No other microtool allows this, and therefore most experiments with optical tweezers in plant cell b...

No other microtool allows this, and therefore most experiments with optical tweezers in plant cell biology are on objects inside living cells,. Protocol 2[r]

The Practical Approach Series

Related Practical Approach Series Titles

light Microscopy in Biology 2/e

Protein Localization by Fluorescence Microscopy Flow Cytometry 3/e

In Situ Hybridization 2/e Cell Separation

Arabidopsis Plant Cell Culture Plant Molecular Biology*

* indicates a forthcoming title

Please see the Practical Approach series website at http://www.oup.co.uk/pas

Plant Cell Biology Second Edition A Practical Approach

Edited by

Chris Hawes

Research School of Biological and Molecular Sciences, Oxford Brookes University, Oxford 0X3 OBP, UK

and

Beatrice Satiat-Jeunemaitre

Institut des Sciences Vegetales, CNRS UPR 40, 91198 Gif-sur-Yvette, France

OXFORD

UNIVERSITY PRESS

Great Clarendon Street, Oxford OX2 6DP

Oxford University Press is a department of the University of Oxford It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide in

Oxford New York

Athens Auckland Bangkok Bogota Buenos Aires Calcutta Cape Town Chennai Dares Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paulo Singapore Taipei Tokyo Toronto Warsaw

with associated companies in Berlin Ibadan

Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries

Published in the United States by Oxford University Press Inc., New York © Oxford University Press, 2001

The moral rights of the author have been asserted Database right Oxford University Press (maker) First edition published 1994

Second edition published 2001

All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above

You must not circulate this book in any other binding or cover and you must impose this same condition on any acquirer

A catalogue record for this title is available from the British Library

Library of Congress Cataloguing-in-Publication Data Plant cell biology / edited by Chris Hawes and Beatrice Satiat-Jeunemaitre.-2nd ed (Practical approach series; 250) Includes bibliographical references (p.)

1 Botanical microscopy-Technique Plant cytochemistry-Technique Plant cells and tissues I Hawes, C R II Satiat-Jeunemaitre, Beatrice III Series

1

QK673 P58 2001 571.692dc21 00-054847 ISBN 19 963866 7(Hbk)

ISBN 19 963865 9(Pbk)

Typeset in Swift by Footnote Graphics, Warminster, Wilts Printed in Great Britain on acid-free paper

Preface

It is just seven short years since the first edition of Plant Cell Biology was published, yet in that time there have been a number of significant advances in the technologies used by cell biologists The start of the new millennium coincides with the dawn of the so-called 'post-genomics' era where biologists will struggle to cope with the almost overwhelming flow of information that will emerge from the various sequencing projects We predict that one consequence of this will be a surge in demand for the techniques of cell biology to aid in the interpretation of the function and location of the myriad of proteins and macromolecules that make up the cell Ironically this demand will necessitate the application of technologies that have somewhat dropped out of favour in the past decade or two such as classical histochemistry and electron microscopy These will, however, have to be combined with new techniques such as the in vivo expression of fluorescent markers and the subcellular manipulation of organelles and physiological measurements from the cytoplasm

When taking over as editors of the new volume we were posed with several serious problems As stated by Harris and Oparka in 1993, in the production of such a volume one has to be extremely selective and miss out many important areas of the topic This has been even more difficult considering the excellent coverage of the original volume, so this edition contains a mixture of up-dated chapters from original authors, new authors covering topics that were in the original volume, and a raft of totally new chapters

now famous green fluorescent protein and its derivatives for in vivo imaging of cell dynamics We now have the ability to directly manipulate cytoplasm and organelles and Chapters and give an introduction to the latest technologies in microinjection and micromanipulation followed by the use of electrophysio-logical techniques to complement cell bioelectrophysio-logical investigations

Moving away from the study of living cells, in Chapter we revisit some of the more useful histological techniques that are now much in demand by molecular and developmental biologists The following three chapters cover the various methods for localizing macromolecules and nucleic acids by light and electron microscopy, whilst also including the basic techniques of specimen preparation for electron microscopy A technology that has suffered with the inexorable rise of molecular biology and that we lose at our peril Finally, we finish on a biochemical note covering cell fractionation and organelle isolation, useful techniques that have become ever more powerful with the development of a wide range of marker antibodies

Obviously in many instances authors of the various chapters have only been able to dip into the vast array of techniques at hand and we apologize in advance if we have missed your favourite technique or propose a different protocol to the one you may routinely use And finally, we would like to thank the various con-tributors for giving their valuable time in preparing their chapters and divulging their laboratory hints and tips and for succumbing to the onerous demands of the editors

Oxford C H

Gif-sur-Yvette B S-J.

Contents

List of protocols page xv Abbreviations xix

1 Introduction to optical microscopy for plant cell biology 1

P J Shaw

1 Introduction 2 Explanation of terms 4 Recording images

Image resolution Film recording Electronic cameras Microscope imaging modes 10

Bright field imaging 10 Phase contrast 12

Differential interference contrast (Nomarski) 24 Dark field 16

Epifluorescence and reflected light microscopy 5 Confocal and 3D microscopy 19

The problem of out-of-focus light 19

The confocal principle: explanation by ray optics 20 Practical confocal microscopes 22

Imaging and the point spread function 23 Deconvolution 24

Two photon imaging 26

6 Comparison of conventional, wide-field fluorescence imaging with confocal fluorescence imaging 27

Noise and resolution 27

When should confocal microscopy be used? 29 Objective lenses for confocal imaging 30 Specimen preparation for confocal imaging 30

References 33

2 Fluorescent probes for living plant cells 35

Mark Fricker, Andrew Parsons, Monika Tlalka, Elison Blancaflor, Simon Gilroy, Andreas Meyer, and Christoph Plieth

2 Selecting probes with high brightness 35 Spectral considerations 37

3 Fluorescence lifetime imaging microscopy (FLIM) 38 Fluorescence polarization anisotropy 38

5 Fluorescence resonance energy transfer (FRET) 38

6 Photobleaching and fluorescence redistribution after photobleaching (FRAP) 39

7 Optimization of fluorescent systems for live cell imaging 40 Selection of the excitation wavelength 41

The dichroic mirror 41

Selection of the emission wavelength 41 Choice of measurement system 42 8 Securing the specimen for microscopy 43

9 Perfusion systems 45

10 Loading strategies for plant cells 46

Extracellular and permeant intracellular dyes 46 Ester loading 47

Low pH loading 48

Cutinase pre-treatment and low pH loading 49 Electroporation 50

Loading via detergent permeabilization 50 Loading tissues with phloem-mobile probes 50

11 Intracellular dye concentrations, viability, and toxicity 53 12 Selection and use of fluorescent probes 54

Vital stains 55 Mortal stains 56

Cell permeant nuclear stains 57 Chloroplasts 57

Mitochondria 58 Vacuoles 60

Endoplasmic reticulum 62 Golgi 62

Cytoskeleton 63

The plasma membrane and endocytosis 64 The cell wall 65

13 Physiological probes 66 Calcium 66

Measurement of apoplastic, cytoplasmic, and vacuolar pH 72 Potassium 76

Aluminium 76

Measurement of cytoplasmic glutathione levels 77 Reactive oxygen species 77

14 Data analysis 78

Attenuation correction for optical sections deep into tissues 80 Acknowledgements 83

References 81

3 Flow cytometry 85

2 Cytometry demonstrated through cell cycle analyses 86 How to understand monoparametric DNA histograms 87

Developing multiparametric DNA histograms and immunofluorescence 90 Extracting intact plant nuclei 91

Which DNA fluorochrome is appropriate? 94 Running and reading the cytometer 94

BrdU incorporation to identify DNA synthesis by fluorescence quenching 95 The particular application of genome size calculation and 'DNA ploidy' 99

Terminology 99

Internal or external standards 99 Calculating base composition 100

4 Sorting of protoplasts and cellular organelles 101 5 Tests for cell viability during functional assays 104

6 Conclusion 104 Acknowledgements 104 References 105

4 Transient expression, a tool to address questions in plant cell biology 107

Jane L Hadlingtan and jurgen Denecke Introduction 107

2 Current methods of transient expression 108 Naked DNA transfer 108

Biological vectors 108

3 Application of transient expression 109 Promoter analysis 109

Cell biology and biochemistry 110

4 Practical considerations for cell biologists 111 Naked DNA transfer 112

Measurement of protein secretion and cell retention 116 Large scale transient expression for cell fractionation 118 Specialized applications 122

5 Conclusions 124 References 124

5 The green fluorescent protein (GFP) as reporter in plant cells 127 Jean-Marc Neuhaus and Petra Boevink

1 Introduction 127

2 The green fluorescent protein 127 Structure 127

GFP variants 128

3 GFP as a reporter for gene expression 128 GFP as a reporter for protein location 129

Cytoplasm and nucleus 129 Chloroplasts and mitochondria 130 Secretory pathway 130

5 Transformation methods 131

PEG-mediated transient expression in protoplasts 131 Agroborterium-mediated transient expression in planta 134 Virus-mediated transient expression 136

6 Visualization and microscopy of GFP 140 Future perspectives 141

References 141

6 Microinjection 143 Michael Knoblauch

1 Introduction 143 Equipment 143

Environment and injection-table 143 Microscope 144

Objectives 144 Glass capillaries 145 Tip puller 145 Micromanipulator 146 Injection techniques 146

Iontophoresis 146 Pressure injection 146

The galinstan expansion femtosyringe (GEF) 147 Cell types 149

Epidermal cells 149

Guard cells and trichomes 150 Mesophyll cells 150

Ground parenchyma cells 151

Sieve elements and companion cells 151 Algae 152

Bacteria and organelles 153 Plant tissue cultures 153 Material suitable for injection 154

Fluorochromes 154 Dextran conjugates 154 Proteins and antibodies 155 Nucleic acids 155

6 Tips to make life easier 155 The chuck 155

Syringe to delete air bubbles 156 Flexible fused silica capillaries 156 Petri dishes 157

References 157

7 Micromanipulation by laser microbeam and optical tweezers 159

Karl Otto Greulich 1 Introduction 159

3 Physical background 160 Generating extreme heat 160

Why can light be used to move microscopic objects? 160

4 How to build laser microtools 161

The choice of lasers 161

Building a laser microbeam or optical tweezers 162 Applications of laser microbeams in plant biology 163

Laser-induced microinjection 163

Ablation to study cell fate during plant development 164 Protoplast fusion 165

Preparation of cell membranes from root hairs 165 Applications of optical tweezers to plant biology 166

Capturing subcellular organelles for inspection 166 Simulating microgravity 167

7 Conclusion 168 References 168

8 Electrophysiological methods: monitoring exo- and endocytosis in real time 171

Gerhard Thiel, Jens-Uwe Sutter, and Ulrike Homann

1 Introduction 171

2 Theoretical background 171

The membrane is equivalent to a capacitor 171 A cell as an equivalent circuit 172

3 Techniques for the measurement of membrane capacitance 172 Square-wave stimulation: time-domain technique 173

Saw-tooth stimulation 174 Capacitance cancellation 174 Sinusoidal excitation 174

4 Capacitance measurements as an assay for exo- and endocytosis: practical considerations 175

What kind of cells can be examined? 176 Estimation of the specific capacitance 176 Recording of single fusion and fission events 176

What kind of information can be extracted from the measurements? 179 Macroscopic measurement of membrane capacitance 181

Acknowledgements 186 References 187

9 Plant histology 189

Jackie Spence

1 Introduction 189

2 Conventional chemical fixation methods 189

3 Conventional embedding methods and sectioning 191 Embedding in a matrix 192

4 Conventional staining methods 195 General tissue stains 195

Cell wall stains 397

Carbohydrate and starch stains 200 Lipid stains 201

Nucleic acid stains 202

Miscellaneous staining methods 203 References 206

10 Immunocytochemistry for light microscopy 207 Beatrice Satiat-Jeunemaitre and Chris Howes

1 Introduction 207

2 Principles and use of immunocytochemistry 208 Direct and indirect immunostaining 208 The antibody-antigen complex 208 Whole molecules or fragments 209 Polyclonal and monoclonal antibodies 211 When to perform in situ immunoreaction 214 Antibodies to epitope tags 215

3 Basic methods for immunostaining 215 Preparing plant material 216

Attaching material to slides and coverslips 220 Accessing epitopes in cells 220

Counterstaining and mounting 230

Interpreting the immunostaining pattern 230 Multiple staining 231

Multiple immunostaining 231

Combining immunostaining with other affinity techniques 231 Conclusion 232

References 232

11 Electron microscopy 235

Chris Hawes and Beatrice Satiat-Jeunemaitre Introduction 235

2 Transmission electron microscopy 235 Conventional methods 235

Low temperature methods 250 Rotary shadowing of proteins 257 Scanning electron microscopy 258

Ambient temperature SEM 258 Low temperature SEM 262 Immuno-SEM 263 Acknowledgements 263 References 264

12 In situ hybridization 267

2 Applications of in situ hybridization 269

DNA:DNA in situ hybridization 269 RNA:RNA in situ hybridization 269

3 Background to the methods 269

4 Chromosome preparation for DNA:DNA in situ hybridization 270

Animal material 270 Plant material 271

5 Material preparation for RNA:RNA in situ hybridization 273

Specimen preparation 273 Controls 275

6 Labelling the nucleic acids 276

DNA labelling 276

Checking probe incorporation 282

7 Material pre-treatment 283

Pre-treatment for DNA:DNA in situ hybridization 283 Pre-treatment for RNA:RNA in situ hybridization 284

8 In situ hybridization reaction 285

DNA:DNA in situ hybridization 285 RNA:RNA in situ hybridization 286

9 Post-hybridization washes 287

Washes for DNA:DNA in situ hybridization 288 Washes for RNA:RNA in situ hybridization 288

10 Probe detection and visualization 289 11 Visualization 290

Epifluorescence microscopy 290 Image capture 291

Image manipulation 292

References 292

13 Organelle Isolation 295 David G Robinson and Giselbert Hinz

1 Introduction 295 General methodology 295

Homogenizing media 295 Methods of homogenization 298 Methods of organelle separation 299

3 Isolation of chloroplasts 300 Isolation of mitochondria 300 Isolation of nuclei 301 Isolation of microbodies 303 Isolation of plasma membrane 304 Isolation of tonoplast 306

12 Assays for marker enzymes 325

13 Antibodies for organelle recognition 319 References 320

A1 List of suppliers 325

Protocol list

Microscope imaging modes

Adjustment for bright field imaging with Kohler illumination 11 Adjustment for phase contrast 14

Adjustment for DIC 15 Adjustment for dark field 16

Specimen preparation for confocal imaging 30 Collection of confocal images 32

Fluorescent probes for living plant cells

Growth of Arabidopsis thaliana seedlings in Phytagel for in situ observation of roots 44

Loading strategies for plant cells 46

Loading dyes by vacuum infiltration of leaf pieces 47

Loading dyes as acetoxymethyl ester or acetate ester derivatives 47 Low pH loading of root tissues 49

Increasing the permeability of the cuticle by cutinase digestion 50 Loading root tissues with phloem-mobile probes 51

Direct observation of phloem transport in bean leaves 52 Selection and use of fluorescent probes 54

Labelling of actin filaments in living cells using fluorescent phallotoxins 63 Labelling of microtubules in epidermal cells using fluorescent analogue

cytochemistry 64 Physiological probes 66

In vitro calibration of calcium ratio dyes 68 In situ calibration of calcium dyes 71

In situ calibration of ratio pH dyes using nigericin 75 Measurement of H2O2 in situ using H2DCF 77 Data analysis 78

Processing ratio images 79

Quantitative analysis of ratio data from regions of interest 80

Measurement of signal attenuation with depth through a permeabilized specimen infiltrated with a fluorochrome 'sea' 80

Flow cytometry

Sorting of protoplasts and cellular organelles Sorting subclasses of plant nuclei 102

Practical considerations for cell biologists

Preparation of electroporation competent protoplasts from tobacco leaves 113 Electroporation and subsequent incubation 115

Harvesting cells and medium from electroporated protoplasts 116 Extraction and concentration of proteins from cells and medium 117 Isolation of vacuoles from protoplasts after short transient expression 119 Quantifying the recovery of vacuoles 120

Cell fractionation by sucrose density centrifugation 121 Assessment of membrane association 123

Assessment of membrane orientation 123

Transformation methods

Preparation of protoplasts from sterile tobacco plants 132 Preparation of protoplasts from an Arabidopsis cell suspension 134 PEG-mediated transient expression in protoplasts 135

Agrobarterium-mediated transient expression in tobacco leaves 136 PVX vector expression in whole plants 137

Biolistic inoculation of PVX vector constructs 138

Microinjection

Epifluorescence observation of pressure injection of Alexa 488 into epidermal cells 149

Iontophoretic microinjection of Lucifer Yellow-CH potassium salt (LYCH) into a mesophyll cell 150

GEF-mediated injection of 40 kDa dextran-LYCH conjugate into the nucleus of a Mougeotia cell 152

GEF-mediated injection of LYCH in a single chloroplast of a guard cell of a young tissue culture leaf 153

Applications of laser microbeams in plant biology Injection of DNA into plant cells 164

Applications of optical tweezers to plant biology Basic operation of a laser tweezers set-up 166

Capacitance measurements as an assay for exo- and endocytosis: practical considerations

Measurement of single fusion and fission events in the cell-attached configuration using a dual-phase log-in amplifier 177

Measurement of single fusion and fission events in the whole-cell configuration using a dual-phase log-in amplifier 179

Monitoring of macroscopic changes in surface area using whole-cell membrane capacitance measurements 181

Combined fluorescent and electrical measurements of exo- and endocytosis 186

Plant histology

Standard paraformaldehyde-glutaraldehyde fixation 190 Standard FAA fixation 191

Conventional embedding methods and screening 191 Silanized slides 191

Embedding in and sectioning wax 193

Embedding in and sectioning LR White acrylic resin 194 Conventional staining methods 195

Toluidine Blue 195 Acridine Orange 196

Haematoxylin with counterstains 196 Ruthenium Red for pectin 198 Calcofluor for cellulose 198 Phloroglucinol for lignin 198 Aniline Blue for callose 199 Resorcinol Blue for callose 199 PAS for total carbohydrate 200 Iodine staining for starch 201 Sudan staining for lipids 201 Nile Blue for acidic lipids 202

Methyl Green-Pyronin for RNA and DNA 202 Fluorescent stains for DNA 203

Stains for cell viability 204 Methyl Blue for fungi 204 Thionin-Orange G for fungi 204 Gram stain for bacteria 205

Principles and use of immunocytochemistry Fractionation of IgGs from serum 211

Tissue printing for observing immunological and protein profiles 216 Generic protocol for immunofluorescence 218

Immunofluorescence on root squashes 220

Immunofluorescence on suspension culture cells or protoplasts 222 Immunostaming of pollen tubes 223

Gold labelling and silver enhancement of mung bean hypocotyl microtubules (IGSS method) 224

Immunostaining on de-waxed sections 225

Methacrylate embedding, sectioning, and immunostaining 226 Immunofluorescence on cryosections of roots 228

Immunostaining by freeze-shattering of plant cell walls 229

Transmission electron microscopy

Preparation of a typical double aldehyde fixative (1% paraformaldehyde/ 2% glutaraldehyde) 237

Fixation and dehydration 237 Resin embedding 240

Progressive lowering of temperature and acrylic resin embedding 242 Preparation of Reynold's lead citrate 244

Zinc iodide/osmium tetroxide impregnation (ZIO) 245 Phosphotungstic acid staining of plasma membranes 245 PATAg staining of carbohydrates 246

On grid negative staining 247

Standard immunogold labelling and silver enhancement 249 Typical freeze-substitution schedule 253

Freeze-fracture and replication including deep-etching and rotary replication 256 Rotary metal shadowing of isolated proteins 258

Basic preparation of material for ambient temperature SEM 259 Critical-point drying 260

Fracturing and osmium maceration 261 Cryo-SEM 263

Chromosome preparation for DNA:DNA In situ hybridization

Preparing mammal chromosome spreads 270 Preparing plant chromosome spreads 272

Material preparation for RNA:RNA in situ hybridization 273 Coating of slides for cryosections 274

Preparation of cryosections 274

Acridine Orange staining to check retention of nucleic acids in cryosections 275

Labelling the nucleic acids

Labelling of DNA by nick translation 277 Labelling DNA using PCR 278

Labelling RNA by in vitro transcription 280

Precipitation of probe (for DNA or RNA probes) 281 Checking label incorporation: dot blot 282 Material pre-treatment 283

Pre-treatment for DNA:DNA in situ hybridization

Pre-treatment for RNA:RNA in situ hybridization: cryosections 284

In situ hybridization reaction

DNA:DNA in situ hybridization reaction 285 RNA:RNA in situ hybridization 287

DNA:DNA post-hybridization washes 288 RNA:RNA post-hybridization washes 288

Probe detection and visualization

Probe detection 290

Organelle isolation

Detergent-assisted homogenization of protoplasts 298 Isolation of chloroplasts 300

Isolation of intact chloroplasts from spinach leaves 300 Isolation of mitochondria from spinach leaves 301 Isolation of nuclei from tomato leaves 303 Isolation of peroxisomes from spinach leaves 304

Isolation of plasma membranes by aqueous phase-partitioning 305 Large scale preparation of vacuoles from storage tissue 307 Preparation of tonoplast membranes from isolated vacuoles 308 Isolation of endoplasmic reticulum 310

The isolation of intact dictyosomes 312 Isolation of clathrin-coated vesicles 313

Abbreviations

A surface area AM acetoxymethyl

AOTF acousto-optic tuneable filter BSA bovine serum albumin c specific capacitance CBN conjugate bond number CCD charge coupled device CCV clathrin-coated vesicles CFP cyan fluorescent protein Cm membrane capacitance Cp capacitance of membrane patch Cpip (stray-) capacitance of patch electrode Cpm capacitance of whole plasma membrane CV coefficients of variation

CWL centre wavelength DCB 2,6-dichlorobenzonitrile DEPC diethylpyrocarbonate

DIC differential interference contrast DPI dots per inch

DTT dithiothreitol DW distilled water EM electron microscopy ER endoplasmic reticulum

Erev reversal voltage of total plasma membrane currents FAA formalin/acid/alcohol mixtures

FA-BSA fatty acid-free bovine serum albumin FACS fluorescence activated cell sorter FALS forward angle light scatter FDA fluorescein diacetate

FLIM fluorescence lifetime imaging microscopy FLIP fluorescence loss in photobleaching

FWHM full width at half-maximum GEF galinstan expansion femtosyringe GFP green fluorescent protein GST glutathione S-transferase GUS B-glucuronidase

HA haemagglutinin HBW half band width

I total current passing the plasma membrane Ic capacitive current

Ir resistive current LM light microscopy LED light emitting diode LUT look-up table

MS Murashige and Skoog salt mixture NA numerical aperture

NLS nuclear localization signal NRA Naturstoffreagens A PCR polymerase chain reaction PEG polyethylene glycol PHA phytohaemagglutinin PI propidium iodide pm plasma membrane psf point spread function PVX potato virus X QE quantum efficiency QY quantum yield Ra access resistance RALS right angle light scatter Rm membrane resistance r.m.s root mean square ROI regions of interest ROS reactive oxygen species Rp resistance of membrane patch Rpm resistance of whole plasma membrane RS seal resistance

SDS sodium dodecyl sulfate SEM scanning electron microscopy S/B signal-to-background

S/N signal-to-noise SV secretory vesicles

Chapter 1

Introduction to optical

microscopy for plant cell biology

P.J Shaw

Department of Cell Biology, John Innes Centre, Colney, Norwich NR4 7UH, UK

1 Introduction

The development of the optical microscope in the seventeenth century opened up new areas of study in many fields of science In particular, the observations of plant and animal tissues and micro-organisms gave rise to cell biology, although our modern idea of what a 'cell' actually is arose somewhat later Some of the data recorded by the early microscopists, using only primitive microscopes and drawing freehand what they saw, are quite remarkable Hooke, Malpighi, and Leuwenhoek are the best known seventeenth century microscopists, but Nehemiah Grew was the first true specialist in plant microscopy He produced detailed descriptions of plant microanatomy which have proved remarkably accurate An example is shown, along with Grew's original legend (1), in Figure 1. Although there is only space to show one such picture, the original volumes contain many pages of such detailed, painstaking hand-engravings With today's access to photography, video cameras, and computer image processing, this should serve as a reminder of how much can be achieved with careful observa-tion and a very simple microscope Many biologists overlook the useful informa-tion that can be rapidly obtained by even a very simple laboratory microscope

Figure A transverse section of a horse-radish root, hand-drawn and hand-engraved by Nehemiah Grew (1) in 1673 Grew's original legend is as follows:

Fig A slice of the lower part of the root of Horse-radish cut traversly, as it appeareth to the bare eye a The skin ac The bark, with the succiferous vessels therein represented by the smaller specks Within stand the air-vessels represented by the larger and blacker specks e. The pith.

Fig The same slice, as it appeareth through a microscope AA The skin A.8 The bark B.L. The succiferous vessels therein postured in the form of a glory B.G The air-vessels postured in a thick ring; the several conjugations whereof are radiated G.E Other succiferous vessels within the air-vessels postured in a thin ring E The pith ee The bubles of the pith.

(Reproduced from ref with permission.)

refractive index within the specimen In some types of biological specimens This has proved a very valuable technique

is a relatively non-destructive method of probing cells, and this makes optical studies of living cells possible, often coupled with microinjection (Chapter 5) or other loading of fluorescent markers In the past few years the introduction of the green fluorescent protein (GFP) has started a revolution in the microscopy of living cells and organisms This naturally fluorescent protein can now be ex-pressed transgenically in virtually any organism or cell of interest, and can be used for a wide variety of in vivo studies Most exciting, the GFP gene can be fused to the gene for another protein of interest to express a GFP-protein fusion Surprisingly, in almost all cases, the chimeric protein behaves in the same way and is localized as the original protein This means that it is becoming possible to analyse virtually any protein of interest in living cells, opening up enormous opportunities for future cell biology research

In parallel with these developments in fluorescent probes, there have been radical developments in optical microscopy itself, notably the invention of the confocal microscope This was suggested by Minsky in 1957 as a method for increasing the resolution of an optical microscope It was first implemented by Petran for reflection imaging Although confocal methods have been also de-vised for transmission imaging, it is with fluorescence imaging that confocal microscopy has had the greatest impact in biological microscopy Fluorescence microscopy is a dark field imaging mode, with very bright structures contrasted against a black background In conventional epifluorescence imaging the con-tribution of each part of the specimen to the coarse features in the image extends a long way either side of the focal plane whereas the fine image detail is rapidly attenuated away from the focal plane This gives rise to a high back-ground out-of-focus contribution that tends to obscure the fine structure and makes the images generally hard to interpret in detail The confocal arrange-ment, which is described below, excludes nearly all of the out-of-focus com-ponent and thus produces clearer fluorescence images—clean 'optical sections' It is also the best technique for measuring focal section series, and thereby obtaining true three-dimensional reconstructions of cellular and subcellular structures

The confocal microscope produces optical sections by manipulation of the light in the microscope An alternative method for 3D optical imaging is to use computer image processing to remove the out-of-focus light from each image by calculation This method, invented nearly twenty years ago, is called deconvolu-tion or deblurring, and is now becoming much more widespread as electronic cameras improve, and computers become more powerful and less expensive

2 Explanation of terms

Before describing the different types of optical imaging, a brief description some of the technical terms that are encountered in optical microscopy will be given, particularly in the description of the optical components A more complete glossary is given in ref The primary image forming lens is called the object-ive It is invariably an assembly of several optical components Another lens assembly, the condenser, focuses illuminating light on the specimen for trans-mitted light microscopy The final image is produced by the eyepiece or ocular. Each objective lens has a type description, and some numbers engraved on the side of the barrel; for example 'Plan 25/0.08 160/0.17' The type description word denotes the level of correction of aberrations of the objective

Optical lenses suffer from various aberrations, and these are corrected to different extents in different types of objective In achromats, the chromatic aberration (bringing light of different wavelengths to different focal planes) is minimized for two wavelengths (usually one below 500 nm and one above 600 nm) In apochromats, the chromatic aberration is minimized for three wavelengths (generally about 450 nm, 550 nm, and 650 nm) In plan objectives the curvature of field (most noticeable at the edge of the field of view) is minimized In the objective terminology this is also used as a prefix—e.g plan-achromat or plan-apochromat Other terms are used by individual manu-facturers to describe special features; for example, Zeiss call objectives which are designed for UV transmission neofluar and plan-neofluar, whereas other manufacturers use different terms such as fluor, UV UV transmission is neces-sary for the excitation of some fluorochromes, such as the DNA dye DAPI Plan-apochromats from some manufacturers transmit light in the near UV and can be used for DAPI, those from others not Generally plan-apochromats are the best objectives (and the most expensive)

The first number after the type is usually the magnification Objective magnifications range from X l to X l00 The most useful magnifications for cell biological work are: a low magnification (X16 or X25); intermediate X40; and high X63 (or X60) Next to the magnification is the numerical aperture, a number greater than zero and less than 1.5 Technically it is the refractive index of the immersion medium multiplied by the sine of the aperture angle of the lens—in essence it tells you the angle of the cone of scattered light the objective collects The larger the numerical aperture, the greater the light collection efficiency and the greater the resolution obtainable

approximately 0.25 um (depending on the wavelength of the light) Some object-ives have an adjustment collar to allow the numerical aperture to be changed The highest available numerical aperture should normally be used, but it is occasionally useful to be able to decrease it, for example for dark field imaging (see below), or to increase the depth of field Many objectives are designed for a tube length of 160 mm, which is usually the next number engraved on the side of the objective This is the distance between the objective and the eyepiece Some objectives are termed infinity-corrected, which has the advantage of allow-ing an arbitrarily long distance between the objective and the eyepiece As far as the optics are concerned, 160 mm tube length objectives are interchangeable between different microscopes, although some of the correction of chromatic aberration is often in the eyepiece, so eyepiece and objective should strictly be matched for optimal performance But infinity-corrected objectives cannot be interchanged with 160 mm ones, unless a special correction lens is used Thus, until recently objectives could be interchanged between microscopes fairly easily, because all the manufacturers used a common thread (defined by the Royal Microscopical Society—RMS) for mounting them Unfortunately all the major manufacturers now use their own, different, threads on their objectives, making it virtually impossible to exchange objectives between different microscopes This development is anti-competitive and regrettable

A final number may be engraved on the objective, usually 0.17 This corres-ponds to a coverglass thickness of 0.17 mm It means the objective has been designed for a coverglass of this thickness, and you should not use any other thickness—using any other coverglass thickness, whether thinner or thicker, will degrade the optical performance of the objective, mainly by increasing the spherical aberration It is a good idea never to have any other thicknesses in the laboratory to avoid confusion

particularly easy to see with epifluorescence Focus on a small bright spot and observe how the spot changes either side of focus In the absence of spherical aberration it should expand and fade equally either side Almost invariably you will find it is asymmetric, disappearing quickly one side, but slowly the other side An image of a single point such as this is called the point spread function It is an important characteristic of a microscope, especially if you are interested in detailed 3D imaging or image processing Within certain limits, if you know how a single point is imaged, then you can determine the image which should result from any specimen (at least for fluorescence and bright field imaging)

A final objective characteristic which is often important is the working distance This is the distance between the focal plane and the front of the objective It is usually very small (perhaps 0.2 mm) for high numerical aperture lenses, since to obtain a large collection angle it is necessary either to have the lens close to the specimen or to have very large diameter lenses, which are expen-sive It is possible to obtain objectives with particularly long working distances, of the order of tens of millimetres (for example the Nikon SLWD range, which still have quite large numerical apertures) This type of lens is normally used on inverted microscopes for observation of samples in containers such as Petri dishes, but is also very useful for micromanipulation and microinjection on a non-inverted microscope, since it gives reasonable access to the specimen during observation

The condenser is crucially important in transmission imaging Its role is obvious in phase contrast and dark field imaging (see below), but it is often not appreciated that its contribution to the overall resolution is equal to that of the objective in bright field and DIC imaging For high resolution, you should use a high quality plan-apochromat condenser If the objective is an oil immersion one, then the condenser should also be an oil immersion lens if possible A large research microscope will have a number of different condenser positions A wheel in the condenser is rotated to bring different phase rings, dark field apertures, and DIC prisms into the optical path

The eyepieces on a microscope are rarely, if ever, changed Their magnifica-tion is calculated so that the image is magnified enough for the human eye to see all the significant detail This is usually X10 or X20 In binocular microscopes there is a focusing collar on one or both eyepieces If the microscope has a camera attached, one eyepiece will probably have a photoscreen graticule in it, or there may be an eyepiece graticule If so, focus on the specimen and, looking only through this eyepiece (shutting the other eye), adjust the eyepiece collar until the graticule is sharply in focus Then the specimen and graticule should both be sharply in focus Now look through the other eyepiece and adjust it to bring the specimen sharply into focus This will have adjusted the eyepieces for your eyes, and ensured that when the image looks in focus to you, it will be in focus at the camera (provided the camera has been set up correctly)

there is another intermediate magnification-changing lens (sometimes called an optovar) It is best to measure the actual magnification of your microscope To this you need a stage micrometer—a microscope slide with a scale engraved on it An eyepiece graticule is also useful This is a glass plate with a dimen-sionless scale engraved on it, which is inserted in the eyepiece It is calibrated for each objective magnification using a stage micrometer, and is used to estimate the size of objects through the eyepiece

3 Recording images

For all but very routine microscopy it is a good idea to use photography or other image recording almost as a matter of course Somehow you can never find such a good field of view as the one you saw when taking a quick look! The moral is— record it, and take each image as if it was a picture for publication, because you never know in advance which ones you will actually want to publish

3.1 Image resolution

The microscope eyepiece in its normal configuration produces a virtual image, which is then projected as a real image onto the retina by the lens of the eye Recording the image onto film or electronically requires that a real image is produced by the microscope This is usually achieved by a projector lens, which is optically often the same as an eyepiece ocular However this lens needs to be at a slightly different distance from the objective in order to form a real image of the same focal plane as the eye sees through the regular eyepiece The position of the projection ocular is usually adjustable, so that the plane of focus seen directly and that projected onto the camera are the same and are both accurately in focus The total magnification produced by the microscope is the product of the objective magnification, the eyepiece or projector magnification, and the microscope intermediate magnification factor (including optovar magnification if present) The magnification of the regular eyepiece is chosen so that the total magnification produced is sufficient to enable the human eye to see the finest details that the objective can produce Any further magnification will not pro-duce any more image detail, and is often called 'empty' magnification However, an image recording device will have different resolution characteristics from the human eye, and a different projector magnification will be needed, which may be more or less than the human eye requires

magnification of XI25 is needed Given an objective magnification of X60, the rest of the optics—the intermediate magnification and the projector—need to produce a total of about X2 It is not always easy to find the intermediate magnification figures needed for the foregoing calculation Therefore it is best to use a stage micrometer to measure the magnification of the whole microscope optical system directly

Although photographic film is not as sensitive as the best low light level cameras, it does have an extremely high capacity as an image recording medium With a pixel resolution of 10 um, a 35 mm film frame records 3500 x 2400 pixels This is much more than most cameras A typical video camera records 512 X 768 pixels, and a 'high resolution' camera may record something around 1024 X 1200 pixels Thus in calculating the projector magnification needed for an elec-tronic camera, a compromise must be made If it is necessary to retain the full resolution of the image, only a small part of the field of view visible through the eyepiece can be recorded At 512 x 768 pixels this would be a field of around 40 x 60 um Alternatively, if the projector is chosen to keep the full field of view, the resolution will be seriously degraded The only real solution is to use different projector lenses for different purposes

3.2 Film recording

For immunofluorescence, the author's laboratory uses Kodak TMAX film quite extensively for black and white photography; it has good contrast and low grain, and is available in a range of speeds (TMAX 400 is the most often used in the author's laboratory) Kodak Technical Pan film is much slower and has lower contrast, but is more suitable for bright field or DIC images Technical Pan film with liquid Technidol low contrast developer is useful for photographing video and computer displays; most other films have too high a contrast for this applica-tion For colour slides, use tungsten colour rated films if using a tungsten lamp, daylight rated films for a mercury arc lamp (for example when photographing immunofluorescence specimens in their original colours, or photographing colour video or computer screens) For colour prints, it is recommended that slide film is used and prints from the diapositives (Cibachrome prints)

Most modern exposure meters use centre-weighted average metering This is usually fine for bright field types of image, since the brightness of the area of interest is about the same as the rest of the field However, it is not appropriate for other types of image, particularly fluorescence, since the objects of interest are much brighter than the dark background, and the average brightness is near to the background value Here, the best solution is to use spot metering, where the exposure is set in a small area in the centre of the field Move the object of interest into the central area to set the exposure If you not have spot meter-ing you will have to use trial and error As a guide, usmeter-ing 800 ASA film, a reason-ably bright cytoskeleton immunofluorescence image might need an exposure of 30 seconds, DAPI stained nuclei may need only second

handled in a computer and later printed for publication or made into a slide To obtain the maximum amount of information on the film, it will need to be scanned at something like a 10 um raster Film scanners and printers are usually rated in dots per inch (DPI) Thus scanning at 10 um requires about 2500 DPI or better Specialized negative scanners will provide this resolution, although most cheaper flat-bed scanners will not However, it is not always necessary to scan images at the maximum resolution, and handling the very large files that result can cause problems The best approach is to work back from whatever final out-put is needed So, for example, if the final image is to be on a printer which has 400 DPI resolution and is to be inches X inches, this will require 800 x 800 pixels in the image file If the area to be printed is 20 mm X 20 mm on the original negative, it will only need to be scanned at 40 pixels/mm, which is about 1000 DPI

3.3 Electronic cameras

The first video cameras used a semiconductor target, such as silicon, to record the incoming light The pattern of light projected onto the target left a corres-ponding pattern of charge distribution which was read out by scanning an electron beam across it and amplifying the resulting current This raster scan was defined by the broadcast television standards—typically 25 frames per second at 625 horizontal lines The resulting analogue signal could be displayed directly to a monitor To convert it to a form that could be handled by a com-puter, it had to be digitized by a device called a frame-grabber—a fast A to D converter and computer image memory

Alternative electronic imaging devices began to appear in the early 1980s These devices, called charge coupled device (CCD) cameras effectively record the incident light digitally from the start Made by semiconductor microfabrication methods, CCD chips contain an array of 'charge wells' Photons falling on the chip cause electrons to be produced and trapped in the wells After a given exposure time, the charge in each well is read out and recorded The early CCD cameras used a direct digital interface to send the electron counts for each well directly to computer memory This is still the circuitry used in expensive scien-tific slow scan CCD cameras Today virtually all video cameras use CCD chips as the recording element However, to make them compatible with CCTV and broadcast standards, the image is read out at video frame rate and converted to an analogue video signal Thus, ironically, what was originally a digital image is converted to a poorer analogue signal, and must still be digitized into a computer by a frame-grabber

electrons The readout noise increases with the readout speed, and so is very much higher in video rate cameras The noise can be reduced by accumulating multiple frames in a computer frame-grabber/frame-store, but the readout noise is introduced into every video frame read, so it is better to have 'on-chip' integration, where the chip is exposed for a longer period of time before the image is read out Thus, readout noise is only introduced once Secondly, CCD chips accumulate 'dark current'—thermal electrons are produced and are trapped in the charge wells along with the photon-produced electrons In video rate cameras, the dark current is small because of the short exposure time of each frame, but with long on-chip exposures, the dark current can become very large The solution to this problem is to cool the CCD chip The dark current can be reduced to insignificant levels for microscopical imaging using thermoelectric cooling to -40°C

A final application of CCD chips is in the current computer digital cameras In these cameras the chip is exposed and then read out as a digital image They are therefore intermediate between slow scan, scientific cameras and video cameras Currently even relatively cheap consumer cameras are available with resolutions around 2000*2000 pixels Since they are not cooled, they will suffer from dark current, and the maximum exposure will be limited compared to a cooled CCD camera However they are becoming an attractive option for a cheap and simple way of recording images electronically from a microscope The current genera-tion is capable of recording bright field microscope images, and fluorescence images bright enough to be visible by eye, but not very weak fluorescence images and definitely not bioluminescence images The main problem is that most of the current consumer cameras have a built-in lens for standard photography, and this makes it hard to mount them satisfactorily on a microscope What is needed is a camera without a lens which has a standard bayonet or C-mount fitting

4 Microscope imaging modes

Microscope imaging may be divided into transmission modes, such as bright field, dark field, phase contrast, and differential interference contrast (Nomarski), and epi-illumination modes, primarily epifluorescence and reflection contrast Con-focal microscopes have been developed which use either type of illumination, but epifluorescence and reflection contrast are much simpler to implement in a confocal arrangement and to date have been by far the most widely used in biology Deciding what type of imaging to use is sometimes straightforward and sometimes difficult It is important to be aware of the different imaging modes available and of their capabilities, and to be sufficiently familiar with the microscope to try different imaging methods with a specimen

4.1 Bright field imaging

using any microscope Always check this adjustment before using the micro-scope for anything else (except epifluorescence, dark field, and reflection), and check it every time you change objectives or insert any other element into the light path

Light from each point on the bulb filament is focused by the collector lens to a point at the condenser aperture At the field aperture this gives a wide area of even illumination Since the field aperture is at an equivalent focal plane to the specimen, the specimen is evenly illuminated Light from the specimen is focused by the objective at a plane known as the primary image plane The eye-piece then forms an image a small distance from the top lens of the eyeeye-piece

You should refer to the handbook for your microscope for a detailed descrip-tion of the parts of your microscope and how to adjust it A descripdescrip-tion is given here which should apply in general to any microscope, Examples of bright field images are shown in Figures 2A and 3A.

Adjustment for bright field Imaging with Kohler illumination

Equipment

• All microscopes with a standard bright field condenser should be capable of this type of imaging

Method

1 Focus on a specimen with the condenser and field apertures wide open,

2 Close the field aperture until you can see its edges

3 Focus the condenser up or down until the edges of the field aperture are sharp If necessary, centre the image of the field aperture with the condenser centring

adjustments

5 Close the condenser aperture until the glare around the field aperture just dis-appears, and then reopen it a little (if in doubt, open the aperture about halfway) Alternatively readjust the condenser aperture during observation of the specimen to obtain the best linage Opening the condenser aperture increases the resolution, but decreases image contrast and increases bright flare around objects Closing it decreases flare and increases image contrast, but decreases resolution

Figure Filament of the alga Spirogyra grevettiana (A) Bright field image The contrast is fairly low but the characteristic spiral chloroplasts are visible (B) Phase contrast image of the same cells The contrast is greatly enhanced, but note the prominent haloes around all the structures, particularly the cell wall (C) DIC image of the same cells This is probably the optimal imaging mode for this specimen, showing fine structure inside the cell, good contrast, without the haloes present in the phase image (Reproduced from ref with permission.)

4.2 Phase contrast

Figure A wax section of wild-type flower of Antirrhinum majus labelled with an in situ probe to the transcript of the gene floricaula detected by alkaline phosphatase-conjugated antibody (A) Bright field image, showing the distribution of the transcript as revealed by the insoluble dark-coloured product of the alkaline phosphatase reaction, but poor detail of the cell structure (B)The equivalent dark field image The crystalline product of the alkaline phosphatase reaction reflects light strongly, and is thus seen very clearly (C) For comparison the phase contrast image of a portion of the same specimen is shown The contrast is better than (A) but poorer than (B) (Reproduced from ref with permission.)

have- had a small specimen-dependent phase change introduced and most of it will be scattered so that it no longer passes through the phase ring—thus it will be additionally retarded by another quarter wavelength, and will interfere de-structively with the unscattereci light So the specimen-dependent phase changes will be turned into changes in amplitude, and be visible

areas are related to the refractive index within the specimen Small structures can appear bright or dark at slightly different focus levels A dark structure might be a dense object, or it might be something else like a hole or a vacuole Rings and haloes around structures are also highly visible, and the resolution is less than in bright field For these reasons, although phase contrast can often be very useful, the resulting images should be interpreted with caution Sec Figures 2B and 3C for examples of phase contrast images.

