Signal Transduction: A Practical Approach 2nd ed

Signal Transduction The Practical Approach Series SERIES EDITOR B. D. HAMES Department of Biochemistry and Molecular ...

Author: G. Milligan (Editor)

86 downloads 1570 Views 19MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Signal Transduction

The Practical Approach Series SERIES EDITOR B. D. HAMES Department of Biochemistry and Molecular Biology University of Leeds, Leeds LS2 9JT, UK

See also the Practical Approach web site at http://www.oup.co.uk/PAS * indicates new and forthcoming titles

Affinity Chromatography Affinity Separations Anaerobic Microbiology Animal Cell Culture (2nd edition) Animal Virus Pathogenesis Antibodies I and II Antibody Engineering * Antisense Technology Applied Microbial Physiology Basic Cell Culture Behavioural Neuroscience Bioenergetics Biological Data Analysis Biomechanics—Materials Biomechanics—Structures and Systems Biosensors Carbohydrate Analysis (2nd edition) Cell-Cell Interactions The Cell Cycle Cell Growth and Apoptosis * Cell Separation

Cellular Calcium Cellular Interactions in Development Cellular Neurobiology * Chromatin if Chromosome Structural Analysis Clinical Immunology Complement * Crystallization of Nucleic Acids and Proteins (2nd edition) Cytokines (2nd edition) The Cytoskeleton Diagnostic Molecular Pathology I and II DNA and Protein Sequence Analysis DNA Cloning 1: Core Techniques (2nd edition) DNA Cloning 2: Expression Systems (2nd edition) DNA Cloning 3: Complex Genomes (2nd edition) DNA Cloning 4: Mammalian Systems (2nd edition)

* Drosophila (2nd edition) Electron Microscopy in Biology Electron Microscopy in Molecular Biology Electrophysiology Enzyme Assays Epithelial Cell Culture Essential Developmental Biology Essential Molecular Biology I and II * Eukaryotic DNA Replication Experimental Neuroanatomy Extracellular Matrix Flow Cytometry (2nd edition) Free Radicals Gas Chromatography Gel Electrophoresis of Nucleic Acids (2nd edition) * Gel Electrophoresis of Proteins (3rd edition) Gene Probes 1 and 2 Gene Targeting Gene Transcription if Genome Mapping Glycobiology * Growth Factors and Receptors Haemopoiesis * High Resolution Chromatography Histocompatibility Testing HIV Volumes 1 and 2 * HPLC of Macromolecules (2nd edition) Human Cytogenetics I and II (2nd edition)

Human Genetic Disease Analysis * Immobilized Biomolecules in Analysis Immunochemistry 1 Immunochemistry 2 Immunocytochemistry * In Situ Hybridization (2nd edition) lodinated Density Gradient Media Ion Channels if Light Microscopy (2nd edition) Lipid Modification of Proteins Lipoprotein Analysis Liposomes Mammalian Cell Biotechnology Medical Parasitology Medical Virology MHC Volumes 1 and 2 if Molecular Genetic Analysis of Populations (2nd edition) Molecular Genetics of Yeast Molecular Imaging in Neuroscience Molecular Neurobiology Molecular Plant Pathology I and II Molecular Virology Monitoring Neuronal Activity Mutagenicity Testing * Mutation Detection Neural Cell Culture Neural Transplantation Neurochemistry (2nd edition)

Neuronal Cell Lines NMR of Biological Macromolecules Non-isotopic Methods in Molecular Biology Nucleic Acid Hybridization Typesetter to balance columns Oligonucleotides and Analogues Oligonucleotide Synthesis PCR 1 PCR 2 *PCR3:PCR In Situ Hybridization Peptide Antigens Photosynthesis: Energy Transduction Plant Cell Biology Plant Cell Culture (2nd edition) Plant Molecular Biology Plasmids (2nd edition) Platelets Postimplantation Mammalian Embryos * Post-Translational Modification Preparative Centrifugation

Protein Blotting if Protein Expression Protein Engineering Protein Function (2nd edition) Protein Phosphorylation Protein Purification Applications Protein Purification Methods Protein Sequencing Protein Structure (2nd edition) Protein Structure Prediction Protein Targeting Proteolytic Enzymes Pulsed Field Gel Electrophoresis RNA Processing I and II * RNA-Protein Interactions Signalling by Inositides Subcellular Fractionation Signal Transduction if Transcription Factors (2nd edition) Tumour Immunobiology

Signal Transduction A Practical Approach Second Edition Edited by

G. MILLIGAN Institute of Biomedical and Life Sciences, Division of Biochemistry and Molecular Biology, University of Glasgow, Glasgow

OXFORD UNIVERSITY PRESS

OXFORD UNIVERSITY PRESS Great Clarendon Street, Oxford OX2 6DP Oxford University Press is a department of the University of Oxford and furthers the University's aim of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Athens Auckland Bangkok Bogota Buenos Aires Calcutta Cape Town Chennai Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paulo Singapore Taipei Tokyo Toronto Warsaw and associated companies in Berlin Ibadan Oxford is a registered trade mark of Oxford University Press Published in the United States by Oxford University Press Inc., New York © Oxford University Press 1999 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. Within the UK, exceptions are allowed in respect of any fair dealing for the purpose of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, or in the case of reprographic reproduction in accordance with the terms of licenses issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms and in other countries should be sent to the Rights Department, Oxford University Press, at the address above. This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out, or otherwise circulated without the publisher's prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser Users of books in the Practical Approach Series are advised that prudent laboratory safety procedures should be followed at all times. Oxford University Press makes no representation, express or implied, in respect of the accuracy of the material set forth in books in this series and cannot accept any legal responsibility or liability for any errors or omissions that may be made. A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Signal transduction : a practical approach / edited by G. Milligan. — 2nd ed. p. cm.—(The Practical approach series ; 209) Includes bibliographical references and index. 1. G proteins—Research—Methodology. 2. Cellular signal transduction—Research—Methodology. 3. G proteins—Receptors —Research—Methodology. I. Milligan, Graeme. II. Series. QP552.G16S54 1999 571.6—dc21 99-26199 CIP ISBN 0-19-963721-0 (Hbk) 0-19-963720-2 (Pbk) Typeset by Footnote Graphics, Warminster, Wilts Printed in Great Britain by Information Press, Ltd, Eynsham, Oxon.

Preface Since production of the first Edition of Signal Transduction: A Practical Approach in 1992 areas covered by the term "signal transduction" have expanded greatly in range and number. Furthermore, the widespread application of molecular biology to this field has revealed almost unimaginable diversity and complexity in such systems with large families of proteins involved in the production and destruction of second messenger molecules and information transfer via kinase cascades and their associated regulatory proteins. With such complexity known to occur and with many isoforms of signalling proteins frequently being co-expressed in individual cells, assays to measure second messenger production and regulation and the phosphorylation status and activity state of key kinases has become even more important. Within a single volume it would be impossible to attempt to cover the range of techniques currently employed in this field and indeed even within the Practical Approach series specific volumes have since been produced which concentrate in considerable detail on different methodologies employed to measure and analyse the production of inositol phosphates and related molecules or intracellular [Ca2+] for example. The current book thus seeks to update and expand the areas covered by the first Edition of Signal Transduction: A Practical Approach and to give a broad perspective of techniques currently in use, with particular emphasis of studies of systems regulated by agonists at G protein-coupled receptors (GPCRs). However, as the overlap of systems modified by GPCRs and many other classes of receptors has become virtually all encompassing then the techniques and approaches described should find broad applicability. Glasgow May 1999

G.M.

This page intentionally left blank

Contents List of Contributors Abbreviations 1. Direct assessment of conformational changes in G protein-coupled receptors, using fluorescence spectroscopy Ulrik Gether and Brian Kobilka 1. Introduction 2. Expression and purification of the B2 adrenergic receptor 3. Fluorescence labelling and spectroscopic analysis of the purified B2 adrenergic receptor 4. Site-selective fluorescent labelling of the B2 adrenergic receptor Labelling of cysteines Site-specific labelling at lysines 5. Concluding remarks References

xix xxiii

1 1 2 6 11 12 15 16 16

2. Probing the structure of receptor-binding sites by the substituted-cysteine accessibility method 19 Merrill M. Simpson, George Liapakis, and Jonathan A. Javitch 1. Introduction

19

2. Applications of the substituted-cysteine accessibility method (SCAM) Systems studied with SCAM SCAM versus site-directed mutagenesis and affinity labelling

19 19 20

3. Critical parameters Reactivity of wild-type background Cysteine substitution Determination of function of cysteine-substituted mutants 4. Chemistry of SCAM Mechanism of reaction

21 21 21 23 24 24

Contents Description of reagents Reaction with the MTS reagents Determination of rates of reaction of the MTS reagent Protection of substituted cysteines by bound ligand 5. Interpretation of results Assumptions of SCAM Interpreting the effects of reaction with the MTS reagents Secondary structure Conformational changes associated with receptor activation

30 30 30 31 32

References

33

3. Post-translational acylation of signal transducing polypeptid.es: palmitoylation of G protein-coupled receptors 1.

35

Michel Bouvier, Ulla Petaja-Repo, Thomas Loisel, and Charlene Belanger Introduction

35

2. Detection of receptor palmitoylation Heterologous expression systems Metabolic labelling Characterisation of incorporated radiolabel

36 37 39 45

3. Kinetics of palmitoylation Pulse labelling Pulse-chase labelling Agonist-mediated regulation of receptor palmitoylation

47 47 49 50

4. Assessing the palmitoylation state of the receptor at the cell surface

51

5. In vitro palmitoylation of synthetic peptides

54

References

56

4. G protein-coupled receptor phosphorylation and desensitization

59

Andrew B. Tobin, Angela Rae, and David C. Budd 1. Introduction

59

2. Techniques in the investigation of GPCR phosphorylation Determination of GPCR phosphorylation in intact cultured cells by immunoprecipitation Identification of GPCR phosphorylation in a crude membrane preparation

60 60 64

Contents 3. Identification of desensitization of phospholipase C-coupled receptors References

5. G proteins and their identification Ian Mullaney 1. Introduction 2. Production of crude plasma membrane fractions for analysis of G proteins 3. Gel electrophoresis of G proteins Mono-ADP-ribosylation of G proteins by bacterial exotoxins Gel electrophoresis of G proteins

4. Immunological methods

66 71

73 73 75 77 77 81

87

Immunization and serum collection Immunoblotting and immunoprecipitation

5. Quantification of G protein a subunits 6. Functional aspects of G protein signalling Determination of GTP hydrolysis in membrane preparations Measurement of receptor-stimulated [35S]GTP-yS binding in membrane preparations

Acknowledgements References

87 91

95 98 99 100

101 101

6. Construction and analysis of receptor-G protein fusion proteins Alan Wise 1. Introduction 2. Methods to study G protein function Second messenger production G protein activation Use of pertussis toxin-resistant G protein mutants

3. Receptor-G protein fusions

103 103 105 105 105 105

107

Background Construction of receptor-G protein fusion proteins

4. Expression of receptor-G protein fusions in cultured cells Choice of recipient cell line Choice of vector Transient expression of receptor-G protein fusions Xi

107 109

110 110 111 111

Contents Preparation of DNA-Lipofectamine mix Cell harvesting and plasma membrane production 5. Assays used for functional characterization of receptor-G protein fusions Background Receptor-promoted binding of guanosine-5'-[-y-35thio]-triphosphate ([35S]GTP-yS) Measurement of GTPase activity Receptor binding studies

112 112 113 113 114 115 117

6. Receptor-G protein fusions as research tools 117 Measurement of agonist-induced guanine nucleotide turnover byGi1a 117 Measurement of agonist efficacy 121 Elucidating the role of N-terminal acylation of Gi1a 124 Study of interactions between the A1 adenosine receptor and multiple Gi-family G proteins 128 Receptor-G protein fusion regulation of effectors 132 7. Summary and future perspectives

134

Acknowledgements

135

References

135

7. Application of the baculoviral expression system to signal transduction

139

Andrew Paterson 1. Introduction

139

2. Sf9 cell culture Media Reviving Sf9 cells from frozen Maintaining monolayer cultures of Sf9 cells Adaptation to suspension culture Freezing Sf9 cells Adaptation to serum-free medium

140 141 141 141 142 143 143

3. The baculoviral life cycle, and constructing recombinant baculoviral vectors Time course of viral infection and the polh locus Construction of a recombinant transfer plasmid Construction of recombinant baculovirus by cotransfection Recombinant baculoviral construction with baculoviral shuttle vector Baculoviral passage and titre

144 144 145 145 152 155

4. Recombinant protein expression and purification Assessing a recombinant virus

156 156

xii

Contents Scaling-up infection of Sf9 cells as monolayers Infecting suspension cultures Harvesting suspension cultures Hypotonic lysis Nitrogen cavitation Purification of recombinant protein Rapid purification with glutathione- or Ni2+/NTA-agaroses Alternative cell lines Protein complexes Alternative promoters Further scale-up of culture volume References

159 159 160 160 161 162 163 166 166 167 167 168

8. Reporter gene systems for the study of G protein-coupled receptor signal transduction in mammalian cells 171 Stephen Rees, Susan Brown and Jenny Stables 1. Introduction

171

2. What is a reporter gene? Construction of a reporter gene Reporter proteins

171 173 177

3. Reporter gene systems for the study of GPCR signal transduction Reporter genes for GPCRs which couple to members of the Gas and Gai G protein families Reporter genes for GPCRs which couple to members of the Gaq/11 G protein family Reporter genes for G protein B-y signalling

178 178 180 184

4. Factors influencing the design of a mammalian cell reporter-gene assay Choice of cell line Choice of expression vectors Choice of expression protocol Choice of reporter gene and reporter enzyme Optimization of reporter assay conditions

189 189 190 190 195 195

5. Preparation of cells for reporter-gene assays Agonist assays Antagonist assays Constitutive activity and inverse agonist assays

197 198 199 200

6. Reporter enzyme assays Firefly luciferase Renilla luciferase and dual luciferase assays

201 201 203

xiii

Contents LuFLIPRase Secreted placental alkaline phosphatase (SEAP) Chloramphenicol acetyltransferase B-galactosidase B-lactamase

205 205 207 210 212

7. Reporter protein assays Aequorin GFP

214 214 215

8. Summary and future perspectives

217

Acknowledgements

217

References

218

9. Adenylyl cyclases and cyclic AMP

223

Maurice K. C. Ho and Yung H. Wong 1. Introduction

223

2. Mammalian expression of recombinant ACs DEAE-dextran/chloroquine mediated transfection Adenovirus-mediated transfection Other methods of transfection

226 226 228 233

3. Measurement of intraeellular cAMP level Metabolic labelling of ATP [3H]cAMP assay Isolation of [3H]cAMP Data collection and interpretation

233 234 235 236 239

4. cAMP-responsive bioluminescence assays using firefly luciferase cAMP-responsive transcription of the firefly luciferase gene PKA-responsive luciferase mutant Indirect cAMP assay using luciferase

239 240 243 244

5. Miscellaneous systems for monitoring AC activity cAMP-responsive transcription of the chloramphenicol acetyltransferase gene Other methods of cAMP-induced enzymatic activity Pigment movement in Xenopus laevis melanophores cAMP-induced inward current in Xenopus oocytes Functional rescue of CYR1 -defective Saccharomyces cerevisiae mutant Functional rescue of the Escherichia coli Acya TP2000 mutant

244

6. Perspectives

248

References

249

xiv

244 246 246 246 247 248

Contents

10. Analysis of the polyphosphorylated inositol lipids of Saccharomyces cerevisiae Stephen K. Dove and Robert H. Michell 1. Introduction 2. Structures and nomenclature of inositol glycerophospholipids 3. Analysis of inositol glycerophospholipids: general considerations 4. Radioactive labelling of the phosphoinositides of yeast cells 5. Extraction of inositol lipids from yeast 6. Resolution and identification of yeast phosphoinositides Analysis of polyphosphoinositides by TLC Deacylation of inositol glycerolipids, and HPLC analysis of the resulting water-soluble GroPInsPns

7. Anion-exchange HPLC analysis of inositol lipid-derived GrojPInsPns from yeast HPLC: sample injection and detection of radioactivity Commercially available standards for HPLC analysis of inositol lipid-derived GroPInsPns

References

255 255 256 257 259 263 266 266 271

275 277 278

280

11. Phosphoinositide 3-kinases K. E. Anderson, L. R. Stephens, and P. T. Hawkins 1. Introduction 2. Methods for identifying a role for the PI3K signalling pathway in cellular events Inhibition of PI3K Constitutively active alleles of PI3K and downstream effectors

3. Measurement of the activation of PI3K 4. Measurement of PI3K lipid products in the cell 32

P-labelling Adherent cells [3H]-inositol labelling of cells Separation and quantitation of radiolabelled PI3K lipid products Mass analysis of PtdIns(3,4,5)P3

5. Summary References

283 283 285 285 286

287 290 291 293 293 294 297

298 299 xv

Contents

12. Phospholipase D and phosphatidylcholine metabolism Kathryn E. Meier and Terra C. Gibbs 1. Introduction 2. PLD assays in intact mammalian cells Metabolic labelling of cells with [3H]-fatty acids Incubation of cells with agonists Extraction of phospholipids from intact cells Thin-layer chromatography Quantification of PLD activity Interpretation of results 3. PLD assays with broken-cell preparations Preparation of cell membranes Fluorescent PLD assay Acknowledgements References

13. Signal transduction by sphingosine kinase Dagmar Meyer zu Heringdorf, Chris J. Van Koppen and Karl H. Jakobs 1. Introduction G protein-coupled sphingolipid receptors Receptor regulation of sphingosine kinase and the role of intracellular SPP 2. How to study involvement of sphingosine kinase in a certain signalling pathway 'Positively': testing the intracellular activity of SPP 'Negatively': inhibiting sphingosine kinase 3. How to study regulation of sphingosine kinase Measurement of sphingosine kinase activity in intact cells Measurement of sphingosine kinase activity in subcellular fractions Final remarks References

301 301

309 311 312 317 318

321 321 321 323 325 325 326 327 328 334 335 335

14. Detection, isolation, and quantitative assay of mitogen-activated protein kinases in intact cells and tissues 337 Neil G. Anderson 1. Introduction 2. General mechanism of activation of MAP kinases

xvi

337 339

Contents 3. MAP kinase substrates 4. Testing for activation of known MAP kinases Cultured cells Other sources

339 340 340 342

5. Assessment of MAP kinase activity by gel electrophoresis and immunoblotting Detection of phosphotyrosine in whole cell extracts Mobility shifts Phospho-specific antibodies

6. Measurement of MAP kinase enzyme activity Assay of MAP kinase in partially purified cell extracts Assay of MAP kinases following immunoprecipitation Assay conditions Other MAP kinase assays

7. Determination of MAP kinase subcellular localization and activation by immunocytochemistry 8. Assay of upstream activators of MAP kinases 9. Chemical inhibitors of MAP kinase pathways References

15. Measuring inositol 1,4,5-trisphosphateevoked 45Ca2+ release from intracellular Ca2+ stores Colin W. Taylor and Jonathan S. Marchant 1. Introduction 2. Preparation, permeabilization, and 45Ca2+ loading of hepatocytes Isolation of rat hepatocytes Cell permeabilization, loading of intracellular stores with45Ca2+, and the effects of Ins(l,4,5)P3 45

Acknowledgements References

343 344 348

349 350 353 355 356

356 358 358 359

361 361 362 362 364

2+

3. Rapid kinetic measurements of Ca release from intracellular stores Rapid superfusion methods Rapid superfusion apparatus Rapid responses to Ins(l,4,5)P3

343

368 368 370 378

382 382

Appendix Index

385 389 XVii

This page intentionally left blank

Contributors K. E. ANDERSON

The Babraham Institute, Babraham, Cambridge CB2 4AT, UK. N. G. ANDERSON

Division of Cancer Studies, School of Medicine, University of Manchester, Room G38, Stopford Building, Oxford Road, Manchester M13 9PT, UK. C. BELANGER

Department of Biochemistry and Groupe de Recherches sur le Systeme Nerveux Autonome, Universite de Montreal, P.O. Box 6128, Down-town station, Montreal, Quebec, Canada, H3C-3J7. M. BOUVIER

Department of Biochemistry and Groupe de Recherches sur le Systeme Nerveux Autonome, Universite de Montreal, P.O. Box 6128, Down-town station, Montreal, Quebec, Canada, H3C-3J7. s. BROWN Molecular Pharmacology Unit, Glaxo-Wellcome Research and Development, Stevenage SG1 2NY, UK. D. c. BUDD Department of Cell Physiology and Pharmacology, University of Leicester, University Road, Leicester LE1 9HN, UK. S. K. DOVE Centre for Clinical Research in Immunology and Signalling and School of Biochemistry, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. U. GETHER

Division of Molecular and Cellular Physiology, Department of Medical Physiology 12-5-22, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark. T. C. GIBBS Department of Pharmacology, Medical University of South Carolina, 171 Ashley Avenue, Charleston, South Carolina 29425-2251, USA. p. T. HAWKINS The Babraham Institute, Babraham, Cambridge CB2 4AT, UK. M. K. c. HO Department of Biology, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China.

Contributors K. H. JAKOBS

Institut fur Pharmakologie, Universitat GH Essen, Hufelandstrasse 55, D45125 Essen, Germany. J. A. JAVITCH

Columbia University College of Physicians and Surgeons, Center for Molecular Recognition, P&S 11-401, 630 West 168th Street, New York, New York 10032, USA. B. K. KOBILKA

Howard Hughes Medical Institute, Division of Cardiovascular Medicine and Department of Molecular and Cellular Physiology, Stanford University Medical School, Stanford, CA 94305-5428, USA. G. LIAPAKIS

Columbia University College of Physicians and Surgeons, Center for Molecular Recognition, P&S 11-401, 630 West 168th Street, New York, New York 10032, USA. T. LOISEL

Department of Biochemistry and Groupe de Recherches sur le Systeme Nerveux Autonome, Universite de Montreal, P.O. Box 6128, Down-town station, Montreal, Quebec, Canada, H3C-3J7. J. S. MARCHANT

Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1QJ, UK. K. E. MEIER

Department of Pharmacology, Medical University of South Carolina, 171 Ashley Avenue, Charleston, South Carolina 29425-2251, USA. D. MEYER ZU HERINGDORF

Institut fur Pharmakologie, Universitat GH Essen, Hufelandstrasse 55, D45125 Essen, Germany. R. H. MICHELL

Centre for Clinical Research in Immunology and Signalling and School of Biochemistry, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. I. MULLANEY

Department of Pharmacology, University of Otago, PO Box 913, Dunedin, New Zealand. A. PATERSON

Division of Signal Transduction Therapy, MSI/WTB Complex, University of Dundee, Dundee DD1 5EH, UK.

xx

Contributors U. PETAJA-REPO

Department of Biochemistry and Groupe de Recherches sur le Systeme Nerveux Autonome, Universite de Montreal, P.O. Box 6128, Down-town station, Montreal, Quebec, Canada, H3C-3J7. A.RAE

Department of Cell Physiology and Pharmacology, University of Leicester, University Road, Leicester LE1 9HN, UK. S. REES

Molecular Pharmacology Unit, Glaxo-Wellcome Research and Development, Stevenage SG1 2NY, UK. M. M. SIMPSON

Columbia University College of Physicians and Surgeons, Center for Molecular Recognition, P&S 11-401, 630 West 168th Street, New York, New York 10032, USA. J. STABLES

Molecular Pharmacology Unit, Glaxo-Wellcome Research and Development, Stevenage SG1 2NY, UK. L. R. STEPHENS

The Babraham Institute, Babraham, Cambridge CB2 4AT, UK. C. W. TAYLOR

Department of Pharmacology, University of Cambridge, Cambridge CB2 1QJ, UK. A. B. TOBIN

Department of Cell Physiology and Pharmacology, University of Leicester, University Road, Leicester LE1 9HN, UK. C. J. VAN KOPPEN

Institut fur Pharmakologie, Universitat GH Essen, Hufelandstrasse 55, D45125 Essen, Germany. A. WISE Molecular Pharmacology Unit, Glaxo-Wellcome Research and Development, Stevenage SG1 2NY, UK. Y. H. WONG

Department of Biology, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China.

xxi

This page intentionally left blank

Abbreviations AC AcMNPV ADP AP-1 App(NH)P APS B-gal B2AR B2 BAPTA BEVS BFP BSA BPBt BPC BTK CAM CaM CaMK cAMP CAT cDNA CDTA CFTR CGRP CHO CLM CMV CO2 CRE CREB Cys DAG DBM ddH2O DG DMEM DMSO DTE DTT

adenylyl cyclase Autographa californica multiple nuclear polyhedrosis virus adenosine 5' diphosphate activator protein 1 5' adenylimidodiphosphate ammonium persulphate B-galactosidase adrenergic receptor l,2-Bis(2-amino-phenoxy)ethane-N,N,N',N'-tetracetic acid baculoviral expression vector system blue fluorescent protein bovine serum albumin BODIPY-phosphatidylbutanol BODIPY-phosphatidylcholine (Note: BODIPY is a trademark of Molecular Probes, Inc.) Brutons tyrosine kinase constitutively active mutant calmodulin Ca2+ calmodulin-dependent protein kinase 3',5' -cyclic adenosine monophosphate chloramphenicol acetyltransferase complimentary deoxyribonucleic acid cyclohexanediamine tetraacetic acid cystic fibrosis transmembrane regulator calcitonin gene related peptide Chinese hamster ovary cytosol-like medium cytomegalovirus carbon dioxide cAMP-responsive element cAMP-responivse element binding protein cysteine sn-1-2 diacylglycerol n-dodecyl-B-D-maltoside double-distilled water diglyceride Dulbecco's Modified Eagle's Medium dimethyl sulfoxide dithioerythreitol dithiothreitol

Abbreviations E. coli EC50 EDTA EGTA ELISA ERK FACS FCCP FLIPR fMLP FRET FSH G protein GDP GFP GM-CSF GPCR GRE GRK GroPIns(3,4)P2 GroPIns(3,5)P2 GroPIns(4,5)P2 GroPIns3P GroPIns4P GroPIns5P GroPInsP GroPInsP2 GroPInsPn GST GTP GTPs 3 H-DHA HEK Hepes hGPHa HSV-TK IANBD IBMX ICAM IgG IMAC Ins(l,4,5)P3 IP3

Escherichia coli concentration causing half-maximal effect ethylenediaminetetraacetic acid ethylene glycol bis(B-aminoethyl ether) N,N,N',N'tetracetic acid enzyme-linked immunosorbant assay extracellular signal regulated kinase fluorescence activated cell sorter carbonyl cyanide p-trifluormethoxyphenylhydrazone fluorescence imaging plate reader formylated-met-leu-phe fluorescence resonance energy transfer follicle-stimulating hormone guanine nucleotide binding protein guanosine 5'-diphosphate green fluorescent protein granulocyte macrophage-colony stimulating factor G protein-coupled receptor glucocorticoid-responsive element G protein-coupled receptor kinase glycerophosphoinositol(3,4)bisphosphate glycerophosphoinositol(3,5)bisphosphate glycerophosphoinositol(4,5)bisphosphate glycerophosphoinositol(3)monophosphate glycerophosphoinositol(4)monophosphate glycerophosphoinositol(5)monophosphate glycerophosphoinositolmonophosphate glycerophosphoinositolbisphosphate glycerophosphoinositolphosphates glutathione S-transferase guanosine 5'-triphosphate guanosine 5'-(O-thio)triphosphate [3H]dihydroalprenolol human embryonic kidney N-[2-hydroxyethyl]piperazine-N'-[2-ethane-sulfonicacid] human glycoprotein hormone-a Herpes simplex virus thymidine kinase nitrobenzdioxazol iodoacetamide 3-isobutyl-1-methylxanthine intercellular cell adhesion molecule immunoglobulin G immobilised metal anion chromatography inositol 1,4,5-trisphosphate inositol 1,4,5-trisphosphate xxiv

Abbreviations IPTG IRES IU JNK L-BPBt LPA LPC LSC luc MAP kinase MAP-2 MAPK MAPKK MAPKKK MBP MBS mcs MEK MEM MG MMTV MOI MTS MTSEA MTSES MTSET NAD NEM NFAT NP40 ONPG OPD ORF PA PACAP PAGE PAP PBS PBt PC PCA pCMBS pcor PCR

isopropyl-B, D-thiogalactopyranoside internal ribosomal entry sites international units c-Jun N-terminal kinase lyso-BODIPY-phosphatidylbutanol lysophosphatidic acid lyso-phosphatidylcholine liquid scintillation counting luciferase gene mitogen-activated protein kinase microtubule-associated protein 2 kinase mitogen-activated protein kinase MAP kinase kinase MAP kinase kinase kinase myelin basic protein Mes-buffered saline multiple cloning site MAP kinase/ERK kinase Eagle's minimal essential medium monoglyceride mouse mammary tumour virus multiplicity of infection methanethiosulfonate methanethiosulfonate-ethylammonium methanethiosulfonate-ethylsulfonate methanethiosulfonate-ethyltrimethylammonium nicotinamide adenine dinucleotide N-ethylmaleimide nuclear factor activator of transcription NonidetP40 o-nitrophenyl B-D-galactopyranoside o-phenylenediamide dihydrochloride open reading frame phosphatidic acid pituitary adenylyl cyclase activating peptide polyacrylamide gel electrophoresis phosphatidate phosphohydrolase phosphate-buffered saline phosphatidylbutanol phosphatidylcholine perchloric acid p-chloromercuribenzene-sulfonate p6.9 basic core-associated protein promoter polymerase chain reaction XXV

Abbreviations PDE PDGF PDK PEI PEt PH PDK Pipes PK-A PKB PKC PLA2 PLC PLCB PLD PMSF ppolh p.s.i. PtdIns Ptdlns PtdIns(3)P PtdIns3p PtdIns(3,4)P2 PtdIns(3,5)P2 PtdIns(3,5)P2 PtdIns(3,4,5)P3 PtdIns(3,4,5)P3 PtdIns(4)P PtdIns4P PtdIns(4,5)P2 PtdIns(4,5)P2 PtdIns5P PTX Ro 20-1724 r.p.m. RS RSH RT SAPK SCAM SDS SDS-PAGE SEAP SH2

phosphodiesterase platelet-derived growth factor phosphoinositide dependent kinase polyethyleneimine-cellulose phosphatidylethanol pleckstrin homology phosphoinositide 3OH-kinase piperazine-N,N'-bis[2-ethanesulfonic acid] protein kinase A protein kinase B protein kinase C phospholipase A2 phospholipase C phospholipase CB phospholipase D phenylmethylsulfonyl fluoride polyhedrin promoter pounds per square inch phosphoinositide phosphatidylinositol phosphoinositide(3)phosphate phosphatidylinositol(3)monophosphate phosphatidylinositol(3,4)bisphosphate phosphoinositide(3,5)bisphosphate phosphatidylinositol(3,5)bisphosphate phosphoinositide(3,4,5)trisphosphate phosphatidylinositol(3,4,5)trisphosphate phosphoinositide(4)phosphate phosphatidylinositol(4)monophosphate phosphoinositide(4,5)bisphosphate phosphatidylinositol(4,5)bisphosphate phosphatidylinositol(5)monophosphate pertussis toxin [4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone] revolutions per minute ionized thiolate unionized thiolate room temperature stress activated protein kinase substituted-cysteine accessibility method sodium dodecyl sulphate SDS polyacrylamide electrophoresis secreted placental alkaline phosphatase src-homology region two xxvi

Abbreviations SIE SOS SRE STAT SV40 Tiam-1 T. ni TBAS TC TCA TCF TE buffer TEMED TLC TM Tn7 TRE TRH Tris TTBS UAS X-gal

serum-inducible element son-of-sevenless serum response element signal transducer and activator of transcription Simian Virus 40 T lymphoma invasion and metastasis-1 Trichoplusia ni tetrabutylammonium sulphate tissue culture trichloroacetic acid ternary complex factor Tris-EDTA buffer N,N,N',N'-tetramethylethylenediamine thin layer chromatography transmembrane Transposon 7 TPA response element thyrotropin releasing hormone 2-amino-2-(hydroxymethyl)-l,3-propandiol Tween Tris-buffered saline upstream activating sequence 5-bromo-4-chloro-3-indolyl-B,D-galactoside

xxvii

This page intentionally left blank

1

Direct assessment of conformational changes in G protein-coupled receptors, using fluorescence spectroscopy ULRIK GETHER and BRIAN KOBILKA

1. Introduction G protein-coupled receptors (GPCRs) represent the largest class of transmembrane signalling molecules. The mechanism by which hormones and neurotransmitters activate intracellular signalling cascades through these homologous proteins is of great interest. While GPCRs all share a common seven-helix transmembrane structure, domains involved in ligand binding are nearly as diverse as the chemical structures of the known agonists. Small molecular weight ligands bind to sites within the hydrophobic core formed by the transmembrane (TM) a-helices, while binding sites for peptides and protein agonists also include the amino terminus, and extracellular hydrophilic loops joining the transmembrane domains (1). This diversity in ligand binding suggests that different classes of receptors may have unique mechanisms of activation. However, when one considers that all these receptors function by activating a highly homologous family of G proteins, it is reasonable to propose that agonist binding is linked to G protein activation through changes in the conformation of the transmembrane domains, and that what we learn about receptor activation from one class of receptors will apply to most, if not all, members of this large receptor family. It is generally assumed that binding of the agonist to the receptor induces a set of finely orchestrated changes in the tertiary structure of the receptor, that are recognized by the associated G protein a-subunit. Many methodological approaches have been applied in the attempt to understand these structural changes that provide the critical link between agonist binding and G protein coupling (2). Until recently models for how GPCRs are activated have been based on indirect evidence; hence, the conformation of the receptor has mostly been inferred from activation of messenger systems and/or from computational simulations (3-7). However, the possibility of establishing purification procedures and applying biophysical

Ulrik Gether and Brian Kobilka techniques in the study of this class of receptors has now allowed novel insight into the molecular mechanisms underlying activation of GPCRs (8-12). Most of these studies have been performed on rhodopsin. There are abundant natural sources of rhodopsin, and its inherent stability makes it possible to produce and purify relatively large quantities of recombinant protein. The elegant use of EPR spectroscopy by Hubbell, Khorana, and co-workers has provided the most substantial insight into conformational changes associated with photoactivation of rhodopsin (8-10). Rhodopsin is also the only GPCR for which direct information about the tertiary structure is available. Projection maps at 8A resolution of bovine and frog rhodopsin, based on electron microscopy of two-dimensional crystals, have provided crucial information about the relative positioning of the transmembrane helices in the seven-helix bundle (13, 14). Recently, we have applied spectroscopic techniques to the B2 adrenergic receptor (B2AR) (11, 12, 15-17). As described in this chapter, we have taken advantage of the sensitivity of fluorescent molecules to the polarity of their molecular environment. A sulfhydryl-reactive fluorescent probe was covalently incorporated into the purified 32AR, and used as a molecular reporter for structural changes occurring following agonist binding to the receptor (11, 12, 16). The background for applying spectroscopic approaches was a wish to develop methods that would allow direct, time-resolved analysis of conformational changes accompanying ligand-induced activation of GPCRs. In contrast to rhodopsin, which is a highly specialized GPCR with its ligand covalently bound, the B2AR is a typical ligand-activated receptor. Developing spectroscopic techniques for the B2AR would thus allow analysis of differences between the conformational states of unbound receptors and receptors bound to different kinds of ligands, including full agonists, partial agonists, neutral antagonists, and inverse agonists. The goal of this chapter is to describe the background and methodology involved in using fluorescence spectroscopic techniques for analysing conformational changes in ligand-activated GPCRs. The currently available data will be reviewed and discussed in the context of the experimental procedures.

2. Expression and purification of the B2 adrenergic receptor Compared with other methods for studying receptor structure, such as crystallography and NMR, fluorescence spectroscopy requires relatively small amounts of pure protein (~10-100 picomoles per assay). However, the total amount of receptor required to perform a series of studies using different ligands can exceed 1 mg. Therefore, spectroscopic analysis requires a reliable method for producing and purifying recombinant receptor protein. We have used the baculovirus/Sf-9 cell system to express the 32AR. Sf-9 cells are easy to maintain, they do not require CO2, and they quickly adapt to growth in

1: Direct Assessment of conformational changes suspension cultures in a standard shaker, using either glass or polyethylene Erlenmeyer or Fernbach flasks (see also Chapter 7). The most prominent problem of the insect cell expression system may be the varying fraction of improperly folded, and thus non-functional, protein (18, 19). In the case of the B2AR, approximately half the synthesized receptor is non-functional (19). The fraction can vary for different proteins; however, high levels of expression of several GPCRs in Sf-9 insect cells have been reported (20, 21). We have expressed a modified form of the B2AR in the baculovirus/Sf-9 cell system (18). The amino terminus contains a cleavable influenza-haemagglutinin signal sequence followed by the Ml antibody 'FLAG'-epitope (18). The signal sequence resulted in an approximately twofold increase in expression (18). At the carboxyl terminus the receptor was tagged with six histidines (SF-hB2-6H) (19). A three-step procedure was developed to purify the receptor from the Sf-9 cells, including an initial nickel chromatography step, followed by anti-FLAG immunoaffinity chromatography, and alprenolol affinity chromatography (19). These three steps together ensure that only fulllength and properly folded receptor are purified. It should be noted that for many purposes the immunoaffinity purification step can be omitted (12). Protocol 1. Receptor purification Equipment and Reagents • pVL 1392 baculovirus expression vector (Pharmingen), or equivalent. • Sf-9 insect cells • Baculo Gold transfection kit (Pharmingen), or equivalent • SF900-11 medium (Gibco) • Alprenolol

• Triple-baffled Fernbach flasks (Bellco Glass Inc.) • Chelating Fast Flow Sepharose Resin (Pharmacia) • Flag™ M1 antibody (Eastman Kodak) • [3H] dihydroalprenolol (Amersham)

A. Expression vector and transfection We express the B2AR under control of the polyhedrin promoter, using the pVL1392 baculovirus expression vector (Pharmingen, San Diego, CA). In our hands this vector consistently results in high and reproducible expression. 1. Epitope-tag the cDNA, encoding the human B2AR, at the amino terminus with the cleavable influenza-haemagglutinin signal sequence, followed by the 'FLAG'-epitope (Eastman Kodak, Rochester, NY), and tag the carboxyl terminus with six histidines (SF-hB2-6H) as previously described (18, 19). 2. Co-transfect the vector, containing the cDNA encoding the modified B2AR with linearized BaculoGold DNA, into Sf-9 insect cells, using the BaculoGold transfection kit, according to the manufacturer's instruction (Pharmingen, San Diego, CA).

Ulrik Gether and Brian Kobilka Protocol 1.

Continued

B. Virus amplification and plaque purification

1. Harvest the virus 4-5 days after the transfection, and amplify once before plaque-purification. 2. The plaque-purified viruses are usually amplified three times to obtain 500 ml of a high titre virus stock (about 1 x 109 pfu). 3. Test each virus stock in small-scale cultures to determine the optimal inoculum for the large-scale infections. C. Culturing Sf-9 cells

1. Maintain Sf-9 insect cells in SF900-II medium (Gibco, Grand Island, NY) supplemented with 0.1 mg ml-1 gentamicin (Gibco) and 5% heatinactivated fetal calf serum (Gibco). The optional addition of serum allows the cells to grow to higher densities (7-8 X 106 cells ml-1). 2. Keep the cell stock in 250 ml polypropylene Erlenmeyer flasks (Corning Costar, Acton, MA) at 27°C in a shaker set at 125 r.p.m. Each flask contains 70-100 ml medium, and the cells are kept at a density varying from 0.5-6 x 106 cells ml-1. D. Infection for purification

1. Seed cells in 2800 ml triple-baffled Fernbach flasks (Bellco Glass Inc., Vineland, NJ), and grow until they reach a density of 5-7 x 106 cells ml-1 in a total volume of 1000-1200 ml of medium. 2. Remove the culture from the incubator, and keep the flask at room temperature for 1.5-2 h to sediment the cells. 3. Aspirate most of the medium carefully, and resuspend the cells in fresh medium plus 1 mM of alprenolol, to a cell density of 5 x 10s cells ml-1. 4. Infect cells by adding virus stock (1:30 to 1:100 dilution). Determine the optimal inoculum for each virus stock by infecting small-scale suspension cultures (20 ml in 125 ml disposable Erlenmeyer flasks). 5. Incubate cells for 48 h at 27°C in a shaker set at 125 r.p.m., and harvest by centrifugation for 10 min at 2700 g. The resulting cell pellets can be kept at -70°C until purification. E. Purification

For our spectroscopic analyses we have used the following purification procedure. This procedure has been described in detail previously (19). 1. Lyse one or two pellets of cells from 1000 ml infected cultures in 10 mM Tris-HCI buffer, pH 7.5, containing 1 mM EDTA, 10 (mg ml-1 leupeptin (Boehringer, Mannheim, Germany), 10 mg ml-1 benzamidine (Sigma,

1: Direct Assessment of conformational changes St. Louis, MO), and 0.2 mM phenylmetnylsulfonylfluoride using 100 ml per pellet.

(Sigma),

2. Centrifuge the lysed cells at 45000 g for 30 min, discard the supernatant, and weigh the pellets. 3. Resuspend the pellets in 20 mM Tris-HCI buffer, pH 7.5, containing 1.0% n-dodecyl-B-D-maltoside (DBM) (Anatrace Inc., Maumee, OH), 500 mM NaCI, 10mg ml-1 leupeptin (Boehringer), 10 mg ml-1 benzamidine (Sigma), 0.2 mM phenylmethylsulfonylfluoride (Sigma), and 10-6 M alprenolol (Sigma). 4. Solubilize the resuspended pellets in a Dounce homogenizer (20 strokes with a tight pestle), and stir at 4°C for 1.5-2 h. Use 10 ml of buffer for each gram of lysed cells. 5. Separate non-solubilized particulate from solubilized protein by centrifugation at 45000 g for 30 min. Add imidazole from a 2.0 M stock solution (pH 8.0) to the supernatant at a final concentration of 50 mM. Add Chelating Fast Flow Sepharose Resin (Pharmacia) (0.5 ml of packed resin per gram of lysed cells), charged with nickel and equilibrated in high-salt buffer (20 mM Tris-HCI, pH 7.5, with 500 mM NaCI and 0.08% DBM), and incubate for 2-3 h at 4°C with gentle rotation. 6. Isolate nickel resin by centrifugation for 5 min at 2000 g. Wash the resin once in four times the column volume of high-salt buffer, load onto a column, and wash with three times the column volume of high-salt buffer, and twice the column volume of high-salt buffer containing 25 mM imidazole. Elute in 1/4-column volume fractions with 200 mM imidazole in high-salt buffer. 7. Assay fractions for receptor-binding activity, and pool peak fractions. 8. Add CaCI2 to the pooled fractions at a final concentration of 2.5 mM. Load the pooled fractions onto an M1 antibody column (Eastman Kodak) (0.2 ml per nmol of receptor), equilibrate in low-salt buffer (20 mM Tris-HCI, pH 7.5 with 100 mM NaCI, and 0.08% DBM), and recycle four times by gravity flow. 9. Wash the column with four times the column volume of low-salt buffer containing 2.5 mM CaCI2, and elute using low-salt buffer containing 1 mM EDTA in 1/4-column volume fractions. 10. Analyse the fractions for receptor-binding activity, and pool peak fractions. These two purification steps can produce almost pure protein (specific activity around 5 nmol mg-1 of protein). However, approximately half is non-functional (19). To separate the non-functional receptor from the functional, we use alprenolol affinity chromatography, which is a standard

Ulrik Gether and Brian Kobilka Protocol 1. Continued procedure for purification of the B2AR (22-24). It is important to note that we have been able to omit the M1 immunoaffinity chromatography in some applications. This results in a specific activity of the purified receptor of about 5-10 nmol mg-1, compared with 10-15 nmol mg-1 for the three-step purification. Approximately 5 nmol of purified protein can generally be obtained from a 1000 ml culture. Protein is determined using the detergent-insensitive Bio-Rad DC protein assay kit (Bio-Rad, Hercules, CA). Purified receptor is analysed by classical 10% (w/v) SDS-polyacrylamide gel electrophoresis. Note that samples should not be boiled before loading on to the gel, as this may cause receptor aggregation. The receptor is visualized by standard Coomassie blue staining. F. Binding assay The amount of purified B2AR is assessed in binding assays, using 3Hdihydroalprenolol (3H-DHA) as radioligand (Amersham, Arlington Heights, IL). 1. Incubate purified B2AR (10 ml of an appropriately diluted sample) with 10 nM 3H-DHA (10 ml from a 1:100 dilution of 3H-DHA) in a total volume of 100 ml low-salt buffer (20 mM Tris-buffer, pH 7.5, containing 100 mM NaCI and 0.08 % DBM) for 1 h. Determine non-specific binding in the presence of 10 mM alprenolol. 2. Stop the binding assay, and separate free 3H-DHA from bound by loading the binding mixture onto a 2 ml Sephadex G50 (Pharmacia) column (Poly Prep columns, Bio-Rad). 3. Elute columns directly into 20 ml scintillation vials with 1 ml of ice-cold low-salt buffer. 4. Add scintillation fluid, and count in a scintillation counter.

3. Fluorescence labelling and spectroscopic analysis of the purified (32 adrenergic receptor The emission from many fluorescent molecules is strongly dependent on the polarity of the environment in which they are located. Fluorescent labels incorporated into proteins can therefore be used as sensitive indicators of conformational changes and of protein-protein interactions that cause changes in the polarity of the environment surrounding the probe (25-28). Nitrobenzdioxazol iodoacetamide (IANBD) is a highly fluorescent, cysteineselective reagent (26,27). The fluorescence from IANBD increases as the polarity of the solvent decreases, and is more than tenfold stronger in nbutanol and n-hexane than in aqueous buffer (Figure 1A). There is a parallel 6

1: Direct Assessment of confurmutional changes

Figure 1. Fluorescence properties of IANBD and lANBD-labelled p2AR. (A) emission spectra of cysteine-reacted IANBD (0.3 mM) in solvents of different polarity. Excitation was at 481 nm. (B) emission spectrum of lANBD-labelled B2AR (0.15 mM receptor, and 1.2 mol IANBD per mol receptor). Control is emission spectrum of 0.15 mM B2AR, 'labelled' with IANBD prebound to free cysteine instead of free IANBD, to assess possible nonspecific attachment of the probe to the receptor during labelling. Insert: 10% (w/v) SDSpolyacrylamide gel electrophoresis of lANBD-labelied B2AR. Lane 1, 150 pmol IANBDlabelled B2AR; lanes 2 and 3, 150 pmol B2AR preincubated with iodoacetamide before exposure to IANBD (lane 2) and N-ethylmaleimide (lane 3). Left panel of insert: Coomassie blue staining of gel; right panel of insert: gel photographed under UV tight. The weak band with an apparent molecular weight of 32.5 kDa is a degradation product of the receptor. (Reproduced from Gether et al., ref, 11, with permission).

Ulrik Gether and Brian Kobilka blueshift in the emission maximum from 540 nm in aqueous buffer to 530 nm in n-butanol and 510 nm in n-hexane (Figure 1A). Labelling of B2AR purified from Sf-9 insect cells with IANBD revealed a strong fluorescence signal with an emission maximum at 523 nm (Figure 1B). The blueshift in emission maximum, compared with cysteine-reacted IANBD in aqueous buffer, indicates that the modified cysteine(s) are located in an environment that, on the average, is of lower polarity than n-butanol but higher than n-hexane. This would probably involve labelling of one or more of the five cysteine residues that are located in the transmembrane, hydrophobic core of the receptor (see Figure 3). The covalent modification of the receptor was confirmed by SDSpolyacrylamide electrophoresis of the labelled receptor, and the specificity of the labelling was verified by blocking the incorporation of IANBD with the cysteine-specific, non-fluorescent reagents, iodoacetamide and N-ethylmaleimide (Figure 1B, insert). Importantly, the fluorescent labelling did not perturb the pharmacological properties of the receptor, either in terms of agonist or antagonist binding (11). To examine conformational changes induced by agonist binding, we performed time-resolved spectroscopic analyses (11). As illustrated in Figure 2, binding of the full agonist, isoproterenol, to lANBD-labelled B2AR caused a dose-dependent decrease in fluorescence, reaching a maximum amplitude below the extrapolated baseline after 10 min. The response to isoproterenol could be readily reversed by the active (-)isomer of the antagonist propranolol, but not by the less active (+)isomer (Figure 2). The response to isoproterenol was similarly reversed by several other antagonists, including alprenolol, ICI 118 551, pindolol, and dichloroisoproterenol (11). Moreover, the isoproterenol response was dose-dependent and stereospecific (11). Prior to adding ligand, we normally observe a slight but constant decline in baseline fluorescence (Figure 2). This loss of fluorescence over time is probably caused by bleaching of the fluorophore, combined with some loss of protein possibly due to the protein sticking to the inside of the cuvette. The decrease over time was unaffected by addition of 0.1% bovine serum albumin, 10% glycerol, or phospholipids to the cuvette (11); however, precincubation of the receptor in the cuvette for 15 min before performing the experiments minimized (but never eliminated) the constant decline in baseline fluorescence. It should be noted that the decline in fluorescence is unlikely to be due to denaturation of the protein, since a similar loss of fluorescence also was observed with labelled receptor that was intentionally denatured in guanidinium chloride (11). The observed agonist-induced decrease in fluorescence from the IANBDlabelled receptor is most likely to be due to movement of the fluorophore to a more polar environment upon agonist binding. Importantly, the magnitude of the fluorescence changes was found to correlate with the intrinsic biological efficacy of the ligand, as demonstrated by comparing the effect of a series of partial and full agonists on adenylyl cyclase activity with their effect on the magnitude of the fluorescence changes (11). This suggests that the ligand-

1: Direct Assessment of conformational changes

Figure 2. Reversible decrease in fluorescence from the lANBD-labelled B2AR induced by isoproterenol. (A) Control: addition of water (H2O). (B) and (C), Reversal of the response to isoproterenol (ISO) by the active (-)isomer of the antagonist propranolol, (-)PROP (B), but not by the less active (+)isomer, (+)PROP (C). Dotted lines indicate extrapolated baseline. Excitation was at 481 nm, and emission was measured at 523 nm. Fluorescence in all the individual traces shown was normalized to the fluorescence observed immediately after addition of ligand. All traces shown are representative of at least three identical experiments. (Reproduced from Gether et al., ref. 11, with permission).

induced changes in fluorescence are relevant to the receptor activation mechanism. Protocol 2. Labelling of receptor with fluorescent probes Equipment and Reagents . IANBD (Molecular Probes) • Sephadex G50 (Pharmacia) • Centricon-3-filter (Amicon)

' Chelating Fast Flow Sepharose (Pharmacia) ' Bio-Rad DC protein assay kit (Bio-Rad)

Ulrik Gether and Brian Kobilka Protocol 2. Continued A. Fluorescent labelling of the purified B2AR We have used two different protocols for labelling of the purified B2AR (11,12). Results obtained with the two protocols are indistinguishable; however, we recommend the second procedure since it is more effective at removing non-covalently bound fluorophore from the purified protein (11,12). 1. Incubate purified B2AR(1-1.5 nmol) with 10-15-fold molar excess of IANBD (Molecular Probes, Eugene, OR) (150 (mM) in a total volume of 100 ml buffer (20 mM Tris-buffer, pH 7.5, containing 100 mM NaCI and 0.08 % DBM). IANBD is added from a 10 mM stock solution in DMSO. Allow the reaction to proceed for 1 h at room temperature in the dark, and quench by addition of 1 mM cysteine from a 100 mM stock, followed by 5 min of incubation. 2. Remove cysteine-reacted dye using a Sephadex G50 gel filtration column (0.5 cm X 9 cm). Apply the reaction mixture directly to the column, and elute with 2.0 ml of buffer directly into a Centricon-30 filter device (Amicon, Beverly, MA). 3. Concentrate the eluate, containing the labelled receptor, to approximately 50 ml in the Centricon-30 filter device, fay centrifugation for 45 min at 3000 g in a fixed angle rotor (Sorvall SS-34). Either use the labelled receptor directly for the fluorescence spectroscopy analysis, or store it on ice at 4°C. Under these conditions the protein is stable for several days. 4. Bind purified receptor (up to 5 nmol) to a 150 ml nickel-column by recycling by gravity flow six times (Chelating Fast Flow Sepharose Resin from Pharmacia equilibrated in high-salt buffer, 20 mM Tris-HCI, pH 7.5, with 500 mM NaCI and 0.08% DBM). 5. Perform IANBD labelling by recycling 1.0 ml of 0.5 mM IANBD in highsalt buffer several times over the nickel column for 20 min. 6. Remove excess dye by extensive washing of the column with approximately 50 times the column volume of high-salt buffer. 7. Elute labelled p2AR in 50 ml fractions with 200 mM imidazole in highsalt buffer. Assay fractions for protein content, and pool peak fractions. The labelled receptor can be used directly for the fluorescence spectroscopy analysis, or stored on ice at 4°C. Under these conditions the protein is stable for several days. Both labelling procedures result in incorporation of 1.2-2 mol IANBD per mole of receptor, as determined by measuring absorption at 481 nm and using an extinction coefficient of 21 000 M-1 cm-1 for IANBD and a MW of 50 000 Da for the receptor. Protein concentration was determined using the Bio-Rad DC protein assay kit (Bio-Rad, Hercules, CA). 10

1: Direct Assessment of conformational changes B. Fluorescence spectroscopy analysis Fluorescence spectroscopy is performed at room temperature using a SPEX Fluoromax spectrofluorimeter, connected to a PC equipped with the Datamax software package. We use the photon-counting mode, and generally an excitation and emission bandpass of 4.2 nm (11, 12, 16). 1. Use 30-50 pmol lANBD-labelled receptor for emission scan experiments. Add 10 ml of receptor to 390 ml buffer (20 mM Tris buffer, pH 7.4, containing 100 mM NaCI and 0.08% DBM) in a 5 x 5 mm quartz cuvette, and mix by pipetting up and down. Set the excitation wavelength at 481 nm, and measure the emission from 490 nm to 625 nm with an integration time of 0.3 sec nm-1. 2. Like the emission scans, perform time-resolved fluorescence spectroscopy using 30-50 pmol of labelled receptor. Add 10 ml of receptor to 490 ml buffer (20 mM Tris-buffer, pH 7.4, containing 100 mM NaCI and 0.08 % DBM) in a 5 x 5 mm quartz cuvette. 3. Preincubate the mixture for at least 10 min in the cuvette before the experiment is started to stabilize the baseline. Both during this period and during the time-scan experiment, keep the mixture under constant stirring using a 2 x 2 mm magnetic stirring bar (Bel-Art Products, Pequannock, NJ). 4. During time-scan experiments, set the excitation wavelength at 481 nm, and measure emission at a wavelength of 525 nm. The time scan is routinely performed over 30 min, and the first addition of ligand is usually done after 5 min. The volume of the added ligands is one percent of the total volume, and fluorescence is corrected for this dilution. The compounds tested in our fluorescence experiments have an absorbance of less than 0.01 at 481 nm and 525 nm in the concentrations used, thus eliminating inner filter effects.

4. Site-selective fluorescent labelling of the B2 adrenergic receptor Studies of NBD-labelled B2AR provide direct evidence for ligand-induced conformational changes, but do not identify the nature of the structural changes. To obtain information about the movement of specific receptor domains following agonist activation, it is necessary to determine the site or sites of NBD labelling in the wild-type receptor that are responsible for ligand-induced changes in fluorescence. This requires modifying the receptor structure to limit the amino acids that are susceptible to chemical modification. The most reactive amino acids are cysteines and lysines. It is possible to direct a fluorescent probe to either cysteines or lysines by choosing the

11

Ulrik Gether and Brian Kobilka appropriate chemically reactive fluorophore. However, directing the labelling to a specific cysteine or lysine often requires mutagenesis to limit unwanted labelling sites. In the sections below we discuss site-directed labelling of cysteines and lysines.

4.1 Labelling of cysteines The B2AR contains thirteen cysteines, of which five cysteines are not expected to be available for chemical derivitization. In the extracellular loops, four cysteines (106Cys, 184Cys, 190Cys, and 191Cys) form two disulfide bridges (Figure 3) (29-31), and in the intracellular carboxyl terminal tail 341Cys has been shown to be palmitoylated (32, 33). To identify the cysteine(s) responsible for the agonist-induced change in fluorescence, and thus to establish a system that would allow site-selective incorporation of the IANBD fluorophore, we mutated cysteines in the receptor and generated a series of mutant receptors with one, two, or three cysteines available for chemical derivitization (12). All these mutants displayed minimal changes in pharmacological properties compared with the wild-type, both with respect to ligand binding and functional coupling to adenylyl cyclase (12). However, mutation of several cysteine residues led to a reduction in receptor expression (12). Notably, a mutant receptor with all free cysteines substituted expressed so poorly that purification in sufficient quantities for fluorescence spectroscopy analysis was impossible (12). Ideally, it should be possible to take out all endogenous cysteines, and either reintroduce them one by one, or introduce single cysteines in new positions. Unfortunately this was not possible in the B2AR. Nevertheless, as illustrated in Figure 3 and described below, it is possible to obtain site-specific information from a system where it is not possible to remove all cysteines. The mutant receptors containing one, two, or three of the naturally occurring cysteines were all purified and labelled with the IANBD fluorophore. As expected, the lANBD-labelled mutants all demonstrated emission maxima around 525 nm (12). Time-resolved analysis of the mutants revealed that agonist-induced changes in fluorescence are observed only in receptors in which 285Cys or 125 Cys are present (Figure 3). A mutant lacking only these two Figure 3. The effect of isoproterenol on fluorescence from IANBD labelled wild-type and mutant B2AR (A) 'Snake diagram' of the B2AR. The receptor contains thirteen Cys residues of which five ("Cys, 116Cys, 125Cys, 285Cys, 327Cys) are predicted to be in the transmembrane domain. Three Cys residues are predicted to be in the cytoplasmic regions (265Cys, 375Cys, 406Cys). Five Cys residues are not expected to be available for chemical derivatization (small white circles): four residues (106Cys, 184Cys, 190Cys, 191Cys) form two disulfide bridges (refs 29-31), and in the intracellular carboxy terminal tail, 341 Cys, has been shown to be palmitoylated (refs 32, 33). (B) Bar diagram of changes in fluorescence in response to the full agonist isoproterenol (1 mM) for the wild-type receptor and indicated mutants. The ligand concentration was chosen to ensure

12

1:Direct Assessment of conformtational changes

saturation of the receptors, eliminating any influence from different agonist affinities. Excitation was at 481 nm, and emission was measured at 525 nm. Data are given as percent change in fluorescence (mean ' SE, n - 3-6). The percent change was calculated as the change in fluorescence relative to the extrapolated baseline 15 min after addition of ligand. The cysteine-mutants are named according to the cysteines still present in the receptor and available for chemical derivatization. Thus, Cys(285) describes a construct where 2S5Cys is present but where 77Cys, 115Cys,126Cys, 265Cys, 327Cys, 378Cys and 406Cys have been mutated. 106Cys, 164Cys, 190Cys and 191 Cys were excluded from the 'name', since they are not available for chemical derivatization. (Reproduced from Gether et al., ref. 12, with permission).

13

Ulrik Gether and Brian Kabilka

Figure 4. Simplified model of the B2AR with indication of predicted movements in response to agonist binding, NBD bound to 125Cys in TM III and 286Cys in TM VI are, according to a molecular model of the receptor, predicted to lie at the protein-lipid interface, oriented predominantly towards the lipid as illustrated (the NBD fluorophore is indicated by F) (16). 285Cys-NBD is predicted to be at the helix VI-VII interface, in a boundary zone between the lipid bilayer (or the hydrophobic tails of the detergent micelle) and the more polar interior of the protein (16). An agonist-induced counterclockwise rotation of helix VI, as indicated by the arrow (seen from the extracellular side), would lead to movement of 285Cys-NBD from the nonpolar environment of the lipid bilayer (or detergent micelle) to the more polar environment of the interior of the protein, explaining the observed changes in fluorescence. This is consistent with spin-labelling studies in rhodopsin, which also suggest a counter-clockwise rotation of TM VI upon photo-activation of rhodopstn (10). In TM III, 125Cys-NBD is predominantly exposed to the lipid bilayer (or detergent micelle) (16). Our data indicate that the extent of lipid exposure is changed in response to agonist binding. This is consistent with an agonist-induced movement of TM III, causing the fluorophore to be exposed to a more polar face of TM IV and/or the more polar interior of the receptor, as indicated by the arrow.

cysteines (Cys77, 116, 265, 327, 378, 406) showed no response to agonist binding (Figure 3). The data suggest that agonist binding to the B2AR promotes a conformational change in the receptor that exposes NBD, attached to 125Cys in transmembrane segment (TM) III and to 285Cys in TM VI, to a more polar environment. We have attempted to predict the actual structural changes, using molecular modelling and computational simulations (12). Most importantly, the simulations demonstrated a significant conformational restraint for the NBD bound to 285Cys -NBD and 125 Cys-NBD (12), This suggests that the 14

1: Direct Assessment of conformational changes change in molecular environment around the bound NBD reflects movement of the transmembrane helix to which it is attached, rather than movement of the NBD relative to the transmembrane helix. As schematically illustrated in Figure 4, our data are consistent with a counter-clockwise rotation of helix III and VI in response to agonist binding. This is consistent with the suggested rigid body movements of the corresponding helices in rhodopsin (10). It is interesting to note that 285Cys in TM VI is situated one a-helical turn below 288pro, which is highly conserved among GPCRs and provides a flexible hinge in this helix (12). It has therefore been proposed that the movement of 285CysNBD to a more polar environment in the protein interior could be directly facilitated by this flexible hinge connecting the binding site with the putative G protein coupling domain at the cytoplasmic end of the helix (12). A few issues should be emphasized when interpreting data from our fluorescence spectroscopy analysis. It is important to note that the amplitude of the fluorescent change is only a rough indicator of the magnitude of conformational change. For example, we cannot assume that there is a linear correlation between change in fluorescence and magnitude of movements. Therefore, the movement of TM III, for example, may not be of the same magnitude as movement of TM VI. It is also necessary to ensure that the fluorescent probe, when incorporated into the receptor, does not interfere with binding of the ligands. In the B2AR this is highly unlikely. Labelling of the receptor with IANBD does not alter agonist or antagonist binding properties (11), as would be expected if the bound NBD was positioned within the ligand binding pocket. The results from mutagenesis studies have also provided substantial evidence that amino acids involved in forming the ligand-binding pocket are on a different side of the transmembrane a-helix and one to two a-helical turns closer to the membrane surface than 125Cys and 285 Cys. Our results show that even though we were unable to remove all IANBDreactive cysteines from the B2AR, it was possible to deduce the sites of labelling that were responsible for the agonist-induced fluorescence changes observed in the wild-type receptor. While true site-specific labelling is not possible with IANBD, this may be possible by labelling the B2AR with larger, more hydrophillic cysteine-reactive probes which are unable to gain access to transmembrane cysteines.

4.2 Site-specific labelling at lysines Lysine and arginine are positively charged amino acids that are important in dictating topology of membrane-embedded segments of proteins. In many cases lysines and arginines are functionally interchangeable; however, only lysine has a primary amine and is therefore much more chemically reactive towards isothiocyanates and succinimidyl esters. All basic amino acids in the B2AR are predicted to reside in the hydrophillic domains, or at the boundaries 15

Ulrik Gether and Brian Kobilka of transmembrane helices. We replaced the sixteen lysine residues of the 32AR receptor with arginine, using standard PCR mutagenesis (17). The resulting receptor, referred to as the OK-P2AR, retains the ligand binding and G protein coupling properties of the wild-type B2AR receptor (17). While the OK-32AR receptor is expressed at lower levels than the wild-type B2AR, it is possible to purify sufficient quantities for fluorescence labelling studies (17). To explore movement of a specific receptor domain, it is possible to add back one of the original lysines, thereby permitting site-specific labelling at the single lysine. It should be noted, however, that the amino terminus is also reactive towards isothiocyanates and succinimidyl esters. Thus, receptors having single reactive lysines will also be susceptible to labelling at the amino terminus. To overcome this problem, we have recently constructed a OKB2AR receptor having a specific proteolytic cleavage site in the amino terminus. This permits removal of unwanted amino terminal fluorophore.

5. Concluding remarks Using a sulfhydryl-reactive and environmentally sensitive fluorescent probe, IANBD, as a molecular reporter, we have been able to characterize and map ligand-induced conformational changes in the B2AR. Future studies will attempt to determine more precisely the type of conformational changes that take place following agonist binding. Specifically, we would like to determine how specific transmembrane domains move relative to each other and to the lipid bilayer. To accomplish these goals, it will be necessary to make use of site-specific labelling techniques in combination with spectroscopic techniques, such as fluorescence resonance energy transfer or EPR spectroscopy, two methods which can be used to monitor changes in distance between two receptor domains. These dynamic studies will complement high-resolution structural information when it becomes available.

References 1. Ji, T. H., Grossmann, M., and Ji, I. (1998). J. Biol. Chem., 273, 17299. 2. Gether, U., and Kobilka, B. K. (1998). J. Biol. Chem., 273, 17979. 3. Samama, P., Cotecchia, S., Costa, T., and Lefkowitz, R. J. (1993). J. Biol. Chem., 268, 4625. 4. Luo, X., Zhang, D., and Weinstein, H. (1994). Protein Eng., 7, 1441. 5. Ballesteros, J. A., and Weinstein, H. (1995). Meth. Neurosci. 25, 366. 6. Fanelli, F., Menziani, M. C., and De Benedetti, P. G. (1995). Bioorg. Med. Chem. 3, 1465. 7. Scheer, A., Fanelli, F., Costa, T., De Benedetti, P. G., and Cotecchia, S. (1996). EMBO J. 15, 3566. 8. Farahbakhsh, Z. T., Ridge, K. D., Khorana, H. G., and Hubbell, W. L. (1995). Biochemistry, 34, 8812.

