26. Wu S, Koizumi K, MacRae-Crerar A, Gallagher KL (2011) Assessing the utility of photoswitchable fluorescent proteins for tracking intercellular protein movement in the Arabidopsis root. PLoS One 6:e27536 27. Rausin G, Tillemans V, Stankovic N, Hanikenne M, Motte P (2010) Dynamic nucleocytoplasmic shuttling of an Arabidopsis SR splicing factor: role of the RNA-binding domains. Plant Physiol 153:273–284 28. Hamamura Y, Saito C, Awai C, Kurihara D, Miyawaki A et al (2011) Live-cell imaging reveals the dynamics of two sperm cells during

214

Alexandra Lesˇkova´ et al.

double fertilization in Arabidopsis thaliana. Curr Biol 21:497–502 29. Lummer M, Humpert F, Steuwe C, Caesar K, Schu¨ttpelz M et al (2011) Reversible photoswitchable DRONPA-s monitors nucleocytoplasmic transport of an RNA-binding protein in transgenic plants. Traffic 12:693–702 30. Gerlitz N, Gerum R, Sauer N, Stadler R (2018) Photoinducible DRONPA-s: a new tool for investigating cell–cell connectivity. Plant J 94:751–766 31. Wang H, Han S, Siao W, Song C, Xiang Y et al (2015) Arabidopsis synaptotagmin 2 participates in pollen germination and tube growth and is delivered to plasma membrane via conventional secretion. Mol Plant 8:1737–1750 32. Klementieva NV, Lukyanov KA, Markina NM, Lukyanov SA, Zagaynova EV et al (2016) Green-to red primed conversion of Dendra2 using blue and red lasers. Chem Commun 52:13144–13146 33. Schapire AL, Voigt B, Jasik J, Rosado A, Lopez-Cobollo R et al (2008) Arabidopsis synaptotagmin 1 is required for the maintenance of plasma membrane integrity and cell viability. Plant Cell 20:3374–3388

34. Koncz C, Schell J (1986) The promoter of T L-DNA gene 5 controls the tissue-specific expression of chimaeric genes carried by a novel type of Agrobacterium binary vector. Mol Gen Genet 204:383–396 35. Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16:735–743 36. Chudakov DM, Lukyanov S, Lukyanov KA (2007) Tracking intracellular protein movements using photoswitchable fluorescent proteins PS-CFP2 and Dendra2. Nat Protoc 2:2024–2032 37. Adam V, Nienhaus K, Bourgeois D, Nienhaus GU (2009) Structural basis of enhanced photoconversion yield in green fluorescent protein-like protein Dendra2. Biochemistry 48:4905–4915 38. Pakhomov AA, Martynov VI, Orsa AN, Bondarenko AA, Chertkova RV et al (2017) Fluorescent protein Dendra2 as a ratiometric genetically encoded pH-sensor. Biochem Biophys Res Commun 493:1518–1521

Chapter 14 Cellular Force Microscopy to Measure Mechanical Forces in Plant Cells Mateusz Majda, Aleksandra Sapala, Anne-Lise Routier-Kierzkowska, and Richard S. Smith Abstract Cellular force microscopy (CFM) is a noninvasive microindentation method used to measure plant cell stiffness in vivo. CFM is a scanning probe microscopy technique similar in operation to atomic force microscopy (AFM); however, the scale of movement and range of forces are much larger, making it suitable for stiffness measurements on turgid plant cells in whole organs. CFM experiments can be performed on living samples over extended time periods, facilitating the exploration of the dynamics of processes involving mechanics. Different sensor technologies can be used, along with a variety of probe shapes and sizes that can be tailored to specific applications. Measurements can be made for specific indentation depths, forces and timing, allowing for very precise mechanical stimulation of cells with known forces. High forces with sharp tips can also be used for mechanical ablation of cells with force feedback. Key words Plant mechanics, Microindentation, Cell walls, Turgor pressure, Elastic modulus, Stiffness