Adjustment for phase contrast

Equipment

• Phase contrast requires special objectives having a phase ring and a matching

condenser annular aperture—it cannot be done on other types of objectives

Method

1 Set the microscope up for Kohler illumination in bright field (Protocol 1).

2 Open the condenser aperture fully, and either replace one eyepiece with a phase telescope, or, more usually, move a Bertrand lens into the light path (usually by a rotating wheel somewhere between the lens turret and the eyepiece)

3 Observe the back focal plane of the objective, where the phase ring is located You should see two rings—a bright one corresponding to the condenser annulus and a dark one corresponding to the phase ring,

4 Adjust the position of the condenser annulus, so that the phase rings overlap sym-metrically This is accomplished by a centring device (not the same as the condenser centring mechanism) On some modern microscopes the annuli are pre-centred and will not require adjusting If the two rings are of very different radii then the condenser annulus is not the correct one for the objective

5 Remove the phase telescope and return to normal viewing; you should now have phase contrast imaging,

4.3 Differential interference contrast (Nomarski)

vectorially The overall effect of this arrangement is to produce a resultant image which is the sum of the two displaced images Where two points summed have the same phase, the resultant summed image point will have the same plane of polarization as the original incident polarized light, and thus be blocked by the second polarizing filter However, where there is a phase difference between the two points the resultant will have a rotated plane of polarization, and a com-ponent of this light will pass through the analyser The result is an image which maps changes in refractive index within the specimen The Wollaston prisms also introduce a specimen-independent relative phase difference between the two images; this additional phase difference can be changed by displacing one prism, which alters the final image seen This displacement is provided via a knurled knob/lead screw device on most microscopes In some microscopes a quarter-wave plate compensator is used rather than displacing a Wollaston prism

DIC images can be difficult to interpret The method displays an image of changes in refractive index in the direction defined by the orientation of the prisms For this reason it is particularly good at revealing edges in biological structures—for example, organelle and nuclear boundaries, cell boundaries, and cell walls It has also been used very effectively to image fibrous subcellular com-ponents such as microtubules, often in combination with video enhancement One side of an edge will be brighter than the background, the other side will be darker For this reason the images have a false 'shadowed' appearance This is visually very attractive and can be very informative However it can also be very misleading, since it makes the images seem three-dimensional because of the way the human brain interprets lighting effects It is important to realize that these images are not three-dimensional ones, they merely give that impression To obtain true three-dimensional information focal sectioning must be used In a simple way this can be achieved by focusing up and down through the specimen Because DIC can use the highest condenser aperture available it can give smaller depths of field than other conventional imaging modes, and so is better for focal sectioning For highest resolution with DIC, a condenser with a numerical aperture as great as that of the objective should be used If the objective is a high numerical aperture oil immersion one, the condenser should be also, and should be used with oil—this is rarely appreciated, still less actually done A DIC image of a Spirogyra filament is shown in Figure 2C

Adjustment for DIC

Equipment

• For DIC a microscope must have provision for the two polarizing filters and the two prisms in addition to standard bright field optics In principle DIC should work for any high quality objective lens It was

Method

1 First adjust the microscope for Kohler bright field illumination (Protocol 1) with the condenser set to the DIC position (condenser prism in place), but with the objective prism out of the optical path

2 Open the condenser aperture, and insert the polarizing filters On some micro-scopes the polarizing filters are fixed, on others they can be rotated If the polar-izing filters are rotatable, rotate one of them until the field is maximally dark Insert the objective prism, and adjust the translation to obtain the 'best' image It is

up to the microscopist to judge what the best image is However, it is generally best to have the condenser aperture as far open as possible Closing the aperture, while increasing the contrast, will decrease the resolution

4.4 Dark field

All the foregoing techniques, with the exception of polarized light microscopy, are essentially bright field methods; that is, darker structure in the specimen is seen imaged against a bright background This is very familiar and readily interpretable However it can be difficult to see small, weakly imaged features, since they represent small changes against a large background, and therefore provide inherently low contrast In dark field techniques, the background is low and the specimen structure is bright, and so the contrast is much greater It is particularly good for reflective structures such as silver grains in autoradiographs and silver-enhanced, gold-labelled immunocytological specimens (see Chapter 10) Plant cell walls also show up brightly in dark field A dark field image is shown in Figure 3B.

Adjustment for dark field

Equipment

• A microscope with a dark field condenser This is generally a 'patch stop' condenser, which has a central disc to block the direct

light beam If the condenser is an oil immersion type, use oil and use the maximum condenser aperture

Method

1 Adjust the microscope for bright field, Kohler illumination (Protocol 1), and focus on your specimen in bright field,

3 If you are using a patch stop condenser which requires centring, use the phase telescope or Bertrand lens to view the back focal plane of the objective Focus the condenser up and down until the patch stop is seen as clearly as possible, then centre it

4 Switch back to normal viewing and adjust the condenser focus as before for the best dark field image.b,c

a It may be necessary to readjust the centring of the patch stop; look for the location of the

darkest part of the image as the condenser is focused, check the centring of the condenser patch stop, and recentre if it is not central

bIn contrast to bright field, the angle of the illuminating cone of light must be greater than that of the objective aperture, otherwise the unscattered light will not be excluded from the image This may mean that dark field will not work effectively with very high numerical aperture objectives, unless the condenser can provide an even greater effective numerical aperture, and it may therefore not be possible to get a dark background unless you can stop down the objective numerical aperture

cAt low magnification, the largest phase annulus may work as a patch stop Otherwise it is possible to improvise one with a disk of card underneath the bright field condenser—there is usually a suitable filter holder above the condenser accessory lens

4,5 Epifluorescence and reflected light microscopy

The previous methods described are all transmission imaging modes—the speci-men is illuminated from the side opposite the objective by means of a condenser lens In epifluorescence and reflection imaging, the specimen is illuminated by light which is introduced from the same side of the specimen as the objective, usually through the objective, which therefore acts as its own condenser This means that it is not necessary to adjust the condenser for these modes (although it is a good idea to get into the habit of always checking the bright field adjustment before using a microscope)

to cut out UV light, and a heat filter to prevent damage to the polarizer Com-mercially available reflection systems use circularly polarized light, which has some advantages, but cost more For very low magnification reflection imaging a dissecting microscope with an annular illumination collar which fits around the objective lens works well

Epifluorescence is probably the most important optical technique in use in subcellular imaging studies at the moment The details are described in later chapters; only the principles will be discussed here Fluorescent molecules absorb light of particular wavelengths, and then re-emit light at longer wavelengths and in all directions, the energy difference ultimately heating the specimen Epi-fluorescence achieves extremely high specificity and low backgrounds (and thus sensitivity) by a double discrimination—in frequency and direction—against the incident light and in favour of the emitted light A mercury or xenon arc lamp is the most usual light source The light passes through an excitation filter, which matches the absorption characteristics of the fluorescent label being used The light is then reflected down through the objective lens by a dichroic mirror; this is a special type of mirror which has wavelength-dependent reflectivity Light at shorter wavelengths than the mirror's cut-off is reflected, whereas light at longer wavelengths passes through It is thus both more efficient than a semi-silvered mirror and more wavelength selective Back-scattered light from the specimen will have the same wavelength as the incident light and will therefore be re-flected again by the dichroic mirror and will not reach the eyepiece Fluorescently emitted light from the specimen passes through the dichroic mirror then reaches another specific filter, the emission filter, which only passes light of the emission wavelength of the fluorescent marker, and further discriminates against non-fluorescent light

Provided you have the correct components on your microscope, and provided the mercury lamp has been correctly set up, there is no alignment necessary for epifluorescence Some precautions are advisable with the mercury lamps gener-ally used as the light source Many mercury bulbs have a rather short life, and are also expensive The life is considerably shortened by frequent switching on and off It is a good rule not to switch the bulb off for at least 30 minutes after switching on, and not to switch it on again until at least 30 minutes after switch-ing off The recommended life is only 100 hours for the older types of bulb, although it is much longer for some more recent mercury bulbs Resist the temptation to run the bulbs longer than their recommended time; they can explode if this is done, which is highly dangerous since mercury vapour is then released into the air If this happens, evacuate and close the room until the mercury vapour has had a chance to disperse

for DAPI, from others they not, and special UV-transmitting objectives must be used Use the highest numerical aperture objective you can; because the objective also acts as the condenser, the amount of fluorescent light gathered is very highly dependent on the objective numerical aperture It is well worth measuring transmission spectra of all the filters and dichroic mirrors in the microscope, and comparing them with the fluorescent probes you wish to use Do not assume the limited choice provided by the microscope manufacturer necessarily provides the best solution for your particular needs, and make enquiries from a specialist filter manufacturer if necessary

5 Confocal and 3D microscopy

In the following section, only epifluorescence microscopy will be considered, since this is by far the most common technique used in biological confocal and 3D microscopy More detailed discussions are given in refs 7, 9, 11, 12

5.1 The problem of out-of-focus light

In spite of its advantages, conventional epifluorescence microscopy has some troublesome features In principle if all the fluorescent light emitted from a specimen could be recorded, it should be possible to reconstruct a perfect map of the distribution of the fluorochromes which emitted the light, at least to the resolution limit specified by the wavelength In practice this is not the case The reason lies in the geometry of a practical microscope The emitted light must be collected by passing through an objective lens, and this of necessity introduces an aperture into the system The result is that some of the emitted light cannot be recorded, and the image reconstructed by the optical system has the resolu-tion seriously degraded, particularly in the direcresolu-tion of the optical axis (usually denoted z) Furthermore the degradation depends on the level of image detail in a rather complicated way Thus, fine details, often called high spatial frequencies, are relatively little affected, whereas large structures, or low spatial frequencies, are greatly spread out in the z direction This is often called the problem of 'out-of-focus' light A small sphere ends up being imaged in three dimensions as a complicated elongated structure

Currently, there are two feasible methods which overcome the problem The first one is confocal microscopy, in which the optical system is modified so as to minimize the problem by eliminating or at least reducing the contribution of the out-of-focus light The second approach is to measure the conventional image accurately, out-of-focus light included, to measure also the properties of the imaging system in detail, and to use image processing techniques to remove the out-of-focus contribution from the image This section will attempt to describe in as non-technical a way as possible the basis of these approaches There has been much published in recent years about confocal microscopy One of the best sources of information is the second edition of the Handbook of biological confocal

5.2 The confocal principle: explanation by ray optics

The most common way to explain the operation of a confocal microscope is by a ray diagram similar to the one shown in Figure A light source, almost always a laser, is used to provide an effective point source illumination through a pin-hole At the specimen this gives the diffraction limited image of a point With a conventional light source this would be the Airy pattern With a laser beam the focus is somewhat different, but may still be thought of as approximating an Airy pattern Some of the emitted light which originates from this plane passes back up through the objective lens, through the dichroic mirror and emission filter, and through a second pin-hole, finally being detected by a photomultiplier behind the detector pin-hole The two pin-holes—illuminating and detecting-are both located in planes conjugate to the plane of focus in the image ('confocal' is derived from a contraction of 'conjugate' and 'focal') However, the out-of-focus problem arises because parts of the specimen above and below the plane of focus are also illuminated and also emit light Now consider rays of light origin-ating from above the plane of focus (the light shaded rays in Figure 4) Originorigin-ating nearer to the objective lens, these rays are brought to a focus further away from the objective on the other side, behind the detector pin-hole At the level of the detector pin-hole, these rays have not converged sufficiently to pass through the pin-hole, and so they are eliminated from the resulting image A similar argu-ment applies to light from parts of the specimen below the plane of focus Therefore the confocal arrangement—the interaction of the illuminating and

Figure Comparison of conventional, wide-field and confocal imaging (A) Conventional fluorescence image, collected using a cooled CCD camera, of a Drosophila embryo (gastrula stage) in which the actin network has been labelled with rhodamine-phalloidin The large amount of out-of-focus light masks much of the fine image detail (B) Confocal image of the same specimen The exclusion of the out-of-focus light leaves the image detail easily visible

Figure Stereo-pair of a 3D confocal image stack A slice of pea root tissue has been stained with the DMA stain DAPI, and imaged using a Bio-Rad 1024 confocal microscope with a UV argon ion laser A depth of about 30 um has been projected to give the stereo-pair

detecting pin-holes—has the effect of discriminating against light originating away from the plane of focus

Figure gives an example which shows the difference that excluding the out-of-focus light by confocal microscopy can make It should be realized that there will only be such a dramatic difference between conventional, wide-field and confocal images in cases where there is a substantial amount of out-ot-focus light present Figure shows a stereo-pair view of a confocal image stack of consider-able depth Notice how all the different layers arc in focus

5.3 Practical confocal microscopes

up an image This is done in several different ways in different designs, and is one of the main differences between the different types of instrument In principle, the illuminated spot can be scanned across the specimen in a raster, in a manner very similar to the scanning electron beam in a TV screen, or the speci-men can be moved through a stationary light path The latter design has the advantage of a simple and accurate optical design, but suffers from lack of speed in scanning an image, particularly at low magnification Some of the first con-focal microscopes built were of this stage scanning type However, virtually all current biological confocal microscopes use a scanned beam design An angular deflection of the light beam at a diffraction plane becomes transformed into a translation in the specimen/image planes Thus, a system of two deflecting mirrors scanning back and forth about two axes positioned at or near a diffraction plane can be used to scan the light beam in a 2D raster across the specimen The scan driving and measurement circuitry are interfaced together so that light intensity measurements are taken which cover the specimen area in a regular raster These intensities are digitized into a computer to produce a digital image Gen-erally image accumulation and averaging are provided, either frame by frame, or line by line, or both Often, the scanning rate can also be changed This principle is used in confocal microscopes made by several companies, including Bio-Rad, Leica, and Zeiss With currently available mirror deflection systems, the maxi-mum scanning rate is a few frames/second (i.e well below 'real time' or video scanning rates) A detailed description of the various scanning systems and a discussion of their relative merits is given by Stelzer (10)

In order to increase the scanning to video frame rates or higher, slit scanning designs have been developed Instead of a pin-hole aperture giving a diffraction limited spot, a narrow slit of light is scanned in a direction at right angles to its length across the specimen This is achieved in a similar way to spot scanning by using one or more scanning mirrors and a stationary slit aperture The emitted light is then passed through a narrow detector slit This has the advantage that since only a one-dimensional scan is required, the scanning rate can be much faster than in a point scanning system Furthermore, since a line of the specimen is imaged at one time, the rate of light accumulation from the specimen is much higher However, the disadvantage is that a proportion of the out-of-focus light— that component which is distributed in the direction of the slit—is also detected, and so the optics are only partially confocal This design, therefore, is inherently not capable of producing such clean optical sections as a point scanning system, but the fast scanning speed and bright image produced mean that the image can be observed directly through an eyepiece, as in a conventional microscope

light passes back through the same pin-holes The condition necessary for con-focal imaging is that no emitted light should pass through the 'wrong' pin-hole, and this means that the pin-holes must be spaced far apart relative to their diameters This means that only a limited set of points are imaged by the disc in each position The full image is obtained by spinning the disc rapidly, so that the pin-holes, which are usually arranged in a spiral pattern, scan across the whole image area The main problem with this design is that the light source has to be spread out over the whole of the disc, and so is many orders of magnitude less bright that the single pin-hole/laser arrangement Furthermore, only a very small proportion of the available light passes through the disc to illuminate the specimen, and more seriously, only a very small proportion of the reflected or emitted light passes back through the disc to be detected The result is that it is difficult to record enough light for a satisfactory image, particularly in the case of fluorescence, and much of the available fluorescent light is wasted This has substantially limited the use of this design in biological applications However, recent designs have improved the efficiency and sensitivity of this type of instrument, and it may be used much more in the future for biology

5.4 Imaging and the point spread function

spread out away from the central plane, producing a complicated set of concentric conical structures This is described in refs 9, 12, 13

Thus, in conventional fluorescence microscopy each point of the specimen is replaced by a copy of the psf, whose intensity is directly proportional to the intensity of the point In a confocal microscope, the same argument shows that the illuminating pin-hole, effectively a point source, produces a light distribu-tion at the specimen which is also given by the psf Light is detected from the specimen only if it passes through the detector pin-hole Since light paths are reversible, this means light is detected from a region of the specimen corres-ponding to the image of the detector pin-hole at the specimen—again given by the psf Thus, the overall effect is that the psf is applied twice in confocal imaging—once because of the illuminating pin-hole, and again because of the detector pin-hole As far as the detected confocal image is concerned, applying the psf twice means that the effective confocal psf is the original, conventional psf squared Since the psf comprises a central maximum with lower surrounding maxima, squaring it reinforces the central peak while weighting down the surrounding components It is the subsidiary maxima that give rise to the out-of-focus contribution in conventional microscopy, and so decreasing their relative weight in the confocal psf has the effect of removing a large part of the out-of-focus light

5,5 Deconvolution

As explained in the previous section, if the imaging process is linear and shift-invariant, the image can be described mathematically as the convolution of the specimen with the psf Provided that the psf is the same all over the specimen, then knowing the psf is enough to characterize the imaging properties of the microscope Given the distribution of light emission in the specimen, the resulting image can be calculated as a convolution Conversely, given the image and the psf, it should be possible to calculate what specimen structure gave rise to the observed image This deblurring process is called restoration, and since it generally involves reversing the convolution operation, it is often also called deconvolution There are many different computer algorithms for achieving this calculation, and there is still debate in the field as to which is the optimal method to use Generally there is a trade-off between the time and computing power needed, and the reliability of the results The best methods are based on statistical and probabilistic assumptions, but take a very large amount of computation A few years ago, even the simpler methods required expensive computers, but now that computers are becoming powerful and cheap, this type of image processing approach is being more and more widely used Deconvolution can be applied whenever the imaging can be described by an invariant point spread function Within certain limits (see above for a discussion about the psf) this applies to conventional wide-field and confocal fluorescence imaging, and to conventional bright field transmission imaging, but not to phase contrast and differential interference contrast imaging

noise level in the image With no noise present, a perfect reconstruction can be carried out trivially But, as discussed below, all images are certain to contain noise, often at relatively high levels In the presence of noise, image restoration becomes a technically difficult problem Computer methods for handling this problem have been extensively developed, and suitable programs for optical microscopy are now commercially available (see ref 13 for a more detailed discussion of this topic and a survey of currently available software) The prin-ciple advantage of the confocal microscope is that it eliminates the unwanted out-of-focus light from measurement This is entirely equivalent to the differ-ences in the psfs between confocal and wide-field microscopy that have been described above Thus, it is possible to collect wide-field images taking advantage of the excellent imaging characteristics of scientific grade CCD cameras, and using image processing to counteract the effects of the wide-field psf However, it should be realized that while it may be possible to eliminate or reduce the out-of-focus component from each focal plane by restoration methods, this un-wanted light is present in the measured image, and contributes to the noise in the measured image The more out-of-focus light there is in the wide-field image, the more noise it adds to the image, and the smaller is the fraction of the measured photons corresponding to the in-focus image The confocal arrange-ment eliminates this light from measurearrange-ment and, along with it, the associated noise So the more out-of-focus light is present, the greater the benefit of elimin-ating it from measurement by the confocal arrangement However, although the confocal optics eliminates the out-of-focus light from detection, it does not pre-vent parts of the specimen out of the plane of focus from being illuminated by the excitation light Photodamage is therefore likely to be as great with confocal as with conventional imaging, other factors being equal

Figure shows an example of deconvolution applied to a wide-field

escence image focal section stack from a nucleus of Vicia faba The results of the deconvolution are comparable to what would be obtained by confocal microscopy

5.6 Two photon imaging

Two photon imaging has attracted much attention recently, and commercial systems are available from several manufacturers In conventional fluorescence microscopy, the specimen is illuminated with photons of the correct wavelength to raise an electron in the fluorophore to a higher energy level However, if a high enough intensity of light at double the required wavelength can be used, then the fluorophore can absorb two photons almost simultaneously to produce the required energy level change The combined probability of the double absorption depends on the square of the light intensity distribution Thus, the illuminating psf is the square of the conventional psf (i.e very similar to a confocal psf) Effectively, this means that the exciting intensity is only high enough to allow two photon absorption very close to the focal plane, and so only a small region of the specimen around the focal point either absorbs or emits light (see Figure 8) In practical instruments, a detector pin-hole has also been used to give an improved psf The advantages of this method are that light should only be absorbed near the focal spot, and thus should only cause fading and photodamage at this position in the specimen

Another potential advantage of two photon imaging is that the long illumin-ating wavelengths used—about 1000 nm—are relatively good at penetrillumin-ating deep into specimens Also, UV fluorophores can be excited with visible wavelengths, which avoids the problems associated with UV optics The disadvantage is that a powerful and expensive pulsed laser system must be used Currently, such laser systems cost in the region of £100 000, but it is hoped that more affordable lasers will soon be available Another potential problem is damage to the specimen by the high intensity long wavelength incident light Experience with different

cells is so far limited; some living cells have been imaged without problem, but the presence of any cellular constituents which absorb in the far red/near IR region of the spectrum would have disastrous effects, and cause the specimen to be rapidly destroyed by heating

6 Comparison of conventional, wide-field

fluorescence imaging with confocal fluorescence imaging

A more detailed comparison of wide-field and confocal imaging is given else-where (12), and only a brief discussion of some of the more important aspects of this topic is given here A major limitation of most current confocal microscopes, apart from the spinning disc designs, is that they must use laser light sources, with consequent severe restrictions on the available wavelengths While de-velopments in laser technology will reduce these limitations, it is unlikely that they will disappear in the foreseeable future In contrast, conventional, wide-field fluorescence microscopy can use virtually any wavelength from the visible spectrum and beyond

6.1 Noise and resolution

The major advantage of confocal microscopy over conventional microscopy (often now called wide-field microscopy to distinguish it from confocal) is the elimination of the out-of-focus light This is broadly equivalent to an increase in resolution in the z direction—the optical axis In principle, a confocal micro-scope is capable of better resolution in the image (x, y) plane as well, but this will only be realized if very small pin-holes are used, and if the signal-to-noise ratio is large enough The image detail actually observable will also depend on the noise level in the image; the higher the relative noise, the more the image compo-nents will be lost beneath it Since the image contrast decreases at high resolu-tion this means that increasing the noise level will have the effect of decreasing the resolution in the image; coarse, large scale structure remains above the noise, while fine detail is lost In practice, the noise in a confocal image can easily be very much higher than the noise in a wide-field image as is shown below, and so it is entirely possible for the effective resolution to be worse for a confocal image

dis-tribution of the recorded photons In general terms, detector noise is most im-portant at very low photon levels, Poisson noise at higher levels In a typical cooled CCD camera, the measurement noise, mostly readout noise from the A/D converter in the camera, is about 10 electrons (r.m.s.) per pixel This means that if many measurements of exactly the same number of incident electrons were made, the standard deviation (root mean square deviation from the mean value) would be 10 So, if exactly 100 electrons had accumulated in a particular pixel in the array, there would be a probability of about 70% that the measured number would be between 90 and 110 The detective quantum efficiency (i.e the fraction of incident photons which produce a detectable electron in the charge wells— DQE) varies between 20% and 80% depending on the wavelength and the CCD chip design Let us assume 50% This means that the detector noise is equivalent to about 20 incident photons Poisson noise arises because the emission and detection of photons is a random process and fluctuates according to the Poisson distribution If many measurements of a 'uniform' flux of photons were made using a perfect measurement device, then the standard deviation of these measurements would be given by the square root of the mean value So, if the detected number of photons is 100, we can regard this as a sample of a prob-ability distribution whose mean is 100 and whose standard deviation is 10 Thus, at a signal level of 400 photons, the Poisson noise would be 20 photons—the same level as the detector noise At this signal level we could expect an r.m.s noise level of around 7% of the signal (400 photons measured with 20 photons Poisson noise and the equivalent of 20 photons detector noise—since the two sources of noise are independent we can add their contributions as the square root of the sum of their squares) With a higher signal level, the Poisson noise would be greater than the (signal-independent) detector noise

The photomultipliers used in confocal microscopes, if they are kept cool, or at least not allowed to get warm, and if the detection circuitry is optimal, are capable of very low detector noise levels—probably to counts or lower Thus, to obtain the same 7% signal-to-noise ratio as for the CCD camera we would only need to measure about 200 counts (Only Poisson noise will contribute signifi-cantly and V200/200 is about 0.07.) However, the DQE is much lower—perhaps only 10% or less for a typical photomultiplier—so detecting 200 counts would require 2000 incident photons at this sort of efficiency level

a single scan Therefore, to obtain the numbers of photons in the image that a CCD camera records in a single second exposure may require hundreds of suc-cessive confocal scans to be accumulated This is usually not feasible, and thus confocal images generally have a considerably higher noise level than wide-field images recorded by CCD cameras

6.2 When should confocal microscopy be used?

Some of the factors that should be considered in deciding whether confocal microscopy is the most appropriate technique for a particular imaging experi-ment will be briefly discussed

Since confocal microscopes not currently compare favourably with high grade CCD cameras in detection sensitivity and speed of image capture, they are not well suited to imaging very weak or photosensitive specimens It is entirely possible that a confocal microscope will produce an image for a weak specimen that is actually worse than a conventional fluorescence microscope with an attached CCD camera The only real advantage of a confocal microscope is that the out-of-focus light is eliminated Therefore, there is only any real point in using confocal microscopy where this advantage is important; in practice this means for specimens with substantial thickness But it is not always obvious what 'substantial' means Certainly we have obtained good confocal images from specimens several hundred micrometres in thickness in which the in-focus signal was virtually obscured by the out-of-focus light, making the conventional fluor-escence image so poor as to be useless On the other hand it is pointless using a confocal microscope to image thin microtome sections (a fraction of a micro-metre in thickness) Better results will almost certainly be obtained by the better image detection capabilities of a good camera on a conventional microscope

However, most real specimens are somewhere in between these two extremes, and it is not always easy to predict what imaging method will perform best for a given specimen To some extent what is used will depend on what facilities are available If a very good cooled CCD camera, coupled with state-of-the-art decon-volution software is available, and, importantly, if there is access to expertise in using it, then it is likely that better images can be produced in this way in many cases, at least for specimens which are no more than one cell thick There are good examples in the literature of outstanding 3D reconstructions produced in this way, and several integrated imaging and restoration packages are now commercially available

computer and a high quality CCD microscope imaging system, in both micro-scope and computer hardware, and in the software needed for image collection and subsequent analysis and processing It may be hoped that one of the various manufacturers might take the step of integrating both types of imaging into a single instrument

Figure shows an example of the same specimen—a Drosophila embryo labelled with rhodamine-phalloidin to show the actin distribution—imaged by both con-ventional fluorescence using a cooled CCD camera and by confocal fluorescence microscopy Because this is a relatively thick specimen with a great deal of out-of-focus light arising from other planes than the plane of focus, the difference between the two methods of imaging is dramatic

6.3 Objective lenses for confocal imaging

In principle any objective which can be used for conventional fluorescence can also be used for confocal fluorescence imaging As with all fluorescence imaging the brightness of the image is strongly dependent on the numerical aperture (NA) of the objective and so objectives with the highest available NAs should generally be used As discussed above, confocal imaging is more seriously de-graded by aberrations than conventional imaging, and so it is preferable to use high quality plan-apochromat objectives if possible However it should be noted that while all plan-apochromat objectives are suitable for fluorescence imaging with visible light (e.g FITC, rhodamine, Texas Red, Cy3), the objectives of this type made by some manufacturers not transmit the UV light required for excitation of the common DNA dyes like DAPI In fact very few available objectives transmit the commonly used UV laser wavelengths from high power argon ion lasers efficiently (351 and 363 nm) In the author's laboratory oil immersion objectives are nearly always used for confocal imaging, in spite of the spherical aberration and refractive index mismatch problems described above Optically very good results have been obtained from a water immersion, coverglass-free objective (Zeiss, X63,1.2 NA) In fact the lower aberration in use of this objective means that optically better images are obtained with it than with the oil immersion objectives having higher numerical apertures (e.g Nikon X60,1.4 NA, and Leitz X63, 1.4 NA) However, using these coverglass-free object-ives poses several problems: the specimens are not well protected, and are diffi-cult to keep; focusing the objective transfers forces to the specimen which tends to cause it to move around and often to break up; it is difficult to find a satis-factory water-based antifade mountant A water immersion lens with a cover-glass would be a better alternative Several manufacturers are now making these objectives, but at the moment they are very expensive

7 Specimen preparation for confocal imaging

In general terms, confocal fluorescence microscopy simply requires a escent specimen, which can be prepared in a similar way to any other fluor-escently labelled specimen However, the ability of the confocal microscope to produce clean optical sections from thick, three-dimensionally well preserved specimens should not be wasted In practice, this means that some care should be taken to preserve the three-dimensional structure of the specimen as well as possible For living specimens, methods that keep the tissue or cells alive and active will almost certainly preserve three-dimensional structure well For dead, fixed specimens the best fixative that is consistent with the labelling method should be used Formaldehyde solutions (which should be freshly made by dis-solving solid paraformaldehyde) are nearly always used in the author's labora-tory, sometimes with a small percentage of glutaraldehyde Electron microscopy has shown that the bifunctional glutaraldehyde is a better fixative than formal-dehyde, but it often interferes with penetration of antibodies and other probes It also causes a high autofluorescent background, which can be alleviated to some extent by a subsequent treatment with sodium borohydride Fixed tissues often also need extra permeabilization to allow penetration of probes In the cases of plant material, partial digestion of the cell wall with cellulase and other cell wall degrading enzymes is usually needed (see Chapter 10) Producing good specimens usually depends on achieving a balance between preservation of the structures of interest and disrupting them so as to allow probes in to visualize the structures

In general, it is best to leave the physical form of tissues as unaltered as possible In many cases the tissue of interest is within a relatively large organism or organ and must be physically removed or sectioned to make labelling and imaging possible In these cases, a vibratome, which can cut quite thick sections (50-100 um) from living or fixed tissues and plants, is routinely used in the author's laboratory Finally, the specimen should be mounted with care, so that the coverglass does not squash the carefully preserved structure For thin speci-mens the thickness of the coating between the wells on multiwell slides is usually sufficient to protect the specimen from the coverglass For thicker speci-mens it may be necessary to support the coverglass Either use nail varnish or more coverglasses to make a platform to raise the coverglass away from the specimen

is required to interpret the vast amounts of image data that this technique can produce The developments in computer technology mean that the raw com-puting power needed for this is now affordable; what is needed are better software packages to accomplish this analysis, and in particular better human/ machine interfaces

Collection of confocal Images

This is not really a protocol so much as a set of procedures that should be used in optimizing the parameters for collection of confocal images In principle it should apply to any confocal microscope, but the details will differ on different machines The settings described in steps 1-4 all interact with each other, and must often be iterated to produce the best image What constitutes the 'best' image also depends on the imaging experiment being undertaken Ideally, for multiprobe imaging of several different fluorochromes, steps 1-4 should be optimized for each fluorescent probe in turn, although this is not always possible

Method

1 Optimize the laser illumination intensity This depends on the specimen and the fluorescent labelling In general the highest intensity that will not cause problems should be used Problems from too much illumination light include fading of the fluorochrome and photodamage to living cells It should be remembered that the very high laser light levels used can easily cause a complete population inversion of the fluorescent molecules to the excited state; any further excitation will only cause photodamage Often it is better to use a lower illuminating light level, and average the image over a longer time,

2 Optimize the detector pin-hole diameter The smaller the pin-hole, the better the resolution—primarily the better the exclusion of out-of-focus light, but the less light is detected and the poorer the statistical properties of the resulting image It is often better to open up the pin-hole for a weak specimen, and compromise on the confocality of the image

maximum, and again may result in the loss of image information or the intro-duction of artefacts

4 Optimize the image averaging Single image scans usually show poor image statistics—i.e high levels of image noise—simply because of Poisson noise given the low numbers of photons detected in a single scan The solution is to accumulate many scans of the image, or to scan at a slower rate, or both The image signal-to-noise ratio increases as the square root of the number of scans accumulated The disadvantage to accumulating many scans is the time taken and the increased light dosage, causing fluorochrome fading and photodamage In the case of living cells, dynamic changes may occur and be blurred by long accumulation series The requirement for dynamic information and good image signal-to-noise must be balanced against each other

5 Set up the other data collection parameters, such as for focal series (z series), time series, etc., and collect the image data

References

1 Grew, N (16731 An idea of a phytological history propounded Together with a continuation of the anatomy of vegetables, particularly prosecuted upon roots With an account of the vegetation of roots grounded chiefly thereupon Published by Nehemiah Grew

2 Rawlins D J (1992) Light microscopy Bios Scientific Publishers Ltd., Oxford.

3 Lacey, A.J (ed.)(1999) Light microscopy in biology a practical approach 2nd edn IRL Press, Oxford

4 Bradbury, S (1989) An introduction to the light microscope Royal Microscopical Society Handbooks Oxford University Press, Oxford

5 O'Brien, T P and McCully, M E (1981) The study of plant structure Principles and selected methods Termacarphi Ply Ltd Australia

6 Taylor, D L and Wang, Y.-L (1989), fluorescence microscopy of living cells in culture Vol A and B Academic Press, London/New York

7 Pawley.J B (ed.) (1995) Handbook of biological confocal microscopy Plenum Press, New York and London

a RMS dictionary of light microscopy (1989) Royal Microscopical Society Handbooks 15. Oxford University Press, Oxford

9 Shaw P J [1999) Ill Light microscopy in biology: a practical approach (ed A J Lacey) 2nd edn, p 45 IR1 Press, Oxford

10 Stelzer E H K (1995) In Handbook of biological confocul microscopy (ed J B Pawley). Plenum Press, New York and London

11 Petran, M., Hadravsky, M., Egger, M, D, and Galambos, R (1968) J Opl Soc Am., 58, 661

12 Shaw, P J (1995) In Handbook of biological confocal microscopy (ed J R Pawley), p 373, Plenum Press, New York and London,

14 Pawley, J B (ed.) (1995) In Handbook of biological confocal microscopy, p 19 Plenum Press, New York and London

15 Sheppard, C J R., Gan, X., Gu, M, and Roy, M (1995) In Handbook of biological confocal microscopy (ed J B Pawley), p 363 Plenum Press, New York and London.

Chapter 2

Fluorescent probes for living plant cells

Mark Fricker, Andrew Parsons, and Monika Tlalka

Department of Plant Sciences, Oxford University, South Parks Road, Oxford 0X1 3RB, UK

Elison Blancaflor and Simon Gilroy

Penn State University, Biology Department, 208 Mueller Laboratory, University Park, PA 16802-5301, USA

Andreas Meyer

Institute fur Forstbotanik und Baumphysiologie, Am Flughafen 17, D-79085 Frieburg i.Br Germany

Christoph Plieth

Zentrum fur Biochemie und Molekularbiologie, Christian-Albrechts-Universitaet, Leibnizstrasse 11, D-24118 Kiel, Germany

1 Introduction

There are several reviews that cover the biological questions that have been addressed with fluorescent techniques in plants (1-3) and there is extensive on-line documentation on the properties of the fluorescent probes themselves (e.g Molecular Probes, http://www.probes.com) This chapter focuses on optimization of the systems used for live cell imaging, the methods used to mount plant specimens and load them with dyes, a compendium of dyes that have proved useful in plant tissues, and a series of protocols to convert qualitative observa-tions to quantitative measurements The hazards associated with these dyes are, in the main, not known, so all should be treated as potentially harmful

2 Selecting probes with high brightness

wavelength The molar extinction coefficient (E or EC, units: cm-1 M-1) is defined as the absorption of a 1M solution measured through a cm path length, usually at the wavelength that shows the maximum absorption in a particular solvent (Emax) In the case of more hydrophobic probes, the solvent used is often ethanol, DMSO, or DMF, thus literature values of E may not be entirely appropriate to the conditions expected in living cells Some electrons decay by radiationless trans-itions that compete to depopulate the excited state Alternatively, the singlet electron may convert to a triplet state (T1) by intersystem crossing Typically, electrons remain in T1 for an extended period (> 10-6 sec) before decay to S0 ground state (phosphorescence), as this transition also requires the spin on the electron to reverse Occasionally, electrons in T1 receive sufficient thermal energy to excite them back to S1, giving rise to delayed fluorescence, occurring in around 10-6 sec

The probability that light will be emitted provides a measure of the relative extent to which fluorescence versus other competing energy dispersive pro-cesses occur and is termed the quantum efficiency (QE) or quantum yield (QY, O)

of the fluorochrome according to Equation 1:

Some of the best current fluorophores have a QE of 0.7 or higher, however, the QE of many probes is rarely quoted The QE is often dependent on the en-vironment around the probe, meaning that changes in a physiological parameter can be inferred from changes in fluorescence

The brightness of a fluorophore is the product of QE and Emax In most applica-tions high values of both QE and e are advantageous, however, at increasing concentrations, self-absorption (A) by the dye itself can significantly decrease the expected fluorescence yield The extent of absorption for a dye in solution is given by the Beer-Lambert law (Equation 2):

minimize self-absorption Self-absorbance is also greatly reduced in multi-photon imaging, as the dye has essentially no absorption at the primary red or IR wave-length outside of the region where multi-photon excitation can occur

At high dye concentrations the fluorescence emission may also be reduced by molecular collisions that dissipate energy by radiationless transitions before it can be emitted as fluorescence, although the dye concentrations required rarely occur in live cell imaging

2.1 Spectral considerations

The energy levels in both S0 and S1 are spread slightly, so it is possible to excite the molecule with photons with a range of different energies (i.e different wave-lengths) represented by the excitation spectrum This may differ from the absorption spectrum if not all wavelengths absorbed contribute to fluorescence In some cases, the excitation spectrum shows additional peaks at shorter wave-lengths that indicate transition of electrons from S0 to higher excited states (S2 => Sn)

Decay of the electron from the different energy levels in these excited states to S0 gives rise to photons with different energies represented by the emission spectrum The shift between the peak excitation wavelength and the peak emission wavelength is termed the Stokes shift

The QE, and both the excitation and emission spectra are critical consider-ations when selecting the optimum fluorophore for a particular application For example, in multiple labelling experiments, a fluorochrome with a large Stokes shift facilitates separation of the excitation and emission wavelengths and may even allow a single excitation wavelength to be used with two probes that can be separated by their emission spectra Alternatively, a long tail in the emission spectrum gives bleed-through of fluorescence from the probe with the shorter emission, to contaminate the signal from the probe with the longer emission

3 Fluorescence lifetime imaging microscopy (FLIM)

Electrons not decay instantaneously from the excited state but follow a first-order kinetic with a rate constant, kF The reciprocal of the rate constant gives the relaxation time or fluorescence lifetime, TF One lifetime is the time required for the number of excited electrons to decrease to 1/e or 37% of the initial number, and is in the nsec range for many fluorophores (The fluorescence half-time is also often quoted as the half-time for the number to decay to 50% and equals 0.693 x TF) The fluorescence lifetime is variable for different fluorophores and is also sensitive to the local environment around the probe Thus measurement of TF using short, pulsed excitation and time-gated detectors can be used to separate signals from fluorophores with considerable spectral overlap, if they have differ-ent values of TF Alternatively, it is possible to infer physiological information on the cellular environment if the lifetime alters in a predictable way Thus a number of Ca2+ indicators not show spectral shifts upon binding Ca2+ but show changes in TF on ion binding, allowing ratio measurements in the time-resolved domain rather than the spectral domain It is currently rather expensive and rare to combine fluorescence lifetime measurements with imaging micro-scopy (FLIM), so most time-resolved measurements in plant systems have used photometric measurements For example, time-resolved measurements have been used to measure the viscosity of the cytoplasm (4) and plasma membrane (5)

4 Fluorescence polarization anisotropy

Many fluorophores have extended ring structures that can be preferentially excited by light with a matching angle of polarization (photoselection) If the molecule is immobilized, the fluorescence emission retains an amount of the initial polarization If the molecule can move during the lifetime of the excited state the emission is randomized to a greater extent with respect to the ex-citation light (motional depolarization) The degree of polarization (P) is usually measured by comparing the fluorescence intensity with polarizers oriented parallel (I||) and perpendicular (I1 to the direction of the exciting beam:

For this technique to work effectively it is important that the other optical components in the system preserve the polarization angle of the excitation and emitted light

Fluorescence polarization measurements have only rarely been used in plant systems, but can provide information on microviscosity of membrane lipids (5) or the orientation of cellulose microfibrils (6)

5 Fluorescence resonance energy transfer (FRET)

in a process called fluorescence resonance energy transfer or FRET FRET is en-hanced if the two fluorophores are within a few nanometres of each other, but falls off very rapidly at approximately the reciprocal of the sixth power of the separation distance (around 2-6 nm for 50% decrease) Typically, increased FRET is measured as a decrease in donor fluorescence coupled to an increase in accep-tor fluorescence FRET measurements provide an indication of the proximity or binding between two separate fluorescently-tagged molecules at physical dis-tances far below the resolution limit of the light microscope More recently, FRET probes have been developed where the separation of two spectral variants of green fluorescent protein (and hence the extent that FRET occurs) is depend-ent on Ca2+-induced conformational changes in a linker polypeptide composed of calmodulin and a calmodulin-binding peptide (7, 8) Binding of Ca2+ to cal-modulin causes the calcal-modulin to fold and associate with the CaM-binding peptide This reduces the separation of the CFP and YFP fluorophores, increases FRET, and gives rise to a shift in the emission spectrum towards the YFP peak This important development heralds a new era of ratioable molecular fluor-escent reporters whose ligand specificity, expression level, tissue and subcellular targeting are all encoded in gene constructs and can be readily manipulated using molecular biology techniques

6 Photobleaching and fluorescence redistribution after photobleaching (FRAP)

In principle, increasing the excitation intensity gives rise to a stronger fluor-escence signal and better signal-to-noise (S/N) ratio, however, increased excitation has major drawbacks in live cell imaging The fluorescence process is cyclical allowing each fluorochrome to emit many photons, however, the small but finite amount of time spent in the excited state means the process can be satur-ated at sufficiently high illumination intensities Further increasing the level of excitation does not provide any increase in fluorescence A second consequence of excess excitation is the increased frequency of inter-system crossing to T1 The excited electron tends to remain in T1 for a much longer period of time com-pared to S1, which substantially increases the likelihood that it will interact with another species, such as molecular oxygen, to form damaging free radicals capable of destroying the fluorophore (photobleaching) or damaging the tissue (phototoxicity)

Fluorophores have markedly different tendencies to photobleach and this may be another important consideration governing the choice of fluorochrome Photobleaching can be reduced in fixed preparations by antifade reagents or by reducing oxygen levels, but these tricks are rarely possible with live cell prep-arations unless anoxia is acceptable

coefficients of membrane lipids (9), ER-mediated intercellular communication (10), rates of phloem unloading (11), and the dynamics of cytoskeletal rearrange-ments following microinjection of fluorescently-tagged tubulin or actin mono-mers (2, 12) In a variant of the FRAP technique, continued bleaching of a specific area can be used to probe connectivity with adjacent regions in, for example, the ER network or stromules between chloroplasts In addition to the loss of signal from the bleached area, molecules diffusing from the connected regions are also bleached in a process termed fluorescence loss in photobleaching or FLIP (13)