16

1: Direct Assessment of conformational changes 9. Altenbach, C., Yang, K., Farrens, D. L., Farahbakhsh, Z. T., Khorana, H. G., and Hubbell, W. L. (1996). Biochemistry, 35, 12470. 10. Farrens, D. L., Altenbach, C., Yang, K., Hubbell, W. L., and Khorana, H. G. (1996). Science, 274, 768. 11. Gether, U., Lin, S., and Kobilka, B. K. (1995). J. Biol. Chem., 270, 28268. 12. Gether, U., Lin, S., Ghanouni, P., Ballesteros, J. A., Weinstein, H., and Kobilka, B. K. (1997). EMBO J., 16, 6737. 13. Unger, V. M., Hargrave, P. A., Baldwin, J. M., and Schertler, G. F. (1997). Nature, 389, 203. 14. Schertler, G. F., Villa, C., and Henderson, R. (1993). Nature, 362, 770. 15. Lin, S., Gether, U., and Kobilka, B. K. (1996). Biochemistry, 35, 14445. 16. Gether, U., Ballesteros, J. A., Seifert, R., Sanders-Bush, E., Weinstein, H., and Kobilka, B. K. (1997). J. Biol. Chem., 272, 2587. 17. Parola, A. L., Lin, S., and Kobilka, B. K. (1997). Anal. Biochem., 254, 88. 18. Guan, X. M., Kobilka, T. S., and Kobilka, B. K. (1992). J. Biol. Chem., 267, 21995. 19. Kobilka, B. K. (1995). Anal. Biochem., 231, 269. 20. Grisshammer, R., and Tate, C. G. (1995). Quart. Rev. Biophys., 28, 315. 21. Tate, C. G., and Grisshammer, R. (1996). Trends Biotechnol, 14, 426. 22. Benovic, J. L., Shorr, R. G. L., Caron, M. C., and Lefkowitz, R. J. (1984). Biochemistry, 23, 4510. 23. Caron, M. G., Srinivasan, Y., Pitha, J., Kociolek, K., and Lefkowitz, R. J. (1979). J. Biol. Chem., 254, 2923. 24. Parker, E. M., Kameyama, K., Higashijima, T., and Ross, E. M. (1991). J. Biol. Chem., 266, 519. 25. Phillips, W. J., and Cerione, R. A. (1991). /. Biol. Chem., 266, 11017. 26. Dunn, S. M. J., and Raftery, M. A. (1993). Biochemistry, 32, 8608. 27. Gettins, P. G. W., Fan, B., Crews, B. C., and Turko, I. V. (1993). Biochemistry, 32, 8385. 28. Cerione, R. A. (1994). In Methods in enzymology (ed lyengar, R.) Vol. 237, p 409. Academic Press, London. 29. Fraser, C. M. (1989). J. Biol. Chem., 264, 9266. 30. Dohlman, H. G., Caron, M. G., DeBlasi, A., Frielle, T., and Lefkowitz, R. J. (1990). Biochemistry, 29, 2335. 31. Noda, K., Saad, Y., Graham, R. M., and Karnik, S. S. (1994). J. Biol. Chem., 269, 6743. 32. O'Dowd, B. F., Hnatowich, M., Caron, M. G., Lefkowitz, R. J., and Bouvier, M. (1989). J. Biol. Chem., 264, 7564. 33. Mouillac, B., Caron, M., Bonin, H., Dennis, M., and Bouvier, M. (1992). J. Biol. Chem., 267, 21733.

17

This page intentionally left blank

2

Probing the structure of receptor-binding sites by the substituted-cysteine accessibility method MERRILL M. SIMPSON, GEORGE LIAPAKIS, and JONATHAN A. JAVITCH

1. Introduction The substituted-cysteine accessibility method (SCAM) provides an approach to map systematically the residues on the water-accessible surface of a binding site, such as the binding-site crevice of a G protein-coupled receptor (GPCR) which binds ligand within the transrnembrane portion of the receptor. Consecutive residues in putative membrane-spanning segments are mutated to cysteine, one at a time, and the mutant receptors are expressed in heterologous cells. The water-accessible residues are identified by assessing the reaction of charged, hydrophilic, sulfhydryl reagents with the engineered cysteines.

2. Applications of the substituted-cysteine accessibility method (SCAM) 2.1 Systems studied with SCAM Cysteine substitution and covalent modification have been used to study structure-function relationships and the dynamics of protein function in a variety of membrane proteins (1-6). Moreover, charged, hydrophilic, lipophobic, sulfhydryl reagents have been used to probe systematically the accessibility of substituted cysteines in putative membrane-spanning segments of a number of proteins. This approach, SCAM, has been used to map channel-lining residues in the nicotinic acetylcholine receptor (7-9), the GABAA receptor (10, 11), the cystic fibrosis transmembrane conductance regulator (12), the UhpT transporter (13), and potassium channels (14), among

Merrill M. Simpson et al. others. We have adapted this approach to map the surface of the binding-site crevice in the dopamine D2 receptor (15-19), a member of the GPCR superfamily. SCAM can also be used to determine differences in the structures of the membrane-spanning segments in different functional states of proteins, to map electrostatic potential in membrane-spanning domains, and to size a channel or binding-site crevice.

2.2 SCAM versus site-directed mutagenesis and affinity labelling Site-directed mutagenesis is the most common approach to the determination of critical binding site residues. Insights into critical binding domains have also been gained from the construction of chimeric receptors. The interpretation of the functional effects of typical mutagenesis experiments, however, is complicated by the difficulty of differentiating local effects at the site of the mutation from indirect effects of the mutation on protein synthesis, folding, processing, and structure. It is often assumed that functional changes caused by a mutation, such as changes in binding affinity, are due to local effects at the site of the mutation, and not due to indirect effects of the mutation on protein structure. The validity of this assumption is rarely proven for individual mutations, but the structure of the y phage receptor, maltoporin, showed that 50% of the residues, which on the basis of mutagenesis experiments had been implicated in \ phage recognition, are actually buried (20). Thus mutation of these buried residues alters X phage recognition indirectly. Likewise, the crystal structures of several dihydrofolate reductase mutants have demonstrated that a mutation approximately 15 A from the substrate-binding pocket exerts an effect on catalytic activity through an extended structural perturbation (21). Affinity labelling, an alternative approach to the determination of bindingsite residues, has the advantage that the residues identified are likely to be very near the binding site. Disadvantages of this approach, however, include the limited number of affinity reagents available for a particular site, the limited number of residues that can be labelled, and the significant technical difficulties involved in identifying labelled residues. SCAM does not rely on the functional effects of a given mutation, and allows one to determine whether a residue is at the water-accessible surface of the binding-site crevice even when the mutant has near normal function. In contrast to affinity labelling, SCAM can be applied systematically to any binding site, and, because the engineered cysteine is known, the labelled residue does not need to be laboriously identified with protein chemical techniques. Other advantages of the approach include the ability to probe binding sites, by assessing the ability of agonists or antagonists to retard the reaction of sulfhydryl reagents with particular substituted Cys, and the ability to probe the steric constraints and electrostatic potential of sites by comparing the rates of reaction of reagents of varying size and charge. 20

2: Probing the structure of receptor-binding sites

3. Critical parameters 3.1 Reactivity of wild-type background We typically use small, charged, hydrophilic derivatives of methanethiosulfonate (MTS) for SCAM. The function of the protein used as the background for SCAM must not be affected by these polar sulfhydryl reagents. In some cases, such as the [32 adrenergic receptor, endogenous cysteines are not accessible to reaction with the MTS reagents (or reaction causes no functional effect) (22), while in other cases, such as the dopamine D2 receptor, endogenous cysteines are accessible and must first be identified and mutated to other residues (18). We typically substitute serine or alanine for endogenous cysteine, although at particular positions other residues are better tolerated (23). The mutant protein must retain near normal function, and must be relatively insensitive to the MTS reagents. The ideal starting point would be to create a cysteine-less pseudo-wild-type protein with normal expression and function. Such a construct has been possible with the lactose permease (6), the NhaA-Na+/H+ antiporter (24), and a glutamate transporter (25).

3.2 Cysteine substitution Using a wild-type or pseudo-wild-type protein insensitive to the MTS reagents as background, SCAM can now be used to determine the residues that line the water-accessible surface of the protein. Cysteines are substituted, one at a time, for residues in a putative membrane-spanning segment. The mutant receptors are expressed in heterologous cells either transiently or stably. The affinity of ligand binding to intact cells expressing the cysteine-substituted receptors is determined to ensure that the mutant has near wild-type structure. The effect of the MTS reagents on ligand binding is then determined.

Protocol 1. Cysteine substitution Equipment and Reagents • Mutagenesis or PCR kit

Method 1. Generate a series of Cys mutations in the wild-type receptor or the pseudo-wild-type background by site-directed mutagenesis. The oligonucleotides should be designed to incorporate a change in a restriction site, to facilitate screening and verification of mutants after subcloning.

21

Merrill M, Simpson et al. Protocol 1. Continued 2. Screen the mutations by restriction mapping. 3. Confirm mutations by sequencing.

Protocol 2.

Transient transfection of cells

Equipment and Reagents • Dulbecco's modified Eagle's medium/ Ham's nutrient mixture/F-12 (DMEM/F12), containing 3.15 g 1-1 glucose with 10% defined supplemented bovine calf serum (Hyclone)

• Lipofectamine (Gibco) . Optimem (Gibco) * PRSVTag

Method 1. Grow HEK 293 cells in Dulbecco's modified Eagle's medium/Ham's nutrient mixture/F12 (1:1) (DMEM/F12), containing 3.15 g l-1 glucose with 10% defined supplemented bovine calf serum (Hyclone) at 37°C and 5% C02. 2. Co-transfect 35 mm dishes of HEK 293 cells at 70-80% confluence with 1.6 mg of wild-type or mutant receptor cDNA in an appropriate expression vector and 0.4 mg pRSVTag, using 9 ml of lipofectarnine (Gibco) and 1 ml of OPTIMEM (Gibco). 3. Change the medium 5 h after transfection, and again 24 h after transfection. 4. Harvest the cells 48 h after transfection (see Protocol 4).

Protocol 3.

Stable transfection of cells

Equipment and Reagents • DMEM-F12 + 10%BCS (+/- 0.7 mg ml-1 G418 . Lipofectamine (see Protocol 2} • (see Protocol 2) • Optimem (see Protocol 2)

Method 1. Grow HEK 293 cells in DMEM/F12 containing 3.15 mg ml-1 glucose with 10% defined supplemented bovine calf serum (Hyclone) at 37°C and 5% CO2. 2. Transfect 35 mm dishes of HEK 293 cells at 70-80% confluence with 2 mg of wild-type or mutant DNA, in a bicistronic vector expressing the receptor cDNA from a CMV promoter and the neomycin resistance

22

2: Probing the structure of receptor-binding sites

3. 4. 5. 6. 7.

gene from an internal ribosomal entry site (IRES) (26), using 9 ml of lipofectamine (Gibco) and 1 ml of OPTIMEM (Gibco). Change the medium 5 h after transfection. Split the cells into 100 mm plates 24 h after transfection. After another 24 h select in medium containing 0.7 mg ml-1 G418. Change the medium about every 2 days throughout the selection process to remove dying cells. Following selection, maintain the stable cells in medium containing 0.3 mg ml-1 G418.

3.3 Determination of function of cysteine-substituted mutants For further application of SCAM the affinities of the mutants for radiolabelled ligand must be similar to those of the wild-type or pseudo-wild-type receptor to ensure that the structure of the mutant is close to that of the wildtype receptor, and thus that the substituted Cys is an accurate reporter of the accessibility of the wild-type side chain. Affinity for the ligand can be determined by saturation analysis in intact, dissociated cells.

Protocol 4.

Harvesting of transiently or stably transfected cells

Equipment and Reagents • Phosphate-buffered saline (PBS: 8.1 mM NaH2P04, 1.5 mMKH2P04, 138 mM NaCI, 2.7 mM KCI, pH 7.2)

• Binding buffer (140 mM NaCI, 5.4 mM KCI, 1 mM EDTA, 25 mM Hepes, pH 7.4, containing 0.006% bovine serum albumin)

Method 1. Wash the cells with phosphate-buffered saline (PBS: 8.1 mM NaH2PO4, 1.5 mM KH2PO4, 138 mM NaCI, 2.7 mM KCI, pH 7.2). 2. Treat the cells briefly (30-60 s) with PBS containing 1 mM EDTA. 3. Remove the PBS-EDTA, and dissociate the cells in PBS. 4. Pellet the cells at 1000 g for 5 min at 4°C. 5. Resuspend the cells in an appropriate buffer for binding or treatment with MTS reagents.

The ability to substitute cysteine residues for other residues and still obtain functional receptor is central to this approach. In the dopamine D2 receptor, 91 of 96 cysteine-substitution mutants tested to date bound agonist with nearnormal affinity (15-17, 19). These tolerated substitutions were for hydro23

Merrill M. Simpson et al. phobic residues (alanine, leucine, isoleucine, methionine, and valine), polar residues (asparagine, serine, and threonine), neutral residues (proline), acidic residues (aspartate), aromatic residues (phenylalanine, tryptophan, and tyrosine), and glycine. There are several reasons why cysteine substitution may be so well tolerated. Cysteine is a relatively small amino acid with a volume of 108 A; only glycine, alanine, and serine are smaller (27). In globular proteins, roughly half the non-disulfide linked cysteines are buried in the protein interior, and half are on the water-accessible surface of the protein (28). Furthermore, cysteine has little preference for a particular secondary structure (29, 30). A cysteinesubstitution mutant that does not function cannot be studied by SCAM (or by traditional site-directed mutagenesis). The residues that cannot be mutated to cysteine are either accessible and make a crucial contribution to binding, or make a crucial contribution to maintaining the structure of the site and/or to the folding and processing of the protein. The determination by SCAM of the accessibility of the neighbours of a crucial residue may allow us to infer the secondary structure of the segment containing this residue, and thus whether or not it is likely to be accessible as well. If it is not accessible, then the functional effect of its mutation is probably due to an indirect effect on structure.

4. Chemistry of SCAM 4.1 Mechanism of reaction In the membrane-spanning segments of a GPCR, the sulfhydryl group of a native or engineered cysteine will be either at the water-accessible surface of the protein, at the lipid-accessible surface, or in the protein interior. We assume that the surface of the binding-site crevice is part of the water-accessible surface. We further assume that small, charged, hydrophilic reagents react much faster with sulfhydryls facing into the water-accessible surface than with residues facing the lipid or protein interior. Moreover, the thiosulfonate reagents react with the ionized thiolate (RS-) more than a billion times faster than with the unionized thiol (RSH) (31), and only cysteines accessible to water are likely to ionize to a significant extent.

4.2 Description of reagents For such polar sulfhydryl-specific reagents, we use derivatives of methanethiosulfonate (MTS): positively charged MTS ethylammonium (MTSEA) and MTS ethyltrimethylammonium (MTSET), and negatively charged MTS ethylsulfonate (MTSES) (Figure 1) (32). These reagents differ somewhat in size with MTSET > MTSES > MTSEA. The largest, MTSET, fits into a cylinder 6 A in diameter and 10 A long. The MTS reagents form mixed disulfides with the cysteine sulfhydryl, covalently linking -SCH2CH2X, where X is NH3+, N(CH3)3+, or SO3+, The 24

2: Probing the structure of receptor-binding sites

Figure 1. The structure of the methanethiosulfonate derivatives, and their reaction with cysteine.

MTS reagents are specific for cysteine sulfhydryls, and do not react with disulfide-bonded cysteines or with other residues. MTSET is a quaternary ammonium with a fixed charge, and MTSES is fully ionized at neutral pH. Both are membrane-impermeant (24, 33). MTSEA, however, is a weak base and has been reported to enter membrane vesicles readily (33); it presumably crosses the membrane in the unprotonated state. Thus, although MTSEA should react very much faster with ionized thiolates and should, therefore, only react with water-accessible cysteines, the reagent may gain access to these residues by passing through the membrane and then reacting from the cytoplasmic side. Therefore, a residue which is water-accessible from the cytoplasmic side, but not from the extracellular side of the membrane, should not react at an appreciable rate with extracellularly applied MTSET or MTSES, but might react with MTSEA, although the cytoplasmic reducing environment would be expected to scavenge much of the cytoplasmic MTSEA and thereby reduce the apparent rate of reaction. The hydrophilic, negatively-charged, organomercurial p-chloro-mercuribenzene sulfonate (pCMBS) also has been used to probe the accessibility of substituted cysteines in membrane-spanning segments of a number of membrane proteins (13, 24, 34). Protocol 5. Storage and preparation of the MTS reagents Reagents • 2-aminoethyl methanethiosulfonate hydrobromide (MTSEA), MW 236.2 . [2-(trimethylammonium)ethyl] methanethiosulfonate bromide (MTSET), MW 278.2

• sodium (2-sulfanatoethyl) methanethiosulfonate, MW 233.2 (Toronto Research Chemicals, Toronto, ON, Canada)

25

Merrill M. Simpson et al. Protocol 5.

Continued

Method 1. Store the stock reagent desiccated at 4°C, and keep a frequently used stock desiccated at room temperature (RT). 2. The frequently used stock can be replenished from the 4°C stock after waiting for the desiccator to warm to RT before opening. 3. Weigh an appropriate amount of the desired MTS reagent. 4. Immediately before use dissolve the MTS reagents in water.a ' At pH 7 and 22°C, the MTS reagents rapidly hydrolyse with a half-time of 5-20 min (35). Thus, the MTS reagents hydrolyse quickly in buffer. If it is necessary to dissolve the reagents or perform an intermediate dilution in buffer at physiological pH, this should be done immediately before starting the reaction. At lower pH and lower temperature, hydrolysis is appreciably slower. A solution in distilled water is stable for several hours at 4°C.

4.3 Reaction with the MTS reagents Reaction can either be detected directly or indirectly, by measuring the effect of reaction on a functional property of the protein. Because of the very small quantities of protein produced in most heterologous expression systems, we cannot rely on the direct detection of reaction. Instead, we use the irreversible modification of function to assay the reaction (Figure 2). Additionally, reaction with a cysteine near a binding site should be retarded by the presence of inhibitor and/or substrate.

Protocol 6. Reaction of MTS reagents with substituted cysteines Equipment and Reagents • MTS reagant (see Protocol 5) • Binding buffer (see Protocol 4}

• Glass fibre filters (grade 934AH) (Brandel, Inc.)

Method 1. Add 5 ml of the appropriate concentration (10x) of freshly prepared MTS reagent into aliquots (45 ml) of cells suspended in binding buffer (140 mM NaCI, 5.4 mM KCI, 1 mM EDTA, 25 mM Hepes, pH 7.4, containing 0.006% bovine serum albumin).a 2. Mix cells and reagent, and incubate at room temperature for 2 min. 3. Dilute the cell suspensions 10-20-fold in binding buffer to slow the reaction.b 4. Use 200 ml aliquots to assay for radioligand binding as appropriate.

26

2: Probing the structure of receptor-binding sites 5. Separate the bound from free radioligand by filtration through glass fibre filters. 6. Count the binding samples by liquid scintillation spectrometry. 7. The fraction of inhibition can be calculated as follows: 1- [(specific binding after MTS reagents)/(specific binding without reagent)] a For screening, use final concentrations of 2.5 mM MTSEA, ImM MTSET, and 10 mM MTSES to normalize for the intrinsic reactivities of the reagents with sulfhydryls in solution (32). b Do not attempt to stop the reaction with a reducing agent such as dithiothreitol or 2mercaptoethanol, as these will reduce the newly formed disulfide bonds between the MTS reagents and the substituted cysteine. The reaction should also not be quenched with solutions containing free sulfhydryls, as these can undergo disulfide exchange with the newly formed disulfide bond. Thus the concentration of the MTS reagents must be decreased by dilution, centrifugation, or filtration.

4.4 Determination of rates of reaction of the MTS reagent For the Cys mutants which are inhibited by an MTS reagent, the rate of reaction can be calculated by determining the effect of reaction at several different concentrations of reagent. The extent of reaction is taken to be the extent of inhibition of binding after a fixed time, typically 2 min, with 5-6 concentrations of reagent. The fraction of initial binding, Y, is fitted to (span*e-kct) + plateau, where k is the second-order rate constant (inM-1S-1),c is the concentration of MTS reagent, t is the time (120 s), plateau is the residual binding at maximal concentration of the MTS reagent, and (span + plateau) = 1. A number of factors, including steric constraints and electrostatic potential, can contribute to differences in the rates of reaction of different substituted Cys, or of a given substituted Cys with different reagents (35). An affinity of the reagent for a site near the substituted Cys can also increase the rate of reaction by boosting the local concentration of reagent. We have observed this with MTSEA at particular positions in the D2 receptor, presumably because of an affinity of the ethylamine moiety for the dopamine binding site (17). We have also observed this with MTSET at a number of positions, and this effect can be explained by an interaction of the hydrophobic cation of MTSET with nearby aromatic groups in the receptor, which increases the local concentration of MTSET (15, 19).

4.5 Protection of substituted cysteines by bound ligand Reaction with MTS reagents of substituted cysteines at or near the binding site should be retarded in the presence of bound ligand. To assess the ability of substrate or inhibitor to retard the rate of reaction of the MTS reactions, cells are incubated in the presence or absence of ligand, and then MTS reagents are added in the continued presence or absence of ligand. 27

Merrill M. Simpson et al.

Figure 2. Schematic representation of the reaction of the MTS reagents with a cysteine accessible in the binding-site crevice. The membrane is represented by a stippled rectangle; the binding-site crevice is indicated within the plane of the membrane, and the oval represents ligand. SSEX represents -SSCH2CH2X (where X is NH3-, N(CH 3 ) n ', or SO3-), which is covalently linked to the water-accessible cysteine sulfhydryl. In the bound state, represented in the upper right panel, ligand is reversibly bound within the bindingsite crevice. In the unbound stale, represented in the upper left panel, the binding site is unoccupied. After irreversible reaction with MTSEX, represented in the middle left panel, ligand binding is altered, as is shown in the bottom left panel. The cysteine sulfhydryl facing lipid or the interior of the protein does not react with MTSEX. Ligand retards the rate of reaction of receptor with MTSEX (middle right panel), thereby protecting subsequent ligand binding (bottom right panell.

Protocol 7.

Protection of substituted Cys by bound ligand

Equipment and Reagents • MTS reagant (see Protocol 6) . Binding buffer (see Protocol 4)

• 96-woll multiscreen plate with GF/B filters (Millipore)

Method 1. Pre-wet with 100 ml binding buffer a 96-well multiscreen plate containing GF/B filters (Millipore). 2, Remove the buffer by filtration under vacuum. 28

2: Probing the structure of receptor-binding sites 3. Add dissociated cells (100 ml) and 50 ml binding buffer, containing ligand or buffer, and incubate for 20 min at RT. 4. Add the appropriate MTS reagent (50 ml of 4x)a in the continued presence or absence of ligand, and incubate for 2 min at RT. 5. Stop reaction by removing the reagents by filtration. 6. Add 250 ml binding buffer per well. Gently mix the plate for 5 min on an orbital shaker at RT and then filter. 7. Repeat step 6 three times.b 8. Add 100 ml binding buffer, and 50 ml of cold ligand in buffer (for the determination of non-specific binding) or buffer (for the determination of total binding) per well. 9. Shake the plate for 5 min at RT to resuspend the cells, and add 100 ml of radioligand per well. 10. Incubate as appropriate, separate the bound from free radioligand by filtration, and wash twice with cold wash buffer. 11. Punch the filters into vials, and measure the radioactivity by liquid scintillation spectrometry.c 12. Protection is calculated as 1 - [(inhibition in the presence of ligand)/ (inhibition in the absence of ligand)]. a

It is important to note that this protocol examines the ability of a reversible reaction to slow an irreversible reaction. Thus, too high a concentration of reagent or too long a time of reaction will obscure the presence of protection. To facilitate determination of a change in the rate of reaction, the concentrations of sulfhydryl reagent should be chosen to produce, in the absence of ligand, approximately 70% of the maximal effect. The concentration of protecting ligand should be approximately 1000-fold greater than its K1 in each particular mutant. b The protecting ligand should be relatively hydrophilic to facilitate complete removal prior to the determination of residual binding. Controls testing for inhibition by residual protecting ligand without MTS reagent should be included. c The hydropure membrane support in Millipore 96-well filter plates avidly binds many small molecule radioligands. Thus, it is imperative to determine if the radioligand used binds to the hydropure backing. If so, the glass fibre filters must be counted without the backing. This is facilitated by complete drying of the filters prior to punching.

A decrease in the rate of reaction of a substituted Cys with an MTS reagent in the presence of ligand is most simply explained by steric exclusion. Not every residue which is protected, however, need contact ligand, as the ligand might protect residues deeper in the binding-site crevice by binding above them and preventing access of the MTS reagents deeper in the crevice. It is also difficult to rule out indirect conformational changes in the mechanism of protection. Such conformational changes are less likely to be caused by an antagonist than by an agonist, but are nonetheless still possible. 29

Merrill M. Simpson et al.

5. Interpretation of results 5.1 Assumptions of SCAM To interpret the results of SCAM we make the following assumptions: (1) the highly polar MTS reagents react much faster at the water-accessible surface of the protein than in lipid or in the protein interior; (2) in membrane-spanning segments, access of highly polar reagents to side-chains is only through the binding-site crevice; (3) the addition of -SCH2CH2X to a cysteine at the surface of the binding site alters binding irreversibly, and, reciprocally, (4) that for substituted cysteines that line the binding site, antagonists or agonists should retard the reaction with the MTS reagents.

5.2 Interpreting the effects of reaction with the MTS reagents The effects of the addition of -SCH2CH2X to the engineered cysteine could be a result of steric block, electrostatic interaction, or indirect structural changes. Thus, reaction could inhibit or potentiate binding. Regardless, an irreversible effect is evidence of reaction, and, therefore, of the accessibility of the engineered cysteine. This can be illustrated in the dopamine D2 receptor by the mutation of Asp108 (17). Mutation to cysteine of this residue at the extracellular end of the third membrane-spanning segment reduced the receptor's affinity for antagonist binding about threefold. Reaction of the positivelycharged MTSEA or MTSET at this position significantly inhibited binding. In contrast, reaction of the negatively charged MTSES restored the negative charge at this position, and shifted the affinity towards that of wild-type receptor, thereby increasing occupancy and potentiating binding. The fact that reaction can potentiate function necessitates care in experimental design; a potentiation of binding could be missed by measuring binding at too high a ligand concentration relative to the KD (36). In the dopamine D2 receptor, we have observed that reaction of MTSEA at certain positions has a much greater effect on the binding of particular ligands: for example, reaction of MTSEA with the highly reactive endogenous cysteine, Cys118, caused only a negligible decrease in the affinity for the antagonist haloperidol, but caused a 3000-fold decrease in the affinity for the antagonist sulpiride (37). If reaction of a substituted cysteine with an MTS reagent results in a negligible effect on the affinity of binding, it might be inferred that the residue was inaccessible, resulting in a false negative determination. Thus, while it seems unlikely that a residue forming the surface of the binding-site crevice could be covalently modified by the addition of the charged -SCH2CH2X without interfering with binding, such a result is possible. Moreover, a residue which is water-accessible might not react with the reagents due to steric factors. These potential complications demonstrate the importance of systematically mutating to cysteine consecutive residues along 30

2: Probing the structure of receptor-binding sites an entire membrane-spanning segment; while mutation of any individual residue might be subject to potential misinterpretation due to silent reaction or steric factors, this is unlikely to be a systematic problem affecting the overall pattern of accessibility of multiple residues in a membrane-spanning segment.

5.3 Secondary structure To infer a secondary structure, we must assume that, if binding to a mutant is not affected by the MTS reagents, then no reaction has occurred, and that the side chain at this position is not accessible in the binding crevice (as discussed above, however, this may not universally be so). In an a-helical structure one would expect the accessible residues to form a continuous stripe when the residues are represented on a helical net. For example, in the third membrane-spanning segment of the dopamine D2 receptor, the pattern of accessibility is consistent with this membrane-spanning segment forming an ahelix with a stripe of about 140° facing the binding-site crevice (Figure 3) (17). In contrast, in a B-strand, one would expect every other residue to be accessible to the reagents. More complex or irregular patterns of accessibility can be more difficult to interpret, but these findings can also be rather informative, and suggest

Figure 3. Helical wheel (left) and helical net (right) representations of the residues in and flanking the third membrane-spanning segment of the dopamine D2 receptor, summarizing the effects of MTSEA on [3H]YM-09151-2 binding. Reactive residues are represented by squares, where the fill indicates the range of the second-order rate constants in M-1S-1 for reaction with MTSEA: solid squares = k > 20; hatched squares = 20 > k > 10; striped squares = 10 > k > 3; open square = 3 > k > 1. Small open circles indicate that MTSEA had no effect on binding. The solid circle indicates no binding after cysteine substitution. D108 and 1109 are represented outside the a-helix in the loop from the second membrane-spanning segment.

31

Merrill M. Simpson et al. the presence of kinks, twists, and dynamic changes in the structure of the membrane-spanning segments (15,16,19, 22).

5.4 Conformational changes associated with receptor activation Conformational changes in a protein may result in changes in the accessibility of substituted cysteines as assessed by their rates of reaction with polar sulfhydryl-specific reagents. For example, residues lining the channel of the nicotinic acetylcholine receptor change in accessibility upon activation of the receptor and opening of the channel (7, 8). Similarly, it should be possible to determine changes in the accessibility of residues in GPCRs in different functional states. To identify activation-induced structural changes in the residues forming the surface of the binding-site crevice, we sought to determine the relative accessibilities of a series of engineered cysteines in the resting and activated receptor. Agonist cannot be used to activate receptor, however, because the presence of a ligand within the binding site would interfere with access of the MTSEA to the engineered cysteines. Alternatively, the activated state of the receptor can be achieved by using a constitutively active mutant (CAM) receptor as a background for further cysteine substitution. A CAM receptor is intrinsically active, and has a higher affinity for agonist than does the wildtype receptor (38). The high affinity state for agonist is typically associated with the activated receptor-G protein complex. That agonist affinity is higher in the CAM even in the absence of G protein suggests that the structure of the binding site of the CAM is likely to be similar to that of the agonist-activated wild-type receptor binding site (or more easily isomerizes to the active state). Thus, we can compare the resting and active forms of the receptor by determining the accessibility of substituted cysteines in the binding-site crevice in these two states, using wild-type receptor and a CAM as background constructs. MTSEA had no effect on the binding of agonist or antagonist to wild-type P2 receptor expressed in HEK 293 cells. This suggested that no endogenous cysteines are accessible in the binding-site crevice. In contrast, in the CAM B2 receptor, MTSEA significantly inhibited antagonist binding, and isoproterenol slowed the rate of reaction of MTSEA (22). This implies that at least one endogenous cysteine becomes accessible in the binding-site crevice of the CAM B2 receptor. Cys285, in the sixth transmembrane-spanning segment (TM6), is responsible for the inhibitory effect of MTSEA on ligand binding to the CAM (22). The acquired accessibility of Cys285 in the CAM may result from a rotation and/or tilting of TM6, associated with activation of the receptor. This rearrangement could bring Cys285 to the margin of the binding-site crevice, where it becomes accessible to MTSEA. Such a movement of TM6 upon receptor activation is 32

2: Probing the structure of receptor-binding sites consistent with the results of fluorescence spectroscopy studies in the B2 receptor (39, 40) and spin-labelling studies in rhodopsin (41), and suggests that the substituted-cysteine accessibility method in a CAM background is a powerful approach for probing conformational change in these receptors. The membrane-spanning segments not only form the binding-site crevice, but also constitute the transduction pathway from the binding site to the intracellular loops which interact with G protein. Agonist binding in the crevice must, therefore, alter the conformations or orientations of at least some of the membrane-spanning segments. As described above, we have identified a cysteine in TM6 which becomes accessible in a CAM background. By extending this method, it should now be possible to map the activation-related changes in accessibility of all the residues that form the surface of the bindingsite crevice.

References 1. Todd, A. P., Cong, J., Levinthal, F., Levinthal, C, and Hubbell, W. L. (1989). Proteins, 6, 294. 2. Altenbach, C., Marti, T., Khorana, H. G., and Hubbell, W. L. (1990). Science, 248, 1088. 3: Jakes, K. S., Abrams, C. K., Finkelstein, A., and Slatin, S. L. (1990). J. Biol. Chem., 265, 6984. 4. Careaga, C. L., and Falke, J. J. (1992). Biophys. J., 62, 209. 5. Pakula, A. A., and Simon, M. I. (1992). Proc. Natl. Acad. Sci. USA, 89, 4144. 6. Jung, K., Jung, H., Wu, J., Prive, G. G., and Kaback, H. R. (1993). Biochemistry, 32, 12273. 7. Akabas, M. H., Stauffer, D. A., Xu, M., and Karlin, A. (1992). Science, 258, 307. 8. Akabas, M. H., Kaufmann, C., Archdeacon, P., and Karlin, A. (1994). Neuron, 13, 919. 9. Akabas, M. H., and Karlin, A. (1995). Biochemistry, 34, 12496. 10. Xu, M., Covey, D. F., and Akabas, M. H. (1995). Biophys. J., 69, 1858. 11. Xu, M., and Akabas, M. H. (1993). J. Biol. Chem., 268, 21505. 12. Akabas, M. H., Kaufmann, C., Cook, T. A., and Archdeacon, P. (1994). J. Biol. Chem., 269, 14865. 13. Yan, R. T., and Maloney, P. C. (1995). Proc. Natl. Acad. Sci. USA, 92, 5973. 14. Pascual, J. M., Shieh, C. C., Kirsch, G. E., and Brown, A. M. (1995) Neuron, 14, 1055. 15. Fu, D., Ballesteros, J. A., Weinstein, H., Chen, J., and Javitch, J. A. (1996). Biochemistry, 35, 11278. 16. Javitch, J. A., Fu, D., and Chen, J. (1995). Biochemistry, 34, 16433. 17. Javitch, J. A., Fu, D., Chen, J., and Karlin, A. (1995). Neuron, 14, 825. 18. Javitch, J. A., Li, X., Kaback, J., and Karlin, A. (1994). Proc. Natl. Acad. Sci. USA, 91, 10355. 19. Javitch, J. A., Ballesteros, J. A., Weinstein, H., and Chen, J. (1998). Biochemistry, 37, 998. 20. Schirmer, T., Keller, T. A., Wang, Y. F., and Rosenbusch, J. P. (1995). Science, 267, 512. 33

Merrill M. Simpson et al. 21. Brown, K. A., Howell, E. E., and Kraut, J. (1993). Proc. Natl. Acad. Sci. USA, 90, 11753. 22. Javitch, J. A., Fu, D., Liapakis, G., and Chen, J. (1997). J. Biol. Chem., 272, 18546. 23. Ferrer, J. V., and Javitch, J. A. (1998). Proc. Natl. Acad. Sci. USA, 95, 9238. 24. Olami, Y., Rimon, A., Gerchman, Y., Rothman, A., and Padan, E. (1997). J. Biol. Chem., 272, 1761. 25. Seal, R. P., and Amara, S. G. (1996). Soc. Neurosci. Abstr., 22, 1575. 26. Rees, S., Coote, J., Stables, J., Goodson, S., Harris, S., and Lee, M. G. (1996). BioTechniques, 20, 102. 27. Creighton, T. E. (1993). Proteins: structures and molecular properties, W.H. Freeman and Co., New York. 28. Chothia, C. (1976). J. Mol. Biol, 105, 1. 29. Levitt, M. (1978). Biochemistry, 17, 4277. 30. Chou, P. Y, and Fasman, G. D. (1977). J. Mol. Biol, 115, 135. 31. Roberts, D. D., Lewis, S. D., Ballou, D. P., Olson, S. T., and Shafer, J. A. (1986). Biochemistry, 25, 5595. 32. Stauffer, D. A., and Karlin, A. (1994). Biochemistry, 33, 6840. 33. Holmgren, M., Liu, Y., Xu, Y., and Yellen, G. (1996). Neuropharmacology, 35, 797. 34. Yan, R. T, and Maloney, P. C. (1993). Cell, 75, 37. 35. Karlin, A., and Akabas, M. H. (1998). In Methods in Enzymology (ed. P. M. Conn.). Vol. 293, p. 123. Academic Press, London. 36. Javitch, J. A. (in press). In Methods in Enzymology (ed. S. G. Amara.). Vol. 296. Academic Press, London. 37. Javitch, J. A., Fu, D., and Chen, J. (1996). Mol. Pharmacol, 49, 692. 38. Samama, P., Cotecchia, S., Costa, T., and Lefkowitz, R. J. (1993). /. Biol Chem., 268, 4625. 39. Gether, U., Lin, S., Ghanouni, P., Ballesteros, J. A., Weinstein, H., and Kobilka, B. K. (1997). EMBOJ., 16, 6737. 40. Gether, U., Lin, S., and Kobilka, B. K. (1995). J. Biol. Chem., 270, 28268. 41. Farrens, D. L., Altenbach, C., Yang, K., Hubbell, W. L., and Khorana, H. G. (1996). Science, 274, 768.

34

3

Post-translational acylation of signal transducing polypeptides: palmitoylation of G protein-coupled receptors MICHEL BOUVIER, ULLA PETAJA-REPO,THOMAS LOISEL, and CHARLENE BELANGER

1. Introduction In recent years, a number of proteins involved in signal transduction have been found to be acylated by fatty acids or prenyl derivatives. Three major classes of fatty acylation have been characterized. These are: (a) the Nterminal myristoylation of glycine residues through amide linkage; (b) the prenylation of cysteine residues by the formation of thioether links to farnesyl or geranylgeranyl moieties; and (c) the palmitoylation which occurs through the thioesterification of cysteine residues. In contrast to myristoylation and prenylation, which typically occur co-translationally, palmitoylation is a genuine post-translational modification (1). The chemical reactivity of sulfhydryls and thioester linkages have led several investigators to propose that protein palmitoylation may represent a reversible modification which could undergo dynamic regulation. In fact, rapid turnover of the covalently attached palmitate has been reported for several proteins, and this turnover is modulated by external stimuli (2-4). Palmitoylation has been found to be particularly prevalent for proteins involved in processes such as cell adhesion, cell growth, and signal transduction, raising the intriguing possibility that it could play regulatory roles in these events. Palmitoylated proteins include p21ras (5), GAP-43 (6), the endothelial nitric oxide synthase (7), several tyrosine kinases belonging to the p60src family, such as p56lck (8), p59fyn, p55fgr, and p56hck (9), as well as many G protein a-subunits (10), their cognate receptors (11), and the G protein-coupled receptor kinases (12, 13). Palmitoylation of G protein-coupled receptors (GPCRs) was first demonstrated for the visual receptor rhodopsin (14). Indeed, as early as 1984, it was

Michel Bouvier et al. observed that bovine rhodopsin could covalently incorporate palmitate following incubation of rod outer segments with [3H]palmitate. Four years later, physico-chemical analysis of a C-terminal peptide fragment derived from bovine rhodopsin led to the identification of cysteine-322 and -323 as the main palmitoylation sites for this protein (15). Primary sequence comparison revealed that at least one of these two cysteine residues is conserved in a similar position within the carboxyl tail of most GPCRs. Such a high level of conservation led several groups to investigate whether palmitoylation could represent a general modification with functional consequences for this class of signalling polypeptides. Although the occurrence of palmitoylation has been shown for a number of hormonal GPCRs, including the B2- and the a2-adrenergic (16, 17), the Dl and D2 dopamine (18, 19), the serotonin 5-HT1B and 1A (20, 21), the thyrotropin-releasing hormone (22), the V2 vasopressin (23), the m2 muscarinic (24), the lutropin-choriogonadotropin (25) as well as the endothelin receptors A and B (26, 27), and the glutamate metabotropic-4 receptor (28), progress in studying additional receptors and the dynamic regulation of this modification has been hampered by a number of technical difficulties, some of which are inherent to the nature of the GPCRs. For example, the low level of expression of most hormonal GPCRs makes it difficult to assess the palmitoylation state of endogenously expressed receptors. Studying their palmitoylation is also complicated by the fact that, contrary to most other palmitoylated proteins which are soluble (the palmitoylation being proposed as essential for their targeting from the cytosolic fraction to the inner face of the plasma membrane), the receptors are integral plasma membrane proteins that are fairly hydrophobic, thus complicating their purification and handling. Finally, the sparse knowledge concerning the enzymes and molecular processes that control palmitoylation has complicated the development of standardized in vitro palmitoylation assays. In this chapter, we describe a number of experimental approaches and protocols that have been developed over the last ten years or so to circumvent some of the problems described above, and which allow the study of various aspects of GPCR palmitoylation.

2. Detection of receptor palmitoylation One of the major difficulties in directly studying any post-translational modification of GPCRs is the low amount of these proteins that can be isolated from tissues or cell lines naturally expressing them. This problem is particularly important when considering palmitoylation, because of the relatively low specific activity of [3H]- or [14C]palmitate, the tracers used to detect palmitoylation, in comparison with the 32Pi used to detect phosphorylation, for example. One approach that we and others have selected to circumvent this problem has been to use heterologous expression, allowing various levels 36

3: Post-translational acylation of signal transducing polypeptides of over-expression. In addition, to facilitate purification of sufficient quantities of receptor, these systems permit the study of selected receptor subtypes in chosen genetic backgrounds.

2.1 Heterologous expression systems Two major heterologous expression systems have been used to study the palmitoylation of GPCRs: baculovirus-insect cells, and mammalian cells. 2.1.1 Insect cells In the last few years, recombinant Autographica californica nuclear polyhedrosis viruses (baculovirus) have been successfully used to direct high-level expression of many GPCRs in insect cells, in most cases using Spodoptera frugiperda (Sf9) cells (see also chapter 7). Proteins expressed using this system have largely been found to maintain their normal characteristics, including specific post-translational modifications such as palmitoylation and phosphorylation. Receptor levels reaching 20-100 pmol mg-1 of protein have been reported for many GPCRs using the insect system. This high level of expression, rarely reached in mammalian heterologous expression systems, has been perceived as a major advantage for the study of palmitoylation, and many studies have used the insect system to facilitate the detection of the palmitoylated receptor. As an example, the first demonstration of [3H]palmitate incorporation into the human B2-adrenergic receptor required more than 6 weeks of autoradiography, following metabolic labelling of mammalian cells expressing 2 pmol of receptor per milligram of membrane proteins (16). The use of Sf9 cells expressing 10-50 times higher levels significantly reduced the time required to visualize the modification. However, since gene expression occurs through the infectious cycle of the virus, it is transient, the cells dying from the infection several days later. Defining a time window that allows for high levels of expression, while maintaining close to 100% cell viability, is therefore of primary importance for these systems. The optimal time of infection should be determined for each recombinant baculovirus but, in most cases, maximal levels of expression are found at 48 h post-infection, a time at which the proportion of viable cells, as assessed by their ability to exclude Trypan Blue, is usually greater than 90%. Also, to maintain a reproducible level of expression from one experiment to the next, virus stocks must be properly maintained and propagated, and their titres verified regularly. Additional details about baculovirus and Sf9 cells can be found in O'Reilly et al. (29) (see also Chapter 7). i. Construction of the recombinant baculovirus. Typically, to prepare the recombinant baculovirus encoding the GPCR of interest, the coding region of the baculovirus polyhedrin gene is replaced with that of the selected receptor. For this purpose, the receptor DNA fragment is ligated into a baculovirus recombination vector such as pJVETLZ at the NheI site, yielding the pJV-receptor vector. Transfer of the receptor and B-galacto37

Michel Bouvier et al. sidase coding sequences from pJV-receptor to the baculovirus genome is then achieved by homologous recombination with the linear AcMNPV genome, by cotransfection in Sf9 cells using a Cationic Liposome Mediated Transfection kit (Invitrogen). Recombinant baculoviruses are purified by plaque assay, using the B-galactosidase detection system (30). Confirmation of receptor expression can then be assessed by radio-ligand binding assay or Western blot analysis. The availability of commercial systems to clone and purify recombinant baculovirus, such as the Bac-to-Bac kit (Life Technologies) makes the generation of these viruses relatively easy. ii. Culture and infection of Sf9 cells. Sf9 cells are grown in Grace's insect medium (Gibco) supplemented with fetal bovine serum (FBS, 10% v/v) at 27°C. Although Sf9 cells can also be grown as monolayers in flasks, metabolic labelling is more easily carried out when they are grown in suspension using Erlenmeyer flasks placed in an orbital shaker in the presence of pluronic acid (0.001%) to prevent cell tearing due to agitation. To permit expression of the receptor at sufficient levels in the entire cell population, infection should be carried out at a multiplicity of infection (MOI) of approximately 5 recombinant baculovirus molecules per Sf9 cell when they are in a logarithmic phase of growth at a density of 1-2 X 106 cells ml-1 (see Chapter 7 for further details). 2.1.2 Mammalian cells Although the levels of expression achieved in mammalian cells following either transient or stable transfection protocols rarely reach those obtained in Sf9 cells, both these systems have been used successfully to study palmitoylation. The major advantage of transient over stable expression systems is that one does not need to go through the tedious process of selection and cellular cloning (4-6 weeks), and one can readily use the cells 48 h after transfection. Moreover, higher levels of expression are generally obtained using transient systems. However, stable expression systems provide homogeneous populations of cells that all express the receptor at the same level and allow for repeated experiments under identical conditions. In general, we feel that stable cell lines offer a better model system to study palmitoylation of GPCRs. These cells are generally obtained by transfecting the cells with expression plasmids that encode the receptors of interest, as well as a selectable marker such as an antibiotic resistance gene. Selection of the expressing cells is then realized by treatments with appropriate concentrations of the antibiotic. A large number of cell lines can be used to generate stable expression systems. Considerations in selecting a cell line should include the ease of cultivation of large quantities of these cells, the absence of endogenous expression for the receptor to be transfected, the presence of the signalling pathway(s) to which the receptor should be functionally coupled, the presence of other 38

3: Post-translational acylation of signal transducing polypeptides signalling molecules which could be considered in the course of the study, etc. Chinese hamster fibroblasts (CHW-1102, NIGMS GM0459; and CHO, ATCC CRL 9096), and human embryonic kidney cells (HEK-293, ATCC CRL 1573), stably expressing various GPCRs, have been used successfully to detect palmitoylation. For the B2-adrenergic receptor, stable expression levels reaching 10 pmol per mg of protein were obtained in HEK-293 cells using the pCDNA3 plasmid (Invitrogen), greatly facilitating the study of palmitoylation in mammalian cells. Once the cell line of choice has been generated, the culture conditions should be maintained as usual until one day before the experiments. In our case, CHW or HEK-293 cells are grown as monolayers in Dulbecco's minimum Eagle's medium (DMEM), supplemented with 10% FBS, penicillin (100 units ml-1), streptomycin (100 mg ml-1), fungizone (0.25 mg ml-1), and glutamine (2 mM), in an atmosphere of 95% air-5% CO2 at 37°C. In the case of stably transfected cell lines, we found that maintaining the selective pressure with the selectable antibiotic (in most cases neomycin) helps to maintain the level of receptor expression at stable levels for longer period of times (between 30 and 60 passages, depending on the clones).

2.2 Metabolic labelling The approach most widely used to study receptor palmitoylation is directly monitoring the covalent attachment of radiolabelled palmitate molecules to the receptor in situ, following isotopic labelling of the cellular palmitoyl-CoA pool. The next section describes the method used to perform such metabolic labelling. 2.2.1 Optimization of labelling The substrate involved in the palmitoylation reaction in cells is the metabolic intermediate palmitoyl-CoA. However, the polar characteristic of this activated lipid prevents its direct utilization in metabolic labelling experiments, since it can not cross biological membranes. The labelling of the cellular pool of palmitoyl-CoA is therefore achieved by incubating the cells with [3H]palmitate, which penetrates into the cells, and is transformed by ATP-driven esterification with extra-mitochondrial CoA-SH to yield palmitoyl-CoA. This newly formed palmitoyl-CoA can then be used as the palmitate donor in protein acylation reactions, but is also used in the fatty acid oxidation and elongation pathways in various cellular compartments, including the mitochondria. As a consequence of the large cellular pool of palmitoyl-CoA, its rapid metabolism in distinct metabolic pathways, and its subcellular compartmentalization, it is virtually impossible to determine the specific activity of the acyl-CoA pool that is used for the palmitoylation reactions. Moreover, its rapid turnover makes it very difficult to reach isotopic equilibrium for extended period of times. It follows from the previous observations that establishing the conditions that allow optimum steady-state labelling of the receptor is of 39

Michel Bouvier et al. primary importance, and that the time of labelling is a crucial parameter. The optimum time of labelling varies with the cell type used, and therefore must be established for each system considered. For instance, maximum [3H]palmitate incorporation into the B2-adrenergic receptor was observed following 1 h labelling in Sf9 cells (31), whereas 3 h labelling is required to attain maximal incorporation in HEK-293 cells (Lynda Adam and M. B., unpublished observation). When possible, labelling longer than 4 h should be avoided, to reduce the incorporation of the label into amino acids following metabolic degradation of the fatty acid, and use of the labelled acetyl-CoA in the amino acid biosynthetic pathways. Because tritium is a relatively weak radioisotope that is required in large quantities for detection by autoradiography, selecting a cell line that expresses a large number of receptors is of primary importance. For this reason and as indicated above, Sf9 cells have often been used as the model of choice to study palmitoylation. However, mammalian cells expressing 1 pmol of receptor per mg of membrane protein or more can also be used. 2.2.2 Solubilization and purification of the receptor Even in Sf9 cells that express as much as 50 pmol of receptor per mg of membrane proteins, the presence of a fairly large number of palmitoylated proteins prevents the detection of the palmitoylated receptor in crude membrane preparations. It follows that, after metabolic labelling, receptors must be solubilized and purified. Solubilization represents an important and delicate step in the isolation of the receptors. Indeed, only a few detergents (e.g. digitonin and n-dodecyl maltoside) preserve the ligand-binding activity of most GPCRs. This is important, since this biological activity may be used in affinity purification schemes. Moreover, maintaining the ligand-binding activity allows the measurement of the number of functional receptors throughout the purification steps. Such quantification is important for comparing the palmitoylation states in different conditions. The best detergent (or detergent mixture) and optimum Solubilization conditions should be determined for each receptor in a given cell type. However, either digitonin or n-dodecyl maltoside represent good starting choices, as they have been found to solubilize many distinct receptors successfully. The various approaches that are generally used to purify GPCRs can be classified in three categories: (a) methods that take advantage of the ligandbinding properties of the receptor and use an affinity resin; (b) immunoprecipitation techniques using polyclonal or monoclonal antibodies raised against the receptor sequence; and (c) methods based on the construction of fusion recombinant proteins between the receptor and epitopes specifically designed to allow purification. The use of an affinity resin requires that the receptor maintains its binding properties after Solubilization. When this is not the case, immunoprecipitation or an epitope-based purification scheme should be used. 40

3: Post-translational acylation of signal transducing polypeptides The following protocol describes a procedure to detect the palmitoylation of the human B2-adrenergic receptor expressed in Sf9 cells, using an affinity chromatography purification scheme. Procedures using mammalian cells and alternative purification procedures are described later in this chapter.

Protocol 1. Pulse metabolic labelling of the human B2-adrenergic receptor with [3H]palmitate in Sf9 cells Equipment and Reagents • Sf-9 cells • Recombinant baculovirus expressing the human B2-adrenergic receptor • [9, 10-3H] palmitic acid • Grace's insect medium (Gibco)

.

• Autoradiographic film Fluorographic enhancer . Centriprep and Centricon cartridges (Amicon) • Alprenolol-sepharose resin

Method 1. Infect 100 to 200 ml of Sf9 cells (2 X 106 cells ml-1), grown in suspension, with a recombinant baculovirus encoding the human 32AR at a multiplicity of infection (MOD of approximately five recombinant baculovirus molecules per Sf9 cell. 2. At 30 h post-infection, transfer the cells to serum-free medium for 18 h to stop cell growth and reduce metabolic activity.a 3. Following this starvation period, incubate 25 ml of Sf9 cell suspension (50 X 106 cells) for 1 h at 27°C in the presence of Grace's insect medium (Gibco) supplemented with 1% FBS, and add [9,103 H]palmitic acid [60 Ci mmol-1] dissolved in dimethyl sulfoxide (DMSO) to a final concentration of 0.2 mCi ml-1, for periods of time varying between 15 min and 4 h.b 4. Terminate the labelling period by centrifugation at 500 g for 5 min at 4°C. From this point the samples are always kept at 4°C unless indicated otherwise. 5. Wash the pelleted cells twice with 50 ml of ice-cold phosphatebuffered saline (PBS), and disrupt them by sonication in 5ml of icecold buffer containing 5 mM Tris-HCI, pH 7.4, and a protease inhibitor cocktail composed of 2 mM EDTA, 5 mg ml-1 leupeptin, 10 mg ml-1 benzamidine, and 5 mg ml-1 soybean trypsin inhibitor (buffer A).c 6. Centrifuge the lysate at 500 g for 5 min, and centrifuge the supernatant again at 45000 g. Wash the pelleted membrane three times in buffer A. 7. To solubilize the receptor, resuspend the pelleted membrane preparations to a final concentration of 2 mg of protein per ml, in 2-4 ml of a 41

Michel Bouvier et al. Protocol 1.

Continued

buffer containing 100 mM NaCI, 10 mM Tris-HCI, 5 mM EDTA, pH 7.4, 0.3% digitonin (or 0.3% n-dodecyl maltoside) and protease inhibitors (as described above). Stir the suspension gently, using a magnetic stirring bar, for 90 min. 8. Remove the non-solubilized material by centrifugation at 100000 g for 60 min. You can use the supernatant containing the solubilized receptor immediately for the purification procedures, or quickly freeze it and keep it at -80°C for several days if necessary. 9. To purify the receptor, equilibrate an alprenolol-Sepharose resin (32, 33), 6 ml of gel per sample, with 50 ml of buffer B (100 mM NaCI, 10 mM Tris-HCI, 2 mM EDTA, pH 7.4, 0.05% digitonin [or 0.02% ndodecyl maltoside] and protease inhibitors). Then add the solubilized receptor (2-4 ml) to the affinity resin, and shake gently for 2 h at room temperatured to allow binding of the receptor to the matrix. 10. Following this batch-wise loading procedure, transfer the resin to 15 ml columns, let the soluble preparation flow through, and place the columns at 4°C.d Once the temperature of the resin has reached 4°C, wash the columns with 15 ml of an ice-cold buffer containing 500 mM NaCI, 50 mM Tris-HCI, 2 mM EDTA, pH 7.4, 0.05% digitonin (or 0.02% n-dodecyl maltoside). After the high salt washing, restore the original ionic strength by washing with 30 ml of buffer B. 11. In preparation for the bio-specific elution, return the columns to room temperature.d Then elute with 15 ml of buffer B containing 60 mM of alprenolol at a rate of 5 ml h-1 at room temperature, and collect the eluate on ice to reduce proteolytic degradation 12. Concentrate the eluted receptor to a final volume of 50 ml by ultrafiltration, using Centriprep and Centricon cartridges (Amicon). After assessing the number of receptors isolated,e prepare the samples for SDS-PAGEf immediately, or keep them at -80°C for a few days following quick-freezing in liquid nitrogen. 13. Following electrophoresis, incubate the fixed gels with a fluorographic enhancerg for 30 min, dry, and expose the gel to Kodak XAR5 or Biomax-MS films with intensifying screens at —80°C for several days. 14. Analyse the fluorogram obtained by densitometry, using a ladder of [3H]palmitic acid blotted on a Whatman paper as an internal fluorography control.h a

Good labelling of the receptors can also be achieved without the starvation period. Once the optimum labelling time has been established, experiments can be carried out using that time. For the p2-adrenergic receptor expressed in Sf9 cells, the maximal labelling was observed following a 1 h labelling period, b

42

3: Post-translational acylation of signal transducing polypeptides c

ln some cases, adding PMSF, bestatin, and aprotinin may be necessary to prevent proteolysis completely. d To favour rapid binding on- and off-rate of the receptor to the resin, the loading and elution procedures are carried out at room temperature. However, to slow down the off-rate during the washing procedures, the resin temperature is reduced to 4°C to prevent non-specific elution of the receptor. e After desalting the purified sample on a G-50 gel filtration column to remove alprenolol, B2AR recovery can be measured by soluble binding using [125I] cyanopindolol (CYP) as the radioligand (34). f The thioester link between the receptor and the palmitate moiety is somewhat sensitive to reducing agent, and it may be preferable to carry out the SDS-PAGE under non-reducing conditions. However, many GPCRs will tend to form high molecular weight aggregates that do not penetrate the polyacrylamide gels under non-reducing conditions. Using a low concentration of reducing agents, such as 1% (B-mercaptoethanol or 5 mM DTT, is a good compromise that does not significantly affect the palmitoylation of most proteins. Boiling the samples is not recommended, as it may promote more aggregation. g Using this protocol, incorporation of tritiated palmitate can be detected after one to two weeks of exposure. A commercial fluorographic enhancer can be advantageously substituted by a solution of 1 M salicylic acid. h Fluorography is linear only over a small range of signal intensity. Blotting standardized amounts of [3H]palmitate onto a Whatman paper allows the generation of a standard curve that facilitates quantification.

Unfortunately, affinity resins that permit bio-specific purification are not available for all GPCRs and, as mentioned above, solubilization does not always permit preservation of the binding activity of the receptor. This has led to the use of alternative approaches. Although immunoprecipitation, using antibodies raised against receptors, would be the technique of choice to purify GPCRs, attempts to raise antibodies against these receptors have generally failed to produce high affinity antibodies that can be used satisfactorily in immuno-affinity purification procedures. To date, only a few such antibodies have been generated. To circumvent this problem, many investigators have constructed recombinant receptors fused in frame with sequences encoding specific epitopes that can be recognized by commercially available antibodies. Epitopes derived from parts of the c-myc and the haemagglutinin coding sequences, as well as the FLAG™ epitope, are among the most popular. Antibodies that have been used successfully to immunoprecipitate epitopetagged receptors include the anti-c-myc 9E10 antibody (Santa Cruz Biotechnology), the anti-haemagglutinin 12CA5 antibody (Roche Diagnostics), and the anti-FLAG M2 antibody (Sigma). In addition to immuno-reactive epitopes, stretches of six or more histidines have also been used. These polyhistidine epitopes interact by a chelation reaction with high affinity to nickel, and thus allow purification of proteins bearing this epitope using an agarose resin coupled to nickel (Ni-NTA-agarose, Qiagen). The poly-Histagged receptors can be eluted from the resin using neutral imidazole, which preserves the binding activity of the receptor. Theoretically, the epitopes can be introduced anywhere within the coding sequence of the receptor. However, positioning them either at the N- or C-terminus has generally been favoured over an insertion within the core of the receptors, with the idea that 43

Michel Bouvier et al. (a) it would result in a greater accessibility of the epitopes to the antibodies or nickel, and (b) it should have less detrimental effects on the 3D structure of the receptors. Successful purification schemes have been developed using both N- or C-terminally positioned epitopes. In addition to permitting immunopurification of the receptors, the presence of the epitope allows visualization of the receptor by immunoblotting throughout the purification steps. Protocol 10 will describe how the FLAG epitope positioned at the Cterminus of the human 8-opioid receptor can be used to study the palmitoylation of this receptor in HEK-293 cells. Now, we will describe how the polyHis epitope can also be used to study palmitoylation of the |32-adrenergic receptor in Sf9 cells.

Protocol 2. Purification of the B2-adrenergic receptor using nickel-NTA-chromatography Equipment and Reagents • Recombinant baculovirus expressing the human B2-adrenergic receptor containing a hexa-histidine tag

• Sf-9 cells . Nickel-NTA-agarose (Qiagen)

Method 1. Infect 100-200 ml of Sf9 cells (2 X 106 cells ml-1), grown in suspension, with a recombinant baculovirus encoding the human |32-adrenergic receptor which is tagged at its N-terminus with six histidine residues. Use a multiplicity of infection (MOD of approximately five recombinant baculovirus molecules per Sf9 cell. 2. Repeat Steps 2 to 8a of Protocol 1, replacing the Tris-HCI buffer with a 25mM Hepes buffer, pH 7.4. 3. Load the solubilized membrane preparations on columns packed with 10 ml of Ni-NTA-agarose (Qiagen), at a rate of 0.5 ml min-1 in the presence of 10 mM imidazole to reduce the non-specific binding of proteins to the resin. All the chromatographic steps are carried out at 4 °C to prevent proteolytic degradation. 4. Wash the columns with 20 volumes of a buffer containing 25 mM Hepes (pH 7.4), 500 mM NaCI, 10 mM imidazole, 0.02% n-dodecylmaltoside, and the protease inhibitor cocktail (buffer C) at a rate of 1 ml min-1. 5. Re-equilibrate the columns to a lower ionic strength with 5 volumes of buffer C in which the NaCI concentration has been reduced to 100 mM. 44

3: Post-translational acylation of signal transducing polypeptides 6. Elute the poly-His-tagged receptor at 0.5 ml min-1 with 3 volumes of a buffer containing 25 mM Hepes (pH 7.4), 100 mM NaCI, 100 mM imidazole, 0.02% n-dodecyl-maltoside, 20% glycerol (v/v), and the protease inhibitor cocktail. 7. Treat the purified receptor as in Protocol 1 (steps 12-14). a

The solubilization and all chromatographic steps should be carried out in the absence of EDTA or EGTA. These chelating agents would bind to nickel and strip it away from the resin.

Immuno-purification and nickel-based chromatography methods usually lead to overall purification yields that are higher than those attained by affinity chromatography (yields can easily reach 85% of the solubilized receptor, compared to ~50% for the alprenolol-Sepharose affinity method described in Protocol 1). Generally, however, more contaminating bands may accompany the receptor following these epitope-tag-based procedures. Identification of the palmitoylated band(s) that correspond to the receptor thus becomes particularly important. Western blot analysis of the same samples, using either an anti-tag antibody or an antibody directed against the receptor itself, can be used to identify directly the receptor band. Another approach that permits the identification of the receptor is photo-affinity labelling using radiolabelled photo-reactive ligands, or chemical cross-linking if peptide ligands are available for the receptor of interest.

2.3 Characterization of incorporated radiolabel The most direct way to identify the nature of the fatty acid incorporated into the receptor would certainly be analysis by mass spectrometry (35). However, simpler approaches, requiring considerably less purified protein, can satisfactorily confirm the identity of the [3H]palmitate incorporated. First, the covalent nature of the attachment of the labelled lipid to the receptor can be confirmed by assessing the resistance of the labelling to organic extraction. Noncovalently bound lipids would be extracted from the purified receptor.

Protocol 3. Organic extraction of non-specifically adsorbed [3H]lipids Method 1. Incubate a purified and concentrated receptor preparation, obtained as in Protocol 1 or 2 from metabolically labelled cells, in a mixture of chlorofornrmethanol (2:1, v/v), and mix vigorously.

45

Michel Bouvier et al. Protocol 4.

Continued

2. After a 30 min incubation at room temperature, centrifuge the mixture at 4500 g to pellet the proteins. 3. Repeat the above extraction and centrifugation steps twice with the same solvents, three times with a mixture of chloroform:methanol: water (1:1:0.3, v/v/v), and finally once with methanol alone. After each extraction step, incubate the mixture for 10 min at room temperature. 4. Recover the pelleted proteins, solubilize them in SDS-PAGE sample buffer, and prepare the samples for electrophoresis and fluorography as in Protocol 1.a a The detection of tritium incorporation into the protein following the organic extraction would confirm the covalent attachment of the lipid moiety to the receptor.

The nature of the covalent link between the label and the receptor can also be deduced from the fact that, unlike the N-terminal amide linkage or the Cterminal thioether linkage of myristoylation and prenylation, respectively, thioesterification of cysteine by palmitic acid is sensitive to hydroxylamine. Sensitivity to hydroxylamine can be determined by treating either the soluble purified protein or, alternatively, the polyacrylamide gel following electrophoresis.

Protocol 4.

Hydroxylamine treatment of purified receptor

Equipment and Reagents • 1M hydroxylamine, pH7.0

• Centricon 30 cartridges (Amicon)

Method 1. Incubate the purified and concentrated receptor preparation, obtained in Protocol 1 or 2 a, with a 1 M Tris solution, with or without hydroxylamine, at a final concentration of 1 M, and pH of 7.0. 2. After an incubation for 1 h at room temperature, desalt and concentrate the samples using Centricon 30 cartridges (Amicon). 3. Prepare the concentrated samples for SDS-PAGE and fluorography as above.b • Samples that have been extracted with organic solvent, as in Protocol 3, can also be used. b Alternatively, the sensitivity of the labelling to hydroxylamine can be tested directly on fixed polyacrylamide gels following SDS-PAGE of the purified labelled receptors. For this, incubate the fixed gel with a 1 M hydroxylamine solution containing 1 M Tris, at a final pH of 7.0, for 16h, wash extensively, dry, and expose for fluorography.

46

3: Post-translational acylation of signal transducing polypeptides Third, analysis of the [3H]lipids incorporated into the receptor can also be performed using ascending chromatography. Protocol 5. Thin layer chromatography (TLC) of the[3H]lipid incorporated into the receptor Equipment and Reagents • Trypsin . Silica gel TLC plates (Aldrich)

. pH] lipid standards . 1M sodium salicylate

Method 1. Following SDS-PAGE of the purified labelled receptor, identify the receptor band using the fluorogram, and excise it from the polyacrylamide gel.a 2. Homogenize and digest the band in 5 ml of a solution containing 0.1 M (NH4)HCO3, (pH 7.7) with 0.3 mg ml-1 of trypsin. Incubate for 15 min at 37°C. 3. Acidify the digest with HCI, and extract with 1 volume of hexane. 4. Lyophilize the extract, and treat with 1 ml of 1 M KOH for 12 h at 37°C to cleave the attached lipids. 5. Concentrate the extract under nitrogen, and apply to a silica gel TLC plate (Aldrich) along with tritiated lipid standards in parallel lanes.b 6. Develop the chromatogram with hexane:ethyl acetate:acetic acid (80:20:1, v/v/v). 7. Air-dry the silica plates, immerse them in a 1 M salicylate solution for 1 h, and seal them into a hybridization bag prior to autoradiography.c a

The same protocol can be apply to the purified labelled receptor before it is resolved by SDSPAGE. b The identity of the palmitate is confirmed by directly comparing the chromatographic mobility of the labelled lipid cleaved from the receptor with that of commercial standards. Reverse-phase HPLC could be used instead of TLC. c Alternatively, the solid support of the silica gel can be cut into small slices, and directly counted by liquid scintigraphy.

3. Kinetics of palmitoylation 3.1 Pulse labelling As indicated above, studying the dynamics of palmitoylation is complicated by the fact that the cellular pool of palmitoyl-CoA is large and turns over very rapidly. It follows that the kinetics of receptor palmitoylation observed in pulse-labelling experiments is largely dependent on the rate of incorporation of the [3H]palmitate into the cells, its thioesterification with Co-A-SH, and the 47

Michel Bouvier et al. metabolic fate of [3H]palmitoyl-CoA once it is formed. Until now, it has not been possible to demonstrate that true isotopic equilibrium of the palmitoylCoA pool occurs in usual metabolic labelling experiments. Consequently, useful information can be gathered by assessing the entry of [3H]palmitate into cells, and its incorporation into cellular lipids. For this purpose, wholecell lipid labelling must be studied under the same metabolic labelling conditions that are used for assessing receptor palmitoylation. Protocol 6. Whole cell lipid labelling in Sf9 cellsa Equipment and Reagents • 1M Salicylate . [3H] lipid standards • Silica gel TLC plates (Aldrich)

• Speedvap (or similar) centrifugal vacuum evaporator

Method 1. Aliquot 1 ml of Sf9 cells that were metabolically labelled as in Protocol 1, and harvest them by centrifugation at 500 g for 5 min at 4°C.b 2. Rinse the cells twice with ice-cold PBS, and resuspend them in 200 ml of this buffer before lysing the cells by sonication.c 3. Extract total cell lipids by adding 200 ml of CHCI3:methanol:H20 (5:5:1, v/v/v) and mixing vigorously. 4. Separate the organic and the aqueous phase by centrifugation at 500 g for 5 min. Quantitatively recover the organic phase, repeat the extraction procedure twice, and pool the fractions. 5. Dry the pooled organic phases under vacuum, and resuspend the dried material in 100 ml of CHCI3:methanol:H20 (6:6:1, v/v/v). 6. Apply 20 ml of the resuspended lipid extract to a silica gel TLC plate for ascending chromatography, using n-butanol:CH3COOH:H20 (5:2:3, v/v/v) as the mobile phase. In parallel lanes load commercially available [3H]lipid standards.d 7. Immerse the silica gel plates for 1 h in a 1 M salicylate solution, and seal them into hybridization bags prior to autoradiography.a a

The same protocol can be used for mammalian cells. An aliquot of the supernatant can be collected and counted by liquid scintigraphy to determine the proportion of radiolabelled tracer that was not incorporated into the cells. c An aliquot of the cell lysate can be collected and counted by liquid scintigraphy to determine the proportion of radiolabelled tracer that was incorporated into the cells. d At least a [3H]palmitate standard should be included to identify the labelled band corresponding to cellular palmitate. In that system, fatty acids will migrate at the top of the chromatogram, while phospholipids will be found at the lower end of the TLC. e Because of the high level of radioactivity incorporated in the lipids, the use of salicylate is not required, and rapid direct detection can be achieved using the Biomax-MS films from Kodak with a low energy (LE) screen. b

48

3: Post-translational acylation of signal transducing polypeptides In the systems that we have studied, [3H]palmitate was found to enter the cells and to be rapidly metabolized. The rapid incorporation of tritiated fatty acids into cellular phospholipids and into the receptor indicates a fast turnover of both the cellular palmitoyl-CoA and receptor-bound palmitate. Comparing the labelling of the free palmitate with that of the receptor-bound palmitate over time allows determination of the time at which an apparent steady-state of labelling has been reached. A steady state is assumed when the labelling intensity of neither palmitate nor receptor changes with time. For palmitoylation of the human B2-adrenergic receptor in Sf9 cells, the steady state was reached between 40 and 80 min of labelling (31). Although determining the time required to attain the steady state is useful to select the optimal labelling conditions, it does not provide any real kinetic information about the receptor palmitoylation process. However, one such kinetic parameter, the turnover rate of the receptor-bound palmitate, can be obtained by performing pulse-chase labelling experiments.