1

Introduction Plant cells are surrounded by a rigid cell wall that must be able to withstand the substantial turgor pressure within. Plant cell walls must also be dynamically extensible, as cells can quickly grow several orders of magnitude in size. Growth in plants is symplastic; cells are attached to their neighbors and cannot move with respect to each other. This implies that plant shape and form emerges from regulated growth of the cell wall. Understanding how genes perform this regulation, and its interaction with cell wall mechanics, requires methods to quantify mechanical properties at the cellular level. Growth is driven by cellular turgor pressure, which induces tensile stress on the surrounding cell walls. The wall is a complex polysaccharide structure made of strong cellulose microfibrils embedded in a highly hydrated matrix made of polysaccharides such as hemicelluloses (mainly xyloglucans), pectins, and structural

Fatima Cvrcˇkova´ and Viktor Zˇa´rsky´ (eds.), Plant Cell Morphogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1992, https://doi.org/10.1007/978-1-4939-9469-4_14, © Springer Science+Business Media, LLC, part of Springer Nature 2019

215

216

Mateusz Majda et al.

proteins. During plant cell growth, elastic strain in the cell wall caused by turgor pressure becomes plastic (irreversible), as interactions between different polysaccharides within the wall are modified in a process called remodeling. Combined with the biosynthesis of new wall components, these processes determine the cell wall stiffness and its ability to grow [1]. Different methods have been used to measure stiffness in plant tissue, the simplest of which is direct stretching of isolated tissues with an extensometer. Another approach for single cells is to manipulate turgor pressure with a pressure probe and measure the resulting deformation [2]. However, these techniques are destructive for the sample and cannot be used to measure the mechanical properties in live tissue (reviewed in ref. 3). Other methods such as noncontact Brillouin microscopy [4, 5], microextensometer [6, 7] or indentation techniques are less invasive and bring a potential to study the plant material in vivo [8–12]. Indentation techniques allow for the quantification of the stiffness of a material based on the force required to deform the material with a probe. The character of the sample, such as its dimensions and stiffness, determines the type of indenter used, and the size and shape of the tip. In addition to measuring stiffness, many microindentation systems are also able to read the topography of the cells by scanning the surface with the probe. Perhaps the most common indentation technique used on plants is Atomic Force Microscopy (AFM, refs. 8, 13–15). Originally developed for imaging and force measurement at the nanoscale, AFM is particularly well suited for the study of molecules and subcellular structures requiring the use of very small tips (tens of nanometers radius) and low forces (nano-Newton), such as those seen in animal cells. Attempts have been made to increase the range of the AFM by using very stiff cantilevers and larger tips, for example by gluing 1 or 5 μm beads onto the cantilever [16], or by using specially made large and stiff cantilevers [17]. Here we will present a protocol for Cellular Force Microscopy (CFM, ref. 10), which is a noninvasive microindentation technique designed to measure forces in plants at the cellular level. Like AFM, it is a scanning probe-based method, but it is much larger, and designed to operate with larger displacements and higher forces typically found in turgid plant cells. The CFM is comprised of a force sensor and indenter that is attached to a microrobotic actuator that can move the sensor along 3 axes (Fig. 1a). The force sensors currently supported in the software are the interferometer-based sensors from Optics11 (http://www.optics11.com; Fig. 1b), and the microelectromechanical system (MEMS) sensors from FemtoTools (http://www. femtotools.com; Fig. 1c). Both convert the measured force to a voltage that is sent to the computer for processing. Software on the PC controls the movement of the actuator, and by recording the