7 Optimization of fluorescent systems for live cell imaging

Live cell imaging is particularly demanding as both the amount of fluorochrome that can be used and the intensity of the excitation are heavily constrained by the need to minimize phototoxicity and keep the cells alive and functioning normally This means that the level of fluorescence signal will be at the limits of detection by the human eye The noise ratio (S/N) and the signal-to-background ratio (S/B) become of increasing importance in such studies, particu-larly for quantitative measurements With this in mind, it is worth optimizing each component in the optical system for each application The basic com-ponents of different fluorescent systems have been described in Chapter 1: the brief comments below focus on the additional requirements of a live cell imaging system

(a) Lasers provide well-collimated beams at defined wavelengths, but are limited in the range of the spectrum that can be conveniently covered in terms of cost or the technical difficulties in combining several different lasers on the same instrument

(b) Mercury arc lamps have spectral lines at 366, 405, 436, 546, and 578 nm with substantial intervening regions of relatively lower, but even intensity DC operation can be used for some lamps to increase the stability of the output

(c) Xenon arc lamps are often preferred choices for multi-wavelength systems as they have an almost flat spectrum from UV to red allowing the user to select the most appropriate wavelength Lamps need to be burnt in under con-ditions of high mechanical stability to prevent the arc wandering

(d) Quartz-halogen lamps are cheap sources of visible wavelengths, but cannot be used for applications requiring UV

(e) Super-bright light emitting diodes (LEDs) give out sufficient power to excite fluorescence and are now available at selected wavelengths from blue to IR These may well provide a very low-cost and versatile illumination system

illumin-ation controlled by a shutter system or a blank filter position are often needed to minimize photobleaching and phototoxicity

7.1 Selection of the excitation wavelength

Single-line lasers not require an excitation filter, however, for broad-band light sources or multi-line lasers, the excitation wavelength(s) is usually selected by an interference band-pass filter, often mounted in a computer controlled filter-wheel More recently, very flexible spectral excitation sources have been developed that use either monochromators with a galvonometer-driven holo-graphic grating, or an acousto-optic tuneable filter (AOTF)

Interference filters are characterized by a centre wavelength (CWL) and a bandwidth measured as the full width at half-maximum (FWHM) or half band width (HBW) of the peak The number of cavities in the interference coatings affects the slope of the transition from attenuation (defined as 5% of peak trans-mission) to transmission (defined as 80% of peak transtrans-mission)

Judicious choice of the centre wavelength, bandwidth, and slope can signifi-cantly improve the ratio of fluorescence signal-to-background or specimen autofluorescence A wide range of filters is available off-the-shelf and most manu-facturers offer facilities for customer-specified filters (e.g http://www.chroma.com or http://www.omegafilters.com)

7.2 The dichroic mirror

In epi-illumination microscope systems, the excitation beam is directed towards the sample using a dichroic mirror that reflects the excitation beam but trans-mits the longer wavelength emission Typically, the dichroic mirror is specified as a long pass filter with the cut-on wavelength at 50% of the peak transmission when oriented at 45°, the angle used in the microscope A sharp transition from reflection to transmission is essential for many fluorochromes that not have large Stokes shift, to maximize the amount of the emission peak that can be collected In multiple-labelling experiments, double- or triple-dichroics allow simultaneous excitation and collection of signals from different probes with no mis-registration

7.3 Selection of the emission wavelength

7.4 Choice of measurement system

Several different techniques are available to visualize or quantify fluorescent signals The main task is to match the measurement system with the biological question being addressed

(a) Fluorimetry is useful for populations of (single) cells in suspension Sampling is rapid (interval ca 0.05 sec or better) and spectra are easy to measure Auto-fluorescence is easy to correct in a parallel sample or prior to dye loading There is no spatial resolution and heterogeneous responses from different cells cannot be distinguished Signals from dead and dying cells also included

(b) Flow cytometry is also appropriate for populations of single cells or (robust) protoplasts in suspension Sampling is rapid, but of a different cell for each data point Many systems allow measurement of multiple parameters simul-taneously and there is potential for preparative sorting of cells by their re-sponse There is no subcellular spatial resolution other than through targeting of the probe and heterogeneous responses in the population appear as an increase in variance (see Chapter 3)

(c) Micro-photometry is typically used for microscope-based measurements on entire single cells or occasionally subcellular regions in large cells An average spatial measurement is usually defined by a (variable) mechanical aperture and the specimen or aperture may be moved to sample different regions or different cells Sampling is rapid (interval 0.05 sec or better minimum, typically sec in practice) An average autofluorescence correction from comparable unloaded cells is straightforward More sophisticated systems can be coupled to a spectral analysis system Photometry measurements are prone to errors from heterogeneous dye distribution (or redistribution)

(d) Camera imaging is useful to map subcellular spatial heterogeneity and/or variation in populations of cells and provides visual cues on the morphology and well-being of the cells Subcellular regions can be measured typically down to 0.3-0.4 um in (x, y), however (z) is poorly defined and can give rise to quantitation errors The best systems use cooled, back-thinned CCD detectors with a fast digital (12-bit) readout Dual-excitation or dual-emission involves sequential switching of the excitation or emission wavelength This gives a delay between the two wavelengths and can cause artefacts in ratioing applications if the specimen grows, moves, alters shape, or the cytoplasm is streaming Simultaneous dual-emission imaging is possible with either a high-quality colour camera, split-view optics onto a single camera faceplate, or two monochrome cameras with appropriate coupling optics The sampling interval is typically every 1-2 sec—more expensive systems can run at video rate, but in practice an extended integration period is often required to in-crease the S/N ratio Autofluorescence subtraction is difficult as the autofluor-escence may also have structure within the image and cannot, therefore, be subtracted as a single value

of cells, or cells in intact tissue when the out-of-focus blur significantly re-duces the image contrast and interferes with quantitative measurements The (x, y) and (z) resolution are relatively well defined (best around 0.2 X 0.2 x 0.6 um, typically 0.4 X 0.4 x 1.2 um) The fastest temporal resolution is dependent on the instrument and the volume sampled, but range from milliseconds for a line scan, seconds for 2D section, and seconds to minutes for 3D data stack Sophisticated sampling is also possible, for example to photobleach user-defined areas in FRAP experiments

(f) Multi-photon microscopy also achieves optical sectioning However, instead of using the absorption of a single photon at a short wavelength to excite an electron from S0, it is possible to combine the energy of two or even three red or IR photons to achieve the same electronic transition if the excitation photons arrive within the order of 10-12 to 10-15 sec Pulsed femto- or pico-second lasers give a flux density that is sufficiently high at the focal point for two-photon excitation to take place, providing optical sectioning without the use of the pin-hole required for confocal imaging and reduces photobleach-ing UV fluorophores can also be excited with red or IR light, in the range 660-1047 nm, that generally penetrates further into intact specimens and is likely to less damage to the tissues, provided that the specimen shows negligible absorption at the primary wavelength Typically the peak for the two-photon excitation spectrum is broader and blue-shifted in comparison to the equivalent single-photon excitation spectrum (14)

8 Securing the specimen for microscopy

The specimen needs to be securely fixed down to prevent movement during microscope observation, perfusion, and especially if microinjection is to be attempted (see Chapter 6) The procedure to immobilize roots in Phytagel is given in Protocol (15) Other techniques include:

(a) Immobilization in agarose Embed tissues in 1-2% (w/v) low melting point agarose (gel point 26-30°C; Sigma type VII) warmed to 40°C; or mix equal volumes of cell/protoplast suspension with 1-2% agarose at 40°C on a pre-warmed coverslip

(b) Immobilization in gelatin Concentrations of gelatin up to 18% (w/v) in nutrient media are useful for embedding single cells Samples are mixed with molten gelatin warmed to 40°C and spread thinly on a pre-warmed coverslip (c) Immobilization in alginate Alginic acid (~ 1.5%, w/v) forms a gel at room temperature in the presence of excess (mM) CaCl2 that can be used to trap protoplasts or cells The high Ca2+ concentrations may perturb the cell physiology

(e) Contact adhesives (e.g Dow Corning No 355, Reading, UK; Secure B401, Factor II, Lakeside, AZ, USA)

(f) Adhesive tape (e.g Cellux) can be fixed to the microscope slide with the sticky side up using double-sided tape For microinjection, the tissue can be injected dry or covered with an inert oil such as Voltalef PCTFE Oil (Atochcm, Pierre-Benite, France) The oil prevents evaporation and improves the optics of the system for observation (16),

(g) Coating the coverslip with poly-L (or D)-lysine Poly-lysine can be applied at 0.01-0.1% (w/v) in 10 mM Tris-HCl pH 8.0 tor 5-60 min, followed by wash-ing (see Chapter 10) Alternatively slides pre-coated with poly-L-lysine arc available (Sigma)

(h) Suction pipettes with 10-20 um diameter tips can be used to hold proto-plasts or single cells These may be filled with inert solutions (e.g siliconc fluid Dow Coming 200/100 CS)

(i) Mechanical clips or restraints can be used to secure large cells or tissues [e.g ref 17)

(j) Rosettes of Ambidopsis plants can be mounted with the stem in water-filled silicon tubing inserted into a small (15 X 30 X 15 mm) box on the microscope stage (18)

(k) Epidermal fragments can be trapped under folding 100/100 mesh (100 lines per inch) EM grids (Ted Pella Inc., Redding, CA) and held down by vacuum grease (19)

It is critical to ensure that the immobilization protocol does not markedly affect cell or tissue responses For example, poly-lysine may induce KT-channel activity (20), and hot agarose or gelatin may heat shock protoplasts before it cools enough to form a gel All embedding and immobilization procedures are likely to reduce diffusion to and from the tissue as well as rapid exchange of media, although this is generally less of a problem with mechanical restraints

Growth of Arabidopsis thallana seedlings in Phytagel for in situ observation of roots

Equipment and reagents

• Growth cabinet or equivalent suitable for

Arabidopsis

• Nutrient medium: mM KNO3, mM Ca(NO3)2+4H20, 0.5 mM MgSO4.7H20,1 mM (NH4)2PO4,1 mg/ml thiamine, 0.5 mg/ml pyridoxine-HCl, 0.5 mg/ml nicotinic acid

0.56 mM myo-inositol, 2.3 mM Mes, 0.1 g/litre sucrose, 25 uM KCl, 17.5 uM H3BO3, uM MnSO4.H20,1 uM

Method

1 Autoclave the media supplemented with 1% Phytagel (Gellan gum agar substitute, Sigma) for 25 at 121 °C and pour ml to mm depth onto autoclaved coverslips (48 x 65 mm) contained in sterile 90 mm Petri dishes The gel should polymerize at room temperature within 10

2 Surface sterilize seeds of Arabidopsis in 95% ethanol (5 min), followed by 10% sodium hypochlorite (5 min), and five washes in sterile water

3 Plant seeds by pushing through the gel onto the surface of the coverslip and chill for

24 h at 4oC

4 Germinate Arabidopsis seeds for four days under continuous light (36 umol m-2 s-1) at 22-24°C with the coverslip at an angle of 45° to promote growth of the root down to and then along the surface of the coverslip

5 Prior to dye loading, ensure the gel matrix is fully hydrated by adding excess nutrient media for 15

9 Perfusion systems

The best clarity or brightness of image for live cell imaging is achieved with high numerical aperture (NA) water immersion microscope objectives and immersion of the specimen in aqueous buffer Submersion reduces light scattering from highly reflective surfaces in the sample and may assist in efficient dye loading and application of many stimuli and calibration solutions Perfusion is required to prevent anoxia developing and allow addition of test compounds or calibration solutions As a guide, rapidly respiring cells, such as guard cells, become anoxic within 10-30 minutes without perfusion (21) The composition of the bathing medium should ideally mimic the environment around the cells in vivo, particu-larly with respect to ionic composition, water potential, and gaseous environment

In the simplest case solutions can be exchanged by drawing excess solution under the coverslip or through the chamber by capillary action onto an absorb-ent tissue, A more typical perfusion system comprises;

(a) A Perspex chamber with a coverslip held onto the base with glue or grease mounted on the stage of an inverted microscope Alternatively, layers of elec-trical insulation tape stuck onto the coverslip provide a rapid and convenient means to cut out chambers of varying geometry

(b) A gravity feed from 50 ml plastic syringes coupled together with manually operated stopcocks (Sigma) The flow rate can be adjusted by varying the height of the syringes More reproducible control of flow rate can be achieved with a constant head apparatus or peristaltic pumps,

(d) The inlet and outlet tubes to the chamber can be constructed from cut-off stainless steel syringe needles positioned using coarse micromanipulators or magnetic stage clamps

(e) The outlet is connected to a vacuum-assisted sipper with a water trap (f) It is important that the rate of suction is high and the diameter of the outlet

syringe small to minimize oscillations in the flow The noise of the sipper also provides an auditory cue that the perfusion is functioning even if the experiment is in darkness

(g) Upright microscope systems with water immersion objectives and focusing of the objective rather than the stage, can also be used for thick, opaque speci-mens such as roots or leaves

Partially closed perfusion systems can be set up with a rectangular chamber partially covered by a small square coverslip to leave open wells at either end This gives a faster flow across the specimen with fewer unstirred regions and better transmission optics, however, the positioning of the inlet and outlet tubes is more critical to maintain a smooth flow through the chamber Completely sealed chambers are more difficult to set up and are most appropriate for cells or tissues that are being cultured under sterile conditions

10 Loading strategies for plant cells

The objective is to introduce the probe to give good signal-to-noise without causing toxic effects or significantly disturbing the cell physiology In principle, membrane permeant dyes can be loaded directly in solution whilst membrane impermeant dyes require additional technique to pass this barrier In practice, many plant tissues load poorly due to the presence of a cuticle or suberized cell walls that restrict diffusion of the dye to the cells, or through binding of the dye in the cell wall A variety of strategies have emerged, but there are no simple rules as to which will be most effective with a particular tissue Population load-ing techniques are considered here, whilst microinjection techniques are dealt with in Chapter

10.1 Extracellular and permeant intracellular dyes

Loading dyes by vacuum infiltration of leaf pieces

Equipment

• Vacuum desiccator and water pump

Method

1 Cut leaf into small pieces between 4-25 mm2, depending on the size and access-ibility of the intercellular air spaces within the leaf, or punch out discs with a cork borer,

2 Place tissue pieces in the dye solution at the working concentration in a suitable container (e.g Eppendorf tube) and place in a desiccator attached to a water pump Apply vacuum for 2-5 Alternatively, place tissue pieces in a syringe containing liquid, evacuate with the plunger, and ensure the tissue is immersed on release of the vacuum

3 Repeated infiltration, cycles may be necessary for some tissues."

4 Pieces that have been infiltrated will become more transparent and sink down Successful infiltration can also be checked by co-infiltration with a cell impermeant

dye, such as propidium iodide, which stays in the apoplast and the intercellular air spaces unless cells have been damaged

a It should be noted that vacuum infiltration degasses the solutions and is likely to give rise to anoxic conditions

10.2 Ester loading

Most of the ion-selective dyes and a number of other probes arc membrane impermeant due to one or more charged curboxy] groups Esterification of the carboxyl groups in the molecule with acetate or acetoxymethyl (AM) groups masks their charge and venders the dye membrane permeant Hydrolysis by intracellular esterases releases the free dye in an active form in the cytoplasm (Protocol 3)

Loading dyes as acetoxymethyl ester or acetate ester derivatives

Reagents

• Ester form of the dye * DMSO

Method

2 Dilute the indicator stock solution to 1-20 uM with deionized water or media immediately prior to use.a Keep unused dye on ice in the dark

3 Incubate cells for 10-120 minb,c,d at room temperature.c

4 Wash out excess dye prior to observation

aAddition of 0.01-0.2% (w/v) Pluronic F-127 may maintain dye solubility and aid tissue penetration

b Varying the external pH in the range pH 5-8 may facilitate cytoplasmic loading of some probes

cPre-incubation for 0.5-1 h in 0.1% B-escin (saponin) (22) or 0,1% digitonin (23) may improve loading

d Pre-incubation for 0.5-1 h with esterase inhibitors, such as 0.1 mM eserine (Sigma), may also help prevent external hydrolysis of the dye and improve loading (22), although care is required as esterase inhibitors can be highly toxic to humans,

e Changing the temperature to either 4°C or raising it to 30°C (22) may improve loading

Although widely used, there are numerous potential problems with the ester loading technique, including:

(a) Labelling of any compartment that has appropriate esterases (e.g mitochon-dria and vacuoles)

(b) Release of acetic acid and formaldehyde within the cell following hydrolysis of AM esters

(c) Incomplete hydrolysis releasing partially activated fluorescent intermediates with different spectral properties

(d) Sequestration in the vacuole or other compartments after release in the cytoplasm

10.3 LowpH loading

Low pH loading of root tissues

Reagents

• Salt form of the dye • KOH

• DGMA; 25 mM dimcthyglutaric acid pH 4.5

Method

1 Make a mM stock solution of the salt form of the dye in deionized water Free-acid forms of the dye have to be titrated to - pH 7,0 with KOH to make them soluble Aliquot into small volumes and store at - 20°C in darkness

2 Dilute the stock solution prior to use to a final dye concentration of 20-50 uM in 25 mM DMGA pH 4.5,

3 For tissues or cells embedded in a gel, hydrate the gel matrix by adding excess nutrient media for 15 prior to adding the dye

4 Add the dye and incubate tissue in the dark for 1-2 ha,b,c at room temperature

5 Wash out unloaded dye by rinsing in fresh media.d,e

aIncreased temperature (30°C) and the presence of 0.1% B-escin (saponin) (22) or Pluronic F-127 (0.01-0.2%, w/v) may assist dye loading in some tissues

b If dye penetration appears to be a problem, access of the dye to the tissue can be facilitated by a cutinase pre-treatment (see Protocol 5),

c When loading some tissues, such as stomatal guard cells, incubation in low light may be required to maintain the physiological state of the cells (19)

d Washing out the unloaded dye is difficult for plant roots supported in gels and may result in high background fluorescence Carefully removing roots from the gel matrix or allowing the roots to grow into a gel-free zone prior to intubation in the dye reduces this problem,

e In some intact cells the dye appears to stick in the wall, either through co-ordination with other charged groups in the apoplast or possibly through precipitation in localized regions of low pH Charge masking with high levels of other ions might reduce this problem

10.4 Cutinase pre-treatment and low pH loading

Increasing the permeability of the cuticle by cutinase digestion

Reagents

• Purified cutinasea (1 mg/ml equivalent to 100-300 mmol/min/mg protein) stored in 10 mM Tris-HCl pH 7.6 with HC1 at 4°C

Method

1 Incubate tissues in a small drop of cutinase at cutinase activities ranging from 0.1-10 mmol/min/mg protein for 5-30 For larger tissues, such as an intact leaf, small drops of cutinase can be placed on the leaf surface

2 Wash tissues in water or buffer (2 x 10-30 min) 3 Load dye using the low pH method (see Protocol 4).

a Cutinase can be prepared according to Coleman et al (24),

10.5 Electroporation

Pores of variable size can be selectively induced in the plasma membrane of proto-plasts by short, high voltage pulses Resealing is spontaneous, but can be slowed sufficiently at low temperature to allow diffusion of dye or other macro molecules into the cytoplasm A cocktail of low molecular weight factors is normally included to replace cytoplasmic components diffusing out of the permeabil-izcd protoplasts The precise conditions for successful and reversible electro-permeabilization of the plasma membrane, such as field strength (0.1-5 kV/cm), number of pulses applied (one to five), and capacitor used (1-50 ul) require careful optimization Aim for roughly 60-80% permeabilization efficiency and > 80% viability after resealing Many cells not survive and the remainder are loaded with variable concentrations of dye In addition, protoplasts should be analysed with independent cellular assays of function to ensure- that protoplasting and electroporation have not altered the cellular response Details for electroporation are given in Chapter

10.6 Loading via detergent permeabilization

Loading of dyes into large tissues, such as somatic embryos, can be facilitated by treating with a low concentration of detergents such as 0.1% (v/v) B-escin (saponin) (22) or digitonin (23) to partially permeabilize the plasma membrane The detergent is then washed from the sample to allow the membranes to reseal

10.7 Loading tissues with phloem-mobile probes

wound in the leaf as acetate or AM esters at very high concentration (Protocol 6), The dyes are loaded into the phloem and the ester groups cleaved to give free dye that is translocated to the roots A number of dyes can be successfully loaded including BCECF-AM, carboxy SNARF-1 DA, CFDA, calcein-AM, carboxy SNAFL-1 DA, fluorcscein diacetate (FDA), HPTS-acetate, and sulfofluorescein diacetate

The pattern of symplastic dye unloading from the phloem and its subsequent distribution and sequestration can be followed using confocal optical sectioning (11, 18, 25) In the case of HPTS-acetate, the method of uptake into the phloem is unclear as the acetate group does not render the dye uncharged, however, HPTS currently provides the best symplastic tracer following unloading as it appears to be sequestered in root vacuoles far more slowly than the other probes It should be noted that HPTS (pyranine), along with several of the other probes, has a strong pH sensitivity (pKa - 7.3) and is also used as a cytoplasmic pH indicator (26, 27) Dye accumulated in acidic vacuoles is less fluorescent, and, unless the isoexcitation wavelength at - 425 nm is used, the fluorescence intensity reflects both the dye distribution and the pH of the compartment A variety of other dyes have also been tested that not load via the phloem, including the Ca21 and pH indicating dyes Calcium Orange AM, Calcium Green-1 AM, chloromethyl SNARF-1 acetate, Fluo-3 AM, and Rhod-2 AM (25) The probability of loading dyes into the phloem can be predicted from their physicochemical properties, including molecular or ionic weight, log octanol-water partition coefficient (log P), con-jugated bond number (CBN), and pKa, using a structure-activity relationship

(SAR) model (25),

The ability to load dyes into the phloem can also be used to observe phloem transport directly (28), although this requires removal of overlying tissues to permit access for microscopy (Protocol 7).

Loading root tissuesa with phloem-mobile probes

Reagents

• Agarose medium: full strength Murashige and Skoog (MS) salt mixture, 3% sucrose (pH 5.8), and 1% agarose

Method

1 Surface sterilize seeds of Arabidopsis thaliana (ecotype 'Columbia') in 5% sodium hypochlorite and rinse in water

2 Plate seeds on near-vertical 1% agarose plates.

3 Chili for 24 h at 4°C

4 Germinate for 4-15 days under continuous light,

6 Place moist filter paper around seedling to maintain a high humidity

7 Apply dyes at 100-500 uM from stock solutions at 10 mM in DMSO (BCECF-AM, carboxy SNARF-1 DA, CFDA, calcein-AM, carboxy SNAFL-1 DA, and FDA) or 8-10 mM from stock solutions in water (HPTS-acetate and sulfofluorescein diacetate) either by:

(a) Crimping a leaf with forceps to make a small wound and loading 0.5-1 ul probe (b) Cutting the tip from the cotyledon and continuously loading dye the cut edge via

a microcapillary positioned using 'blu-tack (bostik)1 for support

8 Cover Petri dish and illuminate with white light (370 umol m-2 s-1 for - 30 min)

9 Observe unloading of the dye from the protophloem in the tip of the primary root after 0.5-3 h using long working distance (X10 or x20) objectives

10 Dye within selected regions of the root tip can be bleached using a 100 mW laser

and 10-20 sec scanning at high zoom to asses the rate of dye movement from recovery of fluorescence after photoblcaching (FRAP)

aThis technique has been used to label symplastic fields in the developing Arabidopsis shoot apex following loading of HPTS (2.5 mg/ml} through silicon tubing (1 mm inner diameter) attached to the cut petiole of older leaves (18)

Direct observation of phloem transport in bean leaves

Reagents

• Bathing medium: 10 mM KC1, 10 mM CaCl2, and mM NaCl (unbuffered)

• Loading buffer: l0mM KCl, l0 mM CaCl2, and mM NaCl pH 6.3 with HC1

Method

1 Make two shallow, paradermal cuts in the major vein of a mature leaf from a 17-21 day-old Vicia faba plant using a new razor blade Cuts should be approx 10 mm long. by mm wide, and separated by about cm

2 Immediately add unbuffered bathing medium to the cut surfaces

3 Mount the leaf (still attached to plant) upside down on the stage of an upright confocal microscope using double-sided tape on a convex surface

4 Add ul RH-160 (Molecular Probes) from a 25 mg/ml stock in EtOH to the basal well to stain the plasma membrane and visualize the cell morphology

5 Observe the intact phloem in the basal cortical window using a X63, 1.2 NA water immersion lens with a working distance — 220 um (ex 567 nm, em > 590 nm). BIot the apical window and apply HPTS-acetate (20 mg/ml), CFDA (0.5 mg/ml), or

7 Allow the dye to translocate for 30 before observation (ex 488 or 476 nm, em >510 nm)

a Dyes were initially dissolved in a small volume of 500 mM KOH

11 Intracellular dye concentration, viability, and toxicity

Once successfully loaded into the plant cell the fluorescent dye may interfere with the normal function of the cells as the concentration is increased In addition, the interaction of illumination with the fluorescent dye may cause phototoxic damage, particularly if the excited dye reacts with oxygen to give highly reactive free radicals Thus, the concentration of fluorochromc intro-duced should be kept low to minimize buffering effects and reduce any potential non-specific chemical or photochemical side-effects

It is important to have accessible markers of cell function to compare in loaded and unloaded cells, such as cytoplasmic granularity, organelle morphology, cytoplasmic streaming, membrane potential, rate of cell division or elongation, response rate, or level of gene expression Alternative strategies include the monitoring of cell viability using other vital or mortal staining techniques, such as fluorcscein diacetate (see Section 12.1) or propidium iodide (see Section 12.2), respectively, as these probe different aspects of membrane integrity and metabolic activity

The best approach is to determine the upper limit of dye loading consistent with minimal disruption of cell physiology Typically for ion indicators this is be-tween 3-50 uM intracellular dye Once the cells are loaded with the appropriate dye, the image collection protocol can be optimized Keeping cells alive may be at odds with maximal spatial or temporal sampling and compromises have to be made to balance the spatial resolution, temporal resolution, and spectral resolu-tion It is important to optimize the instrument to maximize the signal-to-noise ratio (S/N) and minimize phototoxiciry under these conditions It is also import-ant to minimize the light exposure to the sample, hence even when finding the cells to study this should be done as fast as possible The amount of excitation illumination presented at the sample is probably the most critical parameter, however, it is not easy to predict the appropriate intensities for each systems and specimen It is useful to be able to measure the illumination intensity in the specimen plane when the imaging conditions have been optimized to act as a guide for other experiments Typically four sets of controls should be run:

(b) Sample plus illumination: to test the biological effects of the excitation illumination and to measure the levels of autofluorescence

(c) Sample plus dye (but without fluorescence excitation): to test the effects of dye loading on physiological function

(d) Sample plus dye plus illumination: to test the potential phototoxic effects of illumination levels and dye concentrations

12 Selection and use of fluorescent probes

Two main applications of live cell imaging have emerged: first to follow the morphology and dynamics of different cell compartments and organelles, and secondly to infer physiological information from quantitative analysis of fluor-escence intensity, wavelength shifts, polarization angle, or lifetime measurements In qualitative measurements, the emphasis is on generating an image of sufficient brightness and contrast to visualize the structures of interest Quantitative measurements are significantly more demanding if they are to be reliable A number of criteria can act as a guide to select appropriate dyes for plant cells:

(a) The excitation/emission wavelengths in relation to the spectral sensitivity of the tissue (i.e will illumination of the dye also trigger a phytochrome or blue light response)

(b) The level of autofluorescence of the tissue at the measurement wavelengths (c) The instrument configuration for multiple-labelling or multiple-excitation or

emission dyes

(d) The ease of loading the dye into a defined compartment

(e) The behaviour of the dye within the cell, including unwanted compartment-alization, metabolism, and physiological perturbation

(f) Compatibility with other optical techniques, such as UV photolysis of caged compounds

Additional criteria become important for dyes used to measure ion activities in plant cells including the Kd of the dye (i.e how close is the dissociation con-stant of the dye to the ion level in the cell compartment to be monitored), and the possibility of interference by changes in other ions in the cell of interest Dyes with spectral shifts are preferable for ion measurements as they permit ratio measurements that distinguish fluorescence changes due to ion binding from those due to dye leakage, bleaching, or uneven distribution

is used, however, it must be borne in mind that only a limited number of species can be readily transformed There is also a considerable overhead in time and resources required to construct and express each GFP-probe and there are currently very few GFP-probes for physiological studies A description of the methods and full details of GFP constructs are given in Chapter The following sections provide a guide to chemical probes that have proved useful for labelling particular organelles in live cells, followed by probes that have been used to assess particular aspects of cell physiology such as Ca2+ signalling or pH regulation

12.1 Vital stains

Most of the intracellular probes described in the following sections can be used as indicators that one or more aspects of cell function are still operational In the case of the most commonly used vital dyes, such as fluorescein diacetate (FDA), fluorescence requires the action of intracellular esterases to cleave off the ester groups and an intact plasma membrane to ensure retention of the dye (30) De-tails of common classes of probe are given in Table Esterified dyes are typically

Table Representative classes of fluorescenta probes for vital staining

Dye

Fluorescein diacetate (FDA)

5-(and-6)-carboxyfluorescein diacetate (CFDA) 5-(and-6)-carboxy-2',7'-dichloro-fluorescein diacetate (carboxy-DCFDA) Calcein-AM Chloromethyl fluorescein diacetate (CMFDA, CellTracker Green™)

CellTracker Green BODIPY (8-chloromethyl-4,4-difluoro- l,3,5,7-tetramethyl-4-bora-3a,4a-diaza-s-indacene)

Neutral Red (NR)

Ex nm 490 492 495 494 492 522 541 Em nm 514 517 529 517 517 529 640

E 10-3

88 75 32 76 78 72 39

Mode of action

Cleavage of the acetate groups by intracellular esterases releases free fluorescein that is fluorescent Similar to FDA except the additional carboxy group gives better cellular retention

Lower pKa (~ 4.8) means carboxy-DCF

is less pH sensitive than fluorescein at cytoplasmic pH values

Essentially pH-insensitive fluorescein derivative Fluorescence is quenched by some heavy metals at uM concentrations

The Chloromethyl group makes this a substrate for GST catalysed The conjugation to GSH conjugate may then be a substrate for transport into the vacuole by glutathione conjugate (GS-X) pumps Although the Chloromethyl group functions in a similar manner to that in CMFDA, CellTracker Green BODIPY has no ester groups that require cleavage to give fluorescence

NR is a weak amine that accumulates in acidic compartments such as vacuoles Spectra are pH-dependent with a pKa ~ 6.7

made up in dry DMSO at 1-10 mM and diluted to 1-25 uM for use (see Protocol 3). Although the initial and major site of release of fluorescein from FDA may be the cytoplasm, any cell compartment that has appropriate esterases can be labelled, furthermore, fluorescein may be transported into vacuoles or leak back out across the plasma membrane in its uncharged (protonated) form Modified forms of FDA have additional carboxyl or halogen groups to increase the retention of the probe within the cell once the acetate groups have been cleaved, although the anion forms are still effectively sequestered in the vacuole of many cell types (31) One of the better fluorescein derivatives used in animal systems is calcein-AM, as the cleavage product is both highly charged and relatively pH-insensitive around pH The Bodipy, Alexa, and Oregon Green dyes are also relatively pH-insensitive and are available as ester derivatives suitable for use as vital dyes, but have not yet been widely tested in plant systems

An alternative approach to aid cellular retention is to include a chloromethyl group that makes the released dye a target for conjugation to glutathione (GSH), if an appropriate glutathione S-transferase (GST) is present (see Section 13.5) The glutathione-dye conjugate may remain cytoplasmic or itself be a substrate for vacuolar glutathione S-conjugate (GS-X) pumps

Fluorescein derivatives have a strong pH-dependent shift in their excitation spectrum This can be exploited to load pH indicator dyes with a range of differ-ent pKa values into particular compartments (see Section 13.2) If pH-dependent variation in fluorescence adds an unwanted variable into the analysis, probes with little or no pH-dependence, such as CellTracker Green BODIPY can be used Neutral Red is a weakly basic dye that has been used extensively as a vital probe in bright field microscopy as it readily accumulates in acidic compart-ments, usually the large central vacuole, of living cells (32), staining them red at acid pH (shifting to yellow at pH 6.8-8) It is also weakly fluorescent and can be imaged with excitation around 540 nm and emission at 640 nm

12.2 Mortal stains

PI is usually used at 5-10 ug/ml, although higher concentrations (30 ug/ml) tend to be used when it is excited away from its excitation maximum It is loaded into the apoplast of intact tissues along with other dyes of interest Root tissues load readily as far as the Casparian strip, shoot and leaf tissues load more slowly and typically require vacuum infiltration (Protocol 2) PI does penetrate intact cells slowly, so it is preferable to keep incubation times relatively short (10-20 min) Other DNA dyes, such as YO-PRO-1 (1 uM) may also prove useful as cell im-permeant DNA stain that labels nuclei in dead cells (33)

12.3 Cell permeant nuclear stains

DAPI and Hoechst dyes are excellent cell permeable UV-excited dyes that label DNA in living cells (see Table 2) Confocal imaging applications usually require dyes that can be excited at longer wavelengths, such as Acridine Orange (AO) or the SYTO dyes (Molecular Probes) So far the green SYTO dyes have been used successfully in plants (31, 33), however, the SYTO series encompasses a broad wavelength range, suitable for multiple-labelling applications SYTO dyes not exclusively label DNA and give some background cytoplasmic and mitochondrial staining DNA stains are potential mutagens and it is recommended that double gloves be worn when handling stock solutions Before disposal, dye solutions should be poured through activated charcoal and the charcoal incinerated

12.4 Chloroplasts

Chlorophyll shows a strong autofluorescence in vitro or in vivo Chlorophyll dissolved in ether can be excited at either of its major absorption peaks in the

Table Cell permeant nuclear stains used in plant systems

Dye Acridine Orange DAPI Hoechst 33342 SYTO 11 SYTO 16 Stock solution

1 mg/ml in water, dark

0.1 mg/ml in water, 4°C, dark

1 mg/ml in water, 4°C, dark

5 mM in DMSO -20°C, dark

1 mM in DMSO, -20°C, dark

Loading Ex nm conditions

1 ug/ml for 500 10-30

0.1-1 ug/ml 358 for 10

0.5-5 ug/ml 352 for 10

1-5 uM for 508 h for

protoplasts 488 Em nm 526 461 461 527 518

E 10-3

53 21 45 75 42 Comments

Labels double-stranded DNA green and single-stranded RNA red (ex 460 nm, em 640 nm) Also accumulates in acidic compartments such as the vacuole

Preferentially binds to AT-rich regions with a 20-fold increase in fluorescence

Preferentially binds to AT-rich regions of DNA with essentially no cytoplasmic labelling

Similar levels of fluorescence from DNA and RNA

red (662 nm) or blue Soret band (430 nm) There is a radiationless transition from the upper (blue) excited state to the lower (red) excited state within 10-12 sec so the fluorescence is usually > 662 nm In vivo there is a range of slightly different chlorophyll pigments that occur in complexes with proteins giving a red shift to the absorbance peaks and the fluorescence emission to > 686 nm In addition, the presence of accessory pigments means that chlorophyll can be excited across a much broader spectral range than its own spectrum would suggest The amount of chlorophyll fluorescence inversely tracks the flow of electrons from the reaction centres and has been used as a powerful tool to monitor the activity of the photosystems and dark reactions in vivo It is possible to monitor a number of parameters, such as the photochemical yield of PS II or the level of non-photo-chemical quenching, with imaging techniques at the tissue level (34), fluorimetry on individual protoplasts (21), or even imaging single chloroplasts within cells in intact leaves (35)

It is possible to use other fluorescent probes to monitor photosynthetic activity in chloroplasts For example, fluorescein tetrazolium (Polysciences) added at 0.01-0.001% (w/v) from a 0.1% stock in methanol is reduced within the chloroplasts over a 30 period to precipitate a fluorescein-tagged diformazan salt (36)

12.5 Mitochondria

NADH can be excited at 365 nm and fluoresces at 460 nm In animal tissues, most of the fluorescence attributable to NADH comes from the mitochondria as cytosolic NADH fluorescence is often quenched and fluorescence from NADPH has a much lower quantum yield (37) It is possible to image NADH fluorescence using fluorimetry, camera-based, confocal, or multi-photon imaging (37) As many other cellular components will also fluoresce at these wavelengths, typical treatments include changing the oxygenation state of the tissues, with anoxia leading to increased fluorescence as NADH levels increase Although UV imaging is not easy in plants, the development of multi-photon systems may facilitate NADH measurements in intact systems

Several probes label mitochondria with reasonable specificity at low concen-trations, including DiOC6, DiOC7 (38), Rhodamine 123 (39), MitoTracker, and JC-1

(Table 3) The fluorescence intensity depends on the mitochondrial membrane

potential that drives accumulation of the dye In the case of JC-1, the dye mo-lecules further stack-up at certain concentrations to form J-aggregates which shift the emission from green to red This process appears to be promoted by specific interactions with cardiolipin in the mitochondria (J Ermantraut, P Spanu, and M D Flicker, unpublished) JC-1 can, in principle, be used as a ratiometric indi-cator for mitochondrial membrane potential, however, it is very difficult to cali-brate the response of JC-1 in intact tissues unless the concentration of dye in the cell can be guaranteed to match that in the surrounding medium

Table Vital probes for plant mitochondria Dye DiOC6a H2DCF-DA JC-1 MitoTracker Green FM MitoTracker Orange CMH2-TMRos

MitoTracker Orange CMTMRos MitoTracker Red CMH2-XRos MitoTracker Red CMXRos Rhodamine 123 Stock solution mg/ml in DMSO,-20°C, dark

5 mM stock in ethanol mg/ml in DMSO

1 mM DMSO, -20°C, dark

1 mM DMSO, -20°C, dark Store under

Ar or N2C mM DMSO, -20°C, dark

1 mM DMSO, -20 °C, dark Store under

Ar or N2c

1 mM DMSO, -20°C, dark

1 mg/ml in H20, MeOH, or 10 mg/ml in DMSO Loading conditions 0.1 ug/ml, 10-30 10 uM 0.1-1 MM (sparingly soluble)

100 nM for

30-60

in dark 500 nM for 15-30 E

100 nM for 30-60 in dark 100 nM for 30-60 in dark

100 nM for 30-60 in dark 0.1-10 uM, 10-30 Ex nm 484 495 514 490 at 24°C 554 579 507 Em nm 501 529 529/59 516 576 599 529

E 10-3

154 32 '0 119 102 116 101 Comments

Labels mitochondria at low concentrations and ER at higher concentrations

Non-fluorescent until oxidized

195 JC-1 forms J-aggregates with Abs/Em = 585/590 nm at concentrations above 0.1 uM in aqueous solutions (pH 8.0) Mitochondrial accumulation is insensitive to membrane potential, but gives weaker labelling than the other MitoTracker dyes Non-fluorescent until oxidized Localizes to the mitochondria and vacuole

Can be fixed with 4% paraformaldehyde in PBS and retains localization and fluorescence

Non-fluorescent until oxidized Does not appear to label plant mitochondria wellb

Can be fixed with 4% paraformaldehyde in PBS and retains localization and fluorescence

More specific for mitochondria at low concentrations than DiOC6

a The related probe 3,3'-diheptyloxacartoocyanine iodide (DiOC7(3)) is also reported to give good mitochondrial staining in plants (38)

bJ Balk, personal communication

c Packaged as 50 ug lyophilized solid Store at-20°C in dark It is best to make up the stock solution and use on the same day

inhibit mitochondrial electron transport and reduce fluorescence, whilst monen-sin (0.1 mM) stimulates fluorescence through hyperpolarization of the membrane potential (39)

The MitoTracker probes are cell permeant mitochondrion-selective dyes They also have a mildly thiol-reactive chloromethyl moiety that increases retention of the dye following fixation MitoTracker dyes are also available in reduced forms that are essentially non-fluorescent until they are oxidized, in a similar manner to H2DCF The fluorescent probe then accumulates in the mitochondria

12.6 Vacuoles

A considerable number of vacuolar pigments, such as flavonoids and antho-cyanins, show strong autofluorescence, particularly with UV excitation Generic classes of compound can be identified by their spectral properties and how these alter in response to alkalinization or treatment with Naturstoffreagenz A (NRA) which induces secondary fluorescence from flavonoids (41) Thus, the green vacuolar fluorescence of flavonoids can be enhanced by exposure of the tissue to 0.5% (w/v) NH3 or by incubating fresh cut sections for in 0.1% (w/v) NRA diluted from a 2.5% (w/v) stock in ethanol (41)

Multi-photon laser scanning microscopy may be a useful tool to study auto-fluorescence of heavily pigmented cells The primary infrared excitation wave-length is affected far less by attenuation and scattering in comparison with single-photon UV or visible excitation, but can excite a wide range of antho-cyanin and flavonoid components with subcellular resolution (A J Meyer and M D Fricker, unpublished)

Several different mechanisms can be exploited to load vacuoles with fluor-escent dyes, including cation trapping, active transport of anions, or glutathione S-conjugates The different metabolic requirements for vacuolar accumulation of these different classes of probes can be useful to identify different types of vacuoles (42), equally, however, none of these probes can be considered as a generic marker for all types of vacuole

Weakly basic dyes, such as Acridine Orange (AO), Neutral Red (NR), and the LysoSensor probes (Table 4), are membrane permeant in their uncharged form but become trapped in acidic compartments upon protonation AO can be excited at a range of blue/green wavelengths and accumulates in some, but not all, vacuoles depending on the pH gradient and extent of binding to intra-vacuolar components AO also binds to other cellular components, notably DNA (see Section 12.3), and can be highly phototoxic at high concentrations

Many anionic dyes, such as Lucifer Yellow, BCECF, CDCF, or carboxy-SNARF-1, are also effectively sequestered in the vacuole of some cell types when loaded as membrane permeant esters or at low pH (30) In the case of dyes loaded as esters, localization may reflect the distribution of the appropriate esterase activity (30) In other cases, sequestration probably involves specific transporters These may include multi-drug-resistance pumps, glutathione (GS-X) pumps (43), or sulfon-ate transporters (44) At least some of these transporters can be inhibited by pre-incubation for h with mM probenecid (45) Probenecid has to be solubilized under alkaline conditions (pH 11.5) and then adjusted to pH with HC1 to give a 10 mM stock solution (45)

5-chloro-Table Fluorescent probes used for vacuolar labelling Dye Acridine Orange BCECF-AM CDCF-DA CMAC (CellTracker Blue) CMAC, CBZ-Phe-Arg CMFDA (CellTracker Green) LysoSensor Yellow/Blue DND-160 Monochloro-bimane (MCB) Neutral Red Stock solution

1 mg/ml in H20

1 mM in DMSO

1 mM in DMSO

10 mM in DMSO l0min

20 mM in DMSO or EtOH

5 mM in DMSO

1 mM in DMSO, -20°C, dark

5 mM in DMSO

1 mg/ml in water

Loading conditions

1 ug/ml for

3-10 uM for 10-60 at 18-22 °C

5 uM for 10-60 at 18-22oC

100 uM for

20 uM, h

50 uM, h

10 uM, h (medium pH increased to pH 8)

50 uM, h

70 uM, 20 Ex nm 489 503 504 354 353 475 384 395 541 Em nm 520 529 529 466 466 517 540 477 640

E 10-3

65 90 90 14 14 29 21 39 Comments

Weak amine that accumulates in acidic compartments Also used as a vital DNA stain

Initially fluorescent in the cytoplasm Can be used as a dual-excitation pH indicator

Labels the vacuole and small vesicles in the cytoplasm Can be used as a dual-excitation pH indicator

Conjugated to GSH and transported into the vacuole by a GS-X pump

Cleavage of the peptide gives a substantial shift in the absorption and emission spectrum

Conjugated to GSH and transported into the vacuole by a GS-X pump Only fluorescent after cleavage of the acetate groups Can be used as a dual-excitation pH indicator

Weak amine that accumulates in acidic compartments Ratioable pH indicator with pH-dependent shifts in both excitation and emission spectra

Only fluorescent after conjugation to glutathione Transported into vacuole by a GS-X pump

Spectra are pH dependent with a pKa~6.7

methylfluorescein diacetate (CMFDA, CellTracker Green), or 7-amino4-chloro-methylcoumarin (CMAC, CellTracker Blue) are substrates for conjugation to GSH in the cytoplasm The GS-X conjugate is then transported into the vacuole by a GS-X conjugate pump MCB is not fluorescent until conjugated to GSH (see Section 13.5), CMFDA is not fluorescent until the acetate groups are cleaved off, whilst CMAC is fluorescent whether conjugated or not

in-hibited by pre-incubation with the cysteine protease inhibitors E-64 or leupeptin at 100 ug/ml (42)

12.7 Endoplasmic reticulum

A variety of fluorescent probes have been used to label the endoplasmic reticu-lum in plants, including DiOC6, Rhodamine B hexyl ester, and chlortetracycline (CTC, aureomycin, see Table 5) Typically, cells or tissues are loaded by incubation in the dye for 5-30 min, followed by a series of washes in dye-free medium None of these dyes are specific for the ER; typically they label mitochondria at low concentrations and can label other intracellular membrane compartments at high concentrations In species that can be transformed, ER-targeted versions of GFP provide a more reliable ER marker (see Chapter 5) Other probes may initially insert into the plasma membrane but become redistributed into internal branes following flip-flop from the outer to inner leaflet of the plasma mem-brane This has been suggested for NBD-phosphatidylcholine, which labels the ER after metabolism to the diacyglycerol form (NBD-DAG) by putative plasma membrane Ca2+-dependent phospholipase C activity (10)

Cells labelled with the ER dyes DiOC6, Rhodamine B hexyl ester or AFC12 have been photobleached to show that ER lipids can diffuse through plasmodesmata between neighbouring cells (10) In contrast, probes that are restricted to the plasma membrane, such as NBD-SM, not appear to move (10)

12.8 Golgi

Lipid-based fluorescent probes used to label the Golgi in animal cells, such as NBD-ceramide, not appear to label the Golgi in plant tissues Thus at the moment the most practical route to visualize the Golgi in living cells is through

Table Vital probes for plant endoplasmic reticulum

Dye CTC DiOC6 Dodecanoyl-amino fluorescein (AFC12)

NBD-PCa

Rhodamine B hexyl ester

Stock solution

1-5 mg/ml in EtOH or DMSO

1-5 mg/ml in EtOH or DMSO 10-15

1 mg/ml in EtOH

1-5 mg/ml in EtOH or DMSO

Loading conditions

50-200 uM for 0.5-1 h

Ex nm

405

1-5 ug/ml 484 for 10-60

2 (Ag/ml at 4-25 °C for

40 ug/ml at 25°Cfor 10-15

1 ug/ml at 4-25°C for 10-15 495 465 556 Em nm 530 501 518 534 578

E 10-3

154

85

22

123

Comments

ER, mitochondria, and spherosomes Also used as a Ca2+ indicator

Labels ER, mitochondria, and lipid bodies

Labels ER, mitochondria, and lipid bodies.The fluorescein moiety confers pH fluorescein sensitivity to the fluorescence

Labels intracellular membrane after conversion to NDB-DAG by a Ca2+ -dependent phospholipase (10)

Labels ER, mitochondria, and lipid bodies Shows better photostability than DiOC6

targeted GFP-derivatives (see Chapter 5) Fluorescent Bodipy BFA (Molecular Probes) labels the Golgi at very low concentrations (0.2 ug/ml from a 100 ug/ml stock in DMSO) and does not disrupt its morphology (C, Hawes and C Steele-King, personal communication),

12.9 Cytoskeleton

The fungal toxins, phalloidin and phallacidin, bind to and to some extent stabilize F-actin Several fluorescently-tagged derivatives are available that can be intro-duced into cells by microinjection or after partial permeabilization of the plasma membrane (46-48) (Protocol 8) At low concentrations they not appear to greatly perturb cell function and dynamic changes in microfilament arrays can be followed Actin filaments may be partially stabilized using m-maleimido-benzoic acid N-hydroxysuccinimide ester (MBS), particularly prior to fixation

Fluorcscently-taggcd actin and tubulin have also been introduced into cells, usually by pressure microinjection (e.g 12, 46-51) (Protocols and 9) in a tech-nique known as fluorescent analogue cytochemistry (reviewed in refs 2, 50).