3.2 Pulse-chase labelling Protocol 7. Determining the half-life of the receptor-bound palmitate Equipment and Reagents • [9, 10-3H] palmitic acid

• Grace's insect medium (Gibco)

Method 1. Incubate the cells with [9,10-3H]palmitic acid (60 Ci mmol-1) dissolved in DMSO to a final concentration of 0.2 mCi ml-1 (as indicated in Protocol 1) for the period of time required to attain steady-state labelling (as determined in Protocol 6). 2. At the end of the labelling period, harvest the cells by centrifugation at 500 g for 5 min, and rinse them three times with 50 ml of PBS at room temperature. 3. Initiate the chase by adding 100 ml of fresh Grace's insect medium with 1% FBS, containing 0.2 M of unlabelled palmitate. 4. Continue the incubation at 27°Ca for times varying between 5 and 60 min. 5. Stop the chase by placing the cells on ice, harvest the cells by centrifugation, prepare membranes, and purify the receptor as indicated in Protocol 1. 6. Following SDS-PAGE and fluorography, calculate the half-life of the receptor-bound palmitate by analysing the decay of the labelling 49

Michel Bouvier et al. Protocol 7. Continued intensity of the receptor band, using the one compartment metabolic turnover equation: q(t) = q(t—>oo) + q(t = 0)e(-R)t where t is the time of incubation (in minutes), R is the rate of decay, and q represents the level of labelling (as a percentage of the control). The half-life is estimated as f when q(t) = 50%. * Incubation should be carried out at 37oC with the appropriate medium if mammalian cells are used.

To determine if the decay in [3H]palmitate labelling is the consequence of receptor-bound palmitate turnover, and does not reflect the turnover of the receptor protein itself, pulse-chase labelling with [35S]methionine-cysteine should be carried out in parallel experiments. For this purpose, cells are preincubated in Grace's supplemented medium deprived of methionine and cysteine for 30 min. Tran35S-Label (ICN) is then added at a concentration of 100-200 mCi ml-1 to the medium for the time of labelling, and the chase is initiated by adding complete Grace's medium or DMEM containing 1 mM methionine and 1 mM cysteine.

3.3 Agonist-mediated regulation of receptor palmitoylation The effect of agonist stimulation on receptor palmitoylation can be assessed by various approaches. In pulse-labelling experiments, the effect of the agonist can be assessed by comparing the extent of receptor labelling observed in the presence and absence of various concentrations of agonists. The agonists can be added to the medium for the entire labelling period, or for shorter times (5-15 min) once steady-state labelling has been achieved. Alternatively, the effect of agonist stimulation on the turnover rate of the receptor-bound palmitate can be studied by adding agonists at the time of the chase in pulse-chase experiments. In any case, it is important to verify that the effects observed do not result from an agonist-promoted change in the specific activity of the palmitate donor pool. This is easily assessed by verifying that the treatment does not non-specifically influence the incorporation of [3H]palmitate into all the proteins.

Protocol 8. Incorporation of [3H]palmitate into total proteins Equipment and Reagents • 1M sodium salicylate

• Densitometer

50

3: Post-tmnslational acylation of signal transducing polypeptides Method 1. Aliquot 30 ml of Sf9 cells that have been metabolically labelled or chased in the presence and absence of agonist, and harvest them by centrifugation at 500 g for 5 min at 4°C, 2. Rinse the cells twice with ice-cold PBS, and resuspend them in 5ml of this buffer before lysing the cells by sonication. 3. Prepare membranes as indicated in Protocol 1, and directly solubilize them in SDS-PAGE sample buffer. 4. After SDS-PAGE separation, incubate the fixed polyacrylamide gel for 30 min in a solution of 1 M salicylate, and expose it to an autoradiographic film for several hours. 5. Assess the intensity of labelling of the various bands by densitometric analysis.

4. Assessing the palmitoylation state of the receptor at the cell surface Several studies have shown that agonist stimulation can selectively modulate the palmitoylation state of GPCRs (19, 36). More specifically, it has been proposed that agonists promote an increase in the turnover rate of receptorbound palmitate (31). This suggests that palmitoylation-depalmitoylation cycles can occur at the cell surface where the receptor is accessible to the agonists. To assess directly the palmitoylation occurring at the plasma membrane level, we developed a procedure that allows detection of the incorporation of [3H]palmitate into only those receptors that are expressed at the cell surface (U. P-R and M.B., manuscript in preparation). For this, we took advantage of the possibility of biotinylating cell surface proteins, using a water-soluble reagent that does not cross the plasma membrane, before carrying out metabolic labelling with [3H]palmitate. Following the metabolic labelling, cell-surface proteins can be purified through a streptavidin affinity chromatographic step, and the receptor subsequently purified from this mixture. The following protocols describe this approach, using the human 8opioid receptor expressed in mammalian HEK-293 cells. Because the 8-opioid receptor does not readily retain its binding properties following solubilization, a receptor construct harbouring a Flag epitope at its C-terminus was used to allow immunopurification.

51

Michel Bouvier et al. Protocol 9.

Biotinylation of cell-surface receptor and metabolic labelling

Equipment and Reagents • HEK293 cells expressing a Flag™-tagged 8-opioid receptor

• Sulfo-NHS-biotin (Pierce) • [9,10-3H] palmitic acid

Method 1. Grow HEK-293 cells expressing the Flag-tagged 8-opioid receptor in 150 cm2 flasks until nearly confluent, and wash them three times with PBS containing calcium and magnesium.a 2. Place the flasks on ice, and add 0.5 mg of sulfo-NHS-biotin (Pierce) per ml to the cells into a final volume of 6 ml of PBS. Incubate for 30 min at 4°C while shaking gently.b 3. Stop the biotinylation reaction by adding Tris-HCI (1 M, pH 7.4) to a final concentration of 50 mM, and incubate for 10 min at 4°C. 4. Wash the cells three times with 10 ml of PBS containing 50 mM TrisHCI, pH 7.4; finally add 10 ml of serum-free DMEM, and let the cells warm up by placing the flask at 37°C. 5. Initiate the metabolic labelling by adding [9,10-3H]palmitic acid (60 Ci mmol-1) in serum-free DMEM (10ml per 150 cm2 flask) to a final concentration of 0.4 mCi ml-1 in the presence or absence of the agonist Leu-enkephalin, for times varying between 15 and 120 min.c 6. Stop the labelling by aspirating the medium and washing the cells three times with 10 ml of ice-cold PBS. 7. The cells can then be quickly frozen and kept at —80°C, or processed immediately for purification. aThe presence of calcium and magnesium in the washing buffer helps to maintain the attachment of the cells to the flask. bWater-soluble biotinylation reagents with several different reactive groups are available from Pierce. Biotin can thus be coupled to either lysine or cysteine residues, or even to carbohydrate groups of the receptors. Reagents should therefore be selected keeping in mind which coupling would be the least detrimental to receptor function. cln HEK-293 cells, optimal labelling is obtained following a 60 min labelling period.

Protocol 10. Solubilization and purification of cell surface palmitoylated 5-opioid receptor from HEK-293 cells Equipment and Reagents • Magnetic stirrer • Streptavadin-agarose (Pierce)

• Anti-Flag™ peptide (Sigma) • Centricon cartridges (Amicon)

52

3: Post-translational acylation of signal transducing polypeptides Method 1. Following biotinylation and metabolic labelling, lyse the cells by sonication in 3 ml of a buffer containing 25 mM Tris-HCI (pH 7.4), 2 mM EDTA, 0.5 mM PMSF, 2 mM 1,10-phenanthroline, 2 mg ml-1 aprotinin, 5 mg ml-1 leupeptin, 5 mg ml-1 soybean trypsin inhibitor, 10 mg ml-1 benzamidine (buffer D). 2. Sediment the cell membranes by centrifugation at 45000 g for 20 min at 4°C, and wash the pellet twice in buffer D. 3. Solubilize the membrane proteins in 1 ml of a buffer containing 0.5% n-dodecyl maltoside, 25 mM Tris-HCI (pH 7.4), 140 mM NaCI, 2 mM EDTA, 0.5 mM PMSF, 2 mM 1,10-phananthroline, 2 mg ml-1 aprotinin, 5 mg ml-1 leupeptin, 5 mg ml-1 soybean trypsin inhibitor, 10 mg ml-1 benzamidine (buffer E) for 60 min at 4°C on a magnetic stirrer. 4. Remove the non-solubilized material by centrifugation at 100000 g for 60 min. 5. To the solubilized fraction, add 50 ml of streptavidin-agarose resin (Pierce)a in a 1.5 ml Eppendorf tube, and incubate for 2 h at 4°C to allow batchwise binding of the biotinylated proteins to the resin. 6. Centrifuge the Eppendorf tube in a microfuge for 1 min at 1000 g to sediment the resin. Aspirate and discard the supernatant. Wash the resin twice with 0.5 ml of Buffer E, four times with 0.5 ml of buffer E in which the concentration of n-dodecyl maltoside has been reduced to 0.1% (Buffer F), and finally once with 0.5 ml of a 25 mM Tris-HCI (pH 7.4) solution. After each washing procedure, recover the resin by centrifugation. 7. Initiate the elution of the biotinylated proteins by adding 100 ml of a solution containing 25 mM Tris-HCI (pH 7.4), and 1% SDS to the resin. Incubate for 15 min at room temperature, and 5 min at 95°C. Centrifuge and recover the eluate. Wash the resin with an additional 900 ml of Buffer E, centrifuge again, and pool this last supernatant with the 100 ml eluate. 8. To purify the Flag-tagged 8-opioid receptor from this mixture of cell surface biotinylated proteins, add 25 ml of anti-Flag(M2) antibodyagarose (Sigma), and incubate overnight at 4°C while maintaining gentle agitation. 9. Following the loading of the immuno-affinity resin, centrifuge and wash twice with 0.5 ml of Buffer E, and four times with 0.5 ml of Buffer F. 10. Initiate the final elution by adding 150 ml of 175 mg of Flag-peptide (Sigma) per ml of Buffer F, and incubate for 10 min on ice. Centrifuge the resin and recover the supernatant. Repeat this elution procedure 53

Michel Bouvier et al. Protocol 10.

Continued

two more times, and pool the three eluates. Concentrate the samples using Centricon cartridges, and prepare them for SDS-PAGE.b 11. Detection of the palmitoylated receptor is carried out by fluorography, as described in Protocol 1. aImmobilized monomeric avidin (Pierce) can also be used as an affinity matrix if it is advantageous to maintain functional activity of the receptor. Indeed, biotinylated proteins can be eluted from this resin with biotin in non-denaturing conditions. b This elution procedure allows specific elution of the Flag-tagged receptor. However faster and less expensive elution can be achieved by incubating the immuno-affinity resin directly in 100 MI of SDS-PAGE sample buffer, 15 min at room temperature and 2 min at 95°C. The sample is then centrifuged, and the supernatant loaded on the PAGE. However, elution of non-specific bands can occur, and needs to be controlled.

5. In vitro palmitoylation of synthetic peptides As indicated in the introduction, very little is known about the mechanism controlling the palmitoylation state of GPCRs. Despite considerable efforts to characterize and purify the enzyme(s) catalysing the palmitoylation of GPCRs, palmitoyl-transferase activity directed at GPCRs has remained elusive. Several investigators have even suggested that a non-enzymatic reaction may play a role. In fact, autocatalytic transfer of palmitate from palmitoyl-CoA to specific cysteines within a protein has been shown to occur in vitro for at least one GPCR, rhodopsin (14), for G a subunits (37), and for several other proteins (38, 39). Interestingly, the sites found to be autocatalytically palmitoylated in vitro were identical to those identified in whole cell palmitoylation experiments, suggesting that the molecular determinants of palmitoylation selectivity may be contained in the primary sequence of the protein, independently of the presence of an enzyme. Therefore, in an effort to identify the primary sequence required to direct palmitoylation of a specific cysteine within a protein (C. B. and M. B., manuscript in preparation), we developed an in vitro autocatalytic palmitoylation assay using synthetic peptides of various compositions with [3H]palmitoyl-CoA as the palmitate donor. Protocol 11. [3H]palmitoyl-CoA synthesisa Equipment and Reagents • [9,10-3H] palmitic acid . CoA-SH, lithium salt (Sigma) • Acyl-CoA synthetase (Sigma)

• Slilica gel TLC plates (Aldrich) • Film and screens for autoradiography

54

3: Post-translational acylation of signal transducing polypeptides Method 1. In a final volume of 1 ml of a buffer containing 0.1 M MOPS-NaOH (pH 7.4), 20 mM MgCI2, 2 mM CHAPS, 10 mM ATP, and 1 mM DTT, incubate 0.5 mmol of [9,10-3H]-palmitate (~25 mCi) with 25 mmol of CoA-SH lithium salt (Sigma), in the presence of 0.2 units of acyl-CoA synthetase (Sigma) at 28°C for 1 h. 2. To assess the efficacy of the enzymatic [3H]palmitoyl-CoA synthesis, take a 1 ml aliquot for TLC analysis. 3. Deposit the aliquot on a silica-gel TLC plate, and develop the chromatogram using n-butanol:pyridine:acetic acid:H2O (45:30:9:36, v/v/v/v) as the mobile phase.b 4. Air-dry the silica plates, and expose them directly to Biomax-MS films from Kodak with a low energy (LE) screen for autoradiography.c 5. Keep the synthesis products at -20°C.d a3

[ Hlpalmitoyl-CoA is commercially available, but is significantly more expensive than [3H]palmitate. b In this chromatographic system, the palmitate migrates with the solvent front, whereas the palmitoyl-CoA is found half-way in the chromatogram (Rf = 0.5). On a routine basis, the synthesis yield is ~80%. cAlternatively, detection can be achieved by fluorography, following immersion of the plates in 1 M salicylic acid as previously described. dThe [3H]palmitoyl-CoA produced is stable for several months.

Protocol 12. Auto-palmitoylation of synthetic peptides Equipment and Reagents • [3H] palmitoyl-CoA (Amersham)

• Silica gel TLC plates (Aldrich)

Method 1. Following synthesis,a solubilize the peptide in 100% DMSO, and dilute it in a buffer containing 0.1 M MOPS-NaOH (pH 7.4)b and 5 mM DTT, to obtain a final concentration of 10 mM peptide and 10% DMSO. Keep the stock solution of the peptide at -20 °C under nitrogen. 2. Add 5 ml of peptide to 5 ml of [3H]palmitoyl-CoA (~0.1 mCi) in a final volume of 30 ml of a buffer containing 0.1 M MOPS-NaOH (pH 7.4). Incubate at 37°C for times varying between 15 and 150 min.c 3. To visualize the products of the reaction, separate the palmitoylated peptides from palmitoyl-CoA by TLC chromatography on silica gel plates, using n-butanol:pyridine:acetic acid:water (60:30:5:15, v/v/v/v) as the mobile phase.d 55

Michel Bouvier et al. Protocol 12.

Continued

4. Detect the palmitoylated peptide by direct autoradiography as described above, and quantify the [3H]palmitate incorporated in the peptides by liquid scintillation counting, after cutting the band of the autoradiogram corresponding to the palmitoylated peptides.e aPeptides as short as eight amino acids were found to be good auto-palmitoylation substrates. The inclusion of DTT in the peptide stock solution and the conservation under nitrogen are required to prevent the formation of intermolecular disulfide bonds. bThe pH at which the reaction is carried out greatly influences the level of palmitoylation achieved, lower pH inhibiting the reaction. cWith most peptides, maximal palmitoylation is observed after 120 min. d ln this system, palmitoyl-CoA does not migrate away from the origin, facilitating the separation of the palmitoylated peptides. However, peptides with very polar properties may not migrate well. eThe covalent attachment of the palmitate, and the nature of the thioester linkage with the peptide, can be assessed by analysing the isolated palmitoylated peptide as indicated in Protocols 3 and 4.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Omary, M. B., and Trowbridge, I. S. (1981). J. Biol. Chem., 256, 12888. Bonatti, S., Giovanni, M. S., and Simons, K. (1989). J. Biol. Chem., 264, 12590. Omary, M. B., and Trowbridge, I. S. (1981). J. Biol. Chem., 256, 4715. Alvarez, E., Girones, N., and Davis, R. J. (1990). J. Biol. Chem., 265, 16644. Hancock, J. F., Paterson, H., and Marshall, C. J. (1990). Cell, 63, 133. Skene, J. H. P., and Virag, I. (1989). J. Cell Biol., 108, 613. Robinson, L. J., Busconi, L., and Michel, T. (1995). /. Biol. Chem., 270, 995. Paige, L. A., Nadler, M. J. S., Harrison, M. L., Cassady, J. M., and Geahlen, R. L. (1993). J. Biol. Chem., 268, 8669. Shenoy-Scarcia, A. M., Dietzen, D. J., Kwong, J., Link, D. C., and Lublin, D. M. (1994). J. Cell Biol., 126, 353. Linder, M. E., Middleton, P., Hepler, J. R., Taussig, R., Oilman, A. G., and Mumby, S. M. (1993). Proc. Natl. Acad. Sci. USA, 90, 3675. Morello, J.-P., and Bouvier, M. (1996). Biochem. Cell Biol., 74, 449. Stoffel, R. H., Randall, R. R., Premont, R. T., Lefkowitz, R. J., and Inglese, J. (1994). J. Biol. Chem., 269, 27791. Premont, R. T., Macrae, A. D., Stoffel, R. H., Chung, N., Pitcher, J., Ambrose, C., Inglese, J., MacDonald, M. E., and Lefkowitz, R. J. (1996). J. Biol. Chem., 271, 6403. O'Brien, P. J., and Zatz, M. (1984). /. Biol. Chem., 259, 5054. Ovchinnikov, Y. A., Abdulaev, N. G., and Bogachuk, A. S. (1988). FEBS Lett., 230, 1. O'Dowd, B. F., Hnatowich, M., Caron, M. G., Lefkowitz, R. J., and Bouvier, M. (1989). J. Biol. Chem., 264, 7564. Kennedy, M., and Limbird, L. E. (1994). J. Biol. Chem., 269, 31915. Ng, G. Y. K., O'Dowd, B. F., Caron, M., Dennis, M., Brann, M. R., and George, S. R. (1994). J. Neurochem., 63, 1589. 56

3: Post-translational acylation of signal transducing polypeptides 19. Ng, G. Y., Mouillac, B., George, S., Caron, M., Dennis, M., Bouvier, M., and O'Dowd, B. (1994). Eur. J. Pharmacol. Mol. Pharmacol. Sec., 267, 7. 20. Ng, G. Y., George, S. R., Zastawny, R. L., Caron, M., Bouvier, M., Dennis, M., and O'Dowd, B. F. (1993). Biochemistry, 32, 11727. 21. Butkerait, P., Zheng, Y., Hallak, H., Graham, T. E., Miller, H., Burris, K., Molinoff, P. B., and Manning, D. R. (1995). J. Biol. Chem., 270, 18691. 22. Kosugi, S., and Mori, T. (1996). Biochem. Biophys. Res. Commun., 221, 636. 23. Sadeghi, H. M., Innamorati, G., Dagarag, M., and Birnbaumer, M. (1997). Mol. Pharmacol., 52, 21. 24. Hayashi, M. K., and Haga, T. (1997). Arch. Biochem. Biophys., 340, 376. 25. Kawate, N., and Menon, K. M. (1994). J. Biol. Chem., 269, 30651. 26. Horstmeyer, A., Cramer, H., Sauer, T., Muller-Esterl, W., and Schroeder, C. (1996). J. Biol. Chem., 271, 20811. 27. Okamoto, Y., Ninomiya, H., Tanioka, M., Sakamoto, A., Miwa, S., and Masaki, T. (1997). J. Biol. Chem., 272, 21589. 28. Alaluf, S., Mulvihill, E. R., and Mcllhinney, R. A. (1995). J. Neurochem., 64, 1548. 29. O'Reilly, D. R., Miller, L. K., and Luckow, V. A. (1992). Baculovirus expression vectors, a laboratory manual, W. H. Freeman and Co., New York. 30. Vialard, J., Lalumiere, M., Vernet, M., Briedis, D., Alkhatib, G., Henning, D., Levin, D., and Richardson, C. (1990). J. Virol., 64, 37. 31. Loisel, T. P., Adam, L., Hebert, T., and Bouvier, M. (1996). Biochemistry, 35, 15923. 32. Benovic, J. L., Shorr, R G. L., Caron, M. G., and Lefkowitz, R. J. (1984). Biochemistry, 23, 4510. 33. Caron, M. G., Srinivasan, Y., Pitha, J., Kociolek, K., and Lefkowitz, R. J. (1979). J. Biol. Chem., 254, 2923. 34. Bouvier, M., Hnatowich, M., Collins, S., Kobilka, B. K., De Blasi, A., Lefkowitz, R. J., and Caron, M. G. (1988). Mol. Pharmacol., 33, 133. 35. Papac, D. I., Thornburg, K. R., Bullesbach, E. E., Crouch, R. K., and Knapp, D. R. (1992). /. Biol. Chem., 267, 16889. 36. Mouillac, B., Caron, M., Bonin, H., Dennis, M., and Bouvier, M. (1992). J. Biol. Chem., 267, 21733. 37. Duncan, J. A., and Gilman, A. G. (1996). J. Biol. Chem., 271, 23594. 38. Bizzozero. O. A., McGarry, J. F., and Lees, M. B. (1987). J. Biol. Chem., 262, 13550. 39. Mack, D., Berger, M., Schmidt, M. F. G., and Kruppa, J. (1987). J. Biol. Chem., 262, 4297.

57

This page intentionally left blank

4

G protein-coupled receptor phosphorylation and desensitization ANDREW B. TOBIN, ANGELA RAE, and DAVID C. BUDD

1. Introduction Phosphorylation of G protein-coupled receptors (GPCRs) has been linked with receptor desensitization, primarily through work on the B-adrenergic receptor(1-3). Extensive studies have demonstrated that agonist stimulation of the B-adrenergic receptor results in phosphorylation of the receptor by two kinases; protein kinase A (PKA), and the receptor-specific kinase B-adrenergic receptor kinase (GRK-2) (1-3). Each of these kinases phosphorylates distinct sites on the receptor. PKA sites are at RRSS motifs in the third intracellular loop and C-terminal tail (4), and GRK-2 sites are in an acid serine/threoninerich region of the C-terminal tail (5). Phosphorylation at both PKA and GRK-2 sites results in receptor desensitization (1). In the case of GRK-2mediated desensitization the mechanism has been demonstrated to involve the intermediary protein B-arrestin, which binds to the phosphorylated receptor and in doing so is able to 'displace' the Gas-protein, thereby uncoupling the receptor from adenylyl cyclase (6). Cloning studies have revealed that GRK-2 is a member of a protein kinase family called the G protein-coupled receptor kinase (GRK) family (7). There are at present six members of this family (GRK-1 to GRK-6); each is thought to be able to mediate agonist-stimulated phosphorylation of certain GPCRs (7). The process of agonist-mediated phosphorylation and desensitization of the B-adrenergic receptor is considered to represent a general regulatory phenomenon applicable to a wide range of GPCRs. There is now compelling evidence that nearly all GPCRs, including both adenylyl cyclase and phospholipase C (PLC)-coupled receptors, undergo agonist-dependent phosphorylation (8, 9). Whereas receptor phosphorylation appears to be a universal regulatory mechanism, the role that receptor phosphorylation plays in receptor function across this broad range of receptors is still unclear. However, because of the analogy with the B-adrenergic receptor system, most laboratories focus on a possible role in receptor desensitization. In the present chapter we shall describe methods we have developed for

Andrew B. Tobin et al. investigating receptor phosphorylation in intact cultured cells and broken cell preparations. We shall also describe approaches we have adopted in the investigation of desensitization of PLC-coupled receptors.

2. Techniques in the investigation of GPCR phosphorylation 2.1 Determination of GPCR phosphorylation in intact cultured cells by immunoprecipitation Early studies on the B-adrenergic receptor and m2-muscarinic receptor relied on ligand-based affinity-purification techniques to isolate phosphorylated receptors solubilized from tissues or cell lines that had been labelled with [32P]orthophosphate (e.g. 10, 11). The ability to purify these two particular receptor subtypes has allowed in vitro assays to be developed, in which purified receptors reconstituted in phospholipid vesicles were used as substrates for the receptor-specific kinases (e.g. GRK-2). Not only were these reconstitution techniques used in the early studies on the purification of GRK-2 (10), but they have also been employed recently to determine the modes of GRK-2 regulation, in particular the regulatory roles of G protein By-subunits (12, 13) and the phospholipid phosphatidylinositol 4,5-bisphosphate (14, 15). Similar affinity chromatography techniques have been used to isolate phosphorylated forms of other receptor subtypes, for example the CCK receptor (16). However, these techniques are extremely technically demanding due to the low expression levels of GPCRs, and the problems associated with solubilization of the receptor in a form that still allows ligand binding. The universal role of receptor phosphorylation in GPCR regulation has necessitated the development of less demanding techniques to monitor receptor phosphorylation that could readily be applied to a number of different receptor subtypes. Our laboratory was one of the first to use receptor-specific antibodies to immunoprecipitate a GPCR in the study of receptor phosphorylation (17). The use of receptor-specific antisera or epitope-tagged receptors is now widespread, and has been directly responsible for revealing the diversity of receptor subtypes that undergo phosphorylation (9). We have applied this technique in the investigation of phosphorylation of the PLC-coupled m3muscarinic (17-19) and ml-muscarinic receptors (20), expressed as recombinant proteins in Chinese hamster ovary cells (CHO cells), and more recently with the B2-adrenergic receptor expressed in CHO cells (Figure 7). The protocols we describe here are based on our experience with these receptor systems. Any immunoprecipitation protocol naturally centres on the quality of the primary antibody. By far the best anti-receptor antiserum we have used is that against the m3-muscarinic receptor, which was raised against a glutathione-Stransferase (GST) bacterial fusion protein containing a region of the third 60

4: G-protein coupled receptor phosphorylation

Figure 1. Time course for the agonist-mediated phosphorylation of the B2-adrenergic receptor. CHO cells expressing recombinant B2-adrenergic receptor were labelled with [32P]-orthophosphate and stimulated with the adrenergic agonist isoproterenol (1 mM). The reaction was stopped with ice-cold RIPA buffer, and the B2-adrenergic receptor immunoprecipitated using a receptor-specific antiserum. The positions of molecular weight markers (kDa) are shown.

intracellular loop of the human m3-muscarinic receptor (17). Although construction of bacterial fusion proteins can he lime-consuming and somewhat unpredictable, due to the potential problems with expressing certain receptor regions (particularly very hydrophobic regions), they do generally make excellent antigens. Alternatively, we have raised a number of receptor antisera (e.g. ml-muscarinic and metabotropic glulamate la receptors) against peptides conjugated to keyhole limpet haemocyanin. The procedures we use for antisera production are based on those described in the excellent methods book by Harlow and Lanc (21). The problem encountered with many GPCRs is that they are poor immunogens, so that raising high quality antisera for immunoprecipitation protocols has been difficult. An alternative approach to raising receptor subtype-specific antibodies for immunoprecipitalion studies is to epitope-tag the receptor, usually at the C-terminus. The solubilized receptor can then be immunoprccipitated using commercially available monoclonal antibodies raised against the epitope tag (see e.g. ref. 22). One epitope tag employed in GPCR studies is the influenza hacmagglutinin (HA) tag YPYDVPDYA, which is recognized by the 12CA5 monoclonal antibody available from BabCo, Berkeley Antibody Co, and Boehringer Mannheim. A shorter IIA tag, DVPDYA, has also been reported to be recognized by the 12CA5 antibody. Also the FLAG tag DYKDDDDK, which is recognized by the FLAGM2 monoclonal antibody available from Kodak IBI, is commonly used. 61

Andrew B. Tobin et al. It is essential to characterize the antiserum correctly, even (or especially!) when it has been obtained from a commercial source. The glycosylation of GPCRs results in unpredictable migration on SDS-PAGE, and their hydrophobic nature often results in broad 'fuzzy' bands (Figure 1). The use of transfected cell lines is probably the most reliable test for antiserum specificity. The antiserum should be tested against transfected and sham-transfected controls in both immunoprecipitation and Western blots. In testing the most appropriate receptor solubilization conditions, we have found that the RIPA buffer described here was the most efficient. The advantage of immunoprecipitation over methods that employ purification of the receptor on a ligand affinity matrix is that maintaining the ligand-binding properties is not important for immunoprecipitation. Therefore, the solubilization conditions can be harsh, ensuring total solubilization of the receptor. This is particularly important in the case of GPCRs that appear to be differentially solubilized by mild detergent solutions designed to maintain the ligand binding properties of the receptor (23). In our early studies we labelled cells with [32P]-orthophosphate in suspension (17). Recently, however, we have found it more convenient and reliable to label plated-down cells. Both methods are given here in Protocols 1 and 2 respectively. Protocol 1. [32P]-labelling of the cellular ATP pool and receptor solubilization (of cells in suspension) Equipment and Reagents • [32P]-orthophosphate (Amersham) • Cells expressing receptor of interest

• Microcentrifuge

Method 1. Harvest cells using HBS/0.5 mM EDTA (HBS: 10 mM Hepes, pH 7.4; 0.9% NaCI) and wash twice in phosphate-free Krebs/Hepes buffer (10 mM Hepes, 118 mM NaCI, 4.3 mM KCI, 1.17 mM MgS04.7H2O, 1.3 mM CaCI2.2H20, 25 mM NaHC03, 11.7 mM glucose, pH 7.4) 2. Resuspend the cells in phosphate-free Krebs/Hepes buffer at a density of 1-3 x 106 cells ml-1, and aliquot into 1ml aliquots. 3. Add [32P]-orthophosphate (50 mCi ml-1), and incubate the cells at 37°C for 60 min. Under these conditions we have calculated that the specific activity of the intracellular ATP pool is 840 ±161 c.p.m. per pmol ATP (in the case of CHO cells; ref. 17). 4. The cell suspensions can now be challenged with experimental reagents for the required time. Terminate the reaction by rapid centrifugation at ~2000 r.p.m. on a bench top Microfuge for 30 s. Aspirate the medium 62

4: G-protein coupled receptor phosphorylation and resuspend the cell pellet in 1 ml ice-cold RIPA buffer (10 mM TrisHCI, 10 mM EDTA, 500 mM NaCI, 1% NP-40, 0.1% SDS, 0.5% deoxycholate, pH 7.4). 5. Following 30 min solubilization on ice, clear the sample by centrifugation in a Microfuge (maximum speed for 3 min). Save the supernatant for immunoprecipitation of the solubilized receptor, see below.

Protocol 2. [32P]-labelling of the cellular ATP pool and receptor solubilization (of adhered cells) Equipment and Reagents • Cells expressing receptor of interest • [32P]-orthophosphate (Amersham)

• Microcentrifuge

Method 1. Grow cells to ~70% confluence on six-well dishes. It is important not to allow cells to become too confluent, since the solubilization procedure will precipitate the genomic DNA to give a stringy mass that will result in high background if the cell density is too high. 2. Remove cell culture medium, and wash cells twice in phosphate-free Krebs/Hepes buffer. 3. Incubate the cells in 1 ml phosphate-free Krebs/Hepes buffer containing 50 mCi [32P] orthophosphate for 60-120 min at 37 °C. 4. Add stimulatory reagents for the required time. 5. Terminate the reaction by aspirating the medium and adding 1 ml RIPA buffer (4°C). 6. Following 30 min solubilization on ice, transfer the reaction to a 1.5 ml Eppendorf centrifuge tube, clear by centrifugation on a bench top microfuge for 3 min at maximum speed. Remove the supernatant and save for immunoprecipitation of the solubilized receptor.

Protocol 3. Receptor immunoprecipitation Equipment and Reagents • Appropriate antiserum or antibody • Protein-A-sepharose CL-4B (Pharmacia) • Rotating mixer

• Microcentrifuge • X-ray film

63

Andrew B. Tobin et al. Protocol 3.

Continued

Method 1. Add antiserum (0.2-1.0 mg) to the solubilized cell extract obtained from Protocols 1 or 2 above. Incubate on ice for 60-90 min. (Note: it is preferable to purify the antiserum in some way, e.g. by protein A before use) 2. Add 175mlof protein A-sepharose slurry (Protein A-sepharose CL-4B (Pharmacia) 1.5 g resuspended in 50 ml of TE buffer (2.5 mM Tris-HCI, 2.5 mM EDTA, pH 7.4)) to the sample and mix on slowly rotating rollers for 15 min at 4°C. 3. Pellet the protein A-sepharose by brief centrifugation (maximum speed in a Microfuge for 30 s), aspirate the supernatant, and wash the pellet 3-5 times with TE-buffer. (It is important to note that the wash steps can be changed to eliminate background. High-salt and low-salt washes can be included, as well as detergent washes. For example, in the immunoprecipitation of the B2 adrenergic receptor (see Figure 1) from CHO cells expressing recombinant receptor, we wash the protein A pellet twice in a Tween-20 buffer (100 mM Tris-HCI, 0.5% Tween 20, 1.5 M NaCI, pH 7.4) followed by a further wash in TE-buffer. 4. At the final wash step remove as much of the supernatant as possible using a fine pipette tip. 5. Resuspend the pellet in 20 ml of 2 X SDS-PAGE sample buffer. Heat the sample to ~85 °C for exactly 2 min, and then resolve the proteins by SDS-PAGE. Do not boil the sample since this results in aggregation of the receptors. 6. Stain the gel with 0.2% Coomassie blue in 40% methanol-10% acetic acid, and destain in 40% methanol-10% acetic acid. This procedure will stain the immunoprecipitated antibody, thereby confirming equal immunoprecipitation/loading; however, the receptor will not be visible due to the low quantity. Dry the gel, and obtain an autoradiograph.

2.2 Identification of GPCR phosphorylation in a crude membrane preparation We have been able to demonstrate that the kinase responsible for phosphorylation of the m3-muscarinic receptor expressed in CHO cells (CHO-m3 cells) is, at least in part, associated with the plasma membrane (18, 24). Using a membrane preparation from CHO-m3 cells we have reported that the m3muscarinic receptor can still be phosphorylated in an agonist-dependent manner (24). This membrane kinase assay has allowed the characterization of the endogenously expressed receptor kinase using cell-impermeable inhibitors (24). Furthermore, we have been able to reconstitute purified receptor kinases 64

4: G-protein coupled receptor phosphorylation with the membrane preparation, and investigate the ability of exogenously added kinase to increase the level of agonist-sensitive receptor phosphorylation (18, 19). Similar studies have used reconstituted membrane preparations from cell lines to investigate the phosphorylation of the B-adrenergic receptor (25), the A3-adenosine receptor (26), and the bradykinin B2 receptor (27). The method outlined here for identification of receptor phosphorylation in membrane preparations from cloned cell lines is drawn from our work on the m3-muscarinic receptor expressed in CHO-m3 cells (expression levels ~1 pmol mg-1 protein). The technique can be divided into three distinct sections; (i) preparation of the membrane fraction, (ii) stimulation of receptor phosphorylation, and (iii) solubilization and immunoprecipitation of the receptor.

Protocol 4. Preparation of a membrane fraction from cell lines Equipment and Reagents • Polytron

Method 1. Harvest cells using HBS/0.5 mM EDTA (use at least two confluent 175 cm2 flasks), and resuspend the cell pellet in ice-cold TE-buffer (15 ml) plus protease inhibitors (1 mM phenylmethylsulfonyl fluoride (PMSF), 10 mg ml-1 soybean trypsin inhibitor, 1 mg ml-1 leupeptin, 1 mg ml-1 pepstatin A, 100 mg ml-1 benzamidine, 100 mg ml-1 iodoacetamide). 2. Leave cells on ice for 10 min to swell and homogenize (15 s pulse in a Polytron). 3. Remove cell debris by centrifugation at 2000 g for 3 min. Collect the supernatant which contains the membrane fraction, and pellet the membranes by centrifugation at 15 000 g for 10 min. 4. Resuspend membrane pellet in kinase buffer (20 mM Tris-HCI, 10 mM MgCI2, 1 mM EGTA, pH 7.5) plus protease inhibitors. Adjust protein concentration to 1 mg protein per ml.

Protocol 5. Stimulation of receptor phosphorylation in crude membranes Equipment and Reagents • Microcentrifuge

65

Andrew B. Tobin et al. Protocol 5.

Continued

Method 1. To 50 ml of the membrane preparation obtained using Protocol 4 (e.g. for CHO-m3 cells this is equivalent to ~0.1 pmoles of receptor) add stimulatory reagents and/or exogenous protein kinase preparation or buffer blank. Add kinase buffer to bring the final volume to 100ml,and start the reaction by adding 100 mM [y32P] ATP (1-4 c.p.m. fmol-1 ATP). 2. Continue the reaction at 32°C for the required time. It is important to remember that the membrane preparation contains ATPase activity. We have calculated that in our experiments using CHO-m3 membranes ~80% of the ATP is consumed within 10 min. 3. Stop the reaction either by adding 1 ml ice-cold RIPA buffer directly, or by first pelleting membranes by a brief 30 s spin in a Microfuge (maximum setting) followed by aspiration of the supernatant and resuspension of the pellet in 1 ml ice-cold RIPA buffer. 4. Solubilize the membranes on ice for 30 min. Clear the supernatant by centrifugation (13 000 g, 3 min), and immunoprecipitate the solubilized receptor as described in Protocol 3. Resolve the immunoprecipitated proteins by SDS-PAGE gel electrophoresis. Stain and destain the gel (as outlined in Protocol 3) before drying and obtaining an autoradiograph.

3. Identification of desensitization of phospholipase C-coupled receptors Many of the techniques that report desensitization of PLC-coupled receptors analyse the accumulation of total inositol phosphates in the presence of a lithium block of inositol monophosphatase (28). Although this technique can give a good indication of the rate of phosphatidylinositol polyphosphate hydrolysis at early time points (particularly within the first few seconds and minutes of agonist exposure), prolonged receptor stimulation results in problems associated with lipid-pool depletion and changes in the specific activity of the radiolabelled phosphatidylinositol polyphosphate pool (27). The technique we describe here measures the product of phosphatidylinositol 4,5bisphosphate hydrolysis, namely inositol 1,4,5-trisphosphate (Ins(l,4,5)P3). Using analysis of Ins(l,4,5)P3 production, we have been able to identify rapid desensitization of the peak Ins(l,4,5)P3 response that occurs after 10 seconds of agonist stimulation (29) (Figure 2). In contrast, the sustained Ins(l,4,5)P3 response appears to be resistant to desensitization. Since the sustained component contributes the majority of the inositol phosphates when measuring total inositol phosphates in the presence of lithium, any receptor desensitiza66

4:

G-protein

coupled

receptor

phosphorylation

Figure 2. Desensitization of the lns(1,4,5)P3 response in CHO-m3 cells. CHO-m3 cells expressing ~1 pmol mg-1 protein of recombinant human m3-muscarinic receptors were pre-incubated for 5 min with 1 mM muscarinic agonist carbachol (open circles) or vehicle (closed circles). Cells were then washed three times, and allowed to recover for 5 min before stimulation with 1 mM carbachol. Reactions were stopped with TCA (1 M), and lns(1,4,5)P3 extracted and assayed using a radio-receptor binding assay. Data represent the mean ± SEM (n = 3).

tion that results in a reduction in the early peak Ins(l,4,5)P3 response would not be detected if total inositol phosphate accumulation is employed. It is for this reason that we mainly measure Ins(l,4,5)P3 production using a radioreceptor assay to assess desensitization of PLC coupled receptors. This procedure can be divided into three components: (i) stimulation of the cultured cells; (ii) extraction and neutralization of Ins(l,4,5)P3; and (iii) Ins(l,4,5)P3 radio-receptor assay. (Although methods for (ii) and (iii) are given here, they have also been described in detail previously, refs 30,31)

Protocol 6. Stimulation of cultured cells in the investigation of PLC-coupled receptor desensitization Equipment and Reagents • Cells expressing at appropriate receptor e.g. M3 muscarinic acetylcholine receptor

Method

1. Seed cells on to 24-well dishes, and allow to grow overnight. 67

Andrew B. Tobin et al. Protocol 6.

Continued

2. Remove medium and wash once in Krebs/Hepes buffer (10 mM Hepes, 118 mM NaCI, 4.3 mM KCI, 1.17 mM MgS04.7H20, 1.3 mM CaCI2.2H20, 25 mM NaHC03, 11.7 mM glucose, 1.17 mM KH2PO4, pH 7.4). 3. Allow cells to stabilize for 10 min at 37°C. 4. Pre-expose cells to agonist/vehicle for the appropriate time with 200 ml agonist in Krebs/Hepes buffer (e.g. 5 min with 1 mM of the muscarinic agonist carbachol). 5. Stop pre-incubation by washing cells rapidly with 200 ml of Krebs/ Hepes buffer (37°C) three times. This should take ~2-3 min per dish. 6. Allow cells to incubate for a further 2-3 min, so that the total wash and recovery phase is at least 5 min. This recovery time is the minimum employed since calcium pool refilling takes ~2-3 min, and any depletion of the calcium pool will result in a reduced lns(1,4,5)P3 response, due to disruption of calcium feedback on PLC activity (32). 7. Stimulate cells by replacing buffer with 100 ml of Krebs/Hepes buffer containing agonist for the appropriate time (general time course employed 0, 5, 10, 60, 500 s). 8. Stop reaction with equal volume of 1 M trichloroacetic acid.

Protocol 7.

Extraction and neutralization of lns(1,4,5)P3

Method 1. Remove 160 ml of the reaction mixture obtained in Protocol 6, step 8. 2. Add 40 ml of 10 mM EDTA and 200 ml of Freon/tri-n-octylamine (1:1 v/v). Vortex the mixture and separate the phases by brief centrifugation in a microfuge. 3. Remove 100 ml of the upper phase, and to this add 50 ml of 25 mM NaHCO3. 4. Prepare buffer blanks for use in making up lns(1,4,5)P3 standards (see Protocol 9). Mix 400ml of Krebs/HEPES buffer with 400 ml 1 M TCA and then add 200 ml of 10 mM EDTA. This mixture is extracted with 1 ml of Freon/tri-n-octylamine. Remove 500 ml of the upper phase and to this add 250 ml of 25 mM NaHCO3.

68

Figure 3. Typical standard curve obtained for the lns(1,4,5)P3 radio-receptor assay. Cold lns(1,4,5)P3 at concentrations ranging from 0.036-36 pmol was used to displace [3H]lns(1,4,5)P3 (7349 d.p.m. per assay) from binding proteins present in an adrenal cortex preparation. Bound lns(1,4,5)P3 was separated from free by rapid filtration through GF/B filters. (Non-specific binding is defined at 1200 prnol cold lns(1,4,5)P3).

Protocol 8. Preparation of adrenal cortex lns(1,4,5)P3 binding protein Equipment and Reagents • Bovine adrenal glands

• Polytron

Method 1. Obtain bovine adrenal glands and cut each gland longitudinally. Remove the medulla and scrape the cortex from the outer capsule. 2. Dispense the cortex into 50 ml centrifuge tubes (~5 g tissue per tube), and homogenize in 30-40 ml of ice-cold buffer A per tube (20 mM NaHCO3, 1 mM dithiothreitol, pH 8.0). 3. Centrifuge the homogenate (5000 g, 10 min, 4oC). Pool the supernatants and extract the pellet again by rehomogenization in 30-40 ml of buffer A and recentrifugation. 4. Combine the supernatants from the two extractions, and discard the twice-extracted pellet. 69

Andrew B. Tobin et al. Protocol 8.

Continued

5. Obtain a P2 fraction by centrifugation of the supernatant (40000 g, 20 min, 4°C). Discard the supernatant and rehomogenize the pellet in buffer A. Recentrifuge (40 000 g, 20 min, 4 °C). Repeat this wash step of the P2 pellet three times. 6. Resuspend the washed P2 pellet in buffer A, and adjust protein concentration to 15-18 mg ml-1. 7. Aliquot the adrenal cortical preparation, and store at -20 °C. The preparation can be stored for 6-12 months.

Protocol 9. lns(1,4,5)P3 radio-receptor assay Equipment and Reagents • [3H]-inositol (1,4,5) trisphosphate (Amersham or NEN) . Inositol (1,4,5) trisphosphate K

« Adrenal cortex-binding protein (see Protocol 8) . Filtration apparatus or cell harvester (Millipore)

Method 1. To prepare the lns(1,4,5)P3 standards, dilute 40 mM stock lns(1,4,5)P3 using the buffer blank (Protocol 7) to give the following concentrations: 0, 1.2, 4, 12, 40, 120, 400, and 1200 nM. Non-specific binding is determined using 40 mM stock. Note that by using 30 ml of each standard in a final reaction volume of 120ml,the concentration of lns(1,4,5)P3 in the assay tubes ranges from 0.036-36 pmol lns(1,4,5)P3 per assay. An example of a typical standard curve is given in Figure 3. 2. Take 30 ml of neutralized extract obtained from Protocol 7, step 3, or 30 ml of lns(1,4,5)P3 standard, and add to 30 ml of a 4x concentrated assay buffer (100 mM Tris-HCI, 4 mM EDTA, pH 8.0). 3. To this mix add 30 ml of 3H-lns(1,4,5)P3 diluted in water to give 60008000 d.p.m. per assay (Stock [3H]-lns(1,4,5)P3 obtained from Amersham, 30-50 Ci mmol-1, or from NEN, 17-20 Ci mmol-1). 4. Start the reaction by adding 30 ml of adrenal cortex-binding protein. Total reaction volume is 120 ml. 5. Vortex the mixture and incubate on ice for 30 min. 6. Stop the reaction by adding 3 ml of ice-cold wash buffer (25 mM TrisHCI, 1 mM EDTA, 5 mM NaHC03, pH 8.0) and immediately filter through GF/B filter disks using rapid vacuum filtration (Millipore vacuum manifold). 70

4: G-protein coupled receptor phosphorylation 7. Wash filter disks with 2 x 3 ml of ice-cold wash buffer. Note that filtering procedure should be done as quickly as possible. 8. Transfer the filters to scintillation vials, add 4mls of scintillation fluid, and allow to extract overnight before counting.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Hausdorff, W. P., Caron, M. G., and Lefkowitz, R. J. (1990). FASEB. J., 4, 2881. Lohse, M. J. (1993). Biochim. Biophys. Acta, 1179, 171. Haga, T., Haga, K., and Kameyama, K. (1994). J. Neurochem., 63, 400. Hausdorf, W. P., Bouvier, M., O'Dowd, B. F, Irons, G. P., Caron, M. G., and Lefkowitz, R. J. (1989). J. Biol. Chem., 264, 12657. Fredericks, Z. L., Pitcher, J. A., and Lefkowitz, R. J. (1996). J. Biol. Chem., 271, 13796. Sterne-Marr, R., and Benovic, J. L. (1995). Vitamins and hormones, 51, 193. Lefkowitz, R. J. (1993). Cell, 74, 409. Tobin, A. B., and Waugh, M. G. (1996). In The phospholipase C pathway: its regulation and desensitisation (ed. A. B. Tobin) p. 65. Springer Press, New York. Tobin, A. B. (1997). Pharmacol. Ther., 75, 135. Benovic, J. F., Mayor, F., Staniszewski, C., Lefkowitz, R. J., and Caron, M. G. (1987). J. Biol. Chem., 262, 9026. Haga, K., and Haga, T. (1990). FEBS Lett., 268, 43. Haga, K., and Haga, T. (1992). J. Biol. Chem., 267, 2222. Pitcher, J. A., Inglese, J., Higgins, J. B., Arriza, J. K., Casey, P. L., Kim, C., Benovic, J. L., Kwatra, M. M., Caron, M. G., and Lefkowitz, R. J. (1992). Science, 257, 1264. Pitcher, J. A., Fredericks, Z. L., Stone, C. W., Fremont, R. T., Stoffel, R. H., Kock, W. J., and Lefkowitz, R. J. (1996). J. Biol. Chem., 271, 24907. DebBurman, S. K., Ptasienski, J., Benovic, J. L., and Hosey, M. M. (1996). J. Biol. Chem., 271, 22552. Klueppelberg, U. G, Gates, L. K., Gorelick, F. S., and Miller L. J. (1991). /. Biol. Chem., 266, 2403. Tobin, A. B., and Nahorski, S. R. (1993). J. Biol. Chem., 268, 9817. Tobin, A. B., Keys, B., and Nahorski, S. R. (1996). J. Biol. Chem., 271, 3907. Tobin, A. B., Totty, N. F., Sterlin, A. E., and Nahorski, S. R. (1997). J. Biol. Chem., 272, 20844. Waugh, M. G., Burford, N. T., Nahorski, S. R., and Tobin, A. B. (1995). Brit. J. Pharmacol., 114, 143P. Harlow, E., and Lane, D. (1984). Antibodies: a laboratory manual. Cold Spring Harbor Laboratory Press, NY. Ali, H., Richardson, R. M., Tomhave, E. D., Didsbury, J. R., and Snyderman, R. (1993). J. Biol. Chem., 268, 24247. Rinken, A., Kameyama, K., Haga, T., and Engstrom, L. (1994). Biochem. Pharmacol., 48, 1245. Tobin, A. B., Keys, B., and Nahorski, S. R. (1993). FEBS Lett., 335, 353.

71

Andrew B. Tobin et al. 25. Pei, G., Tiberi, M., Caron, M. G., and Lefkowitz R. J. (1994). Proc. Natl. Acad. Sci. USA, 91, 3633. 26. Palmer, T. M., Benovic, J. L., and Stiles, G. L. (1995). J. Biol. Chem., 270, 29607. 27. Blaukat, A., Alla, S. A., Lohse, M. J., and Muller-Ester, W. (1996). J. Biol. Chem., 271, 32366. 28. Wojcikiewicz, R. J. H., Tobin, A. B., and Nahorski, S. R.(1993). Trends Pharmacol. Sci., 14, 279. 29. Tobin, A. B., Lambert, D. G., and Nahorski, S. R. (1992). Mol. Pharmacol., 42, 1042. 30. Challiss, R. A. J., Batty, I. H., and Nahorski, S. R. (1988). Biochem. Biophys. Res. Commun., 157, 684. 31. Challiss, R. A. J. (1995). In Signal transduction protocols. (ed. D. A. Kendall, and S. J. Hill). p. 167. Humana Press Inc, Totowa, NJ. 32. Willars, G. B., and Nahorski, S. R. (1995). Mol. Pharmacol., 47, 509.

72

5

G proteins and their identification IAN MULLANEY

1. Introduction The demonstration that G proteins mediate many intracellular signalling processes has led to the development of specific techniques that can be used to identify which of these polypeptides is involved upon receptor activation by ligand. This chapter deals with the methodology involved to produce specific immunological tools, and the use of these reagents to probe functionally the specificity of receptor-G protein interaction. Many hormones and neurotransmitters exert their actions on target cells through changes in the levels of a number of intracellular second messengers via activation of cell surface receptors. Activation of these receptors initiates a sequence of biochemical cascades, leading to certain physiological events such as contraction of smooth muscle, neurotransmitter release, etc. The discovery of the critical role that guanine nucleotides play in mediating hormonal stimulation of adenylyl cyclase, the enzyme directly responsible for modulating intracellular cyclic AMP levels, led to the isolation and purification of the first of a number of closely related guanine nucleotide-binding proteins [or G proteins], Gs (1). These polypeptides, which function through the cyclical binding and hydrolysing of GTP, have proved to be central in transducing the effects of activated heptahelical integral membrane receptors. The 'classical' G proteins exist as heterotrimers comprising non-identical a, B, and y subunits. Initial identification of G proteins involved the ability of the a subunit to act as a substrate for mono-ADP-ribosylation, catalysed by the ADP-ribosyl transferase activity of a number of bacterial exotoxins, a modification that functionally altered the involvement of the G protein in signal transduction. The use of [32P] NAD+ as a substrate allowed the visualization of ADP-ribosylated polypeptides following separation in SDS-PAGE gels and autoradiography. In this way Gs, the G protein involved in the hormonal stimulation of adenylyl cyclase, was identified as a substrate for ADPribosylation by cholera toxin (2). Similarly, Gi, the inhibitory G protein, was characterized by its ability to act as a substrate for pertussis toxin (3-5). Although the use of toxins yields little information about the molecular identity of the G proteins involved, it can be useful in the initial investigation

Ian Mullaney Table 1. The peptide sequences which have been widely used to generate a series of antipeptide antisera directed against the a subunits of various G proteins Peptide used

G protein sequence

Antiserum identifies

*RMHLRQYELL TPEPGEDPRVTRAKY *KENLKDCGLF LERIAQSDYI *KNNLKECGLY *ANNLRGCGLY GCTLSAEERAALERSK NLKEDGISAAKDVK *QLNLKEYNLV EKVSAFENPYVDAIKS EKVTTFEHQYVNAIKT *QENLKDIMLQ *HDNLKQLMLQ *QLNLREFNLV *ARYLDEINLL *QNNLKYIGLC

Gsa 372-381 Gsa 325-339 TD1a 341-350 Gi2a 160-169 Gi3a 345-354 G0a 345-354 G0a 1-16 G0a 22-35 Gqa 350-359 Gqa 119-134 G11a 119-134 G12a 370-379 G13a 368-377 G14a 346-355 G15a 365-374 Gza 346-355

14 19 T D , a , TD2a, Gi1a, Gi2a20 14 Gi2a 14,21 Gi3a 14 G0a G0a 21 22 G0a 23 Gqa,G11a (G14a?) 24 Gqa 24 G11a G12a 6 G13a 25 7 G14a, Gqa, G11a 7 G16a, G16a 26 Gza

Reference

Gsa Gsa

* denotes C-terminal sequences. Amino acids are represented using the one letter code. TD = transducin

into G protein function in a particular system. However, a number of G protein a subunits have shown themselves to be refractory to influence by these toxins. The ability of agonists to activate phospholipase CBl, and cause hydrolysis of inositol-containing phospholipids, has been shown to be unaffected by treatment with both pertussis and cholera toxin in the vast majority of cells and tissues studied, implying the involvement of toxininsensitive G proteins (6, 7). Indeed, the use of the polymerase chain reaction, based on conserved sequence domains across the G protein family, and the isolation of cDNAs has now identified at least 17 G protein a subunits, including members of two new, toxin-insensitive subfamilies (Gq and G12). The elucidation of G protein primary amino acid sequences has allowed for the production of specific antisera directed against these proteins, using specific peptide stretches known to be both unique and functionally important (see Table 1). In this chapter, methods designed to probe the identity of G protein a subunits will be outlined. These will include the use of bacterial toxins, resolution of polypeptides by SDS-PAGE, and production of specific G-protein antisera. Further, the use of these tools to probe receptor-G protein interactions will be presented, specifically, the immunoprecipitation of [35S]-labelled product, and agonist-induced cholera toxin-mediated ADP-ribosylation. Finally, functional assays that use the guanine nucleotide activation-deactivation cycle will be described. 74

5: G proteins and their identification

2. Production of crude plasma membrane fractions for analysis of G proteins Since the majority of G protein assays use cell-free systems, crude plasma membrane fractions remain the most common biological preparation in which to study the function of these polypeptides. Isolation of plasma membrane sheets by the differential centrifugation method described by Koski and Klee has become the standard technique in the field (8). Immunoblotting of supernatant fractions produced during membrane production routinely shows little immunoreactivity, reinforcing the concept that the G protein a subunits are located at membranes. Protocol 1 describes the preparation of plasma membrane from cells grown in culture, whilst Protocol 2 outlines a method for preparing these fractions from intact animal tissue. We have found it beneficial, in terms of final yield, to store tissue and cell pastes at -80°C before homogenization. This freezing procedure fractures the cells, making them more amenable to rupture by the tissue grinder. However, it is advisable to keep the pellet produced from the first centrifugation step on ice. This pellet contains mainly unbroken cells and can, if the final product yield is low, be rehomogenized and re-fractionated. Although early preparations did not include the use of protease inhibitors in the homogenization buffer, it has become clear that some elements of the signalling cascade can be sensitive to proteolytic degradation. The inclusion of common inhibitors in the buffer is recommended. Since many of the techniques involved in investigating G protein function rely on resolution by electrophoretic techniques, the method involved in preparation of samples for analysis is outlined in Protocol 3. Protocol 1. Preparation of crude plasma membrane fraction from cells grown in cell culture Equipment and reagents • Refrigerated ultracentrifuge and benchtop centrifuge • Tight fitting Teflon-on-glass homogenizer • Phosphate-buffered saline (PBS) (0.2 g KCI, 0.2 g KH2PO4, 8 g NaCI, 1.14 g Na2HP04 (anhydrous), pH 7.4, H20 to 1000 ml) . TE buffer (10 mM Tris-HCI, 0.1 mM EDTA, pH 7.5)

• Spectrophotometer . plasma membrane buffer (TE buffer containing the following protease inhibitors: 10 mM NaF, 100 MM Na3VO4, 1 mM phenylmethanesulfonyl fluoride, 3 mM benzamidine, 0.1 mM soybean trypsin inhibitor, 10 mM. Ieupeptin, 0.2 mM aprotinin, 1.5 mM antipam)

Method 1. Gently remove cells from the surface of the flasks with a Pasteur pipette or rubber policeman, collect in a 50 ml conical centrifuge tube on ice, and centrifuge at 1000 r.p.m. at 4°C for 5 min on a benchtop centrifuge.a

75

Ian Mullaney Protocol 1. Continued 2. Discard the supernatant, resuspend the cell pellet in 30 ml ice-cold PBS, and centrifuge as before. 3. Repeat this procedure twice, and store the resultant washed cell paste at -80°C until needed. 4. Thaw the frozen cell pastes, and resuspend in 2 ml of ice-cold TE buffer containing the protease inhibitor cocktail, and homogenize with 20 strokes of a Teflon-on-glass tissue grinder. 5. Centrifuge the homogenates for 10 min at 500 g in an ultracentrifuge. 6. Discard the pellet, and recentrifuge the supernatant at 48000 g for 10 min. 7. Discard the supernatant, resuspend the pellet in 5 ml of plasma membrane buffer, and recentrifuge for 10 min at 48000 g. 8. Finally, discard the supernatant, and resuspend the pellet in plasma membrane buffer. 9. Triturate the resuspended pellet with a syringe with a fine-gauge needle, and aliquot into appropriate volumes and store at -80°C until needed. 10. Determine protein concentrations. Membranes should have protein concentrations of 1-2 mg ml-1. a

The protocol is suitable for between 107 to 1012 cells per preparation.

Protocol 2. Preparation of crude plasma membrane fraction from whole animal tissue Equipment and reagents • Benchtop centrifuge and refrigerated ultracentrifuge • Polytron • Spectrophotometer

• PBS (see Protocol 1) . TE buffer (see Protocol 1) . plasma membrane buffer (see Protocol 1)

Method 1. Remove tissue (up to 10 g), and briefly wash in 10 volumes of PBS 2. Chop the tissue with scissors, rinse in two washes of PBS, and homogenize on ice for 60 s (4 X 15s bursts) in 10 volumes of ice-cold plasma membrane buffer with a polytron at setting number 4. 3. Follow steps 5-10 in Protocol 1. 76

5: G proteins and their identification Protocol 3. TCA-deoxycholate precipitation of samples for SDSPAGE Equipment and reagents • Microcentrifuge and benchtop centrifuge . 2% (w/v) 7-deoxycholic acid, sodium salt, in H2O . 24% (w/v) TCA in H20 • 1 M Tris base (do not adjust the pH)

• Laemmli sample buffer (0.605 g Tris, 30 g urea, 5 g SDS, 6 g dithiothreitol, 10 mg bromophenol blue, in 100 ml H2O, pH adjusted to 8.0 with HCI)

Method 1. Take 25-150 ug of membranes, and place on ice in a 1.5 ml Eppendorf centrifuge tube. 2. Centrifuge for 5 min at 12000 r.p.m. on a microcentrifuge, remove the supernatant, and resuspend the pellet in 20 ul of TE buffer. 3. Add 6.5 ul of 2% (w/v) 7-deoxycholic acid to each tube, followed by 750 ul of double-distilled H2O, then 250 ul 24% (w/v) TCA. a 4. Vortex mix each sample, and spin in a benchtop microcentrifuge for 10 min at 12 000 r.p.m. 5. Carefully discard the supernatant, and bring the pellet to weakly alkaline pH by addition of 20 ul 1 M Tris base.b 6. Add 20 ul Laemmli sample buffer, and load sample onto gel. a

The solutions must be added in the order 7-deoxycholate, H2O, and then TCA. Solutions can be kept indefinitely at room temperature. b If the blue Laemmli buffer solution turns brown-orange then the sample is still acidic. Add 5 ul amounts of 1 M Tris base until the sample turns blue.

3. Gel electrophoresis of G proteins 3.1 Mono-ADP-ribosylation of G proteins by bacterial exotoxins Until specific G protein antisera became available, the ability of a subgroup of G proteins to act as targets for bacterial exotoxins from Vibrio cholerae (cholera toxin) and Bordetella pertussis (pertussis toxin or islet-activating protein) was the most widely used approach to identify and probe the function of these polypeptides. These toxins, which exert their effects by catalysing the covalent transfer of an ADP-ribose moiety on to the G protein a subunit, were initially noted for their ability to disrupt receptor-mediated control of adenylyl cyclase. Whilst cholera toxin is able catalyse ADPribosylation of a variety of proteins in vitro, its key substrates are the splice variant forms of the stimulatory G protein of adenylyl cyclase, Gsa where an 77

Ian

Mullaney

Figure 1. G12 but not G9 is a substrate for in vivo pertussis toxin-catalysed ADP-ribosylation. Membranes from U937 cells which were either untreated (lanes 1, 3) or pretreated with pertussis toxin (25 ng ml-1, 16 h) as described in Protocol 4 were resolved by SDS-PAGE, and immunoblotted with antisera directed against G12 (panel A) or G9 (panel 2) (see Protocols 7, 75). (Data from Mitchell et al., 1991, ref. 23).

arginine residue acts as the substrate amino acid. In contrast, pertussis toxin modifies a number of G proteins, including the inhibitory G protein of the adenylyl cyclase cascade, which have acceptor cystcinc residues located four amino acids from the carboxyl terminus. To date, this family of pertussis toxin substrates includes G11 G12, G13, G01, G02, and both forms of transducin. Protocol 4 outlines the method for toxin-mediated ADP-ribosylation of cultured cells in vivo. Both toxins are members of the A-B type toxin family, comprising two components, the A (active) component that consists of the ADP-ribosyltransferase that catalyses the transfer of ADF-rihose to the acceptor amino acid residue, and the R (binding) component lhat binds to the ceil surface, enabling the A component to enter the cell. The use of these toxins in conjunction with specific immunological tools has proved invaluable in identifying the G protein family. It should be recognized that although many G proteins are toxin-sensitive, a sizeable number are refractory to the toxins. Figure l shows an immunoblot of untreated and pertussis toxin-pretrealed membranes prepared from the monocyte-dcrived cell line, U937. The specific G12 antiserum detected a single immunoreactive species that migrated more slowly in membranes prepared from the toxin-pre-treated cells. In contrast, the mobility of G q , which does not contain the acceptor cysteine, was unaffected by the toxin, confirming its insensitivity to pertussis toxin, 78

5: G proteins and their identification

Protocols

Mono-ADP-ribosylation of cultured cells by bacterial toxins (in vivo)

Equipment and reagents • Pertussis toxin (0.44 mg ml-1) • Cell culture medium

• Cell culture facilities • Cholera toxin (1 mg ml-1)

Method 1. Use cells at 70-80% confluence, and replace with culture medium containing either cholera or pertussis toxin at a final concentration of 100 ng toxin per ml of medium. 2. Incubate for 16-24 h, and harvest the cells. 3. Prepare the plasma membrane fraction (see Protocol 1).