Cellular Force Microscopy in Plant Cells

217

Fig. 1 The CFM hardware. (A) A 3-axis SmarAct positioner is attached to an inverted light microscope. The sensor is attached to the positioner with a custom-made arm. A digital camera from the side allows for the monitoring of the distance between the sensor and a sample. (B) An interferometer-based Optics11 sensor. It is comprised of a cantilever attached to the bottom of a plastic block that contains an optical fiber. Cantilever tips can be ordered in different shapes, sizes and stiffness values (here: a spherical bead on a 200 μm fiber). (C) A FemtoTools sensor with a long, thin tungsten needle attached on the bottom. The blue lines in B and C indicate a suitable liquid level for the experiment. Scale bars, 1 mm in B, 0.6 mm in C

change in force during displacement, a force-indentation curve can be produced. This is then used to extract the stiffness of the sample. The CFM actuator can be used in combination with an inverted light microscope (Fig. 1A) allowing for the simultaneous acquisition of forces and optical images. For upright microscopes, near simultaneous acquisition can be accomplished by switching between the CFM and a microscope objective. This gives the possibility to link mechanical forces with specific cellular events from traditional light microscopy or laser confocal imaging. The 3-axis positioner is made by SmarAct (http://www. smaract.com) and uses 3 linear actuators combined on orthogonal axes (x–y–z). These piezo actuators can perform very smooth movements of up to several microns, and have internal sensors that monitor their location with nanometer precision. The fine movements involving only a voltage change on the piezo are called scanning mode. Since the range of piezo motion is only several microns, it is combined with a stick-slip mode that allows the actuator to move several centimeters. The stick-slip motion introduces vibrations and is used for coarse positioning, while the scanning mode is very smooth and used for precise stiffness or force measurement. CFM measurements can be performed in different environments such as air or liquid medium (e.g., water, buffer). The Optics11 sensors can be fully submerged, but must be calibrated while immersed in the medium (Fig. 1B). The FemtoTools sensors cannot be submerged but use a longer tip (Fig. 1C); however, the surface tension of the medium can create forces on the sensor probe of up to several micro-Newtons. The long tip of the FemtoTools sensors makes them advantageous for samples where

218

Mateusz Majda et al.

accessibility is an issue. The Optics11 sensors on the other hand are less fragile and do not suffer from surface tension issues. Although there are many variations possible, a typical CFM experiment begins with a quick surface scan, after which points are selected to make stiffness measurements. Since the surface curvature can affect the sample measurement, the selected points are normally on the top in the center of the cells. The CFM is mostly sensitive to pressure (and geometry) in turgid plant cells, so usually it is sufficient to select only a few points per cell. The stiffness is measured by performing several indentation cycles in scanning mode and then fitting a section of the resulting force indentation curves to a line. CFM gives new possibilities to measure mechanical properties of biological systems in vivo. Surface scans are used to reconstruct sample shape in high resolution in position (typically micrometer to nanometer scale) but also records the force-indentation curves, which are further used to quantify the stiffness and turgor pressure. Moreover, fully adjustable probe size, controlled force, and timing allow for a very precise mechanical stimulation of cells with known forces. In case of high forces applied, programmed mechanical ablation of specific single cells is also possible. Among many applications, the main advantages of CFM are its ability to measure mechanical properties in the large areas at cellular and multicellular levels (in centimeter scale), and its ability to detect and apply a large range of forces (from several to hundreds of micronewtons). Moreover, customized probes (e.g., long and thin tungsten tips) allow measuring the properties of very narrow regions of the samples. All the measurements can be carried in living plant tissues in vivo over extended time periods, without fixation, in single cells and tissues [10, 18–22]. In the following paragraphs, we will focus on stiffness measurements in live plant material submerged in water.