Labelling of actin filaments in living cells using fluorescent phallotoxinsa

Reagents

• Fluorescein-phalloidin, rhodamine-phalloidin,b or Bodipy FL phallacidin stock solutions (6.6 uM in MeOH) stored frozen in 20 ul aliquots

• Microinjection buffer: 100 mM KCI • PBS buffer: 0.14 M NaCl, 2,7 mM KC1, 1.5

mM KH2PO4, 8.1 mM Na2HPO4 pH 7.0

• Actin stabilization buffer: 100 uM m-maleimidobenzoic acid

N-hydroxysuccinimide ester (MBS), 50 mM Pipes pH 7.0

• Permeabilization buffer: 20 uM MBS, mM EGTA, mM MgSO4 0.01% (v/v) Nonidet P-40, 1% DMSO, 50 mM Pipes pH 6,8

A Microinjection

1 Evaporate the MeOH from the stock solutions of fluorescent phallotoxin and redissolve in buffer to 0.66 uM

2 Sonicate and centrifuge (5 at 10000g).

3 Load cytoplasm to a final concentration of 0.02-0.05 uM by pressure microinjection (see Chapter 6)

B Permeabilization

1 Lightly cross-link F-actin in actin stabilization buffer

3 In some tissues direct incubation in 30 nM labelled phallotoxin may give sufficient labelling

a Methods based on refs 46-48.

b Rhodamine-phailoidin increases its fluorescence on binding F-actin, is more photo stable than fluorescein-phalloidin, and is better retained within cells after fixation

Labelling of microtubules in epidermal cells using fluorescent analogue cytochemistrya

Reagents

• Sample buffer: 50 mM potassium glutamate, 0.5 mM MgCl2, and 2.75% (w/v) sucrose

• Tubulin buffer: 20 mM sodium glutamate, 0.5 mM MgSO4, and mM EGTA

• Carboxyfluorescein-tubulin prepared according to Zhang et al (49), or

tetramethyl-rhodamine tubulin prepared according to Wymer et al (51), and frozen at -80°C at 3-10 mg/ml in tubulin buffer

Method

1 Immobilize an epidermal peel from a pea stem on a slide using adhesive tape (Cellux) and construct a tape chamber around the specimen,

2 Wet the specimen with sample buffer and leave for 15 to recover

3 Select cells with cytoplasmic streaming using a X63/1.2 NA water immersion objective Thaw CF-tubulin stock and dilute to mg/ml in tubulin buffer supplemented with

1 mM GTP on ice

5 Load injection pipettes (0.75 um outer diameter) by tip filling from a ul drop

6 Load cells by pressure micro injection with to a final concentration of 0.3 uM Remove injection needle within

a Modified from the methods in refs 12 and 48

12.10 The plasma membrane and endocytosis

Table Vital probes for the plasma membrane Dye FM1-43 NBD-PE (NBD- phosphatidyl-ethanolamine) NBD-SM (NBD- sphingosyl-phosphocholine) Stock solution

5 mg/ml in DMSO

1 mg/ml in EtOH (may need sonication)

1 mg/ml in EtOH (may need sonication)

Loading conditions

1 uM

40 ug/ml at 4°C for 15 ug/ml at 4°C for 10-20 Ex nm 510 463 436 Em nm 626 536 566

E 10-3

66

21

22

Comments

The absorption and emission shift to shorter wavelengths by 20 nm and 80 nm respectively in polar environments

Labels internal membranes with time and at higher temperatures (9)

Does not appear to be metabolized to a DAG-derivative and remains in the plasma membrane (10)

The cationic styryl FM1-43 and FM4-64 dyes are essentially non-fluorescent in the external medium but become brightly fluorescent when incorporated into the outer leaflet of the plasma membrane The dyes not spontaneously re-orient in the membrane so that dye appearing internally should represent inter-nalization of the plasma membrane and provides an estimate of the membrane flux into the cell and the compartments that trafficking is directed Internal-ization of plasma membrane has been followed after treatment with uM FM1-43 for 5-60 (52, 53) The use of protoplasts also permits parallel electro-physiological measurement of changes in membrane capacitance

Other probes have been used to label surface components, for example, rhodamine-labelled fucose-specific lectin (UEA I-TRTTC, Sigma) binds to densely cytoplasmic protoplasts from the hypersecretory leyer of maize root tips when used at mg/ml for h (52) Alternatively, fluorescent markers can be covalently attached to membrane lipids or proteins (54)

12,11 The cell wall

Many cell walls show strong autofluorescence from lignin components, particu-larly with excitation at UV wavelengths (41) or using two-photon excitation from 750-800 nm In some cases, the spectrum can be attributed to specific compo-nents, such as equisetumpyrone in Equisetum arvense (41).

Calcofluor White MR2 (0.1 mg/ml, 10 min) is a very bright label for cellulose, but requires UV excitation Primulin (200 uM, h) is less bright, but can be excited with a 442 nm He-Cd line in confocal applications Propidium iodide (1-10 ug/ml, 10 min) can be excited using a broad range visible wavelengths and has increasingly found applications as both a stain for cell walls and a viability indicator (see Section 12.2)

determined from confocal fluorescence anisotropy measurements with polarized excitation at 514 nm and simultaneous dual-channel detection equipped with crossed polarizers (6)

In addition to labelling cell wall polymers, it is possible to probe the physio-logy of the apoplast with cell impermeant dyes, loaded directly by incubation, vacuum infiltration, or through the transpiration stream Optically, the very thin dimensions of the wall make quantitative measurements more difficult, how-ever, the activities of a number of ions have been successfully measured and/or imaged in the apoplast (see Section 13.2)

13 Physiological probes

The basics of physiological measurements are presented here, primarily for Ca2+ and pH Additional information and biological applications of the techniques are covered in refs 1-3, 16

13,1 Calcium

A wide range of Ca2+ dyes are available (Table 7), but the majority of cytoplasmic measurements have used the single wavelength Ca2+ indicators Fluo-3 and Calcium Green-1 (notably for confocal measurements) (55), or the dual-excitation ratio dye Fura-2, notably for camera imaging experiments CTC is the only freely membrane permeant dye that can be used to infer changes in calcium (56) and BTC is the only probe so far used for apoplastic Ca2+ measurements (57) The other dyes are typically loaded as AM esters (Protocol 3) or at low pH (Protocol 4), however, Ca2+ dyes are notorious for either not loading well into plant tissues or for rapidly compartmentalizing following loading In some cell types, the level of intracellular compartmentalization can be estimated following selective per-meabilization of the plasma membrane with 10-100 uM digitonin to release the cytosolic dye The fluorescence decreases to a stable value over a few minutes that represents the amount of dye trapped in intracellular compartments Sub-sequent addition of Triton X-100 (0.1%, v/v) is sufficient to release all the intra-cellular dye The fluorescence from dye accumulated in the ER can be used to estimate ER Ca2+ levels (58), however, the concentrations fall outside the useful reporting range of most cytosolic indicators Attempts to measure localized plasma membrane fluxes of Ca2+ with the lipophilic derivative Fura-C18 were not successful as the dye rapidly killed the cells (17)

Table Fluorescent calcium dyes used in plant cells Dye BTC Calcium Green-1 Calcium Green-1 dextran 10 000

CTC Fluo-3 Fura-2 Fura-2 dextran 10000 Fura-C18 lndo-1 lndo-1 dextran 3000 Quin-2 Yellow Cameleon Yellow Cameleon 3ER Loading protocol Vacuum infiltration of apoplast Microinjection; low pH; AM ester

Pressure microinjection

Direct incubation

Microinjection; low pH; AM ester, detergent assisted

Microinjection; low pH; AM ester

Pressure microinjection

Pressure microinjection

Microinjection; low pH; AM ester

Pressure microinjection

Microinjection, AM ester, or electroporation Via transformation Via transformation Ex nm -Ca +Ca 464 401 506 510 405 503 363 335 364 338 364 338 346 330 341 408 353 333 Em nm -Ca +Ca 533 529 531 535 530 525 512 505 500 500 475 401 356 466 495

433 480 535

433 480 535 Kd uM 0.19 0.26 10-440 0.39 0.14 0.24 0.15 0.23 0.32 0.06 0.07 and 11 4.4 Properties

Dual-excitation, low affinity Ca2+ dye,

used to measure apoplastic Ca2+

levels.(57)

Single wavelength dye Brighter and more stable than Fluo-3

Membrane impermeant version of Calcium Green well retained in the cytosol after microinjection

Low affinity, permeant Ca2+ indicator

Also responds to hydrophobicity, pH, Mg2+ (56)

Single wavelength dye with large (40-fold) enhancement in fluorescence on binding Ca2+

Dual-excitation UV ratioable dye

Membrane impermeant version of Fura-2 well retained in the cytosol after microinjection

Measures Ca2+ adjacent to membranes

but appears to clot the cytoplasm and kill algal cells (17)

Dual-emission, UV-excited dye Rather prone to photobleaching

Membrane impermeant version of lndo-1 well retained in the cytosol after microinjection

Now superseded by other Ca2+ dyes

FRET-based transgenic cytoplasmic Ca2+ indicator (7, 8)

FRET-based transgenic cytoplasmic Ca2+ indicator targeted to the ER

13.1.1 In vitro calibration—calcium dyes

to define the dynamic range expected for the dye response tinder the collection conditions used for the experiment It is rarely used to calibrate the Ca2 re-sponse in vivo, as the calibration is only valid for a single dye concentration and path length For ratio dyes, an in vitro calibration provides an indication of the Ca2* concentration in vivo, if the calibration solutions matches the conditions ex-perienced by the dye in the cytosol with respect to pH, ionic strength, viscosity, hydrophobiciry, temperature, and protein binding Unfortunately, values for Kd measured in situ can be markedly different from that estimated in vitro (62).

The spectra for dextran conjugates is often different from the free dyes and the Kd also varies from batch to batch depending on the length of the dextran molecules and the degree of substitution (62) For example, Rira-dextran has a dissociation constant around 350 nM compared to 150 nM for Pura-2 and the peak in the excitation spectrum for the Ca2 free form is shifted to shorter wave lengths (from 380 nm to 364 nm) (63) In general, definition of the appropriate composition of the calibration solution can be best estimated from comparison of the dye spectra in vivo with that determined in vitro under a variety of hydro-phobicity, viscosity, and ionic composition regimes

In vitro calibration of calcium ratio dyes3

Equipment and reagents

• Scanning fluorimeter, such as a Perkin-Elmer LS50B with ml quartz cuvettes • CaEGTA buffer 100 mMKCl, 10 mM Mops

pH 7.2, and 10 mM CaEGTA made up in deionized water

• K3EGTA bufrer:100mM KCl,10 mM Mops pH 7.2 and 10 mM K3EGTA made up in deionized water

A Measurement of the Kd of the dye

1 Make up the indicator dye as a mM stock solution in buffer without Ca2+ or EGTA Add 20 ul dye to 1.98 ml K2EGTA

3 Measure the appropriate excitation or emission spectrum of the dye in zero-Ca2+ Add 60 ul dye to 5.94 ml CaEGTA solution

5 Remove 0.2 ml solution from the K2EGTA + dye sample in the cuvette and replace with 0.2 ml of the CaEGTA + dye sample to give a total Ca2+ concentration of mM Measure the spectrum again

7 Continue the sequential dilution procedure according to the following table:

[CaEGTA] required (mM) Volume to remove/replace (ml) Approx free Ca2+(nM)

8 The free Ca2: concentration is calculated from the Kd of EGTA for Ca2' using the following equation:

9 The Kd for EGTA at 20°C pH 7.2 and 100 mM KC1 is 150.5 X 10-9 M however, the Kd is highly sensitive to the temperature, pH, ionic strength, and presence of other ions and has to be established for other calibration solutions Detailed protocols are given in refs 60, 61

10 In addition to changing the pH and ionic strength of the calibration solution to match that expected in the cytoplasm, the viscosity may be increased by addition of 20-60% sucrose or 500 mM mannitol, and hydrophobicity altered with 25% ethanol 11 The actual free Ca2- is calculated using an iterative computer program that accounts for all the ionic interactions in the calibration buffer A program called Maxchelator (MAXC) is freely available on the Web at http://www.stanford.edu/-cpatton/maxc html

B Calibration of the response of the imaging system

1 Set up a calibration series as in part A but reserve a small volume (ca 100 ul) at each step and adjust the volume removed and replaced accordingly Alternatively use Ca2+ buffer kits with pre-diluted solutions

2 For confocal or multi-photon measurements, the sampling volume is defined by the instrument settings and so accurate control of the path length in the calibration solutions is not required For other systems, measurements should be made using volumes of calibration solution similar in size to loaded cells in the experimental apparatus by either:

(a) Vortexing 100 u1 of immersion oil (Fisher Scientific, Type FF) with ul of dye solution in a microcentrifuge tube to obtain cell-sized droplets

(b) Using flat sided rectangular capillaries (W5005, Vitro Dynamics Inc., Rockaway, NJ USA) with a defined (50 um) internal path length

3 Measure the fluorescence at both excitation or emission wavelengths for each cali-bration solution, subtract the value measured at each wavelength in the absence of dye and then calculate the ratio

4 Experimentally derived values are calibrated using the following equation, sub-stituting in appropriate values from the in vitro calibration:

5 It is more convenient to rearrange Equation for use with a non-linear curve fitting package as follows:

The measured ratio at each pCa value can be fitted by this function with Rmax Rmin, and pKd, as variables (63) The fitted sigmoidal response is used to estimate Ca 2-concentrations from the ratios measured in the living system and also gives the Kd

of the dye

a Method based on Calcium Calibration Buffer Kits from Molecular Probes (62).

13.1.2 in situ calibration—calcium dyes

Dye loaded cells are permeabilized with a Ca2 ionophore, such as ionomydn, or 4-bromo A23187 (a non-fluorescent analogue of A23187) at the end of the experiment (Protocol 11) lonomydn is less effective than Br-A23187 in plant cells (59) as it requires alkaline (pH 9) rather than acidic (< pH 7) conditions normally encountered in perfusion solutions Cytosolic Ca24 is then set by extracellular Ca2- -EGTA or Ca2'-BAPTA buffer solutions Ideally cytosolic conditions experi-enced by the dye should not have changed significantly between the calibration and in vivo measurements of Ca2' during an experiment In practice, it is often only practical to perform an in situ calibration at two points, typically at very low Ca2 concentration to determine Rmin and FmaxY2, and at a saturating Ca2 concen-tration to determine Rmax and FminY2 In this case the measured ratios can be con-verted to concentration values only if the Kd of the dye is known inside the cell This is often not the case in plant cells

Results from single wavelength dyes are difficult to interpret, as there is no inherent correction for dye leakage, redistribution, or bleaching An alternative approach to using EGTA to deplete Ca2 is to add mM Mn2' that permeates Ca2 channels and quenches the fluorescence The Mn24 quench appears to give more consistent results than determining Fmin with EGTA (64) Even so, calibra-tion is extremely difficult to perform accurately and in general data from single wavelength dyes is more qualitatively useful than quantitatively accurate

A second major problem in plant cells is that ionophores not equilibrate Ca2 concentrations to a sufficient or reproducible extent across the plasma membrane lonophore concentrations greater than uM often act like deter-gents and damage the cell, however concentrations up to 100 uM have to be applied to penetrate into the tissues and to shift the internal ion concentration Thus, in many cases in situ calibration is not a reliable method for plant cells, leaving an in vitro calibration as the only alternative.

calibrations may be required during the experiment or for different regions in the cell It is possible to determine the shift in Kd of the dye in appropriate cali-bration solutions (once the effect of pH on the calicali-bration buffer itself has been taken into account) The modified Kd is then used in Equation (63)

In situ calibration of calcium dyes

Reagents

* Ionomycin or Br-A23187

Method

1 Add l-10 uM ionomycinor Br-A23187at the end of the experiment

2 Increase external Ca2+ to -1 mM and allow the signal to stabilize at Rmax (ratio-metric dye) or Fmax (single wavelength dye)

3 Replace the high Ca2+ medium and perfuse with medium containing mM EGTA and allow fluorescence to stabilize at Rmin or Fmin

4 An alternative to determining Fmin for single wavelength dyes is to use Mn2+ to quench the fluorescence from the dye Add 0,1-1 mM MnCl2 with 10 uM ionophore for 10 Fluo-3 fluorescence is quenched to 0,2 x Fmax This set point can then be used in a modified form of the equation below (64)

5 Calculate [Ca2+]cyt using published Kd values or Kd values measured in vitro using Equation for ratio dyes or the following equation for single wavelength dyes;

13.1.3 The manganese quench technique

The fluorescence quenching observed with M n2' can also be used to discriminate between Ca2- influx and Ca2' mobilization from internal stores during a Ca2+ response If Mn2+ is present in the external medium, opening of plasma mem-brane Ca2+ channels allows Mn24 influx and consequent fluorescence quenching In the case of fura-2, this can be readily measured at the iso-excitation wave-length around 360 nm (66)

13.1.4 Dissipation of intracellular calcium gradients using buffers with varying pKd

In root hairs, pollen tubes, and rhizoids a steep tip-focused gradient in Ca2+

cor-relates with tip growth (55, 67) The Ca2+ gradient can be dissipated by

micro-injection of 0.1-1 mM Ca2+-BAPTA buffers with varying Kd arising from

substitutions on the BAPTA moiety These buffers are thought to increase the mobility of Ca2+ in the cytoplasm, particularly if the Kd falls between the

con-centrations expected at the high and low points of the gradient (68, 69)

13.1.5 Calcium measurements using aequorin

In addition to fluorescent probes, luminescent techniques have also been ex-tensively used to measure calcium in plants using aequorin Aequorin is a Ca2+

-dependent photoprotein extensively used to measure Ca2+ dynamics in plants

Active aequorin is reconstitued in vivo from a 22 kDa apoaequorin and a low molecular weight luminophore called coelenterazine in the presence of oxygen When calcium is bound, the coelenterazine is oxidized to coelenteramide and the protein undergoes a conformational change accompanied by the release of carbon dioxide and emission of blue (462 nm) light Aequorin is highly selective for Ca2+, for example Mg2+ and K+ not trigger luminescence, although these

ions may depress the Ca2+ sensitivity Aequorin can potentially detect free

cal-cium levels of up to 100 uM, although in practice, most measurements are made in the range of 10 nM-10 uM Aequorin has proved a very versatile Ca2+

re-porter, particularly in intact transgenic seedlings (70, 71), although it can also be used at the single cell level with difficulty (72) Aequorin has several potential advantages over fluorescent dyes as an indicator for Ca2+ Luminescence

measurements usually have an intrinsically high signal-to-background ratio as there is relatively little endogenous luminescence under optimal conditions As a natural protein, aequorin is expected to be non-toxic and remain in the cyto-plasm unless specifically targeted elsewhere Light emission is unaffected by pH values greater than pH Photodamage associated with excitation illumination for fluorescence is also avoided For full methodological details on the use of aequorin in plants, the reader is referred to refs 70 and 71

13.2 Measurement of apoplastic, cytoplasmic, and vacuolar pH

Measurement of cytoplasmic pH follows similar principles to cytosolic Ca2+

measurements A range of dyes based on fluorescein has been developed with different substituent groups that alter the pKa of the probe (Table 8) Several of

dextran 10000 and perform dual-excitation (488 and 543 nm), dual-emission (515-540 and > 590) ratioing (74) The SNARF and SNAFL dyes have even more complex pH-dependent excitation and emission spectra and can be used in either dual-excitation or dual-emission mode (3, 75-77)

Table Fluorescent pH dyes used to measure apoplastic, cytoplasmic, or vacuolar pH

Dyea 5-(and-6)-carboxy SNAFL-1 5-(and-6)-carboxy SNARF-1a 5-(and-6)-carboxy-2',7'-dichlorofluorescein 5-(and-6)-carboxy-4',5'-dimethylfluorescein 5-(and-6)-carboxy-fluorescein BCECFa CI-NERFa DM-NERF FITC-dextran 4000 HPTS (pyranine) LysoSensor Yellow/Blue DND-160

Oregon Green 488 carboxylic acid Ex (nm) Low High pH pH 508 548 495 500 475 482 504 497 478 403 384 478 540 576 504 507 492 503 514 510 492 454 329 492 Em (nm) Low PH 543 587 I 520 527 540 High pH 623 635 529 537 517 528 540 536 517 511 440 518 pKa 7.8 7.5 4.8 7.0 6.4 7.0 3.8 5.4 6.2 7.3 4.2 4.7 Comments

Dual-excitation or dual-emission pH indicator

Dual-excitation or dual-emission pH indicator

Dual-excitation ratio indicator using iso-excitation wavelength ~ 440 nm for vacuolar pH

Dual-excitation ratio indicator using iso-excitation wavelength

~ 440 nm

Dual-excitation ratio indicator using iso-excitation wavelength ~ 440 nm for acidic

compartments

Dual-excitation ratio indicator using iso-excitation wavelength ~ 440 nm Used for cytoplasmic and vacuolar pH measurements

Dual-excitation ratio indicator using iso-excitation wavelength ~ 440 nm for vacuolar pH

Dual-excitation ratio indicator using iso-excitation wavelength ~ 440 nm for vacuolar pH

Dual-excitation ratio indicator using iso-excitation wavelength ~ 427 nm for apoplastic pH

Dual-excitation ratio indicator Used for vacuolar pH measurements and as a symplastic tracer

Dual-excitation, dual-emission ratioable pH indicator used for vacuolar pH measurements

Dual-excitation ratio indicator using iso-excitation wavelength

~ 440 nm for vacuolar pH

Apoplastic pH measurements involve loading dyes with pKa values in the

range pH 3-7, such as Cl-NERF, Cl-NERF dextran 10 000, or FITC-dextran 4000, by infiltration or via the transpiration stream, at concentrations of 30-100 uM (74, 78-80) It may be necessary to keep the dye present in the medium to avoid washout from the apoplast and use confocal optical sectioning to visualize signal from the apoplast Dual loading with a pH-insensitive dye (Rhodamine or Texas Red dextran 3000) has also been used to facilitate ratio pH measurements from the apoplast (74, 79) Linking the dye to dextran ensures that there is no loading into the protoplast, however, some fluorescence properties of the dextran-linked dyes differ from the free fluorophore Thus FITC-dextran does not have a true iso-excitation wavelength, but shows strong fluorescence quenching at acidic pH values (78) Apoplastic calibration can be performed in situ provided the buffer strength is increased to at least 50 mM to overcome the pH buffering of the cell wall (74)

Other indicators may be useful for apoplastic pH measurements, for instance 6-glucoxy-7-hydroxycoumarin (0.1-1 mM from M stock in M KOH) was loaded by vacuum infiltration into the apoplast of a range of species and tissues and remained apoplastic Other coumarins tested had a higher membrane permea-bility and were judged not to be useful for apoplastic measurements (81)

Vacuolar pH measurements rely on uptake and compartmentalization of appropriate dyes in the vacuole (e.g esculentin, pyranine, and fluorescein derivatives, CF, CDCF, BCECF) Estimates of pH have been based on the ability of intact cells to take up a variety of fluorescent pH indicators with differing Kd

values The changes in fluorescence can be attributed to pH in particular com-partments on the basis of the distribution of the dye and the pH range over which the dye is responsive (26) Care has to be taken with such measurements from intact tissues as changes in fluorescence may result from changes in other parameters, such as light scattering (27) More conventional ratioing approaches may suffer contamination of the vacuolar signal with cytoplasmic signal, and therefore represent a complex average of the pH in the two compartments (82) The calibration response of dyes in the vacuole and cytoplasm may be markedly different as the ionic strength, viscosity, and protein-binding interactions will vary

13.2.1 Calibration of ratio pH dyes

The approach for calibration of pH dyes, such as BCECF, is essentially similar to those outlined for Ca2+ dyes In vitro calibration solutions are similar to those

used for Ca2+, and the ionic strength, hydrophobicity, and viscosity may all be

modified to better mimic the response of the dye in the cytosol For example Feijo et al (65) used 100 mM KC1, 30 mM NaCl, 500 mM mannitol, 40% (w/v) sucrose, 25 mM Mes, 25 mM Hepes at varying pH values

For in situ calibrations (Protocol 12), the K+/H+ exchanger, nigericin, has been

used as an ionophore to equilibrate internal and external pH in the presence of high K+ (e.g ref 76) Simultaneous addition of the K+ ionophore, valinomycin,

effective (65), However, two point in situ calibrations arc only useful if the Kd can be established The overlap between in situ and in vitro calibrations can be improved by additions such as deproteinized coconut water supplemented with 1% ovalbumin (83) pH intervals can be monitored fairly accurately even if The absolute level cannot be determined The relatively small shift in ratio values means the limits of reliable detection lie between 0.05 and 0.15 pH units (75, 76, 83), Parallel measurements using pH-sensitive microelectrodes provide a check of the fluorescence calibration,

It is also difficult to overcome the intrinsic pH buffering and cellular pH regulation using ionophores and treatments also stress cells rapidly Alternative approaches based on equilibration of permeant weak acids and bases may be more effective and less damaging to the cells (3, 75, 76, S3 84)

In situ calibration of ratio pH dyes using nigericin

Reagents

• Nigericin

Method

1 Increase the external K- to a value close to the anticipated internal K', typically 100-120 mM at the end of the experiment

2 Add nigericin to a final concentration of 10 ug/ml and adjust the pH to pH < 6.0 (for BCECF}.a

3 Allow the ratio value to stabilize (2-15 min) and measure Rmin.

4 Shift the external pH to pH 8,5 in steps to give Rmax and allow the ratio to stabilize.b

5 Estimate the calibration pH values from Equation substituting appropriate values for the pH dye used

6 Determine the experimental pH values from a sigmoidal fit to the in situ calibration data with an assumed Kd or, if there are enough intermediate data points, from the fitted parameters

a Addition of valinomydn (2 uM) may give a more effective clamp (65),

b As it is difficult to shift pH to Rmax and Rmin in situ, an alternative calibration between 0.5-1 pH

unit above and below the pKa to cover the near-linear region of the ratio response can be used

13.2.2 pH clamping with weak acids

the acid anion is trapped in the cytosol after release of a proton Assuming the plasma membrane is permeable only for the undissociated form (HBA) of the acid, an equilibrium will be established between bathing medium and cytoplasm such that the concentration of HBA will be equal in both compartments The amount of anion that dissociates under these conditions is given by Equation 8, provided no transport systems exist for the anion and that the molecule is not metabolized (63):

Acid anion accumulate in the cytoplasm and other compartments with an equivalent release of protons Equation gives an estimate of the amount of protons imported into the cell Most of these protons equilibrate with cytoplas-mic buffer sites Only a few lead to the measured pHcyt decrease Normally there are other cations in the bathing medium (K+, Na+, Mg2+, Ca2+) which can also bind to BA- forming neutral salt molecules (KBA, NaBA, MgBA2, CaBA2) These salt molecules are membrane permeable to an unknown extent, but will cause an overestimate in the calculated amount of protons imported into the cell

The weak acid can be washed out of the cell and, at low concentrations, ex-hibits no severe side-effects on the cell physiology However, cells respond to weak acid loading treatments and significant changes in ion transport systems have been reported (86)

13.3 Potassium

Both cytoplasmic and apoplastic potassium levels have been measured with the potassium binding fluorescent indicator (PBFI, benzofuran isophthalate) For cytoplasmic measurements, PBFI is loaded as the AM ester from a mM stock made up in DMSO/EtOH 1:7.5 The stock is first diluted 2:1 with Pluronic F-127 and then diluted in buffer to a final concentration of uM Protoplasts are loaded for h at 4°C (87) For apoplastic measurements, leaves are vacuum in-filtrated with 50 uM PBFI buffered with 80 mM Mes, 35 mM Tris pH 6, to control for pH sensitivity of PBFI (88) Plants were grown in low Na+ to prevent Na+ interference in the measurements and calibration (88)

13.4 Aluminium

13.5 Measurement of cytoplasmic glutathione levels

Monobromobimane (MBB) has long been used to derivatize low molecular weight thiols to give fluorescent products that can be analysed by HPLC The less reactive analogue monochlorobimane (MCB) can be used at 10-100 uM to fluorescently label GSH in intact cells, if the cells contain an appropriate glutathione S-transferase (GST) to catalyse the conjugation reaction (91) The requirement for a GST also makes MCB labelling far more specific for GSH over other thiols in com-parison to MBB Although the excitation peak of glutathione S-bimane (GSB) is at 395 nm, GSB can be imaged using a confocal microscope equipped with a high-powered UV (364 nm) argon ion laser, the 442 nm line of a HeCd laser (92), or with two-photon excitation at 770 nm (93) Labelling is followed until a plateau is reached, typically within 10-60 The amount of GSB is calibrated against GSB standards prepared by conjugating mM MCB to excess GSH in the pres-ence of U/ml of rabbit liver GST (Sigma) As the GSB formed in the cytoplasm is usually sequestered into the vacuole as part of the normal detoxification path-way, the volume ratio of the two compartments has to be known to calculate the initial cytoplasmic GSH concentration This ratio can be rapidly estimated from a uniform random set of serial optical sections using the Cavalieri estimator of volume Although this can be done manually, commercial software (Digital Sterology, Kinetic Imaging, Liverpool, UK) greatly facilitates analysis (93) An alternative approach is to prevent vacuolar sequestration of GSB by depleting cellular ATP with 1-5 mM NaN3 in which case, the cytoplasmic fluorescence can be directly calibrated against GSB standards Measurements of GSB fluorescence deeper into tissues also require correction for sample- and depth-dependent attenuation according to Protocol 16 (92)

To demonstrate that GSH is labelled specifically the GSH level can be reduced to zero by pre-incubation with mM BSO for 24 h to prevent GSH synthesis or with 10 mM CDNB (a competitive substrate for GSH conjugation) for 10 Subsequent addition of MCB should not result in significant amounts of fluor-escence Labelled tissue can also be analysed by HPLC to confirm that GSH is the predominant thiol labelled and the GSH pool has been labelled to completion

13.6 Reactive oxygen species

Dichlorodihydrofluorescein (H2DCF) is non-fluorescent until oxidized to dichloro-fluorescein (DCF), preferentially by H2O2 in the presence of peroxidases (94) H2DCF can be loaded into cells in the diacetate form (H2DCF-DA) and releases free dye after the ester groups are cleaved (94) (Protocol 7.3) Oxidized dye accumulates in the chloroplasts, mitochondrion, cytosol, and nucleus

Measurement of H202 In situ using H2DCF

Reagents

Method

1 Peel abaxial epidermis from the first fully expanded leaf of Nicotiana tabacum Load with 50 uM H2DCF-DA in buffer for 10 in darkness

3 Wash peels by floating on fresh buffer

4 Immobilize tissue on a microscope slide with silicone grease and immerse in 0.5 ml buffer

5 Observe and/or measure the rate of fluorescence increase using fluorescein filter sets

6 Reactive oxygen species can be artificially increased by:

(a) Addition of exogenous H202, which gives a rapid and linear increase in fluor-escence in the range uM to mM

(b) Addition of Rose bengal (4,5,6,7-tetrachloro-2',4',5',7'-tetraiodofluorescein) at 50 uM Rose bengal forms singlet oxygen (1O2) on irradiation with white light (95) In the plant, 1O2 is rapidly converted to 02 and H2O2

(c) Addition of xanthine (0.5 mM) and xanthine oxidase (0.2 U/ml) to form Q2 which dismutes to give H2O2 spontaneously or is catalysed by SOD (25 U/ml),

14 Data analysis

Measurements from photometry systems intrinsically average the signal from a large area (volume) of the specimen The key stages in analysing the data are to ensure that the dark-current and autofluorescence are correctly measured and subtracted before calculation of the ratio value In photometry measurements, autofluorescence is estimated from the signal measured either prior to loading dye or after quenching the dye at the end of the experiment for the same measurement area However, errors can be introduced into photometry measure-ments from uneven dye distribution within the cells For example, changes in localized regions of the cytoplasm may be swamped by the large signal derived from the nucleus, which may comprise 30-50% of the total A pragmatic approach to the autofluorescence problem is to calculate a mean autofluorescence value from many cells and ensure that this auto fluorescence is less than 10% of the dye signal from the loaded cell

two normally distributed populations gives a highly skewed distribution of ratio values unless expressed on a log scale Thus it is usually more appropriate to visualize changes using ratio images, but to perform quantitative analysis on the original intensity data from the individual wavelength images directly from regions of interest (ROIs) (sec Protocol 15).

Processing ratio images

1 Align images taken at each wavelength in (x, y) to correct for any minor mis-registra-tion between the two wavelength images The extent of misalignment can be determined by imaging a standard fluorescent bead sample with both wavelengths

2 If necessary, increase the S/N ratio in the raw images at the expense of spatial resolution by an averaging filter (e.g averaging over a x box reduces noise threefold)

3 Subtract the instrument background for each wavelength, measured in the absence of the specimen, from all images

4 Correction for tissue autofluorescence is more difficult One approach is to measure autofluorescence from an adjacent region of tissue that is unloaded An alternative is to record an autofluorescence image at a different wavelength that does not inter-fere with the loaded dye and subtract the appropriate 'bleed-through' component from the dye images

5 Mask pixels with low values or those outside the object by setting the intensity to zero with a spatially defined mask Three protocols may be used to define the mask:

(a) An intensity value at a fixed number of standard deviation (s.d.) units above the mean background intensity, typically s.d units

(b) The 50% threshold between the fluorescence intensity within the object and the background

(c) A morphological boundary, such as the edge of the cell, defined from a separate image, such as a bright field view

6 Mask regions in each image that approach saturation of the digitization range A pragmatic approach is to measure the distribution of intensities in a fluorescent area at about the concentration of fluorochrome encountered in vivo and determine the highest mean value where the distribution is not clipped

7 Calculate the ratio image pixel-by-pixel and apply the masks

Quantitative analysis of ratio data from regions of Interest

1 Define regions of interest on one image in the pair

2 Measure the average pixel intensity for this region from both images

3 Subtract the average background values independently for each ROI at each wave-length,

4 Calculate the ratio of the averaged intensities after background subtraction.a

aAn alternative analysis to estimate the average ratio and the confidence limits can be made using application of Bayes theorem, where a priori information can be incorporated into the analysis (76)

14.1 Attenuation correction for optical sections deep into tissues

The intensity response in the axial (z) plane is affected by the axial geometric and chromatic aberrations present along the entire optical path including the speci-men Tissues contain many additional refractive index boundaries which will all contribute to further chromatic and spherical aberration of the confocal probe geometry The consequences of these effects will be increasing signal attenuation with increasing depth through the specimen The complex spatial distribution of the refractive material and the overall geometry of the tissue currently pre-vent development of universal models for lissue-dependent attenuation How-ever, partial correction can be achieved by a more pragmatic approach based on determination of the axial intensity profile of a permeabilized specimen filled with a fluorochrome 'sen' (96) (Pnotncol 16) The resultant response combines the effects of depth-de pendent 'sea' response and the additional contribution of the permeabilized tissue In vivo, the effects are likely to be marginally worse even than this case as a number of refractive boundaries, parlicularly membranes, will be distorted or extracted during the fixation/permeabilization procedure

Measurement of signal attenuation with depth through a permeabilized specimen infiltrated with a fluorochrome 'sea'

1 Fix and permeabilize" the specimen in either:

(a) 4% paraformaldehyde in PBS with 10% DMSO for 30 (b) Fresh ethanol/acetic acid (3:1} for 30

2 Rehydrate through an ethanol series (70/50/30/10/PBS) for 30 each at room temperature

3 Incubate fixed and permeabilized tissue in PBS containing 10-50 uM of fluorescein or Rhodamine B for 24-48 h with gentle agitation to ensure good tissue penetration 4 Collect axial (x, z) optical sections simultaneously in fluorescence and reflection mode

into two channels through the permeabilized, infused specimen, using the same imaging conditions as used for the experiment Sampling should start ca 10 um outside the fluorescent medium and continue at 0.4-0.5 um z-step intervals through the medium plus specimen

5 Measure the average fluorescent intensities in regions of the specimen at varying depth to determine the in situ 'sea' response The most useful regions to measure are in vacuoles or spaces between cells where there is a large volume of homogeneous dye concentration

6 Normalize the attenuation profile to the start of the tissue, defined from the reflection images

7 Profiles can be fitted with a variety of functions: typically a quadratic or cubic function provides a reasonable description of the attenuation

8 The inverse of the parametized attenuation profile is used to generate a correction factor to apply to successive z planes in the experimental data.

0 Allowing the dye to penetrate the tissues is more important than high quality tissue preserva-tion, so crude fixation regimes are acceptable

The magnitude of the attenuation will depend on the depth, lens (particularly NA), immersion medium, bathing medium, and wavelength (96) In structures that have a relatively constant organization in the x, y plane a single axial cor-rection equation may be sufficient (96], For tissues that show a more complex and variable organization, a series of correction equations may be required related to each zone of the tissue (92),

Acknowledgements

Research in the authors' laboratories has been supported by BBSRC, DFG, ELI, NERC, USDA NSF, Nuffield Foundation, Royal Society, INTAS, and Aventis Crop Science,

References

1 Gilroy S (19971 Annu Rev- plant Physiol Plant Mui Bzol 48, 165. 2 Hepler P, K and Gunning B E S, (1998) Protoplasma, 201 121, 3 Roos W (2000) Planta, 210, 347.

5 Shripathi, V., Swamy, G S., and Chandrasekhar, K S (1997) Biochim Biophys Acta,

1323, 263

6 Verbelen, J P and Kerstens, S (2000) J Microsc., 198,101.

7 Miyawaki, A., Griesbeck, O., and Tsien, R Y (1999) Proc Natl Acad Sd USA, 96, 2135. 8 Allen, G J., Kwak, J M., Chu, S P., Llopis, J., Tsien, R Y., Harper, J F., et al (1999) Plant

]., 19, 735

9 Metcalf, T N III, Wang, J L, and Schindler, M (1986) Proc Natl Acad Sd USA, 83, 95. 10 Grabski, S., de Feijiter, A W., and Schindler, M (1993) Plant Cell, 5, 25.

11 Oparka, K J., Duckett, C M., Prior, D A M., and Fisher, D B (1994) Plant]., 6, 759. 12 Hush, J M., Wadsworth, P., Callaham, D A., and Hepler, P K (1994) J Cell Sd., 107,

775

13 Kohler, R H., Cao, J., Zipfel, W R., Webb, W W., and Hanson, M R (1997) Science,

276, 2039.

14 Xu, C., Zipfel, W., Shear, J B., Williams, R M., and Webb, W W (1996) Proc Natl Acad.

Sd USA, 93,10763.