Cholera and pertussis toxins can facilitate the transfer of ADP-ribose to plasma membrane fractions in vitro (Protocol 5). The use of these cell-free

Figure 2. Cholera and pertussis toxin-catalysed ADP-ribosylation in membranes of NG108-15 cells. Membranes (25 (i.g) of control (a, c, d, f) or 6-day dibutyryl cAMP differentiated (b, e) NG108-15 cells were treated with [32P]NAD1 and cholera toxin (a, b), pertussis toxin (d, e), or without toxin (c, f) as described in Protocol 4. Samples were recovered by deoxycholate-TCA precipitation and resolved by SDS-PAGE (10% w/v acrylamide) (see Protocols 5, 76). (Data from Mullaney et al., 1988, ref. 22),

79

Ian Mullaney systems has made the need for the binding component of the toxin redundant. Preactivation of the toxins with dithiothreitol to dissociate the binding apparatus from the catalytic component is recommended. Further, the use of the radiolabelled substrate [32P]NAD+, combined with SDS-PAGE and autoradiography, allows direct visualization of the polypeptides (9). Figure 2 shows plasma membranes prepared from the neuroblastoma X glioma cell line, NG108-15, and subjected to mono-ADP-ribosylation with radiolabelled NAD+ and either pertussis or cholera toxin. Protocol 5. Mono-ADP-ribosylation of membranes by bacterial toxins (in vitro) Equipment and reagents • Benchtop centrifuge, microcentrifuge, and refrigerated ultracentrifuge • Water bath • Cholera toxin (1 mg ml-1) • Pertussis toxin (0.44 mg ml-1) . Nicotinamide adenine dinucleotide, diftriethylammonium) salt, [adenylate-32P] (product no. NEG023), specific activity 10-50 Ci mmol-1 (NEN-Dupont)

1.5 M sodium phosphate buffer ADP-ribosylation cocktail mix (for 15 samples): 75 ul 0.2 M thymidine, 30 ul 1 mM GTP, 125 ul 1.5 M sodium phosphate buffer, 10 ul 0.04 M ATP, 15 ul 1 M arginine hydrochloride, 30 uCi [32P] NAD+ to 300 ul H2O Laemmli sample buffer (see Protocol3) 100 mM dithiothreitol (DTT)

Method 1. Pre-activate toxins by addition of equal volumes of 100 mM DTT for 60 min at room temperature prior to assay. 2. Take 25-50 u,g of plasma membrane fraction (see Protocol 1), and place on ice in a 1.5 ml Eppendorf centrifuge tube. 3. Centrifuge for 5 min at 12000 r.p.m. in a microcentrifuge, remove the supernatant, and resuspend the pellet in 25 ul of TE buffer. 4. Add 20 ul of the ADP-ribosylation cocktail mix to each tube, and start the incubation by adding 5 ul of the appropriate pre-activated toxin. a 5. Incubate for up to 90 min in a 37°C water bath. 6. Place on ice and precipitate samples using the TCA-deoxycholate method (Protocol 3). 7. Add 20 ul Laemmli sample buffer, and load the sample onto the gel (Protocol 7). a

Add 5 ul of 50 mM DTT instead of toxin for negative control

Under assay conditions where no exogenous GTP is added, it is possible for pertussis toxin-sensitive G proteins to act as substrates for mono-ADPribosylation by cholera toxin in plasma membrane preparations (Protocol 6). Further, addition of selective agonist markedly enhances incorporation of 80

5: G proteins and their identification radioactivity into 40 kDa (Gi-like) polypeptides, but not Gs, in a dosedependent manner. Used in conjunction with immunoprecipitation with specific G protein antisera, this provides a method to identify direct receptor activation of G protein a subunits (10,11) Protocol 6. Agonist-mediated mono-ADP-ribosylation of membranes by cholera toxin Equipment and reagents • Benchtop centrifuge, microcentrifuge, and • 1.5 M sodium phosphate buffer refrigerated ultracentrifuge . ADP-ribosylation cocktail mix (for 15 sam• Water bath ples): 75 ul 0.2 M thymidine, 125 ul 1.5 M . Cholera toxin (1mg ml-1) sodium phosphate buffer, 10 ul 0.04 M . Nicotinamide adenine dinucleotide, di(tri- ATP, 15 ul 1M arginine hydrochloride, 30 ethylammonium) salt, [adenylate-32P] (produCi [32P] NAD+ to300ulH2O° uct no. NEG023), specific activity 10-50 • Laemmli sample buffer -1 Ci mmol (NEN-Dupont) . 100 mM dithiothreitol

Method 1. Pre-activate toxin by addition of an equal volume of 100 mM DTT for 60 min at room temperature prior to assay. 2. Take 25-50 u.g of plasma membrane fraction (see Protocol 1), and place on ice in a 1.5 ml Eppendorf centrifuge tube. 3. Centrifuge for 5 rnin at 12000 r.p.m. on a microcentrifuge, remove the supernatant, and resuspend the pellet in 20 ul of TE buffer. 4. Add 20 ul of the ADP-ribosylation cocktail mix and 5ul of an appropriate agonist to each tube, and start the incubation by adding 5 ul of the appropriate pre-activated toxin. 5. Incubate for up to 2 h in a 37°C water bath. Terminate the reaction by transfer of tubes to ice. 6. Samples can either be precipitated using the TCA-deoxycholate method (Protocol 3) and separated by SDS-PAGE, or can be immunoprecipitated with G protein antiserum (Protocol 78) prior to separation with SDS-PAGE (Protocol 7)

3.2 Gel electrophoresis of G proteins Adaptations of the discontinuous SDS-PAGE technique developed by Laemmli provide the single most important method by which to separate the various members of the G protein superfamily (11). Although the toxinmediated incorporation of radiolabel appears to consist of single bands under the gel conditions used in Figure 2 (10% (w/v) acrylamide plus 0.27% (w/v) bisacrylamide) (Protocol 7), they actually represent the total pools of substrates for each particular toxin. Indeed the majority of G protein a subunits 81

Ian Mullaney have predicted molecular masses of between 39 and 45 kDa, and gel conditions and sample treatments have been optimized to separate polypeptides within this narrow molecular mass range. The most useful of these has been to reduce the concentration of bisaerylamide in the resolving gel from 0.27 to 0.06% (w/v). Figure 3 shows membranes prepared from NG108-15 cells that have been subjected to [ 32 P]ADP-ribosylation with pertussis toxin, and the radiolabclled products resolved by SDS-PAGE on 12.5% (w/v) acrylamide/ 0.06% (w/v) bisaerylamide gel (Protocol 8). Three pertussis toxin-sensitive G proteins are clearly separated in these cells using this gel system (panel A), that were identified as G(). G12, and G13, using specific antisera (panels B-D). In addition, treatment of the sample with N-ethylmaleimide differentially alkylates the a subunits of the pertussis toxin-sensitive G proteins G11, G12, G13, and the isoforms of G0 on accessible cysteine residues (Protocol 9). This has the effect of altering the migration of these a subunits on SDS-PAGE, with the result that it is possible to obtain greater resolution of the G0 isoforms from the Gi-like G proteins. If sample alkylation is performed in conjunction with resolution on a 12.5% acrylamide/0.06% bisaerylamide gel, the separation achieved can be dramatic. This technique is particularly useful when trying to identify G proteins with antisera that cross-react, a common example being that antisera directed against the carhoxyl terminus of G n can cross-react with G13 because of the presence of an immunodominant tyrosine in the peptide sequence of these two polypeptides. To separate the isoforms of G0a or G proteins of the Gq family, the best strategy is to resolve the membranes on 12.5% acrylamide gels containing 6 M urea (Protocol 10). It is also possible to separate these proteins on SDS-PAGE gels which contain a 4-8 M urea gradient, or on two-dimensional gels. However, both of these methods arc technically move difficult and have little advantage over the 6 M

Figure 3. Resolution and identification of pertussis toxin-sensitive G proteins. Membranes from NG108-15 cells 1100 ug) were treated with thiol-activated pertussis toxin and [32P]NAD1 for 90 min at 37 o C and resolved by SDS-PAGE (12.5% w/v aeryIamide, 0,06% w/v bisacrylamide) (see Protocols 4, 70). Samples were transferred to nitrocellulose, and either autoradiographed (panel Al or immunoblotted with antisera directed against G13 (panel B), G12 (panel B), and Gn (panel B).

82

5: G proteins and their identification Table 2. SDS PAGE gel compositions a Solution Resolving gel Gel buffer 1 Acrylamide solution 1 Acrylamide solution 2 50%(v/v)glycerol 10%(w/v)APS TEMED H2O Stacking get Gel buffer 2 Acrylamide solution 1 10%(w/v)APS TEMED H2O

10% (ml)

12.5% (ml)

6.0 8.0 0.0 1.6 0.09 0.01 8.2

6.0 0.0 10 2.0 0.09 0.01 5.8

3.75 1.5 0.15 0.01 9.75

3.75 1.5 0.15 0.01 9.75

a

The volumes given are for one gel (180 mm x 160 mm with spacers of 1.5 mm) run as part of a Bio-Rad Protean I electrophoresis apparatus.

urea SDS-PAGE gel system. It should be noted that polypeptides do not run normally on SDS-urea gels. They may appear to run at different molecular masses from those seen in the absence of urea, making it more difficult to identify particular a subunits. It is suggested that purified or recombinant a subunits first be resolved on these gels and used as reference standards. Table 2 outlines the amounts of reagents needed to produce the various gels. Protocol 7. SDS-PAGE (10% (w/v) acrylamide) of G protein a subunits Equipment and reagents • Hamilton syringe • Gel buffer 2 solution (6 g Tris, 4 ml 10% . SDS-PAGE gel apparatus and power pack (w/v) SDS in 100ml H2O, pH 6.8) (e.g. Bio-Rad Protean I electrophoresis • 50% (v/v) glycerol in H2O system) . TEMED . 10% (w/v) SDS in H2O . 10% (w/v) ammonium persulfate in H2O . Acrylamide solution 1 (30 g acrylamide, 0.8 . Electrophoresis running buffer (6 g Tris, g bisacrylamide in 100 ml H2O) 28.8 g glycine, 20 ml 10% (w/v) SDS in 2000 • Gel buffer 1 solution (18.17 g Tris, 4 ml 10% ml H2O. Do not adjust pH) (w/v) SDS in 100 ml H2O, pH 8.8)

Method 1. Set up the gel apparatus according to the manufacturer's guidelines. Add all gel reagents in the order and amounts given in Table 2 into a 250 ml conical flask and mix gently. Cast the gel using a Pasteur pipette. 83

Ian Mullaney Protocol 7.

Continued

2. Carefully overlay the cast gel with approximately 1 ml of 0.1% (w/v) SDS, and allow the gel to polymerize. This should take between 1 and 2 h at room temperature. 3. After polymerization, remove SDS overlay, and wash the gel with distilled water to remove any remaining traces of SDS. Add all stacker gel reagents in the order and amounts given in Table 2 into a 100 ml conical flask and mix gently. Pour the stacker gel on top of the resolving gel, and place the well-forming comb in the top of the gel, ensuring that no air bubbles are trapped under the comb, and leave to polymerize for 1 h at room temperature. 4. After polymerization, remove the sample-well comb, and place the gel in the gel tank containing enough running buffer in the base to cover the bottom edge of the gel, and add the remaining running buffer to the top. 5. Load the prepared samples in the preformed wells using a Hamilton syringe. 6. Run the gel overnight (approximately 16 h) at 60 V and 15 mA per plate, until the dye front reaches the bottom of the gel plates.

Protocol8.

SDS-PAGE (12.5% (w/v) acrylamide) of G protein a subunits

Equipment and reagents • Hamilton syringe • Gel buffer 2 solution (6 g Tris, 4 ml 10% . SDS-PAGE gel apparatus and power pack (w/v) SDS in 100 ml H2O, pH6.8) (e.g. Bio-Rad Protean I electrophoresis • 50% (v/v) glycerol in H2O system) . TEMED . 10% (w/v) SDS in H2O . 10% (w/v) ammonium persulfate in H2O . Acrylamide solution 2 (30 g acrylamide, . Electrophoresis running buffer (6 g Tris, 0.15 g bisacrylamide in 100 ml H2O) 28.8 g glycine, 20 ml 10% (w/v) SDS in 2000 • Gel buffer 1 solution (18.17 g Tris, 4 ml 10% ml H2O. Do not adjust pH) (w/v) SDS in 100 ml H2O, pH 8.8)

Method 1. Set up the gel apparatus according to the manufacturer's guidelines. Add all gel reagents in the order and amounts given in Table 2 into a 250 ml conical flask, and mix gently. Follow steps 1-5 from Protocol 7. 2. Run the gel overnight (approximately 18-20 h) at 100 V and 15 mA per plate, until the dye front reaches the bottom of the gel plates. 84

5: G proteins and their identification Protocol 9.

NEM treatment of samples for SDS-PAGE

Equipment and reagents • Microcentrifuge • 100'C heating block

. 100 mM NEM in H2O • Laemmli sample buffer (see Protocol 3)

Method 1. Take 25-150 ug of membranes, and place on ice in a 1.5 ml Eppendorf centrifuge tube. 2. Centrifuge for 5 min at 12000 r.p.m. in a microcentrifuge, remove the supernatant, and resuspend the pellet in 20 ul of TE buffer. 3. Add 10 ul of 5% (w/v) SDS, 50 mM DTT, and incubate at 90°C for 5 min. 4. Cool the samples on ice, add 10 ul of freshly prepared 100mM NEM to each tube, and leave at room temperature for 20 min. 5. Add 20 ul Laemmli sample buffer, and load the sample onto the gel.

Protocol 10. SDS-urea PAGE (12.5% (w/v) acrylamide) of G protein a subunits Equipment and reagents 10% (w/v) SDS in H2O • 6 M urea-gel buffer 2 solution (6 g Tris, 4 ml 10% (w/v) SDS, 36.025 g urea in 100 ml H2O, pH 6.8) > 50% (v/v) glycerol in H2O i TEMED > 10% (w/v) ammonium persulfate in H2O Electrophoresis running buffer (6 g Tris, 28.8 g glycine, 20 ml 10% (w/v) SDS in 2000 ml H2O. Do not adjust pH)

• Hamilton syringe • SDS-PAGE gel apparatus and power pack (e.g. Bio-Rad Protean I electrophoresis system) • 6 M urea-acrylamide solution 2 (30 g acrylamide, 0.15 g bisacrylamide, 36.025 g urea in 100 ml H2O) • 6 M urea-gel buffer 1 solution (18.17 g Tris, 4 ml 10% (w/v) SDS, 36.025 g urea in 100 ml H2O, pH 8.8)

Method 1. Set up the gel apparatus according to the manufacturer's guidelines. Add all gel reagents in the order and amounts given in Table 2 for a 12.5% SDS-PAGE gel into a 250 ml conical flask, and mix gently.8 2. Follow steps 1-5 from Protocol 7. 3. Run the gel overnight (approximately 18-20 h) at 120 V and 50 mA per plate, until the dye front reaches the bottom of the gel plates. ' If the room temperature is on the cold side, there is the possibility that the urea will come out of solution as the gel is polymerizing. To avoid this, allow the gel to set in a warmer room, e.g. tissue culture room, 30°C hot room, etc.

85

Ian Mullaney The detection of radiolabelled polypeptides is best done by autoradiography (Protocol 11} using conventional X-ray film. However, newer technologies such as phosphorimage analysis allow greater sensitivity and easier quantification of results (Protocol 12). Protocol 11. Autoradiography Equipment and reagents • • • • •

Imaging densitometer Gel drier and vacuum pump Whatman 3 mm filter paper X-ray film developing apparatus -80°C freezer

Coomassie stain buffer (0.25% (w/v) Coomassie Brilliant Blue R-250, 45% (v/v) methanol, 10% (v/v) glacial acetic acid) Gel destaining buffer (45% (v/v) methanol, 10% (v/v) glacial acetic acid)

Method 1. Remove the SDS-PAGE gel with the resolved radiolabelled polypeptides from the electrophoresis apparatus, and soak for 1 h in Coomassie stain buffer. 2. Remove the stain buffer, add the destaining solution, and leave for 2-3 h. 3. Remove destaining buffer, and dry the gel onto Whatman 3 mm filter paper under suction from an electric vacuum pump attached to a gel drier at 70°C for 2 h. 4. Transfer the dried gel to a Kodak X-o-matic cassette with intensifying screens (or similar) containing Kodak X-omat S X-ray film, and allow to autoradiograph at -80°C for an appropriate time. 5. Develop the film (e.g. on a Kodak X-o-mat developing machine), and quantify the autoradiograph using an imaging densitometer.

Protocol 12.

Phosphorimaging

Equipment • Phosphorimager

Method 1. Follow steps 1-3 in Protocol 11. 2. Transfer the dried gel to a phosphorimager plate, and leave for an appropriate length of time. 3. Develop the image using a phosphorimager, and quantify the resultant image. 86

5: G proteins and their identification

4. Immunological methods Initial attempts to define the specificity of cellular signalling systems made great use of the bacterial exotoxins produced by Vibrio choleras and Bordetella pertussis (2-5). These toxins were found to modulate the inhibitory and stimulatory mechanisms of cyclic AMP regulation by causing mono-ADPribosylation of those G proteins that transduced receptor interactions with adenylyl cyclase. However, the realization that multiple G protein gene products could act as substrates for these toxins within a single cell or tissue made the need for more selective, discriminatory tools. Immunological probes were first generated against purified G protein preparations, and although these proved extremely useful in the initial identification of G proteins, the problems of obtaining homogeneous polypeptide preparations meant that their use was limited. It was the isolation of cDNA species corresponding to the G protein a subunits that revolutionized our understanding of these molecules, allowing the generation of specific antisera directed against short peptide stretches unique to particular proteins (for a review, see ref. 13). To date, the primary amino acid sequences of seventeen G protein a subunits have been deduced, allowing the generation of these antisera for use as specific immunological tools. Although the G protein superfamily is highly conserved, there are regions of sequence variation, particularly in the carboxy terminal region, which contains the receptor coupling sites, that have been successfully used to produce specific polyclonal antipeptide antisera (see Table 1), Such antisera are generally produced by subcutaneous injection of peptide conjugated to carrier protein into rabbits, and can be tested either in ELISA assays against the antigen peptide, or by immunoblotting using either purified or recombinant G protein a subunits. This section outlines the procedures needed in the production and characterization of specific antipeptide antisera. In addition, methods that use these immunological tools are also described.

4.1 Immunization and serum collection Protocol 13 outlines a method for the production of antipeptide antisera directed against specific peptides corresponding to specific regions of G protein a subunits in commercially purchased New Zealand White rabbits (14). Protocol 14 describes the steps needed to harvest the serum for laboratory use. Pre-immune blood samples should be taken from each of the animals prior to injection, and the serum checked for any significant titre or immunological identification of cellular proteins. All antisera produced by this method should be tested for specificity against other C-terminal G protein sequences and non-related peptides by ELISA, to test for non-specific immunoreactivity (see Protocol 15). ELISA determinations use plates precoated with the immunogenic peptide and with peptides corresponding to the 87

Ian Mullaney

Figure 4. ELISA reactivity of antiserum raised against the carboxyl terminal peptide of Gqa/Gna against other C-terminal peptides of G proteins. ELISA assays were performed by the method described in Protocol 15 with various antiserum dilutions using 100 ng of peptides corresponding to the C-terminal decapeptides of Gq/G11 (open dotted squares), G2 (open diamonds), G11 + G12 (filled squares), G13 (open squares), and G0 (filled diamonds). (Data from Mitchell et al., 1991, ref. 23).

equivalent regions of other G protein a subunits. Half-maximal antiserum dilutions for useful antisera in immunoblots and immunoprecipitations are typically between 1:10 000 and 1:100 000, although it should be stressed that ELISA alone gives little information on the suitability of candidate antisera for these techniques. Figure 4 shows an ELISA performed on an antiserum raised against the terminal decapeptide of Gq with carboxyl terminal peptides corresponding to a variety of G protein a subunits. Immunoreactivity is specific to Gq peptide, with a half-maximal dilution value in excess of 1:100000. Alternatively, immunoblot analysis using antisera with added peptide to compete specific binding may prove useful in determining antiserum specificity. Table 1 contains peptide sequences commonly used in G protein anti-peptide antisera production. 88

5: G proteins and their identification Protocol 13. Immunization of New Zealand white rabbits Equipment and reagents • Probe sonicator . 21 mM glutaraldehyde in H2O • Keyhole limpet haemocyanin • 0.1 M Na phosphate buffer (pH 7.0) • Freund's complete and incomplete adjuvants • Peptide

Method 1. Dissolve 10 mg of keyhole limpet haemocyanin and 3 mg of peptide in 1 ml of 0.1 M Na phosphate buffer (pH 7.0). 2. Add 0.5 ml of 21 mM glutaraldehyde dropwise with stirring, and incubate at room temperature overnight. 3. Mix with an equal volume of Freund's complete adjuvant, and sonicate at full power for 20 s. 4. Immediately after sonication, inject the resultant emulsion in 0.2 ml volumes into multiple subcutaneous sites in the rabbit. Immunizations are normally performed simultaneously into two rabbits to maximize successful antibody production. 5. After two weeks, give each animal a booster immunization with material prepared identically, except that one-half as much peptide and keyhole limpet haemocyanin are injected in Freund's incomplete adjuvant.

Protocol 14. Serum collection Equipment • Benchtop centrifuge

• Glass universal for serum collection (20 ml)

Method 1. Four weeks after the booster injections, bleed the animals from ear arteries, collect the blood into glass universals, and allow to clot overnight at 4 °C. a 2. Remove the straw-coloured serum from the clot, and centrifuge at 1000 r.p.m. for 5 min on a benchtop centrifuge to remove any residual erythrocytes. 3. Aliquot the serums in appropriate volumes (100-200 ul), and store at -20°C. * Use glass universals; plasticware does not allow the clot to shrink to allow serum harvest.

89

Ian Mullaney Protocol 15. ELISA Equipment and reagents • . . . .

ELISA plate reader • ELISA secondary antibody solution (1:1000 96-well ELISA plates dilution of horseradish peroxidasePeptide solution (10ugml-1) conjugated donkey anti-rabbit IgG, diluted in ELISA PBS (see Protocol 1} antibody carrier solution) ELISA blocking solution (1 g powdered milk . Citrate-phosphate buffer (17.9 ml 0.1 M in 100 ml PBS) citric acid; 32.1 ml 0.2M Na2HPO4; 50 ml . PBS-Tween 20 (0.5 ml Tween 20 in 1000 ml H2O, pH 6.0). PBS) • H2O2 solution (10 ul of stock H2O2 in 10 ml

. ELISA antibody carrier solution (0.05 ml Tween 20, 0.1 g powdered milk in 100 ml PBS) . ELISA primary antibody solution (G protein antiserum diluted in ELISA antibody carrier solution)"

H2O) « o-phenylenediamide dihydrochloride (OPD) substrate solution (4 mg OPD, 9 ml citratephosphate buffer, 1 ml H202 solution) • 2 M H2S04

Method 1. Coat a 96-well ELISA plate with antigen peptide by adding 100 ul of 10 ug ml-1 peptide solution to each well, covering the plate with cling film, and incubating overnight at 4°C. 2. Remove the liquid, and wash each well twice with PBS. 3. Blot dry, add 100 ul of ELISA blocking solution to each well, and incubate at 37°C for 1 h. 4. Remove blocking solution, wash each well twice with PBS-Tween 20, and blot dry. 5. Add 100 ul of increasing dilutions of antiserum (from 1:10 to 1:100 000) to each well, cover the plate with cling film, and incubate overnight at 4°C. 6. Remove antiserum dilutions, wash each well twice with PBS-Tween 20, and blot dry. 7. Add 100 ul of secondary antibody solution to each well, and incubate at 37°C for 1 h. 8. Remove the secondary antibody solution, wash each well five times with PBS-Tween 20, and blot dry. 9. Add 100 ul of OPD substrate solution to each well, wrap the plate in aluminium foil, and incubate, in the dark, at room temperature, for 15-20 min. 10. Stop the reaction by addition of 50 ul 2 M H2S04 to each well, and read at 492 nm using an ELISA plate reader. 11. Plot the absorbance reading against antiserum dilution to determine specificity. a

Dilute the antiserum simply by a series of 1:1 dilutions starting from 1:10, 1:20, 1:40, 1:80, 1:160, etc.

90

5: G proteins and their identification

4.2 Immunoblotting and immunoprecipitation Protocol 16 describes the transfer of proteins from SDS-PAGE gels onto nitrocellulose, and subsequent incubation with antisera. This method, which essentially follows the strategies reported by Towbin (15), can be used in conjunction with the variety of gels, including low bisacrylamide and SDSurea gels, that have been previously described. Although we prefer to use a fully immersed blotting procedure, semi-dry transfer can also be successfully used. Protocol 16. Electroblotting of proteins onto nitrocellulose Equipment and reagents • Electroblotting apparatus and power pack (e.g. Bio-Rad Trans-Blot Cell) • Whatman 3mm chromatography filter paper • Nitrocellulose • Blotting buffer (15 g Tris, 72 g glycine, 1000 mlmethanol, made up to 5000 ml with H20. Do not adjust pH) . Ponceau S solution (15 g trichloroacetic acid in 500 ml H20; allow to dissolve and add 0.5 g Ponceau S). Keep stock solution at room temperature and re-use . PBS-NP40 (2 ml Nonidet P40 in 1000 ml PBS)

• Blocking buffer (5 g gelatin in 100 ml PBS) . First antiserum solution (appropriate anti-G protein antiserum dilution in PBS-NP40 containing 1% (w/v) gelatin) 'Second antibody solution (1:500 dilution of commercial horseradish peroxidaseconjugated donkey anti-rabbit IgG in PBSNP40 containing 1% (w/v) gelatin) • o-dianisidine solution (10 mg o-dianisidine hydrochloride solution in 1 ml H20) • Sodium azide solution (1 g NaN3 in 100 ml H2O) . 30% (v/v) hydrogen peroxide

Method 1. Transfer proteins, separated by SDS-PAGE, onto nitrocellulose using electroblotting apparatus according to the manufacturer's instructions.3 2. Transfer the electroblotted nitrocellulose sheet into a dish, cover with 100 ml immunoblot blocking buffer, and incubate for 2 h at 30°C. 3. Remove blocker, wash the nitrocellulose using copious amounts of double-distilled water, add the first antiserum solution (normally the specific anti-G-protein antiserum), and incubate overnight at 30°C. 4. Remove primary antiserum, and wash the blots thoroughly with double-distilled water to remove all the unbound antiserum. 5. Wash the blot with PBS-NP40 for 2 X 1 0 min, then incubate at 30°C for 2 h in the second antiserum solution. 6. Thoroughly wash the blot with double-distilled water, then with PBSNP40 for 2 x 1 0 min, and finally with two washes each of 10 min with PBS. 7. Place the blot in the dish containing 40 ml PBS. Add 1 ml newly prepared o-dianisidine solution, then 10 n-l of stock hydrogen peroxide. 91

Ian Muilaney Protocol 16.

Continued

8. Remove the developer, terminate the reaction by addition of sodium azide solution, and leave for 2 mm. Pour off the sodium azide and wash with water. a

To check if transfer is complete, remove the nitrocellulose from the blotting sandwich, place in a clean container, and cover with Ponceau S solution. Gently rock until protein banding appears. Discard the staining solution, and wash the blot with blotting buffer from the electroblotting tank until the bands disappear.

Sustained exposure of cell surface receptors to agonist frequently results in downregulation of both the receptor and the activated G protein in a coordinated manner. We have used the ability of many of these anlisera to immunopreeipitate G protein a subunits to probe the mechanisms involved in the turnover of these polypeptides. Protocol 17 describes pulse-chase assay techniques, in which cells are first incubated with [ 35 S]-methionine labelled medium, then chased in the presence or absence of a receptor agonist. Subsequent immunoprecipitation of these radiolahelled polypeptides with specific antisera has revealed that the mechanism of downregulation for a variety of G protein a subunits is agonist-induced accelerated turnover of the a sutmnit (Protocol 18). As an example, we have demonstrated that activation

Figure 5. The effect of agonist on the rate of degradation of Gqa/G11a in Chinese hamster ovary cells expressing the human M1 muscarinic acetylcholine receptor. Cells were labelled with Trans[ 35 S]-label (Protocol 77) and the rate of turnover of Gqa/G11a was measured by immunoprecipitation (Protocol 181with the specific antiserum, CQ2, following various periods in which the ceils were maintained in the presence (+) or absence (-) of 1mM carbachol. Panel A depicts a fluorogram from a typical experiment. Panel B shows the resultant quantification of the fluorogram by densitometric scanning. Signals from imrnunoprecipitants from cells maintained in the absence (open square) or presence (filled circle) of carbachol. pirn = preirnmune serum, (Data from Mitchell et al., 1993, ref. 27).

5: G proteins and their identification

of the human ml muscarinic receptor by the selective agonist carbachol results in an enhanced degradation of Gqa/G11a, the G proteins which interact with this receptor (Figure 5). Protocol 17. [35S]-trans pulse-chase Equipment and reagents i 35S labelling medium: methionine- and cysteine-free Dulbecco's modified Eagle's medium (ICN Biomedicals, Inc.) supplemented with 50 uCi ml-1 Tran35S-label (ICN) and 1% (v/v) heat-inactivated, dialysed, fetal bovine serum •

• Tissue culture facilities • Benchtop centrifuge, microcentrifuge, and refrigerated ultracentrifuge • 100°C heating block • 2 ml screw-cap Eppendorf tubes . 2% w/v SDS

Method 1. Seed cells into 75 cm3 flasks or 6-well culture dishes. 2. When cells are approximately 60% confluent, replace growth medium with 35S labelling medium, and incubate cells in 35S-label for 20-48 h (pulse). 3. Wash the cell monolayer twice with normal growth medium, and leave in fresh growth medium (chase). 4. At appropriate times, gently wash the cells off the surface of the flask with a Pasteur pipette or rubber policeman, collect in a 12.5 ml plastic centrifuge tube on ice, and centrifuge at 1000 r.p.m. at 4°C for 5 min in a benchtop centrifuge. 93

Ian Mullaney Protocol 17.

Continued

5. Take off medium, resuspend the cells in 100 ul H2O, and add 100 ul 2% (w/v) SDS.b Transfer to a 2 ml screw-cap Eppendorf tube. 6. Tighten the screw cap onto the tube, and heat to 100°C for 20 min.c 7. Transfer the tube to ice, pulse spin to collect all moisture at the bottom of the tube, and proceed to the immunoprecipitation method (see Protocol 18). a

Caution: 35S is volatile. Stocks should be opened and aliquotted in a fume hood. Put aliquots in 2 ml screw-cap Eppendorf tubes and store at -80°C. b To ensure proper solubilization of the cell pellet, always resuspend in H2O, then add the SDS. Do not resuspend the cell pellet directly in SDS. c If the sample is still viscous at this stage, pass it through a 25-gauge needle and syringe, and re-boil as before for 10 min. At this stage it is possible to freeze the sample at -20°C and store until further use.

Protocol 18.

Immunoprecipitation

Equipment and reagents • Benchtop centrifuge, microcentrifuge, and refrigerated ultracentrifuge . 100°C heating block ., . .. . u .Rotating wheel inhibitor, 10 . 1.33% (w/v) SDS . Pansorbin or protein A-sepharose . IP wash buffer (80 ml IP buffer, 20 ml 1% (w/v) SDS).

• IP buffer (1% Triton X-100, 10 mM EDTA, 100 mM NaH2PO4, 10 mM NaF, 100 ixM Na2VO4, 50 mM Hepes, 1 mM PMSF, 3 mM benzamidine, 0.1 u.M soybean trypsin uM leupetin, 0.2 uM aprotinin, 1.5uMantipain, pH 7.2, at 4°C) . IP final wash buffer (50 mM Tris, pH 6.8, at 4°C) . Laemmli sample buffer

Method 1. To each sample in a final volume of 50 ul, add 150 ul of 1.33% (w/v) SDS in 2 ml screw-cap Eppendorf tubes. 2. Heat the samples to 95°C for 5 min, then place on ice to cool and centrifuge the samples to the bottom of the tubes by briefly spinning the samples at maximum speed in a microcentrifuge. 3. Add 0.8 ml of ice cold IP buffer containing protease inhibitors to each sample, mix by inverting, and leave on ice for 1 h. 4. Spin samples in a microcentrifuge at 12000 r.p.m. for 10 min at 4°C, and transfer the supernatant to a new tube. 5. Add an appropriate amount of antibody (between 2 and 20 ul, depending on the antiserum) and incubate with rotation at 4°C overnight. 6. Add 50 ul of Pansorbin (or 20 ul protein A-sepharose) to each sample, and incubate at 4°C with rotation for a minimum of 4 h. 94

5; G proteins and their identification 7. Spin samples in a microcentrifuge at 12000 r.p.m. for 2 min at 4°C, and wash three times with 1 ml IP wash buffer, pelleting the Pansorbin complex between each wash by centrifugation for 30 s at 12000 r.p.m.. 8. After the final spin, remove the supernatant and wash the pellet with 1 ml IP final wash buffer, and repeat the centrifugation step. 9. Remove the final wash buffer, and add 50 ul Laemmli buffer. Heat samples to 100CC for 10 min, centrifuge for 10 min at 12000 r.p.m. to pellet the Pansorbin, then load the supernatant onto the gel.

5. Quantification of G protein a subunits The ability to produce high-level expression of recombinant mammalian G protein a subunits in Escherichia coli has revolutionized our ability to measure accurately the cellular levels of these polypeptides. Protocol 19 describes a method published by Wise and Milligan that uses plasmid pT7.7 into which the various Ga genes have been subcloned (16). This expression system uses the promoter for bacteriophage T7 RNA polymerase in the plasmid vector pT7.7. This expression vector contains an initiation codon ATG, and a ribosome binding site positioned downstream from the T7 promoter such that maximal expression is ensured. Expression constructs were transformed into the lysogen BL21 (DE3), which contains a single chromosomal copy of the gene for T7 RNA polymerase under control of the isopropyl-p-D-thiogalactopyranoside (IPTG)-inducible lac UV5 promoter. It should be noted that preparations of partially purified G protein a subunits are available commercially.

Protocol 19. Preparation of competent E. constrain BL21 DE3, transformation of cells with the expression vector pT7.7, and expression of mammalian G protein a subunits Equipment and reagents • Tissue culture facilities and shaking • Sterile L-broth-glucose (10 g tryptone, 5 g incubator yeast extract, 10 g NaCI, 3.6 g glucose in . Clinical autoclave (e.g. Prestige medical 1000 ml H2O, pH 7.0) series 2100) • Sterile ampicillin stock (50 mg ml-1) • Benchtop centrifuge • Sterile L-broth-agar (10 g tryptone, 5 g . Spectrophotometer yeast extract, 10 g NaCI, 15 g agar in 1000 o . 100 C heating block mlH2O,pH7.0) . SDS-PAGE gel apparatus and power pack ' 100 mM isopropyl-B-D-thiogalactopyrano. E. coli strain BL21 DE3 with G protein cDNA side (IPTG) • Laemmli sample buffer (0.605 g Tris, 30 g jnsert urea, 5 9 SDS, 6 g DTT, 10 mg bromosterile L-broth (10 gtrytone, 5g yeast

phenol blue in 100 ml h2o

extract, 10 g NaCl in 1000 ml H2,lo

95

Ian Mullaney Protocol 19.

Continued

Method 1. Take 50 (il of E. coli stock, add 5 ng of plasmid DNA, and incubate on ice for 15 min. 2. Heat the cells at 42°C for exactly 90 s, then return to ice for 2 min. 3. Add 450 ul of L-broth-glucose, and allow the cells to recover by incubation at 37°C for 1 h in a shaking incubator. 4. Spread 100 ul of the transformants onto L-broth-agar plates containing 100 ul ml-1 ampicillin, and incubate overnight at 37°C. 5. Select single colonies from the plate, and put them into 10 ml L-broth containing 100 ug ml-1 ampicillin, and incubate in a shaking incubator overnight at 37°C. 6. Take 0.5ml of the overnight culture, inoculate into 50 ml L-broth containing 100 ug ml-1 ampicillin, and incubate in a shaking incubator at 37°C until an absorbance at 550 nm of between 0.3 and 0.5 absorbance units is reached. 7. Remove 1 ml of the cell suspension as a control, add 0.5 ml of 100 mM IPTG to the remainder, and incubate in a shaking incubator at 37°C for 4h. 8. Remove 1 ml of the cell suspension, spin in a benchtop microcentrifuge for 10 min at 12000 r.p.m., and discard the supernatant. 9. Add 25 ul of Laemmli sample buffer to the pellet, heat to 90°C for 10 min, and resolve the whole cell extracts by SDS-PAGE.

Quantification is most conveniently achieved by immunoblotting various amounts (0-100ng) of either E. coli-expressed G protein or purified G protein along with known amounts of the plasma membrane fractions. Data from these immunoblots can then be obtained either by densitometric scanning (Protocol 20), or overlay techniques that use commercial radiolabelled second antibody (Protocol21). A standard curve can be constructed, and levels of G protein in the sample can be assessed and expressed in terms of membrane protein, tissue amount, or even cell number. Figure 6 shows such an experiment where the cellular levels of Gsa were assessed in NCB20 cells.

96

5; G proteins and their identification

Figure 6. Quantification of levels of E. coli-expressed Gsa, (A). Various amounts (0-25 ng) of E coli-expressed Gsa (long isoform) were resolved by SDS-PAGE, and immunoblotted with an antipeptide antiserum directed against G s a. Lane 1, 2.5 ng; lane 2, 5 ng; Iane3, 7,5 ng; lane 4: 10 ng; lane 5, 15 ng; lane 6, 20 ng; lane 7, 25 ng. Membranes (20 ug) from untreated (lane 8) and isoprenaline-treated (10 uM, 16 h) (lane 9} NCB-20 cells stably transfected to express the human B2-adrenoceptor were also immunoblotted. (B). The developed immunoblot was subjected to densitometric analysis as described in Protocol 20. and a standard curve was constructed, (Data from Mullaney et al., 1995, ref. 28).

Protocol 20.

Densitometric quantitation of immunoblots

Equipment • Imaging densitometer

Method 1. Place the developed immunoblot onto filter paper and allow to dry.

2, Scan the blot into an imaging densitometer (e.g. Bio-Rad GS-670) and quantify.

Ian Mullaney Protocol 21.

Continued

3. Plot the standard curve showing amount of recombinant or purified G protein against arbitrary densitometric values, and extrapolate the values for the unknowns from the curve.

Protocol 21. [125l]-labelled donkey anti-rabbit immunoglobulin overlay technique Equipment and reagents • Gamma counter . [126l]-labelled donkey anti-rabbit immuneglobulin, (product no. IM 134), specific activity 750-3000 Ci mmol-1 (Amersham International plc., Amersham, UK)

• [125l]-overlay solution: (50ml PBS-NP40 containing 1% (w/v) gelatin and 5 (iCi [125l]-labelled donkey anti-rabbit immunoglobulin)

Method 1. Place the developed blots in a dish containing 50 ml [125l]-overlay solution, and incubate for 1 h at 30°C. 2. Remove the overlay solution, wash the blots thoroughly with doubledistilled water to remove all the unbound label, and then wash twice for 30 min with PBS. 3. Allow the blot to air dry, excise the immunoreactive bands, and measure by liquid scintillation counting. 3. Plot the standard curve showing amount of recombinant or purified G protein against the c.p.m. obtained from the bound [125l]-overlay solution, and extrapolate the values for the unknowns from the curve.

6. Functional aspects of G protein signalling The ability to bind and hydrolyse guanine nucleotides as integral parts of their activation-deactivation cycle is perhaps the most prominent functional characteristic of signal-transducing heterotrimeric G proteins. Measurement of the effects of receptor activation on high-affinity GTPase activity provides the basis of a variety of methods used to probe G protein function. Activation of receptor by agonist initiates the release of guanosine 5'-diphosphate (GDP) from the a subunit of the heterotrimer, followed by the binding of guanosine 5'-triphosphate (GTP). This leads to the dissociation of the aBy heterotrimer into GTP-liganded a subunit and free By dimer, both of which can interact with a variety of effectors to modulate intracellular second messenger concentrations. Deactivation occurs from hydrolysis of GTP by the intrinsic GTPase activity of the a subunit, and subsequent reassociation of the heterotrimer (1,13). 98

5: G proteins and their identification

6.1 Determination of GTP hydrolysis in membrane preparations Using a modified version of the method developed by Cassel and Selinger (17), it is possible to measure the rate of agonist-mediated GTP hydrolysis (Protocol 22). Determination of high affinity GTPase activity is achieved by measuring the agonist-mediated breakdown of substrate y[32P]-GTP to GDP and [32Pi], which is counted as an index of enzymatic activity. Although receptor stimulation of GTPase as a measurement of activity for Gsa has been reported, the degree of stimulation is often modest. On the other hand, robust GTPase activity measurements mediated through pertussis toxin-sensitive G proteins are common, a reflection of their cellular abundance and higher intrinsic rate of GTP hydrolysis.

Protocol 22. Determination of agonist-stimulated high-affinity GTPase activity in plasma membranes Equipment and reagents • Liquid scintillation counter • 2x stock reagent assay mixture (2 mM . Refrigerated benchtop centrifuge App(NH)P, 2 mM ATP, 2 mM ouabain, 20 mM . TE buffer (see Protocol 7) creatine phosphate, 5 units ml-1 .,,_. ___ , ... creatine kinase, 200 mM NaCI, 10 mM Qn ...... . ^FP] GTP (product no-1 PB 10244), specific MgCl2, 4 mM2 mM EDTA, activity >5000 Cimmol (Amersham Phar- 20 mM Tris-HCl (pH 7.5), 1 uM GTP 32 a,b macia Biotech UK Limited., Amersham, UK) containing -y[ P] GTP). • Charcoal slurry (5% (w/v) activated charcoal in 20 mM phosphoric acid, pH 2.3)

Method 1. Take an appropriate amount of plasma membranes (typically between 2 and 10 ug in 20 ul TE buffer), place in a 1.5 ml Eppendorf centrifuge tube, and make up to 20 ul with TE buffer.c 2. Add 50 ul of assay mixture and 10 ul of the appropriate agonist to each tube. Blank values are determined by replacing membranes with 20 ul TE buffer, and non-G protein-mediated low-affinity GTPase activity is assessed by incubating in parallel tubes that include 100 uM GTP. 3. Make up each sample to 100 ul final volume with water, briefly vortex mix, and initiate the assay by transferring the tubes to a 37°C water bath. After 20 min, terminate the assay by removal of the tubes to ice.d 4. Add 900 ul of charcoal slurry to each tube, mix, and centrifuge at 12000 gfor 20 min in a refrigerated benchtop centrifuge 99

Ian Mullaney Protocol 21.

Continued

5. Carefully remove 500 u1 of the clear supernatant from each sample to a scintillation vial, and count by either liquid scintillation or Cerenkov counting. a

The reaction mix is twice the final reagent concentration -y[32P] GTP is added as a trace amount to the cold GTP. Each assay tube should contain approximately 50000 c.p.m. Count an aliquot of the final reaction mix to determine the exact number of counts per assay tube C AII manipulations should be performed on ice. d Further hydrolysis of GTP is negligible on ice. b

To decrease non-specific hydrolysis of GTP, ATP, creatine phosphate, and creatine kinase are included in the assay as an ATP regenerating system to prevent the nucleoside diphosphokinase-mediated transfer of [32Pi] to endogenously present ADP, which can in turn be hydrolysed by specific ATPases. In addition, App(NH)P is added to inhibit nucleoside triphosphatases and ouabain is included as an inhibitor of Na +/ K + ATPase. Exogenous Mg2+, an important co-factor in G protein activation, is also added. Results are expressed as pmol GTP hydrolysed per minute per mg of membrane protein.

6.2 Measurement of receptor-stimulated [35S]GTPyS binding in membrane preparations The binding of radiolabelled GTP analogues which are not hydrolysed by the GTPase activity of the G protein provides a convenient tool with which to probe the initial steps of G protein activation. Measurement of receptormediated increases in 5'-0-(y/-[35S]thio)triphosphate ([35S]GTP-yS) binding is most frequently used (Protocol 23). Protocol 23. Determination of agonist-stimulated [35S] GTP-yS binding in plasma membranes Equipment and reagents • • • .

Liquid scintillation counter Refrigerated benchtop centrifuge TE buffer (see Protocol 1) [35S] GTPyS, (product no. NEG-030H)), specific activity >1250 Ci mmol-1 (NENDupont). To avoid decomposition, the stock should be diluted 100x in 10 mM Tricine, 10 mM dithiothreitol (pH 7.6), aliquotted, and stored at -80°C

• Brandel cell harvester or other binding drum apparatus . Whatman GF/C glass fibre filters • 2X reagent assay mixture (6 mM MgCI2, 200 mM NaCI, 20 uM GDP, 0.4 mM ascorbic acid, 40 mM Hepes, pH 7.4) containing [35S] GTPyS at 50 nCi per assay point a, b • Ice-cold filter wash buffer (3 mM MgCI2, 20 mM Hepes, pH 7.4)

Method 1. Take an appropriate amount of plasma membranes (typically between 10 and 50 ug in 25 ul TE buffer), and place in a 5 ml disposable glass test tube.c 100

5: G proteins and their identification 2. Add 50 ul of assay mixture and 10 ul of the appropriate agonist to each tube. Blank values are determined by replacing membranes with 25 ul TE buffer, and non-specific binding is assessed by incubating parallel tubes that include 10 uM unlabelled GTP-yS. 3. Make up each sample to 100 ul final volume with water, briefly vortex mix, initiate the assay by transferring the tubes to a 4°C ice slurry bath, and incubate for 60 min. 4. Incubation is terminated by addition of 3 ml filter wash buffer to each tube, and immediate passing of the sample through a GF/C filter. Wash the filter with 2 x 5 ml filter wash buffer. 5. Remove the filter into scintillation vials containing 10 ml scintillation cocktail, and count by liquid-scintillation counting. a

The reaction mix is 2x final reagent concentration Each assay tube should contain approximately 100 000 c.p.m of [35S] GTP-yS. Count an aliquot of the final reaction mix to determine the exact number of counts per assay tube c All manipulations should be performed on ice. b

One attraction of this method is that it can be adapted to give optimal responses for receptor stimulation of different G proteins in a variety of cells and tissues. When measuring receptor-mediated [35S] GTP-yS binding to Gi or G0 with incubations either at 4°C or 25-37°C, addition of GDP (0.1-10 uM) and Nad (100-150 mM) is recommended. In contrast, measurement of receptor-mediated [35S]GTP-yS binding to Gs should be performed at 4°C in the absence of both GDP and NaCl, since these reagents act to decrease binding to this G protein (18). Results are expressed as pmol [35S]GTP-yS bound per mg of membrane protein.

Acknowledgements I would like to thank Professor Graeme Milligan for providing laboratory space and helpful discussion in the preparation of this manuscript. I would also like to thank Drs F. R. McKenzie, F. M. Mitchell and A. Wise for their contributions towards the information contained in this chapter.

References 1. 2. 3. 4. 5. 6.

Gilman, A. G. (1987). Annu. Rev. Biochem., 56,615. Gill, D. M., and Meren, R. (1978). Proc. Natl. Acad. Sci. USA, 75,3050. Katada, T., and Ui, M. (1982). J. Biol. Chem., 257,7210. Katada, T., and Ui, M. (1982). Proc. Natl. Acad. Sci. USA, 79,3129. Kurose, H., Katada, T. Amano, T., and Ui, M. (1983). J. Biol. Chem., 258,4870. Strathmann, M., and Simon, M. I. (1990). Proc. Natl. Acad. Sci. USA, 87,9113. 101

Ian Mullaney 1. Wilkie, T. M., Scherle, P. A., Strathmann, M. P., Slepak, V. Z., and Simon, M. I. (1991). Proc. Natl. Acad. Sci. USA, 88,10049. 8. Koski, G., and Klee, W. A. (1981). Proc. Natl. Acad. Sci. USA, 78,4181. 9. Hudson, T. H., and Johnson, G. L. (1980). J. Biol. Chem., 255,7480. 10. Milligan, G., Carr, C., Gould, G. W., Mullaney, L, and Lavan, B. E. (1991). J. Biol. Chem., 266, 6447. 11. Georgoussi, Z., Merkouris, M., Mullaney, L, Megaritis, G., Carr, C., Zioudrou, C., and Milligan, G. (1997). Biochim. Biophys. Acta, 1359,263. 12. Laemmli, U. K. (1970). Nature, 227,680. 13. Milligan, G. (1988). Biochem. J., 255,1. 14. Goldsmith, P., Gierschik, P., Milligan, G., Unson, C. G., Vinitsky, R., Malech, H. L., and Spiegel, A. (1987). J. Biol. Chem., 262,14683. 15. Towbin, H., Staehelin, T., and Gordon, J. (1979). Proc. Natl. Acad. Sci. USA, 76, 4350. 16. Wise, A., and Milligan, G. (1994). Biochem. Soc. Trans., 22,12S. 17. Cassel, D., and Selinger, Z. (1976). Biochim. Biophys. Acta, 452,538. 18. Wieland, T., and Jakobs, K. H. (1994). In Methods in enzymology (ed. lyengar, R.) Vol. 237, p. 3. Academic Press, London 19. Milligan, G., Unson, C. G., and Wakelam, M. J. O. (1989). Biochem. J., 262,643. 20. Falloon, J., Malech, H., Milligan, G., Unson, C., Kahn, R., Goldsmith, P., and Spiegel, A. (1986). FEBS Lett., 209,352. 21. Mullaney, L, and Milligan, G. (1990). /. Neurochem., 55,1890. 22. Mullaney, L, Magee, A. L, Unson, C. G., and Milligan, G. (1988). Biochem. J., 256, 649. 23. Mitchell, F. M., Mullaney, L, Godfrey, P. P., Arkinstall, S. J., Wakelam, M. J. O., and Milligan, G. (1991). FEBS Lett., 287,171. 24. Milligan, G. (1994). In Methods in enzymology (ed. Iyengar, R.). Vol. 237, p. 268. Academic Press, London. 25. Milligan, G., Mullaney, L, and Mitchell, F. M. (1992). FEBS Lett., 297,186. 26. Casey, P. J., Fong, H. K. W., Simon, M. L, and Gilman, A. G. (1990). J. Biol. Chem., 265,2383. 27. Mitchell, F. M., Buckley, N. J., and Milligan, G. (1993). Biochem. J., 293,495. 28. Mullaney, I., Shah, B. H., Wise, A., and Milligan, G. (1995). J. Neurochem., 65, 545.

102

6

Construction and analysis of receptor—G protein fusion proteins ALAN WISE

1. Introduction It is now well established that a multitude of diverse extracellular signals such as neurotransmitters, hormones, odorants, and light elicit intracellular responses via vectorial signal transduction mechanisms across the lipid bilayer, following binding to specific cell surface receptors possessing seven transmembrane-spanning domains. Agonist-occupation of these receptors leads to activation of effector proteins, and the concomitant mobilization or production of second messengers that initiate the desired physiological response within the cell. In all eukaryotic organisms the family of heterotrimeric G proteins plays a pivotal role, acting as intermediates between such cell surface receptors and cytoplasmic or membrane-bound effector molecules, leading to the regulation of humoral, neural, metabolic, and developmental functions. Examples of intracellular effectors which are regulated by G protein activation include adenylate cyclase, phosphoinositidase C, cyclic GMP phosphodiesterases, phospholipase A2, and a number of ion channels (1,2). Structurally, the heterotrimeric G proteins are composed of distinct a, (3, and -y subunits of molecular masses 39-52, 35-36, and 7-8 kDa, respectively. The a subunit binds GTP, possesses intrinsic GTPase activity, and determines the specificity of the holoprotein for its receptor(s) and effector(s), whilst the By complex exists as a tightly associated complex which can itself activate effector molecules. All heterotrimeric G proteins studied so far function in a similar manner: agonist-occupation of G protein-coupled receptors (GPCRs) increases the rate of exchange of GTP for GDP in the nucleotide-binding pocket of the G protein a subunit. The GTP-liganded a subunit can then dissociate both from the receptor and from the G protein By complex, allowing effector regulation. Deactivation of the system is achieved by the intrinsic capacity of the a subunit to serve as a GTPase, resulting once again in the presence of GDP in the binding site (1). Molecular cloning techniques have allowed identification of at least sixteen genes encoding a subunits, and six and twelve encoding B and y subunits,

Alan Wise

Figure 1. Shows homology of the G protein a subunits at the amino acid level. The four major subclasses, Gs, Gi, Gq, and G12 are grouped.

respectively. The a subunits can be divided into four major subclasses, termed Gs, Gi; Gq and G12, based on structural homology at the amino acid sequence level (Figure 1). In addition, the number of cDNAs encoding GPCRs so far identified exceeds 400 and is still growing rapidly (2). Given that many cells express multiple G proteins and GPCRs, the potential signalling circuitry is immense. The fact that multiple receptors can couple to a single G protein, that single receptor subtypes can couple to more than one G protein, that individual ligands can bind to multiple functionally distinct receptor subtypes, and that G protein species can stimulate more than one effector, demonstrates that G protein-mediated signalling pathways form 104

6: Construction and analysis of receptor-G protein fusion proteins highly complicated networks. The advent of heterologous expression systems has proved invaluable in the characterization of signalling components and in elucidating receptor-G protein-effector coupling specificities. The techniques and methodologies developed to study and delineate the activation, function, and receptor-effector coupling specificities of individual G protein a subunits have proved extremely useful in unravelling the complexities of signalling across the plasma membrane (see Chapter 5). This chapter describes a recent approach taken by us and others (3-11) which allows the interaction between a receptor and its cognate G protein to be studied in isolation, following heterologous expression of both entities as constrained fusion proteins. In particular, GPCR coupling to members of the pertussis toxin-sensitive Gi family will be examined.

2. Methods to study G protein function 2.1 Second messenger production Over the last two decades a number of assay systems have been developed which measure changes in levels of second messenger molecules, including cAMP, IP3 and Ca2+, following exposure of whole cells to agonists. Hence they provide compelling evidence for the identification of effector proteins involved in particular G protein-mediated signalling cascades. Such technologies are now widely used to identify and unravel G protein and effectorcoupling specificities of receptors. More recently, reporter gene technologies have been identified and used as alternative ways of studying G protein function, and will be discussed elsewhere in this book (Chapter 8).

2.2 G protein activation One of the earliest points in the signalling cascade at which G protein function can be measured is at the level of G protein activation by receptor. This may be achieved using assays that measure the exchange of GDP for GTP, and/or the subsequent hydrolysis of GTP induced by agonist ligands. Such assays are, however, limited to members of the Gi family (with the exception of Gz), since they possess significantly higher rates of guanine nucleotide exchange (turnover numbers in the region of 4 min-1) than other G protein family members (1).

2.3 Use of pertussis toxin-resistant G protein mutants The a subunits of many individual G proteins are substrates for bacterial toxins elucidated by Vibrio cholerae and Bordetella pertussis. Each toxin mediates the transfer of an ADP-ribose moiety from NAD+ to specific amino acid residues (Figure 2). The site of ADP-ribose incorporation catalysed by cholera toxin has been identified as a single Arg residue, corresponding to 105

Alan Wise

Figure 2. Cholera and pertussis toxins catalyse the transfer of an AOP-ribose group from NAD1 to specific amino acid residues on G5 and G1 family G protein a subunits, respectively. In the case of the Gi G proteins, this modification occurs at a conserved cysteine residue four amino acids from the C-terminus, and leads to receptor-G protein uncoupling. Mutation of this cysteine residue to glycine renders the G protein insensitive to pertussis toxin modification.

position 188 of Gsa, and such modification leads to inhibition of the intrinsic GTPasc activity of the a suhunit and constitutive activation of the signalling pathway (12). Pertussis toxin-mediated ADP-ribosylation occurs at a cysteine residue at the fourth position from the C-terminus on members of the Gi-family (13, 14). This abrogates functional contact between receptor and Gi, thus maintaining the G protein in its inactive aBy heterotrimeric conformation (15). Since other G protein a subunits do not serve as substrates for pertussis toxin-catalysed ADP-ribosylalion, it is possible to determine whether a particular agonist at a GPCR invokes activation of members of the Gi family of G proteins by measuring attenuation of function following pertussis toxin treatment of cells or tissue (16). The pertussis toxin-sensitive G proteins Gl1a, G12a, and G13a are often co-expressed in cells, hence definition of the specificity of interactions of a receptor with individual G proteins from this family cannot be attempted using such a limited approach. To circumvent this problem, we (17) and others (18. 19) have mutationally altered these proteins, such that the cysleine residue which is the target for pertussis toxin-catalysed ADP-ribosylation was exchanged for a glycine residue, thus rendering these proteins refractory to pertussis toxin-catalysed ADP-ribosylation (Figure 2). These pertussis toxin-insensitive C y s > G l y mutants of G11a, G12a, and G13a can then be introduced into cells with a receptor of interest. Specificity of interaction between expressed receptor and Cys—»Gly G1a can be studied in isolation, following exposure of cells to pertussis toxin to eliminate any potential coupling to endogenous G 1 -family G proteins. We have previouslyused such technology to investigate interactions of the porcine a2 A -adrenoceptor with individual members of the G1 G protein family following transient expression in COS-7 cells (17). 106

6: Construction and analysis of rcceptor-G protein fusion proteins

3. Receptor—G protein fusions 3.1 Background The direct receptor-stimulated GTPase activity of G proteins has historically been difficult to measure, due to a lack of information on the absolute levels of expression of the receptor and its cognate G protein, and their localization in relation to one another in cells and at the plasma membrane. Indeed, such measurements have often resulted in estimates too low to account for the rapid kinetics of ligand-indueed activation and deactivation of signal transduction cascades (1,20.21). These problems can he overcome by constraining receptor and G protein within a single fusion protein, thus defining the stoichiomelry of expression of the two entities as 1:1 and ensuring their colocalization following expression (Figure 3). We and others (3-11) have used this approach to study thc mechanisms and specificities governing receptor-G protein interaction. In particular, we have generated fusion proteins between the a 2A -adrenoccptor and G11a, and between the A] adenosine receptor and Cys—>Gly pertussis toxin-insensitive variants of G11a, G12a, and G13a. These constructs have proved useful tools to study the enzymic capacity of G1 G proteins, to measure ligand efficacy, and to evaluate receptor coupling specificities of related G proteins. Strategies employed in the construction, expression, and measurement of function of such fusions will be described herein.

Figure 3. Receptor-G protein fusion protein. Construction involves fusion of the Nteminus of the G protein a subunit to the C-terminus of the receptor. 107

Alan Wise

Figure 4. Construction of receptor-G protein fusion cDNA. In this case the cDNAs encoding the a2 A -adrenoceptor and C y s - G i y pertussis toxin-insensitive variant of G11a were fused together, using a PCR-based approach as described in Protocol 7. This procedure results in production of an in-frame construct, whereby the 3' end of the a2Aadrenoceptor ORF is exactly adjacent to the 5' end of the Cys >Gly G11a ORF. Finally- the entire construct is subcloned into an expression vector such as pCDNA3. The letters E, N, and K denote restriction enzyme sites for EcoRI, A/col and Kpnl, respectively.

108

6: Construction and analysis of receptor-G protein fusion proteins

3.2 Construction of receptor-G protein fusion proteins The following protocol details the construction of a fusion protein made between the a2A-adrenoceptor and a Cys—»Gly pertussis toxin-insensitive variant of G11a (4), which is depicted in Figure 4. The protocol takes advantage of the use of the polymerase chain reaction (PCR), which significantly expedites the manufacture of such constructs. Pfu DNA polymerase (Stratagene) from Pyrococcus furiosus is also employed, which exhibits a 12-fold higher fidelity of DNA synthesis than the more traditionally used Taq DNA polymerase (22). It also utilizes the existence of an Ncol site, which is found straddling the ATG initiation codon of G, family G proteins. Hence a similar strategy can be applied to the manufacture of other fusions that harbour GPCRs linked to other members of the Gj G protein family. Basically, the strategy involves the PCR-mediated addition of an Ncol restriction site at the 3' end of the ORF of the cDNA encoding the a2Aadrenoceptor, such that receptor and G protein cDNAs can be ligated together. Introduction of the Ncol site at the 3' end of the ORF results in the C-terminal amino acid of the receptor being altered from valine to alanine, and to removal of the stop codon. Rat Cys—»Gly G11a cDNA contains two Ncol sites, one straddling the ATG start codon and the other 268 bp downstream from this. Therefore, this 268 bp fragment must be excised from G11a to permit ligation to the 3' end of the receptor ORF. Once this ligation has been achieved, the 268 bp fragment can then be reinserted to yield the full fusion construct. A suitable host vector is pBluescript (Stratagene), since it does not possess an Ncol restriction site. This procedure results in production of an in-frame construct, whereby the 3' end of the at2A-adrenoceptor ORF is exactly adjacent to the 5' end of the Cys—>Gly G11a ORF. The full fusion construct can then be excised and ligated into a suitable eukaryotic expression vector. Protocol 1. Construction of receptor-G protein fusion cDNA The porcine a2A-adrenoceptor (23) was obtained from Dr. L. E. Limbird, Vanderbilt University, Nashville, TN, USA. The Cys-»Gly pertussis toxininsensitive variant of G11a was made using site-directed mutagenesis, which is detailed in ref. 17. Equipment • PCR machine

• 37°C shaking incubator

Method 1. PCR-amplify the ORF of the a2A-adrenoceptor DNA using the oligonucleotides: sense, 5'-TTGGTACCATGTATCCTTACGACGTrC-3'; anti109

Alan Wise Protocol 1. Continued

2.

3.

4.

5.

6.

sense, 5'-AAGAATTCCATGGCGATCCGTTTCCTGTCCCCACGGC-3'. The restriction sites for Kpnl, EcoRI and Ncol are underlined. PCR is performed in a reaction volume of 50 ul containing 20 mM Tris-HCI, pH 8.2; 10 mM KCI; 6 mM (NH4)2S04; 2 mM MgCI2; 1% (v/v) Triton X-100; 10 ug ml-1 BSA; 25 pmol of each oligonucleotide; 250 uM dNTPs; 2.5 ng of supercoiled a2A-adrenoceptor miniprep plasmid DNA in the vector pCDNAS (Invitrogen); and 2.5 units of native Pfu DNA polymerase. DMSO at a final concentration of 5% (v/v) is also included to assist primer annealing. The temperature cycling conditions (30 cycles) are: 94°C for 42 s (denaturation), 60°C for 1 min (annealing), and 72°C for 5 min (polymerization). Restriction digest a 1% (w/v) agarose gel-purified PCR-amplified fragment of approx. 1.5 Kb with Kpnl and EcoRI, and ligate into pBluescript through these restriction sites. Restriction digest rat Cys—»Gly G11a in pBluescript with Ncol, and remove the 268 bp fragment from the 5' end of the ORF by 1% (w/v) agarose gel electrophoresis. Religate the shortened Cys-»Gly G11a cDNA, and excise from pBluescript by digestion with EcoRI, and clone into the EcoRI site of the PCRamplified a2A-adrenoceptor in pBluescript, adjacent to the 3' end of the receptor ORF. Subclone the Ncol excised 268 bp fragment of rat Cys-»Gly G11a between the Ncol sites at the 3' end of the a2A-adrenoceptor ORF and at the 5' end of the Cys->Gly G11a ORF. The full fusion construct is then excised from pBluescript with Kpn\ and EcoRI, and ligated into the eukaryotic expression vector pCDNA3.

4. Expression of receptor-G protein fusions in cultured cells 4.1 Choice of recipient cell line Most of the functional studies that we have performed so far on receptor-G protein fusions have been carried out following transient transfection into two recombinant cell lines stably harbouring the SV40 large T antigen: human embryonic kidney cells (HEK293T), and the simian line COS-7. Both cell lines have been widely used to study receptor-G protein function following transient expression of foreign receptor cDNA; they are particularly appropriate for such studies as they do not express many GPCRs endogenously. They can also be cultured in monolayers for ease of manipulation, and possess rapid doubling times (<20 h), which is highly correlated with the ability to take up DNA (24). 110

6: Construction and analysis ofreceptor-G protein fusion proteins

4.2 Choice of vector We routinely use the eukaryotic expression vector pCDNA3 for all transient transfection studies, since it possesses the four major components necessary to ensure ease of use and high level transfection of eukaryotic cells: bacterial origin of replication (Ori) and an antibiotic resistance marker to allow amplification and selection in a bacterial host; SV40 polyadenylation signal to allow processing of DNA transcripts; constitutive eukaryotic promoter sequence derived from cytomegalovirus (CMV); and a multiple cloning site to allow insertion of foreign cDNA downstream from the eukaryotic promoter.

4.3 Transient expression ofreceptor-G protein fusions A number of methods are now widely used to deliver foreign genes into cultured cells in a highly efficient manner. These include the more traditional methods such as DEAE-Dextran (25), and calcium phosphate precipitation (26). However, we find that more recent approaches involving the use of liposomes (27) (in particular Lipofectamine reagent, Life Technologies) to mediate DNA transfection allow higher and more consistent transfection efficiencies to be achieved. Protocol 2 details the procedure. Cells are maintained in log-phase growth prior to transfection, and can be harvested 48-72 h following exposure to DNA and assayed for receptor-G protein fusion expression and function. Purity of plasmid DNA is critical to ensure efficient transfection. We recommend DNA prepared using Wizard Maxi, Midi, or Mini preparation kits from Promega for this purpose; however, alternative commercially available DNA preparation kits will suffice, as will traditional CsCl gradient preparations. We routinely expose cells to pertussis toxin at 50 ng ml-1 for 16-24 h prior to harvest, to ensure complete ADP-ribosylation of endogenous Gi family G proteins and thus prevent potential interactions between these and the fusion protein. Protocol 2. Transfection by Lipofectamine reagent Equipment and reagents • OptiMEM medium and Lipofectamine reagent (Life Technologies)

• 60-mm cell culture dishes (Costar)

Method 1. On the day prior to transfection (day 1), trypsinize and replate cells in 60-mm culture dishes, aiming to grow to 60-80% confluency in 24 h. 2. On day 2 remove the culture medium, and wash cells with OptiMEM. 3. Prepare the DNA-Lipofectamine mixture, using the protocol described in Section 4.4.

111

Alan Wise Protocol 2.

Continued

4. Add the DNA-Lipofectamine mix to the cells, and return to the incubator for 5 h. 5. Add 2 ml of culture medium supplemented with 20% (v/v) fetal calf serum. 6. Replace the medium 18-24 h following transfection. 7. Harvest the cells 48-72 h following transfection. COS-7 and HEK293T cells are maintained in Dulbecco's modified Eagle's medium containing 10% (v/v) fetal-calf serum and 2 mM L-glutamine.

4.4 Preparation of DNA-Lipofectamine mix A total of 3 ug of DNA and 10 ul of Lipofectamine are used to transfect each 60-mm culture dish. Basically, an appropriate amount of pCDNA3 containing the relevant cDNA species is diluted to 0.1 mg ml-1 in sterile H20, and 30 ul of this diluted mix is added to 170 ul of OptiMEM in a sterile 15 ml tube. Lipofectamine (10 ul) is diluted with 190 ul of OptiMEM, and this mix is then added to the DNA, giving a final volume of 400 ul. Lipofectamine and DNA are gently mixed by flicking the tube, and liposome-DNA complexes are allowed to form for 30 min at room temperature. OptiMEM (1.6 ml) is then added, and the whole is applied to each 60-mm culture dish devoid of medium, as described in Protocol 2.

4.5 Cell harvesting and plasma membrane production Transfected cells transiently expressing receptor-G protein fusions can be harvested simply by scraping from the culture dish in phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KC1, 10 mM Na2HPO4,1.8 mM KH2PO4, pH 7.2). Cells are recovered by centrifugation at 2000 g for 2 min in a microcentrifuge at 4°C. Cells must be thoroughly washed in PBS to ensure complete removal of any residual serum, as this will act as a G protein stimulator and may interfere with the functional assays described later. Harvested cells may be stored as a pellet at —80°C for up to two months, or can be used directly for the production of plasma membranes as described below. Protocol 3. Plasma membrane preparation from cultured cells Equipment and reagents • Microcentrifuge capable of speeds of up to 14 000 r.p.m. which can accept 1.5 ml Eppendorf tubes . TE buffer (10 mM Tris HCI, 0.1 mM EDTA, pH 7.4)

• Either a small homogenizar of the PotterElvehjem type having a glass mortar of approx. 10 ml capacity and a Teflon pestle, or a Polytron-style homogenizer with a cutting head of 15 mm maximal diameter

112

6: Construction and analysis of receptor-G protein fusion proteins Method All procedures must be performed at 4°C. 1. Harvested cells from one 60-mm culture dish are resuspended in 1 ml of TE buffer. 2. Homogenize cells using either a hand-held Potter-Elvehjem type homogenizer (20-30 strokes), or a Polytron-type homogenizer (3 x 10s pulses). 3. Pass the homogenate through a fine-gauge syringe needle (26-G). 4. Centrifuge the homogenate at 500 g for 10 min in a microcentrifuge to pellet unbroken cells and nuclei. 5. Recover plasma membranes from the supernatant by centrifugation at 16000 g for 30 min in a microcentrifuge.a 6. Resuspend the plasma membranes in TE buffer to a final concentration of 2-4 mg ml-1 (approx. 200 ul buffer), and store in aliquots at -80°C.b a

Most protocols detailing isolation of plasma membranes stipulate a centrifugation step in excess of 40000 g to ensure membrane recovery. However, we find equivalent yields are achieved using 16 000 g for 30 min, which has the advantage of obviating the need for an ultracentrifuge, thus expediting isolation of membranes. b We have not found any evidence of proteolytic degradation of transiently expressed proteins from either HEK293T or COS-7 cells, hence we do not routinely include anti-protease cocktails in membrane preparation buffers.