2

Materials

2.1 CFM Components

1. The sensors provided by Optics11 (optical fiber sensor, https://www.optics11.com) consist of a tip (made of a borosilicate glass or silica), which is glued on a fiber or directly on the cantilever. The cantilever is mounted on an optical glass probe (made of borosilicate base and acrylic probe holder). Tips can be ordered for a range of sizes (3–500 μm radius) and shapes (spherical or cylindrical) as well as different stiffness (0.01 N/ m–10 kN/m). The sensors are relatively durable and can be reused in multiple experiments. Sensors can operate in any environment, from liquids to vacuum as well as in different ranges of temperature. FemtoTools (http://www.femtotools. com) sensors can be ordered with different ranges, 0–100,

Cellular Force Microscopy in Plant Cells

219

1000, 10,000, and 100,000 μN. There are several options for tips ranging from 1 to 250 μm radius. The can be reused in many experiments, but cannot be completely submersed in liquid. 2. The actuator robot is a modular system composed of three orthogonal linear micropositioners moving the sensor (SmarAct GmbH, SLC-1780) and a manually operated control unit (SmarAct GmbH, MCS-3D, http://www.smaract.com). The hand controller allows for the coarse positioning of the sensor prior to experiments, and displays the current position of the sensor. The controller is connected to the computer via USB. The sensor is mounted on the robot with a custom plastic arm (CAD plans for 3D printing available upon request). The actuator and sensor setup is mounted on an inverted Zeiss Axiovert H DIC microscope. 3. When using an Optics11 sensor, the sensor is connected via an optical fiber to the interferometer (OP1550, Optics11, https://www.optics11.com) which measures the displacements of cantilever and maximizes the sensitivity and dynamic range of the wavelength, gain and offset. The output of the interferometer is represented as analog signal, typically 10 to +10 V, that can also be seen in interferometer display. FemtoTools sensors directly output a voltage from 5 to +5 V. 4. A data acquisition card in the computer collects the signal (National Instruments NI PCIe-6321, http://www.ni.com). 5. The computer requires an nVIDIA graphics card that supports CUDA. A card with 2 GB or more dedicated video memory is recommended. The computer itself should have at least 8 GB of main memory and a 4 core CPU. 6. Install the MorphoRobotX control and visualization software for the CFM, which can be downloaded at http://www. MorphoRobotX.org. MorphoRobotX has been tested on Linux Mint 18, but should also work on other Debian-based Linux systems. You will need to install the libraries for the SmarAct controller (available from SmarAct) as well as the Comedi library package that interfaces with the National Instruments data acquisition card. MorphoRobotX is an extension of a 3D image processing software MorphoGraphX. For basic information on how to operate this software see [23] or visit www.MorphoGraphX.org. 7. The light source for the microscope is replaced with LED rings (AmScope, http://www.amscope.com) placed around the sensor. 8. A digital microscope camera (DigiMicro 2.0, dnt GmbH, http://www.dnt.de) connected to Cheese software (GNOME

220

Mateusz Majda et al.

webcam application, https://wiki.gnome.org/Apps/Cheese) is used to provide a side view of the sensor tip. 9. An objective camera (Point Grey Research FL2G-50S5C-C, https://www.ptgrey.com) is connected to the inverted microscope to visualize the sample from below. The camera is connected to the software (Coriander 2.0.2, http://damien. douxchamps.net/ieee1394/coriander/), which can be used to take images and record movies during the experiment. 10. The whole setup sits on an antivibration table (TMC 63-530 series, http://www.techmfg.com) and is covered with a custom plexiglass box and grid cage, to isolate CFM from acoustic vibrations and air currents. The table and setup are also electrically grounded. 2.2 Biological Samples

The CFM is suitable to make measurements on outer plant tissues, such as epidermal cells in leaves, cotyledons, hypocotyls, and roots in vivo. Samples must be immobilized in medium-sized plastic Petri dishes and submersed in liquid. 1. The viability of the cells can be monitored by the observation of cytoplasmatic streaming (in transparent tissues such as isolated epidermal onion tissues) in the light microscope or using propidium iodide (0.1%) dye (Sigma-Aldrich, MilliporeSigma) in confocal microscope. 2. Flat samples such as leaves or isolated epidermal tissues can be stabilized using permanent double-sided tape (Scotch) or laboratory tags (Tough-Tags, Diversified Biotech). 3. Nonflat or very small (1) are displayed in white, while red is used to draw the direction of shrinkage (stretch ratio