15 Legue, V., Blancaflor, E., Wymer, C., Perbal, G., Fantin, D., and Gilroy, S (1997) Plant

Physiol, 114, 789.

16 Wymer, C L., Beven, A P., Boudonck, K., and Lloyd, C W (1999) In Confocal microscopy:

methods and protocols (ed S W Paddock), p 103 Humana Press, Totowa, New Jersey.

17 Plieth, C and Hansen, U.-P (1996) J Exp Bat., 47,1601.

18 Gisel, A., Barella, S., Hempel, F D., and Zambryski, P C (1999) Development, 126,1879. 19 Allen, G J., Kuchitsu, K., Chu, S P., Murata, Y., and Schroeder, J I (1999) Plant Cell, 11,

1785

20 Reuveni, M., Lerner, H R., and Poljakoff-Mayber, A (1985) Plant Physiol, 79, 406. 21 Goh, C.-H., Schreiber, U, and Hedrich, R (1999) Plant Cell Em., 22,1057.

22 Tretyn, A., Kado, R T., and Kendrick, R E (1997) FoliaHistochem Cytochem., 35, 41. 23 Timmers, A C J., Reiss, H.-D., and Schel, J H N (1991) Cell Calcium, 12, 515. 24 Coleman, J O D., Hiscock, S J., and Dewey, F M (1994) Physiol Mol PlantPathol, 43,

391

25 Wright, K M., Horobin, R W., and Oparka, K J (1996) J Exp Bot., 47,1779. 26 Yin, Z.-H., Neimanis, S., Wagner, U., and Heber, U (1990) Planta, 182, 244. 27 Yin, Z.-H., Hiive, K., and Heber, U (1996) Planta, 199, 9.

28 Knoblauch, M and van Bel, A J E (1998) Plant Cell, 10, 35. 29 Haseloff, J (1999) MethodsCellBiol, 58,139.

30 Huang, C.-N., Cornejo, M J., Bush, D S., and Jones, R L (1986) Protoplasma, 135, 80. 31 Brauer, D., Uknalis, J., Triana, R., andTu, S.-I (1996) Protoplasma, 192, 70.

32 Timmers, A C J., Tirlapur, U K., and Schel, J H N (1995) Protoplasma, 188, 236. 33 Bethke, P C., Lonsdale, J E., Path, A., and Jones, R L (1999) Plant Cell, 11,1033. 34 Genty, B and Meyer, S (1994) Aust.J Plant Physiol, 22, 277.

35 Oxborough, K and Baker, N R (1997) Plant CellEnv., 20,1473. 36 Robertson, D and Earle, E D (1987) Plant Cell Rep., 6, 70.

37 Masters, B R and Chance, B (1999) In Fluorescent and luminescent probes for biological

activity (ed W T Mason), 2nd edn, pp 361-74 Academic Press, London

38 Liu, Z., Bushnell, W R., and Brambl, R (1987) Plant Physiol, 84,1385.

39 Vannini, G L., Pancaldi, S., Poli, F., and Fasulo, M P (1988) Plant CellEnv., 11,123. 40 Maxwell, D P., Wang, Y., and Mclntosh, L (1999) Proc Natl Acad Sd USA, 96, 8271. 41 Hutzler, P., Fischbach, R., Heller, W., Jungblut, T P., Reuber, S., Schmitz, R., et al

(1998)J Exp Bot., 49, 953.

42 Swanson, S J., Bethke, P C., and Jones, R L (1998) Plant Cell, 10, 685.

43 Rea, P A., Li, Z.-S., Lu, Y P., Drozdowicz, Y M., and Martinoia, E (1998) Annu Rev Plant

44 Klein, M., Martinoia, E., and Weissenbock, G (1997) FEES Lett., 420, 86.

45 Cole, I., Coleman, J O D., Kearns, A., Morgan, G., and Hawes, C (1991) ] CellSci., 99,

545.

46 Schmit, A.-C and Lambert, A.-M (1990) Plant Cell, 2,129.

47 Zhang, D., Wadsworth, P., and Hepler, P K (1993) Cell Mold Cytoskel, 24,151. 48 Braun, M and Wasterneys, G O (1998) Planto, 205, 39.

49 Zhang, D., Wadsworth, P., and Hepler, P K (1990) Proc Natl Acad Sci USA, 87, 8820. 50 Hepler, P K., Cleary, A L, Gunning, B E S., Wadsworth, P., Wasteneys, G O., and

Zhang, D H (1993) Cell Bio! Int., 17,127.

51 Wymer, C L., Shaw, P J., Warn, R M., and Lloyd, C W (1997) Plant]., 12, 229. 52 Carroll, A D., Moyen, C., Van Kesteren, P., Tooke, F., Battey, N H., and Brownlee, C

(1998) Plant Cell, 10,1267.

53 Brownlee, C., Manison, N F H., and Anning, R (1998) Exp Biol Online, 3,11. 54 Van Kesteren, W J P and Tempelaar, M J (1993) Cell Biol Int., 17, 235.

55 Camacho, L., Parton, R., Trewavas, A J., and Malho, R (2000) Protoplasma, 212,162. 56 Plieth, C., Sattelmacher, B., and Hansen, U.-P (1998) Planta, 207, 42.

57 Muhling, K H., Wimmer, M., and Goldbach, H E (1998) Physiol Plant, 102,179. 58 Bush, D S., Biswas, A K., and Jones, R L (1989) Planta, 178, 411.

59 Callaham, D A and Hepler, P K (1991) In Cellular calcium: a practical approach (ed J G. McCormack and P H Cobbold), pp 383-410 Oxford University Press

60 Fabiato, A (1991) In Cellular calcium: a practical approach (ed J G McCormack and P H. Cobbold), pp 159-76 Oxford University Press

61 Bers, D M., Patton, C W., and Nuccitelli, R (1994) Methods Cell Biol., 40, 3. 62 Haugland, R P (1999) Handbook of fluorescent probes and research chemicals, 7th edn

Molecular Probes, Eugene, Oregon

63 Plieth, C., Sattelmacher, B., and Hansen, U.-P (1997) Protoplasma, 198,107. 64 Kao, J P Y (1994) Methods Cell Biol., 40,155.

65 Feijo,J A., Sainhas, J., Hackett, G R., Kunkel, J G., and Hepler, P K (1999) J Cell Biol.,

144,483.

66 Plieth, C., Sattelmacher, B., Hansen, U.-P., andThiel, G (1998) Plant]., 13,167. 67 Felle, H H and Hepler, P K (1997) Plant Physiol, 114, 39.

68 Speksnijder, J E., Miller, A L, Weisenseel, M H., Chen, T H., and Jaffe, L F (1989) Proc Natl Acad Sci USA, 86, 6607.

69 Pierson, E S., Miller, D D., Callaham, D A., Shipley, A M., Rivers, B A., Cresti, M., et al (1994) Plant Cell, 6,1815

70 Knight, H and Knight, M R (1995) Methods Cell Biol., 49, 201.

71 Knight, H., Trewavas, A J., and Knight, M R (1997) In Plant molecular biology manual (ed S B Gelvin and Schilperoort), C4,1 Kluwer Academic Publishers

72 Messerli, M and Robinson, K R (1997)J CellSci., 110,1269. 73 Scott, A C and Allen, N S (1999) Plant Physio!., 121,1291.

74 Bibikova, T N., Jacob, T., Dahse, L, and Gilroy, S (1998) Development, 125, 2925. 75 Roos, W., Evers, S., Hieke, M., Tschope, M., and Schumann, B (1998) Plant Physiol., 118,

349

76 Parton, R M., Fischer, S., Malho, R., Papasouliotis, O., Jelitto, T C., Leonard, T., eta! (1997)J Cell Sci., 110,1187.

77 Messerli, M A and Robinson, K R (1998) Plant J., 16, 87

78 Muhling, K.-H., Plieth, C., Hansen, U.-P., and Sattelmacher, B (1995) J Exp Bot., 46, 377

81 Pfanz, H and Dietz, K.-J (1987) J Plant Physiol, 129, 41 82 Brauer, D., Otto, J., andTu, S.-I (1995) J Plant Physiol, 145, 57 83 Pheasant, D J and Hepler, P K (1987) Ew.J Cell Biol, 43,10.

84 Brauer, D., Uknalis, J., Triana, R., and Tu, S.-I (1997) Plant Physiol Biochem., 35, 31. 85 Franchisse, J.-M., Johannes, E., and Felle, H H (1988) Biochim Biophys Acta, 938,199. 86 Reid, R J and Whittington, J (1989)J Exp Bot., 40, 883.

87 Lindberg, S (1995) Planto, 195, 525.

88 Miihling, K and Sattelmacher, B (1997).J Exp Bot, 48,1609.

89 Browne, B A., McColl, J G., and Driscoll, C T (1990) J Environ Qua!., 19, 65. 90 Vitorello, V A and Haug, A (1997) Plant Sci., 122, 35.

91 Coleman, J O D., Randall, R., and Blake-Kalff, M M A (1997) Plant Cell Env., 20, 449. 92 Flicker, M D., May, M., Meyer, A J., Sheard, N., and White, N S (2000) J Microsc., 198,

162

93 Meyer, A J and Flicker, M D (2000) J Microsc,, 198,174. 94 Allan, A C and Ruhr, R (1997) Plant Cell, 9,1559. 95 Knox, P and Dodge, A D (1984) Plant Sci Lett., 37, 3.

Chapter 3

Flow cytometry

Teodoro Coba de la Pena

Departamento Bioloxia vexetal e Ciencias Solo, Facultade de Ciencias, Universidade de Vigo, 36200 Vigo, Spain

Spencer Brown

Institut des Sciences Vegetales, Laboratoire de Cytometrie, CNRS, 91198 Gif-sur-Yvette, France

1 Introduction

Single cell biochemistry, single cell screening, and assessment of heterogeneity in cell populations have become feasible in both fundamental and industrial situations These strategies have benefited from the following happy marriage On one hand we have development of thousands of fluorescent markers for probing cellular functions, physiological states, or gene expression (Chapters 2, 4, 5), to which one must add the diagnostic tools of in situ hybridization (Chapter 12) and immunolabelling (Chapters 10, 11) On the other hand we have the numerical power, simplicity, and sensitivity of cytometry This term conveys cell-by-cell quantification of structure or, typically, of fluorescent signals This may be achieved by quantitative microscopy—or by flow cytometry, in which case no image is obtained but fluorescent intensities are measured with remarkable ease and precision There is not space to handle cell physiological studies here, although vacuolar properties (1), pH (2), relative membrane potential (3), and membrane viscosity (4) can be assessed by cytometry

Scanning cytometry involves optical devices scanning a slide or filter for several parameters, usually based on fluorescence, thereby building up popula-tion statistics Many workstapopula-tions then offer the possibility of revisiting given 'events' (e.g cells) with a mark-and-find tool for an interactive examination of their morphology or labelling

Obviously, the major constraint of this technique in plant sciences is that it does not handle tissues, but only individual objects However, the following protocols demonstrate situations where simplicity plus sampling power make flow cytometry an efficient complement to imagery, and where heterogeneity can be handled rather than being compressed into average values of bulk prep-arations Our protocols concern nuclei since cell cycle analyses and ploidy screening have become widespread Although used at low concentrations, many of the reagents mentioned are toxic or mutagenic They should be disposed of formally Likewise, a laser accident potentially produces blindness No reflective object should be left near or above a laser beam

Cytometers can be large and expensive sorters, or small benchtop analysers (FAGS, fluorescence activated cell sorter, is a proprietary term that should not replace cytometer.) The technical options must be known in order to choose a protocol Notably, the light sources may be lasers, with or without ultraviolet (see also Chapter for laser descriptions), or inexpensive mercury lamps The familiar parameters are forward angle light scatter (FALS) or right angle light scatter (RALS) which empirically define the size, morphology, or granularity of an object (cell, organelle, debris, etc.) as it flashes past the light beam, and three to four fluorescent signals associated with these

Signals may be amplified and converted to intensity values in arbitrary units using both linear and logarithmic scales, the latter being particularly useful in the case of polyploid series or polysomaty (Section 3.1), or for immunofluorescence where negatives, dims, and brights occur together Data will be represented as monoparametric histograms (one parameter versus frequency), biparametric histograms (also termed cytograms), and multiparametric cluster analyses (usually clarified by colour coding) Data are normally also stored in listmode (a huge spreadsheet of values cell-by-cell) so that subpopulations can be defined a

posteriori with gates (or bitmaps, windows on biparametric graphs, figure 1) and

data can be replayed as often as necessary with different definitions, even once the sample is finished This is 'virtual purification' Pulse compensation is an electronic module allowing the cross-talk between channels due to fluorescence spectral overlap to be corrected, object by object

2 Cytometry demonstrated through cell cycle analyses

Figure Atypical example of virtual purification in cytometry, using gating and a bitmap to

target the plant nuclei present in a crude suspension The resultant monoparametric DNA histograms are clean

a method for distinguishing imbricated cell cycles in plant material is presented in detail, namely Lhe fluorescence1 Hoe-chst-quenching analysis of bromo-deoxyuridine (BrdU) incorporation The reader should consult standard texts (5, 6) on general flow cytometry, as well as the machine manuals Other references explain flow cytometry with plant material (7), plant organdies (8), or plant nuclei (9,10) for genome size analysis or biotechnology (11)

2,1 How to understand monoparametric DNA histograms

Figure Monoparametric DNA histograms of nuclei from a lucerne cell culture, (a) The curve has been deconvolved with software to establish the frequency of nuclei in the compartments Gl, S, G2, as shown However, all cells are not necessarily moving through the given compartments, and the G2 position (at 4C DNA) in fact may contain Gl disomatic cells if there is mixoploidy in the culture Another possible artefact is a shoulder at early S ('SI') if nuclei are incompletely

isolated, (b) After release from a chemical synchronization with hydroxyurea for 30 h, a major cohort (88%) of cells are moving through S (left) and several hours later (right) 83% are in G2.

The reliability of cytometry can normally be appreciated from several factors

(a) First, the tightness of the dominant Gaussian distributions reflects the im-precision of isolation, staining, and measure Their standard deviations (<r, with mean m) in channel units are less useful than their relative dispersion noted as coefficients of variation (CV), where:

An excellent analysis will have CV = 1-2%; 3% can be routine This CV says nothing about the reproducibility of the sample nor its representativity

(b) In practice, symmetry of these peaks is critical Notably, instability of the cytometer or poorly isolated nuclei with occasional attached cytoplasm may produce a shoulder at 2C (right-skewed) that is statistically indistinguishable from the early S compartment, SI Unstable 4C nuclei will produce a left-skewed 4C peak (kurtosis) that artefactually raises the S phase calculation

(c) Another factor is the linearity of staining: the ratio of the observed 4C position relative to the 2C should theoretically be 2.0 but may fall to 1.90 without compromising the estimation of cycle phases

(d) Cellular and nuclear fragments contribute baseline noise which can, as a last resort, be subtracted by modelling, but with low confidence for the S phase calculation It is clearly preferable to improve one's cytology

(e) Doublets of 2C nuclei will be assessed as 4C These must be eliminated by better preparation, by filtration, and minimizing centrifugations Fortunately, singlets and doublets are also discriminated by analysis of the form of the electronic pulse in the cytometer, taking fluorescence pulse integral versus pulse height (Figure I).

(f) Mitotic figures may partly be dispersed When the mitotic index is expected to be high, it must be assessed microscopically as well Analysis of whole cells or protoplasts does not suffer from this problem

Effective algorithms (e.g Multicycle, Phoenix Flow Systems, San Diego) are available for deconvoluting histograms to determine components such as Gaussian 2C and 4C subpopulations, the S-fraction, noise subtraction, imbricated aneuploid cycles, or a supplementary Gaussian distribution from a cohort of synchronized cells (Figure 2) The algorithm will also calculate x2 the residue between the final model and the real data However, a low global residue such as a x2 of a few per cent can arise from excellent modelling of the dominant population, Gl, despite unsatisfactory estimation of minor but essential sub-populations like S A report sheet should indicate not only the histogram and the frequencies Gl, S, G2, possibly M; but also the CV, ratio of positions G2/G1, additional characteristics of the S population (e.g its mean and mode), the residual x2, and the inputs chosen for the algorithm (e.g S phase of 0, 1, or orders, smoothing, background subtraction, the proposition that a cohort of synchronized cells is also present in S, etc.) (see ref 11)

The histogram is a static image of a dynamic event In cells growing asyn-chronously, the frequencies of Gl, S, G2 simply reflect the proportion of time spent in these phases Two cultures with very different generation times may have identical proportions of Gl, S, G2-M, and identical histograms Moreover, the histogram may come from both active and quiescent cells Within the elementary phases GO (inactive), Gl, S, G2, M (mitosis) of the proliferative cycle lie further variations (11) such as reversible or irreversible quiescence from GO, persistent S phase, quiescent G2, G2 to Gl switching without M prior to endo-reduplication, apoptosis, etc These are major phenomena in plant tissue and will not be discriminated by the simple monoparametric DNA analysis Con-ditions that recruit cells may initially increase the frequency of phase S; yet con-ditions that accelerate a phase will eventually lessen its duration and hence the proportion of cells found in that phase The sampling times should also be sufficiently close to follow a dynamic event such as a cohort moving through S to G2, back to Gl, and on to another S Clearly this cytometric analysis must be complemented with parameters such as mitotic index, growth curve, packed cell volume, and additional cytological or molecular markers (12) As a routine precaution, we recommend archiving a drop of cells into Carnoy fixative for assessing the mitotic index Alternatively, for immediate observation of nuclei or mitoses, prepare jjig/ml DAPI or Hoechst 33342 in 5% Triton X-100 and add 10 (jil of this to 50 M-l cell culture

2,2 Developing multiparametric DNA histograms and immunofluorescence

Several examples of bivariate analyses address the identification of quiescent cells amongst active populations The key example (Section 2.6; Protocol and Figure 3) uses BrdU incorporation as an analogue of thymidine in DNA synthesis for insight into the kinetics of more or less active populations distinguished by quenching of Hoechst fluorescence In a heavier classical approach, cells that have incorporated BrdU can be identified by microscopy or cytometry with im-munofluorescence using anti-BrdU antibodies (13) A bivariate analysis gives both the BrdU labelling index and the distribution of the nuclear DNA content with DAPI so that, for instance, late replicating DNA or endoreduplicating cells may be identified Reactivation of nuclei at all levels of ploidy can also be simply followed by assessing total nuclear proteins with sulforhodamine 101 versus DNA staining (14, 15), or, nuclear RNA versus DNA with Acridine Orange which reveals subcompartments (G1A, GIB, S, G2) according to RNA levels, distinguish-ing quiescent (GO), apoptotic, and less active cells (10,16)

fixation, and permcabilization, or at least 'squared protoplasts' (with partially re-duced cell walls) Follow the protocols for immunostaining of whole protoplasts in suspension (see Chapter 10)

2,3 Extracting intact plant nuclei

In general, the nucleus is best isolated from the cytoplasm before fixation to re-duce non-specific staining or interfere nee with the measure The major options presented in Protocol i are chopping (11, 18) (by far the most common), proto-plast lysis (19), or squash of formaldehyde-fixed tissue (20) Precision is inferior (CV > 7%) wirh fixed protoplasts or cells compared to isolated nuclei (CV 2-4%) Leaves or other plant material may be stored unfixed at 4"C in humidified paper within plastic envelopes Although less satisfactory, tissue may be frozen for storage, then subsequently chopped for immediate analysis Isolated nuclei may be lightly fixed and despatched at ambient temperature

No single isolation buffer is universally optimal (Protocol I) Although the two buffers from Galbraith (18) and Marie (11) seem adequate for the majority of plant materials, Dolezel's (9) is key for some situations (e.g Frugaria spp.) The isolated nucleus must be stabilized (with osmoticum and chelators), with or without mild fixation (e.g 1% formaldehyde) A surfactant (usually Triton X-100) will reduce adhesion of cellular debris, disperse chlorophyll from chloroplasts, and reduce non-specific binding of lipophilic dyes Adding extra Triton, 0.5% (w/v) or even 1%, may improve cell cycle resolution or nuclear extraction for certain material The morphology of nuclei is changed by the buffer Citrate will condense DNA Where browning occurs, it may be minimized with a reductanl or protectant against polyphenols For some material, the buffering concentration may need to be doubled: when problems arise, check the pH of the preparation

General toolkit for isolation of diverse nuclei for cytometric analysis

Equipment and reagents

Alternative Extraction buffers*

• Dolezel Nuclear Buffer: 20 ntn Nad, 80 mM KC1, 20 mM MgSO4, mM EDTA.Na2,

0.5 mM spermine.HCl, 15 mM Tris pH 7.5, 0.1% (w/v) Triton X-100.15mM

p-mercaptoethanol (1 ^1/ml) added daily3

• Galbraith Nuclear Buffer: 45 mM MgCl2,

30 mM sodium citrate 20 mM

4-morpholinepropane sulfonate pH 7.0, 0.1% (w/v) Triton X-100

Marie Nuclear Buffer: ISmMNaCl 15 mM KC1, 50 mM sodium citrate, mM

EDTA.Na2, 50 mM glucose, 50 mM Hepes

pH 7,2, 0,5%(w/v)Tween 20, mM metabisulfate addedfor4 h Sgorbati Nuclear Buffer (19) only for prefixed material: 100 mM NaCl, 10 mM EDTA.Na2,10raMTrispH7.4, 0.1%(w/v)

Protectants

• M sodium metabisulfite (syn pyrosulfite, Na2S205 Mr 190) is frozen in I ml aliquots.

Added freshly to solutions at 5-10 mM, this antioxidant is active for about h but can progressively acidify poorly buffered solutions;

and/or,

• 10% (w/v) aqueous stock polyvinyi pyrrolidone (-10000Mr Sigma P6755),

autoclaved and stored at 4°C, To reduce browning induced by phenols and polyphenols, this may be added to solutions at 1% (w/v) final concentration,

Nuclear stains (see Section 2.4) Supplementary material

• 10% (w/v) Triton X-100 stock, autoclaved and stored at4"C

• Sorbitol M stock (autoclaved 110°C, min} Using this sugar alcohol, one avoids the crystals associated with rnannitol which may hinder cytometry and microscopy

• Formaldehyde EM grade 38% (w/v) Store this at room temperature, avoiding direct sunlight Do not disturb any sediment of

polymer that may form (accelerated by cold) and discard the stock once this sediment becomes important as both the fixation and buffering will be severely compromised Purchase small stocks Ribonuclease A Use DNase-free RNase (Boehringer 109169,50 Units/nig) Prepare Tris saline buffer: 12 g/1 Tris and g/1 NaCl (final concentrations 100 mM), adjusted to pH 7.6 with HC1 Prepare a 1% (w/v) RNase stock in this Tris buffer, boil for 20 to inactivate any residua! DNase, aliquot and freeze

Double-edge razor blades (Gillette blue) Filtrettes; cut 20 x 20 mm squares of nylon filter (Nylon Scrynel NYHC; 10, 30,48, 75, 100 or 200 ixm pore size from ZBF, 8803-Ruschlikon, Switzerland or PoJyLabo, Paris), Cut across a i m ! blue pipette tip (e.g Treff) in two places: 27 mm from its tip (discard this extremity) and mm from the top Using the latter as a collar, fix the nylon filter onto the central section In most situations the assembly may be washed and re-used Larger filtrettes may be prepared with ml pipette cones

A Chopping method

1 Place a 1-2 cm2 leaf (avoiding major vascularisation) or about 150 mg callus, roots,

stems, apices, anthers, cellsb, etc on a 90 mm plastic or glass Petri dish Although

unnecessary for cell cycle analysis, when using plant tissue as an internal standard (e.g Petunia, wheat}, add a smaller quantity of its leaf at this stage

2 Chop with 500 ^1 ice-cold nuclear buffer/ We routinely use Galbraith buffer Chop with a fine double-edge razor blade, limiting dispersion or drying Work

quickly: 30 sec should suffice If using cell suspensions as an internal standard (e.g chicken red blood cells), add \j.\ here: not after staining!

4 Transfer this to a 30n.m filtrette pre-moistened with buffer, in an ice-cold cytometer sample tube Remove filter

5 Add 1% (w/v) formaldehyde to stabilise unless the cytometric analysis follows imme-diately

7 Add DNA stain Keeping samples in ice, wait at least for Hoechst, 20 for Mithramycin or Chromomycin A3,15 for intercalating stains

8 Measure at room temperature as detailed under Cytometry (Section 2,5)

B Protobtast lysis

1 Pipette 200 Ul of about 10° protoplasts/ml into a tube.d

2 If the cell wall has re-formed, a hypotonic pre-treatment may be required: add up to an equal volume of water Nuclei may be isolated as follows from protoblast-derived petunia cells up till 72 h culture,"

3 Add an equal volume of cold Galbraith buffer at 0.5% (w/v) Triton X-100 Mix by pipetting three times

4 After min, again pipette three times Filter through 30 or 48 uM nylon

6 Stabilise with 1% (w/v) formaldehyde and store at °C Add stain and measure as detailed under Section 2,5

C, Squash'

1 Fix material in ice-cold 4% (w/v) formaldehyde in Sgorbati Nuclear Buffer for 10 min at 4°C If using a standard (wheat embryos, chicken Petunia leaf, etc.), also fix this tissue in the same vial

2 Rinse twice, at least 10 min, with nuclear buffer

3 Crush tissue in cold 500 ul nuclear buffer with a rough-ended glass rod on a glass Petri dish or in a plastic multiwell plate

4 Filter, stain and read

"For health reasons, p-mercaptoethanol is no longer recommended,

bIf using this chopping method on cell suspensions, these should first be centrifuged, resus-pended for rinsing once in fresh medium, centrifuged again, the wet pellet shifted by spatula to an equivalent volume (300 ul) of cold nuclear isolation buffer with 0.5% Triton, chopped with a new razor-blade, prefiltered through 100 um large filtrette taking 200 u1 extra buffer to rinse, then through a 30 um nylon standard filtrette, and stabilised with 1% formaldehyde Chopping to fixation should take less than

c'The same approach can be adapted to smaller amounts of material For instance, 10 root apices may be chopped in 100 ul buffer This is then taken up in 200 u1 for rinsing, etc

dWhen studying the cycle in cell suspensions, one must be aware that DNA synthesis may continue during protoblast preparation although further cells apparently not cross the Gl/S transition Therefore, the cellulase incubation should be as short as possible and the conse-quence of the duration of cellulase incubation should be checked

(So that the nuclei are isolated, it is essential that the protoplasts burst outwards rather than collapsing inwards

2.4 Which DNA fluorochrome is appropriate?

Factors influencing choice of a DNA fluorochrome are resolution, stability, in-cubation time, excitation wavelengths available on the cytometer, compatibility with other simultaneous staining, cost, and DNA stoichiometry In whole cells, cytoplasmic DNA will also be labelled: for instance, the AT-specific dyes have a high quantum yield with mitochondrial and plastidial DNA All the common dyes should be treated as health risks The following examples are traditional Many new dyes (Chapter 2) are interesting for their spectral properties, their contrast and brilliance, and for penetration into tissue (SYTO series, and the homo- or hetero-dimeric cyanines YOYO, YOPRO, TOTO, from Molecular Probes Inc.)

The Hoechst bisbenzimides 33342 (Mr 616), the most common, or 33258 (Mr 624) require UV excitation at 365 nm, with blue emission 455 nm They are AT-specific, inexpensive, and staining is readable within They should be used at (ug/ml, and higher concentrations for species with high nuclear DNA content (2C > pg DNA) Aqueous stocks of mg/ml may be aliquoted, stored frozen, or kept for weeks at 4°C Many laser-based cytometers cannot operate in UV whereas the cheaper lamp-based machines give excellent results in this analysis (see examples in ref 11) DAPI (4',6-diamidino-2-phenylindole; Mr 350) has similar qualities, with excitation at 340 nm, emission 465 nm, and is more resistant to photo-quenching in microscopy

Mithramycin (Mr 1085) and Chromomycin A3 (Mr 1183) have excitation 420-450 nm, emission 560 nm Used at 50-100 ug/ml, they are GC-specific anti-biotics, require 20 incubation, and high magnesium such as 45 mM in Galbraith's original buffer (18) Stocks may be prepared as mg/ml in M MgCl2 and stored frozen

Propidium iodide (Mr 668) with excitation 535 nm, emission 617 nm and ethidium bromide (Mr 394) with excitation 518 nm, emission 605 nm, are inter-calating dyes having broad excitation bands in UV and blue/green, compatible with both the UV and visible excitation of all flow cytometers Aqueous stocks at mg/ml may be stored frozen Nuclear staining is stable within 10 but not after several hours For dual analysis of DNA versus fluorescein-labelled anti-bodies, propidium iodide is preferred due to its longer emission wavelength As these dyes also intercalate with double-stranded RNA, a 20 RNase treatment may be necessary (Protocol 1): for plant nuclei, U/ml is adequate, less than one-fifth that indicated for mammalian cells

2.5 Running and reading the cytometer

(arbitrary units) A series of equidistant peaks (22) will indicate each population of nuclei corresponding to ploidy increments 2C, 4C, 8C, 16C, etc (see Section 3.1) When a cytometer does not enable simultaneous log/linear amplification from one photomultiplier, it may be possible to split the emission with a mirror in order to use two photomultipliers, or to split the photomultiplier output with a T junction so as to process two signals in parallel

Use forward angle light scatter (FALS) to identify isolated nuclei and to gate out debris, and pulse analysis to eliminate doublets In summary, a typical proto-col of monoparametric histograms and biparametric cytograms (Figure 1) would

be:

(a) Accumulate data for 10 000 to 20 000 nuclei (~ min) counted on Histogram

(b) Cytogram 1: FALS versus log DNA fluorescence—set a generous gate around nuclei

(c) Cytogram 2: fluorescence pulse area versus height—to gate out doublets and debris

(d) Histogram 3: linear DNA fluorescence gated on cytograms and 2—the key result

(e) Histogram 4: log DNA fluorescence gated on cytograms and 2—the over-view

(f) Histogram (not shown): log DNA fluorescence gated only on cytogram 1—to survey the effect of doublet elimination

Where feasible, use band-pass optical filters to avoid extraneous fluorescence and notably that from chlorophyll whose emission is above 670 nm, although Triton usually has dispersed chloroplasts To avoid clumps, refilter stored samples through 30 |j,m filtrettes just before cytometry Every effort is necessary to optimize protocols and stability of the cytometer, and to resolve ambiguities on the spot prior to data processing with software For instance, it is essential to identify whether an asymmetric 2C peak or a split peak is an artefact Is it re-producible? Does re-running a control sample produce the same phenomenon? Are the nuclei stable according to the cytograms? Under the microscope, are nuclei unstable or attached to debris?

2.6 BrdU incorporation to identify DNA synthesis by fluorescence quenching

Figures BrclU incorporation to identify DNA synthesis by fluorescence quenching (Protocol 2). Cytograms of nuclei isolated from lucerne A2 suspension cultures < BrdU, double stained with Hoechst 33258 and propidium iodide (PI), (a) Control culture without BrdU All nuclei, active or not, lie on a single correlation of the two nuclear stains, (b) Synchronized cells to which BrdU has been added for h upon release from a chemical block: active nuclei increase their PI staining much more than their Hoechst staining, moving through S* to G2* A trace of inactive 4C° ceils and many inactive 2C° are evident, (c) Asynchronous cells pulsed with BrdU for h with the typical curvilinear population of labelled S" nuclei between 2C and 4C, some that have reached G2*r and others that, having been in Swhen the nucleotide was added, form a

continuum back to 4C° (d and e) 30 h treatments of slow cultures without auxin (d) or upon its addition (e) The growth regulator has pushed many more cells through to the position G2* and then, with cytokinesis, to Gl* exactly halfway back to the origin (0,0), The frequencies and probabilities are shown in Table Other experiments have sometimes revealed a second round of synthesis, towards G2+ + , and even a subpopulation at ' - t - A ' (in e) which turned out to

be apoptotic nuclei, (ft Using logarithmic amplification, the some processes can be observed in mixoploid cultures where monosomatic and disomatic cells are distinct In some cases not only HO but afso PI intensity decreases in subsequent generations This does not compromise the analysis The essential aspect is that the slope PI/HO must increase with each generation; these slopes are shown

and relative fluorescence quenching (loss of HO intensity) due to DNA synthesis in presence of BrdU

Table Proliferation compartments identified by BrdU in Figures 3d and 3e, expressed as both frequencies and probabilities3 data line a b c d e f g h i j h Proliferation compartment 2C° 4C° 8C° S*

G2* [ followed by cytokinesis)

Gl*

S+*

* > 4C

Check total: l(a n)

Total active during 30 h: i (d h) Cytokinetic events: 0.5 x X (f g)

Frequency (%) Auxin starvation 49 25 2.6 5.2 13 0.4 2.0 100 23.2 7,7

+ NAA

30 hours 17 7.4 1.5 4.9 18 44 2.3 4.9 100 74.1 25.6 Probability Auxin starvation 0.531 0.271 0.030 0.028 0.056 0.070 0.002 0.011 1.000 0.167 0.072 + NAA 30 hours 0,228 0.099 0.020 0,066 0.242 0.296 0.015 0.033 1,000 0.652 0.314

"Trie division of some cells will automatically modify the relative frequency of all the other populations, even those whose num-ber has effectively remained constant It is therefore wiser H cytokinesis is occurring, to convert any relative frequency (percent-age) at a given time mto a probability that a cell would, during the pulse, have moved through to such a phase In short, considering the sub populations of Protocol (step 7) with their respective frequencies n after 30 h, the relation with the number of cells initially present at time (N,,) is:

NO = (!„; «;., + n,s 6~ i 0,5 * n,B t., , GS , - 0.25 * n (C: etc [3| This calculation becomes more elaborate if there is also en reduplication orapoptosis

BrdU incorporation to identify DMA synthesis by fluorescence quenching

Equipment and reagents

• Multiparameter flow cytometer with UV source

• Filtrettes 48 JJUTL (Protocol!)

• Proliferating cell culture." protoplasts,

tissue in vitro QT in planta

• Bromo-deoxyuridine (Mr 307), freshly prepared 15 mM aqueous stock (4.6 nig/ml) If the culture will exceed 12 h sterilize the BrdUby filtration It is critical that this solution be less than 48 h old

• Cleaning fluid for the cytometer: Coulter Clenz (Beckman-Coulter)

l%(w/v)hypochlorite 70% (v/v) ethanol

Galbraith nuclear buffer at 0.5% (w/v) Triton X-100 (Protocol 1)

100 n.g/ml aqueous bisbenzimide Hoechst 33258

100 |Ag/ml aqueous propidium iodide RNase (Protocol1)

Method

1 Add 30 uM BrdU to cultures: time zero The duration of the pulse (4-30 h) will de-pend upon the experimental objective, but may cover several cycles.1'

2 Isolate the nuclei either via rapid production of protoplasts (Protocol IB) or by chopping (Protocol 1A) Filter, and stabilize with 1% (w/v) formaldehydec

3 Keeping samples over ice, add U/ml RNase and M-g/ml Hoechst 33258 (HO) for 10 min; then add n.g/ml propidium iodide (PI) for at least 10 (stock solutions are weak in order to facilitate precision) Protect from strong light

4 To ensure that competing fluorochromes are not present, the sample line is initially rinsed for 30 sec with 1% hypochlorite, immediately chased with water, then 70% ethanol for min, and water again for Then rinse with Coulter Clenz or equivalent

5 Before each sample, either the line is stabilized for 30 sec with nuclear isolation buffer containing ug/ml HO and ng/nil PI or, if quantity permits, the sample is left to run for equilibration (The PI is one-tenth normal saturating concen-tration.)

6 Set up the cytometer for doublet discrimination, listmode and using both linear and logarithmic amplifiers in the case of endoreduplicating cells It is essential to generate a biparametric histogram of blue fluorescence (x: 400 nm < HO < 500 nm) versus red (y: PI > 610 nm), preferably on log as well as linear scales The amplifiers of HO and PI are set to give almost equivalent signals so that their slope (y/x) is ~ 0.8 BrdU-free nuclei then lie on an almost diagonal line on the HO versus PI cytogram The listmode data may be replayed to assess the number of cells in each clustered

subpopulation We propose terminology which distinguishes by asterisks the num-ber of rounds of DNA synthesis undertaken or completed: 2C°, 4C°, 8C° means inactive; S*, C2*, Gl' first generation; S**, G2**, Gl** second generation Numbers in these 'proliferation compartments' may either be expressed as frequencies or may be converted to probabilities that a given cell at time zero would progress to each phase (Table 1).

8 Routine control samples without BrdU must give a tight single correlation between the two fluorescent emissions Run these early and late in a series so as to quantify the background of false positives (generally several per cent of nuclei)

aCell suspensions need to be rinsed in cell culture medium to reduce degradative activities in their extracellular medium

N To learn the method and the inteipretation, preliminary tests with drugs added to block at mitosis (with oryzalin, propyzamide) or at Gl (with olomoucine) (24)

3 The particular application of genome size calculation and 'DNA ploidy'

This major research and industrial application of flow cytometry is extensively treated elsewhere (9-11, 25) The calculation of base composition (Section 3.3) is less common but simple

3.1 Terminology

The number of chromosomes present in the gametes without replication, con-stituting the haploid complement, is represented as n The basic chromosome number for a given taxon is represented as x, a unit of ploidy; this may or may not be respected in a given specimen or species, depending upon evolutionary loss or gain of chromosomes or short-term ecological variation Ploidy is strictly defined by cytogenetic description of the karyotype as the number of copies of the chromosome complement for that species The quantity of DNA correspond-ing to the haploid complement is represented as C, in arbitrary or in absolute units such as picograms or base pairs (Initially, C derived from the notion of Constant quantity, but was subsequently modified by Swift to Complement quantity.)

The simple flow cytometric assay of interphasic nuclei (below) yields an esti-mate of DNA quantity, 2C, rather than chromosome number, 2n This value C is compared to that of a reference plant of known ploidy (true C) or to published values of C for the given species Therefore, the DNA ploidy reported from cyto-metry equates a constant DNA quantity with a complete chromosome comple-ment This certainly may not always be true (e.g an aneuploid specimen having lost one chromosome but gaining another), yet it is a highly advantageous simplification The terminology 'DNA ploidy' is used to respect this logic We rarely use 'DNA index' between plants as this term gives too much credence to the notion of a fixed value, C, for a species whereas interplant variation can indeed be significant

Most plant families display developmentally regulated endoreduplication so that tissue is polysomatic (26) However, a plant that is genetically diploid will in its leaves show at least some diploid cells, although may be only a trace

3.2 Internal or external standards

The fluorescence intensities indicated on a histogram are normally in arbitrary units of channel numbers derived from photomultiplier outputs In order to assess genome size or ploidy of an unknown, this scale must be calibrated with a reference The simplest calibration is to analyse a reference plant of known ploidy (e.g 2x) of the same species, note the position of its Gl nuclei (2C), and characterize all other samples by the relative position of their Gl peak How-ever, an internal reference within the preparation is more reliable: for instance, use of a tetraploid reference to screen for diploids

Table Plants convenient for regular standard leaf tissue (11) Common namea Arabidopsis Barrel medic Tomato Petunia Lucerne A2 Sweet pea Wheat

"These are diploid i "Note that pg D composition of 42."

Specific name (GC%)

Arabidopsis thaliana L Heynh ecotype

Bensheim

Medicago truncatula Gaertn cv R108-1 Lycopersicon esculentum Mill, cv Roma Petunia hybrids (Hort.) PxPc6

Medicago saliva L subsp x varia (Martyn)

Arcangeli cv Rambler A2

Pisum sativum L cv Express long Triticum aestivum L cv Chinese Spring

except tetraploid lucerne and hexaploid wheat iNA = ~ 965 Mbp The 2C value of chicken is

7% GC (11).

2C DNA (pg)A

0.33 0.98 1.99 2.85 "3.47 8.37 30.9

taken as 2.33 pg Base composition 40.3 38.1 40.0 41.0 38.7 40.5 ""43.7

(18) with a base

Figure The genome size of a plant can be simply assessed by flow cytometry using DNA

fluorochromes and the ratio, R, of nuclear fluorescence between the unknown (here, wheat) and an internal standard such as pea (2C = 8.37 pg, 40.5% GC) However, intercalating dyes (such as propidium iodide) must be used For the ratio will be fluorochrome-dependent if base-specific dyes (like Hoechst or chromomycin) are used and if the base compositions of the two genomes differ, as in this overlay of three separate analyses of wheat (43.7% GC) with pea Fortunately, as shown in the exercise at right, the relationship between ratios can be used to deduce the base composition of the unknown (wheat) relative to a standard, pea (Section 3.3)

plants of known genome size (Table 2) or chicken erythrocytes (10) are included as internal standards (Figure 4; and examples in ref 11) They must be fixed and stained along with the unknown The DNA quantity is then deduced from the fluorescence ratio, R, of the two species of nuclei

3.3 Calculating base composition

increase in fluorescence intensity This is irrelevant in cell cycle analysis of a given genome Likewise, for DNA ploidy comparisons within a given genus, base composition may be ignored

It is obvious that genome size evaluation by fluorometry will be distorted if a specimen genome is calibrated with a reference species whose DNA has a different base composition Fortunately, this observation can be used to estimate the base composition of genomes of higher plants (11, 27, 28) and algae (29) Comparing the experimental ratio, R, of 2C intensities for an unknown speci-men and a known standard such as petunia, observed with Hoechst (HO), Chromomycin (CH; alternatively, Mithramycin), and an intercalating dye propidium iodide (PI; alternatively, ethidium bromide):

for

RPI = Intensityspecimen / Intensitystandard

and similarly RHO and RCH

then

AT%specimen = AT%standard *(RHO/RPI)

and

GC%specimen = GC%standard * (RCH/RPI)

4 Sorting of protoplasts and cellular organelles

Many flow cytometers allow physical isolation of objects being analysed, accord-ing to their fluorescence and light scatteraccord-ing In plant biology this includes separating protoplasts of different cellular types, such as those rich in particular secondary products (30) or expressing GFP (31), as a prelude to their culture or biochemical and microscopic analysis It also includes pollen (5, 7), chromo-somes (32, 33), and nuclei (3, 34)

and the use of anticoincidence (where, to ensure purity, a target object will be rejected because of the proximity of another in the stream)

Other systems exist where sorting is done by deviating the stream with a wavefront or with a catching cup The material will then be recovered with high dilution,

Protocol explains a simple sort used to obtain nuclei from different phases of the cell cycle, as LI way of getting pre-messenger RNA, viral DNA, or proteins from specific nuclei (34-36) In a day's work several million nuclei may be recovered from various points of the cycle, defined us a series of windows set on fluorescence correlated with nuclear DNA content

Sorting subclasses of plant nuclei

Equipment and reagents

• Flow cytometer with sorting module • Epifluorescent microscope, microscope

slides, and non-fluorescence cleaning paper (Kimwipes, Kimberly-Clark) • Black cellulose" 13 mm filters (Millipore

HABP 01300, 0.45 um) as a non-fluorescent antireflex support for epifluorescence observation of sorted material • Coverslips, 20 x 20 mm, handled with

tweezers

• 13-15 mm round or square nylon nucleic acid hybridization filters, generally white • Photographic films: 400 ASA TMAX (Kodak)

black and white; 400 or 1000 ASA Fujichromo for slides; 1600 ASA Fujicolor for colour prints

• RNase-free labware

• Potri dishes or similar clean recipients for posing objects such as tweezers

• DEPC-treated 30 or 48 um mesh nylon filtrettes (Protocol 1)

Cleaning fluid for the cytometer: Coulter Clenz (Beckman-Coulter)

1% hypochlorite bleach 70% ethanol

Water treated overnight with 0.1% diethyl pyrocarbonate (DEPC) then autoclaved Nucleic acid specific fluorescent dyes (Section 2.4)

Sheath fluid (2 litres per day): 75 mM KC1, mM EDTA,Na2,100 mM sorbitol, mM Hepes pH 7, treated overnight with 0,1% DEPC then autoclaved The cytometer will filter this in-line

TH buffer: 10 mMTris-HCl, mM Na2EDTA pHS

TLES buffer: 50 mMTris-HCl pH 9,150 mM LiCl, mM EDTA-Na2, 5% SDS

Plant nuclei in suspension, as dense as possible—up to x 106 nuclei/ml (Protocol 1). Filter through 30 um nylon (48 um if there are large polyploid nuclei) Keep in ice

Method

1 Warm-up, clean, and align the flow cytometer with the specific sheath fluid Run test sorts with fluorescent polystyrene alignment beads: recovery should be

3 Run a conventional cytoraetric cell cycle analysis (Section 2), This must be stable and precise, with a tight coefficient of variation (2-4%) on the GO-G1 peak Consider the target nuclear classes: what are their frequencies relative to all nuclei and relative to the rate of all objects analysed per second? Working at 2000 objects analysed/ second, what will be the sort rate? What quantity is needed? How long will this take?