5. Assays used for functional characterization of receptor-G protein fusions 5.1 Background Agonist-liganded receptors increase the rate of exchange of GDP for GTP on the G protein a subunit, and thereby increase the rate of GTP hydrolysis (1,2). Members of the Gi family of G proteins, with the notable exception of Gz, possess the highest rates of guanine nucleotide exchange of all the heterotrimeric G proteins (turnover numbers in the region of 3-4 min-1) (1). This property allows the function and ability of only Gi G proteins in a complex membrane mixture activated by receptor agonists to be assessed, by measuring the exchange of GDP for GTP and/or the subsequent hydrolysis of GTP. We have used two assays which measure (1) agonist-stimulated binding of the poorly hydrolysable analogue of GTP, GTPyS, and (2) agonist-stimulated GTPase activity, in order to study the mechanisms of receptor-Gi G protein coupling when constrained within a fusion protein. Measurement of G protein GTPase activity is essentially achieved by following the breakdown of y[32P]GTP with the concomitant production of [32Pi]. The [32Pi] is then 113

Alan Wise separated from the -y[32P]GTP, allowing the rate of GTP hydrolysis to be measured (28). General applications and uses of both assays to delineate receptor-G protein coupling mechanisms have been described in-depth previously (29). Therefore, only the basic protocols for performing such assays are detailed below. Radioligand binding assays can also be performed to determine accurately expression levels of both receptor and G protein, as both signalling molecules are constrained within the same polypeptide and are therefore present in a 1:1 molar ratio. Protocol 6 describes how a selective radiolabelled a2A-adrenoceptor antagonist can be used to determine levels of expression of the a2A-adrenoceptor-Cys—>Gly G11a fusion protein. Section 6 documents how these assay technologies have been used and adapted to provide insights into the mechanisms and specificities of receptor-G protein coupling, and how a fusion approach can be employed as a means of measuring agonist efficacy.

5.2 Receptor-promoted binding of guanosine-5'-[y-35thio]triphosphate ([35S]GTPyS) Binding of [35S]GTP-yS to membranes can be performed using two techniques, which differ only in the means by which bound nucleotide is separated from free: (1) using wheat germ agglutinin scintillation proximity assay (SPA) bead technology, and (2) using traditional separation of bound nucleotide from free by filtration. The assay using SPA technology is amenable to a 96well format, is more convenient than the filtration assay, and is detailed below in Protocol 4. The footnote to Protocol 4 briefly outlines the alternative filtration assay. Protocol 4. High affinity [35S]GTP-yS binding - SPA format Equipment and reagents • Assay buffer (20 mM Hepes, 100 mM NaCI, 10 mM MgCI2, pH 7.4) . [35S]GTP-yS (1170 Ci mmor') (Nycomed Amersham) . Wheatgerm agglutinin SPA beads (Nycomed Amersham) . 96-well clear-bottomed polystyrene plates (Wallac)

• Scintillation counter capable of reading in 96-well format (e.g. 1450 Microbeta Trilux, Wallac) • If using the filtration assay, additional equipment includes: vacuum filtration apparatus capable of filtering greater than 12 samples individually; filters of 2.5 cm diameter.

Method All membrane manipulations should be carried out at 4°C. Assays are performed in a 96-well format, using a method modified from that described in ref. 30. 1. Dilute membrane proteins (Protocol 3)(5 ug per point) to 0.083 mg ml-1 in assay buffer supplemented with saponin (10 mg l-1). 2. Preincubate diluted membranes with 40 uM GDP for 2-5 min.

114

6: Construction and analysis of receptor-G protein fusion proteins 3. Apply 60 ul of diluted membrane mix (5 ug) to each well of a clearbottomed 96-well plate. 4. Add 20 ul of agonist or vehicle. 5. Determine non-specific values by the addition of 20 ul of 3 mM GTP (0.6 mM final concentration). 6. Start the assay following the addition of 20 ul of [35S]GTPyS (1170 Ci mmol-1) at 0.3 nM final concentration (100 ul final assay volume), and allow to proceed at room temperature for 30 min. 7. Add wheatgerm agglutinin SPA beads (Nycomed Amersham) (0.5 mg ml-1) in 25 ul assay buffer, and incubate the whole at room temperature for 30 min with agitation.a 8. Centrifuge the plates at 1500 g for 5 min, and determine bound [35S]GTP-yS by scintillation counting. a

The binding assay using filtration to separate bound nucleotide from free is essentially the same from steps 1-5. However, the reaction is terminated by rapid filtration through Whatman GF/C glass-fibre filters under vacuum. Filters are then washed three times with 5 ml of assay buffer. Filters are placed in a scintillation vial with 5-10 ml scintillation fluid, and radioactivity is quantified by scintillation counting.

5.3 Measurement of GTPase activity Protocol 5. Agonist stimulation of GTPase activity Equipment • 37°C incubator or water bath

• Scintillation counter

Method All manipulations are performed at 4°C. 1. Membrane proteins (Protocol 3) are diluted in 10 mM Tris HCI, 0.1 mM EDTA, pH 7.4 (TE buffer) to a concentration of 0.25 mg ml-1 (5 ug per assay). 2. An assay mixture is made with the following reagents: Reagent

Concentration in mixture

Final concentration in assay

App(NH)p ATP Ouabain Creatine phosphate Creatine phosphokinase Sodium chloride

2 mM 2 mM 2 mM 20 mM 5 units ml-1 200 mM

1 mM 1 mM 1 mM 10 mM 2.5 units ml-1 100 mM

115

Alan Wise Protocol 5. Continued

3.

4. 5. 6. 7.

8.

9. 10.

Magnesium chloride 10 mM 5 mM Dithiothreitol 4 mM 2mM EDTA 0.2 mM 0.1 mM Tris-HCI 20 mM 10mM y[32P]GTP 1uM 0.5 uMa Final pH of the assay mix is 7.5. An aliquot of this is retained for radioactive counting, so that the exact number of counts in each tube is known. Add aliquots (50 ul) of the above reaction mixture to Eppendorf tubes, together with 20 ul of diluted membrane protein and 10 ul of appropriate agonist. Final assay volume is 100 ul, which may be made up with water. Assess low affinity hydrolysis of y[32P]GTP by incubating parallel tubes in the presence of 100 uM GTP. Initiate the reaction by transferring the tubes to 37°C. Terminate the assay after 20 min by removal of the tubes to ice (hydrolysis of y[32P]GTP on ice is negligible). To separate free [32Pi] from the unhydrolysed y[32P]GTP, add 900 ul of a 5% (w/v) activated charcoal slurry in 20 mM phosphoric acid (pH 2.3) to each tube, giving a total volume of 1000 ul. Centrifuge tubes at 12000 g for 20 min at 4°C to pellet the charcoal along with unhydrolysed -y[32P]GTP. The free [32Pi] is present in the supernatant. Remove 500 ul aliquots of supernatant, and add to scintillation vials for radioactive counting. Radioactivity may be assessed by either liquid scintillation or Cerenkov counting.

a

Since the radioactive GTP will be present in trace amounts, it is necessary to add cold GTP up to the required concentration.

5.3.1 Calculation of GTPase assay results Each assay tube contained 50 pmol of GTP, in addition to approx. 50000 c.p.m. of y[32P]GTP. The amount of y[32P]GTP should be calculated each time by counting 50 u1 aliquots of the assay mixture. Hence the specific activity of the GTP is approx. 1000 c.p.m. per pmol. The rate of hydrolysis is calculated by: (C/S.A.) X 2 X (1000/P) X (1/T) where: C = counts in 500 ul sample S.A. = specific activity of GTP P = amount of protein in ug T = duration of assay 116

6: Construction and analysis of receptor-G protein fusion proteins This will give the rate of hydrolysis of GTP in pmol per min per mg of membrane protein.

5.4 Receptor binding studies Levels of expression of the a2A-adrenoceptor-Cys—»Gly G11a fusion protein can be assessed using the a2-adrenoceptor-selective antagonist [3H]RS-79948197 (Nycomed Amersham). The following protocol should be easily adaptable to permit measurement of expression levels of other GPCR-G protein tandems for which a radiolabelled antagonist ligand is commercially available.

Protocol 6. Measurement of a2A-adrenoceptor-Cys—>Gly G11a fusion protein levels Equipment • 30°C incubator or water bath • Scintillation counter

• Brandel cell harvester

Method All manipulations are to be performed at 4°C. 1. Dilute plasma membrane proteins (Protocol 3) to 0.04 mg ml-1 in 10 mM Tris HCI, 50 mM sucrose, 20 mM MgCI2, pH 7.4 (assay buffer). 2. Add 50 ul (2 u,g) of diluted membrane mix to tubes on ice. 3. Determine non-specific binding by adding 50 ul of 500 uM idazoxan (final 100 uM) to parallel tubes. 4. Start the reaction by adding 50 ul of [3H]RS-79948-197 (final concentration 1 nM) to diluted membrane mix. The assay volume of 250 ul is made up with assay buffer. 5. Incubate for 45 min-1 h at 30 °C, and then filter under vacuum through Whatman GF/C filters to separate bound from free. 6. Wash filters with 3 x 4 ml of cold assay buffer. 7. Determine radioactivity in each filter by liquid scintillation counting. Specific binding is the difference between total and non-specific binding.

6. Receptor-G protein fusions as research tools 6.1 Measurement of agonist-induced guanine nucleotide turnover by G11a Quantification of the agonist-induced stimulation of GTPase activity of G proteins, whether performed in complex membrane preparations or in recon117

Alan Wise

118

6: Construction and analysis of receptor-G protein fusion proteins Figure5. Expression of an a2A-adrenoceptor-Cys—»Gly G11a fusion protein. Membrane fractions of pertussis toxin-treated COS-7 cells transfected with the a2A-adrenoceptorCys—»Gly G11a were subjected to saturation binding studies using [3H]RS-79948-197 (upper panel) as described in Protocol 6. Specific binding is shown. The data were then converted to a Scatchard plot (lower panel). From ref. 4, with permission.

stituted systems, has been problematic, since absolute levels of receptor and/or G protein are difficult to measure and it is unclear whether agonistoccupied receptor has the capacity to activate the full complement of its cognate G protein, due to differences in cellular localization. Indeed, measurements of the GTPase activity of isolated G proteins have routinely resulted in estimates which are too low to account for the rapid kinetics of ligand-induced activation and deactivation of signal transduction cascades (20,21, and reviewed in ref. 1). Use of the a2A-adrenoceptor-Cys—»Gly G11a fusion protein as a tool to measure agonist-induced catalytic-centre activity of G11a constrained therein (4) is described as follows. Basically, transient transfection of the a2A-adrenoceptor-Cys-»Gly G11a fusion protein cDNA in COS-7 cells resulted in high levels of expression of this construct (6-15 pmol mg-1 of membrane protein in separate transfections) when measured by saturation analysis of the specific binding of the a2Aadrenoceptor selective antagonist [3H]RS-79948-197 (Figure 5) as described in Protocol 6. The Kd for this interaction was 0.35 nM, similar to that obtained following individual coexpression of the a2A-adrenoceptor and Cys—»Gly protein was then measured by performing a GTPase assay, as outlined in Protocol 5, with various concentrations of cold GTP in the absence and presence of the a2A-adrenoceptor agonist UK14304 (10 uM). By varying the concentration of substrate (i.e. GTP), it is possible to measure Vmax and Km values for G11a in the absence and presence of agonist, using MichaelisMenten kinetics. The results of such an experiment are shown in Figure 6, with data being presented as an Eadie-Hofstee transformation with the y axis detailing Vmax measurements. In the experiment shown, the increase in Vmax produced by UK14304 was 18.8 pmol per min per mg of membrane protein. Since receptor and G protein within the fusion are by definition present in a 1:1 molar ratio, and the fusion protein was expressed at 6.2 pmol mg-1 in this experiment, then agonist-induced turnover number was calculated to be 3.0 min-1. The measured Km for GTP of the fusion protein was 0.37 uM, which is in agreement with previous estimates of 0.1-0.5 uM (1). In all experiments, cells were exposed to pertussis toxin (50 ng ml-1) for 24 h to eliminate any potential interactions between the receptor element of the fusion protein and endogenously expressed Gi family G proteins. To conclude, Figure 6 shows the utility of using such a receptor-G protein fusion construct to measure G protein enzymic capacity. Such an approach 119

Alan Wise

Figure 6. Agonist stimulation of the high affinity GTPase activity of the a2A-adrenoceptorCys-»Gly G11a fusion protein. Membrane fractions from pertussis toxin-treated COS-7 cells transfected to express the c^A-adrenoceptor-Cys—>Gly G11a fusion protein were used to measure high affinity GTPase activity (Protocol 5). Upper panel: high affinity GTPase activity was measured over a range of concentrations in the absence (filled circles) or presence (open circles) of UK14304 (10 uM). Lower panel: the data are presented as an Eadie-Hofstee transformation. From ref. 4, with permission.

120

6: Construction and analysis of receptor-G protein fusion proteins may offer novel means of studying the ability of GTPase-activating proteins such as the RGS family members (31, 32) to modify the kinetics of G protein signalling mechanisms.

6.2 Measurement of agonist efficacy Agonist efficacy is a measure of the capacity of a ligand to induce a functional response (33, 34). Adequate measurement of agonist efficacy at GPCRs has proven troublesome, because estimates of efficacy can vary with tissue, with expression levels of receptor, and also with the point within a signalling cascade at which the response is measured (35, 36). One of the earliest, and therefore most direct points at which efficacy can be measured is at the level of G protein activated by a GPCR. Hence receptor-G protein fusions, particularly those harbouring Gs family G proteins, offer an alternative means by which efficacy can be measured. The use is described of the a2A-adrenoceptor-Cys—»Gly G11a fusion protein to evaluate efficacy of a number of agonists at this GPCR, by measuring the ligand-induced rate of GTP turnover by the physically associated G11a, following expression in COS-7 cells (4). GTPase measurements were performed on membranes from pertussis toxin-treated cells expressing the a2Aadrenoceptor-Cys—>Gly G11a fusion protein, essentially as described in Protocol 5, in the presence of varying concentrations of a range of a2Aadrenoceptor agonists, as shown in Figure 7. Of those examined, only adrenaline and noradrenaline were found to function as full agonists, with UK14304, xylazine, clonidine, and BHT933 acting as partial agonists. For comparison, individual cDNAs encoding the a2A-adrenoceptor and Cys—»Gly G11a were coexpressed in COS-7 cells, and the efficacy of the same agonists assessed following pertussis toxin treatment of the cells. Table 1 shows that all

Table 1. Capacity of known agonists at a2-adrenoceptors to stimulate high affinity GTPase activity of a2A-adrenoceptor-Cys—»Gly G11a fusion protein or of the a2A-adrenoceptor plus Cys->Gly G11a was assessed following transient expression in COS-7 cells Efficacy to activatea

Ligand

a2A-adrenoceptor-Cys-»Gly G11a a2A-adrenoceptor + Cys-»Gly G11a (% of adrenaline) (% of adrenaline) Adrenaline Noradrenaline a-methylnoradrenaline UK14304 BHT933 Xylazine Clonidine

100 97 105 56 20 16 15

100 99 93 85 44 28 26

'Efficacy was determined relative to the function of adrenaline (1 x 10-4 M) in parallel assays.

121

Alan Wise

Figure 7. Efficacy measurements of ligands at the a2A-adrenoceptor-Cys—>Gly G11a fusion protein. The capacity of varying concentrations of adrenaline (filled triangles), noradrenaline (open squares), UK14304 (filled squares), xylazine (open triangles), BHT933 (open circles), and clonidine (filled circles) to stimulate the high affinity GTPase activity of the a2A-adrenoceptor-Cys-»Gly G11a fusion protein was assessed as described in Protocol 5. Data are presented as the % of the stimulation produced by 1 x 10-4 M adrenaline. From ref. 5, with permission.

of the ligands found to be partial agonists at the a2A-adrenoceptor-Cys—»Gly G11a fusion protein were also partial agonists for separated receptor and G protein in comparison to adrenaline: the rank order of efficacy was also similar in the two systems. Similar enzymic capacity measurements were made, as described in Section 6.1, to elucidate whether the differences in efficacy of the agonists were not just a reflection of distinct agonist-induced variations in the Km of the fusion construct for GTP. Figure 8 shows that this indeed was not the case. The data shown provide compelling evidence that a receptor-G protein fusion can provide a novel means to measure agonist efficacy at a GPCR. Two sources of variation which have until now proved problematic to the measurement of efficacy - the accurate assessment of receptor/G protein levels and localization within the cell - have been circumvented by fixing the receptor and G protein stoichiometry to 1:1, and by defining their co-localization by physical linkage. 122

6: Construction and analysis of receptor-G protein fusion proteins

FigureS. Partial agonism at the a2A-adrenoceptor-Cys—>Gly G11a fusion protein is manifest in varying values for the Vmax of the GTPase activity without alteration in Km for GTP. High-affinity GTPase activity was measured at varying concentrations of GTP (upper panel) in membranes of pertussis toxin-treated COS-7 cells transfected to express the a2A-adrenoceptor-Cys^>Gly G11a fusion protein in the absence of ligand (filled circles), or in the presence of 1 x 10-4 M adrenaline (filled triangles), a-methylnoradrenaline (open circles), UK14304 (filled squares), or xylazine (open triangles), as detailed in Protocol 5. Data were then transformed as an Eadie-Hofstee plot (lower panel) to allow direct estimation of V max , Km for GTP, and the efficacy of the ligands compared to adrenaline. From ref. 5, with permission.

123

Alan Wise

6.3 Elucidating the role of N-terminal acylation of G11a All known G protein a subunits are known to be modified at their N-terminus by co-translational myristoylation and/or post-translational palmitoylation (37-39). These acylations are considered to play key roles in targeting G protein a subunits to the plasma membrane, and may also contribute to proteinprotein interactions between the G protein a subunit and both receptors and the 37 complex (40, 41). Addition of the C14 fatty acid myristate is restricted to a subunits of the Gi family, because they contain the consensus sequence (MGXXXS) for the enzyme W-myristoyl-CoA-transferase. Glycine at position 2 serves as the acceptor residue for the covalent attachment of this fatty acid group, following removal of the initiating methionine. Addition of the C16 fatty acid palmitate occurs on either one or two cysteine residues within the first ten amino acids of the a subunits of all the widely expressed G proteins (37-39). Prevention of myristoylation by site-directed mutagenesis of the acceptor glycine reduces the affinity of interaction of Gi G proteins and the By complex, and renders them cytosolic (37,42). Prevention of palmitoylation by mutation of the N-terminal cysteine acceptor residues also limits membrane association, and has also been reported to abrogate receptor interaction (40, 41, 43, 44). Figure 9 shows the lack of interaction of the various acylationdeficient forms of Cys—»Gly G11a with the a2A-adrenoceptor, following coexpression in COS-7 cells and exposure to pertussis toxin. In this case a GTPase assay, as described in Protocol 5, is employed to measure receptor-G protein coupling (45). However, it has been unclear whether poor receptor regulation of acylation-defective G protein a subunits is simply due to lack of appropriate targeting and thus proximity to a receptor, or is inherently due to the acylation status of the G protein. To address this question use was again made of the a2A-adrenoceptorCys-»Gly G11a fusion protein (4). However, fusion chimaeras were also constructed between the a2A-adrenoceptor and mutant forms of Cys-^Gly G11a in which the sites that are normally palmitoylated and myristoylated, Cys3 and Gly2, were mutated to Ser and Ala, respectively. In addition, the a2Aadrenoceptor is also known to be a target for post-translational palmitoylation, at Cys442 within the C-terminal tail (46). Because this acylation is proposed to create a 'fourth intracellular loop' in the receptor structure, and because this receptor has a relatively short C-terminal tail, fusion proteins were made between a C442A mutant of the a2A-adrenoceptor and the various forms of G11a detailed above. All fusion constructs were made using Protocol 1 and are depicted in Figure 10. Expression of each of these a2A-adrenoceptor-Cys—»Gly G11a fusion proteins in COS-7 cells resulted in similar, high levels of membrane expression of the receptor binding site (15-25 pmol mg-1) as measured by specific binding of the selective and high affinity a2-antagonist [3H]RS-79948-197 as described in Protocol 6. To assess coupling between the a2A-adrenoceptor and its fused 124

6: Construction and analysis of receptor-G protein fusion proteins

Figures. Lack of receptor interactions of acylation-deficient mutants of GJ,(I. High affinity GTPase measurements (Protocol 5) were performed in the presence (hatched barsl or absence (open bars) of UK14304 (10 uM) on membranes from pertussis toxin-treated COS-7 cells transfected with vector alone (1) or with the a2A-adrenoceptor in combination with C y s - G l y G11a (2), C3SCys >Gly G11a. (3), G2ACys-*Gly G11a W or G2A/C3SCys-^GIy G11a (5). From ref. 45, with permission.

Figure 10. Construction of a2A-adrenoceptor/acylation-deficient Cys^Gly G11a fusion proteins. Fusion of the N-terminus of Cys-Gly G11a to the C-terminus of the a 2A adrenoceptor resulted in the receptor C-terminal amino acid (valine) being converted to alanine, as described in Protocol 1. Arrows represent the amino acids altered from the wild-type sequence in the various fusion proteins generated. Marked cysteine residues are known targets for post-translational palmitoylation in both receptor and G protein. Glycine 2 in G11a undergoes cotranslational myristoylation.

125

Alan Wise acylation-deficient pertussis toxin-resistant G11a partners, GTPase assays, as described in Protocol 5, were performed on membranes of transfected COS-7 cells which had been exposed to pertussis toxin (50 ng ml-1) for 24 h prior to harvest. Figure 11 shows the results of such an assay. All of the expressed chimaeric fusion proteins were able to stimulate pertussis toxin-insensitive high affinity GTPase activity in membranes of transfected COS-7 cells upon addition of the a2A-adrenoceptor agonist UK14304. The fusion protein approach guarantees the physical proximity of receptor and cognate G protein. Hence such data demonstrate that the lack of functional activation of acylation-negative mutants of G proteins by co-expressed receptors is related to deficiencies in cellular targeting and location, rather than an inherent inability to produce appropriate protein-protein interactions for signal transmission. Figure 11 also shows that all of the fusion chimaeras required similar concentrations of UK14304 to cause half-maximal effects. Similar data were also generated using fusion proteins containing the C442A mutant version of the a2A-adrenoceptor. These findings also support the argument that acylation

Figure 11. UK14304 stimulates high affinity GTPase activity of a2A-adrenoceptor/acylationdeficient Cys-»Gly G11a fusion proteins. Effects of varying concentrations of UK14304 are shown for a2A-adrenoceptor-Cys^>Gly G11a (open circles), a2A-adrenoceptor-C3SCys^> Gly Gna (open diamonds), a2A-adrenoceptor-G2ACys—»Gly G11a (closed diamonds), or a2A-adrenoceptor-G2AC3SCys^»Gly G11a (filled squares) fusion proteins. GTPase measurements were performed as described in Protocol 5. From ref. 6, with permission.

126

6: Construction and analysis of receptor-G protein fusion proteins is not required to produce effective protein-protein contacts between receptor and a subunit, but to position the G protein appropriately. In addition, co-expression of each of the fusion proteins with the 37 complex, pl-yZ, resulted in greater maximal UK14304 stimulation of GTPase activity, as shown in Figure 12. This was not simply due to enhanced levels of

Figure 12. Expression of B1y2 increases UK14304 stimulation of fusion protein GTPase activity. a2A-adrenoceptor-Cys—»Gly G11a (1 and 2), a2A-adrenoceptor-C3SCys—»Gly G11a (3 and 4), a2A-adrenoceptor-G2ACys-»Gly G11a (5 and 6), or a2A-adrenoceptorG2AC3SCys^»Gly G11a (7 and 8) fusion proteins were expressed with (2, 4, 6, and 8) or without (1, 3, 5, and 7) B1y2. GTPase activity (open bars, basal; filled bars, UK14304, 10 uM) (A) or fusion protein levels (B) were measured on membranes from transfected pertussis toxin-treated COS-7 cells, as described in Protocols 5 and 6, respectively. From ref. 6, with permission.

127

Alan Wise fusion protein expression in the presence of B1y2, as [3H]antagonist binding studies showed no such alterations in expression levels (Figure 12). These results are interesting given that the N-terminus of the G protein a subunit is known to play a central role in By interaction (47, 48), that the By complex may play a key role in receptor interactions with the a subunit (49), and that a subunit acylation is important in defining the strength of interaction with the By complex (37, 42). Such data demonstrate interaction between receptor-G protein fusions and the By complex, and that formation of a receptor-aBy complex is required to permit the most efficient signal transduction between receptor and G protein. Overall, the data presented further document the utility of receptor-G protein fusions as tools to study receptor-G protein interactions, and provide compelling evidence for the role of G protein acylation in directing G protein cellular targeting, but not in the transmission of information between receptors and G proteins.

6.4 Study of interactions between the A1 adenosine receptor and multiple Gi-family G proteins Agonist stimulation of the A1 adenosine receptor leads to modulation of numerous intracellular signalling events, such as inhibition of adenylate cyclase, stimulation of phosphoinositidase C, activation of inwardly rectifying K+-channels, and inhibition of neuronal calcium channels as a consequence of coupling to multiple pertussis toxin-sensitive Gia-famity G proteins (50-52). To study further the interactions between the Al adenosine receptor and individual members of the Gi- family of G proteins, fusion proteins were generated between the A1 adenosine receptor and pertussis toxin-resistant Cys—*Gly variants G11a, G12a and G13a, using essentially the same strategy as described in Protocol 1 for the construction of the a2A-adrenoceptorCys—»Gly G11a fusion protein (4). Minimal disruption was caused to each polypeptide following manufacture of the fusions, with only the C-terminal amino acid of the receptor being altered from aspartic acid to alanine, and the initiator methionine of the G protein (which would normally be removed) remaining in the new protein. All studies were conducted in HEK293T cells, following transient expression of each of the A1 adenosine receptor-Gja fusion proteins as described in Protocol 2, and exposure of the cells to pertussis toxin (50 ng ml-1) for 24 h prior to harvest. Parallel experiments were also performed in which the A1 adenosine receptor was co-expressed along with each of the Cys—»Gly variants G11a, G12a and G13a, to compare agonist-mediated receptor-G protein interactions when expressed as constrained proteins within a fusion and as separate signalling polypeptides. Expression levels of introduced fusion proteins were measured by [3H]antagonist binding studies, essentially as described in Protocol 6, but with the use of the A1 adenosine receptor selective antagonist 128

6: Construction and analysis of receptor-G protein fusion proteins

Figure 13. Agonist-mediated stimulation of [35S]GTP-yS binding to A, adenosine receptor-Cys->GlyGiix fusion proteins and to coexpressed A, adenosine receptor and Cys—»GlyGia G proteins. HEK293T cells were transfected with A1 adenosine receptor alone, together with Cys—»Gly variants of G11a, G12a, and G13a, or as fusion proteins with each of these Gi-family G proteins. Cells were treated with pertussis toxin prior to harvest, and [35S]GTP-yS binding was measured on membrane fractions without (open bars) or with (hatched bars) exposure to NECA (10 uM), as described in Protocol 4. * denotes Cys-»Gly variant of Gia. From ref. 11, with permission.

[3H]DPCPX (Dupont-New England Nuclear, 120 Ci mmor1) as radiolabel, and inclusion of NECA (10 uM) rather than idazoxan to assess non-specific binding. Receptor-G protein coupling was measured by [35S]GTPyS binding in the absence and presence of the adenosine receptor agonist NECA, as described in Protocol 4. Figure 13 shows that introduction of each fusion protein into HEK293T cells led to robust agonist-mediated stimulation of [35S]GTP-/S binding activity, which was of significantly greater magnitude than observed when receptor and G protein were co-expressed. This is probably due to more efficient coupling between receptor and G protein when constrained within a fusion protein, rather than being over-expressed in a cell as separate entities. Certainly, such enhanced G protein activation cannot be attributed to altered levels of receptor expression, since saturation [3H] antagonist binding studies 129

Alan Wise Table 2. Expression levels of transiently transfected A, adenosine receptor-Cys-»Gly G,a fusion proteins Transfections A,AR + A1AR + A1R + A1RGiI a A1ARG12a a

Gi1a Gi2a Gi3a

Antagonist binding (pmol 5.8 ±0.3 6.3 ±0.7 6.1 ±0.3 9.6 ±3.7 7.9 ±2.8 7.9 ±1.2

mg-1)

Dissociation constant, Kd, of antagonist binding (nM) 0.68 ±0.07 0.78 ±0.18 0.64 ±0.10 1.40 ±0.55 0.90 ±0.53 0.96 ±0.26

denotes pertussis toxin-resistant Cys-»Gly variant of da.

revealed similar quantities of A1 adenosine receptor in membranes from cells transfected with cDNAs encoding receptor-G protein fusions, and also in cells transfected with receptor, together with the individual Gi-family G proteins (Table 2). The affinity of interaction between the A1 adenosine receptor and the Cys—»Gly forms of G11a, G12a, and G13a, when the receptor-G protein interaction is constrained within a fusion protein, were then measured by [35S]GTP-yS binding following exposure of membranes to increasing concentrations of NECA. Figure 14 shows that similar EC50 values were attained for all three fusion proteins. Recently, evidence has arisen implying that particular agonists can interact with GPCRs in a defined manner to harness activation of specific G proteins (reviewed in refs 53, 54). The above fusion proteins were employed to study this phenomenon of 'agonist trafficking', since the A1 adenosine receptor causes activation of a variety of intracellular signalling pathways via interaction with Gi-family G proteins. Hence a range of A1 adenosine receptor agonists were characterized for their ability to promote activation of a particular Gi family G protein following expression of the individual fusions in HEK293T cells. No preferential activation of any expressed Gi G protein via the A1 adenosine receptor was observed with any of the compounds tested (representative data for ten compounds out of 40 studied are shown in Table 3). Potency of the same series of ligands was also measured following coexpression of the A1 adenosine receptor with each Cys—>Gly Gia (Table 4). Hence, no agonist-induced 'channelling' of the human A1 adenosine receptor to cause activation of distinct members of the Gi-family of G proteins was observed. To investigate 'agonist trafficking' more rigorously, it may be more pertinent to select GPCRs which are known to couple to less closely related G proteins such as the a2A-adrenoceptor, which has been shown to interact with both Gi and Gs subtypes (55). 130

6: Construction and analysis of receptor-C protein fusion proteins

Figure 14. Dose-dependent agonist-mediated stimulation of [35S]GTP-yS binding to A, adenosine receptor-Cys->GlyG,a fusion proteins and to coexpressed A, adenosine receptor and Cys->GlyGja G proteins. The ability of varying concentrations of NECA to stimulate the binding of [35S]GTP-yS was measured as described in Protocol 4, in membranes of pertussis toxin-treated HEK293T cells transfected to express (A) A1 adenosine receptor-Cys—>GlyGiOi fusion proteins containing G11a (filled triangles), G12a (open circles), and G13a (filled squares), and (B) coexpressed A, adenosine receptor and Cys-»Gly variants of G11a (filled triangles), G12a (open circles) and G13a (filled squares). Modified from ref. 11, with permission.

131

Alan Wise TableS. Efficacy measurements of ligands at the A, adenosine receptor when coexpressed with Cys—>Gly variants of Gia or as fusion proteins with these G proteins Ligand

Potency (EC50 in nM) A,ARGi1 •

NECA 56.4 ±17 CPA 5.3 ±1.2 GR79236X 19.0 ±4.5 CCI4019 46.6 ±3.6 GR56071 7.7 + 0.9 GR56072 3.5 ±1.1 33.2 ± 5.4 GR66683 S-PIA 300 ± 39 R-PIA 8.5+1.2 CHA 11.9 + 3.1 a

42.0 ± 6

49.2 ± 9

14.5

7.8 + 2.8 21.0 + 6.2 66 ± 7.5

±5.5

22.0 ± 6.3 63.7

±6.1

11.8±2.7 3.5 ±1.6 52.1 + 13.4 357 + 80 9.3 ±2.3

46.0

23.4

10.3+1.8

±3.1

10.8

128±

11.7

31.6 + 1.6 96.6 ± 5.6

317 + 38 46 + 2.8

±1.3

4.5 ±2.5

A,ARG,2a 145.8 ±10.1 41. 3 ±3.5 129 ±10.0 340+18 56+1.6

20.7 ±2.8

25.2

320 ±41

214±18.6 1850 ±300

10.9

53.6

247 + 13.8 1330 + 167 64.3 ± 3.3 90.2 ± 8.6

±6.4 ±1.7

±4.1

64.2 ± 6.0

±1.2

159.8+15.3 39.7 + 2.0 125 ±9.6 312 + 10.9 65 ± 2.8 25.5 ± 4.0 241 ± 15.0 1470+147 73.3 + 3.7 98 ± 6.7

denotes pertussis toxin-resistant Cys-»Gly variant of G^

Table 4. Efficacy measurements of ligands at the A1 adenosine receptor when coexpressed with Cys—>Gly variants of Gia or as fusion proteins with these G proteins Ligand

Efficacy to activate (% of NECA) A,ARGi3a

CPA GR79236X CCI4019 GR56071 GR56072 GR66683 S-PIA R-PIA CHA a

90 ±29 121±7 114 + 2 115±4 125±10 111±5 136 ±2 118±4 110±8

113±10 119±8 119 + 3 125 + 7 126±15 116±3 141 ±7 156 ±6 129 ±4

96 ±10 125 ±8 112±3 107 + 4 105 ±19 107 + 4 117 + 4 123 ±6 129 ±5

98 ±1 100 ±2 96 ±3 101 ±2 103 ±4 103 ±2 102 ±9 95 ±2 96 + 3

102 ±2 105 ±2 100 + 2 104 ±1 106 ±1 101 ±2 85 + 4 103 ±1 111±3

103 ±1 97 ±2 95 ±1 103 ±1 99 ±3 96±2 103+1 103 ±1 104 ±2

denotes pertussis toxin-resistant Cys-»Gly variant of G,a.

In conclusion, these results provide strong evidence that a fusion protein approach can be adopted as a reliable means of studying interactions between particular receptor-G protein tandems.

6.5 Receptor-G protein fusion regulation of effectors The work described in the previous sections of this chapter has focused entirely on the use of receptor-G protein fusions as an alternative means to study the complexities of receptor-G protein interaction. However, a number of recently published articles have looked at the downstream regulation of effector molecules by receptor-G protein fusion proteins (3, 7, 8). Indeed, the first work detailing the use of fusions, by Bertin et al. (3), involved the construction of a B2-adrenoceptor-Gs chimaera. Addition of agonist was able 132

6: Construction and analysis of receptor-G protein fusion proteins to cause activation of adenylate cyclase following expression of this fusion protein in S49 lymphoma eye" cells, which lack endogenous Gsa (3). More recently, Seifert et al. (8) used fusions between the B2-adrenoceptor and the long and short splice variants of Gsa to demonstrate subtle differences in their coupling mechanisms. In this study, regulation of adenylate cyclase by these fusion proteins was used as an important method of measuring G protein activation. Furthermore, a fusion protein between the yeast a-factor receptor, Ste2, and the G protein, Gpal was found to transduce signal efficiently in yeast cells devoid of endogenous STE2 and GPA1 genes (7). However, the fusion constructs between the a2A-adrenoceptor and Cys—»Gly G11a, and between the A1 adenosine receptor and Cys—»Gly forms of G11a, G12a,and G13a do not appear to be able to transduce signals to effector proteins via their attached G proteins (9, 11). For example, stable expression of the a2A-adrenoceptor-Cys—»Gly G11a fusion protein in Ratl fibroblasts results in an a2A-adrenoceptor agonist (UK14304)-mediated stimulation of GTPase activity in membranes from these cells. This stimulation is, however, only partially sensitive to pretreatment of cells with pertussis toxin, demonstrating that the receptor in the fusion protein can couple to both fused pertussis toxin-resistant G protein partner and endogenous pertussis toxin-sensitive Gi G proteins (9). a2A-adrenoceptor agonists can also cause inhibition of forskolin-amplified adenylate cyclase in these cells; however, this response is abolished following exposure to pertussis toxin (Figure 15), suggesting that effector regulation is purely via fused receptor and the endogenous Gi G protein population. These data demonstrate that receptors constrained within fusion proteins have the capacity to regulate both their fused G protein partner and also those endogenously expressed. This is not entirely surprising, since GPCRs that have had other proteins, such as green fluorescent protein, linked to their C-termini still appear to be capable of interacting with and activating cellular G proteins (56). However, this is also an extremely important point to consider when adopting such an approach to measure coupling between specific receptors and G proteins. The inability of the a2A-adrenoceptor-Cys—»Gly G11a fusion protein to mediate inhibition of forskolin-amplified adenylate cyclase activity via the physically linked Gi may simply reflect limited spatial opportunities of the fusion protein-linked G protein, or the relative G protein activation ratios, which in the fusion must be limited to mol per mol of agonist-occupied receptor. Indeed, both the a2A-adrenoceptor and A1 adenosine receptor possess relatively short C-terminal tails compared with the B2-adrenoceptor, and at present little is known the about the cellular distribution of signalling proteins relative to one another (57). The inability of Gi-containing receptor-G protein fusions to regulate effectors may also be due to an unknown property of the fused Gi G protein. This possibility will hopefully be addressed by construction of further Gi-based receptor-G protein tandems. 133

Alan Wise

Figure 15. Agonist occupation of a2A-adrenoceptor-Cys-»Gly G11a results in inhibition of adenylate cyclase via endogenous Gi G proteins, but not via fused Cys—»Gly G11a. Adenylate cyclase measurements were performed on Rat1 fibroblasts stably expressing the a2A-adrenoceptor-Cys-»Gly G11a fusion protein. Basal activity (1 and 5), stimulation by forskolin (50 uM) (2 and 6), and the capacity of UK14304 (100 uM ) (3 and 7), or adrenaline (100 uM) (4 and 8) to inhibit forskolin-amplified activity were then measured in untreated (1-4) and pertussis toxin-treated (5-8) cells. Significant inhibition by agonists was only observed in untreated cells. From ref. 9, with permission.

In conclusion, receptor-G protein fusion proteins can regulate effector pathways; however, careful data analysis is necessary to dissect coupling between receptors and their fused G protein partners, and those endogenously represented. From the data published so far, it would seem that the G protein partner contained within a particular fusion is a crucial determinate in denning effector interaction capability. In addition, the capacity of receptor-G protein fusions harbouring Gq and G12 family G protein members to regulate effectors is also eagerly awaited.

7. Summary and future perspectives The generation and use of fusion proteins comprising GPCRs and G protein a subunits is beginning to provide an alternative means to explore and delineate 134

6: Construction and analysis of receptor-G protein fusion proteins the complexities of mechanisms governing receptor-G protein interaction. This technology provides a direct means of studying the catalytic activation of G proteins by their cognate receptors, and will hopefully provide novel means to examine specificity in the pharmacology of cell signalling systems. The use of pertussis toxin-resistant G, G proteins also provides a means of studying interactions between particular receptor-G; G protein tandems in isolation. The generation of receptor-G protein fusions containing members of the Gq and G12 families is restricted due to endogenous expression of these G proteins, and hence the inability to demonstrate unequivocally whether a particular response results from activation of the G protein partner of the fusion protein or from activation of endogenous G proteins. This may be overcome by the use of chimaeric G, G proteins with altered receptor-coupling specificities, in which the C-terminus is exchanged with that of other G protein a subunits. Other future work will involve the generation of receptor G; G protein fusions which can efficiently regulate effector proteins. Such work should permit a more widespread use of this technology to study G proteinmediated signalling pathways.

Acknowledgements I would like to thank Professor Graeme Milligan for providing the facilities to perform most of the experiments described herein. I would also like to thank members of the Molecular Pharmacology Group at the University of Glasgow, especially Craig Carr, for excellent technical assistance. In addition, I would like to thank Stephen Rees and members of the Receptor Systems Unit at GlaxoWellcome Research and Development for their help and support to continue this work.

References 1. Oilman, A. G. (1987). Anna. Rev. Biochem., 56,615. 2. Bourne, H. R., Sanders, D. A., and McCormick, F. (1990). Nature, 348,125. 3. Berlin, B., Freissmuth, M., Jockers, R., Strosberg, A. D., and Marullo, S. (1994). Proc. Natl. Acad. Sci. USA, 91,8827. 4. Wise, A., Carr, I. C., and Milligan, G. (1997). Biochem. J., 325,17. 5. Wise, A., Carr, I. C., Groarke, D. A., and Milligan, G. (1997). FEBS Lett., 419, 141. 6. Wise, A., and Milligan, G. (1997). J. Biol. Chem., 272, 24673. 7. Medici, R., Bianchi, E., Di Segni, G., and Tocchini-Valentini, G. P. (1997). EMBO. J., 16,7241. 8. Seifert, R., Wenzel-Seifert, K., Lee, T. W., Gether, U., Sanders-Bush, E., and Kobilka, B. K. (1998). J. Biol. Chem., 273,5109. 9. Burt, A. R., Sautel, M., Wilson, M. A., Rees, S., Wise, A., and Milligan, G. (1998). J. Biol. Chem., 273,10367.

135

Alan Wise 10. Carr, I. C, Burt, A. R., Jackson, V. N., Wright, J., Wise, A., Rees, S., and Milligan, G. (1998). FEBS Lett., 428,17. 11. Wise, A., Sheehan, M., Rees, S., Lee, M. G., and Milligan, G. (1999) Biochemistry, 38,2272. 12. Gill, D. M., and Meren, R. (1978). Proc. Nad. Acad. Sci. USA, 75, 3050. 13. Katada, T., and Ui, M. (1979). J. Biol. Chem., 254,469. 14. Katada, T., and Ui, M. (1981). J. Biol. Chem., 256, 8310. 15. Kurose, H., Katada, T., Haga, K., Ichiyama, A., and Ui, M. (1986). J. Biol. Chem., 261, 5423. 16. Milligan, G. (1988). Biochem. J., 255,1. 17. Wise, A., Watson-Koken, M.-A., Rees, S., Lee, M., and Milligan, G. (1997). Biochem. J., 321,721. 18. Senogles, S. E. (1994). J. Biol. Chem., 269, 23120. 19. Hunt, T. W., Carroll, R. C., and Peralta, E. G. (1994). J. Biol. Chem., 269,29565. 20. Berstein, G., Blank, J. L., Smrcka, A. V., Higahijima, T., Sternwies, P. C., Exton, J. H., and Ross, E. M. (1992). J. Biol. Chem., 267, 8081. 21. Casey, P. J., Fong, H. K. W., Simon, M. I., and Gilman, A. G. (1990). J. Biol. Chem., 265, 2383. 22. Scott, B., Nielson, K., Cline, J., and Kretz, K. (1994). Strategies 7, 62. 23. Guyer, C. A., Horstman, D. A., Wilson, A. L., Clark, J. D., Cragoe, Jr., E. J., and Limbird, L. E. (1990). J. Biol. Chem., 265,17307. 24. Gorman, C. (1985) In DNA cloning: a practical approach, (ed. D. M. Glover). Vol. II, p. 143. IRL Press, Oxford. 25. Sompayrac, L., and Danna, L. (1981). Proc. Nad. Acad. Sci. USA, 78, 7575. 26. Graham, F., and van der Eb, A. (1973). Virology, 52, 456. 27. Feigner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. N., Wenz, M., Northrop, J. P., Ringold, G. M., and Danielson, M. (1987). Proc. Natl. Acad. Sci. USA, 84, 7413 28. Ferguson, K.M., Higashijima, T., Smigel, M. D., and Oilman, A. G. (1986). J. Biol. Chem., 261, 7393. 29. McKenzie, F. R. (1992). In Signal transduction: a practical approach. (ed. G. Milligan), p. 31. IRL Press, Oxford 30. Wieland, T., and Jakobs, K. H. (1994). In Methods in enzymology (ed R. lyengar). Vol. 237, p. 3. Academic Press, London. 31. Hunt, T. W., Fields, T. A., Casey, P. J., and Peralta, E. G. (1996). Nature, 383, 175. 32. Berman, D. M., Kosaza, T., and Gilman, A. G. (1996). J. Biol. Chem., 271,27209. 33. Stephenson, R. P. (1956). Br. J. Pharmacol, 11, 379. 34. Kenakin, T. P. (1989). Trends Pharmacol. Sci., 10,18. 35. MacEwan, D. J., Kim, G. D., and Milligan, G. (1995). Mol. Pharmacol., 48, 316. 36. MacEwan, D. J., Kim, G. D., and Milligan, G. (1996). Biochem. J., 318,1033. 37. Jones, T. L. Z., Simonds, W. F., Merendino, J. J., Brann, M. R., and Spiegel, A. M. (1990). Proc. Natl. Acad. Sci. USA, 87, 568. 38. Wedegaertner, P. B., Wilson, P. T., and Bourne, H. R. (1995). J. Biol. Chem., 270, 503. 39. Milligan, G., Parenti, M., and Magee, A. I.(1995). Trends Biochem. Sci., 20,181. 40. Wedegaertner, P. B., Chu, D. H., Wilson, P. T., Levis, M. J., and Bourne, H. R. (1993). J. Biol. Chem., 268,25001.

136

6: Construction and analysis of receptor-G protein fusion proteins 41. Edgerton, M. D., Chabert, C., Chollett, A., and Arkinstall, S. (1994). FEBS Lett., 354,195. 42. Mumby, S. M., Heuckeroth, R. O., Gordon, J. I., and Gilman, A. G. (1990). Proc. Natl. Acad. Sci. USA, 87,728. 43. Parenti, M., Vigano, M. A., Newman, C. M. H., Milligan, G., and Magee, A. I. (1993). Biochem. J., 291,349. 44. Galbiati, F., Guzzi, F., Magee, A. I., Milligan, G., and Parenti, M. (1994). Biochem. J., 303, 697. 45. Wise, A., Grassie, M. A., Parenti, M., Lee, M. G., Rees, S., and Milligan, G. (1997). Biochemistry, 36,10620. 46. Milligan, G., Parenti, M., and Magee, A. I. (1995). Trends. Biochem. Sci., 20,181. 47. Wall, M. A., Coleman, D. E., Lee, E., Iniguez-Lluhi, J. A., Posner, B. A., Gilman, A. G., and Sprang, S. R. (1995). Cell, 83,1047. 48. Lambright, D. G., Sondek, J., Bohm, A., Skiba, N.P., Hamm, H. E., and Sigler, P. B. (1996). Nature, 379,311. 49. Taylor, J. M., Jacob-Mosier, G. G., Lawton, R. G., Remmers, A. E., and Neubig, R. R. (1994). J. Biol. Chem., 269,27618. 50. Stiles, G. L. (1992). J. Biol. Chem., 267, 6451. 51. Fredholm, B., and Dunwiddie, T. V. (1998). Trends. Pharmacol. Sci., 9,130. 52. Garwins, P., and Fredholm, B. (1992). J. Biol. Chem., 267,16081. 53. Kenakin, T. (1995). Trends Pharmacol. Sci., 16, 232. 54. Kenakin, T. (1995). Trends Pharmacol. Sci., 16,188. 55. Eason, M. G., Jacinto, M. T., and Liggett, S. B. (1994). Mol. Pharmacol, 45,696. 56. Barak, L. S., Ferguson, S. S. G., Zhang, J., Martenson, C., Meyer, T., and Caron, M. G. (1997). Mol. Pharmacol., 51,177. 57. Neubig, R. R. (1994). FASEB J., 8,939.

137

This page intentionally left blank

7

Application of the baculoviral expression system to signal transduction ANDREW PATERSON

1. Introduction The Baculovirus Expression Vector System (BEVS) is an extremely versatile and powerful laboratory tool (1). In simplest terms, the BEVS allows the forced heterologous expression of recombinant protein in susceptible host strains of insect cells. Gene products from the simplest prokaryotes through to the higher eukaryotes have been expressed successfully in this system. Recombinant proteins are recovered in a folded and active form, and posttranslation modifications, such as phosphorylation, lipid acylation, and simple glycosylation are often preserved. Oligomerization is preserved also, as is the ability to form holomeric complexes. This has been demonstrated clearly with expression of the ~ 40 kDa G proteins in their native «p-y heterotrimeric complexes (2), and extended to the activation of the kinase, c-Raf, upon coexpression with Ras and Src tyrosine kinase (3). Since the host insect cells maintain the ability to process eukaryotic pre-mRNA, heterologous expression from genomic DNA is possible. However, the most attractive facet of BEVS must be the high level of heterologous protein expression with values of between 0.1 and 30% of total cell protein being reported, and often with gene products that fail to express with activity or reasonable yield in other systems. The rationale of the BEVS is fairly simple. The baculoviral genome is engineered to incorporate the coding sequence of a foreign protein. Incorporation is directed to a position downstream, and under the control, of a strong viral promoter. Activation of the promoter during the normal progression of baculoviral infection results in expression of the foreign protein. This is seen as an accumulation of foreign protein, which is available for harvest before completion of the viral life cycle and lytic death of the cell. It is now a fairly simple matter to introduce our chosen coding sequence into the genome of the most commonly used baculovirus, Autographa californica multiple nuclear polyhedrosis virus (AcMNPV). The ability of AcMNPV to

Andrew Paterson Table 1. Signalling proteins expressed in the BEVS Reference Human B2-adrenoceptor Human M1 and M2 muscarinic acetylcholine receptor EGF receptor extracellular domain EGF receptor tyrosine kinase domain Gaq family Gai1,Gai2, Gai3, GaD, Gas Ga12, Gaz Free By-subunits Adenylyl cyclase (type I) PhospholipaseC-p Phospholipase C-y cAMP phosphodiesterase Protein kinase C isoenzymes

4 5 6 7 2 8 9 10 11 12 13 14 15,16

p60 src PI 3-kinase (p85/p110 complex) B2-adrenoceptor kinase (GRK2) Protein kinase B isoenzymes

17 18 19 20

replicate, package progeny virions, and express virally-encoded proteins in cells derived from Spodoptera frugiperda, the fall army worm, has provided the technology we require both to engineer recombinant baculovirus and to express the encoded protein. Over the last decade proteins expressed in the BEVS have been applied to many aspects of signal transduction research (see Table 1). In most of these examples the knowledge gained has been critical and, in may cases, where other expression systems fail to provide suitable material. Several excellent dedicated laboratory manuals describing the molecular biology of the baculoviridae and their use in the laboratory are available currently (21, 22). They are recommended as further reading, as are the laboratory manuals provided with baculoviral products purchased from Life Technologies, Pharmingen, and Invitrogen.

2. Sf9 cell culture The clonal Spodoptera frugiperda line, Sf9 (ATCC # CRL 1711), is most often first choice for use in the BEVS. Its maintained culture is well documented, and information regarding its growth characteristics and requirements are available readily (21). The following description of methodology for establishing novel recombinant baculoviral vectors and protein expression will focus on their use. Alternative cell lines will be discussed briefly at the conclusion of this chapter. In culture, Sf9 cells grow either as monolayers or as stirred suspensions. In 140

7: Baculoviral expression both cases, they grow as monodisperse cells exhibiting logarithmic growth kinetics, and doubling times of less than 30 h. Slow-growing cultures that aggregate to clusters should be discarded. Use of Sf9 cells with sub-optimal health is a common source of problems. Rigid discipline must be adopted with continuous cultures of Sf9 cells. Maintaining Sf9 cells in logarithmic, suspension growth and preventing cell clustering is best achieved over what is an extremely narrow range of cell density (1.0-2.5 X 106 cells mH). Healthy Sf9 cultures may double in density in less than 27 h, and this demands subculturing every 36 h or so.

2.1 Media Choice of basal medium is a matter of personal preference. Many formulations have evolved, and these include TNM-FH, IPL-41, Ex Cell 400, and TC 100. These have been reviewed extensively elsewhere (21, 22). Much of the methodology following relies heavily on complete TNM-FH. The suggested formulation for complete TNM-FH is Grace's Antherean Culture medium (Vaughn's modification) supplemented with 3.3 g l-1 lactalbumin hydrolysate, 3.3 g l-1 Tissue Culture Yeastolate, 10% (v/v) fetal bovine serum, 100 ug ml-1 streptomycin, 100 IU ml-1 penicillin, and 0.25 ug ml-1 amphotericin B. It should be possible to substitute complete TNM-FH with other proprietary serum-containing media in most of the following methodology.

2.2 Reviving Sf9 cells from frozen Thaw a vial of frozen cells in a 27°C water bath. This must be done rapidly and care taken to prevent the thawed culture warming much above freezing point. Dilute the thawed cells in 10 ml ice-cold complete TNM-FH and transfer to a 75 cm2 TC flask. Allow the cells to attach for 15-30 min at 27°C. Replace the medium and unattached cells with 12 ml complete TNM-FH, and incubate at 27°C until the monolayer reaches confluence.

2.3 Maintaining monolayer cultures of Sf9 Cells Monolayers should be passaged 12-24 h after confluence. Detach the monolayers into their growth medium with a sterile cell scraper. Dilute the suspension fourfold with warmed complete TNM-FH, and replate 12 ml in a 75 cm2 TC flask. A cell density of approximately 6 X 104 Sf9 cells cm2 is desired. Incubate the monolayer at 27°C overnight, replace the medium with fresh complete TNM-FH, and further incubate the cells at 27°C until confluence (approximately 2-3 days). As the culture establishes it is possible to increase the dilution at each passage. After 3-4 passages it will be possible to split the monolayer 8- to 10-fold without loss of cell viability. The culture can be maintained in 75 cm2 TC flasks until scale-up is required. For instance, it will be desirable to seed 141

Andrew Paterson at least two 150 cm2 TC flasks in preparation for adaptation to suspension growth.

2.4 Adaptation to suspension culture Growing Sf9 cells as monolayers is convenient. It is relatively simple, inexpensive, and does not require capital investment in specialized equipment. Provided that a temperature of 25-27°C can be established, Sf9 cells will grow. Many low-cost incubators are available, and there is no requirement to supply gassed atmospheres as in mammalian cell culture. Disused 37°C, CO2 incubators are adapted readily for use with Sf9 cells. These are convenient approaches but, in truth, most laboratories have a draught-free area that will maintain a temperature between 25 and 27°C and can be used. Enterprising laboratories have been known to grow these cells in disused cupboards or cabinets fitted with a heat lamp. Splitting Sf9 monolayers by scraping or other means results in loss of significant numbers of cells. The resulting dead cells are removed easily by washing the monolayer with fresh medium on the day following plating. However, their presence does prevent use of monolayer-derived cells for transfections (see Protocol 1), plaque assays (see Protocol 2), etc. It is recommended strongly that monolayers are adapted to and maintained in suspension as soon as possible. 2.4.1 Equipment Suspension cultures of Sf9 cells should be maintained in microcarrier spinner flasks fitted with an impeller blade. Those manufactured by the Bellco Glass Co are recommended. These are available in a range of sizes. Continued culture of Sf9 cells from frozen aliquots to the production of milligram amounts of protein will entail the manipulation of cell suspensions in volumes ranging from 40 to 400 ml. In each case an adequate surface area to volume ratio is required, and the cell suspension should never exceed half the spinner flask's capacity. For example, in the routine culture of Sf9 cells it is recommended that 100 ml of culture be grown in a 250 ml flask (Bellco #1965-00250). A range of spinner flasks should therefore be available to accommodate this. To facilitate stirring, and to help inhibit formation of cell clusters or clumps, the spinner flask impeller blades should have only 5 mm clearance from the vessel wall. Impeller blades marketed for larger spinner flasks (Bellco #A523-199) can be cut to size and fitted in 250 ml spinner flasks. Microcarrier spinner flasks are operated in conjunction with a magnetic stirring plate. Those intended solely for use with microcarrier spinner flasks are recommended. The stirring plate should be non-heating, and capable of maintaining a range of speeds from 50 to 150 r.p.m. These are marketed in a range of models (Bellco ranges #7760-, 7761-, and 7765- ). Although not essential, models featuring tachometers remove the chore of calibrating impeller speeds. 142

7: Baculoviral expression 2.4.2 Suspension growth of Sf9 cells Detach the confluent monolayers from at least two 175 cm2 TC flasks, and collect by centrifugation at 500 g for 5 min at room temperature. Aspirate the medium and cell debris, and resuspend the cell pellet to a density of 1 X 106 cells per ml with complete TNM-FH. Seed 50 ml of this suspension in a 100 ml microcarrier spinner flask (Bellco #1965-00100) and maintain the suspension at 70 r.p.m. in a 27°C incubator. Spinner flask side-arm caps may be loosened to allow gaseous exchange, provided the incubator is humidified. As the cell density reaches 2.5 X 106 per ml (approximately 48 h) the culture should be diluted to 1.0 X 106 per ml. Care should always be taken that the cell density does not exceed 3 X 106 per ml. A cell suspension may be maintained in this manner for several months. The simplest procedure is to maintain a given volume of culture, and to discard the surplus suspension generated with each dilution. The surplus cells grown in the first week after adaptation to suspension culture should not be employed for baculoviral work. Superior results will be obtained with cells that have grown in suspension for at least 7 days. Adapted cells should grow without clumping, and exhibit greater than 90% trypan blue exclusion, with doubling times of 24-30 h. Satisfied with the health of the suspension, the cells may then be used for generating novel virus, amplification of viral stocks, or for expression of recombinant protein.

2.5 Freezing Sf9 cells For practical reasons it is suspension cultures that are frozen for storage. The culture to be frozen must be in optimal health. Collect the cells from at least 50 ml of culture by centrifugation at 500 g for 2 min at room temperature. Aspirate the medium and cell debris, resuspend the cells to 1 x 107 cells per ml in DMSO:fetal bovine serum (1:9), and place 1 ml aliquots in sterile screwtop cryovials. Prepare 5-10 vials for storage. Wrap the filled cryovial in cotton wool, place in a polystyrene box, and freeze slowly to -70°C overnight. Remove the frozen vials, and place in a liquid nitrogen cell vat. Confirm viability of the frozen stock by reviving a vial (see Section 2.2) after at least 2 days in storage.

2.6 Adaptation to serum-free medium Serum-free medium is employed widely for culture of Sf9 cells. Apart from the obvious advantage of cost, the absence of serum-derived protein, and inclusion of detergent to prevent cell shearing in stirred culture is beneficial. However, reduced cell viability and decreased levels of recombinant protein expression are experienced with cells maintained for extended periods in serum-free medium. A compromise is to maintain cells in serum-containing medium, and to adapt a portion to serum-free medium at least a week before required. 143

Andrew Paterson Formulation of serum-free medium is a matter of choice (21). Several excellent proprietary serum-free media are available. These include Ex-Cell 401 and Ex-Cell 420 from JRH Biosciences. Before use these should be completed with 100 IU ml-1 penicillin, 100 ug ml-1 streptomycin, and 0.25 ug ml-1 amphotericin B. Cells grown in complete TNM-FH are weaned gradually onto serum-free medium over a period of 10 to 20 days. Cells should be diluted to 0.5 X 106 per ml in a 1:1 mixture of complete TNM-FH and serum-free medium, and incubated at 27°C until a density of 2.5 X 106 per ml is reached. Cells are then diluted to 0.5 X 106 per ml in a 1:4 mixture of complete TNM-FH and serumfree medium, and allowed to grow as before. Finally, the cells are diluted in a 1:10 mixture, and grown for a further 3 days until they reach 2.5 X 106 per ml. At this stage the cells can be assumed to have adapted to serum-free medium, and can be grown in the absence of TNM-FH. An alternative and less involved approach is to adapt the cells in a single step. Collect cells grown in complete TNM-FH by centrifugation at 500 g, and resuspend them to 1.0 X 106 per ml in ExCell 401 completed with antibiotic and antimycotic. Stir the cells at 70 r.p.m. for two days at 27 °C until the cell density reaches 2.5 X 106 per ml. Dilute the cells to 1 X 106 per ml in fresh complete ExCell 401 and grow as before. The cells tend to clump as they adapt to serum-free medium. Increasing the rate of stirring from 70 to 120 r.p.m. reduces this.

3. The baculoviral life cycle, and constructing recombinant baculoviral vectors 3.1 Time course of viral infection and the polh locus Baculoviral infection leads to lytic death of an infected insect cell. This will take approximately 4 days in AcMNPV-infected Sf9 cells, and results in release of many competent progeny virions. During infections baculoviralencoded genes are transcribed and expressed in a cascade of co-ordinated events. These can be separated and classified into the three phases of viral gene activity: early, late, and very late. These are described further below. Early: Transcriptionally active in the absence of any viral gene expression, and is thus dependent on the host cell's transcriptional machinery. Early genes are active in the first 10 h of infection. Late: Dependent on early viral gene expression and DNA replication. Late genes are active approximately 8-18 h post-infection. For instance, the p6.9 basic core-associated protein promoter (pcor) is active at this time. Very late: Dependent on early viral gene expression and viral DNA replication. Transcription of very late genes will initiate approximately 18 h postinfection, but will probably be at their highest levels 27-48 h post144

7: Baculoviral expression infection. The polyhedrin protein promoter (ppolh) is active during this phase of viral infection. A recombinant baculoviral expression vector is constructed by modifying the AcMNPV genome to encode a foreign protein. In the vast majority of cases the coding sequence is inserted at the viral polyhedrin (polh) locus and its translation controlled by the polh promoter (ppolh). This promoter is active in the very late phase of viral infection and drives polyhedrin expression until it predominates in the infected cell. This protein is not essential for viral function in cultured Sf9 cells, and the coding sequence can be substituted without affecting viral viability. AcMNPV is a double-stranded DNA virus. The circular genome is large (approximately 120 kb), and foreign cDNA cannot easily be introduced at the viral polh locus with techniques as convenient as restriction digestion followed by ligation. Rather, insertion of cDNA is achieved either by recombination between a transfer plasmid harbouring the sequence and AcMNPV, or by Tn7-mediated transposition into a baculoviral shuttle vector.

3.2 Construction of a recombinant transfer plasmid Some transfer plasmids which are available commercially are given in Table 2. Their application to virus construction either by cotransfection or through baculoviral shuttle vectors is indicated also. They offer several possibilities. The recombinant protein can be expressed either with its full-length, native sequence, or fused to a tag which facilitates rapid purification. In addition, all transfer plasmids encode polyadenylation and transcription termination sequences downstream of the multiple cloning site (mcs). When constructing the recombinant transfer plasmid, coding sequence should be introduced with the minimum of 5'-untranslated DNA. The 5'untranslated mRNA should not form significant secondary structure nor possess any transcription termination sequences. Any sequence to be inserted 3' to the translation termination site should be minimized, and should not contain a transcription termination sequence.

3.3 Construction of recombinant baculo virus by cotransfection Possibly the most common method for constructing recombinant baculovirus is to cotransfect insect cells with a transfer plasmid construct and AcMNPV genomic DNA. The mcs of the transfer plasmid is flanked by considerable tracts of sequence derived from the AcMNPV polh locus. This facilitates recombination and integration of the mcs at the viral polh locus when cotransfected with AcMNPV genomic DNA (Figure 1A). The double recombination required for successful integration occurs with low, but finite efficiency. The efficiency of recombination can be increased with linearized 145

Table 2. Transfer plasmids Name

Promoter

Pre-existing tag

Compatible with cotransfection"

Compatible with baculoviral shuttle vectors6

Supplier

pVL1392/pVL1393 pBacPAK8/9 pBac-1 pAcG1/2T/3X pAcHLT-A/B/C pAcGHLT-A/B/C pBacPAK-His1/2/3 pBac-2cp pAcMP2/3 pFastBAC 1 pFastBAC HTa/b/c

polh polh polh

none none none N-terminal GST A/-terminal His6 /V-terminal GST/His6 A/-terminal His6 A/-terminal His6 none none A/-terminal His6

yes yes yes yes yes yes yes yes yes no no

no no no no no no no no no yes yes

Invitrogen/Pharmingen Clontech Novagen Pharmingen Pharmingen Pharmingen Clontech Novagen Pharmingen Life Technologies Life Technologies

a b

see Section 3.3. see Section 3.4.

polh polh polh polh polh cor polh polh

Figure 1. Recombinant baculovirus construction. (A) Homologous recombination between AcMNPV-derived sequence in the transfer plasmid and the polh locus of AcMNPVwt genomic DNA. A double recombination event involving ORF 603 and ORF 1629 is given as an example. (B) AcMNPV DNA linearized at a novel Bsu36l site within ORF 1629 is nonviable. The virus is rescued by recombination with transfer plasmid. (C) Transposition in E. coli. The transfer plasmid mini-Tn7 element transposes to the baculoviral shuttle vector affTn7 site. Transposon function is supplied by a third helper plasmid. Transposition results in disruption of lacZa, and E. coli, which fail to stain blue when grown in the presence of IPTG and X-gal.

147

Table 3. AcMNPV viral preparations Name

Compatible with cotransfection3

Compatible with baculoviral shuttle vectorsb

Supplier

Linearized AcMNPV viral DNA

Linearize AcRP23.lacZ

yes

no

Pharmingen

Bsu361 -cut viral DNA

Baculogold BacPakB BacVector-1000 BacVector-2000 BacVector-3000

yes yes yes yes yes

no no no no no

Pharmingen Clontech Novagen Novagen Novagen

Baculoviral shuttle vector

Bacmid (bMON14272c)

no

yes

Life Technologies

' See Section 3.3. b See Section 3.4. c Supplied cotransformed with transposition helper plasmid, pMON7124, in the competent E. coli strain DH10Bac.

7: Baculoviml expression preparations of AcMNPV genomic DNA (see Table 3); however, less than 30% of the virus recovered from a cotransfection will be recombinant. The background of non-recombinant virus can be removed. Non-viable, linearized viral DNA preparations are now available. The AcMNPV genome has been modified by introduction of a novel Bsu36I site within ORF1629. This open reading frame is essential for viral viability, and maps close to the viral polh. Exhaustive digestion with Bsu36l both linearizes the viral DNA and truncates ORF1629. The resulting deletion destroys viral viability, and must be rescued by recombination with polh-based transfer plasmid (Figure 1B). Protocol 1 describes the cotransfection of Sf9 cells with Bsu36I-digested AcMNPV. Bsu36I-cut viral DNA preparations are available commercially form Pharmingen (BaculoGold), Clontech (BacPak6), and Novagen (BacVector-1000, -2000, and -3000). Although the high background of non-recombinant virus can be reduced significantly with Bsu36I-digests of AcMNPV, but the resulting progeny virus is still not clonal. Kitts and Posee (23) report that up to 5% of the virus constructed by cotransfection with Bsu361-digested AcMNPV did not express the protein encoded by the transfer plasmid. This can be eliminated and clonality ensured by purifying recombinants expressing virus to homogeneity. Purification by plaque assay is described in Protocol 2, and this should be performed with the supernatant recovered from cotransfected Sf9 cells (Protocol 1, Step 11). Plaques formed by virus expressing recombinant protein will be identified after SDS-PAGE analysis. Protocol 1. Cotransfection of Sf9 cells with Bsu36l-digested AcMNPV DNA Equipment and reagents • 25 cm2 TC flask (Corning # 430639) • Sf9 cells adapted to suspension growth in . Sterile microfuge tubes complete TNM-FH. The culture should be in -,,.,., . . ., . ,.- , exponential growth and at a density of #2059) <*ntr,fuge tubes (Falcon L^.5 x 106 cells per ml. _.,.._.. ,_ . , ., .. • Transfer plasmid construct (the com.TNM-FH (Grace's Antherean medium P (Vaughn s Modification) supplemented patibility of several with 3.3 g l-1 lactalbumin hydrolysate, " ' ... 1 3.3 g I- Yeastolate, 100 IU ml-1 penicillin, • 100 ug ml-1 Bsu36l-digested baculoviral DNA 100 ug ml-1 streptomycin, 0.25 ug ml-1 fungizone) • Transfection buffer (25 mmol I-1 Hepes. Complete TNM-FH (TNM-FH completed NaOH, pH 7.1; 140 mmol l-1 NaC|; 125 with fetal bovine serum to 10% v/v) mmol (-1 CaCI2)

Method 1. Plate 2 x 106 Sf9 cells in a 25 cm2 TC flask with 5 ml TNM-FH. Allow the cells to attach for 1 h at 27°C. 2. Mix 5 ug transfer plasmid construct DNA with 5 ul Bsu36l-digested

149

Andrew Paterson Protocol 1. Continued AcMNPV DNA in a sterile microfuge tube. Incubate for 5 min at room temperature. 3. Add 1 ml transfection buffer to the plasmid-AcMNPV DNA mix. 4. Add 1 ml TNM-FH to a sterile 14 ml centrifuge tube. 5. Add the transfection buffer-plasmid-AcMNPV mix to the TNM-FH. Add dropwise and tap the tube to mix every few drops. Do not vortex. 6. Incubate the mixture at room temperature for 30 min. A fine, suspended precipitate will form. 7. Aspirate any unattached cells and medium from the plated monolayer, and wash twice with 5 ml TNM-FH. 8. Aspirate any residual TNM-FH, and add the DNA suspension to the cells. 9. Incubate the cells with the DNA suspension for 4 h at 27°C. 10. Aspirate the transfection buffer-DNA suspension, and wash the monolayer twice with 5 ml supplemented TNM-FH. 11. Replace the medium with 5 ml fresh, complete TNM-FH, and incubate for 5 days at 27°C. 12. Aspirate the conditioned medium from the flask, and remove the cell debris by centrifugation at 500 g for 5 min. Store the supernatant at 4°C in the dark. This should be considered passage zero (pO) viral stock. Alternatively, store 1 ml in a sterile vial at -70°C.