4 Set sorting gates on the Gl peak, including fluorescence and light scatter para-meters Set anticoincidence and droplet-deflection number, affecting purity versus recovery Sort 50 nuclei onto a slide for immediate assessment on a microscope: recovery should be > 90%

5 Set up the sorting gates on the pertinent zones (Figure 1)

6 Sorted nuclei may be collected in Eppendorf tubes containing 500 ul TIES buffer and frozen at -80°C prior to nucleic acid extraction Most cytometers produce

~ 100-300 X 105 sorted objects/ml recovered volume

7 Alternatively, sorted nuclei may be captured on 13 mm black cellulose filters pre-moistened with TE buffer and maintained on a permeable filter support (Swinnex 13) linked to a syringe for light aspiration so that excess sheath fluid may be gently remove as sorting progresses Sorted nuclei then form a tight spot, ~ mm for 105 nuclei The filter is re-moistened with fluorochrome in TE prior to microscopy

8 Using the same Swinnex syringe set-up, nuclei may be directly and quantitatively sorted onto white hybridization filters (which are, however, poor for microscopy) Always handle filters with clean tweezers Several spots may be sorted per filter, marking these with a lead pencil Visualize spots under the low power objective of a fluorescence microscope or with a UV lamp

9 Sort as fast as possible (typically 1000-2000 objects/sec) Verify that the sort recovery rate is as expected (from step above) and occasionally verify the sort efficiency by taking 50 nuclei on a microscope slide

10 Sort appropriate controls, e.g the whole set of nuclei For extraction or blotting,

another control involves processing an equivalent aliquot of unsorted nuclei (that have never seen the flow cytometer)

" These black cellulose filters are not compatible with nucleic acid hybridization,

5 Tests for cell viability during functional assays

Cytometric assays of cell function or reactivity are particularly sensitive for revealing cellular heterogeneity Moreover, the flow cytometry returns data on-line so that dozens of experiments may be done in an afternoon A method is available for estimating cytoplasmic pH in plant protoplasts and monitoring rapid changes (2, 38)

Any flow cytometric analysis of living cells (e.g for pH, oxidative burst, for surface antibodies, for mitochondrial potential) should include simultaneous identification of dying or dead cells, so as to exclude these from the population statistics At a low concentration of n-g/im, propidium iodide (Section 2.4) does not permeate living cells or protoplasts, but it can penetrate disrupted mem-branes of dead or dying cells which will give a strong signal on the red photo-multiplier In this way, red fluorescing objects can be «gated out» as dead cells Chlorophyll fluorescence can still be distinguished at longer wavelengths (665-735 nm) Conversely, viable cells can usually be identified with fluorescein diacetate (FDA) or preferably its more stable lactone derivative, carboxyfluor-escein diacetate (CFDA) These permeant esters will subsequently be hydrolysed by intracellular esterases to their weak acid forms of fluorescein (maximum excitation at 492-495 nm, emission ~ 520 nm) and are retained behind the membrane Store as mg/ml stocks in DMSO at -20°C and add |xg/ml 5-10 before analysis (Chapter 9)

6 Conclusion

The plant protocols of this practical chapter have been amply tested In many cases samples may be despatched by post for analysis Routine tools of mam-malian cell biology, cytometers are now generally available They can be half the price of their essential companion, an epifluorescence microscope (Chapter 1) Another companion is a spectrofluorometer, to explore spectra and for bulk fluorometry However, multiparametric quantitative fluorometry is remarkably precise, sensitive, and simple on a flow cytometer, if imaging is not needed and if single cells or organelles are available

Acknowledgements

References

1 Brown, S C., Renaudin, J P., Prevot, C., and Guern, J (1984) Physiol Veg., 22, 541. 2 Giglioli-Guivarc'h, N., Pierre,] N., Vidal, J., and Brown, S C (1996) Cytometry, 23, 241. 3 Bergounioux, C., Brown, S C., and Petit, P X (1992) Physiol Plant, 85, 374.

4 Gantet, P., Hubac, C., and Brown, S C (1990) Plant Physiol, 94, 729.

5 Melamed, M R., Lindmo, T., and Mendeldohn, M L (ed.) (1990) Flow cytometry and sorting, 2nd edn Wiley-Liss, NY.

6 Robinson, J P (ed.) (1997) Current protocols in cytometry John Wiley & Sons, NY. 7 Galbraith, D W (1994) In Methods in cell biology Vol 42, p 539 Academic Press,

London

8 Schroder, W P and Petit, P X (1992) Plant Physio!., 100,1092. 9 Dolezel, J., Binarova, P., and Lucretti, S (1989) Bio! Plant., 31,113.

10 Brown, S C., Bergounioux, C., Tallet, S., and Marie, D (1991) In A Laboratory guide for cellular and molecular plant biology (ed I Negrutiu and G Gharti-Chhetri), p 326. Birkhaiiser, Basel

11 Marie, D and Brown, S C (1993) Bio! Ceil, 78, 41.'

12 Roudier, F., Fedorova, E., Gyorgyey.J., Feher, A., Brown, S., Kondorosi, A., eta! (2000) Plant;., 23, 73

13 Levi, M., Sparvoli, E., Sgorbati, S., and Chiatante, D (1987) Physiol Plant., 71, 68 14 Sangwan, R., Bourgeois, Y., Brown, S C., Vasseur, G., and Sangwan-Norreel, B (1992)

Pionta, 188, 439.

15 Citterio, S., Sgorbati, S., Levi, M., Colombo, B., and Sparvoli, E (1992) J Cell Sri., 102, 71

16 Bergounioux, C., Perennes, C., Brown, S C., and Gadal, P (1988) Cytometry, 9, 84 17 Desikan, R., Hagenbeek, D., Neill, S J., and Rock, C D (1999) FBBS Lett., 456, 257. 18 Galbraith, D W., Harkins, K R., Maddox, J M., Ayres, N M., Sharma, D P., and

Firoozabady, E (1983) Science, 220,1049.

19 Bergounioux, C., Perennes, C., Miege, C., and Gadal, P (1986) Protoplosma, 130,138 20 Sgorbati, S., Levi, M., Sparvoli, E., Trezzi, P., and Lucchini, G (1986) Physiol Plant, 68,

471

21 Cuq, F., Brown, S C., Petitprez, M., and Alibert, G (1995) P!ant Cell Rep., 15,138. 22 Gendreau, E., Hofte, H., Grandjean, 0., Brown, S., andTraas, J (1998) Plant/., 13, 221 23 Kubbies, M., Goller, B., and Bockstaele, D R (1992) Cytometry, 13, 783.

24 Glab, N., Labidi, B., Qin, L.-X., Trehin, C., Bergounioux, C., and Meijer, L (1994) FEBS Lett., 353, 207.

25 Galbraith, D W., Lambert, G M., Macas, J., and Dolezel, J (1997) Current protocols in cytometry, pp 7.6.1-7.6.22 J Wiley & Sons, Inc.

26 Traas, J., Hulskamp, M., Gendreau, E., and Hofte, H (1998) Curr Opin PlantBioL, 1,498 27 Godelle, B., Cartier, D., Marie, D., Brown, S C., and Siljak-Yakovlev, S (1993)

Cytometry, 14, 618.

28 Blondon, F., Marie, D., Brown, S C., and Kondorosi, A (1994) Genome, 37, 264. 29 Le Gall, Y., Brown, S., Marie, D., Mejjad, M., and Kloareg, B (1993) Protoplosmo, 173,

123

30 Bariaud-Fontanel, A., Julien, M., Coutos-Thevenot, P., Brown, S., Courtois, D., and Petiard, V (1988) In Plant eel! biotechnology (ed M S S Pais), p 403 Springer-Verlag, Berlin

31 Galbraith, D W., Anderson, M T., and Herzenberg, L A (1999) In Methods in cell Wology, Vol 58, p 315 Academic Press, London

33 Macas, J., Gualberti, G., Nouzova, M., Samec, P., Lucretti, S., and Dolezel, J (1996)

Chrom Res., 4, 531.

34 Accotto, G P., Mullineaux, P M., Brown, S C., and Marie, D (1993) Virology, 195, 257. 35 Perennes, C., Bergounioux, C., and Gadal, P (1990) Plant CeURep., 8, 684.

36 Pfosser, M., Amon, A., Lelley, T., and Heberle-Bors, E (1995) Cytometry, 21, 387. 37 Esser, C., Gottlinger, C., Kremer, J., Hundeiker, C., and Radbruch, A (1995) Cytometry,

21, 382

38 Giglioli-Guivarc'h, N., Pierre, J.-N., Brown, S., Chollet, R., Vidal,]., and Gadal, P (1996)

Chapter 4

Transient expression, a tool to address questions in plant cell biology

Jane L Hadlington and Jurgen Denecke

Leeds Institute for Plant Biotechnology and Agriculture (LISA), School of Biology, Faculty of Biological Sciences, LC Miall Building, University of Leeds,

Leeds LS2 9JT, UK

1 Introduction

Transient expression in plant systems is a rapidly expanding technique, which has opened up new strategies to approach biological problems without the necessary delay and environmental risks that are associated with the production of stable transformants DNA is introduced into cells via naked DNA transfer or biological vectors such as bacteria or viruses, and gene expression is assayed for a limited time after the initial transfer Some methods allow the introduction of RNA or proteins as well Results can be obtained much faster without extensive resources and time, and in case of cell suspensions or protoplasts, experiments are more reproducible due to the high number of cells and the homogeneous nature of the suspensions In some cases, i.e the analysis of cytotoxins or dominant mutant approaches disrupting essential pathways or mechanisms, transient expression is the only option available, because transgenic plants would not be viable Finally, with the current debate on the safety and necessity of genetically modified organisms, it is important to introduce alternative ways of addressing biological questions in plants In many cases, transient expression has been shown to provide ultimately the same information as approaches using transgenic plants, thus justifying a more consequent and widespread use of the technique

systems used In addition, a set of novel protocols is provided for practical applications

2 Current methods of transient expression

2.1 Naked DNA transfer

The most frequently applied technique of achieving transient gene expression is based on the direct transfer of naked DNA, mostly plasmid based, into plant cells These can be protoplasts prepared from a variety of plant cells or tissues (Protocol 1), but also cell suspension cultures or entire plant tissues Protoplasts are transfected by electroporation (1) (Protocol 2) or chemical methods based on the use of calcium and PEG (2) Excitingly, PEG-based methods using protoplasts also allow DNA intake into chloroplasts and subsequent analysis of reporters (3) Cell suspensions, callus tissue, or tissue explants can be electroporated using the same protocol as for protoplasts after subjecting cells to plasmolysis (4) This allows diffusion of the DNA through the cell wall towards the protoplast, but electroporation conditions can be different from those used with protoplasts

Entire tissues can be transfected by particle bombardment, in which the DNA is usually coated onto gold particles, which are accelerated towards the tissue within a vacuum (5) In all these cases, the majority of the introduced DNA will not be integrated into the genome and associated gene expression will be trans-ient For this reason, position effects will not influence the result, which can be an advantage in promoter studies Particle bombardment will thus be useful in those cases where protoplasts not fully maintain biological processes observed in complete tissues

However, cellular factors associated with the process under investigation could be titrated out due to the high copy numbers of the introduced DNA in the transfected cells The limiting factor could be a DNA binding protein that mod-ulates promoter activity, or a protein sorting receptor, which can be saturated by high numbers of ligands causing mis-sorting On the other hand, the phenom-enon of receptor saturation could be exploited to study receptor functions, as for example by co-expression of receptors to overcome saturation and establish a functional assay for receptor action Researchers should thus be aware of the possible problems as well as the benefits associated with high gene copy num-bers and/or overexpression

2.2 Biological vectors

contain many transformed cells with insertions in different locations of the genome, hence resulting in the absence of a defined position effect However, in contrast to naked DNA transfer, only very few gene copies will be present in the transformed cells, thus avoiding titration problems

Similarly, virus vectors are used to infect entire plants, and infected tissues containing uninfected and infected cells will be analysed as a whole (7) The latter technique has been particularly useful for analytical methods based on microscopy, as the infected cells are easily identified and can then be studied (see Chapter 5) In comparison with naked DNA transfer in protoplasts, bio-logical vectors allow less control over the ratio of transformed to untransformed cells and will be less quantitative and reproducible However, they not require specialized equipment such as an electroporation device and allow work with entire tissues This can be an advantage if tissue specific biological processes are studied that may not be functional in protoplasts

3 Application of transient expression

3,1 Promoter analysis

Since the introduction of transient expression assays in the second half of the last decade using protoplasts (1,2), or later whole tissues (8, 9), the technique has mainly been instrumental to the analysis of promoters A given promoter is usually fused to a coding region of a reporter enzyme, the most popular to date are GUS (10) and firefly luciferase (11) Due to the sensitivity of these enzyme assays and the high degree of reproducibility, indirect information about pro-moter transcription can be obtained with a limited amount of biological material The absence of position effects facilitates the rapid analysis of many chimeric constructs, i.e series of promoter deletions The homogeneous nature of the protoplast suspension also allows the aliquoting to identical portions for the subsequent analysis of the influence of physiological conditions under which the protoplasts are cultured These include the effects of hormones such as gib-berelic acid and abscisic acid (12), high (6) and low temperature (13, 14), or the development of inducible promoters

transcript is directly affected by the various physiological conditions and that the resulting reporter activity does not only represent promoter activity For example, if low temperature is inducing a promoter, but the reporter is synthesized at a lower rate during low temperature, the result underestimates the true transcrip-tional induction This problem can be partially overcome by subjecting a similar reporter fusion (containing a constitutive promoter) to the same stress condition and comparing the result This technique is similar to the use of internal markers where one gene is kept constant but the second gene is subjected to analysis (i.e promoter deletions) However, direct analysis of endogenous transcript levels by Northern blot analysis is always the safest method for studying promoter regulation and can never be completely replaced by transient expression

3.2 Cell biology and biochemistry

In contrast to promoter studies, which are perhaps more frequently analysed in whole tissues, the following applications are mostly based on biochemical tech-niques and are performed in protoplasts or sometimes cell suspensions Basic processes such as protein synthesis and targeting are often controlled by con-served mechanisms, which are fully operational in protoplasts or cell suspen-sions In these cases, transient expression experiments can produce exactly the same results as stable transformants

In the past, stable transformants were generated when large quantities of bio-logical material were required, i.e for cell fractionation or protein purification However, it has become clear that transient expression techniques can easily be scaled up to tackle even cell fractionation This possibility has yet to be exploited by the scientific community and protocols will be provided below to stimulate the use of this methodology Finally, there are a number of cases where proteins are toxic to cells when produced in large amounts or as mutant derivatives Stable plant transformation is then impossible but transient expression provides the solution as it allows transient monitoring of cells prior to cell death

3.2.1 Protein targeting in the secretory pathway

Transient expression in tobacco protoplasts has helped to develop the concept of the default pathway for the secretion of soluble proteins in plants (18) and to analyse endoplasmic reticulum (ER) retention signals via point mutagenesis (19) In both cases, confirmation of these results with transgenic plants did not provide any additional information

expression first and restrict the generation of stable transformants to those that provided an altered phenotype

3.2.2 Protein targeting to other organelles

Transient expression techniques have been successful in chloroplast targeting assays using biochemical fractionation techniques (27) as well as microscopical techniques based on the green fluorescent protein (Chapter 5, and refs 28-30) Mitochondrial targeting assays have also been successful (29, 31, 32) Nuclear targeting of proteins has been studied using a very elegant technique involving transiently transformed onion (Allium cepa) epidermal cells via particle bombard-ment This allowed routine microscopical evaluation of the nuclear or cytoplas-mic location of in-frame carboxyl- and amino-terminal fusions of the reporter enzyme GUS (33) Peroxisomal targeting has yet to be tested using transient expression in plants, but it is likely to be feasible as well

3.2.3 Protein and nucleic acid biochemistry

Recent progress on plant chaperones was made possible by co-transfection of plasmids encoding either the chaperone (34) or the ligand (35) Protein-protein interactions are then monitored by in vivo labelling and co-immunoprecipitation of the two interacting molecules (36) The technique could be further explored to routinely assess protein stability, mRNA maturation, transport, stability, and translation efficiency Basic cellular processes that are conserved in all cell types can thus be studied An example is the analysis of the effect of altered codon usage on protein production (37) In this context it can be mentioned that be-sides plasmid DNA other biomolecules can be introduced into plant cells These include in vitro transcripts as well as proteins, antibodies, or drugs.

A very specialized example of the use of transient expression is the analysis of toxins In this case the ability to synthesize a reporter is used as a measure for cell viability and the potentially toxic proteins are produced via co-transfection Using this technique it was possible to show that the B-domain of ricin prevents toxicity of the A-chain while it is synthesized in the ER, thus protecting the plant ribosomes from the otherwise extreme toxicity of this potent toxin (26) The use of co-transfection is surprisingly efficient and most transfected cells will contain all the plasmids included in the mixture Clearly, the technique has few limita-tions and is very powerful to study protein-protein interaclimita-tions in vivo.

4 Practical considerations for cell biologists

Figure Flow chart of experimental procedures to tost protein localization and sorting by the secretory pathway The numbers within the hexagons refer to the protocols given within this chapter

produced as a precursor which is proteolyrically trimmed?' The transient expression method can then be adapted to provide optimal conditions to study the phenomenon

To study protein-protein interactions, co-transfcction is often desired to reach a higher concent ration of ligands and/or receptors to allow detection Similarly, co-transfection can be used to test if overexpression of a receptor or a chaperone has an influence on the rate of synthesis and subcellular fate of a protein Co-transfeclion can be accomplished simply by mixing two phismids, which allows dosage-response analysis and is more flexible Naked DNA transfer usually leads to the uptake of many copies of plasmids in the transfected cells, and the probability is very high that most transfected cells receive both plasmids, Kven mixtures of three plasmids have been successfully co-trans fee ted (26)

Alternatively, it is possible to use plasmids carrying more than one gene as well as transcripts coding for more than one gene product, finally, it is possible to combine the two strategies, for example co-electroporating a plasmid carrying two genes with another plasmid carrying a gene whose gene product will affect one of the two genes on the former plasmid (34), The second gene on the larger plasmid then serves as an internal marker

4.J Naked DNA transfer

due to their ability to synthesize large quantities of protein Protoplasts of sus-pension cultures are easily obtained but often exhibit inferior protein pro-duction capacities In case of many reporter enzymes, the sensitivity of the assay is so high that this is not a problem, but when using Western blotting as a detection method, perhaps in combination of cell fractional ion, leaf protoplasts appear to be the best choice

Due to the ease with which axenic tobacco plants can be grown in large glass jars, tobacco leaves are often used (Protocol 1) From such in vitro grown tobacco plants, it is possible to routinely generate 108 protoplasts each day The same protocol can bo used for Aratidopsis thaliana but due to the small size of the leaves, only 106 protoplasts can be produced in the same time It is possible to produce protoplasts from surface sterilized leaves from greenhouse plants, but the yields are lower and the quality of the resulting protoplasts is inferior, possibly due to the use of rather drastic surface sterilizing agents The use of in ntra grown plants is strongly recommended For regular transient expression work, it is necessary to maintain a population of plants at different developmental stages

Preparation of electroporation competent protoplasts from tobacco leaves

Equipment and reagents

• Manually operated high performance peristaltic pump (10-1000 ml/mm, i.e Watson and Marlowe)

• Bench centrifuge • Petri dishes

• TEX buffer B5 salts (Sigma) supplemented with 500 mg/litre Mes, 750 mg/litre CaCl2.2H;,O, 250 mg/litre NH^NO^, 0.4 M sucrose (13.7%), pH to 5.7 (with KOH)"

Falcon tubes

100 fitn nylon mesh filter

10 x digestion mix in TEX buffer: 2% Macerozyme R1O (Yakult Pharmaceutical),

4% Cellulase R1O (Yakult Pharmaceutical)a,b

Electroporation buffer: 0.4 M sucrose (13.7%), 2.4 g/litre Hepes, g/litre KC1, 600 mg/litre CaCl3, pH to 7.2 (with KOH)0

Method

1 Leaves are chosen from in vitro grown plants Cut gently on the lower surface every 1-2 mm (without cutting through the whole surface) The leaf is held in place with a pair of forceps at the central vein to minimize the damage at the leaf surface It is important to cover the entire leaf surface with parallel cuts to provide an entrance through which the enzymes can diffuse into the intracellular spaces

3 Approx 30 before use, gently shake the plates to release protoplasts from the cuticle Just before use, shake the plate again to release more protoplasts,

4 Filter the digestion mix through a 100 umi mesh nylon filter and wash the filter with clcctroporation buffer This will release further protoplasts from the tissue rem-nants and provides a first step in the adaptation to the new buffer The protoplasts are then centrifuged in 50 ml Falcon tubes for 20 at 60 g in a swing-out rotor, at room temperature Living protoplasts will move to the surface of the solution, whereas cell debris will form a pellet or stay in solution It is important to use a large rotor which slows down over a long period of time (2-3 min) to allow main-tenance of the floating cells in a defined and compact cell layer

5 Connect a long Pasteur pipette to the peristaltic pump and insert the Pasteur pipette through the floating cell layer To avoid many protoplasts sticking to the pipette and moving down with it, a 'window' is created by pushing the cells from the centre outwards by 'brushing' the surface of the protoplasts Once the 'window' is in place, it is important to stait the peristaltic pump immediately after penetra-tion of the pipette through the liquid surface to prevent closing of the 'window' The pellet and the underlying medium are removed until the band of living proto-plasts reaches the bottom One should try to remove as much liquid as possible without losing living (floating) protoplasts.1

6 Add 25 ml of electroporation buffer to the tube and spin again at 100 g for at room temperature At this stage it is possible to pool two samples to reduce the number of tubes, but then centrifugation of the resulting 50 ml should be done for 10 at the same speed Remove the underlying solution as described above and repeat this procedure twice The solution below the protoplasts should become clear and there should be hardly any pellet visible If this is not the case, a third repetition of the washing step can be carried out

7 Resuspend the protoplasts in an appropriate volume of electroporation buffer for the desired number of electroporations This concentrated suspension is now used for electroporation and can be stored for h, but should be used immediately if possible The yield can be lower than indicated but it is not necessaiy to count the cells After the last centrifugation, the appropriate dilution is 1-2 times the volume of the floating band of protoplasts In doubt of the technique, cut more leaves than required to ensure sufficient protoplasts are available for the number of electro-porations planned

"All solutions are sterilized by nitration through 0.2 um filter (Millipore) in a laminar flow bench

" Mix for 30 min, centrifuge (in 50 ml Falcon tube) to remove insoluble particles This solution can be stored in 50 ml Falcon tubes in appropriate aliquots at -80°C and the desired number of aliquots diluted tenfold prior to use In our experience, ml aliquots are ideal

Electroporation and subsequent incubation

Electroporation buffer (see Protocol 1) Purified plasmid DNAa

TEX buffer (see Protocol 1) Equipment and reagents

• Electroporation apparatus operating with standard ml cuvettes with built-in plate electrodes at 3.5 mm distance (i.e Bio-Rad Laboratories)

• Protoplasts (Protocol 1)

Method

1 Pipette 500 ul of competent protoplast suspension gently into the cuvette, to avoid shearing of the protoplasts Up to 50 ug of DNA diluted in 100 ul of electroporation buffer (the DNA should be of sufficient concentration to ensure that no more than 40 ul is used), is added to the cuvette and mixed by gentle shaking Protoplasts will float in the cuvettes during the incubation,b and the cuvette should be shaken from time to time,

2 After a incubation, the DNA will have associated with the cell membrane and

the cells are ready for electroporation Slightly longer incubations (up to 20 min) not reduce the efficiency of DNA transfer Just before electroporation, the suspen-sion is shaken to obtain a homogeneous cell suspensuspen-sion The electroporation is performed with 910 uF and 160 V, after which the cuvettes are kept stationary for

15-30 to allow the cells to repair their plasma membranes

3 Transfer the cell suspension by decanting into a small Petri dish, the cuvette is then rinsed twice with ml of TEX buffer to dilute the cell suspension approximately five times Keep the obtained cell suspension (approx, 2.6 ml) in the dark during an appropriate time period (2-48 h) The incubation temperature has a strong influ-ence on gene expression and should therefore be known 25°C is used when high gene expression is to be obtained after an overnight incubation (16 h),

a Plasmid preparations should be of high quality with minimal amounts of RNA or protein E. coli RNA will cause cell mortality and reduces the efficiency of plasmid transfer The presence

of salts in the DNA solution will also increase the mortality of the cells As such, care must be taken to remove any possible salts in the DNA pellet before re suspension Qiagen preparations are usually sufficiently clean as long as the columns are not overloaded However, because DNA is precipitated from a large volume it may be required to wash with 70% ethanol after precipitation to reduce the amount of salt as described in the Qiagen plasmid purification handbook

• This offers an opportunity to test the quality of the protoplast suspension The protoplast

band should occupy approximately one-third to half of the total volume,

each sample, but pooling two to three may improve reproducibility For cell fractionation, such as sucrose gradients or vacuole preparations, usually 10-20 electroporations are pooled for each sample In addition to electroporation, chemical transfecrion methods based on PEG and calcium are effective naked DNA transfer methods, and may be desirable when an electroporation device is not available However, when large scale experiments are done, the PEC-calcium method is very time-consuming and less reproducible compared to electro-poration The authors strongly recommend electroporation for cell fractionalion studies

4,2 Measurement of protein secretion and cell retention

The following two protocols are designed to test if a protein is secreted to the culture medium or retained in the cells, independent of the site of intracellular localization The method is based on the quantitative recovery of cells and the sampling of clear culture medium that contains essentially no cells To quantify the amount of" the specific protein inside and outside the cells, it is important to work volumetrically and refer every sample as portion of the original suspen-sion If cells from a ml suspension are extracted in a total volume of 200 ul extraction buffer, then this extract is tenfold concentrated compared to the suspension Only a portion of the clear culture medium can be recovered, and this sample will then be at the same concentration as the original suspension When comparing enzyme activities, it is possible to compensate mathematically for the difference in concentration However, if a qualitative protein detection method is used such as Western blotting, it is important to load identical equivalents of the original suspension to allow direct comparison The culture medium m a y b e concentrated t o achieve this (Protocols , )

Harvesting cells and medium from electroporated protoplasts

Equipment and reagents

• Manually operated peristaltic pump (see * Falcon tubes

Protocol1) 250 mM sodium chloride

• Bench centrifuge and microcentrifuge

Method

2 Centrifuge at 100 g for at room temperature The cells will float as in Protocol I A Pasteur pipette is refined by local heating and pulling to reduce the diameter of the opening to 0.5 mm Using the refined Pasteur pipette penetrate the floating cell layer and remove clear culture medium It is important to monitor the cell layer and avoid cells moving down and into the pipette It is possible to remove approx ml of medium before cells will start to move down towards the opening of the pipette Before this happens, remove the pipette and use it to rinse the margins of the tube with a small portion of the medium This is done to wash down protoplasts that stick to the plastic, to prevent mortality and ensure complete recovery of cells in the next step The clear medium is transferred to an Eppendorf tube and is kept on ice The remaining cell suspension is kept until all samples have been processed

4 Dilute the remainder of the cell suspension tenfold with 250 mM NaCl (usually 10 ml), mix gently but thoroughly to avoid a sucrose cushion in the conical bottom of the tube, and then centrifuge for at 100 g The cells will now pellet. Remove the supernatant with a Pasteur pipette connected to a peristaltic pump

This yields a concentrated washed cell pellet, which is placed on ice All subsequent steps are now carried out on ice or at 4°C

6 Centrifuge the clear medium from step at maximum speed in a microcentrifuge for at °C This will remove any cell debris as well as traces of living cells that may have contaminated the medium during recovery Clear supernatant is trans-ferred to a new Eppendorf tube and kept on ice The medium can be concentrated if required and the cell pellet can now be extracted using an appropriate method (see lYotocol 4) From now on it is important to work volumetrically and be aware of all the concentration steps that may be involved

Downstream processing will depend largely on the type of protein that is studied The extraction buffer has to be optimized for each case, but suitable recipes am usually be obtained from the- literature Protocol is an example from which the principle of volumetric work can be understood

Extraction and concentration of proteins from cells and medium

Equipment and reagents

• Bench centrifuge and microcentrifuge • Sonicator

• Extraction buffer; e.g 50 mM Tris-HCl pH 7.5.2mMEDTAa

lOmg/inlBSA

Method

1 For the cellular fraction, resuspend the cell pellet using the appropriate extraction buffer1 to a volume identical to or less than the original cell culture volume If less

volume is used, it should be kept in a sensible proportion, such as 10, or 20 times less The volume of the cell pellet should also be considered Therefore, it is recom-mended to extract in a slightly smaller volume, to measure the exact volume with an automatic pipette, and then top up to the desired volume

2 When all cellular samples are diluted to a known volume, lyse the cells by sonica-tionforS sec (5 um amplitude), and then centrifuge at 25000 g for 10 ininat 4°C Transfer the supernatant into a new Eppendorf rube and place on ice This is now

ready for subsequent enzyme assays or for Western blot analysis

4 The concentration of the culture medium is required for most purposes.b Typically, proteins in the medium are precipitated by ammonium sulfate.v Add 20 ul of a 10 mg/ml BSA solution (as carrier) to the tube containing 600 u1 of culture medium, add 900 ul of a 100% aqueous ammonium sulfate solution, mix well, and incubate h on ice

5 Centrifuge at 25000 g for 10 at 4°C, remove most of the supernatant, and centrifuge again for keeping the tubes in the same orientation in the micro-centrifuge Remove the last supernatant and resuspend the pellet in the appropriate amount of extraction buffer (i.e 10 or 20 times less than the original culture medium volume) The proteins resuspend well on ice with occasional vortexing, this step will take only a few minutes

6 Equal concentrations of original suspension are now analysed If the cell extract and the medium are both tenfold concentrated, equal volumes are used for SDS-PAGE and subsequent Western blotting for direct comparison of intra- and extracellular content

" Refer to relevant literature for information on extraction buffers for eacli particular protein, * If no concentration of the culture medium is required, it can simply be used as it is

'Alternatively, it is possible to concentrate the sample by spin dialysis using an appropriate cut-off value for the molecular weight that passes through the membrane This method is more costly but may provide a full recovery in cases where the protein in question, is not precipitated well by ammonium sulfate This shotild be tested by reconstitution experiments

43 Large scale transient expression for cell fractionation

incubate leaves during the day and clectroporate late in the evening for an overnight incubation This will allow complicated steps associated with cell frac-tionation to be carried out other than in the middle of the night To reduce celt wall re-formation and enhance the yield of vacuolcs, uM of 2,6-dichloroben-zonirrile (DCB) can be added to the cell suspension (38), Care should be taken to avoid degradation of proteins during the fractionation procedure Once cells have been disrupted, all steps should be carried out on ice or at 4CC unless otherwise indicated

Isolation of vacuoles from protoplasts after short transient expression

Equipment and reagents

• Water-bath at 42 0C • Bench centrifuge

• Lysis medium: 0.2 M mannitol, 10% Ficoll 400, 20 mM EDTA, mM DTT, mM Hepes, bring to pH then add 10 ug/ml Neutral Red and l50 ug/mlBSA

Falcon tubes 50 ml and m l Vacuole buffer; 0.6 M betaine, 10 mM Hepes bring to pH 7.5: before use add 150 ug/ml BSA, ug/ml loupeptin, aprotinin, and pepstatinA(Sigma) 250 mM NaCl

Method

1 Due to the scale of the experiment the protoplast suspension needs to be concen-trated after incubation Transfer the protoplast suspension into a 50 ml Falcon tube and centrifuge at 100 g for 10 at room temperature Remove the pelleted dead cells and underlying solution with a peristaltic pump as in Protocol Reduce the volume to less than ml

2 Bring the volume of the suspension to 50 ml using 250 mM NaCI and mix thoroughly Remove ml and transfer to a 15 ml Falcon tube

3 Centrifuge both tubes at 100 g for at room temperature, remove supernatant, and place pellets on ice The pellet in the 15 ml Falcon tube is later to be used for total cell extraction whereas the pellet in the large Falcon tube is immediately used for vacuolar preparation

4 Add ml of pre-warmed lysis medium (42 °C) to the larger portion of cells, place in a

water-bath (42 °C) for 20 sec and swirl gently The solution may be pipetted up and down with a 10 ml disposable plastic pipette once or twice This will help to disrupt cells that are not yet broken due to osmotic or heat shock Remove from water-bath and transfer the suspension to a 15 ml Falcon tube

5 Carefully layer ml of a 1:1 lysis medium/vacuole buffer mix on top of protoplasts, then ml of vacuole buffer The three solutions should be separated by sharp inter-phases

7 Carefully remove the 'red vacuole' layer at the interphase below the vacuoie buffer with a sawn-off pipette The pipette should be held just above this vacuole layer to ensure that besides vacuoles only the overlaying vacuole buffer is aspirated It is im-portant to remove most vacuoles with a minimum volume of vacuole buffer (usually 200 ul)

8 Measure the volume of vacuoles and bring the other cell pellet from step to the same volume using vacuole buffer Sonicate the samples for sec (5 um amplitude) and centrifuge at 25 000g for 10 at 4°C.

9 Transfer half of the supernatant from the vacuoles and cells into Eppendorf tubes and place at -80CC until needed

10 The remaining supernatant is transferred to an Eppendorf and is used for

quanti-fying the recovery of vacuoles (see Protocol 6)

At this point it is again important to work volumetric ally The frozen samples have the same concentration of the vacuolar marker as the solutions that will now be analysed Usually, the recovery of vacuoles from the purification step is 10%, mainly due to bursting of vncuoles as well as incomplete lysis of cells Therefore, approximately ton times less cells are sufficient for the total cell sample In most cases, the vacuolar marker activity will be similar for the puri-fied vacuole sample and the total cell sample It is then possible to dilute one of the two samples to obtain equal activities This h done with the frozen samples prior to the analysis of the protein under study

The following protocol provides a method to quantify the vacuolar marker a-mannosidase To test if a protein under study is localized mainly in the vacuoles, equal vacuolar marker activities are compared for total cells and purified vacuoles If the protein is equally abundant in both samples, then it can be concluded that the majority of the cellular protein is in fact present in the vacuoles and co-localizes with the vacuolar marker

Quantifying the recovery of vacuoles

Equipment and reagents

• Spectrophotometer

• Vacuole preparation (Protocol 5) • Extraction buffer: 250 mM Na-acetate

buffer pH 4.6

• Substrate: mM p-nitrophenyl ô-D-marmopyrannoside (Sigma) in extraction buffer

ã Stop buffer: M Na2CO3

Method

1 To 20 [il of vacuole and cell extracts, add 430 ^1 of extraction buffer.

( Stop the reaction using 800 ul of stop buffer The yellow colour will only become visible at this point

1 Measure the absorbance at 405 nm To obtain readings in the linear range, the absorbance measured should be between 0.1-1.0 It may be possible that the reaction was out of scale (substrate limiting) When the absorbance is over 1.0 it is necessary to repeat the assay with less extract or shorter incubation times

I Equalize the remaining vacuolar and cellular extracts stored at -80°C, so that they contain equal amounts of a-mannosidase activity Equal volumes can then be used for example SDS-PAGE followed by Western blotting If the protein under study is measured via quantitative enzymatic reactions, then simple arithmetic division by the vacuolar marker values will suffice to test if the protein co-purifies with the vacuolar marker

If a protein was localized in the cells but is not recovered in the vacuoles, it may be localized in the ER, the Golgi apparatus, the plasma membrane, or the cytosol To test if a protein is localized in the ER, it is necessary to fractionate cells on sucrose density gradients (Protocol 7) Advantage is taken from the property of ER membrane bound ribosomes to dissociate in the presence of EDTA but remain associated in the presence of Mg2 ions This will cause a density shift characteristic to ER membranes only

This method requires equal quantities of protoplasts as in the vacuolar puri-fication protocol, but the downstream analysis is much more lime-consuming due to the large number of fractions to be analysed for cach sample Whereas it is possible to test four different constructs for a possible vacuolar localization, one should restrict sucrose density fractionation to one or two samples

Cell fractionation by sucrose density centrifugation

Equipment and reagents

• Beckman SW40 swing-out rotor (Beckman) or equivalent

• Peristaltic pump • Fraction collector

• Protoplast preparation (Protocol 1)

MgCl2 buffer: lOOmMTris-HClpH 8, 5mMMgCl2 l m M K C l a

EDTA buffer: 100 mM Tris-HCl pH 8, mM EDTA, 10 mM KC1"

Method

1 Concentration of the protoplasts to a pellet on ice is done as in Protocol 5, steps 1-3, but to obtain two identical pellets Place these on ice

2 ml of 12% EDTA buffer Cells are disrupted using a syringe by repeated (ten times) aspiration

3 Spin the resulting solution at 2000 g for at 4°C to remove cell debris, chloro-plasts, mitochondria, and nuclei This single step is sufficient prior to gradient frac-tionation because protoplasts are very clean compared to plant tissues containing cell walls, starch, and physical debris that need to be removed

4 Layer the slightly turbid supernatant directly onto a linear sucrose gradient (22-50% sucrose)11 containing either MgCl2 or EDTA The gradient occupies 10 ml Spin the tubes at 150 000 g for h at 4°C

5 Take fractions with a peristaltic pump using a fine capillary tube inserted into the gradient to the bottom Typically, 24 fractions of 500 ul are collected from each gradient using a timer and a peristaltic pump When available, a fraction collector can also be used

6 This following step is optional If detection of the proteins is not sensitive, each fraction can be concentrated by ammonium sulfate precipitation (see Protocol 4, steps and 5), Resuspend the pellets in 25 ul of an adequate buffer

7 Detection of the proteins can be via enzymatic reaction or by Western blotting Markers of the ER, the Golgi, the tonoplast, or the plasma membrane are best de-tected by Western blottinga The magnesium shift is crucial to provide evidence for OR localization

a Made up as sucrose solutions either 12%, 22%, or 50% (w/w).

aStep gradients arc made by layering 11 steps of WO ul sucrose solutions between 22% and 50% on top of each other, starting with 50% and ending with 22%, Intermediate solutions are made using 9:1, 8:2, 7:3, until 1:9, mixtures between 50% and 22%, The step gradients are kept in the cold for 2-3 h, after which the steps have diffused and the gradient is linear

'' See Chapter 13 for marker antibodies

4.4 Specialized applications

A simplified method to quickly assess whether a protein is present in micro-somes or whether a protein is membrane associated is based on simple osmotic disruption of the protoplasts (Prtacol 8) Under these conditions, cytosol and vacuolc-s are released from the cells, but ER, Golgi, and transport vesicles remain intact and can be pelleted in a microcentrifuge The pelleted microsomes can be extracted by sonication which will release most of the soluble proteins but only some of the membrane proteins Renewed centrifugation w i l l cause pelleting of microsomes depleted of soluble proteins and it is possible to test if a protein is membrane associated This is just a qualitative assessment but it can be carried out with simple equipment and is worth attempting when a new protein is studied for the first time

mech-anical shearing of protoplasts results in insidc-in microsomes only Proteases will have access to the portion that is exposed on the cytosolic side In the absence of detergents, only this portion will be degraded, the membranes protect the re-mainder of the protein After determining the molecular weight of the protected fragment and comparison with the full-length protein it is often possible to deduce its orientation in the membrane An assay in the presence of detergent will provide a positive conirol for the ability of the protease to digest the entire protein under study

Protocol 8

Assessment of membrane association

Extraction buffer: 100 mM Tris-HCl pH 8, raM MgCl2,10 mM KC1, mM

Equipment and reagents

• Equipment as in Prolocols and

Method

1 Obtain the cell pellet from two electroporations using NaCl (see Protocol 3, steps and 5)

2 Quantitatively measure the volume of the cell pellet and carefully add vol of extraction buffer (see Protocol 4)

3 Vortex briefly and place on ice for 30 sec Repeat five times Centrifuge at 25 000 g for 10 at °C

5 Recover the supernatant and place on ice (S1)

6 Dissolve the pellet in vol (see step 2) of extraction buffer, homogenize by sonicat-ing for sec (5 um amplitude), and centrifuge at 25 000 g for 10 at 4°C.

7 Recover the supernatant and place on ice (S2) Dissolve the pellet in vol of extrac-tion buffer and homogenize by sonicating for sec (5 um amplitude) Do not centri-fuge but place on ice for direct analysis (M) The latter sonication step is merely performed to homogenize the membranes

8 liqual volumes of SI, S2, or M are compared by Western blotting

9 Compare with appropriate markers for the cytosol soluble secretory proteins, and membrane spanning proteins

Protocol 9

Assessment of membrane orientation

Reagents

• Protease digestion buffer: 100 mMTris-HCl pH 8, mM MgCl2, 10 mM KC1,

5 mM CaCl2,12% sucrose (w/w)

Protocol continued

Method

1 Obtain the cell pellet from two electroporations using NaCl (see Protocol 3, steps 4 and 5)

2 Quantitatively measure the volume of the cell pellet and carefully add vol of protease digestion buffer

3 Aspirate using a syringe and add proteinase K to achieve a final concentration of 0.3 rng/ml

4 Incubate on ice for 30

5 Duplicate the digestion with proteinase K in the presence of 1% Triton X-100. Assess molecular weight by Western blotting

5 Conclusions

The protocols described have focused on the specific application of transient expression for the cell biologist who is interested in cell compartmentalization Many of the protocols involve scaling up of the procedures, but this does not involve an increase in the time or work required In fact, the limiting factor is in most cases merely the amount of pure plasmid DNA available Given the potential of the method and the fast and reliable manner in which new results can be obtained, this should not prevent scientists from using the technique in large scale Protocols -4 are generally applicable and will also be helpful for other biological questions, such as the study of protein-protein interactions using co-immunoprecipitation and in vivo labelling (36) In conclusion, the potential of the transient expression technique is far from being fully exploited to date How-ever, with the current debate on the use of genetically modified organisms it is likely that transient expression will gain in popularity, similar to that of anti-bodies generated by phage display to reduce the amount of animals to generate antisera

References

1 Fromm, M, Callis, J L., and Wulbot V (1987) In Mtthodf in cnzymoloxy (ed R Wu and L Grossman), Vol 153.), p 351 Academic: Press, New York.

2 Shiltito, R D., Saul, M W., Paszkowki, J., Mulier, M and Potrykus, L (1985).

Biu/Terhnnlogy, 3, 1099

3 Sporlein, B., Streuhel, M., Dahlfield, G., Wcsthoff P., and Koop, H, U (1991) There. Appl Genet 82.

4 Wu, F S and heng T Y (1999| Plant Cell Rep, 18,381

5 Wang, Y C, Klein T M., Fromm, M, Cao, [ Sanford J C and Wu, R (1998) Plant Mol

Riol, 11,433.

7 Dawson, W O., Beck, D L, Knorr, D A., and Grantham, G L (1986) Proc Not! Acad Sci. USA, 83,1832.

8 Montgomery, J., Goldman, S., Deikman, J., Margossian, L, and Fischer, R L (1993) Proc Natl Acad Sci USA, 90, 5939.

9 Huang, X F., NguyenQuoc, B., Seguin, A., and Yelle, S (1998) Euphyttca, 103,17. 10 Jefferson, R A., Kavanagh, T A., and Michael, W B (1987) EMBOJ., 6, 3901 11 De Wet, J R., Wood, J V., De Luca, M., Helsinki, D R., and Subramani, S (1987) Mol

Celt BtoL, 7, 725

12 Gubler, F and Jacobson, J V (1992) Plant Cell, 4,1435.

13 Dunn, M A., White, A J., Vural, S., and Hughes, M A (1998) Plant Mol BtoL, 38, 551. 14 Ouellet, F., VazquezTello, A., and Sarhan, F (1998) FEBS Lett., 432, 324.

15 Wu, H G., Michler, C H., LaRussa-, L., and Davis,] M (1999) Plant Set., 142,199. 16 Pan, S Q., Sehnke, P C., Ferl, R J., and Gurley, W B (1999) Plant Cell, 11,1591 17 Sessa, G., Borello, U., Morelli, G., and Ruberti, I (1998) Plant Mol BtoL Rep., 16,191. 18 Denecke, J., Botterman, J., and Deblaere, R (1990) Plant Cell, 2, 51

19 Denecke, J., De Rycke, R., and Botterman, J (1992) EMBOJ., 11, 2345

20 Bednarek, S Y.; Wilkins, T A., Dombrowski, J E., and Raikhel, N V (1990) Plant Cell,

2, 1145

21 Hofte, H and Chrispeels, M J (1992) Plant Cell, 4, 995

22 Holwerda, B C., Padgett, H S., and Rogers, J C (1992) Plant Cell, 4, 307

23 Dombrowski, J E., Schroeder, M R., Bednarek, S Y., and Raikhel, N V (1993) Plant Cell, 5, 587

24 Schroeder, M R., Dombrowski, I E., Bednarek, S V., Borkhsenious, O N., and Raikhel, N V (1993) J Exp Bot, 44, 315.

25 Neuhaus, J M., Pietrzak, M., and Boiler, T (1994) Plant/., 5, 45

26 Frigerio, L., Vitale, A J., Lord, M., Ceriotti, A., and Roberts, L (1998) J Biol Chem., 273, 14194

27 Sporlein, B., Streubel, M., Dahfield, G., Westhoff, P., and Koop, H U (1991) Theor Appl. Genet., 82, 717.

28 Hibberd, J M., Linley, P J., Khan, M S., and Gray, J C (1998) Plant;., 16, 627 29 Menand, B., Marechal-Drouard, L, Sakamoto, W., Diertrich, A., and Wintz, H (1998)

Proc Natl Acad Sci USA, 95,11014.

30 Kanamaru, K., Fujiwara, M., Seki, M., Katagiri, T., Kakamura, M., Mochizuki, M., et ol (1999) Plant Cell PhysioL, 40, 832

31 Chang, C C., Sheen, J., Bligny, M., Niwa, Y., Lerbs-Mache, S., and Stern, D B (1999) Plant Cell, 11,911

32 Noji, M., Inoue, K., Kimura, N., Gouda, A., and Saito, K (1998) J Biol Chem., 273, 32745. 33 Shieh, M W., Wessler, S R., and Raikhel, N V (1993) Plant Physiol., 101, 353

34 Leborgne-Castel, N., JelittoVanDooren, E P W M., Crofts, A J., and Denecke, J (1999) Plant Cell, 11,459

35 Pedrazzini, E., Giovinazzo, G., Bielli, A., deVirgilio, M., Frigerio, L., Pesca, M., et al (1997) Plant Cell, 9,1869

36 Denecke, J and Vitale, A (1995) Methods Cell Biol., 50, 335.

Chapter 5

The green fluorescent protein (GFP) as reporter in plant cells

Jean-Marc Neuhaus

Laboratoire de Biochimie, Universite de Neuchatel, Rue Emile Argand 11, 2007 Neuch§tel, Switzerland

Petra Boevink

Scottish Crop Research Institute, Department of Cellular and Environmental Physiology, Invergowrie, Dundee DD2 5DA, UK

1 Introduction

Reporter proteins are an important tool in the study of gene expression in bacteria and eukaryotes Their advantage resides in their easy and reproducible detection under standardized conditions in a large variety of organisms and cell types The detection of most reporter proteins ((5-galactosidase, (3-glucuronidase, luciferase) relies on the enzymatic production of a coloured, fluorescent, or luminescent product which then allows an easy quantification or localization The advantage of enzymatic reporters resides in their sensitivity, as a single protein can produce abundant product molecules, while their disadvantage resides in the need to bring an artificial substrate into contact with the enzyme This introduction may be differentially efficient in different tissues and may even require fixation and permeabilization of the tissue The discovery of spontane-ously fluorescent proteins that not need extraneous fluorophores or substrates has permitted the study of the expression of genes and the localization of proteins in living cells, even in whole living organisms

2 The green fluorescent protein

2.1 Structure

The green fluorescent protein (GFP) originally cloned from the jelly fish Aequorea

victoria is a reporter protein which can be detected in a non-invasive way in

acids Ser65, Tyr66, and Gly67 (2) The only requirement for this reaction is oxygen The wild-type GFP absorbs light both in the UV (absorbance band A, 395 nm) and the blue (absorbance band B, 475 nm) parts of the spectrum and emits green light (508 nm) The GFP is small and compact and can be fused both N- and C-terminally to other proteins without losing its fluorescence and often without disrupting the function of the other protein Since both termini of GFP are close together on the same face of the (J-barrel it may also be inserted into a loop of another protein Conversely proteins may be fused within a loop of GFP itself, as long as their ends are close to each other (3)

2.2 GFP variants

Wild-type GFP has been modified in many ways to improve its qualities as a reporter in various organisms Extensive changes have been made in the GFP sequence to adapt the codon usage to various host organisms In the initial plant transformations experiments type GFP was expressed very poorly The wild-type gene was found to encode consensus plant intron splice sites which re-sulted in aberrant splicing of the GFP mRNA Removal of the intron recognition sequences restored a useful level of expression in plants (4) The cryptic intron was also inadvertently removed in GFP variants in which the codon usage was optimized for mammalian cells (5) Further improvements were obtained by increasing the thermostability, folding properties, and/or solubility of GFP (4, 6) Spectral variants have been generated to increase the quantum yield, to change the ratio of the absorption bands A and B, and to shift the emission band towards the blue or red ends of the spectrum Some of these enhanced GFP variants (EGFP, GFP6, etc.) have lost the excitation by UV light, which may prevent applications such as the identification of transgenic plants with a long wavelength UV lamp The 'blue' variants (such as EBFP, Clontech) are excited only by UV light and appear to bleach rapidly The 'cyan' and 'yellow' variants (ECFP, EYFP, Clontech) can be co-expressed and detected separately, provided suitable excitation and emission filters are available A newly available red fluor-escent protein from the sea anemone Discosoma (Clontech, see http://www clontech.com/) has also been successfully tested in plants (unpublished observa-tions) Its fluorescence can be distinguished from chlorophyll autofluorescence with the appropriate filter sets DsRed gives a much lower quantum yield than GFP and appears to photobleach more readily It is expected that improved DsRed mutants will soon become available

3 GFP as a reporter for gene expression

Table Properties of some GFP variants Variant GFP mgfp4 smGFP mgfp5 GFP6 EGFP EBFP ECFP EYFP DsRed Absorption (nm)

395 / (475)

395 / (475) ' 397 / (480) 395 / 475

'475

488

380

433 / 453

513 558 Emission (nm) 508 508 507 510 510 510 440

475 / 501 '527

583

Other properties

Poor expression in transgenic plants, due to a cryptic intron

No splicing

No splicing , more soluble

No splicing, more thermostable, equal absorption peaks

No UV excitation, higher quantum yield

No UV excitation, higher quantum yield, mammalian codon usage

'Blue', needs UV excitation, bleaches rapidly

'Cyan'

'Yellow'

Sea anemone red fluorescent protein

to depend on the plant species and can be solved by using another variant of GFP (7) or by targeting to the ER (4) GFP can be used as an insertion tag to detect tissue- and development-specific activity of unknown genes in living plants (8) or as a marker of tissue origin when isolating protoplasts from a mixed population

(9)-4 GFP as a reporter for protein location

The independence from exogenous cofactors and the good tolerance of N- and C-terminal fusions means that GFP can be used to analyse intracellular targeting to any compartment of a plant cell

4.1 Cytoplasm and nucleus

Table Examples of targeted GFPII

Compartment Construction

Nucleus (import)

Nucleus (export)

Chloroplast

Mitochondria

ER

Neutral vacuole

Lytic vacuole

Golgi

Microtubules

Actin filaments

Microtubules, plasmodesm?

C2-NLS-GFP-GUS (nuclear localization signal-GFP-GUS) mGFP/VPg (nuclear protein-GFP) (11)

NESofRev

NESof AtRanBPla (24)

GFP expressed in Chloroplast (12) recA-GFP (transit peptide fusion) (13)

CoxlV-GFP (transit peptide fusion) (15)

mgfp5-ER (C-terminal HDEL) (4)

SGFP5T (C-terminal vacuolar sorting determinant) (25)

AGFP6 (N-terminal propeptide containing a sequence-specific vacuolar sorting determinant) (35)

aERD2-GFP (fusion to C-terminus of complete ERD2) ST-GFP (N-terminal transmembrane segment of a mammalian sialyl transferase) (17)

GmManl::GFP (a-l,2-mannosidase) (26)

GFP-MAP4 (mammalian microtubule associated protein ) (27)

GFP-talin (28)

rta M:Gfus (GFP fused to the TMV movement protein) (19)

4.2 Chloroplasts and mitochondria

There are two ways to produce GFP in chloroplasts: either by plastid trans-formation with a GFP under the control of a plastidial or bacterial promotor (12), or by nuclear transformation with a gene coding for a GFP fused to a chloroplast transit peptide (13,14)

Fusion of the transit peptide of the yeast cytochrome oxidase subunit IV (coxIV) to GFP resulted in labelling of plant mitochondria, which was superior to that obtained with mitochondria-specific dyes (15)

4.3 Secretory pathway

4.4 Viral proteins

By fusion with various viral proteins, GFP can be used to visualize normal cellular structures such as cytoskeleton or plasmodesmata (19), and also the changes that accompany the development of a viral infection within a cell and the movement of the virus to neighbouring cells (20-22) When expressed in protoplasts, some GFP-movement protein fusions highlight protrusions at the surface of the protoplasts, creating amazing hairy cells (23)

5 Transformation methods

A wide variety of transformation methods are available for many different species Methods to produce transgenic plants and suspension cultures will not be described here, as they not differ from standard procedures However, it should be appreciated that GFP may be used as a visible screening marker to optimize a transformation protocol without having to wait for the effects of a selectable marker, saving both time and material

5.1 PEG-mediated transient expression in protoplasts

This method is extremely simple to use (Protocol 3) but is restricted to species and cell types for which there is a method of protoplast isolation and for which the conditions have been established (Protocols and 2) It allows a rapid test (within 24 h) of new constructions in simple vectors and provides some indications of the kinetics of a cellular process (see also Chapter 4) The size and shape of proto-plasts means that only a small part of the volume can be imaged at any given time and that mobile organelles often move out of the plane of focus On the other hand, sufficient material can be produced for biochemical analysis

The basic principle is that the PEG destabilizes the plasma membrane and thereby allows the DNA to enter the cell Protoplasts from leaves represent a mixture of cell types differing in size and number of chloroplasts (see Chapter and Protocol 1) Protoplasts from a cell suspension represent a more homogene-ous material and some cell lines are free of green chloroplasts, but they have to be maintained by regular inoculation of fresh medium It must also be remembered that protoplasts are cells under stress and undergoing a process of de- or re-differentiation that may differ strongly from the cells in their parent tissue

The plasmids used for transient expression only need a strong plant promotor such as 35S (or enhanced versions) and a termination sequence (Figure 1) It is preferable to use a high-copy plasmid, as extensive purification is required, such as caesium chloride/ethidium bromide gradient centrifugation or anion ex-change chromatograpy followed by phenol/chloroform extraction and ethanol precipitation

gradually decrease, maximal fluorescence being observed after 20-48 hours As rhe level of expression varies from cell to cell apparent transformation efficiency depends on the intensity of the signal The authors have achieved expression levels in tobacco where 80% of the protoplasts are fluorescent

Protocol 1

Preparation of protoplasts from sterile tobacco plants

Equipment and reagents

• Plant culture vessels, e.g Sigma Phytacon or Magenta vessels

• Sterile stainless steel 100 (um mesh sieve • Hnzyme solution: to 100 ml K3, add 1.2 g

Cellulase Onozuka RIO (Serva, Heidelberg) 0.4 g Macerozynie (Serva), g sucrose; filter sterilize

• Culture medium: K3 (29) with 0.3 M sucrose, 0.5 ^.M 2,4-D, 0,9 \iM BAP, NAA

• W5 osmoticum: 154 mM NaCl, 125 mM CaCl2, mM KC1, mM glucose pH • MMM:0.5Mmannitol, ISmMMgCl^, 0.1%

Mes

Method

1 Grow shoot cultures of tobacco (e.g Nicotiana tabacum variety SRI) on a sterile solid medium in culture vessels Transfer to fresh medium after four to eight weeks

2 Cut leaves, wet them in enzyme solution, remove the midrib, and wound the upper epidermis Add enzyme solution to 10 ml for two leaves in a Petri dish and seal with Parafilm,

3 Incubate overnight at 26°C in the dark

4 Gently agitate for 30 inin, collect the protoplasts with a wide bore pipette, and filter through the stainless steel sieve

5 Divide into ml aliquots and overlay with ml W5 solution Centrifuge at 80 g for 10

6 Collect the protoplasts from the interphase, dilute with W5 (for ml, add 10 ml W5), mix gently, pellet at 80 g for min,

7 Repeat the rinsing once

8 ResuspendinW5,storeat4°Cfor2h

9 Count the protoplasts (Protocol 3).

W Pellet the protoplasts at 80 g for min.

Figure Maps of plasmids and vectors used for transient expression of GFP, emphasizing control elements and useful restriction sites pSGFPST, a plasmid allowing the transient expression in protoplasts of GFP targeted to neutral vacuoles The GFP5 variant was taken from the plasmid pBIN mgfp5-ER including the signal sequence (SS) from an Arabidopsis chitinase. The sequences of thejunction between SS and GFP and of the C-terminal extension containing the vacuolar sorting determinant are indicated The location of the fluorophore is indicated by a star pAGFP6, a plasmid allowing the transient expression in protoplasts of GFP targeted to lytic vacuoles The signal sequence and the propeptide containing the vacuolar sorting determinant were taken from barley aleurain The sequences of thejunction between propeptide and GFP and of the natural C-terminus of GFP are indicated The location of the fluorophore is indicated by a star pBIN mgfp5-ER, a binary plasmid for the Agrobacterium-mediated transient or stable transformation of plant tissues, allowing selection on kanamycin GFP constructions can be introduced as SamHI-SacI (or Xbal-Sacl) fragments, e.g from the former two plasmids pTXS.P3C2, a PVX vector for virus-mediated transient expression The first four PVX ORFs are represented by the sizes of their protein products in kDa (K) The coat protein ORF is shown as CP The multiple cloning site is expanded to show the available restriction enzyme recognition sites for inserting foreign coding sequences A duplication of the coat protein subgenomic promoter (PCP) allows high level of expression of the inserted sequence TheT7 RNA polymerase

promoter (PT7) allows the production of run-off transcripts after linearization by Spel at the

position indicated by the flag Abbreviations: P35S, PNOS, PCP PTT promoters 35S, NOS, PVX coat

protein, T7 T35s, TNOS, terminators 35S, NOS LB, RB, left and right borders of T-DNA SS, signal

Protocol 2

Preparation of protoplasts from an Arabidopsis cell suspension

Equipment and reagents

* Two sterile stainless steel 100 um mesh sieves

• Culture medium: Gamborg's (Sigma) supplemented with 20 g/litre sucrose,

D, pH5.7

• 14 cm Petri dishes

• Enzyme solution, W5 (see Protocol 1) • MCM: 0.4 M mamiitol, 0.2

0.3mMMespH5.6

Method

1 Subculture an Arabidopsis cell suspension (30) weekly by inoculating 300 ml medium with g cells For protoplasting, inoculate 300 ml medium with g cells four days before the experiment

2 Collect cells from 50 ml suspension culture by filtration with a stainless steel sieve, transfer to 14 cm Petri dishes, rinse the filter upside-down with 50 ml enzyme solu-tion, and resuspend gently with a cut plastic pipette to dissociate the clumps Seal with Parafilm

3 Incubate for h at 30 SC in the dark with gentle agitation

4 Collect the protoplasts with a wide bore pipette and filter through a sterile stainless steel sieve

5 Divide into ml aliquots and overlay with ml W5 solution Centrifuge at 80 g for 10

6 Collect the protoplasts from the interphase, dilute with W5 (for ml, add 10 ml W5) mix gently, peliet at 80 g for

7 Repeat the rinsing once

8 Resuspend in W5, store at 4°C for h Count the protoplasts

10 Pellet the protoplasts at 80 g. 11 Resuspend in MCM at 2.5 x 10fi/ml

52 Agrobacterium-mediated transient expression in planta

This method allows the rapid testing of a construction and the observation of the fluorescence,

Protocol 3

PEG-mediated transient expression in protoplasts'

• Water-bath

• E pi fluorescence microscope with appropriate filter sets

• TE buffer: 10 mMTris-HCl pH 7.5.1 mM EDTA (diluted from 10 x stock)

Equipment and reagents • PEG solution: 40% PEG 6000 (Merck,

Darmstadt), 0.1 M Ca(NO3)2 0.4 M mannitol 0.1% Mes pH 7.8 Dissolve Ca(NOa)2,

mannitol, and Mes in 70 ml water, adjust pH to 8.0, add 40 g PEG, stir for h, filter sterilize and freeze in aliquots Before use, thaw in warm water-bath

Method

1 Dispense 10 |o,g plasmid (containing appropriate GFP construct) in jxl TE buffer near the bottom of sterile 15 nil plastic tubes For one-day experiments, no special sterilization procedure is usually required for the DNA For experiments lasting several days, DNA can be etHanoi precipitated and dissolved in sterile buffer shortly before the transformation

2 Pipette 750000 protoplasts (0.3 ml) with a wide bore pipette to the DNA and gently mix

3 After 2-5 add 0,3 ml PKG solution and gently mix

4 After dilute the protoplasts with ml K3 medium and incubate in the dark at 26 °C for h

5 Dilute the protoplasts with nil W5 and pellet for at 70 g Resuspend in ml K3 and incubate for another 10-48 h This rinsing step is not absolutely required, but it removes the PEG which may cause problems if extractions and precipitations are required later Rinsing should occur not later than h after transformation

6 Dilution with W5 makes it easier to pellet the protoplasts before fractionation, homogenization, and any biochemical analysis (Chapter 4)

" See Chapter for other methods of transient expression in protoplasts

the cell type The geometry of epidermal cells allows a good observation of mobile organelles but the amount of expressing tissue may restrict biochemical analysis Expression of GFP in the bacteria may also cause a background fluorescence in the intercellular spaces, depending on the plasmid and the GEP construction

It is also possible to infiltrate leaves or leaf pieces by immersion in the bacteria suspension and application of vacuum Upon return to normal pressure, the suspension infiltrates large areas of the leaf It is even possible to infiltrate whole Arubidapsis plants this way.

Protocol 4

Agrobacterium-mediaied transient expression In tobacco leaves

Equipment and reagents

• Tobacco plants (N tobacum, N, hfnthamfano, N clevelandii ail work) in pots

• Indelible marker pen • ml syringes

• Epifluorescence microscope with appropriate filter sets and/or laser scanning confocal microscope

Long wavelength UV light (365 nm) YEB medium: g/1 each beef extract, peptone, sucrose, g/1 yeast extract, 0.2 mM MgSO4

Infiltration medium: 50 mM Mes, mM Na phosphate pH 5,6, 0,5% glucose 100 uM acetosyringone (from a 2000 x stock in DMSO}

Method

1 Using standard molecular biology techniques clone a GFP construct of choice into a binary vector and mobilize into a suitable Agrobacterium lumefariens strain.

2 Culture Agrobacterium to stationary phase (one to two days) in YEB medium at 28 °C. Pellet ml of the culture, rince once, and resuspend in infiltration medium (discard supernatant in disinfectant) For some constructs it may be necessary to determine empirically what concentration of Agrobacteria results in the desired expression levels in the plant Good levels of expression can be obtained with concentrations giving an absorbance at 600 nm (OD600) of 0.6-0.3 or less

4 Press the nozzle of a ml syringe (no needle) against the lower (abaxial) epidermis of a tobacco leaf holding a finger or palm to the other side of the leaf and inject slowly The infiltrated area turns dark Mark its limits with an indelible pen If the suspension does not penetrate easily, a small cut can be made in the lower epidermis and the nozzle of the syringe applied over the cuta

5 Incubate plants under normal growing conditions for two to three days,

6 After two days, excise the marked area and examine under the microscope UV-absorbing GFP can be directly visualized on the plant with a hand-held long wave-length UV lamp

7 Mount explants from the leaves in water on a microscope slideb and observe with an epifluorescence, or preferably a laser scanning confocal microscope

" Infiltration is easier when stomta are open so subjecting plants to a bright light facilitates penetration of the leaf tissue,

b If necessary covers lips can be secured to the slide with waterproof tape,

5.3 Virus-mediated transient expression

as expression vectors for plant systems, in the same way that baailoviruses have been used for protein expression in insect cells Potato virtis X (PVX) has been used to express the marker proteins GFP and GUS (31, 32) and numerous other foreign proteins, many as fusions with the GFP (17, 18), The main advantage the vims expression system has over transgenie plants is that results are obtained very rapidly; three to four days alter inoculating leaves of the host plant with infectious transcripts of PVX.GFP, lesions can be seen on the leaves under a long wavelength UV lamp (365 nm) and infected cells can be observed even earlier under a microscope (Protocol 5) The levels of protein expression from viral vectors are also generally veiy high, from a combination of the strong viral

pro-moter controlling the expression of the foreign protein, and the replication of the viral RNA providing more template for protein production However, certain aspects of the system must be taken into consideration before experiments are undertaken:

(a) Not all of the cells in the plant will be expressing the gene of interest, as not all will be infected with the virus

(b) Genes expressed from most kinds of viral vectors will be expressed consdtutively at a high level so the system cannot be used for studies of gene regulation

(c) Most viruses are limited in the size of insert they will tolerate

(d) Any pathological effects of the virus infection must be taken into account

Protocol 5

PVX vector expression in whole plants

Equipment and reagents • Controlled temperature, licensed

glasshouse (if needed under local regulations), or growth chamber • Epifluorescence or confocal microscope • Suitable virus host plants (generally

Nicotiana benthamiana or N develandii)

T7 RNA polymerase transcription kit (e.g Ambion mMessage mMachine) or separately purchased enzymes and components required for transcription Aluminium oxide abrasive powder

Method

1 Clone the gene for GFP, a GFP fusion, or other protein of interest into the PVX vector, pTXS.P3C2 (Figure 1) The gene must be cloned in the correct orientation with re-spect to the promoter

2 Prepare high quality, RNase-free modified vector DNA and linearize it with Spel or Sphl

3 Synthesize capped transcripts using cleaned, linearized DNA as template

Protocol continued

Add 3-5 ul of transcript per leaf, depending on leaf size, and gently rub the tran-script over the leaf with a gloved finger Rub in unidirectional strokes towards the leaf tip It is not necessary to clean or precipitate the transcripts before inoculating After inoculation wash the abrasive off the leaves Plants should be kept at approxi-mately 23°C

5 Three to four days after inoculation with PVX.GFP, plants should have small fluorescent lesions on inoculated leaves, which are visible under ul light.a

6 After a suitable time leaf segments can be taken for observation of cells with fluor-escence or confocal microscopes For some fusion constructs a fluorfluor-escence or con-focal microscope might be required to find infected cells

aSome forms of GFP which have been optimized for imaging with the confocal microscope are not visible with hand-held UV lamps (Table 1) and some GFP fusions are not visible under a lamp due to lower levels of protein Insertion of foreign genes into the viral genome can cause the virus to spread more slowly

Protocol 6

Blolistic inoculation of PVX vector constructs

Plastic bombardment grid holders (PALL, Gelman Sciences, MI)

Equipment and reagents

• Bombardment apparatus, either

commercial (e.g Bio-Rad) or home-made (33) • fi.ni gold particles (Bio-Rad)

Method

1 See Protocol 5, steps 1-3.

2 Precipitate transcripts with LiCl as described in the instructions of the transcription kit manufacturer and resuspend in the same volume of sterile water

3 Mix (j,l of transcript with |j,lof ethanoland 11 jilof 50mg/ml ^m gold particles (Bio-Rad)

4 Load 2-5 ul of gold mixture per grid and discharge the gun two or three times per loading, moving the leaf between each discharge such that each 'hit' is in a different place

In the PVX vector the coat protein subgenomic promoter has been duplicated and a multiple cloning site created downstream for insertion of foreign genes ( F i g u r e I) Using the pTXS,P3C2 vector plants arc manually inoculated with transcripts Biolistic inoculation is also possible with transcripts or with DNA from vectors in which the viral genome has been cloned in front of the CaMV 35S promoter ( P r o t o c o l ) ) Various other plant viruses, such as, tobacco mosaic virus (TMV), cucumber mosaic virus, tobacco rattle vims (TKV), groundnut rosette virus, pea seed borne mosaic virus, tobacco etch virus, and gemini viruses, have also been engineered to express CFP and other proteins These viruses differ in the levels of foreign gene expression, ease of manipulation of the viral genomes

in cloning operations, inoculation technique, and host specificity TRV, various strains of TMV, and a variety of other tobamoviruses infect Arabidopsis and have been or are being developed as expression vectors

6 Visualization and microscopy of GFP

GFP expression in plant tissues can be viewed at the macroscopic level with hand-held UV lamps providing that the expression levels are sufficiently high and the variant of GFP used is excited by UV light (see Table 1) Seeing GFP fluorescence in plant cells (e.g an ER-targeted variant) for the first time is an amazing experience, best enjoyed with a standard epifluorescence microscope without narrow waveband barrier filters For microscopic examination of GFP expressing tissues standard fluorescence microscopes with mercury, vapour lamps and confocal laser scanning microscopes (CLSM) with blue argon ion lasers, which excite GFP with 488 nm wavelength light, are commonly used To obtain useful images it is necessary to filter the emission spectrum with appro-priate barrier filters to select against reflection and non-specific fluorescence from chloroplasts and other elements of plant tissues Some examples of barrier filters are 525/550 nm filter ('GFP Plant filter') for a Leica MTFLIII stereomicro-scope, 520-560 nm filter for Nikon Optiphot fluorescence microstereomicro-scope, and a 522 DF 32 nm emission filter for a Bio-Rad MRC 1000 CLSM Many more filters are becoming available as the use of GFP spreads in the scientific community Stand-ard fluorescence microscopes have the advantage of being cheap and simple to use, however the emitted light from all depths of the tissue is visible at the same time This makes it difficult to determine what is occurring in a single cell in a multilayered tissue For crisp, clear images at the cellular and subcellular level, it is best to use confocal laser scanning microscopes, which exclude emitted light coming from all areas of the tissue except the plane of focus

If the compartment you have labelled is mobile, it may move out of the plane of observation too fast for your study, especially in round cells like protoplasts This may necessitate fixation with paraformaldehyde as for immunocyto-chemistry (see Chapter 10) prior to observation GFP survives fixation and allows one to compare the pattern of fluorescence before and after fixation Alterna-tively, rapid image capture can be achieved by optimizing the GFP signal, obviat-ing the need for image averagobviat-ing At the expense of some resolution and con-focality the pin-hole aperture can be widened and (if photobleaching is not a major problem) the laser can be used at high power Using this configuration movies can be made by collecting sequential images of whole frames or regions of interest using proprietary confocal time-lapse software

7 Future perspectives

GFP has only recently been introduced into the field of plant cell biology, but it already revolutionized the study of intracellular localization and converted long-time biochemists used to cell fractionation and and bar graphs of enzyme activities into microscopists We can expect many compartments of plant cells to become visible in the near future A major step will be the definition of patterns typical for these compartments, as already known for animal cells The combination of two or even three different variants in the same cells will allow rapid decisions about co-localization, which may also be quantified by FRET (fluorescence resonance energy transfer) (34) (Chapter 2)

References

1 Chalfie, M., Tu, Y., Euskirchen, G., Ward, W W., and Prasher, D C (1994) Science, 263, 802

2 Niwa, H., Inouye, S., Hirano, T., Matsuno, T., Kojima, S., Kubota, M., et al (1996) Proc. Natl Acad Sri USA, 93,13617.

3 Abedi, M R., Caponigro, G., and Kamb, A (1998) Nucleic Acids Res., 26, 623.

4 Haseloff, J., Siemering, K R., Prasher, D C., and Hodge, S (1997) Proc Natl Acad Sri. USA, 94, 2122

5 Patterson, G H., Knobel, S M., Sharif, W D., Kain, S R., and Piston, D W (1997) Btophys.J., 73,2782

6 Davis, S J and Vierstra, R D (1998) Plant Mol Biol., 36, 521.

7 Niwa, Y., Hirano, T., Yoshimoto, K., Shimizu, M., and Kobayashi, H (1999) Plant]., 18, 455

8 Berger, P., Linstead, P., Dolan, L, and Haseloff, J (1998) Dev Biol, 194, 226. Maathuis, F J M., May, S T., Graham, N S., Bowen, H C., Jelitto, T C., Trimmer, P.,

et al (1998) Plant]., 15, 843.

10 Chiu, W., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H., and Sheen, J (1996) Curr Biol, 6, 325

11 Grebenok, R J., Pierson, E., Lambert, G M., Gong, F C., Afonso, C L, Haldeman-Cahill, R., etal (1997) Plant]., 11, 573.

12 Hibberd, J M., Linley, P J., Khan, M S., and Gray, J C (1998) Plant]., 16, 627. 13 Kohler, R H., Cao, J., Zipfel, W R., Webb, W W., and Hanson, M R (1997) Science,

276, 2039

14 Tirlapur, U K., Dahse, I., Reiss, B., Meurer, J., and Oelmuller, R (1999) Bur J Cell Biol., 78, 233

15 Kohler, R H., Zipfel, W R., Webb, W W., and Hanson, M R (1997) Plant/., 11, 613 16 Siemering, K., Golbik, R., Sever, R., and Haseloff, J (1996) Curr Biol., 6,1653 17 Boevink, P., Martin, B., Oparka, K., Santa Cruz, S., and Hawes, C (1999) Planta, 208,

392

18 Boevink, P., Oparka, K., Santa Cruz, S., Martin, B., Betteridge, A., and Hawes, C (1998) PlantJ., 15,441

19 Heinlein, M., Epel, B L., Padgett, H S., and Beachy, R N (1995) Science, 270,1893. 20 Oparka, K J., Prior, D A., Santa Cruz, S., Padgett, H S., and Beachy, R N (1997) Plant J.,

12, 781

21 Heinlein, M., Padgett, H S., Gens, J S., Pickard, B G., Casper, S J., Epel, B L., etal. (1998) Plant Cell, 10,1107.

23 Grieco, F., Castellano, M A., Di Sansebastiano, G P., Maggipinto, G., Neuhaus, J.-M., and Martelli, G (1999) J Gen Virol, 80,1103.

24 Haasen, D., Kohler, C., Neuhaus, G., and Merkle, T (1999) Plant]., 20, 695. 25 Di Sansebastiano, G P., Paris, N., Marc-Martin, S., and Neuhaus, J.-M (1998) Plant].,

15,449

26 Nebenfuhr, A., Gallagher, L A., Dunahay, T G., Frohlick, J A., Mazurkiewicz, A M., Meehl, J B., etal (1999) PlantPhysiol, 121,1127.

27 Marc,]., Granger, C L, Brincat, J., Fisher, D D., Kao, T., McCubbin, A G., etal (1998). Plant Cell, 10,1927

28 Kost, B., Spielhofer, P., and Chua, N H (1998) Plant J., 16, 393.

29 Goodall, G J., Wiebauer, K., and Filipowicz, W (1990) In Methods in enzymology (ed J E. Dahlberg and J W Abelson) Vol 181, p 148, San Diego, London

30 Axelos, M., Curie, C., Mazzolini, L., Bardet, C., and Lescure, B (1992) Plant Physiol. Biorfiem., 30,123

31 Chapman, S., Kavanagh, T., and Baulcombe, D (1992) Plant]., 2, 549. 32 Baulcombe, D C., Chapman, S., and Santa Cruz, S (1995) Plant]., 7,1045.

33 Gal-On, A., Meiri, E., Elman, C., Gray, D J., and Gaba, V (1997) J Virol Methods, 64, 103. 34 Gadella, T W J., van der Krogt, G N M., and Bisseling, T (1999) Trends Plant Sri., 4,

287

35 Di Sansebastiano, G P., Paris, N., Marc-Martin, S., and Neuhaus, J.-M (2001) Plant

Chapter 6 Microinjection

Michael Knoblauch

Institut fiir allgemeine Botanik und Pflanzenphysiologie, Senckenbergstrasse 17, 35390 Giessen, Germany

1 Introduction

During recent decades, microinjection has developed into an important method used by cell biologists, and there have been several thousand papers published describing applications of this technique The main advantage of microinjection is that it enables the direct introduction of substances into individual cells or organelles whilst being observed Microinjection can be used in intact tissue of whole plants as well as in tissue cultures or tissue slices In essence, for micro-injection a glass pipette (needle) with a tip size of about jim is filled with the substance to be introduced The substance is forced from the capillary by hydro-static pressure (pressure injection) or by an electric current (iontophoresis) into the cell The main problem is that the method is technically demanding On the other hand, it is inexpensive, costs being mainly associated with purchase of the experimental hardware

2 Equipment

A variety of different set-ups is available, and there are many possibilities for modification In the following section, one set-up will be described which is ideal for many applications for injecting into plant cells However, this is just one of many possibilities

2,1 Environment and injection-table

2,2 Microscope

Microscopes for microinjeclion or electrophysiology (which uses essentially the same set-up) are available from different manufacturers A minimal configuration is shown in Figure la Microinjoction is easier when carried out using an inverse microscope, as the cells arc more readily accessible The disadvantage is that it is possible to use the system only for thin tissue In thicker tissue, the pipelle is impossible to see during microinjecton Therefore, the author prefers to use the upright configuration, which is independent of tissue thickness

In upright systems, it is an advantage if focusing is achieved by moving the objectives rather than the stage, as then the specimen stays fixed which in some cases is essential to avoid breaking the pipette Thus, if parts outside the visual field need to be observed the microscope has to be fixed on an additional xy-table

(Figure lb)) If a microscope is already available with a fixed stage, both have to be

placed on the anti-vibration table

23 Objectives

Crucial elements of all set-ups arc the objectives For the upright system one should heed the following points

(a) It is often helpful if the pipette can approach the cell at a steep angle Thus, objectives are necessary which provide enough space between the lens and the specimen Long working distance objectives are available as dry or water immersion lenses For most circumstances water immersion lenses arc pre-ferable since they can be used directly in the bathing medium and allow much better images to be captured

56

(b) If possible, objectives must not be corrected for observations through a cover-slip, as one never uses a coverslip during injection, yet most objectives are corrected, resulting in a loss of image quality

(c) The numerical aperture of the objective should be as high as possible If epifluorescence is used for observation, a high aperture is essential to make low level fluorescence visible or to enable reduction of the excitation energy (see below)

(d) If available, the setting of the objective should be ceramic Ceramics are more corrosion resistant than metal settings, and they are electrically insulated, which enables parallel measurement of the membrane potential (1)

For inverse microscopes several objectives are available, which are corrected for different glass thickness (e.g mm for slides)

2.4 Glass capillaries

Borosilicate glass is the most commonly used and cheapest material available for making microinjection needles Standard glass capillaries have an outer diameter of or 1.5 mm and an inner diameter of 0.58 or 0.9 mm respectively (e.g Hilgenberg, Sutter Instruments, Clark Electromedical Instruments) They are available with and without inner filaments Filamented capillaries are used to facilitate the filling of the pipette tip After backfilling the capillary with a drop of the solution to be injected, the solution moves by capillary forces on the internal filament to the tip Aluminosilicate glass and quartz glass are also avail-able and differ from borosilicate glass due to their increased hardness They are preferable if cells with hard or thick walls have to be injected, but alumino-silicate and quartz glass is more expensive

2.5 Tip puller

2.6 Micromanipulator

Most micromanipulators comprise a coarse manipulator to bring the pipette tip near the target cell, and a fine manipulator to control final impalement Both manipulators can move in three dimensions (x, y, z) With three axis manipulators the pipette can only be moved in one direction at a time (forward-backward, left-right, or up-down) However, some fine manipulators allow simultaneous action of all three axes, i.e the pipette can always be moved in the direction of impalement This is an advantage when forcing the needle tip through the cell wall

Micromanipulators work either by mechanical or by hydraulic mechanisms They can be attached either to the table or to the stage of the microscope The author prefers manipulators which are fastened onto the microscope stage and which are controlled by water or oil pressure The control unit can be placed on a separate table so that turning the micrometer screws does not influence the stage or moving the joystick, which may cause vibration

In an hydraulic apparatus forces are transduced by pressure in a liquid-filled system However, as the whole pressure system is closed, it is temperature sensi-tive Thus, if there is any increase temperature, the oil or water in the system expands and drives the piston forward in an uncontrolled fashion It is therefore possible to impale a cell just by touching a liquid-filled tube with a finger Since small temperature changes can never be excluded, one should if possible insulate any liquid-filled lines in the system

3 Injection techniques

3.1 Iontophoresis

For iontophoretic injection the molecules to be injected has to carry a charge The method uses an electrical current of 0.1-10 nA of the same polarity as the substance, to drive it into the cell For iontophoresis the molecular weight of the material to be introduced is normally not larger than kDa with a maximum of about 10 kDa That excludes the injection of genetic material, most proteins, and all uncharged molecules This method is generally used for the insertion of small charged fluorescent dyes such as Lucifer Yellow-CH Molecules of low charge require higher currents, which may damage the cell and in general the current should be as low as possible It is preferable to inject for a longer time with low currents, than for a short time with high currents Currents can be applied continuously or as pulses (for further details see refs 2, 3)

3.2 Pressure injection

Figure Comparison of tip sizes of microinjcction pipettes Scanning eleclron micrograph of the tip of a conventional microinjection needle (a) with a tip diameter of about 0.7 um, and the tip of a galinstan expansion femtosyringe (b) with a diameter of about 0.1 um Bar um.

From Ref 5,

low Because of the low compressibility of liquids, hydraulic pressure generators are more sensitive, easier to control, and are able to produce higher pressure In such systems, a micrometer screw drives a syringe to produce the necessary pressure The syringe is then connected to the needle holder If the injector is combined with a manometer the generated pressure can be monitored during the injection process, and can be accurately adjusted before injection (i.e epi-dermnl cells) Furthermore, the turgor of the cell can be measured (for further details see ref 4) In the author's opinion, pressure injection is less invasive than iontophoresis as long as the pressure is well adjusted and the injected material moves slowly into the cell The apparatus is not as complicated, and can be constructed in any good workshop The system is normally filled with a low viscosity silicon oil which should be degassed under vacuum before filling the system The pressure required for injecting most plant cell types is about 0.2-0.6 MPa Therefore, low pressure Teflon valves and tubes suffice If higher pressure is required, one should use high performance liquid chromatography tubes and valves For pressure injection the use of pipettes with tip sizes of 0,5-1 um is advised (Figure 2a).