Protocol 2. Purification of recombinant baculovirus by plaque assay Equipment and reagents • 100 mm tissue culture dishes (Corning • Sf9 cells adapted to suspension growth in # 430167) complete TNM-FH. The culture should be in . Glass Pasteur pipettes sterilized by autoexponential growth and at a density of claving at 121°C, 15 p.s.i., for 20 min, and 1.5-2.5 x 10 cells per ml. dried in an oven • Baculoagar (2.5% w/v solution of Baculo. Pipette filler bulb sterilized with 70% (v/v) agar, Invitrogen # B315-50 made with ethanol I tissue culture-grade water, and sterilized by autoclaving at 121°C, 15 p.s.i., for 20 min) 24-well tissuecultureplates(Corning # • Gel sample buffer (25 mmol l-1 Tris-HCI, pH ) 8.0; 2.5 mol I-1 urea; 5% w/v SDS; 6% w/v • Complete TNM-FH (see Protocol 1) DTT)

Method 1. Plate6x 106Sf9 cells with 10 ml of complete TNM-FH in each of several 100 mm tissue culture dishes. Allow the cells to plate for 1 h at 27°C.

150

7: Baculoviral expression 2. For duplicate plates, aspirate all but the last 1 ml of medium, and inoculate the monolayer with 2 ml pO viral stock diluted in complete TNM-FH. A range of dilutions covering 10-3 to 10-8 should be employed. Incubate the monolayers with diluted virus at 27°C for 1 h. 3. Prepare molten 2.5% agarose by autoclaving, and cool to 42°C in a water bath. At the same time warm 30 ml aliquots of complete TNMFH to 42°C in the water bath. 4. One at a time, place the dishes on a sheet of polystyrene, and aspirate the medium and virus. Carefully overlay the monolayer with 10 ml of agarose/TNM-FH prepared immediately before use by mixing 10 ml of warm 2.5% agarose with a 30 ml aliquot 10% FBS/TNM-FH. 5. 'Dry' the plates in a humidified 27°C incubator for 2 h. 6. Seal the plates in a polygrip bag, and incubate at 27°C for 5-6 days until the viral plaques are visible. 7. The plaques are small (0.5-1.0 mm) and should be visible against the grey/white background of the monolayer.a Observed under a lowpower microscope plaques are observed as gaps in otherwise uniform monolayers. Cell debris should be visible and the gaps may be lined with cells which are swollen and of irregular shape. 8. Pick from a plate with 10-50 well separated plaques. Pick the plaques with a sterile Pasteur pipette, carefully withdrawing the agarose with a sterilized Pasteur bulb. 9. Pipette the agarose plug into 1 ml of complete TNM-FH, and use 500 ul to inoculate 2 x 105 Sf9 cells plated in a well of a 24-well tissue culture plate. 10. Repeat with 5-10 plaques. 11. lncubate at 27°C for 5 days. 12. For each well, remove the conditioned medium and transfer to a labelled, sterile centrifuge tube. Clarify the supernatant by centrifugation at 500 g for 5 min. This should be considered passage 0 (pO) viral stock. Solubilize the monolayer remaining with 200 ul gel sample buffer. 13. Heat the solubilized monolayer to 95°C for 2 min.b 14. Analyse expression of recombinant protein by SDS-PAGE and/or immunoblotting of 30-100 ul of the solubilized monolayer. Half or more of the selected plaques should now be identified as recombinant, expressing plaques. 15. Select two plaques on their ability to express recombinant protein.

151

Andrew Paterson Protocol 2. Continued Subject the corresponding viral supernatants to a second round of plaque purification. The tissue culture supernatants collected in Step 12 should be used for this purpose. a Viral plaques can be identified by negative staining. On the fourth day of the plaque assay, the existing Baculoagar is overlaid with 5 ml 50 ug ml-1 neutral red, 0.5% w/v Baculoagar in complete TNM-FH. Plaques are visualized the following day as clear areas within the redstained monolayer. b The solubilized protein sample will be viscous. It should be possible to remove sufficient denatured sample to load 100 ul to an SDS-PAGE gel. Alternatively, the DNA may be sheared by passing through a small-gauge needle.

3.4 Recombinant baculovirus construction with baculoviral shuttle vector Fairly recently a new strategy has developed. Recombinant baculovirus now can be generated by transposon Tn7-mediated insertion of foreign DNA into baculoviral shuttle vectors (24). Baculoviral shuttle vectors are, simply, the AcMNPV genome modified to allow it both to replicate in E. coli as a large, 120 kb plasmid, and to act as a target for transposon Tn7. Foreign cDNA is transposed from a mini-Tn7 element contained within a second, transfer plasmid, and transposition is supported from a third, helper plasmid. Successful transposition is identified by disruption of the lacZa. fragment flanking the transposon Tn7 target sequence (attTn7) in the shuttle vector (Figure 1C). This is seen as failure to stain blue when the transformed E. coli are grown in the presence of IPTG (isopropyl-B,D-thiogalactopyranoside) and X-gal (5bromo-4-chloro-3-indolyl-B,D-galactoside. Baculovirus construction with shuttle vectors is described in Protocol 3. Clonal recombinant baculovirus is generated with comparative rapidity, without plaque purification, and with the facilities present in most modern laboratories. This system is extremely useful and straightforward. It is available commercially in the Bac-to Bac or Bacmid system marketed by Life Technologies. Its major drawback is the availability of only a limited number of transfer plasmids. Protocol 3. Baculoviral shuttle vector (Bacmid) construction Equipment and reagents • Microfuge capable of 12000 r.p.m. . Sterile, wide-bore pipette tips (prepare both 1000 ul and 200 ul pipette tips by removing their ends with a sharp blade. A 2 mm aperture should remain. The tips can then be oracked, sterilized in an autoclave at 121 C and 15 p.s.i. for 20 min, and dried in a low-temperature oven).

• Wire cloning loop, glass spreader, and Bunsen burner for flaming . 14 ml capped culture tubes (Falcon # 2059) .1.5. ml microfuge tubes sterilized by autoclaving at 121°C, 15 p.s.i., for 20 min. • CaCI2-competant E. coli.(DH10Bac) (Life Technologies # 10361-012)

152

7: Baculoviral expression LB medium (prepare Luria-Bertrami medium with 10 g I-1 bacto-peptone, 5 g I-1 bacto-yeast extract, 10 g I-1 NaCI, and adjust pH to 7.0 with NaOH. Sterilize by autoclaving at 121 °C, 15 p.s.i., for 20 min) LB/K/G medium (prepare as described for LB medium. Sterilize by autoclaving at 121°C, 15 p.s.i., for 20 min. Cool to 55°C, and add kanamycin to 50 ug ml-1 and gentamycin to 7 ug ml-1)

• SOC medium (prepare SOC medium with 20 g I-1 bacto-tryptone, 5 g l-1 bacto-yeast extract, 0.5 g I-1 NaCI, 0.168 g I-1 KCI, and adjust pH to 7.0 with NaOH. Autoclave at 121 °C, 15 p.s.i., for 20 min, and cool to 55°C. Before use add 5 ml 0.22 micronfiltered 190 g I-1 MgCI2 and 20 ml 0.22 micron-filtered 180 g I-1 glucose) • Resuspension buffer (50 mmol I-1 Tris-HCI, pH 8.0; 10 mmol I-1 EDTA; 100 p-g ml-1

. LB/K/T/G/I/X agar plates (add 12 gI-1bactoagar to unsterilized LB medium and autoclave at 121°C, 15 p.s.i., for 20 min. Cool the agar to 55°C in a water bath, add kanamycin to 50 ug ml-1, tetracycline to 10 ug ml-1, gentamycin to 7 ug ml-1, IPTG to 40 ug ml-1, and X-gal to 200 ug ml-1. Plate 20 ml aliquots of the molten agar solution in 10 cm bacterial culture plates and allow .

, (200 mmol l-1 NaOH2 • Lysis buffer (200 mmol l-1 NaOH, 1% w/v SDS) • Neutralization buffer (3 mol l1 sodium acetate, pH 5.5) • Phenol:chloroform:isoamylalcohol (25:24:1) • Propan-2-ol . Absolute ethanol

to

cool.).

. TNM-FH (see Protocol 7) . Complete TNM-FH (see Protocol 7)

70%

(v/v)

etnanol

. Sterile H20 . Liposome transfection reagent (CellFectin, Life Technologies # 10362-010)

A. Site-specific transposition in E. coli (DHWBac) 1. Thaw CaCI2-competant DHIOBac E. coli on ice. 2. Add 10-50 ng recombinant pFastBac construct to 25 ul of the E. coli in a sterile microfuge tube. Mix very gently by tapping. Incubate on ice for 30 min. 3. Heat shock the mixture by warming to 42°C in a water bath for exactly 60s. 4. Remove the bacteria from the water bath and chill on ice for 2 min. 5. Add 900 ul SOC to the chilled bacteria, and incubate at 37°C for 4 h. 6. Spread 100 ul of the transformed bacteria on LB/K/T/G/I/X agar plates. Repeat this with 100 ul of the bacteria diluted 10- and 100-fold with SOC. 7. Incubate at 37°C for 18-24 h until isolated, blue colonies are clearly visible. Colonies must be allowed to grow sufficiently such that blue and white phenotypes are clearly distinguishable. The majority of colonies will be blue. 8. Pick a white colony with an inoculating loop and resuspend in 100 ul LB medium. Repeat this with at least four colonies. 9. Divide a fresh LB/K/T/G/I/X agar plate into six segments. Mark these on the base of the plate with a waterproof pen. 10. Streak the resuspended colonies from Step 8 on the divided LB/K/T/G/I/X agar plate. Streak each colony in a fresh segment, and incubate the plate at 37°C for 16 h overnight. 11. Confirm the white phenotype of each colony. If necessary repeat Steps 8-10.

153

Andrew Paterson Protocol 3.

Continued

12. Pipette 3 ml of LB/G/K medium into a capped culture tube and inoculate with a single colony picked from the restreaked LB/K/T/G/I/X agar plate. Repeat for each colony selected in Step 8. Grow the bacteria at 300 r.p.m. in a 37°C shaking incubator for 16 h overnight. B. Preparing 'mini-prep' shuttle vector DNA 1. Pellet 1.35 ml of the overnight culture by centrifugation at 12000 r.p.m. for 5 min in a sterile 1.5 ml microfuge tube. Aspirate the supernatant, add a further 1.35 ml of the overnight culture, and repeat the centrifugation. Repeat this for each colony grown overnight. 2. Aspirate all traces of the LB/G/K medium, and disperse the pellet completely in 250 ul of resuspension buffer. 3. Add 250 ul lysis buffer, mix by inverting the tube five or six times, and incubate at room temperature for 5 min. 4. Add 350 ul neutralization buffer, mix immediately by inversion several times, and sediment the precipitate in a microfuge at 12 000 r.p.m. for 10 min. 5. Transfer 750 ul of the supernatant with a wide-bore pipette tip to a fresh, sterile microfuge tube. 6. Add 750 ul phenol:chloroform:isoamylalcohol to the supernatant, mix gently by inversion, and incubate at room temperature for 2 h. Maximize the phenolxhloroform interface by incubating the microfuge tubes on their side. 7. Split the phases for 2 min by centrifugation in a microfuge at 12000 r.p.m.. Transfer 450 ul of the upper phase to a fresh microfuge tube and precipitate the DNA with 1050 ul ethanol. Allow precipitation to proceed for 3 h at -20°C. 8. Recover the DNA by centrifugation at 12000 r.p.m. in a microfuge for 10 min, and wash the pellet with 70% (v/v) ethanol. 9. Aspirate the supernatant and air-dry the shuttle vector DNA for 10-30 min. Solubilize with 40 ul sterile H2O at room temperature. C. Liposome-mediated transfection of Sf9 cells 1. Plate 9 x 105 Sf9 cells in each well of a six-well tissue culture plate. Allow the cells to attach for 1 h at 27°C. 2. Aspirate the medium and unattached cells, and add 2 ml TNM-FH. 3. Incubate the monolayers at 27°CC for 4 h. During this incubation wash the monolayers twice with 2 ml TNM-FH. 4. Dilute 5 ul of shuttle vector DNA (from Section B, Step 9) with 100 ul TNM-FH. 154

7: Baculoviral expression 5. Dilute 6 ul of liposome transfection reagent with 100 ul TNM-FH in a 24-well TC plate. Several transfections can be prepared simultaneously in a single plate. 6. Add the diluted shuttle vector to the transfection reagent/TNM-FH mixture. Mix by careful pipetting with wide-bore tips, and incubate at room temperature for 30 min. 7. Dilute the transfection reagent/shuttle vector mixture to 1 ml with further TNM-FH. 8. Aspirate the medium from the 6-well dish containing the plated Sf9 cells, and replace with the diluted transfection reagent/shuttle vector mixture. 9. Incubate the mixture on the cells for 5 h at 27°C. 10. Aspirate the diluted liposome transfection reagent/shuttle vector mixture from the cells, and replace with 2 ml complete TNM-FH. 11. Incubate in a humidified atmosphere at 27°C for 5 days. 12. Collect the conditioned medium and clear by centrifugation at 500 g for 5 min. This should be considered passage zero viral stock. 13. Store in the dark at 4°C.

3.5 Baculoviral passage and titre Large volumes of baculovirus will be required for the production of recombinant protein. Such volumes are produced, or amplified, by infecting a large number of Sf9 cells with a limited amount of virus and collecting the conditioned culture medium after 4 days incubation at 27°C. However, each amplification, or passage, provides an opportunity for degeneration of the viral stock. The risks of somatic mutation, accumulation of the spontaneous few polyhedra (FP) phenotype, or multiplication of contaminating, nonrecombinant baculovirus all increase with each passage of virus. Each of these will occur at the cost of recombinant baculovirus replication and ultimately protein expression. These problems are avoided by establishing baculoviral stocks from clonal virus, minimising passage number, and limiting the virusto-cell ratio when infecting cells to produce viral stock. Culture supernatants generated by incubating Sf9 cells with plaques (Protocol 2, Step 12) should be considered passage zero (pO), as should those generated by transfection with shuttle vector (Protocol 3, Step 12). The viral titre of both these stocks will be low and will require amplification. The method described in Protocol 4 employs a low virus-to-cell ratio, also referred to as multiplicity of infection (MOI), and generates only a small volume of pi viral stock. This pi viral stock is then suitable for infecting much larger volumes, and when used at an approximate MOI of 0.1, the 10 ml of pi virus has potential to generate over 10 1 of p2 viral stock. Amplification of virus in 155

Andrew Paterson suspension culture should be considered if litre volumes of viral stock will be required. Suspension cultures adapted to ExCell 401 serum-free medium and seeded at 1 X 106 cells per ml should be inoculated with 0.25 ml l-1of p1 or p2 virus and the viral supernatants collected after 4 days incubation at 27°C. The viral titre of p1 and p2 stock will be high and equals 1 X 108 pfu ml-1 or greater. This can be determined by plaque assay. Protocol4. Amplification of virus Equipment and reagents • Complete TNM-FH (see Protocol 1). . 75 cm2 and 150 cm2 TC flasks

• p0 viral stock generated from Protocol 2 or 3

either

Method 1. Plate 7 X 106 Sf9 cells with 10 ml complete TNM-FH in a 75 cm2 tissue culture flask and allow to attach for 1 h at 27°C. 2. Add 0.1 ml of the p0 viral stock to the plated monolayer and incubate at 27°C for five days. 3. Aspirate the conditioned medium from the flask, and remove the cell debris by centrifugation at 500 g for 5 min. Store the supernatant at 4°C in the dark. This is should be considered passage one (p1) viral stock. 4. Plate 2 x 107 Sf9 cells with 25 ml complete TNM-FH in each of several 150 cm2 tissue culture flasks, and allow to attach for 1 h at 27°C. 5. Add 20 ul of the pi viral stock to each of the plated monolayers and incubate at 27°C for 5 days. Harvest this p2 viral stock as described in step 3 above.

4. Recombinant protein expression and purification 4.1 Assessing a recombinant virus Culturing large quantities of Sf9 cells is time-consuming, and the medium is costly. If milligram quantities of recombinant protein are required, early identification of a low-yield virus, and the construction of a second vector will prove time- and cost-effective. It should be noted that not all proteins accumulate in infected Sf9 cells to high levels. The seven transmembrane-domain receptors are notorious for low levels of expression. For example, the B2-adrenoceptor is expressed at relatively low levels and represents only 0.01% of the membrane protein (4). Thankfully, many of the soluble proteins studied in cell signalling will express at between 1 and 5% of total protein. 156

7: Baculoviral expression Assess both the time course and level of recombinant protein expression before proceeding. Several 100 mm dishes of Sf9 cells should be infected and harvested at time points ranging from 40 to 96 h post-infection. A suitable method for the infection of Sf9 cells plated as monolayers is given in Protocol 5. Recombinant protein expression can be analysed after SDS-PAGF. and staining with Coomassie Brilliant Blue R250. With 30 ug of lysate, proteins expressing at 5% of total will be visualized as a prominent stained band (Figure 2). Viruses constructed with empty transfer plasmid are suitable as non-expressing viral controls. Visualization may not be possible if the expression level is considerably lower than 2% of total. This can be resolved cither by immunoprceipitation with selective antibody, or batch chromatography on a relatively small scale before SDS-PAGE. Proteins encoding hexahistidine or glutathione S-transferase motifs can be purified rapidly with Ni/NTA- or glutathtone-agarose, respectively prior to analysis. The functionality of recombinant protein may alter with the progression of infection also. This may represent the effect either of aggregation or altered post-lranslational modification. Thus, the expressed protein should also be tested tor enzymatic activity or ligand binding. It should now be possible to determine if sufficient recombinant material will be produced, and at what point after infection the St'9 cells should be harvested. Reconstructing the baculovirus should be considered if recombinant protein expression levels are inadequate. Re-examine the 5'untranslated sequence introduced with the foreign coding sequence, shorten any lengthy tracts of untranslated sequence, and possibly replace it with the sequence found in

Figure2. Expression of recombinant turkey phospholipase C-B (PLC-Bt). Sf9 cells were infected with baculoviruses expressing PLC-bt under the control of either the polh or cor promoters. Cells were lysed at either 48 or 72 h after infection. Whole cell lysates were analysed by SDS-PAGE and stained with Coomassie Brilliant Blue R-250. The migration of 200, 116.5, 67, and 43 kDa molecular weight markers is indicated. 157

Andrew Paterson

Figures. Translation initiation consensus in AcMNPV. The translation initiation codon is indicated in bold.

known, high-yield viruses. The sequence surrounding the translation initiation codon can also be altered. The AcMNPV genome has been sequenced, and the viral 5'untranslated regions have some striking features. From these it is possible to recognise a consensus in the sequence surrounding the translation initiation codon. Guanosine is absent from the seven bases preceding the initiating AUG, whilst adenosine is conserved at the -3 position, and pyrimidines at the -1 and +5 sites. This is summarized in Figure 3. Another possibility is that the protein is sensitive to degradation. Several proteases are encoded by AcMNPV and will be expressed during progression of the viral infection. These proteolytic enzymes are not essential for progression of the viral infection. An attractive possibility would be an AcMNPV genome modified to exclude protease-coding sequences. A suitable proteasenull baculoviral DNA preparation is currently marketed by Novagen (Bac Vector 3000). Protocol 5. Infection of Sf9 cells as monolayers Equipment and reagents • 100 mm tissue culture dishes (Corning # • 430167) . Disposable cell scrapers (Greiner # 541070) . Refrigerated microfuge .MBS (20 mmol l-1 Mes-NaOH, pH 6.2 . MBS (20 mmol I 1 Mes-NaOH PH 6.2; 140 mmol l~' NaCI; 40 mmol I KCI)

Triton X-100 lysis butter (50 mmol I"1 Trisacetate, pH 7.5; 1% (w/v) Triton X-100; 270 mmol I-1 sucrose; 2 mmol l-1 EGTA; 2 mmol l-1 EDTA; 0.1% (v/v) pmercaptoethanol; 200 umol l-1 PMSF; 1 F mmo| ,-,benzamidine)

Method 1. Plate 2 x 107 Sf9 cells with 10 ml complete TNM-FH in a 100 mm tissue culture dish. Allow to attach for 1 h at 27°C. 2. Inoculate the medium with 1 ml p2 viral stock. 3. Incubate at 27°C for 48-96 h in a humidified incubator. 4. Aspirate the medium, and wash the monolayer with 10 ml chilled MBS. 5. Aspirate all liquid remaining, and add 1 ml chilled Triton X-100 lysis

158

7: Baculoviral expression buffer. Scrape the monolayer into the buffer with the disposable cell scraper, and hold on ice for 10 min. 6. Clarify the lysate by centrifugation at 12000 g for 5 min at 4°C in a refrigerated microfuge. 7. Transfer the supernatant to a fresh microfuge tube. 8. Analyse 30-50 ug of clarified lysate by Coomassie Brilliant Blue R250 staining and/or immunoblotting after SDS-PAGE.

4.2 Scaling-up infection of Sf9 cells as monolayers The method given in Protocol 5 is adapted readily to monolayer cultures of any size. Circular tissue culture dishes are available in diameters ranging from 35-150 mm, representing from 9-170 cm2 of monolayer. This can be increased to 600 cm2 with 245 X 245 mm square dishes (NUNC # 166508), and to 1700 cm2 with disposable, sterile roller bottles (NUNC # 430852). Monolayers should be plated for infection at a density of 1.8 X 105 cells cm-2 and 1.0 ml complete TNM-FH per 106 cells. After incubation at 27°C for 1 h, the medium is aspirated and the cells infected with 0.1 ml of p2 or p3 viral stock per 106 cells, in a total volume of 0.25 ml complete TNM-FH per 106 cells. The monolayers are incubated at 27°C for an additional hour, and TNM-FH added to a final volume of 1.0 ml per 106 cells. Return the monolayers to a humidified 27 °C incubator until harvest. Baculovirus-infected monolayers often detach, especially as infection progresses. Sufficient cells may detach to warrant their recovery. Collect the detached cells by centrifugation at 500 g for 5 rnin, wash with MBS (see Protocol 5), and lyse by resuspending in Triton X-100 lysis buffer (see Protocol 5) to 5 mg ml-1 protein. Alternatively, the cells remaining adherent can be detached and collected with those already in suspension. This is best done by placing the tissue culture dish on a level surface and tapping it sideways repeatedly. The majority of cells will suspend within a few taps.

4.3 Infecting suspension cultures Monolayer culture is expensive. Both the disposable plasticware and serumsupplemented medium impose substantial costs. Suspension cultures of Sf9 cells adapted to serum-free medium can be grown and infected at a fraction of the cost. The processes of infection and harvest are also simplified. Any decrease in the levels of expression that may be experienced with suspension culture is offset by the ease with which far greater numbers of cells are grown, infected, and harvested. It is common practice to ensure synchronous infection by inoculating suspension cultures at high cell density. Cells are collected by centrifugation, resuspended to 1 X 107 per ml in fresh medium, and then inoculated with high-titre viral stock. Infection is allowed to proceed for 1 h, and the cells 159

Andrew Paterson reseeded with fresh medium at 1-2 X 106 per ml in spinner flasks. This multiple step procedure may be accompanied by an increased risk of contamination. Protocol 6 describes a simplified method. Any reduction in expression levels due to the non-synchronous infection is not significant, and can be avoided by harvesting cells from 54-72 h post-infection. Protocol 6. Infection of Sf9 cells in suspension cultures Equipment and reagents • Complete ExCell 401 (ExCell 401 supple• 500 ml microcarrier spinner flasks (Bellco mented with 100 ug ml-1 streptomycin, 100 #1965-00500) sterilized by autoclaving at IU ml-1 penicillin, and 0.25 u9 m1-1 121°C, 15 p.s.i., for 20 min fungizone) . Non-heating magnetic stirrer • Sf9 cells adapted to complete ExCell 401 • High-titre recombinant virus (virus ampli(see Section 2.6). fied as described in Protocol 4)

Method 1. Dilute the Sf9 cells to 1.5 x 106 per ml in complete ExCell 401. 2. Inoculate a 500 ml microcarrier spinner flask with 300 ml of the diluted suspension. 3. Add 22.5 ml high-titre virus to each flask. 4. Stir the suspension at 110-120 r.p.m. in a 27°C incubator. If the incubator is humidified, the side-arm caps can be loosened to allow gaseous exchange. 5. Harvest the cells 48-84 h post-infection.

4.4 Harvesting suspension cultures Collect the culture by centrifugation at 500 g for 5 min in a refrigerated centrifuge, wash the cell pellet with 5-10 volumes of ice-cold MBS (see Protocol 5), and resuspend the cells to 2-5 mg ml-1 protein in Triton X-100 lysis buffer (see Protocol 5). Incubate the lysate on ice for 20 min, and remove the insoluble material by centrifugation at 20 000 g for a minimum of 15 min at 4°C. If required, the low-speed supernatant can be further prepared by ultracentrifugation at 100000 g for 1 h at 4°C.

4.5 Hypotonic lysis Lysing infected Sf9 cells with Triton X-100 is rapid, convenient, and compatible with immunoprecipitation, immobilized metal anion chromatography (IMAC), and glutathione-agarose affinity chromatography. Unfortunately, this does not allow preparation of subcellular fractions, nor the separation of cytosolic from Triton X-100-soluble protein. Hypotonic lysis is reasonably fast and efficient. It also has the advantage 160

7: Baculoviral expression that crude nuclear, plasma membrane-enriched, and soluble protein fractions can be recovered, each in the absence of detergent. The method can be adapted for use on monolayers. Monolayers should first be detached as described in Section 4.2, the resuspended cells collected, and lysed as described in Protocol 7. Protocol 7. Hypotonic lysis of suspension cultures Equipment and reagents • Glass Dounce homogenizer with loosefitting pestle (Wheaton # 357546) . MBS (20 mmol l-1 Mes, pH 6.2; 140 mmol NaCI; 40 mmol l-1 KCI)

• Hypotonic lysis buffer (20 mmolI-1Tris-HCI, pH 7.4; 2 mmol l-1 EGTA; 5 mmol I-1 MgCI2; l-1 10 mmol I-1 dithiothreitol; 1 mmol l-1 benzamidine; 200 (imol I 1 PMSF)

Method 1. Collect infected cells by centrifugation at 500 g for 5 min at 4°C. The rotor should come to rest with the brake set at zero. 2. Wash the cells with 5-10 volumes of ice-cold MBS, and centrifuge as described in Step! above. 3. Resuspend the washed cell pellet in 20 volumes of hypotonic lysis buffer. 4. Incubate on ice for 10 min. 5. Homogenize the suspension with 10 strokes in a Dounce homogenizer. 6. Remove the cell ghosts and nuclei by centrifugation at 500 g for 10 min at 4°C with the brake set at zero. 7. Clarify the low-speed supernatant by ultracentrifugation at 100000 g for 1 h at 4°C. 8. Transfer the high-speed supernatant to a fresh tube. Snap-freeze in liquid nitrogen, and store as a crude cytosolic fraction. 9. Resuspend the high-speed pellet with the Dounce homogenizer to 5 mg ml-1 protein in hypotonic lysis buffer. Snap-freeze in liquid nitrogen, and store as a plasma membrane-enriched fraction.

4.6 Nitrogen cavitation Nitrogen cavitation is described in Protocol 8. Unlike hypotonic lysis, which favours production of cell ghosts, nitrogen cavitation maximizes production of membranes. This is extremely important when expressing membraneassociated proteins, such as the heterotrimeric G proteins, or the seven transmembrane receptors. Lysis by cavitation is complete. No cell ghosts remain and there is no significant release of DNA. The method also has the advantage of dispersing aggregated soluble proteins that formerly were 161

Andrew Paterson recovered in the whole cell/nuclear particulate upon hypotonic lysis. The major disadvantage of this method is gaining access to a nitrogen cavitation cell. Protocol 8.

Nitrogen cavitation

Equipment and reagents • Nitrogen cavitation cell (# 4635 or 4635, Parr Instrument Company), regulator, and compressed nitrogen gas cylinder . MBS (20 mmol I-1 Mes, pH 6.2; 140 mmol NaCI; 40 mmol l-1 KCI)

• Cavitation buffer (50 mmol l-1 Tris-acetate, pH 7.5; 150 mmol I-1 NaCI; 5 mmol l-1 MgCI2; 0.1% v/v p-mercaptoethanol; 1 mmol I-1 l-1 bezamidine; 200 umol I-1 PMSF; 5 M-9 ml-1 leupeptin)

Method 1. Collect and wash the cells in MBS as described in Protocol 7, Steps 1 and 2. 2. Resuspend the washed cell pellet in cavitation buffer 1 x 107 cells ml-1. 3. Place the cooled, cavitation cell on a magnetic stirring plate, and add the cell suspension. 4. Maintain the suspension with slow stirring, and reassemble the cavitation cell according to the manufacturer's instructions. 5. Pressurize with compressed nitrogen to 500 p.s.i. 6. Incubate the cavitation cell for 30 min at 4°C with continued stirring. Introduce further compressed nitrogen should the chamber pressure fall below 500 p.s.i. 7. Discharge the cell suspension according to the manufacturer's instructions. Maintain chamber pressure at 500 p.s.i. during discharge 8. Prepare cell ghost/nuclear, crude cytosolic, and particulate fractions as described in Protocol 7, Steps 6-9.

4.7 Purification of recombinant protein Purification should proceed by protocols adapted from those employed previously with the non-recombinant protein counterpart. Provided the level of expression is adequate, recombinant protein should be purified to near homogeneity within an abridged protocol of two or three chromatographic separations. Both anion- and cation-exchange chromatographies are extremely powerful. This is enhanced by their modern fast-flow supports. More selective media, such as hydroxyapatite chromatography, should not be overlooked and can provide remarkable purification, particularly as a final step. The relative abundance of recombinant protein may merit use of affinity chromat162

7: Baculoviral expression ography at an stage earlier than would be reasonable with non-recombinant protein. For instance, ATP-, affinity dye-, and heparin-agaroses should be considered with recombinant kinases, and heparin-agarose with phospholipase C isoenzymes. Other fractionation methods should not be disregarded. For example, partial purification by ammonium sulfate precipitation can be extremely effective and can be completed in a fraction of the time taken with many chromatographic procedures. It is easily applied to large volumes and can provide substantial purification and concentration of the working material. However, it is restricted to use with soluble protein and is not compatible with detergent. Purification of membrane-associated recombinant protein should proceed along similar lines. The protein must first be solubilized with detergent. Most chromatographic separations are not affected by the presence of detergents at their working concentrations. The only caveat to this is the complete solubilization of the protein prior to chromatography. Again, protocols for purification of non-recombinant protein will indicate suitable conditions. These will indicate which class of detergent and the necessary information on detergent and protein concentrations. Some refinement may still be required. Varying both final detergent concentration and detergent-protein ratio are advised. The high-speed particulate fraction collected after hypotonic lysis of infected Sf9 cells (Protocol 7, Step 9) is enriched in plasma membrane and is an ideal source of membrane-associated protein. Unfortunately, the yield of plasma membrane is extremely low, as most is retained with the cell ghost. This can be avoided by nitrogen cavitation of infected cells (see Protocol 8). Cell disruption is virtually complete and the resulting particulate fraction has been employed successfully for purification of G protein a- and p-y-subunits (2, 9). Monitor complete solubilization by failure to sediment recombinant material after ultracentrifugation at 100000 g for 1 h at 4°C. Release of recombinant enzyme activity from the particulate fraction should be monitored also.

4.8 Rapid purification with glutathione- or Ni2+/NTAagaroses Introducing a novel sequence designed to aid purification on affinity matrices is a common approach. Glutathione S-transferase (GST) fusion, hexahistidinetagged, and epitope-tagged proteins all have been employed successfully. They minimize the number of chromatographic separations required and this, in turn, increases the final yield of purified protein. Several transfer vectors are suitable for construction of baculoviruses encoding GST fusion proteins (see Table 2). In these vectors, coding sequences are sub-cloned directly into the mcs regions positioned downstream of, and in-frame with, the GST sequence. Sub-cloning 'in frame' with GST is 163

Andrew Paterson facilitated by their availability in three frameshift variants. Purification of GST fusion protein on glutathione-agarose is described in Protocol 9. The methodology is compatible with most non-denaturing detergents, and tolerates a minimum of 1 mol l-1 NaCl. Protocol 9.

Purification of GST-fusion proteins

Equipment and reagents • Reduced glutathione-Sepharose 48 (Amersham Pharmacia Biotech # 17-0756-01) . Disposable chromatography column (BioRad # 731-1550 or 732-1010) umol • Clarified cell lysate containing GST-tagged protein

• Wash buffer (50 mmol l-1 Tris-HCI, pH 7.5; 0.1 mmol I-1 EGTA; 0.1% (v/v) p-mercaptoethanol; 1 mmol I-1 benzamidine; 200 I-1 PMSF) *

Method 1. Wash reduced glutathione-Sepharose 4B into wash buffer, and resuspend to 50% (v/v). 2. Add 2 ml 50% glutathione-Sepharose 4B to every 50 ml lysate, and mix end-over-end at 4°C for 1 h. 3. Pellet the glutathione-Sepharose 4B by centrifugation (500 g, no brake, 5min, 4°C). 4. Aspirate the supernatant, resuspend the glutathione-Sepharose 4B in 2 volumes wash buffer, and pack in a disposable chromatography column. 5. If required, fit a frit or adaptor to the column. 6. Wash the resin with 20 bed-volumes 500 mmol I-1 NaCl in wash buffer. 7. Wash with 10 bed-volumes of wash buffer. 8. Elute the GST-fusion protein with 20 mmol I-1 reduced glutathione pH 8.0, 150 mmol I-1 NaCl, in wash buffer. Collect the eluate in fractions equivalent to approximately 0.3 bed-volumes, and identify the fractions containing protein. Pool these, and store appropriately. aWash buffer can be adapted to suit the requirements of individual proteins. For example, nondenaturing detergents can be included.

Transfer vectors allowing construction of baculoviruses encoding hexahistidine-tagged proteins are available also (see Table 2). Again, the coding sequence is sub-cloned downstream of, and in frame with, an aminoterminal hexahistidine sequence. Transfer plasmids are available for both cotransfection and baculoviral shuttle vector technologies. Hexahistidinetagged proteins are purified by IMAC with Ni2+-charged nitrilotriacetic acid 164

7: Baculoviral expression (Ni2+/NTA) supports (see Protocol 10). IMAC with Ni2+/NTA-agarose is compatible most non-denaturing detergents, and with NaCl up to 1 mol l-1. Hexahistidine-tagging is not limited to commercially available fusion vectors. The motif is small and easy to insert into cDNA. Annealed, complimentary oligonucleotides encoding hexahistidine can be inserted at a convenient restriction endonuclease site within a coding sequence (25). Alternatively, hexahistidine can be introduced by PCR amplification with primers encoding either an amino- or carboxy-terminal hexahistidine extension. With both these approaches the modified cDNA is sub-cloned into a non-tagged transfer vector, e.g. pVL1392, or pFastBACl. Purification of tagged proteins either by glutathione- or Ni2+/NTA-agarose chromatography is fast and extremely effective, but further purification may still be required. Contaminants will copurify on both glutathione- and Ni 2+ / NTA-agarose. Often, the contaminating protein represents less than 5% of the total protein present and can be disregarded. If required, adequate further purification may be possible with a single anion-exchange separation (9). Proteins encoding both GST and hexahistidine motifs and purified by sequential chromatography on glutathione- and Ni2+/NTA-agarose offer an alternative solution. Transfer vectors encoding fusion proteins with amino-GST and hexahistidine sequences, both at the amino-terminus, are available from Pharmingen (pAcGHLT A/B/C). Protocol 10. Purification of hexahistidine-tagged protein Equipment and reagents • Disposable chromatography column (BioRad # 731-1550 or 732-1010) . Clarified cell lysate containing hexahistidinetagged protein . Ni2+/NTA-agarose (Qiagen # 30210)

• Wash buffer (50 mmol I-1 Tris-HCI, pH 7.5; 0.1 mmol l-1 EGTA; 0.07% (v/v) Bmercaptoethanol; 1 mmol I-1 benzamidine; 200 mmol I-1 PMSF) *

Method 1. Wash NTA-agarose into wash buffer, and resuspend to 50% v/v. 2. Add 2 ml 50% NTA-agarose to every 50 ml lysate, and mix end-overend at 4°C for 1 h. 3. Pellet the NTA-agarose by centrifugation (500 g, no brake, 5 min, 4°C). 4. Aspirate the supernatant, resuspend the NTA-agarose in 2 volumes of wash buffer, and pack in a disposable chromatography column. 5. If required, fit a frit or adaptor to the column. 6. Wash resin with 20 bed-volumes of 20 mmol l-1 imidazole pH 7.5, 500 mmol l-1 NaCl in wash buffer. 7. Wash with 10 bed-volumes wash buffer. 165

Andrew Paterson Protocol 10.

Continued

8. Elute hexahistidine-tagged protein with 150 mmol l-1 imidazole pH 7.5, 150 mmol I-1 NaCI in wash buffer. Collect eluate in 0.3 bed-volume fractions. 9. Identify fractions containing protein by Bradford assay and/or SDSPAGE analysis. 10. Pool fractions containing protein, and store as appropriate. aWash buffer can be adapted to suit the requirements of individual proteins. For example, nondenaturing detergents can be included. Compatibility with Ni2+/NTA-agarose should be checked with the manufacturer's instructions.

4.9 Alternative cell lines The Sf21 cell line (Sf21AE, Invitrogen # B821-01) should be considered. It is the parental line from which the Sf9 clone was derived, and remains larger and less regular in size. The Sf21 cell is equally applicable to the methodology described for Sf9 cells, although their size variation may create irregularities in plaque assays. However, they are often favoured for expression studies, due to increased recombinant protein yields and ease of cultivation in large quantities. The BTI-TN-5BI-4 Trichoplusia ni (T. ni) cell line ('High 5' cells, Invitrogen # B855-01) has also been reported by many workers to provide higher recombinant protein yields than Sf9 cells. They are a much larger cell line and will grow with either spherical or fibroblast-like phenotype. Primarily they are grown as monolayers in serum-free medium (ExCell 400 or ExCell 405 from JRH Biosciences), but can be adapted to suspension culture according to the instructions provided by Invitrogen. Although these cells are derived from the cabbage looper (T. ni) they are susceptible to infection by AcMNPV and do not require construction of special viral vectors.

4.10 Protein complexes Multimeric proteins can be expressed and assembled in Sf9 cells from their component subunits. Baculoviruses, each encoding individual components, are employed simultaneously. Expression of the holomeric complex must be optimized. Preliminary experiments titrating the MOI for each virus are essential. Possibly the best examples are heterotrimeric G proteins. Triple infection of Sf9 cells with three different viruses encoding G protein a-, B-, and ysubunits, respectively, results in expression of functional heterotrimer (2). As with other proteins, hexahistidine-tagging allows rapid purification. This can be utilized to provide preparations of pure dissociated subunits. Tagging the 7-subunit, and treating the heterotrimeric complex with AlF4 -/GDP/Mg2+ 166

7: Baculoviral expression whilst still bound to Ni2+/NTA-agarose, selectively elutes the recombinant asubunit (9). Conversely, similar treatment of a heterotrimeric complex bound through a hexahistidine-tagged a-subunit will selectively elute the Byheterodimer (9). An interesting variation on this approach was simultaneous infection of Sf9 cells with three different viruses encoding c-Raf, v-ras, and src kinase (3). Synergistic elevation of c-Raf activity was dependent upon expression of both v-ras and src kinase. This is similar to transfection of NIH3T3 cells, where mammalian expression vectors encoding v-ras and src kinase promote similar synergistic activation of c-Raf.

4.11 Alternative promoters The p6.9 basic core-associated protein promoter, pcor, has been used effectively for the production of high activity PKCE (16). Unlike the polh promoter, the cor promoter is active during the late phase of viral infection (see Section 3.1). It is also weaker than the polh promoter (see Figure 2). Both these factors are significant. Considerable post-translational modification is still possible during the late phase, whilst the lower levels of protein expression may prevent any aggregation and inactivation seen with polh-expressed protein. Baculoviral vectors driving expression from the cor promoter are constructed with pAcMP2 and pAcMPS (see Table 2). These plasmids are compatible with linearized and Bsu36I-cut AcMNPV DNA. The plO promoter also is available as an alternative to ppolh. It is active during both the late and very late phases of viral infection, and drives expression to similar high levels. Its main use, however, is as a second promoter when two proteins are being expressed from a single virus. For instance, the transfer plasmid pAcUW51 (Pharmingen) will construct a virus expressing one protein from the polh and a second from the p10 promoter.

4.12 Further scale-up of culture volume The procedure described in Section 4.3 for growing baculovirus-infected Sf9 cells in microcarrier spinner flasks has been optimized for conditions of passive aeration. Increasing the volume of cell suspension in 500 ml flasks above 300 ml may reduce both the rate of growth and level of recombinant protein expression. Reduction in growth rates and cell viability are experienced also with increasing spinner flask size in the absence of active sparging with oxygen. Oxygen air-lift fermenters have been employed successfully (26). Unfortunately these systems are extremely expensive and require considerable operator attention. A cost-effective alternative is large, stirred culture vessels used in conjunction with serum-free medium and sparged with oxygen. Shear stress is minimized by the lower rates of stirring required in a sparged vessel. 167

Andrew Paterson Additionally, modern serum-free media (ExCell 401, ExCell 420 for example) are not prone to excessive foaming, and include surfactant to reduce shear stress. A reasonably inexpensive system can be adapted from a three-litre microcarrier spinner flask, and stirred on a standard magnetic stirrer (27). Otherwise, dedicated systems with direct impeller drive, and dissolved O2 and pH monitoring, are available commercially from Bellco in 8-36 litre formats

(26).

References 1. Luckow, V. A., and Summers, M. D. (1988). BloTechnology, 6,47. 2. Hepler, J. R., Kozasa, T., Smrcka, A. V., Simon, M. I., Rhee, S. G., Sternweis, P. C., and Gilman, A. G. (1993). J. Biol. Chem., 268,14367. 3. Williams, N. G., Roberts, T. M., and Li, P. (1992). Proc. Natl. Acad. Sci. USA, 89, 2922. 4. Kobilka, B. (1995). Anal. Biochem., 231, 269. 5. Parker, E. M., Kameyama, K., Higashijima, T., and Ross, E. M. (1991). J. Biol. Chem., 266,519. 6. Hurwitx, D. R., Emanuel, S. L., Nathan, M. H., Sarver, N., Ullrich, A., Felder, S., Lax, I., and Schlessinger, J. (1991). J. Biol. Chem., 266,22035. 7. Wedegaertner, P. B., and Gill, G. N. (1989). J. Biol. Chem., 264,11346. 8. Graber, S. G., Figler, R. A., and Garrison, J. C. (1992). J. Biol. Chem., 267, 1271. 9. Kosaza, T., and Oilman, A.G. (1995). J. Biol. Chem., 270,1734. 10. Iniguez-Lluhi, J. A., Simon, M. I., Robishaw, J. D., and Oilman, A. G. (1992). J. Biol. Chem., 267, 23409. 11. Tang, W. J., Krupinski, J., and Gilman, A. G. (1991). J. Biol. Chem., 266,8595. 12. Paterson, A., Boyer, J. L., Watts, V. J., Morris, A. J., Price, E. M., and Harden, T. K. (1995). Cellular Signalling 7,709. 13. Jones, G. A., and Horstman, D. A. (1997). In Signalling by inositides: a practical approach, (ed. S. Shears), p. 69. IRL Press, Oxford. 14. Sonnenburg, W. K., Seger, D., Kwak, K. S., Huang, J., Charbonneau, H., and Beavo, J. A. (1995). J. Biol. Chem., 270, 30989. 15. Burns, D. J., Bloomenthal, J., Lee, M. H., and Bell, R. M. (1990). J. Biol. Chem., 265,12044. 16. Rankl, N. B., Rice, J. W., Gurganus, T. M., Barbee, J. L., and Burns, D. J. (1994). Prot. Exp. Purif., 5, 346. 17. Morrison, D., Kaplan, D. R., Rhee, S. G., and Williams, L. T. (1990). Mol Cell. Biol., 10,2359. 18. Woscholski, R., Dhand, R., Fry, M. J., Waterfield, M. D., and Parker, P. J. (1994). J. Biol. Chem., 269, 25067. 19. Kim, C. M., Dion, S. B., Onorato, J. J., and Benovic, J. L. (1993). Receptor, 3, 39. 20. Walker, K. S., Deak, M., Paterson, A., Hudson, K., Cohen, P., and Alessi, D. R. (1998). Biochem. J., 331,299. 21. O'Reilly, D. R., Miller, L. K., and Luckow, V. A. (1992). Baculovirus expression vectors: a laboratory manual. W.H. Freeman, New York. 168

7: Baculoviral expression 22. Richardson, C. D. (ed.) (1995). Methods in molecular biology. Vol. 39 Baculovirus Expression Protocols. Humana Press, Totowa, NJ. 23. Kitts, P. A., and Possee, R. D. (1993). BioTechniques, 14, 810. 24. Luckow, V. A., Lee, S. C., Barry, G. F., and Olins, P. O. (1993). J. Virol, 67, 4566. 25. Hepler, J. R., Biddlecome, G. H., Kleuss, C., Camp, L. A., Hofmann, S. L., Ross, E. M., and Oilman, A. G. (1996). J. Biol. Chem., 271,496. 26. Rice, J. W., Rankl, N. B., Gurganis, T. M., Marr, C. M., Barna, J. B., Walters, M. M., and Burns, D. J. (1993). BioTechniques, 15,1052. 27. Graber, S. G., Lindorfer, M. A., and Garrison, J. C. (1996). In G-proteins (ed. P. C. Roche), Methods in Neuroscience, 29,207. Academic Press, San Diego.

169

This page intentionally left blank

8

Reporter gene systems for the study of G protein-coupled receptor signal transduction in mammalian cells STEPHEN REES, SUSAN BROWN and JENNY STABLES

1. Introduction Traditional protocols for the measurement of G protein-coupled receptor (GPCR) signal transduction in mammalian cells have relied upon a number of biochemical techniques. These include the measurement of the rate of guanine nucleotide exchange at the activated G-protein alpha subunit (1), the measurement of the changes in the level of one of a plethora of intracellular second messenger metabolites, such as cAMP, calcium, or inositol phosphates (2), or the activation or inhibition of an ion channel (3). In recent years these assays have been supplemented by the development of reporter gene systems for the study of GPCR signal transduction (4-5). In this chapter we shall describe a number of such reporter gene systems, we shall outline the advantages of reporter genes over standard biochemical analyses, and we shall describe a number of protocols developed for the study of GPCR signal transduction using reporter gene systems.

2. What is a reporter gene? A reporter gene is the general term used to describe a DNA sequence in which the expression of a readily measurable enzyme or other protein is under the transcriptional control of a natural or synthetic promoter element (Figures 1 and 2). The transcriptional activity of the promoter element is regulated as a direct consequence of the activation of a specific signal transduction cascade following agonist binding to a GPCR at the cell membrane (5). Such inducible reporter genes are inactive or weakly active with respect to transcription until the transcription-factor proteins that bind to and activate the promoter are themselves activated as the consequence of the signaltransduction event. The activated promoter then drives the transcription of a reporter gene, which, following translation, is quantified by measuring the

Stephen Rees et al.

Figure 1. Structure of a 6CRE-luciferase reporter gene, A synthetic promoter consisting of six copies of the consensus cAMP response element (CREJ (bold) was constructed as described in Protocol 7. This was placed upstream of a minimal HSV-TK promoter (26) to create a synthetic cAMP-responsive promoter, which has been used to control the expression of a number of reporter genes, including firefly luciferase.

Figure2. The mechanism of activation of the c-fos promoter as a result of GPCR signalling through specific G protein classes. The 750 nucleottdes upstream of the transcription initiation site of the c-fos promoter contain a number of transcription-factor binding sites, including a cAMP response element (CRE), a TPA response element (TRE; AP-1 binding site), a serum response element (SRE), and a sis-inctucible enhancer (SIE). AP-1, activating protein 1; CREB, CRE-binding protein; PK-A, protein kinase A; PK-C, protein kinase C; STAT, signal transducers and activators of transcription; TATA, TATA box; TCP, ternary complex factor.

enzymatic activity of the protein product. The reporter protein has a unique activity or structure, to enable it to be distinguished from other proteins present within the cell. The choice of reporter protein, which is often but not always an en/yme, is influenced primarily by the availability of a simple, usually non-radioactive, assay of the activity of the protein product. Further factors influencing the choice of reporter protein include the availability of 172

8: Reporter gene systems suitable assay apparatus, the cost of the assay reagent, the sensitivity of the reporter assay, the stability of the assay signal (the half-life of the signal), and the simplicity of the assay (6, 7). Reporter-gene assays have been used for the study of GPCR signal transduction, and have revealed novel aspects of GPCR signalling (4,5). However, the most widespread use of reporter genes to examine GPCR signalling has been within the pharmaceutical industry, for the screening of novel compounds for agonist or antagonist activity at this class of receptor (8). Reporter-gene assays can be used in miniaturized plate formats (96-well and 384-well microplates) for the rapid assessment of the activity of many thousand, or hundred thousand, novel compounds at a GPCR. Many widely used reporter genes encode enzymes for which simple nonradioactive assay procedures have been developed. At present, the most commonly used enzymes include firefly (Photinus pyralis) luciferase (9), Renilla reniformis (sea pansy) luciferase (10), secreted placental alkaline phosphatase (SEAP) (11), and B-galactosidase (12). In addition to the enzymatic reporter proteins, the Aequorea victoria bioluminescent proteins aequorin and Green Fluorescent Protein (GFP) have been used as reporter proteins (13,14). The calcium-sensitive photoprotein aequorin has been expressed in mammalian cells using the constitutive cytomegalovirus (CMV) immediate/early promoter element as a reporter of calcium signalling (15). In contrast to all other reporter proteins, GFP fluorescence is non-enzymatic, and requires no additional cofactors (13). Upon excitation with blue light at a Xmax of 396 nm, the protein emits green light with a nmax of 509 nm. GFP has been widely used as a marker of protein expression (13), to follow intracellular movement of proteins (16), and in the construction of FRET (fluorescence resonance energy transfer) assays (17). Assay systems for many of these enzymes, as well as vectors and mammalian cell lines containing the reporter genes, are available from a number of molecular biology reagent suppliers. A comprehensive review of the assay protocols available is beyond the scope of this chapter. However we will provide protocols for the use of a number of commonly used reporter enzymes.

2.1 Construction of a reporter gene A reporter gene typically consists of three elements: a promoter which is responsive to the signal-transduction cascade under study, the cDNA encoding the reporter enzyme, and transcription-termination sequences such as the polyadenylation sequence from the SV40 small t intron (Figure 1). This is contained within a plasmid suitable for maintenance in E. coli and transfection into mammalian cells (5). The choice of reporter enzyme is largely the preference of the investigator. However, it is the choice of the promoter element, within the reporter gene construct that defines the specificity of the signalling event to be detected by the reporter gene. Two types of promoter 173

Stephen Rees et al. have been used in reporter gene experiments: natural promoters, and synthetic promoters. 2.1.1 Natural promoter elements All promoter sequences contain a number of DNA motifs, termed transcription factor-binding sites or hormone response elements, required for the correct temporal and spatial control of protein expression (18, 19). The level and site of protein expression is determined by the binding to the promoter element of a plethora of transcription factors, and the interaction of these transcription factors with the RNA polymerase and other proteins that bind to the transcription initiation site. Reporter gene assays were first used in experiments designed to characterize the control of gene transcription by extracellular stimuli. In such experiments, a reporter gene such as firefly luciferase was expressed off a natural promoter element, and the ability of extracellular stimuli to activate or inhibit the promoter element was studied using luciferase luminescence as the read-out (for examples see 20, 21). Such studies led to the identification of promoter elements regulated by specific stimuli, to the identification of the DNA sequence motifs required for promoter activity, and to the identification of the proteins, or transcription factors, which were able to bind to these DNA sequence motifs (termed transcription factor-binding sites or hormone response elements). One such promoter that has been extensively characterized is the promoter for the immediate early gene c-fos (22). A number of studies have shown that, in the 750 nucleotides upstream of the transcription initiation site, this promoter contains a cAMP response element (CRE), a TPA response element (TRE or AP-1 site), a serum response element (SRE), and a serum-inducible element (SIE) (Figure 2). This promoter is activated by transcription factor binding to the SIE following activation of the JAK/STAT signal transduction pathway, by transcription factor binding to the SRE following activation of the MAP kinase cascade, by AP-1 transcription factor binding to the TRE element following protein kinase C activation, and by CREB binding to the CRE element following PK-A activation (Figure 2). As such, the c-fos promoter can be induced following agonist binding to many cell-surface receptors, including GPCRs (22), cytokine receptors, and growth factor receptors. A number of other natural promoters have been used in reporter gene constructs to detect receptor signalling, including the ICAM-1 promoter (23, 24). 2.1.2 Synthetic promoter elements Natural promoter elements often contain many different transcription factor binding sites, and will respond to multiple signalling events. To increase the specificity of the promoter and to increase the signal-to-noise ratio obtained in a reporter gene assay, usually through the lowering of basal promoter activity, a number of synthetic promoter elements have been generated. Synthetic promoters usually consist of multiple copies of a specific transcription factor 174

8: Reporter gene systems binding site placed upstream of a minimal mammalian promoter element, to control reporter enzyme expression (25) (Figure 1). The minimal mammalian promoter element is transcriptionally silent, and contains only the RNA polymerase binding site (the CAAT and TATA boxes), and the transcription initiation site. Sequences derived from the promoter elements of the Herpes Simplex Virus Thymidine Kinase (HSV-TK) promoter (25), the Adenovirus Ela promoter (26), and the human interferon-p promoter (27) have been used as minimal promoter elements in reporter gene studies. Consensus transcription factor binding sites have been identified for many transcription factors, including the CRE to monitor PK-A activation (28), the TRE to detect PK-C activity (29), and the SRE to detect MAP kinase activity (30) (Table 1). The number of response elements, and the precise position of the response elements with respect to each other and to the minimal promoter, is derived from empirical study to optimize the signal and signal-to-noise ratio obtained within the reporter assay, and usually multiple copies of the response element are required (4). For example, to create a reporter gene for the study of PK-A activity, as a result of GPCR regulation of cAMP, we have built synthetic promoters containing 3, 6, 12, or 24 copies of the consensus CRE, placed upstream of the minimal HSV-TK promoter (25) (Figure 1). This promoter was used to control the expression of the firefly luciferase or the SEAP reporter enzymes. The activity of these reporter genes was characterized by examining the ability of the agonist CGRP (calcitonin gene-related peptide) to activate the Gas-coupled CGRP receptor, endogenously expressed in SK-N-MC cells, to promote luciferase expression. As seen in Figure 3, stimulation of these cells with CGRP results in a tenfold stimulation of luciferase activity only in cells transfected with the reporter plasmid containing six copies of the CRE. In further studies across a number of cell lines, we failed to observe any stimulation of the reporter genes containing 12 or 24 copies of the CRE, and occasionally observed a weak stimulation of the reporter gene containing three copies of the CRE. In these studies we routinely observed a 5 to 20-fold stimulation of luciferase activity with the reporter plasmid containing six Table 1. Transcription factor binding sites (hormone response elements) used in the construction of GPCR reporter genes Response element

Transcription factor

CRE (cAMP response element) TRE (TPA response element) SRE (serum response element) NFAT (nuclear factor activator of transcription) Yeast Gal4c

CREB AP-1 complex EIk-1, Sap1ab NFAT Gal4

Gai

Gai

Receptora

Reference

GaiandGas Gaq and Gaq Gaq

41 49 30 31

and Gaq

26

' Reporter can be used to study receptors that activate these G protein classes. bOther ternary complex factor transcription factors can also bind to the SRE. cUsed with chimeric transcription factors to study MAP kinase signalling (see Section 3.3).

175

Stephen Rees et al.

Figures. CGRP receptor activation of CRE-luciferase reporter genes. Synthetic cAMPresponsive promoters containing 3, 6, 12, or 24 copies of the consensus cAMP response element were placed upstream of the minimal HSV-TK promoter to control the expression of the firefly luciferase reporter gene. The reporter vectors were transiently transfected into SK-N-MC cells, together with a CMV-SEAP internal control vector, and the ability of the agonist CGRP to activate the cAMP-responsive reporter genes, as a consequence of activation of the Gas-coupled CGRP receptor endogenously expressed in this cell line, was determined. Luciferase and SEAP values were determined for each transfection. To control for transfection efficiency, the luciferase values were normalized according to the SEAP measurements.

copies of the CRE (Figure 3). The reason for this is unclear, but it may be due to secondary structure within the synthetic promoter. This illustrates the need to build a range of synthetic promoters to identify a reporter gene with suitable characteristics. A number of vectors containing minimal promoters, and synthetic inducible promoters, are now available from molecular biology reagent suppliers. A generic protocol for the construction of a synthetic promoter element containing a reporter gene construct is presented in Protocol 1. Protocol 1. Construction of a 6CRE-luciferase reporter gene Equipment Oligonucleotide synthesizer

• PCR machine • DMA sequencing apparatus

Method 1. A synthetic 123 bp Oligonucleotide, 5'-CCAGAAGCCTACGTAGGC GTCGACCTCCTTGGCTGACGTCAGTAGAGAGATCCCATTTGACGTCAT

176

8: Reporter gene systems ACTGAGACGTAGATCTCCATTGACGTCAAGGAGACICQAQGCTCCAT CGCAGTGATCG-3', containing three copies of the consensus cAMP response element (CRE, bold) and Sail and Xho\ restriction sites (underlined) was used as the template in a PCR reaction using the oligonucleotide primers: sense 5'-CCAGAAGCCTACGTAGGCGTC-3'; antisense 5'-CGATCACTGCGATGGAGCCTC-3'). PCR was performed in a reaction volume of 50 ul containing 20 mM Tris-HCI, pH 8.2; 10 mM KCI; 6 mM (NH4)2SO4; 2 mM MgCI2; 1% (v/v) Triton X-100; 10 ug ml-1 BSA; 50 pmol of each PCR primer;10 pmol of the template primer; 200 uM of each dNTP; and 2.5 units of native Pfu DMA polymerase (Stratagene). PCR conditions were 92°C for 30 s (denaturation), 55°C for 1 min (annealing), and 72°C for 1 min (extension) for 25 cycles.3 2. The PCR product was gel purified, then digested with restriction enzymes Sail and Xhol, and ligated directly into similarly digested pBluescript (Stratagene).b 3. The 3CRE sequence was verified by DNA sequencing. 4. The 3CRE element was excised by digestion with Sail and Xhol, and cloned into the unique Sail site of pTK-SEAP (25), and screened for orientation such that the Sail site is regenerated at the 5' end of the 3CRE element to generate p3CRE-SEAP. 5. To generate a 6CRE element, the 3CRE sequence was again excised from pBluescript following digestion with Sail and Xhol, and religated into Sa/l-digested p3CRE-SEAP to create p6CRE-SEAP. 6. To generate p6CRE-luciferase, the firefly luciferase cDNA was excised from the plasmid pGL3-Basic (Promega) following digestion with restriction enzymes Hindill and Xbal, and inserted into similarly digested p6CRE-SEAP to replace SEAP with firefly luciferase (Figure 7). a

As an alternative to the PCR strategy, the CRE element can be contained within a single oligonucleotide capable of insertion into a Sall restriction site (TCAGCTGACTGCAG). The oligonucleotide is phosphorylated using standard molecular biology methods (32), and ligated into Sail-digested pTK-SEAP under conditions that will promote multiple insertion of the oligonucleotide sequence. b All molecular biology protocols are performed according to ref. 32.

2.2 Reporter proteins In contrast to reporter genes, a reporter protein is defined as a protein which is expressed in mammalian cells, from a constitutive promoter element such as the CMV immediate/early promoter/enhancer, to provide a direct measurement of the changes in the intracellular environment (15). A limited number of reporter proteins have been described. The Aequorea victoria photoprotein aequorin has been used for many years as a reporter of changes in intracellular calcium concentration in mammalian cells or Xenopus oocytes (33).

177

Stephen Rees et al. Aequorin is a 21-kDa photoprotein that forms a bioluminescent complex when linked to the chromophore cofactor coelenterazine (34). Following the binding of calcium to this complex, an oxidation reaction of coelenterazine results in the production of apoaequorin, coelenteramide, CO2, and light, with a nmax for emission of 470 nm. Loading of cells has traditionally involved microinjection of purified aequorin protein, which has limited the usefulness of this reporter system. In recent years, cloning of the aequorin cDNA has allowed expression of this protein, both transiently and stably, in a range of cell types, and has greatly expanded the utility of aequorin as a reporter. The increase in cytoplasmic calcium concentration following agonist binding at a GPCR can be detected by the generation of aequorin luminescence in mammalian cells constitutively expressing this reporter (15). To expand the range of uses of aequorin, a number of modified aequorins have been constructed, in which expression of the protein is targeted to particular cellular compartments to measure calcium changes within those compartments (reviewed in ref. 35). This includes aequorin targeted to the mitochondria (36), nucleus (37), and endoplasmic reticulum (38). Furthermore, the construction of fusion proteins with aequorin has facilitated the analysis of local calcium changes. A calcium sensitive adenylyl cyclase-aequorin fusion protein has been used to report the changes in intracellular calcium concentration that regulate this enzyme (39). The Aequorea victoria photoprotein Green Fluorescent Protein (GFP) has also been used as a reporter protein in mammalian cells (see ref. 40 for review). Uses of GFP are summarized in Section 7.2.

3. Reporter gene systems for the study of GPCR signal trans duction 3.1 Reporter genes for GPCRs which couple to members of the Gas and Gai G protein families Agonist binding to a GPCR which couples to the stimulatory G protein alpha subunit, Gas, or to members of the inhibitory G protein alpha subunit family, Gai/o, will result in the respective activation or inhibition of adenylyl cyclase to cause a change in the level of intracellular cAMP (41). An increase in the intracellular concentration of cAMP results in the activation of protein kinase A (PK-A) to cause an increase in the phosphorylation state of members of the CREB (cAMP response element binding protein) transcription factor family. Once phosphorylated, CREB can recognize and bind to its cognate response element, the CRE (TGACGTCA), to cause an increase in the level of expression of genes containing the CRE within their regulatory sequences (Figure 4). For the study of receptors which couple to the inhibitory Gai/o G protein family, agonist activity results in a decrease in the level of intracellular cAMP to cause a decrease in the activation state of CREB family transcrip178

8: Reporter gene systems

Figure4. Reporter gene system for the detection of GPCR signalling through a Gas- or a Goti/o-family G protein. Agonist activation of a Gas-coupled GPCR results in an elevation of the level of intracellular cAMP, as a consequence of Gas activation of adenylyl cyclase. The increase in intracellular cAMP will result in the activation of protein kinase A (PK-A), with a subsequent increase in the phosphorylation state of CREB (cAMP response element binding protein) family transcription factors. Once phosphorylated, CREB is able to cause an increase in transcription of genes containing a CRE (cAMP response element) within their promoter sequence. If the cell contains a cAMP-responsive reporter gene, this will lead to an accumulation of reporter gene product. In the case of firefly luciferase, reporter activity is assayed following cell lysis, to generate a luminescent signal that is directly related to the efficacy of agonist. The cAMP reporter system may also be used to detect GPCR activation of a Gai/o-coupled GPCR. In this case the assay relies on the detection of agonist-mediated inhibition of forskolin-stimulated cAMP accumulation.

tion factors (41). In a reporter gene assay, this is usually studied by examining the ability of the receptor agonist to cause an inhibition of forskolinstimulated CRE promoter activity. Two types of reporter gene have been developed to monitor receptor regulation of cAMP. The first involves the use of a natural promoter such as the c-fos promoter, which contains a CRE element within the promoter sequence (42) (Figure 2). The second involves the construction of synthetic promoters containing multiple copies of the CRE motif (41, 43-46 for examples) (Figure 1). In either case, the promoter is used to control the expression of a readily detectable reporter protein such as firefly luciferase. For example, signal transduction through the Gai-coupled 5-HT1B serotonin 179

Stephen Rees et al. Table 2. pEC50 estimates for agonist activity at the human Cannabinoid CB1 receptor determined using a 6CRE-SEAP reporter gene Cannabinoid

pEC50a

SEMb

plC50 (cAMP assay)c

Hu210 CP55940 +WIN 55212 (-)-A9-THC Methanandamide Anandamide Merck Frosst Indole no 9 Merck Frosst Indole no 13

10.09 9.94 7.55 6.98 6.63 >5.00 >5.00 >5.00

0.01 0.08 0.09 0.12 0.33 -

9.71 8.73 7.62 7.78 nd 6.49 nd nd

a

n= minimum of three in a CRE-SEAP reporter gene assay. Standard error of the mean (CRE-SEAP assay). Taken from ref. 47.

b c

receptor and the Gcts-coupled calcitonin Cla receptors, endogenously expressed in CHO cells, has been studied following stable transfection into this cell line of a reporter gene, containing the firefly luciferase reporter under the control of a promoter containing six copies of the consensus cAMP response element (41). We have used CRE-luciferase and CRE-SEAP reporter genes to characterize a number of GPCRs, including the Gai-coupled human Cannabinoid CBj receptor (Figures 5 and 6). In this study, the CB1 receptor was stably coexpressed in CHO cells, together with a 6CRE-SEAP reporter construct. The ability of a series of agonists to mediate an inhibition of forskolin-stimulated reporter gene activity was determined (Figure 5). As can be seen from Table 2, the IC50 values obtained for each of these agonists agree with values obtained in a traditional cAMP accumulation assay (47). Furthermore, increasing concentrations of the antagonist SR141716A progressively shift the Hu 210 agonist concentration response curve with a pKB of 8.85, again in agreement with previous studies (47) (Figure 6). These studies demonstrate that potency and affinity estimates can be determined using a reporter-gene assay, and that these values agree with values obtained from traditional second-messenger analyses. In our experience, a series of agonists will have the same rank order of potency in a reporter gene assay and a biochemical second-messenger assay. However, we have observed that the absolute potencies obtained in a reporter gene assay may be different from the values obtained in other assay formats, due to signal amplification obtained when using a reporter gene readout.