3.3 The galinstan expansion femtosyringe (GEF)

The theory: the flux of liquids in capillaries is described by the Hagen-Poiseuille formula

With the GEF it is not possible to calculate the speed of efflux exactly, because in very thin capillaries diffusion and adhesion effects dominate The equation is useful, though, in helping understand how forces are generated The parameters in the formula are the volume (V), the pressure (p), the viscosity of the liquid (T)), the length of the capillary (1), and time (t) The diameter of the capillary (R) has an exponent of which means that if the parameters are constant, the flux through a capillary with a diameter of |j.m will be 10 000 times faster than that in a capillary of 0.1 nm diameter Thus, any influx of cytoplasm into the pipette will be much slower if one uses a capillary with a small tip To prevent such a backshot (turgor-driven rapid influx of cytoplasm in the capillary that may cause injury to the cell, see below), the tip must be so thin, that the influx into the needle is slower than the potential for water uptake into the cell On the other hand, it requires a very high pressure to expel a sufficient amount of material out of such a small tip Therefore, for the GEF, a new pressure generator was de-veloped First GEF pipette tips are filled with a portion of the substance to be injected, then they are partially filled with some silicon oil, and the rest with the liquid metal alloy galinstan (an alloy of gallium, indium, and tin, Geraberger Thermometerwerke, Germany), which is fluid down to — 20 °C The air-free, filled, pipette is then closed at the rear with a tight-fit epoxy resin-filled glass cap (see ref 5) The pipette then is heated by a temperature regulated air stream and due to the expansion of the liquids a pressure is generated which is sufficient to overcome the resistance of the tip The combined filling of silicon oil (high co-efficient of expansion) and galinstan (low coco-efficient of expansion) enables the user to adjust the temperature dependence of the pressure generated inside As more silicon oil is used, higher pressure is generated due to its higher expansion coefficient Furthermore, injection pressure can be regulated by the temperature of the applied air stream or the diameter of the capillary The system is highly flexible but requires more user expertise than alternative more conventional techniques This is due to the fact that the tip size is below the resolution of the light microscope, which means that the real tip is not visible during injection

The GEF system is most suitable for the introduction of substances in the femtolitre range into small cells or compartments For large cells such as paren-chyma it can be used to introduce substances which are not needed at high concentrations inside the cell (e.g DNA)

4 Cell types

The most important factor one has to consider prior to any experiment, is the nature of the cell to be injected In contrast to animal cells, which are much easier to inject, plant cells are surrounded by a cell wall and are under a sub-stantial Turgor pressure The size and strength of the cell wall, and the turgor pressure generated can vary enormously from one cell type to another The main problem with the cell wall is that if pipette tips are not strong enough to penetrate the wall they can break or bend

However, the major problem is turgor pressure After penetration of the plasma membrane, cytoplasm can be forced back into the tip This is easily visible and is known as the 'backshot' With conventional pressure injection systems, pressure is increased after impalement and the cytoplasm is pressed back into the cell followed by the substance to be injected This procedure can cause arte-facts and cellular disorder (6) Thus, one has to prevent the backshot as effect-ively as possible and some methods will be described below

4.1 Epidermal cells

Epidermal cells have the advantage that they not need a bathing medium When using conventional pressure injection systems, backshot can be prevented if pre-pressure in the injection needle is adjusted such that it is slightly higher than the expected turgor pressure By impaling the cell the injected material will slowly move inside the cell without backshot This method is only applicable if no bathing medium is necessary (e.g epidermal cells) because the substance to be injected will rapidly move out of the tip as soon as it comes into contact with water The success rate of injecting epidermal cells mainly depends on the cuticle as thick cuticular layers often can block the tip of the needle and make injection impossible

Protocol 1

Epifluorescence observation of pressure injection of Alexa 488 into epidermal cells

Equipment and reagents

• Microinjection needle with a tip size about um diameter

• 2% (w/v) Alexa 488 (Molecular Probes) in distilled water Centrifuge the Alexa 488

for 10 so that all undissolved crystals are precipitated This stock can be stored by -20"Cfora longtime

Method

1 Fill the tip with Alexa 488 and backfill the rest with silicon oil (for filling with silicon oil use flexible fused silica capillaries, see Section 6.3)

Protocol continued

3 Install the pipette in the holder and raise pressure to about 0.35 MPa (3,5 bar) Select a blue excitation filter block on the microscope

5 Impale the needle carefully into the edge of a epidermal cell The angle between cell wall and needfe tip should be about 90°

6 If Alexa 488 flows slowly inside the cell, not adjust the pressure If however, it flows in fast, then lower the pressure If a backshot occurs (see Section 4), increase the pressure

4.2 Guard cells and trichomes

As both guard cells and trichomes arc part of" the epidermal layer, pro-pressure can be applied The best place to inject guard cells are at the edges whcro the cells are connected to each other The success rate is much higher in this position than when the middle of the cell is impaled Trichomes lend themselves to injection into the cytoplasm or the-nucleus The nucleus is often located at the side of the cell and is easy to reach On the other hand they have thick cell walls, which in some cases makes injection impossible If the walls are too strong, search for young trichomes often located on the edge of a young leaf, as they have much thinner cell walls

4.3 Mesophyll cells

In general mesophyll cells have very t h i n cell walls and are easy to penetrate One disadvantage is, that they are often of high turgor and easily burst if any vibrations occur during the injection procedure They have to be bathed in a medium and therefore it is not possible to apply a pre-pressure Thus, a backshot will occur during impalement by conventional microelectrodes One can lower the backshot by increasing the osmotic pressure of the bathing medium, for example, with mannitol Experts arc able to inject mesophyll cells by impale-ment through open stomata and in this case no bathing medium is necessary The disadvantage is, that the cells are not easy to observe through the epidermal layer and the preferred technique will depend on each particular experiment

Protocol 2

lontophoretic microlnjection of Lucifer Yellow-CH potassium salt (LYCH) into a mesophyll cell

Equipment and reagents

• Standard iontophoretic injection apparatus

• Filamented capillaries (tip size of about 0.5 (xm)

• 0.5MKCI

Protocol continued

Method

1 Use a leaf from a plant, which allows easy peeling of the epidermis (e.g Viaafaba) and remove an epidermal strip

2 Fasten the pealed leaf with twosided adhesive tape in a small Petri dish, so that the mesophyll is exposed

3 Add 10 mM KC1, mM CaClj, mM NaCl, 125 mM mannitol to the exposed leaf 4 Fill the tip of the capillary with 1% LYCH and the rest with 0.5 M KC1.

5 Fasten the pipette in the holder and connect it to the iontophoretic equipment

6 Adjust the current to about nA

7 Impale the capillary into a mesophyll cell, then give negative pulses of about sec in length and observe the needle tip If fluorescent dye cannot be seen moving out of the tip, increase the current in steps of1 nA

1 Ground parenchyma cells

Parenchyma cells arc in general easy to inject The cell wall causes no problems during impalement but is thick enough to make the cells less sensitive to vibra-tions compared to mesophyll cells Furthermore cell turgor is generally low

4.5 Sieve elements and companion cells

Sieve elements and companion cells cause the biggest problems for microinjection In general they are very narrow and they transport solutes in concentrations up to 1.4 M, the turgor pressure is immense They have also developed defence mechanisms against injury, namely phloem proteins and callose Upon impale-ment by a conventional needle the turgor pressure is released into the tip As a reaction to the injury, the1 phloem proteins are dragged to ihe site of impale-ment and immediately occlude the tip Vciy often the pipette is so tightly blocked thar the cell has to be abandoned Sometimes the tip can be reopened by increas-ing the capillary pressure, but ihen the dye can shoot into the sieve element causing further injury In the author's opinion it is impracticable to inject sieve elements without first lowering their turgor by one of the following two methods:

(a) One can use a bathing medium of high osmotic pressure but the disadvantage of this method is that the osmotic solution has to have a minimum of 600 mOsm This concentration changes the optical properties of the bathing medium so much, that the cells are not easily observed with a microscope (b) A better method is to put the plant in total darkness for about two days until

4,6 Algae

Some algae are very easy to inject However, after impalement they may start to cover the pipette with new wall material In some cases (e.g Mougt'otiu) the needle is blocked within one minute, but with skill this is sufficient timo to introduce enough material into the cell The pipette can then be retracted without any further turgor loss or injuiy and the injected material is trapped inside the cell

Hefore injecting filamentous alga, the cells have to be fastened onto a surface otherwise they are simply pushed around by the needle An easy method for filamentous alga is to melt a 2.5-3% low temperature setting agar and when this has cooled down to about 30-35 DC the alga can be added and mixed into the agar, A drop is then placed onto a slide, and this is flattened with a second slide to produce a thin layer This layer must be very thin otherwise the filaments are not on the surface but inside the agar and movement of an injection needle inside the agar with a manipulator is impossible Unicellular organisms can be held with a holder pipette and this requires a second manipulator on the srnge, A holder pipette is a capillary with n large tip (about fj,m, depending on the size of the cells) This pipette is connected to a tube and by applying suction, the cell is held in place at the pipette tip and can be impaled with the injection pipette

Protocol 3

GEF-mediated injection of 40 kDa dextran-LYCH conjugate into the nucleus of a Mougeotia cell

Equipment and reagents

• Galinstan expansion femtosyringe (see Section 3.3)

• 1% (w/v) 40 kDa dextran-LYCH conjugate (Sigma) in HPLC grade water

Method

2.5-3% low temperature setting agar Capillary with filament with a tip size of about 0.1 jjt.m

1 Fill the tip with the conjugate by applying a drop of the stock to the back end, Fill about one-quarter of the needle with silicon oil

3 Fill the rest of the filamented needle with galinstan by pushing the tip of the filling syringe (flexible fused silica capillary, see Section 6.3) into the silicon oil (this prevents the formation of air bubbles inside the needle)

4 Occlude the needle at the back end with a glass cap and a two component epoxy glue

5 Fasten the algae in low temperature setting agar on a slide and gently flatten to a thin layer with a second slide

Protocol cc

7 Impale the needle into a nucleus of Mvugeotia cell.

8 Leave the needle in position until you can see a sufficient amount of the dextran conjugate in the nucleus

9 Retract the needle out of the nucleus but not out of the cell Within one minute a clot should form around the pipette tip in the cytoplasm After a further minute re-tract the needle This procedure prevents a pressure release from the cell

4.7 Bacteria and organelles

Both bacteria and organelles are too small to access by conventional micro-injection methods As they require very small needle tips, they can be injected with the GEF system (Plate la, Ibl The optical system for their injection must be of a high quality otherwise a controlled impalement of these small structures is not possible Bacteria can be fastened onto a slide in the same way as algae (Section 4.6) and it is possible to inject them down to a size of ^m in diameter Obviously ihe rate of success decreases as the cells get smaller, but the technique is applicable for experiments which not need a high rate of success such as the introduction of DNA for transformation of formerly untransformable species

4.8 Plant tissue cultures

The tell layer of plant tissue cultures that is exposed to air is in general sur-rounded by a thick cuticle and strong walls Injection in this layer is difficult and often impossible Before injecting cells from tissue cultures the callus should be cut into pieces which can be fixed in low temperature setting agar The cells below the outer layer are in general easy to inject Older tissue cultures, which have already started to form leaves, are also readily injected For instance the in-jection into guard cells of young leaves of tobacco tissue cultures turned out to be much easier than the equivalent injection in normal leaves The cells are large and astonishingly insensitive to impalement These guard cells show no lurgor loss if the needle is carefully retracted after injection and they can be observed for a long time afterwards

Protocol 4

GEF-mediated injection of LYCH in a single chloroplast of a guard ceil of a young tissue culture leaf

Equipment and reagents

• GEF apparatus and needle • Tobacco callus producing leaves

Method

Protocol continued

2 Take a leaf from a tobacco tissue culture and identify chloroplasts of the guard cells If they contain too many starch grains (which is often the case) leave cultures for some days in total darkness until there is a significant reduction in starch

3 Fasten a leaf with two-sided adhesive tape onto a slide

4 Fasten the GEF in the holder/heater system and pre-heat the pipette to about 10°C above room temperature

5 Impale the tip into a chloroplast of a guard cell Try to reach a chloroplast at the edge of the cell whilst remembering that the tip itself is invisible,

6 When the chloroplast is filled with the dye, retract the needle slowly and carefully in steps Wait some seconds after each step

7 After retraction the cell and the chloroplast should be visibly intact

- Material suitable for injection

:'.<.'! Fluorochromes

'The injection of fluorochromes is not associated with as many problems as the injection of other substances such as proteins and nucleic acids In general they are water soluble and if a fluorochrome re-crystallizes in the slock solution, crystals can be eliminated by centrifugation Some fluorochromes are toxic, but the main problem is the exciting light All work before injecting a cell, such as the positioning of the needle lip, should be performed under bright field observation Never use the exciting light source for the fluorochrome For instance it is possible to bleach all the chlorophyll of a leaf by using a 40 times lens with a 50 W mercury lamp and blue excitation After one minute a transparent disk is left on the leaf at the point of illumination If living cells are to be observed, it is best to inject in bright field and switch briefly to exciting light in order to con-trol the outcome of the injection Too strong a light source may kill the cell and/or bleach the fluorochrome The light source should always be just sufficient to illuminate the structures of interest and filter sets are available for lowering the intensity of the light source

5.:> Dextran conjugates

weights with temperature changes such as freezing of a stock solution There-fore, care has to be taken storing purified dextrans

5.3 Proteins and antibodies

The concentration of proteins and antibodies injected should always be as low as possible As the quality of injection depends on the protein and its isoelectric point, the pH of the solution should be far from the isoelectric point to make the protein more soluble The same rule holds for the bathing medium Since dif-fusion occurs extremely rapidly over short distances, proteins immediately pre-cipitate in the needle tip and form plugs if the pH of the bathing medium is too close to their isoelectric point Therefore, depending on the protein, a very high or low pH may be necessary, which may cause problems for the cells In these cases, because of the intracellular pH, the pressure must be increased immedi-ately after impalement so that no precipitation and occlusion of the tip can occur

If antibodies and secondary antibodies are injected together, it is necessary to use Fab fragments (see Chapter 10) of the secondary antibody to prevent a cross-linking of the molecules Otherwise the conjugates are of a size that makes it impossible for them to pass through the tip

5.4 Nucleic acids

Plasmids or linear DNA can be large molecules To prevent occlusion of the needle tip, especially if small tips are used, such material should be used at a low concentration As it is generally not necessary to introduce a large amount of DNA into a cell, this represents no problem A concentration of about 3.5 mol-ecules per femtolitre was sufficient to deliver DNA into chloroplasts in which about femtolitre was injected (5) Within 24-30 h expression of green fluor-escent protein (GFP) became visible inside the injected chloroplasts (Plate Id)

6 Tips to make life easier

There are several helpful hints and tips for every-day work in the microinjection laboratory, which are easy, and cheap to set up Some of these will be discussed in this section

6.1 The chuck

Figure Schematic drawings of some helpful equipment, (a) Chuck for the needle holder to prevent movement of the needle tip during pressure injection A nut with a cone (1) can be fastened on the opposite cone with a slot (2) on the head of a screw (3) By turning the nut the opposite cone is compressed against the needle to prevent movement A rubber disc (4) is also compressed against the needle and the needle holder (5) by turning the screw to prevent leakage of pressure between them, (b) Vacuum syringe to delete air bubbles in the needle tip A rubber tube approx 1.5 cm long (1) with a slightly smaller inner diameterthan the outer diameter of the needle is fastened onto a yellow pipette tip (2) and to a 100 ml syringe (3) After fastening the needle onto the rubber tube, a vacuum can be induced by the syringe and the air bubble in the needle tip will expand and leave the tip (c) Filling syringe for microinjection needles A flexible fused silica capillary approx 10 cm long (1) is attached to the metal tip of a cannula (2) and fastened onto the syringe The silica capillary is thin enough to reach the needle tip and is flexible and thus prevents breakage of the needle during filling

head of the screw A nut that contains the opposite cone fastens the needle The chuck has to be made of plastic, otherwise the pipettes will be broken With such a chuck the pipette will stay in place with pressures of up to MPa If more pressure is needed, the glass surface of the needle can be roughened with a diamond-file

6.2 Syringe to delete air bubbles

Sometimes air bubbles appear inside the needle tip after filling with the probe to be injected which cannot be eliminated with the flexible fused silica capillary In such cases a vacuum syringe can help (Figure 3b) Take a 100 ml plastic syringe and fix a yellow pipette tip on the tip Fasten a soft silicon or plastic tube about 1.5 cm long, with a slightly smaller inner diameter than the outer diameter of the capillary, onto the pipette tip so that about cm is left If there is an air bubble inside the tip, put the back end of the needle inside the tube and create a vacuum with the syringe The bubble will expand and be extracted from the tip

6.3 Flexible fused silica capillaries

ex-changes the metal tip with the capillary and fixes it with glue (Figure 3c) The inner diameter of the flexible fused silica capillary should be between 150-250 um This is thin enough to reach the tip of the injection needle and wide enough for an easy filling (remember Hagen-Poiseuille, Section 3.3)

6.4 Petri dishes

Microinjection needles are not only able to inject plant cells! Needles should be always kept covered If one impales oneself with a needle, one will not feel it unless one hits a nerve cell, and tips break easily and can start to move inside the body Furthermore, fluorescent dyes or other substances are often toxic Needles can be fastened onto a strip of blue tack or modelling clay in a 8.5 cm diameter Petri dish until needed for injection

References

1 Kempers, R., Prior, D A M., Oparka, K J., Knoblauch, M., and van Bel, A J E (1998)

Plant Biol, 1,61.

2 Read, N D., Allan, W T G., Knight, H., Knight, M R., Mahlo, R., Russell, A., et al (1992). J Microsc, 166, 57.

3 Purves, R D (1981) Microelectrode methods far intracellular recording and iontophoresis. Academic Press, London

4 Oparka, K J., Murphy, R., Derrick, P M., Prior, D A M., and Smith, J A C (1991).J Cell Set., 98, 539.

5 Knoblauch, M., Hibberd, J M., Gray, J C., and van Bel, A J E (1999) Not Biotech., 17, 906

6 Knoblauch, M and van Bel, A J E (1998) Plant Cell, 10, 35 7 Kempers, R and van Bel, A J E (1997) Planta, 201,195.

8 Goodwin, P B and Cantril, L C (1999) In Plasmodesmata (ed A J E van Bel and W J P. van Kesteren), p 68 Springer, Berlin, Heidelberg

Chapter 7

Micromanipulation by laser

microbeam and optical tweezers

Karl Otto Greulich

Institutfur Molekulare Biotechnologie, Beutenbergstrasse 11, Postfach 100813, 07708 Jena, Germany

1 Introduction

In several chapters at the beginning of this book, techniques were described using light for observation of, and making measurements on cells This chapter will describe techniques, using light not only for observation but also for manip-ulation Here the fact that light is a carrier of energy (heat) and a carrier of momentum (force) is exploited With the heat effect one can perforate plant cells, cut subcellular structures such as cytoplasmic strands or chromosomes, and can fuse plant protoplasts in a controlled manner With the force effect it is possible to hold or move cells and subcellular structures as if light were a pair of tweezers, which in contrast to conventional tweezers can also work in the interior of closed objects Books on laser microbeams and optical tweezers (1) or solely on optical tweezers (2) and a number of reviews (for example refs 3-7) are available in the literature

2 What are laser microtools?

of the order of several hundred microjoules and the pulse repetition rates are between 1-20 pulses per second Other lasers that have been used for micro-beams are Excimer lasers, Excimer pumped dye lasers, or green Neodymium YAG lasers (532 nm) Occasionally visible continuous lasers are used (see however the caveat at the end of Section 3.1)

3 Physical background

3.1 Generating extreme heat

In order to understand if and how a laser microbeam or optical tweezers are suitable for work with living objects, one has to estimate their quantitative thermal effects Since this is particularly critical for laser microbeams, a focused pulsed nitrogen laser with an ultraviolet wavelength of 337 nm and a pulse duration in the nanosecond range is considered For the short pulses of this laser one can assume that the energy is first transferred to the material (within nano-seconds) and the generated heat is dissipated into the environment within microseconds In that sense the biological material can be regarded as being thermally isolated, at least in the first stage of the absorption process Let us assume that a single pulse of microcalorie (or approximately microjoules) is used This is the energy which, when completely converted into heat, would raise the temperature of a millilitre of water by millionth of a degree When this pulse of microcalorie is used to heat up a volume with linear dimensions of a focused laser, i.e of micrometre (or femtolitre), the energy density and thus the heat is a factor of 1012 higher This results in an increase in temperature of a million degrees with uncontrollable effects such as generation of physical plasma, a state of matter where electrons are stripped off from their atoms The tempera-ture rise in this extremely hot spot, however, is very local since it expands approximately at the speed of sound in water (1 km per second or micrometre per nanosecond) After approximately 10-20 nanoseconds it would be expanded so much that its temperature is below 100 °C This time is much shorter than it takes a protein to denature Thus, the laser pulse is highly destructive exactly where it hits a target, but negligible in its effect on the rest of the experimental environment This is the reason why pulsed ultraviolet lasers can be used for highly precise micromachining of subcellular structures without damaging the whole cell Longer wavelengths have lower absorption coefficients and are there-fore less precise Continuous lasers never reach the energy density for plasma generation and thus only the thermal effects, with a plethora of unwanted side effects, can be exploited for biological micromachining

3.2 Why can light be used to move microscopic objects? A watt laser falling on a black object exerts a force of billionths of a Newton

compar-able to ultracentrifugation The force related to this light pressure may be used to balance a cell in the microscope against gravity and thereby fix it spatially An additional effect, however, makes life even simpler Due to 'gradient forces', a dielectric biological object such as a cell or a subcellular particle is pulled into the focus of a light source, even if it has to be pulled against the direction of motion of the light This process can in some ways be compared with a boat sailing against the wind Calculating such effects is similarly difficult for sailing boats and for the gradient forces of light The latter are, for example dependent on the refraction indices of the object and of the medium, and on the beam quality of the focused light To date, nobody has found a way to calculate such forces exactly, and gradient forces are usually quantified semi-empirically by Equations or

P = Q,"I/C 1

and

F = Q.»W/c

where P and F are the light pressure and light force, W and I the laser intensity (power per unit area) and laser power, and Q is an empirical quality factor (8) Gradient forces can act in two directions, axially (against the propagation of light) thus pulling an object against light pressure into the focus and transversally (perpendicular to the light propagation) For jim silica microbeads, the axial value for Q,is of the order of 0.05, the transversal value is approx 0.15 (or 15% of the value for pure light pressure, for which Q.is defined as or 100%) The corres-ponding values for 20 jjim polystyrene microbeads are 0.1 and up to 0.4, i.e 10% and up to 40% of the light pressure value As a rule-of-thumb one can say that gradient forces of optimally focused light pull an object into the focus with a force corresponding to a few per cent of the force which one would calculate for pure light pressure with Q = The lateral effect can amount to up to 40% Since for some types of experiments these Q values are still not sufficiently accurate, sophisticated calibration procedures have been developed to calculate exact values for the forces exerted by light An overview of such calibration procedures can be found in ref

4 How to build laser microtools

4.1 The choice of lasers

lighr pressure Thus, for optical tweezers, NdYAG lasers (1064 nm, see above), NdYLF lasers (1047/1953 nm), InGaAsf1 semiconductor lasers (1330 nm}, and continuous titanium sapphire lasers have primarily been used With respect to the choice or" lasers for optical tweezers, one problem must be considered which is that the interactions of these lasers with biological material are highly de-pendent on the specific wavelength Thus, each new type of laser may require new experimentation to ascertain the extent of possible laser damage to the specimen Best known are the effects of the NdYAG laser at 1064 nm and there-fore a plant cell biologist not interested in sophisticated effects but looking for a workhorse instrument should select an (ideally, taut not necessarily diode pumped) NdYAG laser A watt NdYAG laser is available on the market for approximately $15000 (Coherent, Spectra Physics, ELS)

4.2 Building a laser microbeam or optical tweezers

Laser microbeams and optical tweezers can be constructed, separately or simul-taneously, by mounting a sem[transparent mirror into the fluorescence illumination path of the microscope An cpifluoresconce microscope is recom-mended, i.e a microscope with two illumination paths Almost any fluorescence microscope can be used to build a laser microbeam or optical tweezers, although with standard (upright) fluorescence microscopes there are some minor re-strictions in handling the specimens The only strict pre-condition is that the mirror close to the objective should move with the objective during focusing, since the distance between it and the image side of the objective should remain constant For example, the following microscopes have been used or are sug-gested for use with laser microbeams and/or optical tweezers (Table 1), but this list is not exhaustive and does not mean that other microscopes are not suitable

Table Examples of microscopes to build laser microbeams and optical tweezers

Microscope

Olympus IMT

Zeiss IM 35, Axiovert 135

Leica DMIRB : PL

Objective

Plan xlOO

Neofluar X100

Fluotar X100

Numerical aperture

NA = 1.2

NA = 1.3

NA = 1.3

In addition to or as a replacement for the fluorescence illumination light, both lasers can be coupled into the microscope and focused through the objectives One possible configuration of a laser microbeam instrument or optical tweezers is a microscope with a movable X/Y scanning stage, where the laser beams are focused into the optical plane Such an instrument, with which most of the experiments described below were performed, is depicted in Figure Another solution would be the flying spot version, which is however recommended only for optical tweezers Here, the microscope stage is spatially fixed and the laser beam is moved by a system of mirrors This version is faster, but less versatile than the stage-scanning instrument

5 Applications of laser microbeams in plant biology

Since a laser can easily be focused within the depth of an object, it is possible to work in the interior of an unopened (living) cell In this sense, laser microtools are unique While their application is not restricted to plant biology, with the plant cell the advantages of microtools becomes particularly evident, since:

(a) They often have walls, which are too hard for mechanical glass microtools which often break

(b) They are normally transparent, so the possibility of working in the interior of a living cell can be fully exploited

(c) They often have large intracellular spaces (for example the vacuoles) and therefore organelles can be moved relatively easily

(d) They often have particularly vigorous cytoplasmic particle transport (stream-ing) and laser microbeams combined with optical tweezers can be used to stop, observe, and treat moving organelles

5.1 Laser-induced microinjection

For many applications microinjection by help of a glass capillary, as described in Chapter 6, is the technique of choice There are, however, some inherent prob-lems in using this micromechanical technique (see Table and Chapter 6) which in some cases makes laser microinjection the tool of choice

Table Some problems associated with micromechanical injection of materials into cells

Injection holes are too large and thus the viability of the cell is compromised.

Tough cell walls can cause capillaries to break

Molecules (essentially proteins) to be injected adhere to and may block the capillary

Cells in suspension can be difficult to immobilize and are displaced by the injection capillary

Steric hindrance by adjacent cells can prevent access to the target cell

If large numbers of cells per experiment have to be injected, the laser technique is faster

Microinjection into subcellular structures may be required (see also Chapter 6)

cells in order to generate a plant and to locate the transfected genes in the daughter generation of the whole plant generated from the transfected cells (11) Material has also been injected from the cytoplasm into chloroplasts inside living cells (12)

Protocol 1

Injection of DNA into plant cells

Equipment and reagents

• Microscope with laser rnicrobeam • Physiological buffer solution with 0,4 M attachment mannitol added to increase osmolarity

Method

1 Suspend a few thousand DNA molecules carrying the gene to be transfected (often a whole plasmid) in 10 u.1 of buffer (scaling up to larger volumes is possible, provided sufficient DNA is available)

2 Transfer the cells to be transfected onto a glass slide

3 Add a droplet of the DNA suspension onto the cells to be transfected

4 Focus the laser onto the cell wall and make one laser shot If no hole (dark spot) is observed, increase the laser power, repeat steps and on a second cell If the cell is destroyed, reduce the laser power and repeat steps and

5 When a hole (dark spot) is visible on the cell surface and the droplet size on the cell decreases due to flow of liquid into the cell, remove and culture it to regenerate a callus

5.2 Ablation to study cell fate during plant development

mech-anism of development of Arabidapsis fhaliana roots (13) In this tissue, in contrast to the shoot meristem, it was previously assumed that the cell fate was determined clonally

In a differentiated Arabidapsis root, cells of different size are located close to each other When, for example, a large cortical cell is deleted by laser ablation, the resulting space has to be filled (13) Sometimes, the large cell is replaced by daughter cells from the neighbouring pericycle Two outcomes are possible: either the gap will be filled by two small daughter cells since the mother cells are programmed to produce small daughter cells (clonal fate), or one of the daughters of the small pericycle cell will expand in volume as it grows into a position where previously the large (ablated) cell was located (positional fate) The result of the laser ablation experiment was that the daughter of the small mother cell grew into the volume previously occupied by the ablated cell This contradicted the previous belief that in Arabidopsis root meristem cells are clonally programmed, and demonstrated that they have a positional fate

5.3 Protoplast fusion

When a laser microbeam is slightly defocused it will no longer perforate the plasma membranes of mammalian cells or cut subcellular structures It will however still be strong enough to perturb the cell membrane in a similar way as it would be if it were subjected to micromechanical pricking If two cells are in contact with each other, they may be fused by a short series of laser pulses The contact may be established by adhesive forces or due to high cell density For example, protoplasts of Brassica napus L can be prepared from hypocotyls by en-zymatic digestion The laser is focused onto the areas of the plasma membranes in contact with each other and a series of pulses is released After damaging the protoplast membranes of both cells they will typically fuse within a few seconds Occasionally it will be necessary to change the focal plane slightly in order to hit the membrane In principle, a single pulse can induce fusion, but attenuating the laser and using a series of 10-20 pulses is better for practical purposes It is difficult to find the exact z- position where the laser should be focused on the membranes Therefore, by using a series of pulses and varying the focal position in the z axis the correct location can be found when protoplasts can be seen to start fusing

5.4 Preparation of cell membranes from root hairs

to the environment Such studies have been extended to other plants (15) such as tobacco (Nfmtiflim tahacum), an unicellular green alga (Enmtispfaru viritiis), an angiospcrm (Flodea densa) and lily pollen tubes (Lilitum longiflontm).

8 Applications of optical tweezers to plant biology

Optical tweezers permit the movement of whole cells While this in itself" may be quite useful, for example in order to pick up a cell in suspension and to look at it from different angles, this feature alone may perhaps not be exciting enough to warrant investment in optical tweezers technology More attractive is the possibility to work within the interior of living cells No other microtool allows this, and therefore most experiments with optical tweezers in plant cell biology are on objects inside living cells,

Protocol 2

Basic operation of a laser tweezers set-up

Equipment and reagents

• La zer tweezers set-up • Microbeads, 1-10 um in diameter

Method

1 Set the NdYAG (or other tweezers) laser at a power of 0.3- watt Image a preparation of 1-10 um diameter microbeads on a glass slide

3 Modify the position of the beam expander telescope in the fluorescence illumina-tion path until a micro bead is captured by the laser beam

4 Replace the microbead preparation with cells to be investigated

5 Focus the microscope until a clear image of the subcellular structure to be treated is obtained Due to adjustments in steps and the laser focus is now on the sub-cellular object

6 Hold the object with the laser beam and move the X/Y stage (in equipment with fixed laser beam and moveable stage) or move laser beam (in flying spot versions) to displace the captured object

6 Capturing subcellular organelles for inspection

Figure Particles in a eyloplasmic stream stopped by optical tweeters Only in the bottom right figure a major group of particles can be seen in the vigorous stream of ether particles

not only single particles can be captured and observed, but also whole groups of such structures In this micrograph the background is blurred clue to the vigorous motion of all other particles not held by the tweezers

In l'iv,urt' it can easily be seen that it would normally be impossible to image the particles under these experimental conditions In other experiments with similar strategy chloroplasts in Ulwlm densa have been spatially fixed and observed with a laser scanning microscope (17), subcellular vesicles around the nuclear region of the alga Pyrocysiis noclihiuit: have been moved away from their original position After switching off the laser, the vesicles snapped back into their home position From the speed of this movement, the temperature effect on relative intracelkilar viscosities has been determined (18) Furthermore, groups of sub-cellular structures in rape seed Brassica napus protoplasts which are usually localized in the centre of the cell have been repositioned immediately below the plasma membrane (19)

While in these preliminary experiments no further treatment of these struc-ture was performed, such a technique may be used to inicroinjeci molecules into organelles and then allow them to return to their home position

6,2 Simulating microgravity

are some clues: close to the tips of the rhizoids dense structures, 1-2 um in diameter, can be observed These so-called statoliths, are barium sulfate micro-crystals enveloped by membranes They sediment under the influence of gravity and control the direction of rhizoid growth The statoliths have a higher refract-ive index than the environment and thus can be easily caught and moved by optical tweezers in the interior of living rhizoids (20) It is not only possible to move individual statoliths but up to five of such structures can be collected in the focus of the optical tweezers and held permanently The force required for displacement depends on the direction of displacement towards the cell body (basipetal), in the opposite direction (acropetal), or perpendicularly to the axis of the tube like rhizoid (lateral) If the statoliths are removed from their original position by optical tweezers, the rhizoids no longer grow towards gravity, but grow in a disoriented fashion This result is not too surprising, as their gravita-tional sensing is disturbed More unexpected is the fact that the speed of growth is reduced by almost one order of magnitude and that statolith displacement and tip growth are highly correlated These results provoke some speculation Perhaps evolution has optimized tip growth for Earth's gravitational acceleration and any deviation from this results in suboptimal growth rates? An answer can only be given when other plants have been investigated using similar techniques

7 Conclusion

The use of optical tweezers is still relatively novel in plant cell biology However, with the advent of the exploitation of fluorescent proteins as vital markers for organelles (see Chapter 5) combined with the ability to hold or reposition organ-elles, exciting new data on subcellular organization and on the control and development of polarity should soon become available

References

1 Greulich, K O (1999) Micromanipulation by light in biology and medicine: The laser

microbeam and optical tweezers Birkhauser, Basel, Wien, Boston (Monograph).

2 Sheetz, M P (ed.) (1998) Methods in cell biology, Vol 55 Academic Press, San Diego. 3 Berns, M W., Wright, W H., and Wiegand Steubing, R (1991) Int Rev Cytol, 129,1. 4 Greulich, K O and Weber, G (1992) J Microsc., 167, 127.

5 Weber, G and Greulich, K O (1992) Int Rev Cytol, 133, 1. 6 Schiitze, K and Clement-Sengewald, A (1994) Nature, 368, 667. 7 Ashkin, A (1997) Proc Nat! Acad Sri USA, 94, 4853.

8 Wright, W H., Sonek, G J., and Berns, M W (1994) Appl Opt., 33,173. 9 Svoboda , K and Block, S M (1994) Annu Rev Biophys Biomol Struct., 23, 247. 10 Weber, G., Monajembashi, S., Wolfrum, J., and Greulich, K O (1989) Ber Bunsenges.

Phys Chem., 93, 252.

11 Weber, G., Monajembashi, S., Greulich, K 0., and Wolfrum, J (1990) Isr.] Bot., 40, 115

12 Weber, G., Monajembashi, S., Greulich, K O., and Wolfrum, J (1988) Plant Cell Tissue

13 van den Berg, C., Willemsen, V., Hage, W., Weisbeek, P., and Scheres, B (1995) Nature, 378, 62

14 Kurkdjian, A., Leitz, G., Manigault, P., Harim, A., and Greulich, K O (1993) J Cell Sri., 105, 263

15 de Boer, A H., van Duijn, B., Giesberg, P., Wegner, L, Obermeyer, G., Kohler, W., etal. (1994).Protoplosmo, 178,1

16 Greulich, K O., Harim, A., Leitz, G., Endlich, N., Schliwa, M., Muller, O., etal (1995) Laser microbeam and optical tweezers: Physical principles and application in cell biology and biotechnology Videofilm, Institute for the Scientific Film, Gottingen Germany 17 Greulich, K O., Pilarczyk, G., Hoffmann, A., Meyer zu Horste, G., Schafer, B., Uhl, V.,

etal (2000)J Microsc., 198,187.

18 Leitz, G., Greulich, K O., and Schnepf, E (1994) Bot Acta, 107, 90.

19 Greulich, K O., Bauder, U., Monajembashi, S., Ponelies, N., Seeger, S., and Wolfrum, J (1989) Labor Praxis/Labor, 2000, 36.

20 Leitz, G., Schnepf, E., and Greulich, K O (1995) Plonto, 197, 278.

Table Laser suppliers and approximate prices

Spectra Physics Lasers Inc Diode, NdYAG, TiSa

Coherent Laser Group Diode, NdYAG

Elektronik Laser Systems Diode, NdYAG, TiSa

Laser Science Inc (LSI) Nitrogen

Laser Technik Berlin Nitrogen

Diode and nitrogen lasers from $ / Euro 5000

NdYAG lasers from $ / Euro 15000 per watt

Chapter 8

Electrophysiological methods: monitoring exo- and endocytosis in real time

Gerhard Thiel*, Jens-Uwe Sutler, and Ulrike Homann*

Albrecht von Haller Institut fur Pflanzenwissen schaften, Abteilung Biophysik de Pflanzen, Untere Karspule 2, D-37073 Gottingen, Germany

*Present address Darmstadt University of Technology, Inst of Botany, Schnittspahnstr 3, 64287 Darmstadt, Germany

1 Introduction

About two decades ago Neher and Marty (1) showed that the patch-clamp tech-nique in combination with some impedance analysis could be used for monitor-ing the membrane capacitance as a quasi real time assay for exocytosis and endocytosis This method allows recording of exo- and endocytosis in single animal cells in vivo with such a high resolution, that single exo- and endocytotic events from vesicles with a diameter under 100 nm can be recorded with a temporal resolution in the order of some 10 msec This method has been adopted with success in plant physiology laboratories to examine these processes in plant protoplasts (reviewed in ref 2) Here we aim to present the range of problems which can be addressed with this methodology and provide guidance to the methodological approach

2 Theoretical background

2.1 The membrane is equivalent to a capacitor

Exocytosis involves the fusion of secretory vesicle membrane with the plasma membrane (pm) resulting in a corresponding increase in plasma membrane surface area In the process of endocytosis, membrane is reclaimed from the pm which leads to a decrease in surface area Minute surface area changes can be monitored by measuring changes in membrane capacitance, Cm A biological membrane is in electrical terms a capacitor, the value of which is linearly pro-portional to the surface area of the membrane This relationship is described by:

molecular composition) and thickness of the membrane Thus, under the valid assumption that the thickness of the membrane is nearly invariant and that the dielectric constant is in the short-term not changing, Cm can be used as a quan-titative measure of changes in surface area resulting from exo- and endocytotic activity

2.1 A cell as an equivalent circuit

A spherical cell can be represented by a minimal analogous electrical circuit consisting of a parallel connection of a resistor (Rpm) and a capacitor (Cpm) (Figure 1A) The lipid bilayer of the pm is represented by the capacitor The resistor reflects ionic conductances in the pm due to the presence of channels, pumps, and transporters The total current (I) passing the membrane is given by the sum of the resistive current (Ir) through ionic conductances and the capacitive current (Ic) resulting from the change in the amount of charge separated by the membrane:

I = Ir + Ic

with:

Ir = V/Rpm

Ic = Cpm(dV/dt)

Electrical measurements of Cpm require application of a voltage across the membrane This is generally accomplished by the patch-clamp technique where by an electrode is tightly sealed to the membrane with a resistance Rs > Gfi In this configuration (Figure 1A) the electrical parameters of the little membrane area underneath the patch electrode can be measured (Section 4.3 and Protocol 1). When the membrane patch is ruptured, a low resistant electrical connection (Ra) is established between the recording pipette and the cell interior (Figure 1C) This allows estimation of Cm of the entire cell in the so-called whole-cell configur-ation (Figure and Protocols 2,3).

The experimental recording condition complicates the equivalent circuit by introducing additional resistors and capacitors (Figure IB, C) These can either be compensated by appropriate cancellation circuits in the patch-clamp amplifier (Protocols 1,2) or considered in the analysis of the admittance (Protocol 3).

3 Techniques for the measurement of membrane capacitance

the GLI- system, (a) Microinjection of IYCH in a single cell of the filamentous cyanobactcrium Phormidium

Imiiitiosuni Green represents LYCH, red represents the autofluorescence o! chlorophyll Bar - urn ib)

GEF-mediatcr] injection of LYCH in a single chloroplast of a mcsophyll cell of Vicia fahfi- The reddish bodies arc oiitofiuorcscing chloroplasts of the same cell Bar = Mm- (c) Microinjection of a 70 kDa dextran-l.YCH

conjugale in the nucleus of a Xonopus distal kidney tubule culture cell, Bar = 10 urn (d) Image token 30 h after the injection of the pNtcZ7 plasmid into a single chloroplast of a tobacco leaf rnesophyll cell Green represents the production of green fluorescent protein inside the ctiloroplast Red shows the

autofluorescence of the other chloroplasts Bar • 10 pm

Plate 2a Toluidine Blue stained section from a LR While embedded Brassies nspus silque. Lignilied cell walls are stained shades of blue and non-lignified walls are stained red

Bar - 40 um

Plate 2b Acridine Orange stained section from LR White embedded Brassies napus carpel wall, Lignifted cells are stained yellow/green, pectin is stained red Rar - 20 um

of the Golgi apparatus (greer spols) and cell surface (plasma membrane) stained with a polyclonal antibody to a sugar moiety on glycoproteins in an onion root tip cell The cell has been isolated by a root squash technique (Chapter 10 Protocol 4) and counterstained with propidium iodide to reveal prophase chromosones (b) Confocal immunofluorescencc micrograph of a longitudinal section of a methacrylate embedded maize root tip stained with monoclonal antibody to sugar moiety of glycoproteins showing Golgi staining (green spots) and cell surface labelling of al1 cell types (Chapter 10, Protocol 9) (c) Immunostaining of microtubule arrays by

inimunogold silver enhancement ;5 nm goldi using a anti-B tubulin in mung bean hypocotyl cells and visualized with epipolarization optics The epidermis was peeled and treated as in Chapter 10 Protocol

fluorescence: FITC, yellow/green fluorescence; Cy3, rod fluorescence) (a-c) Metaphase from a rool tip merisiem of a hybrid plant (synthetic; tobacco) between Nicoliaria sylvestris and N tomentosiformis (a) Probed with total digox:genin labelled DNA from N sylvestris in a GISH experiment to determine the parental origin of the chronosones (b) Probed with biotm labelled total genomic DNA f'orn N tomcntosifortitis (c) DAPI staining for DNA Note that chrornosones are appropriately labelled according to parental origin, (d) N.

sylvestris probed with digoxigcnin labelled telomere repeat (TTTAGGG) and biolin labelled 5S rDNA (arrows),

using viroid-specific biotinylated cRMA probes (antiscnsc RNA) and strcptavidin Cy detection (red

fluorescence), (a, b) CEVd (red) in Cucumis sativus (cucumber) leaf tissue (green), (c) HSVd in lycopcrsicum

oRcutcntum (tomato) ;d) HSVd in cucumber, ethidium hromide stained xylern shown as blue, (o) Control,

uninfectetJ cucumber with HSVil-spccific cRNA probe: no HSVd specific signal is seen If) ASBVri in Parses

Rgure (A-C) Minimal equivalent circuit for pm (grey line) of a protoplast (A), for patch-clamp measuring condition in which a patch pipette (thick black lines) is tightly sealed to the pm (= cell-attached configuration) (B), and after rupturing patch underneath the patch pipette (= whole-cell configuration) (C) The electrical symbols have the following abbreviations: CPm = membrane capacitance of whole pm, Rpm = resistance of whole pm, Erev = reversal

voltage of total pm currents, Rp = resistance of membrane patch, Cp = capacitance of

membrane patch, Rs = seal resistance; CPiP = (stray-) capacitance of patch electrode, and

Ra = access resistance produced by rupture of membrane patch (D-F) Sketch of membrane

current responses (lower panel(s)) for configuration (C) upon different voltage stimulations (upper panel) with square (D), saw-tooth (E), and sinusoidal (F) shape lc can be separated from

I, as detailed in the text

3.1 Square-wave stimulation: time-domain technique

A simple way for monitoring of Cm is provided by the time-domain technique where the relaxation of the current to a square-shaped voltage perturbation is observed (Rgure ID) The dielectric polarization of a lipid membrane leads to a current transient in response to a voltage pulse From the decay of the current signal, Cm can be determined by fitting an exponential to the current relaxation

(Figure ID) (for details of analysis see refs 3, 4) Analysis can be done after the

3.2 Saw-tooth stimulation

An interesting alternative to square-shaped stimulation is provided by voltage applications in the form of a saw-tooth (Figure IE) This will cause a current response Ir through the resistive elements which is determined by the thermo-dynamics and the gating properties of the channels/transporters present The capacitor is charged at a constant rate as soon as the polarity of the voltage stimulation is changed The membrane capacitance in a circuit such as in Figure 1C is given by Equation The value dV/dt is known from the setting of the saw-tooth stimulation For high values of dV/dt the capacitive current Ic can easily be measured as an instantaneous current step at the onset of the polarity change

(Figure IE) This approach can be pursued with any conventional patch-clamp

amplifier if there is no great demand in resolution of Cm and if temporal resolu-tion is not critical The advantages and disadvantages are the same as those listed in Section 3.1

3.3 Capacitance cancellation

The capacitance cancellation circuit in a patch-clamp amplifier is aimed to suppress the transient currents that charge Cm during voltage steps Because this current is unwanted in studies of ion channels, it is generally compensated by this cancellation circuit The same procedure can be exploited as a benefit if measurements of Cm are the focus of interest Software-driven patch-clamp amplifiers such as the EPC-9 (HEKA Electronics, Lambrecht, Germany) offer a continuous algorithmic cancellation procedure, the 'cap trac' feature The settings of the cancellation circuitry are automatically and continuously updated and can be loaded on-line into a file providing a near real time measure for Cm A similar facility is offered by the Clampex Membrane Test feature (Axon Instru-ments, Foster City, USA) which also allows continuous sampling of membrane parameters including Cm (5) These kind of measurements have a minimal re-quirement of a software-driven patch-clamp amplifier The technique is robust in its estimation of Cm It is also not sensitive to changes in the seal resistance or access resistance However, the resolution is low so that it cannot be used to detect single fusion events Furthermore, this approach requires on-line calcula-tion of the capacitance cancellacalcula-tion and thus demands considerable computing capacity

3.4 Sinusoidal excitation

The most popular technique for real time high resolution measurements of Cm is the application of a sinusoidal voltage stimulus about a hyperpolarized DC potential The principle of this procedure is the following: a sinusoidal signal of the form V0sincot applied across the circuit presented in Figure 1A evokes a current of the form:

I(t) = (V0sinowt)/Rm + coCmV0coso>t [5]