3.2 Reporter genes for GPCRs which couple to members of the Gaq/11 G protein family Agonist binding at a GPCR which couples to a Gaq/11 family G protein results in the activation of the phosphoinositidases of the phospholipase CB class 180

8: Reporter gene systems

Figure5. Use of a cAMP reporter system to assess agonist activity at the Gai-coupled human cannabinoid CB1 receptor. A stable CHO cell line was created expressing both the CB1 receptor and a 6CRE-SEAP reporter gene. Following treatment with forskolin, agonist concentration response curves were constructed for (A) HU210, (D) +Win 55212, (•) anandamide, (A) Delta-9-THC,
Figure 6. Use of a cAMP reporter system to assess antagonist activity at the Gai-coupled human cannabinoid CB1 receptor. A stable CHO cell line was created expressing both the CB1 receptor and a 6CRE-SEAP reporter gene. Following treatment with forskolin, agonist concentration response curves were constructed to Hu 210. The ability of increasing concentrations of the antagonist SR 144528 to shift the agonist concentration response curve was determined. (x), (O)and («•), Hu 210 alone; (A) Hu 210 + 10 nM SR 144528; (•) Hu 210 + 100 nM SR 144528; (D) Hu 210 + 1 pM SR 144528.

181

Stephen flees et al. (PLCp) to catalyse the formation of the second messenger metabolites sn 1-2 diacylglycerol (DAG) and inositol (1, 4, 5)-trisphosphate (IP3). This is followed by the release of calcium from intracellular stores, and the activation of protein kinase C (PK-C) (Figure 7; 48). Reporter genes, or reporter proteins, have been developed to monitor both changes in the level of intracellular calcium and the activation of PK-C. An increase in PK-C activity causes a subsequent increase in the activity of the AP-1 transcription factor complex as a consequence of phosphorylation of c-fos by PK-C. The activated AP-1 transcription factor can activate any gene containing the cognate recognition sequence within the promoter element. As before, two types of reporter gene have been developed, the first involving the use of a natural promoter element, such as the c-fos promoter or the promoter for ICAM-1, which contain AP-1 response elements and are thus responsive to PK-C activation (23, 24). The second involves the construction of synthetic PK-C responsive promoters containing multiple copies of the AP-1 response

Figure?. Reporter gene systems for the detection of GPCR signalling through a Gaq/11family G protein. Agonist binding promotes G protein activation of phospholipase Cp (PLCp), to cause the hydrolysis of phosphatidylinositol (4,5)-biphosphate (PIP2), and to result in the generation of inositol (1,4,5)-trisphosphate (IP3) and diacylglycerol (DAG). DAG will cause an increase in the activity of protein kinase C (PK-C) to result in the phosphorylation and activation of fos family transcription factors, to cause the subsequent activation of reporter genes containing a TRE (AP-1 site) within their promoter sequence (such as the human ICAM-1 promoter). The reporter-gene product, such as firefly luciferase, can be assayed following cell lysis. In addition, the increase in intracellular levels of IP3 will lead to an increase in the levels of intracellular calcium. This can be detected using the bioluminescent reporter protein aequorin.

182

8: Reporter gene systems element (also called the TRE: 0-tetradecanoylphorbol 13-acetate response element) linked to a minimal promoter element as described in Section 2.1.1 (4,49). In either case, the responsive promoter is used to promote expression of a reporter protein, such as firefly luciferase or SEAP. Such reporter vectors have been used to characterize agonist activation of a number of GPCRs, including the serotonin 5-HT2A receptor and the tachykinin NK2 receptor (23,24). The increase in cytoplasmic calcium concentration seen following agonist binding at a GPCR can be detected by the generation of aequorin luminescence in mammalian cells constitutively expressing this reporter (Figure 7; 15). We have expressed aequorin in mammalian cells to examine the signalling of a number of Gaq/11 GPCRs, including the human angiotensin II and endothelin A receptors. Stable cell lines expressing recombinant GPCR have been transiently transfected with mitochondrially targeted aequorin (15). In this study, a concentration-response curve constructed for angiotensin II activity at the AT\ receptor generated an ECso of 100 ± 69 nM. A similar concentration-response curve for endothelin 1 activity at the ETA receptor generated an EC50 of 59 ± 13 nM (15). We have also created stable cell lines expressing both the human oxytocin receptor and cytoplasmic expressed aequorin. Concentration response curves for the agonists oxytocin, vasopressin, and desmopressin generated EC50 values of 12 nM, 82 nM, and 418 nM respectively (Figure 8).

Figures. Use of aequorin to detect agonist activity at the Gaq-coupled human oxytocin receptor. CHO cells stably expressing aequorin and the human oxytocin receptor were constructed. Concentration response curves to the agonists oxytocin (solid line), vasopressin (dashed line), and desmopressin (dotted line) were constructed using aequorin luminescence as the read-out. The EC50 values obtained are 12 nM (oxytocin), 82 nM (vasopressin), and 418 nM (desmopressin).

183

Stephen Rees et al.

Figure 9. Use of aequorin to detect agonist activity at the human ML1A melatonin receptor following coexpression of this receptor with the G protein sub unit Ga16. CHO cells, and CHO cells stably expressing Ga16, were transiently transfected with aequorin and the human ML1A melatonin receptor. Concentration response curves for melatonin-evoked light emission were constructed in CHO (O, broken line) and CHO/G16 (•, solid line) cells. Data were pooled from three experiments, each performed in triplicate, and expressed as Relative Light Units (RLU) (taken from ref. 15).

A number of reports demonstrate the use of aequorin to report agonist activation of a histamine receptor (34), the V1A vasopressin receptor (50), and the a1-adrenoceptor (50). Furthermore, aequorin has been coexpressed with the G protein alpha subunit, Ga16, to generate a generic screening system for agonist activity at a number of GPCRs (15). The Gaq/11-family G protein Ga16 is capable of functional interaction with a wide range of GPCRs. Following coexpression of Ga16 and aequorin, agonist activation of the normally Gai coupled human melatonin ML1A and adenosine A1 receptors and the normally Gas-coupled human B2-adrenoceptor and adenosine A2a receptors was demonstrated to promote functional interaction with Ga16 to generate an increase in aequorin luminescence (15) (Figure 9).

3.3 Reporter genes for G-protein p-y signalling Stimulation of many GPCRs causes a rapid elevation in the activity of a family of closely related serine-threonine kinases, known as the mitogenactivated protein kinases (MAPKs) (Figure 10) (51). There are at least three sub-families of MAP kinase enzymes. The extracellular-signal regulated kinases (ERK1 and ERK2) are activated in response to mitogenic stimuli. In contrast, the family of c-Jun N-terminal kinases (JNKs), also called stress-activated protein kinases (SAPKs), and the more recently identified HOG/p38 kinase, are activated in response to cellular stress (51, 52). Activation of 184

8: Reporter gene systems

Figure 10. Reporter gene systems for the detection of GPCR activation of MAP kinase. Agonist activity of many receptors that normally couple to members of the Gaq or Got, G protein families has been reported to result in the activation of a number of signalling pathways, including the Ras/Raf MAP kinase cascade, as a consequence of the activity of the G protein By complex. G protein p-y complex activation of Ras involves an uncharacterized tyrosine kinase (TK), possibly src kinase, and involves the adapter protein Grb2 and the guanine nucleotide exchange factor SOS. Activation of Ras/Raf will ultimately lead to an increase in the activity of ERK1/ERK2 kinases (also known as p42/p44 MAP kinase) (see Section 3.3). Once activated, these kinases are able to phosphorylate a number of transcription factor substrates, including the ternary complex factor family members Elk-1 and Sapla, to cause an increase in their activity. The activated transcription factor will cause an increase in the expression of genes containing the appropriate transcription factor binding site within their promoter sequences. The site of action of the inhibitors pertussis toxin (P. Tox), which inactivates Garfamily G proteins; transducin, which is a G protein p-y scavenger; and PD 98059 (MEK inhibitor), are indicated.

ERK1 and ERK2 is a consequence of agonist activity at many GPCRs, which couple to members of the inhibitory G proteins, Gai1-3 and Gao. This includes the u, K, and 8 opioid receptors, the ORL1 receptor, and the lysophosphatidic acid (LPA) receptor (53-56), and is thought to result from the regulation of the Ras/Raf pathway by the G protein By complex (52) (Figure 10). Similarly, a number of reports demonstrate activation of both the ERK and JNK kinase cascades by the muscarinic, thrombin, endothelin, and angiotensin receptors, all of which couple primarily to members of the Gaq/11 G protein family (51). In contrast to ERK activation by Gai-coupled GPCRs, JNK activation is thought to be mediated through both the G protein a subunit and the By complex, to result in activation of the small G proteins Rac and Rho. Stimulation of these kinases results in the phosphorylation and activation of 185

Stephen Rees et al. a range of transcription factors, to cause an increase in the expression of genes containing specific transcription-factor binding sites within their promoter elements. These include the ternary complex factor (TCF) transcription factors Elk-1, which is a substrate for ERK1/ERK2 and JNK. (57), Sapla which is a substrate for E R K 1 / E R K 2 (58), and the JNK substrate, c-Jun (59). Mammalian transcription factors such as Elk-1 and Sapla consist of an Nterminal DNA-binding domain and a C-terminal regulatory and transcription activation domain. The regulatory domain contains a number of sites for phosphorylation by the upstream kinase that is activated following the stimulation of a cell-surface receptor. The DNA binding and transcription activation domain is able to recognize, and bind to, the transcription factor binding site, to result in the regulation of any promoter element which contains that sequence (60). As such, it is possible to construct chimeric transcription factors consisting of the regulatory domain of one transcription factor fused to the DNA-binding domain of another. The Saecharomyees cerevisiae

Figure 11. The Gal4/Elk-1 MAP kinase reporter gene system, (A) The ternary complex factor transcription factor Elk-1 consists of four domains. The A domain contains an Nterminal Ets domain which contains the DNA binding sequence. The B domain is responsible for complex formation with the serum response factor at the c-fos SRE. The D domain contains the nuclear localization sequence, and the C domain is the transcriptional activation domain containing the sites for phosphorylation by MEK kinase. The Gal4/Elk-1 chimeric transcription factor was generated by replacing the A domain of Elk-1 with the DNA-binding domain of yeast Gal4 (amino-acids 1-147} to create a chimeric protein that is regulated by MEK but is only able to bind to the Gal4 response element. (B) Structure of the Gal4 responsive reporter gene. Five copies of the yeast Gal4 binding site were placed upstream of a minimal adenovirus Ela promoter element. Binding of Gal4/Elk-1 to the synthetic promoter will cause the expression of firefly luciferase.

186

8: Reporter gene systems Table 3. Chimeric transcription factors used as reporter genes for the study of MAP kinase signalling Reportera

Regulatory

Gal4/Sap1a Gal4/Elk-1

83-431 83-428

Gal4/c-Jun Gal4/ATF-2 Gal4/CREB Gal4/Fos

1-223 1-96 1-283 208-313

domainb

Kinasec

ERK1/ERK2 ERK1, ERK2, JNK, p38 kinase JNK JNK, p38 kinase PK-A FRKa

Receptord Gai, Gaq Gai, Gaq, Ga12, Ga13 Gaq Gai, Gaq Gas, Gai Gaq

Reference 62 58 Stratagene Stratagene Stratagene Stratagene

a All reporter constructs consist of the DMA binding domain of yeast Gal4 (amino acids 1-147) fused to the regulatory domain of the appropriate mammalian transcription factor. b Describes the amino acids of the mammalian transcription factor fused to yeast Gal4. c Describes the upstream kinases responsible for phosphorylation and activation of the mammalian transcription factor. d Reporter can be used to study receptors that activate these G protein classes. e Fos-regulated kinase.

transcription factor Gal4 activates two genes involved in galactose metabolism, as a consequence of binding to a specific transcription factor binding site termed the Gal4 upstream activating sequence (UAS) (61) (Figure 11). To facilitate the study of MAP kinase signalling, several groups have constructed chimeric transcription factors consisting of the regulatory domain of the mammalian transcription factor linked to the DNA-binding and transcription activation domains of the yeast Gal4 protein (Table 3). For example, the Gal4/Elk-l chimera consists of amino acids 1-147 of Gal4, which encompasses the DNA-binding domain, fused to the regulatory domain of Elk-1 (amino acids 83-428) (62) (Figure 12). When expressed in mammalian cells, the regulatory domain derived from the mammalian transcription factor becomes multiply phosphorylated in response to upstream kinase activity. The chimeric protein is now able to activate a reporter construct containing multiple copies of the Gal4 UAS within its promoter element, to promote the expression of a reporter enzyme such as firefly luciferase (Figure 11). Gal4based reporter gene assays have been developed to report the activity of many transcription factors, including Elk-1, Sapla, c-Jun, and ATF-2, and are commercially available from Stratagene (Table 3). We have used reporter gene assays to characterize the regulation of MAP kinase by a number of GPCRs that couple to members of the Gai and Gaq/11 G protein families. In initial studies of MAP kinase activation, CHO cells stably expressing the ORL1 receptor were transiently transfected with three reporter genes, a fos-luciferase reporter gene (22), the Gal4/Elk-l chimeric transcription factor together with a second plasmid containing firefly luciferase under the transcriptional control of a Gal4 responsive promoter (62), and third, the Gal4/Sapla chimeric transcription factor together with a second plasmid containing firefly luciferase under the transcriptional control of a 187

Stephen Rees et al.

Figure 12. ORL1 receptor activation of reporter constructs. CHO cells stably expressing the ORL1 receptor were transiently transfected with a fos-luciferase reporter gene (white bars), the Gal4/Elk-1 reporter gene system (a Gal4/Elk-1 expression plasmid together with a plasmid carrying 5xGal4-luciferase), (black bars), and the Gal4/Sap1a reporter gene system (a Gal4/Sap1a expression plasmid together with a plasmid carrying 5xGal4luciferase) (grey bars). Reporter gene activity was measured following treatment of the cells with 10 uM nociceptin, or 10 uM nociceptin in the presence of 50 ng ml-1 pertussis toxin (PTox), or the MEK inhibitor PD 98059 (50 uM) (PD). All values are the mean counts per second (CPS) of at least three experiments performed in duplicate.

Gal4 responsive promoter (58). The ORL1 agonist nociceptin evoked a concentration-related increase in the activity of all three reporter genes (Figure 12). However, the chimeric transcription factor-based reporter systems exhibited a lower basal signal and a higher signal-to-noise ratio within the assay than the fos-luciferase reporter (Figure 12). This illustrates the advantage of these reporter systems over a natural promoter containing reporter gene. In further studies, nociceptin was shown to promote a concentration-related increase in Gal4/Elk-l and Gal4/Sapla reporter gene activity with an EC50 of 0.49 (range 0.15-1.63) nM for the activation of Gal4/Elk-l (Figure 13). This value agrees with the EC50 for nociceptin regulation of MAP kinase determined in a Western blotting-gel shift assay (55), and with an IC50 of 0.15 nM for nociceptin-mediated inhibition of forskolin-stimulated cAMP (N. Bevan, personal communication; 63). In these experiments, nociceptin activation of the Gal4/Elk-l reporter was inhibited following pretreatment of cells with pertussis toxin, the MEK inhibitor PD 98059, and following coexpression of ot-transducin. This indicates that nociceptin activation of Gal4/Elk-l is a consequence of the release of the G protein By complex following receptor activation of a Gai-family G protein (63). These observations confirmed earlier studies performed using traditional biochemical assays (53). 188

8: Reporter gene systems

Figure 13. Concentration response curves for nociceptin activation of Gal4/Elk-1 and Gal4/Sap1a reporter genes in CHO cells stably expressing the ORL1 receptor. Cells were transiently cotransfected with the reporter vectors p5xGal4/luciferase and one of (A) pSG/Gal4/Elk-1 or (B) pSG/Gal4/Sap1a. Concentration response curves to nociceptin are shown in the absence (•) and the presence (A) of pretreatment for 18 h with 50 ng ml-1 pertussis toxin. All values are the mean counts per second (CPS) of at least three experiments performed in duplicate, ± SEM. (Reproduced from reference 63)

4. Factors influencing the design of a mammalian cell reporter-gene assay A number of factors need to be considered in the design of a reporter gene experiment. This includes the choice of host cell line, the choice of expression vector and expression protocol, and the choice of reporter enzyme. Reportergene assays require cells expressing both the receptor of interest and a suitable reporter, the choice of which is entirely dependent upon the second messenger activity under study.

4.1 Choice of cell line Reporter-gene assays can be performed in any mammalian cell line, providing a suitable high efficiency transfection procedure is available. We routinely use Chinese Hamster Ovary (CHO) cells in reporter-gene assays. CHO cells are easy to transfect to high levels using a range of transfection protocols, the cells are robust, and they are easy to grow. These cells tolerate high-level expression of a wide range of GPCRs, and endogenous GPCR expression in CHO cells is well defined (64). We have also performed reporter-gene assays in a number of other cell lines, including HeLa, Swiss 3T3, and HEK 293 189

Stephen Rees et al. (endogenous GPCR expression in these and other cell lines is detailed in ref. 64). In these studies it is necessary to transfect the host cells stably or transiently with plasmids encoding both the GPCR and an appropriate reporter gene. In the absence of a cDNA clone for the GPCR under study, it is possible to perform reporter-gene assays following the transfection of a plasmid encoding a reporter gene into a cell line which endogenously expresses the receptor of interest. The endogenously expressed Cla calcitonin and 5-HT1B serotonin receptors on CHO cells have been studied using a CRE-luciferase reporter gene (41). We have also studied the endogenous CGRP receptor in SK-N-MC neuroblastoma cells, following stable transfection of this cell line using a CRE-lucif erase reporter gene (Figure 3). Reporter genes have also been used to study receptor signalling in primary cultures. The signalling of the B2-adrenoceptor was characterized following transfection of a CRE-luciferase reporter gene into primary airway epithelial cells (Ian Hall, University of Nottingham, personal communication). For each cell line it is usually necessary to optimize both transfection and reporter-gene assay protocols to obtain a robust signal in the assay.

4.2 Choice of expression vectors A variety of reporter vectors have been described in the literature, and many are available from suppliers of molecular biology reagents. In transient transfection studies the plasmid is required to contain only the reporter-gene sequence. However, for the creation of a stable cell line, we routinely use vectors containing both the reporter gene, and a second mammalian expression cassette containing the gene for an antibiotic selection marker.

4.3 Choice of expression protocol A number of experimental paradigms may be used in a reporter-gene assay. Transient expression protocols are used for experiments to optimize a reporter vector prior to stable transfection, to investigate which signal transduction pathways are regulated by a GPCR, or to examine the activity of a limited number of compounds at a GPCR. Transient expression experiments can involve cotransfection of receptor and reporter gene into a host cell line, the transfection of a receptor into a cell line containing a reporter gene, or the transfection of a reporter gene into a cell line either endogenously or recombinantly expressing a specific GPCR. While transient expression experiments do not require the lengthy creation of a stable cell line, such experiments are labour-intensive and involve the use of large quantities of cells, plasmid DNA, and transfection reagent. Furthermore, the sensitivity of the assay and the signal-to-noise ratio within the assay are often lower than in stable transfection experiments. Stable cell lines expressing both receptor and reporter provide a stable reagent for the analysis of receptor signalling. This is particularly useful if large numbers of reporter gene assays are to be performed, such as within 190

8: Reporter gene systems compound-screening activities in the pharmaceutical industry (8). We have used three strategies for the development of stable cell lines. In the first, a reporter gene is stably transfected into a cell line, either stably or endogenously expressing the receptor. This is the preferred method if the receptor is uncloned or difficult to express. In the second a 'host' reporter cell line is created and screened for low basal reporter-gene activity, and for a robust response to a known inducer of the reporter gene system. In the third strategy, receptor and reporter gene expression plasmids are cotransfected into a host cell line. Following selection for integration of both the receptor and the reporter, clonal cell lines are screened for the desired phenotype. We routinely use the latter two expression strategies. Transfection of a receptor into a cell line capable of robust reporter gene activation will generate cell lines with a high level of reporter protein induction. This can be as high as 25fold greater than the basal level. Cotransfection of receptor and reporter offers the greatest variability. Using this approach, cell lines can be isolated for both the study of inverse agonist and agonist activity. The level of GPCR and reporter gene expression and the degree of inducibility of the reporter gene is mainly governed by the site of integration of the two plasmids into the cell genome. As this process is not controlled, it is often necessary to screen many tens, or even hundreds, of clonal isolates to identify cell lines with a suitable phenotype for assay (Figure 14). 4.3.1 Transient expression systems In transient expression experiments, compound activity is generally assessed within 48 h of transfection. This allows the cells sufficient time to recover from the transfection procedure and to begin to express the receptor. In transient transfection experiments it is common practice to transfect a second reporter vector in addition to the reporter under study. This second vector carries a different reporter gene under the transcriptional control of a constitutive promoter such as the SV40 early promoter or the CMV immediate early promoter enhancer, and is used to control for transfection efficiency across different samples. However, if the investigator has confidence in the reproducibility of transfection, it is not always necessary to include the second reporter. Transient transfections can be performed in any cell culture flask or dish, using one of a range of transfection procedures. In our laboratories we routinely use the Lipofectamine reagent, available from Life Technologies, to transfect CHO cells, due to the very high transfection efficiencies that can be obtained (>80%). However, any standard transfection procedure such as electroporation (65), DEAE-dextran transfection (66), other liposome delivery systems (67), or calcium phosphate transfection will work. While Lipofectamine is a highly efficient reagent for the transfection of CHO cells, this may not be the case for all cell lines, and it may be necessary to perform a series of experiments to optimize transfection efficiency prior to initiating the reporter-gene experiment. In transient expression experiments, it may also be 191

Stephen Rees et al.

Figure 14. The figure illustrates the variation in the basal reporter gene signal, the maximum reporter gene response, and the signal-to-noise ratio observed in different clonal isolates, following the construction of a cell line expressing both a GPCR and a reporter gene. The 6CRE-luciferase reporter gene and the human V2 vasopressin receptor were coexpressed in CHO cells. Following transfection, 19 clonal isolates were generated, and reporter gene responses were determined following treatment with 1p,M vasopressin (dark bars). White bars represent basal reporter gene activity. Diamonds and the right axis represent the signal-to-noise obtained with each clone.

necessary to vary the amount of receptor and reporter-gene expression vector used in the study to optimize the signal obtained in the assay. Protocol 2. Transient transfection of DNA into CHO cells in 96well plates using the Lipofectamine reagent Equipment and reagents • Lipofectamine transfection reagent (Life Technologies) • OptiMEM serum-free medium (Life Technologies)

' 96-well opaque plate with clear bottom i Plasmids to be transfected

Method 1. On the day prior to transfection , trypsinize a confluent flask of CHOa cells, and plate at 1:20 dilution into 96-well plate, aiming for 60% confluence in 24 h (100 u,l cells per well). 2. Return to a C02 incubator for 24 h. 3. Prepare the liposome-DNA transfection mix according to the manu-

192

8: Reporter gene systems facturer's instructions. Briefly; two solutions are prepared. Solution A: for each transfection (each plate), dilute DNA (10 ug) into 1 ml Optimem serum-free medium. Solution B: for each plate, dilute 100 ul Lipofectamine in 1 ml OptiMEM. Combine solutions A and B and incubate at room temperature for 15 min.b 4. During incubation, rinse cells with OptiMEM to remove serum.c 5. Add 8 ml Optimem to the DNA-liposome mix (transfection mix). 6. Add the transfection mix to the plate at 100 ui per well. 7. Incubate the cells at 37°C and 5% CO2 for 5 h. 8. Replace the transfection mix with 100 ul growth medium. 9. Return cells to the incubator for 48 h, then assay. a CHO cells are maintained in Dulbecco's Modified Eagle's Medium/Ham's F12 (50:50) containing 10% fetal calf serum and 2 mM L-glutamine. b Incubation to allow the DNA-liposome complex to form. c Serum may interfere with the transfection procedure.

4.3.2 Stable expression systems Compared with transient transfection assays, pharmacological assays using stable cell lines are less expensive, simpler to perform, and have a lower degree of variability across assays. Following the creation of a stable cell line, cells can be grown in large quantities and aliquoted directly into the assay plate on the day prior to assay (8). Again, a variety of transfection protocols are available, and the choice of protocol may depend on the cell line under study. We routinely express GPCRs and reporter genes stably in CHO cells, and have used the Lipofectamine, calcium phosphate (68), and electroporation (65) transfection protocols for this purpose. We present the protocol for Lipofectamine transfection (Protocol 3). In our laboratory, we routinely transfect cells with receptor expression vectors carrying neomycin selection (69), and reporter gene vectors carrying hygromycin selection. Protocol 3. Stable expression of a receptor and reporter gene in mammalian cells Equipment and reagents • Lipofectamine transfection reagent (Life Technologies) • OptiMEM serum-free medium (Life Technologies)

• Plasmids to be transfected (in suitable stable expression vectors)

Method 1. Plate CHO cells at 1:20 density into a six-well plate in 2 ml normal growth medium.3 Prepare an additional well to act as a selection control.

193

Stephen Rees et al. Protocol 3. Continued 2. Return to a C02 incubator for 24 h. 3. Prepare the liposome-DNA transfection mix according to the manufacturer's instructions. Briefly; two solutions are prepared. Solution A: for each transfection (each well), dilute DMA (1 ug) into 100 ul OptiMEM. Solution B: for each transfection (each well), dilute 10 ul Lipofectamine in 100 ul OptiMEM. Combine solutions A and B, and incubate at room temperature for 15 min.b 4. During incubation, rinse cells with OptiMEM to remove serum.c 5. Add 0.8 ml OptiMEM to the DNA-liposome mix (transfection mix). 6. Add the transfection mix to the plate at 1 ml per well. 7. Incubate the cells at 37°C and 5% C02 for 5 h. 8. Replace the transfection mix with 2 ml normal growth medium. 9. Leave 24 h. 10. Remove medium, and wash cells with PBS. 11. Add trypsin to detach cells. 12. Re-suspend the cells in a small volume of growth medium containing the appropriate selection antibiotic/ Place all cells from a single well into individual 80 cm2 flasks. 13. Change the medium every two days until the cells in the control flask have died. 14. Dilution clone the cells expressing the DNA of interest by placing them at a density of one cell per well into individual wells of a 96-well plate, and leave them to grow to confluence. 15. Expand through passages to grow up until there are sufficient cells to screen colonies for expression of protein, using a suitable reporter assay. a CHO cells are maintained in Dulbecco's Modified Eagle's Medium/Ham's F12 (50:50) containing 10% fetal calf serum and 2 mM L-glutamine. b Incubation allows the DNA-liposome complex to form. c Serum may interfere with transfection procedure. d For transfection of CHO cells we select in 1 mg ml-1 G418 and/or 400 ug ml-1 hygromycin B.

4.3.3 Inducible expression systems It has been demonstrated that expression of some receptors can be toxic resulting in an inability to produce a cell expressing both reporter and receptor. Some receptors can exhibit constitutive activity if expressed recombinantly. This leads to high basal activity which cannot be stimulated further with agonist. In either case, the use of an inducible expression system is recommended for expression of the receptor. A number of inducible expression 194

8: Reporter gene systems systems are commercially available. The ecdysone-inducible expression system (Invitrogen) uses the steroid hormone ecdysone analogue, muristerone A, to activate expression of the gene of interest via a heterodimeric nuclear receptor. Expression is dormant until addition of muristerone A (70). In the case of constitutive activity, the concentration of muristerone A can be titrated until the required receptor expression level is obtained. Similar control of gene expression can be obtained using the tetracycline-regulated expression system available from Clontech (71).

4.4 Choice of reporter gene and reporter enzyme The reporter promoter used in an experiment will be entirely dependent on the nature of the receptor under study (see Section 3). However, any reporter enzyme (firefly luciferase, Renilla luciferase, SEAP, CAT, (3-galactosidase, Blactamase, etc.) can be used in conjunction with any reporter promoter. The choice of the reporter is generally governed by three factors: the sensitivity required in the assay, the cost of the assay reagent, and the availability of specialist apparatus (6). While the bioluminescent or chemiluminescent reporter assays are most sensitive, the assay reagent is expensive and a luminometer will be required for detection. In contrast, the calorimetric SEAP, (3galactosidase, and B-lactamase assay protocols, while less sensitive, are less expensive, and only require a standard spectrophotometer for detection. With reporters such as SEAP and B-galactosidase, both calorimetric and chemiluminescent assay reagents are available (see Section 6) (6, 7).

4.5 Optimization of reporter assay conditions To obtain a robust signal in a reporter-gene assay, a number of parameters have to be optimized. These can vary according to cell type, and the nature of the reporter gene and reporter enzyme. 4.5.1 Quiescence conditions Few reporter genes are transcriptionally silent under normal cell-growth conditions. Serum used in the culture medium contains a wide range of growth factors and other components, many of which are capable of activating cellsurface receptors to cause reporter gene expression. To minimize basal levels of reporter protein expression, it is usually necessary to make the cells quiescent in serum-free or low serum-containing growth medium for a period of time prior to assay. Quiescence conditions need to be optimized for each cell line and each reporter gene. Quiescence conditions can be varied according to the concentration of serum in the medium and the duration of quiescence, which can vary from 2-48 h. For experiments in CHO cells, we treat for 18 h in serum-free normal growth medium. If working with a different cell line a starting point for quiescence would be growth medium supplemented with 0.1% serum for 18 h. 195

Stephen Rees et al. 4.5.2 Signal-to-noise ration in the assay Following construction of a cell line stably expressing a reporter gene construct, it is likely that the clones isolated will exhibit a wide range of both basal and agonist-stimulated levels of reporter expression (Figure 14). When selecting clones for further study, it is important to maximize the signal-tonoise ratio by identifying a clone with low basal expression and a high induction to a known agonist (except when assaying for inverse agonists, see Section 5.3). The signal-to-noise ratio obtained in a reporter assay can sometimes be improved by modifying growth and assay conditions, to maximize the number of viable cells. However, it is often easier to improve the signal by increasing the stringency of quiescence (see Section 4.5.1). 4.5.3 Duration of exposure to drug The time of exposure to the drug may have profound effects on the signal-tonoise ratio, and the need for a large response may have to be balanced against simplicity of assay. All reporter-gene assays require a period of several hours following drug treatment to transcribe and translate measurable quantities of reporter protein. To simplify the assay, it is preferable to leave the drug in contact with the cells during this time. However, chronic exposure of receptor to agonist may lead to down-regulation of receptor or components of the signalling pathway, causing a decrease in signal. Alternatively, there may be a steady increase in basal signal due to non-specific effects of the compound, which is only revealed after prolonged exposure. Should this be a problem, cells can be exposed to drug for a period of 30 min, the drug removed and the cells washed, and fresh medium applied to the cells. The cells are then left for 3-6 h to allow for the expression and accumulation of the reporter enzyme. When evaluating antagonists, the assay should be designed so that the compound is permitted to equilibrate at the receptor prior to addition of agonist. A pre-incubation of 10-30 min is usually sufficient, although this may depend on factors such as lipophilicity of the antagonist. 4.5.4 Necessity for inhibitors In cAMP accumulation assays, it is customary to include phosphodiesterase (PDE) inhibitors such as IBMX (3-isobutyl-l-methylxanthine) to prevent the breakdown of cAMP. In our experience, the reporter gene system is usually sufficiently sensitive to generate a large signal-to-noise ratio in the absence of PDE inhibitors, and their presence may result in the agonist concentrationresponse curve shifting to the left. The need for these inhibitors should be assessed for each cell line. Other inhibitors that may be used to facilitate the study of signal transduction, such as pertussis toxin and wortmannin, are generally added to the cells for between 30 min and 18 h prior to the addition of receptor agonist. 196

8: Reporter gene systems The time of exposure to inhibitor and the inhibitor concentration are the same as would be used in a standard second-messenger assay.

5. Preparation of cells for reporter-gene assays A number of different protocols are available for the assay of drug efficacy using a reporter-gene assay. This is dependent on a variety of factors, including the nature of the signal transduction pathway being assayed, the type of activity being examined (agonist, antagonist, or inverse agonist), and the nature of the reporter protein. We routinely use reporter genes containing firefly luciferase, Renilla luciferase, or SEAP, and have developed the following protocols for the detection of drug efficacy using these reporter enzymes. Protocol 4. Preparation of cells for firefly luciferase, Renilla luciferase, dual luciferase, B-lactamase, and SEAP reporter assays Equipment and reagents • Cells either transiently or stably expressing both a GPCR and a reporter gene (see Protocols 2 and 3) • Trypsin-EDTA (Life Technologies) . PBS

• Black clear-bottomed 96-well plates (Costar) . Growth medium (DMEM-F12 (50:50), 10% PCS, 2 mM glutamine, selection antibiotics if required) . Phenol red-free DMEM-F12 (50:50)a

Method 1. Obtain a 80cm2 flask of confluent CHO cells expressing the GPCRreporter gene combination. 2. Remove growth medium, and wash cell surface with PBS. 3. Add 2 ml trypsin EDTA, and leave cells 5 min at 37°C, 5% C02, to detach. 4. When cells have detached, add 10 ml growth medium. 5. Place cells into reagent reservoir. 6. Using a 12-channel pipette and sterile tips, place 100 ul of cells per well of a black, clear-bottomed 96-well plate. 7. Leave 24 h in an incubator at 37°C, 5% C02, to adhere. 8. Next day, quiesce the cells by replacing the normal growth medium with serum-free phenol red-free growth medium.b 9. Leave 24 h in an incubator at 37°C, 5% C02. 10. Assay reporter gene activity using firefly luciferase, Renilla luciferase, B-lactamase, or SEAP assay (see Section 6). ' Phenol red is a steroid hormone and can induce gene expression when using some reporter gene systems. In addition, phenol red can quench the signal in some chemiluminescent reporter-enzyme assay protocols. b The quiescence conditions will vary with different cell types. C BSA needs to be included in quiescence media for some cell types.

197

Stephen flees et al.

5.1 Agonist assays Reporter assays have been widely used to assess agonist activity (Figures 3, 5, 8, and 13). Protocol 5 should be used to assess agonist activity when such treatment is expected to cause an increase in the level of reporter gene product. Examples include the use of CRE reporter genes to detect agonist activity at a Gas-coupled GPCR, the use of MAP kinase reporter genes to detect agonist activity at Gai and Gaq-coupled GPCRs, and the use of TRE or fospromoter-containing reporter genes to detect agonist activity at Gaq-coupled GPCRs. Agonist activity at receptors which couple to Gai/o-family G proteins leads to an inactivation of adenylyl cyclase. In a cAMP-accumulation assay, this is normally measured as agonist-mediated inhibition of a forskolin-stimulated increase in adenylyl cyclase activity (Figure 5). Reporter-gene assays can be used to detect agonist activity at this class of receptor, as a consequence of agonist inhibition of forskolin-stimulated CRE reporter gene activity. For this purpose Protocol 5 is modified by the inclusion of Step 3. Reporter gene assays can be used to distinguish partial agonists from full agonists. However, in comparison to the assessment of partial agonist activity in a traditional second-messenger assay, the activity of a partial agonist in a reporter assay is usually a much higher percentage of the full agonist response. This is thought to be a consequence of the signal amplification that can be obtained in a reporter-gene assay as a result of measuring changes in gene expression rather than changes in the concentration of a primary second messenger.

Protocol 5. Evaluation of agonist activity Equipment and reagents • CHO cells, expressing the receptor of interest and an appropriate reporter, plated in a 96-well microtitre platea (see Protocols 2, 3, and 4)b

• Test compounds diluted to 10x final required concentration in phenol red-free medium • Forskolin diluted to 100 uMc

Method 1. Remove the medium from the cells, and replace with 40 ul phenol redfree DMEM-F12 (50:10) medium (no additions). 2. Add 5 ul test compounds at 10X the final required concentration. 3. Add 5 ul forskolin at 100 uM (forskolin is only included when studying agonist activity at Gai-coupled GPCRs, using a CRE reporter-gene system).

198

8: Reporter gene systems 4. Return the plate to the incubator for 4 h prior to the firefly luciferase assay, or 6 h prior to the SEAP assay.c a If inhibitor studies are being performed, these compounds should be added to the cells for the appropriate time and at the appropriate concentration prior to initiating this protocol. b Cells plated out 24 h prior to assay. The time of incubation varies according to the choice of reporter protein. The shorter the halflife of the reporter protein, the shorter the time of incubation.

5.2 Antagonist assays As in traditional second messenger assays, antagonist activity is measured in a reporter-gene assay by examining the ability of the antagonist to inhibit an agonist response (Figure 6). The following assay protocol has been developed to study antagonist activity, in CHO cells, at a Gas-coupled GPCR using a CRE reporter gene, at Gai- and Gaq-coupled GPCRs using MAP kinase reporter genes, and at Gaq-coupled receptors using TRE or c-fos promoter containing reporter genes. As in Protocol 5, forskolin treatment is required if assaying a Gai-coupled GPCR using a CRE reporter gene. Protocol 6. Evaluation of antagonist activity Equipment and reagents • Cells, expressing receptor of interest and • Standard agonist diluted to 10x the final appropriate reporter, plated in 96-well required concentration in phenol red-free a microtitre plate (see Protocols 2, 3, and 4) medium • Test antagonist compounds diluted to 10X the final required concentration in phenol red-free mediumb

Method 1. Remove the medium from the cells, and replace with 35 ul phenol redfree DMEM-F12 medium (no additions). 2. Add 5 ul test compounds at 10x the final required concentration. 3. Return to incubator for 10-30 min. 4. Add 5 ul standard agonist compound at 10x the final required concentration. 5. Add 5 ul forskolin at 10 x 10 u-M (forskolin is only used for the study of a Gai-coupled receptor using a CRE reporter gene). 6. Return plate to incubator for 4 h prior to firefly luciferase assay, or 6 h prior to SEAP assay.c ' Cells plated out 24 h prior to assay. b If inhibitor studies are being performed, these compounds should be added to the cells for the appropriate time and at the appropriate concentration prior to initiating this protocol. c The time of incubation varies according to the choice of reporter protein. The shorter the halflife of the reporter protein, the shorter the time of incubation.

199

Stephen Rees et al.

5.3 Constitutive activity and inverse agonist assays The phenomenon of constitutive receptor signalling, that is, GPCR activation of a signal-transduction pathway in the absence of ligand (72, 73), can be studied using reporter genes. Furthermore, reporter-gene assays can be used to characterize inverse agonist ligands. The use of a PK-C-responsive reporter to detect constitutive signalling through the thyrotropin-releasing hormone (TRH) receptor was recently reported (72). In this study, Cos-1 cells were transiently transfected with increasing amounts of a mammalian expression vector encoding the TRH receptor. Increased levels of receptor expression, as determined in a ligand-binding assay, correlated with an increase in basal reporter-gene activity. In contrast, constitutive activity of this receptor was not detected in a parallel study when measuring GPCR-stimulated increases in inositol phosphate stimulation. In our laboratory, we have observed constitutive activation of a CRESEAP reporter gene with a number of GPCRs that couple to the stimulatory G-protein Gas. We have stably expressed the human adenosine A2a receptor in CHO cells, along with a CRE-SEAP reporter gene. In this cell line, the antagonist COS 15943 was able to cause a decrease in basal reporter-gene

Figure 15. Inverse agonism at the human adenosine A2A receptor, detected using a CRESEAP reporter gene. A stable CHO cell line was created expressing both the human adenosine A2A receptor and a 6CRE-SEAP reporter gene. Concentration response curves were constructed for the agonist NECA (5'-/V-ethylcarboxamidoadenosine), and for antagonists 8-sPT, 8-PT, and CGS 15943. All experiments were performed at least three times in the presence of 2 units ml-1 adenosine deaminase.

200

8: Reporter gene systems activity, thus demonstrating this compound to be an inverse agonist (Figure 15). This observation was later confirmed in a cAMP accumulation assay. Inverse agonist assays are performed using Protocol 5.

6. Reporter enzyme assays A large number of enzymes and other proteins have been used in reportergene systems to monitor signal transduction. The choice of reporter enzyme is for the investigator, but will be influenced by factors such as the availability of the cDNA for the reporter enzyme, the cost of the assay protocol, the availability of suitable detection apparatus (e.g., luminometers or fluorimeters), and the sensitivity required in the reporter assay. Plasmids containing reporter enzymes such as firefly luciferase, SEAP, or 3-galactosidase are available from many suppliers of molecular biology reagents, as are assay reagent systems for their detection. In addition, a number of reporter genes, and even mammalian cells stably expressing a reporter gene, are available for purchase from suppliers of molecular biology reagents. In this section we present a number of protocols used in our laboratories for the detection of a number of commonly used reporter enzymes.

6.1 Firefly luciferase Firefly (Photinus pyralis) luciferase is the most commonly used of the bioluminescent reporter enzymes (9). Firefly luciferase is a monomeric enzyme of 61 kDa which catalyses a two-step oxidation reaction to yield light, with an emission maximum in the green to yellow region (550-570 nm). The first step is activation of the luciferyl carboxylate by ATP in the presence of magnesium to yield a reactive mixed anhydride. In the second step, this activated intermediate reacts with oxygen to create a transient dioxetane that breaks down to the oxidized products, oxyluciferin and CO2 (9). On mixing with substrates, firefly luciferase produces an initial burst of light that decays over about 15 s to a low level of sustained luminescence. The 'flash' kinetics of the firefly luciferase reaction have required the use of a luminometer equipped with reagent injectors in order to detect the luminescent signal immediately after addition of the assay reagent. However, in recent years a number of reagent suppliers have developed luciferase assay reagents in which the luminescent signal has been stabilized to generate a signal with a half life in excess of 6 h (74, 75). These assay systems require the single addition of a cell lysis/ luciferase substrate solution to cells expressing the reporter gene. We routinely use the LucLite firefly luciferase assay system available from Packard Biosciences (75), although similar systems are available from Promega, Boehringer Mannheim, and Tropix. Unlike earlier assay protocols, the LucLite protocol does not require any plate-to-plate or tube-to-tube transfers, and the luciferase signal has a half-life in excess of 5 h (75). We have used this reagent in a 201

Stephen Rees et al. number of assay formats, including 96-well and 384-well plates. The sensitivity of this reagent makes it well suited for both research and high throughput screening assays. Luciferase luminescence can be detected in a variety of tube or plate luminometers or scintillation counters. For detection in a microplate format we use the Packard TopCount or the Wallac Microbeta Trilux microplate scintillation and luminescence counters. The main advantages of firefly luciferase are attomolar detection sensitivity, a broad dynamic range of eight orders of magnitude, the development of simple assay methods, and the relatively short half-life of luciferase protein of approximately 4 h (6,7). A reporter enzyme that possesses a long half-life can accumulate within the cell as a consequence of basal promoter activity to produce a high basal signal and a low signal-to-noise ratio in the assay. The two disadvantages of firefly luciferase as a reporter enzyme are the cost of the assay reagent and the requirement for a luminometer. Alternative cheaper assays such as the SEAP assay are available. Protocol 7.

Firefly luciferase assay using the Packard LucLite reagent

Equipment and reagents • Topcount 96-well microplate scintillation and luminescence counter (Packard Biosciences) • Black plate sealers

• LucLite luciferase assay system (Packard Biosciences) . Black 96-well clear-bottomed assay plates (Packard Biosciences)

Method 1. According to the nature of the reporter gene experiment, cells are prepared as described in Protocol 4. Compound addition is performed as described in Protocols 5 or 6. 2. Incubate cells for 4 h at 37°C and 5% CO2 following test compound addition. 3. Dilute each bottle of LucLite substrate with 10 ml of buffer (supplied).3 4. Add 50 ul of reconstituted LucLite substrate into each assay well.b 5. Place a black plate sealer onto the bottom of each plate to produce completely black wells. 6. Place into a Topcount and allow to light adjust in the dark for 2 min.c 7. Read each well for 1 s on single-photon counting mode.d a

One bottle of substrate is enough for two 96-well plates. This makes 100 ul in total (including test compounds). To allow for quenching of plate autoluminescence. d The signal is stable for up to 5 h at room temperature. b c

202

8: Reporter gene systems

6.2 Renilla luciferase and dual luciferase assays Renilla reniformis luciferase is a 31 kDa monomeric enzyme that catalyses the oxidation of coelenterazine to yield coelenteramide and blue light with an emission maximum of 480 nm (10). As a reporter enzyme Renilla luciferase has many of the same features as firefly luciferase, with no particular advantages. Its main limitation is the presence of a low level of nonenzymatic luminescence, termed autoluminescence, which reduces the assay sensitivity. As with firefly luciferase, Renilla luciferase luminescence exhibits 'flash' type kinetics with a signal half-life of minutes. As a consequence, the measurement of Renilla luciferase luminescence has required the use of a luminometer equipped with reagent injectors (10). Together with the high cost of the substrate coelenterazine, this has limited the use of this enzyme. However, within the last year a number of manufacturers have developed Renilla luciferase assay reagents with 'glow' reaction kinetics, in which the signal half-life is extended to 1-2 h, enabling Renilla luciferase luminescence to be detected on a microplate luminometer. Renilla luciferase is not generally preferred over firefly luciferase for reporter assays, although it has recently become popular as a companion reporter for experiments where two different reporters are needed. Dual firefly and Renilla luciferase assays allow the rapid and sequential quantification of firefly and Renilla luciferase activity within a single assay sample. Dual luciferase assay reagents are available from Promega and Packard Biosciences. In each of these assay systems, firefly luciferase is measured first following the addition of a cell lysis/firefly luciferase substrate reagent to the assay sample. After quantifying firefly luciferase luminescence, Renilla luciferase activity is detected following the addition of Renilla luciferase substrate buffer to the same assay sample. This second reagent addition quenches firefly luciferase luminescence and initiates the Renilla luciferase luminescent signal, which is subsequently detected in a Packard Topcount luminometer or similar. The dual reporter assay can be completed in 30 min. In our laboratory, we routinely use the Packard FireLite dual luciferase assay reagent for the measurement of Renilla luciferase, or for the simultaneous measurement of both luciferases in the same assay sample according to Protocol 8. Dual reporter assays are most commonly used for the internal control of reporter gene experiments, to compensate for experimental variables such as transfection efficiency. Dual reporter assays may also be used to study twosignal transduction pathways within the same cell, using cells transfected with two reporter genes each specific for a different signal-transduction cascade. Furthermore, in a pharmaceutical screening programme, dual-reporter assays can be used to screen compounds against two GPCRs simultaneously. For this latter application, two cell lines, each expressing a unique GPCR-reporter gene combination, are mixed within a single well of an assay plate. Following the addition of a compound to that well, a dual-reporter assay is performed to 203

Stephen Rees et al. assess the activity of the compound against both GPCRs. In addition to the dual luciferase assay, protocols have been developed for the measurement of firefly luciferase and B-galactosidase, and firefly luciferase and SEAP, within the same assay sample. Protocol 8. Renilla luciferase and dual firefly and Renilla luciferase assay Equipment and reagents • Topcount 96-well microplate scintillation and luminescence counter (Packard)

• FireLite luciferase assay system (Packard) • Black plate sealers

Method 1. According to the nature of the reporter-gene experiment, cells are prepared as described in Protocol 4. Compound addition is performed as described in Protocol 5 or 6. 2. Incubate cells for 4 h at 37°C and 5% CO2 following test compound addition. 3. Dilute each bottle of LucLite and RenLite substrate with 10 ml of buffer (supplied).a 4. Add 50 ul of reconstituted LucLite substrate into each well.b 5. Place a black plate sealer onto the bottom of each plate to produce completely black wells. 6. Leave 10 min. 7. Place into a Topcount luminometer, and allow to light adjust in the dark for 2 min.c Read each well for 1 s on single-photon counting mode.d 8. Remove the plate from the Topcount. 9. Add 50 ul of reconstituted RenLite substrate into each well.6 10. Place into a Topcount and allow to light adjust in the dark for 2 min.c 11. Read each well for 1 s on single-photon counting mode.f a

One bottle of substrate is enough for two 96-well plates. This makes 100 ul in total (including test compounds). c To allow for quenching of plate autoluminescence. d To measure firefly luciferase activity. This is not required if this protocol is being used to assess Renilla luciferase activity alone. Firefly luciferase luminescence is stable for up to 6 h at room temperature. e This makes 150 ul in total (including test compounds). Addition of RenLite also quenches firefly luciferase luminescence. f This will measure Renilla luciferase luminescence. The signal is stable for up to 5 h at room temperature. b

204

8: Reporter gene systems

6.3 LuFLIPRase The LuFLIPRase assay provides an alternative means of simultaneously measuring two signalling events within the same cells or within a mixed cell population (76). This is a 96-well microplate assay that combines a firefly luciferase reporter-gene assay with the measurement of intracellular ion fluxes, most commonly changes in the level of intracellular calcium concentration, using the Fluorometric Imaging Plate Reader (FLIPR) and a microplate luminometer for signal detection. The FLIPR has been developed by Molecular Devices, and is a 96-well plate fluorescence detector that allows the simultaneous addition of assay reagent or compound to every well of a 96-well plate. This is followed by the immediate detection of any fluorescent signal produced within each well of the assay plate. The FLIPR has been used with a range of fluorescent indicator dyes to detect changes in intracellular calcium concentration, intracellular pH, and intracellular sodium concentration. In the LuFLIPRase assay, cells are plated into a 96-well microplate and loaded with the indicator dye calcium green. Agonist is added using the FLIPR, and the mobilization of intracellular calcium is detected as a consequence of an increase in fluorescence of the indicator dye. The assay plate is removed from the FLIPR, the cells are left for 4 h to allow for transcription and translation of the luciferase reporter gene, and firefly luciferase assay is quantified as described in Protocol 7 (76). The LuFLIPRase assay has been used to study the signalling of GPCRs that regulate more than one signal-transduction event. For example, the human PACAP (pituitary adenylyl cyclase activating peptide) receptor was stably expressed in CHO cells together with a 6CREluciferase reporter gene. Treatment of these cells with the agonist peptide PACAP(l-38) results in the activation of adenylyl cyclase to result in an upregulation of the CRE-luciferase reporter gene, and the activation of phospholipase C to cause the subsequent mobilization of intracellular calcium (76). As with the dual reporter-gene assay, the LuFLIPRase assay can also be used for the simultaneous screening of compound activity at two GPCRs. In the LuFLIPRase assay, activation of each receptor would be linked to a different detection system. For instance, one cell line would contain a luciferase reporter gene responsive to one receptor, and the second cell line would contain a receptor which is coupled to an intracellular ion flux measurable with a fluorescent dye in the FLIPR. The LuFLIPRase assay is a versatile technique that can be adapted to the study of almost any receptor-regulated signalling system through the use of reporter genes and fluorescent dyes.

6.4 Secreted placental alkaline phosphatase (SEAP) Secreted Placental Alkaline Phosphatase (SEAP or sPAP) was derived from placental alkaline phosphatase (PAP) following the removal of approximately 30 amino acids from the C-terminus of the protein to remove a membrane205

Stephen Rees et al. anchoring domain (77-79). When used as a reporter enzyme, the levels of SEAP activity in the culture medium provide a direct measurement of the activity of a SEAP-containing reporter gene (78). The secreted nature of SEAP avoids the requirement for cell lysis to detect enzyme activity. This also offers the advantage that the kinetics of reporter-gene activation can be monitored through the repeated sampling of the same culture medium. One disadvantage of SEAP is the requirement to heat-inactivate the assay sample prior to assay. Mammalian cells express a background of endogenous alkaline phosphatases. Unlike these alkaline phosphatases, SEAP is extremely heat stable. Samples for SEAP measurement must be heat-inactivated by warming to 65°C for 30 min, prior to assay, to inactivate endogenous alkaline phosphatases (81). However, if the cells are plated into fresh serum-free medium on the day of assay, we find that heat inactivation is not necessary (Protocol 9). Calorimetric, fluorescent, and chemiluminescent assay systems are available for SEAP detection. This offers the investigator a number of options for the detection of SEAP activity, according to cost and the availability of suitable detection apparatus. The calorimetric assay is simple and inexpensive. We use the calorimetric assay described in Protocol 9, and find that this is suitable for most reporter-gene assays, including 96-well microplate assays. However, calorimetric SEAP assays may not have the sensitivity required to detect weak signalling events, or for the assay of SEAP activity in miniaturized assay-plate formats (384- and 1536-well). A number of molecular biology reagent suppliers, including Clontech and Tropix, have developed chemiluminescent and fluorescent assay systems for SEAP. The chemiluminescent assay system offers similar sensitivity and dynamic range to the firefly luciferase assay reagents described earlier. As with firefly luciferase assays, chemiluminescent and fluorescent SEAP assays are expensive, and require the availability of appropriate detection apparatus. Protocol 9.

Calorimetric SEAP assay

Equipment and reagents • 1 M DEA buffer (105.4 g I-1 diethanolamine, DEA; 0.28 M NaCI; 0.5 mM MgCI2; pH 9.5; store at 4°C)

• p-nitrophenolphosphate (PNPP) substrate (Sigma 104-0)

Method 1. According to the nature of the reporter gene experiment, cells are prepared as described in Protocol 4. Compound addition is performed as described in Protocol 5 or 6. 2. After 5-6 h incubation with the test compound, cool the plates to room temperature. 3. Dilute PNPP substrate (800 mg per 100 ml) in DEA buffer.3 206

8: Reporter gene systems 4. Add 200 ul of diluted PNPP to each well, and leave 5 min. 5. Read absorbance at 405 nM in a 96-well spectrophotometer, using the kinetic mode.b a

A total of 20 ml is required per 96-well plate. An endpoint can be measured, but the time of incubation with PNPP substrate before measurement must be accounted for using the following calculation: [SEAP] in sample = A/T x 18.5 x V, where T = time in substrate (minutes), A = absorbance at 405 nm, V = volume of sample assayed. b

6.5 Chloramphenicol acetyltransferase The bacterial enzyme chloramphenicol acetyltransferase (CAT) was one of the first enzymes to be used in reporter gene studies (80). However, its use has been largely superseded by other enzymes for which simpler and more sensitive assay procedures are available. Traditionally, CAT activity in transfected cells has been measured using either liquid scintillation counting (LSC) or thin layer chromatography (TLC). Cell extracts are incubated in a reaction mix containing 14C- or 3H-labeled chloramphenicol and n-butyryl coenzyme A. Chloramphenicol acetyltransferase transfers the n-butyryl moiety of the cofactor to chloramphenicol. The radiolabelled, butyrylated chloramphenicol can then be detected with either LSC or TLC. These assay procedures are labour-intensive, involve the use of radioactivity, and have a very low throughput (81). Sandwich ELISA assays for CAT activity have been developed, and are available from suppliers including Boehringer Mannheim and Clontech (82). These assays are non-radioactive and sensitive. CAT activity in the assay sample is directly proportional to the colour generated, and is quantified against a standard curve generated with known amounts of CAT. Protocols 10 and 11 have been used to determine CAT activity in mammalian cells. These protocols involve the assay of cells from a 60 mm or 100 mm tissue culture dish for each sample point. For a transient transfection, CAT reporter gene assay cell extracts are collected 48-72 h post-transfection. The cell extraction procedure is based on rapidly repeated cycles of freezing and thawing of cells in Tris buffer to disrupt cellular membranes. In contrast to other reporter gene assays described in this chapter, this assay involves the use of a larger number of cells, is not amenable to a 96-plate format, and requires multiple reagent addition and plate-to-plate transfers. Protocol 10. Preparation of cells for assay in CAT ELISA Equipment and reagents • PBS buffer (137 mM NaCI, 2.7 mM KCI, 4.3 mM Na2HP04.7H2O, 1.4 mM KH2P04) . 0.25 M Tris-HCI, pH 8.0

• TEN buffer (40 mM Tris-HCI, 1 mM EDTA, 150 mM NaCI, pH 7.5)

207

Stephen Rees et al. Protocol 10.

Continued

Method 1. On the day prior to assay, plate cells at a seeding density of 60% into 60 mm or 100 mm culture dishes, and incubate overnight at 37°C in 5% CO2 in serum-free growth medium. 2. On the day of assay, remove the medium and replace with 5 ml phenol-red free growth medium. 3. Add the test compounds, and incubate the cells for 6 h at 37°C in 5% C02. 4. Remove the culture medium from the cells. 5. Wash the cells three times with PBS buffer.a 6. Add 1 ml of TEN buffer (per 60 mm or 100 mm culture dish) and incubate for 5 min at room temperature. 7. Scrape the plate surface with a rubber policeman, and transfer the cells to a microcentrifuge tube. 8. Spin cells at top speed in a microcentrifuge for 1 min at 4°C. 9. Remove the supernatant, and re-suspend the pellet in 100 ul of 0.25 M Tris-HCI, pH 8.0, per 60 mm culture dish (or 150 ul per 100 mm culture dish of cells). 10. Subject the extracts to three rapid freeze-thaw cycles, vortexing vigorously after each thaw cycle.b 11. Heat the lysates at 60°C for 10 min.c 12. Spin the lysates at top speed in a microcentrifuge for 2 min. 13. Transfer the supernatant to a fresh tube.d * Be careful not to dislodge any of the cells. Remove as much of the final PBS wash as possible with a pipette tip. b Use a dry ice-ethanol bath to freeze, and a 37°C water bath to thaw the cells. c This will inactivate any endogenous deacetylase activity. This heating step will not inactivate CAT, but it may inactivate other reporter gene products (e.g., luciferase or p-galactosidase). Omit this step when assaying for other reporters in the sample. d The extracts may be assayed directly or stored at -70°C. Avoid multiple freeze-thaw cycles.

Protocol 11. CATquantitation Equipment and reagents • 96-well (flat-bottom) ELISA plate • Carbonate coating buffer (0.025 M sodium bicarbonate, 0.025 M sodium carbonate, adjust pH to 9.5 using 1 N HCI or 1 N NaOH) . TEST wash buffer (20 mM Tris-HCI, pH7.6; 150 mM NaCI; 0.05% (v/v) Tween-20)

208

1 M phosphoric acid 50 ml (for better mixing) or 15 ml polypropylene tubes for dilutions Anti-CAT pAb, affinity purified (Promega) CAT ELISA 5x buffer (Promega) CAT standard, 100 ng ml-1 (Promega)

8: Reporter gene systems • Biotinylated anti-CAT pAb (Promega) • Streptavidin-HRP conjugate (Promega)

• TMB solution (Promega) • Peroxidase substrate (Promega)

Method 1. Prepare the primary coating antibody solution by adding 40 ul of anti-CAT pAb to 10 ml of carbonate coating buffer in a 15 ml polypropylene tube per 96-well plate. Mix thoroughly. 2. Add 100 ul of the anti-CAT pAb coating solution to each well of a polystyrene ELISA plate. 3. Seal the wells with a plate sealer, and incubate overnight (14-18 h) at 4°C. 4. Prepare CAT ELISA buffer by placing 40 ml of deionized water in a 50 ml polypropylene tube. Aspirate 10 ml of the CAT ELISA 5X buffer, and add it to the water. 5. Remove the coated plate from the refrigerator. Flick out the contents of the wells and slap the plate upside down three times on a paper towel to remove all liquid. 6. Add 200 ul of CAT ELISA buffer to each well.a 7. Incubate at room temperature for 1 h with shaking (400 r.p.m.). 8. Flick out the contents of the wells over a sink. Slap the plate three times upside down on a paper towel. Designate two columns of wells (16 wells) for the standard curve. To prepare the CAT standard curve within the assay plate, add 100 ul per well of the CAT ELISA buffer to wells in rows B through H in the two columns designated for the standard curve. 9. Dilute the CAT standard 1:100 in CAT ELISA buffer to achieve a concentration of 1000 pg ml-1. 10. Add 200 ul of diluted CAT standard (1000 pg ml-1) to the first well (row A) in each column designated for the standard curve. 11. Immediately perform serial 1:2 dilutions (100 ul per well) in the two columns designated for the standard curve. In the last set of wells for the standard curve, do not add any CAT. The final concentrations (in duplicate) in the CAT control columns will be 0-1000 pg ml-1. 12. Add 100 ul of the test samples to each of the remaining wells. 13. Incubate the plate for 2 h at room temperature with shaking (400 r.p.m.). 14. Wash all wells with TBST wash buffer, using an automated plate washer. Wash four times. 15. In a 15 ml polypropylene tube, add 40 ul of the biotinylated anti-CAT pAb to 10 ml of CAT ELISA 1X buffer (1: 250 dilution) to prepare enough reagent for a full 96-well plate. Mix thoroughly.

209

Stephen Rees et al. Protocol 11. Continued 16. Add 100 ul of the diluted, biotinylated anti-CAT pAb to each well. 17. Incubate for 1 h at room temperature with shaking (400 r.p.m.j. 18. Wash plate 5 times with TBST wash buffer. 19. In a 15 ml polypropylene tube, add 40 ul of the stock streptavidin-HRP conjugate to 10 ml of CAT ELISA buffer (1: 250 dilution), to prepare enough reagent for a full 96-well plate. Mix thoroughly. 20. Add 100 ul of the diluted, biotinylated anti-CAT pAb to each well. 21. Incubate for 30 min at room temperature with shaking (400 r.p.m.). 22. Wash plate 5 times with TBST wash buffer. 23. Prepare the TMB substrate 30 min prior to colour development, by combining 5 ml of TMB solution with 5 ml of peroxidase for each 96well plate. Mix gently. 24. Add 100 ul of this TMB substrate to each well. 25. Incubate at room temperature (or 4°C for a slower reaction) for approximately 10 min.b 26. Stop the reaction by adding 100 ul of 1 M phosphoric acid to each well in the same order that substrate was added.c 27. Record the absorbance at 450 nm on a plate reader within 1 h of stopping the reaction. * Ensure that the surface of the wells where the antibody is bound to the plate is not scratched. b Observe the blue colour that forms in the wells, paying particular attention to the colour formation in the wells used for the standard curve. c The colour should change to yellow as the pH decreases. Plate coating requires an overnight (14-18 h) incubation. The subsequent steps require approximately 5 h to complete and should be performed on the following day.

6.6 B-galactosidase The lacZ gene of E. coli encodes for B-galactosidase (B-gal). This enzyme catalyses the hydrolysis of various B-galactosides, including lactose. B-gal activity can be directly measured in extracts prepared from transfected eukaryotic cells (7). (3-gal has been used as a reporter enzyme in mammalian cells for many years, and a plethora of calorimetric, fluorescent, and chemiluminescent detection reagents have been developed (7). These reagents can be purchased from suppliers such as Boehringer Mannheim, Promega, Clontech, and Stratagene. (3-gal activity in mammalian cells is commonly assayed using the calorimetric substrate ONPG (o-nitrophenyl 3-D-galactopyranoside) (5). This is added to lysed cells, and optical density is measured in a spectrophotometer. To increase the sensitivity of the assay, chemiluminescent detection systems 210

8: Reporter gene systems have been developed. These are based on the use of 1,2-dioxetane substrates, and have a detection limit of around 10 fg B-gal, an increase in sensitivity of several orders of magnitude over the calorimetric assay. A number of these assays have stable luminescence with a half life in excess of 1 h. Sandwich ELISA assays for B-gal activity have also been developed. p-gal activity can also be detected in a whole-cell fluorescent assay by loading cells with the fluorogenic B-galactosidase substrate, fluorescein di-B-Dgalactopyranoside (FDG). This detection method can be combined with analysis on a fluorescence-activated cell sorter (FACS) to quantify B-galactosidase activity directly in individual cells (83). As with the alternative SEAP assays, the choice of B-gal assay is largely determined by the requirement for sensitivity, balanced against the cost of the assay reagent and the availability of suitable detection apparatus. Within our laboratory we have used the calorimetric B-gal assay available from Promega, and present this method in Protocol 12. For the measurement of weak promoters or for the measurement of small assay samples, the use of a chemiluminescent detection system may be necessary. Protocol 12. Measurement of B-galactosidase activity using ONPG substrate Equipment and reagents • B-galactosidase enzyme assay system with reporter lysis buffer (Promega) . PBS buffer (Mg2+- and Ca2+-free)

• Cells expressing receptor and reporter gene of interest plated into 60 mm or 100 mm tissue culture dishes

Method 1. Add 4 volumes of water to 1 volume of 5x reporter lysis buffer to produce a 1X stock solution. 2. Remove the growth medium from the cells. Wash the cells twice with PBS buffer, being careful not to dislodge any of the cells. Remove all the final wash with a pipette tip. 3. Add 1x reporter lysis buffer to the cells (400 ul for a 60 mm culture dish, 900 ul for a 100 mm culture dish). Incubate at room temperature for 15 min with gentle rocking. 4. Scrape the cell lysate from the plate. Using a pipette, transfer the cell lysate to a microcentrifuge tube, and place the samples on ice. 5. Vortex the sample for 15 s, then pellet the cell debris by centrifugation for 2 min at top speed in a microcentrifuge at 4°C. Transfer the supernatants to a fresh tube. These can be assayed immediately, or stored at -80°C for at least 2 months. 6. Thaw the assay kit components, and mix each component well and place the assay 2x buffer on ice. 211

Stephen Rees et al. Protocol 12. Continued 7. Dilute the cell extracts in 1X reporter lysis buffer (if necessary). 8. Mix 30 ul of extract with 20 ul of 1 x reporter lysis buffer in a 96-well microtitre plate.a 9. Prepare the same dilution of a cell extract made from cells which have not been transfected with the B-galactosidase gene.b 10. Pipette 50 ul of the cell extracts into labelled wells of a 96-well plate. 11. Add 50 ul of assay 2x buffer to each well of the 96-well plate. 12. Mix all samples by pipetting the well contents. 13. Place a cover on the plate. 14. Incubate the plate at 37°C for 30 min, or until a faint yellow colour has developed. 15. Stop the reaction by adding 150 ul of 1 M sodium carbonate. Mix by pipetting the contents of each well.c 16. Read the absorbance of the samples at 420 nm in a plate reader spectrophotometer. a

This is a starting dilution. Up to 50ul of cell extract can be used per reaction. This is as a negative control. Avoid producing bubbles, which may interfere with absorbance readings.

b c

6.7 B-lactamase B-lactamase (penicillin amido-B-lactamhydrolase) is a recent addition to the field of reporter enzymes. This enzyme offers further versatility in reporter gene assays, as it can be fused to other proteins or signalling sequences, and still retain activity. A number of recombinant B-lactamases have been engineered to develop a secreted enzyme equivalent to SEAP, to develop an intracellularly expressed enzyme equivalent to firefly luciferase or B-galactosidase, or to generate a plasma membrane-associated enzyme (84). Enzyme activity is usually measured in a calorimetric assay, following the addition of the chromogenic substrates nitrocefin (Becton Dickinson) or PADAC (Calbiochem). In Protocol 13 we describe the use of nitrocefin to detect B-lactamase activity. Nitrocefin is a highly coloured cephalosporin that produces a distinctive colour change from yellow to red, and an increase in absorbance at 495 nm, following B-lactamase-mediated hydrolysis. Secreted (3-lactamase can be assayed in the cell culture medium; membrane-anchored or cytosolic (3-lactamase require cell lysis prior to assay. As with other calorimetric assays, the nitrocefin Blactamase assay is cheap, but does not have the sensitivity of the fluorescent or chemiluminescent reporter gene assays (84). A novel fluorescent detector of B-lactamase activity, CCF2, has recently been developed (85). CCF2 is a cell-permeable fluorophore, and allows the fluor212

8: Reporter gene systems escent detection of fi-lactamase activity in whole cells. This reporter allows the highly sensitive real-time measurement of transcription in individual living cells using confocal microscopy, and is amenable to detection on the FACS apparatus (85). CCF2 is an ester composed of the two fluorescent molecules 7-hydroxycoumarin and fluorescein attached to a cephalosporin backbone. The substrate fluoresces green (520 nm) upon excitation at 409 nm, as a consequence of a fluorescence energy transfer (FRET) event between the 7-hydroxy coumarin and the fluorescein groups in the substrate molecule. After attack by p-lactamase, the fluorescein moiety is released from CCF2. This disrupts the FRET, resulting in a shift in emission wavelength to blue light (447 nm). Thus B-lactamase activity is assessed through an increase in the emission of blue light at 447 nm accompanied by a decrease in the emission of green light at 520 nm. This offers a highly sensitive and versatile assay of reporter gene activity, and is especially suited for the analysis of reporter gene activity in live cells, or in small numbers of cells. CCF2 is available from Aurora Biosciences (85).

Protocol 13. Use of nitrocefin to detect secreted B-lactamase activity in mammalian cells Equipment and reagents • Cells expressing receptor and reporter gene of interest

• Nitrocefin (Becton Dickinson Inc.) • 96-well microtitre plate

Method 1. According to the nature of the reporter gene experiment, cells are prepared as described in Protocol 4. Compound addition is performed as described in Protocol 5 or 6. 2. Dissolve 30 mg nitrocefin in 300 ul DMSO. 3. Add the nitrocefin-DMSO to 75 ml 50 mM KH2PO4 (pH 7.0) containing 0.1% Triton-X, dilute with equal volume of water, and store at -20°C. 4. Transfer 30 ul medium from the cells into 96-well microtitre plate.a 5. Add 150 ul nitrocefin solution, and agitate the plate. 6. Cover the plate to prevent evaporation, and incubate for 30 min at 37°C. 7. Read optical density at 492 nm in a microplate spectrophotometer. a

For assay of secreted B-lactamase. For the assay of cytosolic B-lactamase it would be necessary to add a cell lysis step.

213

Stephen Rees el al.

7. Reporter protein assays In contrast to reporter genes, a reporter protein is defined as a protein which is constitutively expressed in mammalian cells to provide a direct measurement of the changes in the intracellular environment.

7.1 Aequorin Aequorin has been used to detect increases in the level of intracellular calcium following agonist activity at a wide range of GPCRs which couple to G proteins of the Gaq/11 family (15). The principles and uses of aequorin as a reporter protein have been described in Section 3.2. Aequorin luminescence has a signal half-life of less than 30 s. Thus a luminometer equipped with reagent injectors is required for the detection of the luminescent signal. Aequorin offers a highly sensitive method to assess changes in the level of intracellular calcium which also offers significant time savings over existing methods for the measurement of changes in either intracellular calcium concentration or inositol phosphate levels. Aequorin assays can be performed in a 96-well microplate. In Protocol 14 we describe the use of aequorin to detect signal transduction in mammalian cells expressing both the GPCR and the aequorin reporter (15). For high throughput calcium assays, Molecular Devices have developed the Fluorescence Imaging Plate Reader (FLIPR). As described in Section 6.3, this allows the simultaneous assay of intracellular calcium concentration in every well of a 96-well plate. FLIPR assays require the use calcium indicator dyes rather than the cofactor coelenterazine, thus generating a significant reduction in the cost of the assay. However, aequorin offers an alternative to the FLIPR should this apparatus not be available. Protocol 14. Use of the calcium-sensitive photoprotein aequorin to measure Ca2+ mobilization in mammalian cells Equipment and reagents • White tissue-culture treated 96-well microtitre plates . Assay buffer (125 mM NaCI, 10 mM Hepes, 10 mM glutamine, 5 mM KCI, 2 mM MgCI2.6H2O, 0.5 mM NaHCO3, 0.1% BSA) • Cell line stably expressing the apoaequorin protein, either stably or transiently transfected with the GPCR under study

• Coelenterazine (dissolved in methanol at 5 mM, stored under nitrogen at -20°C) . Luminometer with the capacity for direct injection of test compounds into the assay well, such as the Dynatech ML3000 luminometer • Compounds for analysis in assay buffer at double the required final concentration

Methods 1. Trypsinize a confluent T175 flask of cells. Resuspend in 50 ml of growth medium.

214

8: Reporter gene systems 2. Add 100 ul cells in growth medium to each well of a 96-well plate.a 3. Ensure that there is no neomycin in the medium.b 4. Return cells to the incubator for 4-18 h to allow them to adhere to the plate. 5. Reconstitute aequorin by replacing medium with 100 ul fresh medium containing a final concentration of 5 u.M coelenterazine. 6. Incubate at room temperature in the dark for 2 h. 7. Add the test compound to cells using the injection mode of the luminometer, and immediately read the luminescent signal.c 8. It may be preferable to calibrate the luminometer, so that levels of light generated can be correlated with intracellular Ca2+ concentration. a

The appropriate cell concentration depends on the cells used, and varies as a function of cell growth rate, expression levels of reporter, and of receptor of interest. Cells should be seeded so that on the day of assay they form a confluent monolayer. If cells need to be transiently transfected with either apoaequorin or receptor of interest prior to assay, they should be seeded at a lower density (to give 50-80% confluence prior to transfection). b Neomycin is a known inhibitor of phospholipase C, and is also thought to have an effect on plasma-membrane calcium channels. c As the aequorin flash response is very rapid (<10 s duration), it is necessary to measure luminescence immediately after compound addition. Ideal instrumentation permits simultaneous addition and recording.

7.2 GFP GFP is a 238 amino acid photoprotein that emits green light with an emission maximum of 509 nm upon fluorescent excitation at 488 nm. Unlike other bioluminescent reporter proteins, no additional substrates or cofactors are required for light emission (13). GFP fluorescence is stable, and has been measured non-invasively in living cells of many species including mammalian cells, Drosophila, C. elegans, yeast, and E. coli. GFP fluorescence can be detected by fluorimetry, by FACS, and by microscopy. As there is no assay reagent or assay protocol, the attractiveness of GFP as a reporter protein is cost, together with the speed and simplicity of the assay (40). The use of GFP as an inducible reporter gene has been limited due to the brightness of the protein, which while readily detectable by fluorescence microscopy or FACS analysis, has not been easily detectable in a plate fluorimeter. The availability of the cDNA sequence for GFP has resulted in the generation and characterization of several GFP mutants with enhanced fluorescence emission. The active chromophore within GFP is a cyclic hexapeptide spanning amino acids 64-69 (86). Mutation of the serine at amino acid 65 to threonine has resulted in the generation of a protein with a sixfold increase in the intensity of fluorescence emission (87). Furthermore, the presence of the Ser65Thr and the mutation of the phenylalanine residue at position 64 to leucine has resulted in a 35-fold increase in fluorescence intensity (87). 215

Stephen Rees et al. Variants of GFP are now available for which codon usage within the cDNA has been optimized for human cell expression (88). A number of novel mutants of GFP have also been identified, with altered excitation or emission characteristics. For example, mutation of the tyrosine residue at position 66 to histidine has generated a protein with blue fluorescence emission, the so-called blue fluorescent protein (BFP), with a nmax for excitation of 368 nm and for emission of 445 nm (40). These and many other variants of GFP protein are now commercially available from suppliers such as Clontech and Packard Biosciences (17,40,89-91). A further limitation on the use of GFP as a reporter protein is the extremely long half life of the protein molecule (>36 h). Within an inducible reporter gene assay this leads to the accumulation of GFP as a result of basal promoter activity. As a consequence, it has not been possible to detect the drug-stimulated reporter response above the high basal GFP signal. In recent months, Clontech have developed a number of destabilized GFP proteins with a short half-life. These proteins contain the PEST domain from mouse ornithine decarboxylase fused to the C-terminus of GFP. This domain targets GFP for rapid degradation, effectively reducing the half-life of the protein to 2 h. This GFP is now being used in reporter gene assays to detect GPCR signal transduction, with detection of GFP fluorescence by either FACS analysis or microplate fluorimetry. The availability of a range of GFP proteins with fluorescence characteristics that can be detected individually when present in a single assay sample also offers the possibility for the design of dual or multiple reporter assays. As with the dual luciferase assay, it may be possible to study multiple signal transduction events simultaneously, or to study the activity of a compound at multiple receptors contained within the same assay sample. GFP has been widely used in fusion proteins to assess protein trafficking (92), and subcellular localization of recombinantly expressed proteins (93). For example, a fusion protein between the B2-adrenoceptor and GFP has been used to monitor receptor expression, localization at the plasma membrane, and internalization following agonist stimulation (16). Further modified GFPs are also becoming available. For example, a recent report described the development of a pH-sensitive GFP, in which the fluorescence emission alters upon a change in pH (94), and in a further report a reporter of membrane voltage was created following the construction of a Drosophila frizzled ion channel-GFP fusion protein (95). Several groups are investigating the use of GFP and BFP as partners for fluorescence resonance energy transfer (FRET). In such studies, excitation of BFP at 368 nm causes emission of light at 445 nm, which excites a Ser65Cys mutant of GFP to generate light emission at 509 nm (17). A fusion protein consisting of GFP and BFP, with a linker sequence containing a trypsin cleavage site between the two fluorescent proteins, was constructed. When excited at 368 nm, the fusion protein emits light at 509 nm. Upon treatment with trypsin, the fusion protein was cleaved, with the result 216

8: Reporter gene systems that fluorescence excitation at 368 nm now generates fluorescence emission at 445 nm (17). Such FRET partners may be used as reporters of protease activity. In a similar report, Miyawaki and colleagues described the construction of a BFP-calmodulin-GFP fusion construct and its use as a non-enzymatic reporter of calcium concentration in mammalian cells (96). An increase in intracellular calcium concentration resulted in an alteration in the conformation of the reporter protein, to cause an increase in FRET with an increase in fluorescence emission at 509 nm.

8. Summary and future perspectives The use of reporter gene assays for the study of GPCR signal transduction has steadily increased over the last few years. A plethora of reporter plasmids, together with a range of alternative reporter enzymes, have become available, offering huge choice and diversity for the investigator. Reporter systems are available for the study of all the common GPCR signal-transduction cascades. Through the choice of reporter enzyme, signal transduction can be studied in whole cells or in cell extracts. It is also possible to study multiple signaltransduction pathways simultaneously, through the use of two or more reporter genes. These assays have been used to elucidate new mechanisms of GPCR signal transduction, and have become widely used within the pharmaceutical industry for the high throughput screening of random compounds for new activity at GPCRs. In contrast to traditional biochemical secondmessenger assays, reporter-gene systems are usually simpler to perform, are cheaper, involve the use of fewer cells, and are non-radioactive. Looking to the future, the number of reporter enzymes is increasing. Similarly, the simplicity and sensitivity of the assay reagents is increasing, as is the sensitivity of the apparatus available for the detection of the reporter enzyme activity. Perhaps the most exciting opportunities in this area will involve the creation of reporter genes and reporter enzymes using the Aequorea victoria Green Fluorescent Protein. The creation of a GFP-calmodulin-BFP FRET reporter for calcium concentration offers the possibility for the creation of further FRET reporters for other second messenger metabolites and for protein kinase activity. It is possible to imagine the creation of a range of FRET reporters for the study of signalling in whole cells, in which the fluorescence emission characteristics of the reporter are modified as a direct consequence of a specific signal transduction cascade.

Acknowledgements The authors would like to express their thanks to the following members of Receptor Systems, Receptor Pharmacology and Biomolecular Structure Units at GlaxoWellcome Research and Development, Stevenage, UK, for their 217

Stephen Rees et al. contributions to the work presented in this chapter: Nicola Bevan, Jim Coote, Liz Eggleston, Andy Green, Melanie Lee, Fiona Marshall, Sarah Parsons, Sharron Rhodes, Helen Rogers, and Sarah Scott. The authors would also like to thank Dr. Diane Ignar (Receptor Biochemistry, Glaxo Wellcome Inc.), Prof. Graeme Milligan (University of Glasgow, Scotland), Dr. Chris Roelant (RTRT, Leuven, Belgium) and Dr. Dave Burns (Packard Biosciences, Meriden, CT, USA) for assistance over many years, and Prof. Peter Shaw (Institute of Cell Signalling, University of Nottingham) for assistance with the MAP kinase studies.

References 1. Wise, A., Watson-Koken, M.-A., Rees, S., Lee, M., and Milligan, G. (1997). Biochem. J., 321,721. 2. Gudermann, T., Kalkbrenner, F., and Schultz, G. (1996). Ann. Rev. Pharmacol. Toxicol.,36,429. 3. Walker, D., and De Waard, M. (1998). Trends. Neurosci., 21,148. 4. Stratowa, C, Himmler, A., and Czernilofsky, C. (1995). Curr. Opin. Biotechnol., 6, 574. 5. Alam, J., and Cook, J. L. (1990). Anal. Biochem., 188, 245. 6. Wood, K. V. (1995). Curr. Opin. Biotechnol., 6,50. 7. Bronstein, L, Fortin, J., Stanley, P. E., Stewart, G. S. A. B., and Kricka, L. J. (1994). Anal Biochem., 219,169. 8. Suto, C. M., and Ignar, D. M. (1997). J. Biomolecular Screening, 2,1. 9. DeWet, J. R., Wood, K. V., DeLuca, M., Helsinki, D. R., and Subramani, S. (1987). Mol. Cell Biol., 7,725. 10. Lorenz, W. W., McCann, R. O., Longiaru, M., and Cormier, M. J. (1991). Proc. Natl. Acad. Sci. USA, 88,4438. 11. Henthorn, P., Zervos, P., Raducha, M., Harris, H., and Kadesch, M. (1988). Proc. Natl. Acad. Sci. USA, 85, 6342. 12. Chen, W., Shields, T. S., Stork, P. J. S., and Cone, R. D. (1995). Anal. Biochem., 226,349. 13. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Science, 263, 802. 14. Tanahashi, H., Ito, T., Inouye, S., Tsuji, F. I., and Sakaki, Y.(1990). Gene, 96,249. 15. Stables, J., Green, A., Marshall, F., Fraser, N., Knight., E, Sautel, M., Milligan, G., Lee, M., and Rees, S.(1997). Anal. Biochem., 252,115. 16. Barak, L. S., Stephen, S. G., Ferguson, J. Z., and Caron, M. G. (1997). J. Biol. Chem, 272,27497. 17. Heim, R., and Tsein, R. Y. (1995). Curr. Biol., 6,178. 18. Su, B., and Karin, M. (1996). Curr. Opin. Immunol., 8,402. 19. Karin, M. (1994). Curr. Opin. Cell Biol., 6,415. 20. Wang, H., Nelson, S., Ascoli, M., and Segaloff, D. L. (1992). Mol. Endocrinol, 6, 320. 21. Hill, C. S., and Treisman, R. (1995). EMBO J., 14, 5037. 22. Weyer, U., Schafer, R., Himmler, A., Mayer, S. K., Burger, E., Czernilofsky, A. P., and Stratowa, C. (1993). Receptors and Channels, 1,193.

218

8: Reporter gene systems 23. Stratowa, C, Machat, H., Burger, E., Himmler, A., Schafer, R., Spevak, W., Weyer, U., and Wiche-Castanon, M. (1995). J. Recept. Signal Transduct. Res., 15, 617. 24. Northrop, J. P., Ulman, K. S., and Crabtree, G. R. (1993). J. Biol. Chem., 268, 2917. 25. Park, J., Collier, P., Chen, E., and Gibson, C. W. (1994). DNA Cell Biol, 11,1147. 26. Lillie, J. W., and Green, M. R. (1989). Nature, 338,39. 27. Hata, A., Ohno, S., Akita, Y., and Suzuki, K. (1989). J. Biol. Chem., 264, 6404. 28. Chen, W., Shields, T. S., Stork, P. J., and Cone, R. D. (1995). Anal. Biochem., 226, 349. 29. Hata, A., Akita, Y., Konno, Y., Suzuki, K., and Ohno, S. (1989). FEBS Lett., 252, 144. 30. An, S., Bleu, T., Hallmark, O. G., and Goetzl, E. J. (1998). J. Biol. Chem., 273, 7906. 31. Boss, V., Talpade, D. J., and Murphy, T. J. (1996). J. Biol. Chem., 271,10429. 32. Sambrook, I., Fritsch, E. F., and Maniatis, T. (ed.). (1989). Molecular cloning: a laboratory manual (2nd edn), Cold Spring Harbor Laboratory Press, NY. 33. Ashley, C. C., and Campball, A. K. (ed.). (1979). The detection and measurement of free Ca2+ in cells. Elsevier/North Holland, Amsterdam. 34. Brini, M., Marsault, R., Bastianutto, C., Alvarez, J., Pozzan, T., and Rizzuto, R. (1995). J. Biol. Chem., 270,9896. 35. Di Giorgi, F., Brini, M., Bastianutto, C, Marsault, R., Montero, M., Pizzo, P., Rossi, R., and Rizzuto, R. (1996). Gene, 17,113. 36. Rizzuto, R., Simpson, A. W. M., Brini, M., and Pozzan, T. (1994). Nature, 35,325. 37. Brini, M., Murgia, M., Pasti, M., Picard, D., Pozzan, T., and Rizzuto, R. (1993). EMB0 J., 12,4813. 38. Brini, M., De Giorgi, F., Murgia, M., Marsault, R., Massimino, M. L., Cantini, M., Rizzuto, R., and Pozzan, T. (1997). Mol. Biol Cell, 8,129. 39. Nakahashi, Y., Nelson, E., Fagan, K., Gonzales, E., Guillou, J-L., and Cooper, D. M. F. (1997). J. Biol. Chem., 272,18093. 40. Chalfie, M., and Kain, S. R. (ed.). (1998). Green fluorescent protein properties, applications and protocols. Wiley-Liss Press, New York. 41. George, S. E., Bungay, P. J., and Naylor, L. H. (1997). J. Neurochem., 69,1278. 42. Chuprun, J. K., Raymond, J. R., and Blackshear, P. J. (1997). J. Biol. Chem., 272, 773. 43. Castanon, M. J., and Spevak, W. (1994). Biochem. Biophys. Res. Commun., 198, 626. 44. George, S.E., Bungay, P. J., and Naylor, L. H. (1998). Biochem. Pharmacol, 56, 25. 45. Migeon, J. C., Thomas, S. L., and Nathanson, N. M. (1995). J. Biol Chem., 270, 16070. 46. Pepperl, D. J., and Regan, J. W. (1993). Mol. Pharmacol, 44, 802. 47. Felder, C. C., Joyce, K. E., Briley, E. M., Mansouri, J., Mackie, K., Blond, O., Lai, Y., Ma, A. L., and Mitchell, R. L. (1995). Mol Pharmacol, 48, 443. 48. Neer, E. J., and Clapham, D. E. (1989). Nature, 333,129. 49. Sista, P., Edmiston, S., Darges, J. W., Robinson, S., and Burns, D. J. (1994). Mol. Cell. Biochem., 141,129. 50. Button, D., and Brownstein, M. (1993). Cell Calcium, 14, 663. 219

Stephen flees et al. 51. Post, G. R., and Brown, J. H. (1996). FASEB J., 10,741. 52. Sugden, P. H., and Clerk, A. (1997). Cell Signal, 9,337. 53. Fukuda, K., Shoda, T., Morikawa, H., Kato, S., and Mori, K. (1997). FEBS Lett., 412,290. 54. Burt, A. R., Carr, I. C., Mullaney, I., Anderson, N. G., and Milligan, G. (1996). Biochem. J., 320,227. 55. Hordijk, P. L., Verlaan, I., van Corven, E. J., and Moolenaar, W. H. (1994). J. Biol. Chem., 269,645. 56. Fukuda, K., Kato, S., Morikawa, H., Shoda, T., and Mori, K. (1996). J. Neurochem., 67,1309. 57. Gille, H., Kortenjann, M., Thomae, O., Moomaw, C., Slaughter, C., Cobb, M. H., and Shaw, P. E. (1995). EMBO J., 14,951. 58. Strahl, T, Gille, H, and Shaw, P. E. (1996). Proc. Natl. Acad. Sci. USA, 93,11563. 59. Derijard, B., Hibi, M., Wu, I. H., Barret, T., Su, B., Den, T., Karin, M., and Davis, R. J. (1994). Cell, 76,1025. 60. Hill, C. S., and Triesman, R. (1995). Cell, 80,199. 61. Sadowski, I. (1995). Genetic Engineering, 17,119. 62. Kortenjann, M., Thomas, O., and Shaw, P. E. (1994). Mol. Cell. Biol., 14,4815. 63. Bevan, N., Scott, S., Shaw, P. E., Lee, M. G., Marshall, F. H., and Rees, S. (1998). Neuroreport., 9, 2703. 64. see http: //www. biomedcomp. com/biomedcomp. html for a comprehensive list of endogenous GPCR expression in mammalian cells. 65. Lurquin, P. F. (1997). Mol. Biotech., 7,5. 66. Sompayrac, L., and Danna, L. (1981). Proc. Natl. Acad. Sci. USA, 78,7575. 67. Feigner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. N., Wenz, M., Northrop, J. P., Ringold, G. M., and Danielson, M. (1987). Proc. Natl. Acad. Sci. USA, 84,7413. 68. Graham, F., and van der Eb, A. (1973). Virology, 52, 456. 69. Rees, S., Coote, J., Stables, J., Goodson, S., Harris, S., and Lee, M. G. (1996). Biotechniques, 20,102. 70. No, D., Yao, T. P., and Evans, R. M. (1996;. Proc. Natl. Acad. Sci. USA, 93,3346. 71. Gossen, M., and Bujard, H. (1992). Proc. Natl. Acad. Sci. USA, 89, 5547. 72. Jinsi-Parimoo, A., and Gershengorn, M. C. (1997). Endocrinology, 138,1471. 73. Porcellini, A., Ciullo, L, Pannain, S., Fenzi, G., and Awedimento, E. (1995). Oncogene, 11,1089. 74. Williams, T. M., Burlein, J. E., Ogden, S., Kricka, L. J., and Kant, J. A. (1989). Anal. Biochem., 176, 28. 75. Roelant, C. H., Burns, D. A., and Scheirer, W. (1996). Biotechniques, 20, 914. 76. Ignar, D. M., and Yingling, J. (1997). US patent application PU3166US1. 77. Berger, J., Hauber, J., Hauber, R., Geiger, R., and Cullen, B. R. (1988). Gene, 66,1. 78. Kain, S. R. (1997). Methods Mol. Biol., 63,49. 79. Cullen, B. R., and Malim, M. H. (1992). In Methods in enzymology (ed. Wu, R.). Vol. 216, p. 362. Academic Press, London. 80. Gorman, C. M., Moffat, L. F., and Howard, B. H. (1982). Mol. Cell Biol., 2,1044. 81. Cassinoti, P., and Weitz, M. (1994). Biotechniques, 17, 36. 82. Reifel-Miller, A. E., Conarty, D. M., Schacht, P. S., Birch, K. A., Heath, W. F., and Stramm, L. E. (1996). Biotechniques., 21,1033. 220

8: Reporter gene systems 83. Lorincz, M., Roederer, M., Diwu, Z., Herzenberg, L. A., and Nolan G. P. (1996). Cytometry., 24, 321. 84. Moore, J., Davis, S., and Dev, I. (1997). Anal. Biochem., 247,203. 85. Zlokarnik, G., Neglulescu, P. A., Knapp, T. E., Mere, L., Burres, N., Feng, L., Whitney, M. W., Roemer, K., and Tisane, R. Y. (1998). Science, 279,84. 86. Reid, B. G., and Flynn, G. C. (1997). Biochemistry, 36, 6786. 87. Haas, J, Park, E. C., and Seed, B. (1996). Curr. Biol, 6,315. 88. Zolotukhin, S., Potter, M., Hauswirth, W. W., Guy, J., and Nuzyczka, N. (1996). J. ViroL, 70,4646. 89. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Nature, 373,663. 90. Heim, R., Prasher, D. C., and Tsein, R. Y. (1994). Proc. Natl. Acad. Sci. USA, 91, 12501. 91. Heim, R., and Tsein, R. Y. (1995). Curr. Biol., 6,178. 92. Ferrer, J. C., Baque, S., and Guinovart, J. J. (1997). FEBS Lett., 415,249. 93. Wang, S., and Hazelrigg, T. (1994). Nature, 369,400. 94. Miesenbock, G., De Angelis, D. A., and Rothman, J. E. (1998). Nature, 394,192. 95. Siegel, M. S., and Isacoff, E. Y. (1997). Neuron, 19, 735. 96. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsein, R. Y. (1997). Nature, 388,834.

221

This page intentionally left blank

9 9

Adenylyl cyclases and cyclic AMP MAURICE K. C. HO and YUNG H. WONG

1. Introduction Adenylyl cyclase (AC; adenylate cyclase or ATP pyrophosphate lyase; E.G. 4.6.1.1) catalyses the formation of cyclic AMP (cAMP), using ATP as substrate. The theoretical topology of all ACs is similar, with both amino- and carboxyl-termini on the cytosolic side, twelve transmembrane domains divided into two groups, and linked with an extended cytosolic loop (1). The catalytic activity and the regulatory elements of AC appear to be concentrated on the cytosolic loop (C1 domain) and the carboxyl-terminal tail (C2 domain). To date, nine subtypes of mammalian adenylyl cyclases (ACs) have been successfully cloned (1-15), and a partial cDNA fragment corresponding to a novel AC subtype (type X) has also been identified (16). All ACs are subjected to extensive regulatory signals, including Ca2+, G proteins, protein phosphorylation, and small molecules such as forskolin and adenosine analogues (also known as P-site analogues), as reviewed in refs. 17-19, and summarized in Table 1. The regulatory mechanisms for each AC subtype may be far more complex, as indicated in several recent studies. Types II, IV, and VII ACs form a subfamily that is known to be stimulated by GBy complex in the presence of Gas-GTP. However, particular combinations of G protein B and y subunits appear to regulate type II AC differentially, it being stimulated by GB1y2 but inhibited by GB5y2 (20). Every AC has its own characteristic Gas-binding property. Forskolin and Gas can stimulate ACs synergistically (except type IX AC), which may be due to the enhancement of the interaction between two catalytic domains of AC (C1 and C2 domains; ref. 21). Such a synergistic event may also be due to the mutual enhancement of the affinity of the two activators to AC (22). Kinetic studies suggest that multiple Gas-binding sites may be present on the catalytic domains of types II, V, and VI ACs (23, 24), which may be due to their different tendencies to form oligomers (23). Mutational and biochemical analyses of Gas-AC interaction also indicate that Gas binds to several discrete regions of both Q and C2 domains (25, 26). cAMP-

Maurice K. C, Ho and Yung H, Wong Table 1. Modes of regulation of adenylyl cyclase subtypes Type

I

II

III

IV

V

VI

VII

VIII

IX

Ca2+/CaM CaMK Forskolin Gai

+

NE 7 + (-)d

+

NE 7 + 7

-a ? + -

-a 7 + -

NE 7 + 7

+ NE + NE

NE 7 NE 7

-b + -

+

Gas

+

+

+

+

+

+

+

+

+

Gaq

7

+e

7

+e

+e

+e

+e

?

7

Ga0 Gp-y P-site analogues PKA PKC

-

NE +

7 NE

NE + _ 7 +

NE NE _ +

NE NE _ +

7 7 _ 7

? NE _ 7 ??

_

_ 7 +

_ 7 +

7 +

7 + _ 7 +

+, Stimulation; -, inhibition; NE, no effect; ?, not tested. "Ca2+ alone. b CaMKIV-specific. c CaMKII-specific. d not usually seen. e probably through PKC. CaM, calmodulin; CaMK, Ca2+/calmodulin-dependent protein kinase; PKA, cAMP-dependent protein kinase; PKC, protein kinase C.

dependent protein kinase (PKA) phosphorylates and inhibits types V and VI AC (27, 28), and the inhibitory effect may be due to the incorporation of phosphate group into one of the Gas-binding sites (28). There are a number of substances modulating AC activity in a subtypespecific manner. Type II AC can be activated by treating with trypsin, thrombin, or plasminogen, while type V AC is inhibited by trypsinization (29). Oxidized derivatives of adrenaline, dopamine, and isoproterenol can specifically reduce the maximal activity of type V AC, but with no effect on type II AC (30). Lipophilic n-alkanols inhibit both types II and V AC, but type II AC can also be stimulated by ethanol when it is expressed on the plasma membrane (31). The differential regulation of AC activities by n-alkanols appears to be independent of the binding of guanine nucleotides. Obviously, a particular subtype of AC receives enormous streams of regulatory signals, and they are strictly controlled for producing precise levels of intracellular cAMP. Purified membrane-bound proteins such as ACs cannot retain their native three-dimensional structures in detergent-free conditions. However, detergentcontaining buffers hinder the investigation of molecular details, and prevent the crystallization of membrane-bound proteins. Bacterial expression systems facilitate large-scale production of high-purity recombinant proteins. Recent advances in the design of bacterial expression vectors (hexahistidine tags, fusion proteins, etc.) have simplified the purification procedures. However, bacteria cannot provide the proper post-translational modifications for mammalian proteins, which may be important for the proteins' functions. This expression 224

9: Adenylyl cyclases and cyclic AMP system is most suitable for producing soluble proteins of small to medium size. Large proteins (>100 kDa) tend to aggregate and become insoluble. Syntheses of membrane-bound proteins are also not favourable, because of the lack of proper protein transporting machinery, and the different compositions of the membrane lipids. Thus, purification of AC using bacterial expression systems could not be achieved until a recent breakthrough in the synthesis of soluble ACs. Both C1 and C2 domains of AC can be synthesized as soluble proteins in bacterial expression systems with hexahistidine tags for purification (32). Reconstitution of the two domains produces a soluble AC responsive to forskolinand Gas-mediated stimulation (21), as well as Ca2+- and Gai-mediated inhibition (33), in a fashion similar to native membrane-spanning ACs. Several interesting facts are observed. C1 and C2 domains derived from various ACs have very different yields, and thus one may not obtain enough of both domains from the same AC. Moreover, C1 and C2 domains of different ACs can form a fully functional soluble AC (32), which may be due to the high homologies of these domains between various ACs. The molecular structures of soluble forms of AC were subsequently resolved through crystallography. Unexpectedly, the C2 domain of type II AC can crystallize as a homodimer with two forskolin molecules (34), in an almost identical three-dimensional arrangement as that found for theC1-C2heterodimeric soluble AC bound to one Gas, one forskolin, and one ATP (35). The resolution of a Gas-AC complex offers a better view at the molecular level for predicting the interactions between AC and other regulatory components such as the Gas and GBy complexes (35). To study the characteristics of particular AC subtypes, an effective and quasi-native expression system would be more beneficial. Introduction of cDNA encoding ACs into mammalian cell lines has become a routine procedure for this purpose. Two transfection methods are described in detail in this chapter. A simple DEAE-dextran/chloroquine-mediated transfection protocol is provided as a classical method for transient expression of recombinant proteins. Adenovirus-mediated transfection has been developed recently, and is particularly useful for introducing cDNAs into less permissive cell types. Instead of measuring the catalytic activity of AC directly, the accumulation of intracellular cAMP can be easily monitored as a read-out of AC activity. Several methods of assaying cAMP are described. A simplified version of metabolic labelling of ATP and isolation of labelled cAMP is described here, which only requires a single type of radioisotope and much easier calculations. cAMP-responsive reporter systems using firefly luciferase or chloramphenicol acetyltransferase have also been used in various studies. Protocols for measuring the activities of these reporter enzymes are provided (see also Chapter 8). Furthermore, a number of interesting AC activitymonitoring systems are discussed briefly at the end of this chapter, which have distinct advantages in different circumstances. 225

Maurice K. C. Ho and Yung H. Wong

2. Mammalian expression of recombinant ACs Recombinant ACs can be co-expressed with other signalling components in mammalian cell lines for examining the regulation of cAMP production. Reproducible results from studying various types of regulation of AC can be obtained by expressing the recombinant proteins transiently in mammalian cell lines. Transient expression systems usually support high expression levels of recombinant proteins, and are much less time-consuming than the generation of a stable cell line. Moreover, as far as the cDNA constructs are available, four to five different recombinant proteins can be co-expressed simultaneously in the same batch of cells using this method. It is particularly useful for reconstituting and investigating a desired receptor-G protein-AC signalling pathway in an intact cell environment. The desired cDNAs are usually subcloned into mammalian expression vectors containing either cytomegalovirus (CMV) or Rous sarcoma virus promoters, to ensure high expression levels of the recombinant proteins. Plasmid DNA used for transfection should be of high purity, such as those prepared with CsCl or Qiagen columns (Chatsworth, CA). Both human embryonic kidney 293 and simian kidney fibroblasts COS-7 cells are routinely used to investigate the regulatory properties of various AC subtypes. These two cell lines are easily transfected with a variety of reagents such as liposomes, calcium phosphate, and DEAE-dextran/chloroquine. A protocol of DEAE-dextran/chloroquine-mediated transfection modified from a previous protocol from our laboratory (37) is described below.

2.1 DEAE-dextran/chloroquine mediated transfection This transfection method requires only general tissue culture ware and very affordable and easily prepared reagents, and is a good choice for beginners in mammalian expression systems. This method can also be applied to other cell lines. Pilot transfection experiments on the cells of choice should be carried out by varying the concentration of DEAE-dextran but maintaining the concentration of chloroquine. Reporter plasmids such as (3-galactosidase driven by CMV promoter can be used for checking the transfection efficiency. Early protocols for this method require premixing plasmid DNA with DEAEdextran (38), or pre-incubating DEAE-dextran with the cells before adding DNA (39), while the protocol here is much simpler but without reducing the transfection efficiency. In general, 30-50% of 293 or COS-7 cells are transfected successfully. The procedures for transfecting 293 and COS-7 cells are very similar, and a protocol for 293 cells is given below. Minor adjustments to the conditions for COS-7 cells are described in the footnotes following each section. 226

9: Adenylyl cyclases and cyclic AMP Protocol 1. Cell seeding and transfection Equipment and reagents • • . .

100 mm tissue culture dishes 12-well culture plates Sterile pipettes Eagle's minimal essential medium (MEM) with 10% (v/v) fetal bovine serum, 100 units ml-1 penicillin, and 100 ug ml-1 streptomycin (growth medium)a • Trypsin-EDTA solution (0.5% (w/v) trypsin, 4.3 mM EDTA, 10 mM NaCI; Gibco BRL)

• 10 mM chloroquine (100X stock; Sigma) • 40 mg ml-1 DEAE-dextran (100x stock; Pharmacia) • Dimethyl sulfoxide (DMSO) • 0.1 mg ml-1 DMA solution . Phosphate-buffered saline (PBS: 8.10 mM Na2HP04, 1.47 mM KH2P04, 137 mM NaCI, 2.68 mM KCI)

A. Cell seeding 1. Culture 293 cells on 100 mm tissue-culture dishes. One plate of cells at 80-90% confluence is enough for seeding one 12-well plate. 2. Wash 293 cells grown on a 100 mm dish with 5 ml PBS, after aspirating the growth medium. 3. Add 1 ml of trypsin-EDTA solution, and incubate in CO2 incubator for 30 s.b 4. Detach the cells from the dishes by slight tapping, and resuspend thoroughly with 4 ml of growth medium. 5. Estimate the cell density, and adjust to 2 x 105 cells ml-1 with growth medium.0 6. Transfer 1 ml of cell suspension to each well of 12-well plates. Allow the cells to attach for 20-24 h before transfection.

B. Transfection 1. Prepare a series of transfection cocktails by adding DEAE-dextran and chloroquine into growth medium to final concentrations of 400 p,g ml-1 and 100 uM, respectively.d 2. Add 0.15-0.25 |xg ml-1 of each plasmid DNA to the desired transfection cocktails. Mix well. 3. Replace the growth medium in the cell wells with 1 ml of the appropriate transfection cocktails. Incubate for 1.5 h in a C02 incubator.6 4. Before the end of incubation, prepare a shocking solution of PBS with 10% DMSO. 5. Aspirate the transfection cocktails, and shock the cells with 1 ml PBS with 10% DMSO for 1 min. 6. Replace the shocking solution with 1.5 ml of PBS to wash the cells, and finally return to 1 ml of growth medium. aFor COS-7 cells, the basic growth medium is Dulbecco's modified minimal medium (DMEM). bLonger incubation with trypsin-EDTA may be required for detaching COS-7 cells. CFor COS-7 cells, 1 x 105 cells per well should be seeded. d For COS-7 cells, reduce the final concentration of DEAE-dextran to 250 Ug ml-1. 8 For COS-7 cells, shock the cells after 3.5 h of incubation with transfection cocktails.

227

Maurice K. C. Ho and Yung H. Wong The DEAE-dextran/chloroquine-mediated transfection has been widely adopted for transfecting a variety of cell lines. The transfection efficiency is approximately 30-50% for 293 and COS-7 cells. The concentrations of both reagents and the duration of incubation are critical for maintaining a balance between high transfection efficiency and a low percentage of cell death. Seeding cells into 12-well culture plates prior to transfection allows the cells to settle in an evenly distributed and sub-confluent monolayer. As there is no trypsin treatment after transfection, it considerably reduces the number of cell deaths after transfection. After transfection, the cells do not look as healthy. A recovery period of at least 24 h is necessary before any further manipulations, because a portion of transfected cells may become easily detached. Shorter incubation period with transfection cocktail may reduce cell death, but is not recommended, since the transfection efficiency would be greatly reduced. Cells with lower passage numbers are preferred for transfection. Aged cells are much more vulnerable to the transfection conditions, and also less permissive for transfection. As previously described (40), the transfection efficiency of COS-7 cells can be enhanced to 50-80% by keeping the incubation with transfection cocktails airtight. This can be accomplished for 12-well plates by sealing with Parafilm.

2.2 Adenovirus-mediated transfection DEAE-dextran/chloroquine-mediated transfection (as well as calcium phosphate- and liposome-mediated transfection) can only be applied to certain cell lines, and is not as effective for primary cultures and other cell lines. Such limitations on introducing foreign DNA into less-permissive cell types can be ameliorated by a recently developed adenovirus-mediated transfection (41). DEAE-dextran and a replication-defective mutant of human adenovirus-5 facilitate the endocytosis of plasmid DNA, and the subsequent lysis of endosomal vesicles, respectively. The simple procedure enhances the transfection efficiency to 60-90% for common cell lines such as COS-7 cells, and permits the transfection of less-permissive cultures such as rat-1 fibroblasts and rat osteoblasts (41). This technique permits the study of the importance of particular AC subtypes or other signalling components in specific cell environments. The following protocols describe the preparation of adenovirus stock and the transfection procedures. 2.2.1 Preparation of adenovirus stock Human adenovirus-5 becomes non-invasive when the E1A coding sequence is mutated (42). The dI343 mutant (a two-nucleotide deletion in the E1A region, ref. 43), B-gal-Ad5, and GPT-Ad5 mutants (the entire E1A and E1B regions replaced by the CMV promoter-linked 3-galactosidase and guanosine phosphotransferase gene, respectively; J. R. Forsayeth and P. D. Garcia, unpublished) have been constructed for adenovirus-mediated transfection. 228

9: Adenylyl cyclases and cyclic AMP These replication-defective adenovirus mutants infect and grow in 293 cells. A viral culture can be made by infecting a small isolated batch of 293 cells. Subsequent recovery of a single plaque from a plaque assay is used to create a large stock of viral suspension for continuous use. The corresponding protocols for preparing an adenovirus stock solution are described below (Protocols 2 and 3). Protocol 2.

Plaque assay and recovery of virus

Equipment and reagents • Tissue culture ware • 55°C water bath . Autoclaved glass pipettes prewarmed to 50°c . Liquid nitrogen • Adenovirus stock solution

• MEM containing 10% (v/v) fetal bovine serum, serum, 100 units ml-1 ml-1 penicillin, and 100 (ug1ml-1 streptomycin umlstreptomycin (growth (growth medium) medium) • 5% (w/v) SeaPlaque agarose in sterile water (or (orsimilar similar grade grade of oflow-melting low-melting agarose) agarose)

Method 1. Prepare nine 60 mm dishes with 1.5 x 106 293 cells with 4 ml growth medium, and grow for 24 h. 2. Dilute the virus stock solution in tenfold dilutions down to 10-12 with growth medium. Prepare at least 2 ml for each dilution. 3. Aspirate the medium, and infect the cells with 2 ml of diluted virus solutions. Try 10-5 to 10-12, and include a control plate without infection. 4. Incubate in a C02 incubator for 2.5 h. 5. Warm 36 ml of growth medium to 55°C for 30 min. Microwave to melt the 5% (w/v) agarose solution, and keep it at 55°C. 6. Transfer 4 ml melted agarose solution to the bottle of hot medium, using a prewarmed glass pipette. Mix well, and allow to cool for a while. 7. Aspirate the infection medium, and pipette 4 ml of agarosecontaining medium slowly onto the inner side of each culture dish. 8. Cool the dishes in the culture hood until the agarose has set. 9. Grow the soft agar cultures in a CO2 incubator for 5-6 days until plaques become visible. 10. Pick the plaques with a pipette, and put them into a 12-well plate containing 293 cells (2 x 105 cells per well, 1.5 ml medium per well). Incubate for 48 h in a C02 incubator. 11. Infected cells become rounded up and detached. Harvest the entire 1.5 ml of medium and cells by flushing the wells, and transfer them into a sterile centrifuge tube. 229

Maurice K. C. Ho and Yung H. Wong Protocol 2. Continued 12. Freeze-thaw the aliquots three times with liquid nitrogen. Leave one aliquot for scaling up the viral culture, and save the rest at -80°C for future use.

Protocol 3. Scaling up viral culture and titering Equipment and reagents • Sterile PBS • Autoclaved 1.5 ml microcentrifuge tubes

• See also Protocol 2

Method 1. Grow 293 cells on 100 mm plates until confluent and tightly packed. 2. Pass the cells from a 100 mm plate to a 150 mm plate and grow for 24 h. The culture should be just confluent. 3. Use one aliquot of freshly prepared virus solution to infect a confluent 150 mm dish of 293 cells (with 25 ml medium) for 48 h in C02 incubator. 4. Flush the plate by pipetting, and collect the cell suspension in a centrifuge tube. Spin down the cells at 3000 r.p.m. for 10 min. Decant the supernatant. 5. Use 1 ml PBS to resuspend the cell pellet, and transfer into an autoclaved microcentrifuge tube. 6. Freeze-thaw the cell suspension three times with liquid nitrogen. Sonicate for 5 min in a water-sonicator. 7. Spin down cell debris at 10000 r.p.m. for 10 min in a microcentrifuge, and save the supernatant containing the adenovirus. 8. Seed 5 X 105 293 cells into each well of a 6-well plate. 9. Dilute the supernatant 1:25, 1:50, 1:100, 1:200, and 1:400 with growth medium. Each dilution should have at least 3 ml of medium. 10. Aspirate the medium from the wells, and add the diluted infection medium. Leave one well without infected medium. 11. Incubate for 48 h in a CO2 incubator. The maximal dilution of virus (usually 1:50-1:100) that can infect all the cells within this period in the well should be used fortransfection. 12. Scale up the adenovirus by repeating this protocol with twenty to forty 150 mm plates. Keep the stock at -20°C for routine use.

230

9: Adenylyl cyclases and cyclic AMP 2.2.2 Transfection A brief protocol for transfecting COS-7 cells with engineered adenovirus is described below. This protocol can also be applied to other cell types as a pilot transfection for optimization of experimental conditions.

Protocol 4. Transfection of COS-7 cells Equipment and reagents • Tissue culture ware • DMEM with 10% calf serum.100 units ml-1 penicillin, and 100 ug ml-1 streptomycin (growth medium) • DMEM with 2% calf serum and antibiotics • Serum-free DMEM with antibiotics • Trypsin-EDTA solution

8 mg ml-1 DEAE-dextran (100x stock) Adenovirus stock (100x stock) 0.1 mg mr-1 DNA solution DMSO

Method 1. Grow COS-7 cells in DMEM with 10% calf serum.a Split confluent plates in the ratio of 1:4 every 3-4 days. 2. Trypsinize the cells with 1 ml of trypsin-EDTA solution, and incubate at 37°C for 2-3 min for detachment. 3. Flush 2 ml of growth medium, and disperse the cells thoroughly by pipetting. 4. Count cells in the suspension, and calculate the volume containing 4 x 106 cells. One confluent dish may yield up to 1 x 107 cells. 5. Transfer the required volume to a sterile centrifuge tube, and centrifuge the cells at 1000 r.p.m. for 5 min. Resuspend the cell pellet in 1 ml of fresh growth medium. 6. In a new tube, mix 50 ul of 8 mg ml-1 DEAE-dextran and 50 ul of adenovirus stock with serum-free DMEM before mixing the DNA samples. 7. Mix at least 2 ug of each plasmid DNA to the transfection cocktails. The total volume of the transfection cocktail, with DNA added, should be 4 ml. 8. Mix 1 ml of cell suspension with 4 ml of transfection cocktail, and then transfer the mixture into a 100 mm plate. 9. Incubate the transfection mixture in a CO2 incubator for 2.5-3 h. Cells will attach to the dish during this period. 10. Aspirate the transfection cocktail, and shock the cells with 3 ml of freshly made PBS with 10% DMSO for 1 min.

231

Maurice K. C. Ho and Yung H. Wong Protocol 4. Continued 11. Return the cells to 10 ml of growth medium, and incubate in a CO2 incubator for 20-24 h. 8 COS-7 cells appear to attach much more firmly and faster after trypsinization when using calf serum as the growth supplement.

2.2.3 Precautions and comments Since the host cells of adenovirus is 293 cells, cross-contamination of adenovirus-transfected cultures would be detrimental for keeping 293 cell stocks. If both 293 and COS-7 cells are routinely used in the same laboratory, they should be kept at least in different incubators whenever adenovirus is used. Adenovirus-infected 293 cells for preparing virus stock must also be isolated from other 293 cells in a similar fashion. A low level of contamination of 293 cell stocks is difficult to detect, but the effect would be magnified when contaminated stocks are grown to confluence. Normal 293 cells can grow to 100% confluence and become tightly packed. For adenovirus-contaminated cells, a large proportion of cells would be rounded up and loosely attached, and they should be discarded. The protocol is adapted for preparing transfection for 100 mm plates, and the number of transfected cells on each dish is enough to seed at most one and a half 12-well plates for assay. For experiments requiring six wells (half of a 12-well plate) only, 60 mm dishes are recommended for transfection, and the recipe for the transfection cocktails should be scaled down accordingly: Prepare 1.5 X 106 cells per 60 mm dish, and resuspend in 400 ul of growth medium; the volume of transfection cocktail is adjusted to 2 ml, which contains 1 (ug of each plasmid DNA. Shock the cells with 1 ml of PBS with 10% DMSO, and return the cells to 4 ml of growth medium. COS-7 cells are first resuspended in growth medium for infection, probably to increase the efficiency of viral infection. Time required for this transfection method is similar to the method previously described. The final concentration of DEAE-dextran in the transfection conditions is 80 ug ml-1. The transfection can accommodate up to five or six different plasmids. After transfection, the appearance of the cells does not look as good as the untransfected ones. Detachment of the transfected cells may take 3-5 min after adding trypsinEDTA solution. This step also removes the unabsorbed viral particles, which may interfere with the preparation of cells for immunofluorescence or biochemical studies, and diminishes the biohazard concerns. The mutant viral construct (3-gal-Ad5 constitutively expresses p-galactosidase after introduction into the cells. The transfection efficiency can thus be examined by colorimetric or chemiluminescent assays of the (3-galactosidase activity (see Chapter 8). Similarly, one could replace the p-galactosidase coding sequence with another reporter gene as the internal control for the level of protein expression. 232

9: Adenylyl cyclases and cyclic AMP

2.3 Other methods of transfection Calcium phosphate-mediated transfection (44) is traditionally applied for both transient and stable expression of recombinant proteins in certain cell lines. Effective protocols for this method can be found elsewhere (36, 45). Transfection kits based on this method are also available from various suppliers. The rigid requirements for the pH, concentration, and preparation method of the transfection mixture affect the transfection efficiency very much, and it is thus difficult to ensure a consistent experimental condition in different trials. Calcium phosphate-mediated transfection is relatively less hazardous to the transfected cells, but most primary cultures and some cell lines do not tolerate this method. Liposome-mediated transfection (46, 47) has become popular recently because it is easy to use, and facilitates the introduction of foreign DNA, from oligonucleotides to yeast artificial chromosomes (47-51), into various less permissive cell types, and even directly to animals in vivo (52). Liposomemediated transfection supports the establishment of both transient and stable transfectants. Increasing numbers of cationic detergents, with different specific modifications, are available commercially for transfecting different cell types. These reagents are much more expensive and toxic to the cells. Introduction of foreign DNA into mammalian cells can also be achieved by electroporation (53, 54). A high voltage is applied to the cells for a short period, and linearized DNA can then be transferred into the cells. This physical method permits the transfection of certain recalcitrant cell types, but it also causes more damage to the transfected cells. An electroporator, and specialized tissue culture dishes that can conduct electricity, are the necessary equipment. Fine optimization of the conditions is critical for obtaining better transfection efficiency, but this technique offers a rapid means of creating transient and stable transfectants.

3. Measurement of intracellular cAMP level Direct measurement of AC activity usually requires preparation of plasma membranes, and various purification procedures for preserving enzyme activity. Protocols for assaying AC activity of Dictyostelium (55, 56), brain membranes (57), hepatocytes (58), 293 cells (59), fall army worm ovarian Sf9 cells (60), or purified AC (61) have been described in detail. Measuring the concentration of cAMP is more convenient as a reflection of adenylyl cyclase activity. There are a number of assays using different approaches for measuring cAMP levels. In ref. 62, binding protein assays for cAMP are fully described, including the preparation of cAMP-binding protein, several cAMP extraction methods, a competitive binding assay and its interpretation, and the application to measuring AC activity. Radioimmunoassays are mentioned elsewhere in detail (63), and there are a number of 233

Maurice K. C. Ho and Yung H. Wong related assay kits supplied commercially. Metabolic labelling of intracellular adenosine nucleotides using [3H]adenine and the subsequent isolation of [3H]cAMP are described in the following five protocols.

3.1 Metabolic labelling of ATP The intracellular pool of ATP can be labelled with radioisotopes by incubating cells with growth media containing [2-3H]ATP, [o-32P]ATP, or [23 H]adenine (64, 65). Stimulation of adenylyl cyclase converts labelled ATP into cAMP. After extraction by trichloroacetic acid, different pools of labelled nucleotides are then separated by sequential ion exchange chromatography. cAMP levels in the cells treated with various test reagents are compared with the untreated group. Protocol 5.

Labelling DEAE-dextran-transfected cells

Equipment and reagents • Eppendorf repeating pipette and sterile 12.5 ml Combitips (Fisher Scientific Company) • Aerosol-free micropipette tips

• MEM with 1% fetal bovine serum (labelling medium) . [2-3H]adenine (1 mCi ml-1, 15-25 Ci mmol-1, Amersham)

Method 1. Prepare a labelling solution by adding 1 (uCi of [2-3H]adenine to each millilitre of labelling medium, preferably using aerosol-free micropipette tips.a 2. Aspirate the growth medium from the wells. 3. Carefully fill up a 12.5 ml Combitip fitted to a repeating pipette with the labelling solution, and distribute 1 ml to each well of transfected cells. a Prior inactivation of endogenous PTX-sensitive G proteins, such as Gi/0, can be achieved by adding 100 ng ml-1 of PTX to the labelling solution.

Protocol 6.

Labelling adenovirus-transfected cells

Equipment and reagents See Protocol 5 Method 1. Trypsinize the transfected cells, and resuspend in serum-free DMEM at the density of 4 x 105 cells ml-1. 2. Distribute 0.5 ml of cell suspension in each well of a 12-well plate. 234

9: Adenylyl cyclases and cyclic AMP 3. Prepare another tube with enough volume of DMEM, with 2% calf serum and 2 jtCi ml-1 of [2-3H]adenine. This is the 2x labelling solution. 4. Distribute 0.5 ml of 2x labelling solution, and mix well with the transfected cells in the wells. 5. Incubate the cells in a C02 incubator overnight.

The advantages and disadvantages of using 3H or 32P for labelling have been discussed previously (65). Protocols 5 and 6 are designed to suit the two transfection methods described in Sections 2.1 and 2.2, respectively, and they can be linked to Protocol 7 for cAMP accumulation, and Protocol 8 for isolating different pools of labelled adenosine nucleotides.

3.2 [3H]cAMP assay The assay can be performed in a non-sterile but designated environment, such as a common fume hood. Reusable plastic disposable pipettes are much more economical for this purpose. The aspirator should be equipped with a 2 litre reservoir flask to collect radioactive liquid waste. Empty scintillant containers are ideal for storage of radioactive liquid waste. Assays can be completed in less than an hour, and thus can be performed without continuous CO2 supply, but Hepes buffer is preferred for maintaining the pH. Phosphodiesterase inhibitors are included during the assay, and cold cAMP is added to the stop solution to prevent rapid degradation of intracellular cAMP. The cells under extraction can be stored at — 20 °C until one is ready to separate the nucleotides. Protocol 7.

[3H]cAMP accumulation assay

Equipment and reagents • Eppendorf repeating pipette and reusable • Aspirator with reservoir to collect radio12.5 ml Combitips active liquid waste • 100 mM 3-isobutyl-1-methylxanthine (IBMX; . MEM with 20 mM Hepes, pH 7.5 (assay Sigma) in 100% DMSO (100X stock, freshly medium) prepared before use), or Ro 20-1724 {[4-(3- . Chilled 5% trichloroacetic acid with 1 mM butoxy-4-methoxybenzyl)-2-imidazolidiCAMP and 1 mM ATP (stop solution) none]; 100 mM aqueous solution; 100x stock; Research Biochemicals International)

Method 1. Freshly dissolve an appropriate amount of IBMX in DMSO.a Add 10 ul to each millilitre of assay medium. Appropriate agonists for the particular receptors to be investigated should also be added at this stage. 2. Wash the labelled cells with 1 ml of warm assay medium.

235

Maurice K. C. Ho and Yung H. Wong Protocol 7.

Continued

3. Transfer 1 ml of assay solution with IBMX and agonists to each well, and incubate with the cells at 37°C for 30 min. 4. Terminate the reaction by adding 1 ml ice-cold stop solution, using the repeating pipette and a reusable Combitip. Store the 12-well plate in the refrigerator for at least 30 min for the extraction of labelled nucleotides. a IBMX is an antagonist for the adenosine receptors, and can be substituted by Ro 20-1724 if necessary.

Every treatment should be done in triplicate, and their means and standard deviations (or standard errors of means) will be used for data interpretation (see Section 3.4). For 12-well plate setting, four different conditions can be tested. For studying stable transfectants (which usually give more robust responses) or primary cell types (which may be obtained by tedious preparations), 24-well plates are preferred. The quantity of labelling medium with 2 (2uCI ml1 [3H]adenine which is incubated with each well of cells is 0.5 ml, and the volume of assay medium is also reduced to 0.5 ml. The amount of stop solution used for 24-well plate assays is still 1 ml, because this is the bed volume of the resin packed in the ion exchange columns.

3.3 Isolation of [3H]cAMP Sequential ion exchange chromatography on Dowex and alumina columns is employed to isolate [3H]cAMP (65). This simple procedure has been widely used in the separation of cAMP from other nucleotides, and it constitutes a major part of AC and [3H]adenine assays. The principle of separation procedures has been described in detail previously in this series (62). A simplified version is briefly described in Protocols 8 and 9. Customized set-ups for analysing large number of samples are described in detail. A separate bench with large cabinets should be designated for running and storing the columns. Large amounts of water waste with a low level of radioactivity (e.g. storage tank water, =£100 c.p.m.) will be generated, and should be disposed of according to local regulations. 3.3.1 Setting-up To accommodate a large number of samples, two Perspex racks are made to hold 100 columns. A hundred holes with a diameter of 1 cm are drilled in two 40 cm X 40 cm Perspex sheets. They are separated in height by 4 cm and fixed with two 40 cm X 10 cm Perspex sheets on two sides. Such an arrangement allows one rack (with Dowex columns) to sit on top of another one (with alumina columns). New columns are fitted into the racks, and the resins are packed as described below. Large rectangular domestic plastic containers are 236

9: Adenylyl cyclases and cyclic AMP used for collecting liquid waste. Plastic disposable scintillation vials are preferred for convenience of disposal, and are usually supplied with paper racks with comparable dimensions to the column racks. Perspex racks for vials may also be made for long-term use. Large volumes of solutions are distributed to the columns during the separation process. Application of Dispensette® fluid dispensers facilitates repeated distribution of various volumes of solutions. Large-sized plastic reagent bottles are very useful, and usually fit well to the fluid dispensers. A 6 inch silicone tubing can be fitted to the rilling tube of a fluid dispenser if the container is relatively tall (e.g. 5 litre bottles). Protocol8.

Column packing

Equipment and reagents • Perspex racks for holding columns . Large rectangular domestic plastic containers • Dispensette® fluid dispenser, 10 ml (Fisher Scientific Company) • Large plastic bottles (2-5 litres) • PolyPrep columns (10 ml bed volume; BioRad) • 20 ml scintillation vials « Double-distilled water (ddH2O)

• Dowex resin (AG50-X8, hydrogen form, 100-200 mesh; BioRad) • Neutral alumina (BioRad AG7, or Sigma WN-3) . 1 N HCI • 1 M imidazole-HCI, pH 7.5 • 0.1 M imidazole-HCI, pH 7.5 (diluted from 1 M stock) • Scintillant (Optiphase HiSafe 3, Wallac)

A. Preparing Dowex columns 1. Pour —110 ml of Dowex resin into a clean 250 ml measuring cylinder. Wash the resin with ddH20 three to four times, until the supernatant is colourless and without fine suspension. 2. Add ddH20 to the washed resin to form a 50% (v/v) slurry. Pour the slurry into a 250 ml beaker, and stir the slurry continuously using a magnetic stirrer. 3. Use a 5 ml micropipette to transfer 2 ml of slurry into each Polyprep column in a rack with a bottom tray. Let the water run to waste. 4. Wash the Dowex columns with 2 ml of 1 N HCI, and then 10 ml of ddH20. 5. Store the rack of Dowex columns with the tips immersed in a tank of ddH20 until use. B. Preparing alumina columns 1. Weigh out 0.5 g of alumina in a 1.5 ml microcentrifuge tube. 2. Mark the tube at the level of the weighed alumina, and cut off the upper part.

237

Maurice K. C. Ho and Yung H. Wong Protocol 8. Continued 3. Melt the tip of a 1000 Ul micropipette tip in a flame, and stick it to the lower part of the cut microcentrifuge tube to form a scoop. 4. Use the home-made scoop to transfer 0.5 g of dry alumina to each Polyprep column in a rack with a bottom tray. 5. Wash new columns with 10 ml 1 M imidazole-HCI, pH 7.5, and then with 10 ml of 0.1 M imidazole-HCI, pH 7.5. 6. Store the columns immersed in a tank of ddH20 until use.

3.3.2. Separation of labelled nucleotides The settings in the last section can accommodate eight 12-well plates in a single run, which may take about an hour to obtain separated pools of labelled nucleotides in scintillation vials. For every set of assays, three blanks are included by running 1 ml of stop solution directly to each of three extra columns. If the sizes of assays exceed eight 12-well plates, the samples can be run on the same set of columns immediately after regeneration. Over 500 samples can be generated within a day, which should be enough for studying a large number of test conditions. Protocol9. Running the columns Equipment and reagents See Protocol 8 Method 1. Place racks of both Dowex and alumina columns on trays. Regenerate the Dowex columns with 2 ml of 1 N HCI and 10 ml of ddH2O, and the alumina columns with 10 ml of 0.1 M imidazole-HCI, pH 7.5. 2. Place the regenerated Dowex columns on top of a set of 20 ml scintillation vials containing 5 ml of scintillant. Load the entire 1 ml cell lysate onto each column, and allow the lysate to drain through the columns to be collected in the scintillation vials. 3. Add 3 ml of ddH2O to each column, and collect the eluate in the same set of vials. Cap the vials, number them as appropriate, and transfer them to a scintillation counter to measure the radioactivity. 4. Place the Dowex columns over a comparable number of regenerated alumina columns so that the Dowex eluate can drip directly onto the alumina. Add 10 ml of ddH2O to the Dowex columns, and let it run through the alumina columns to waste.

238

9: Adenylyl cyclases and cyclic AMP 5. Remove the Dowex columns, and mount the alumina columns over another set of vials containing 7.5 ml scintillant. Add 6 ml of 0.1 M imidazole-HCI, pH 7.5, to elute the labelled cAMP. 6. Cap the vials, number them as the first fraction, and transfer them to a scintillation counter to determine the radioactivity. 7. Regenerate both racks of columns as described in Step 1, and store immersed in ddH2O.

3.4 Data collection and interpretation In Protocol 9, the first fraction collected in Step 3 contains mostly [3H]ATP and [3H]ADP, and the radioactivity is in the order of 104-105 c.p.m., so 1 minute counting would be enough. The second fraction collected in Step 6 is the [3H]cAMP pool, with the radioactivity ranging from 102 to 104 c.p.m., depending on the type of treatment. A 5 minute counting period is suitable for this fraction. Blanks obtained by running 1 ml of stop solution without radioisotope should be averaged, and subtracted from all the counts before analysing the results. It may be done automatically by the latest models of scintillation counter, such as the Beckman LS6500. Results are most conveniently expressed as the ratio of the second fraction to the sum of both fractions, i.e.

A Beckman LS6500 scintillation counter (or the latest models from other suppliers) can save different sets of data in a spreadsheet format, which can easily be imported to computer programs for calculating the means and standard deviations. The counts of blanks indicate the degree of column contamination, and are especially important when the cAMP level is low. The Step 3 fraction is a fair reflection of the cell density, and thus, for relative determinations, it is not necessary to normalize the results by measuring the protein content of the samples.

4. cAMP-responsive bioluminescence assays using firefly luciferase The application of radioisotopes for monitoring cAMP levels is sensitive, but the radioactive waste requires long-term storage (especially for tritiated substances), and the radioactivity always causes potential hazards. Non-radioactive cAMP assays have been developed using cAMP-responsive reporter enzymes, and colorimetric or bioluminescent substrates. One of the well known reporter enzymes is firefly (Photinus pyralis) luciferase, which catalyses 239

Maurice K. C. Ho and Yung H. Wong the oxidation of its substrate, D-luciferin, in an ATP-dependent fashion, and green light is emitted during the process:

Cloning of the firefly luciferase gene (LUC) (66) allowed genetic manipulations to control the expression of luciferase in heterologous systems. A cAMP-responsive transcription of a luc gene construct has been adopted for investigating intracellular cAMP levels regulated by various G proteincoupled receptors (67-69). Detailed information on this technique has been reviewed recently (121; see also Chapter 8). The regulation of AC activity is usually measured in broken cells, or with purified protein. It may not truly reflect the signalling events that occur in intact cells. All the cAMP assays mentioned before are actually an evaluation of the overall accumulation of cAMP in the cells. They do not accurately reflect the changes of cAMP levels in the cells and their consequences. It would therefore be ideal to have a reporter system that allows monitoring of the changes of AC activity and cAMP levels in living cells. Mutational studies have identified a cAMP-dependent protein kinase (PKA)-inhibitable luciferase mutant (70) which can also be used as a cAMP-responsive bioluminescent indicator. Assays of cAMP-responsive expression of luciferase are described in the following section, while other methods using luciferase as an indicator for measuring cAMP levels are discussed briefly.

4.1 cAMP-responsive transcription of the firefly luciferase gene Activation of AC produces cAMP, which activates PKA. Subsequently, the cAMP-responsive element binding protein (CREB) is phosphorylated (71), which in turn regulates the transcription of various genes through the binding to cAMP-responsive elements (CREs) (72, 73). Transcription-based assays have been developed to study the receptor-mediated changes in second messengers such as cAMP, which in turn activate the expression of easily identified reporters under the control of various responsive elements. A bioluminescence assay has been successfully developed by fusing six CREs to the rabbit p-globin promoter region and luc gene as a reporter (67). A plasmid containing the CRE-luc reporter, pADneo2-C6-BGL, has been stably expressed in Chinese hamster ovary (CHO) cells (67-69). Receptor-mediated activation or inhibition of AC activity is translated into the degree of bioluminescence generated from the luciferase in the cells. Assays using transient expression of a luciferase reporter gene have also proved quite successful in some common cell lines that support high levels of gene expression. Expression of luc under the control of a-inhibin promoter into JEG-3 and COS-7 cells has been applied to study the muscarinic m4 receptor-mediated 240

9: Adenylyl cyclases and cyclic AMP Table2. Studies of the regulatory signals of adenylyl cyclase activity using luciferase receptor gene systems Regulatory signals

Effect on AC

A. G protein coupled receptors AdenosineAl Inhibitory Adenosine A2a and A2b Stimulatory Calcitonin C1a Stimulatory Dopamine D1 and D5 Stimulatory

FSH Muscarinicml Muscarinic m4 PACAP (5 subtypes) Serotonin 5HT1B B. G proteins Gai2Q205L Gai2G43V Gai2G2A Gai2G2A/Q205L Ga0Q205L

Cell line

Promoter of luc

CHO CHO CHO CHO

Reference

68 68 69 67

Stimulatory Stimulatory8 Biphasic* Stimulatory Inhibitory

6 CREs + rabbit (i-globin 6 CREs + rabbit p-globin 6 CREs + rabbit (3-globin 6 CREs + rabbit p-globin, or6or12CREs + HSV-TK CHO hGPHa JEG-3 CREof a-inhibin JEG-3 CREof a-inhibin LLC-PK1 16 CREs + MMTVAGRE CHO 6 CREs + rabbit (3-globin

76 74 74 77 69

Inhibitory Inhibitory No effect No effect Inhibitory

JEG-3 JEG-3 JEG-3 JEG-3 JEG-3

CREof a-inhibin CREof a-inhibin CRE of a-inhibin CREof a-inhibin CREof a-inhibin

75 75 75 75 75

FSH, follicle-stimulating hormone; hGPHa, human glycoprotein hormone a; HSV-TK, Herpes simplex virus thymidine kinase; MMTVAGRE, mouse mammary tumour virus promoter without glucocorticoid-responsive element. "The response is not related to phosphatidyl-inositol turnover, increase of Ca2+, or PKC activation. b Inhibitory at low agonist concentration (5=1 |A.M) and Gi2- and G0-linked, but stimulatory at high agonist concentration and possibly Gs-linked.

or a^-induced inhibition of cAMP production (74, 75). Table 2 summarizes the applications of luciferase reporter genes in studying various regulatory signals of AC activity. Transfection of luciferase reporter genes into target cells can be achieved with various effective protocols such as those mentioned in Section 2. Two simple protocols for assaying luciferase activity are briefly described below. Protocol 10. Luciferase assay3 Equipment and reagents • Sterile 48-well tissue culture plates • Non-sterile white opaque 96-well plate • Packard 9912V Top Count Microplate Scintillation and Luminescence Counter • Appropriate drugs in serum-free DMEM

• Packard Luclite® reagent (for cell lysis and luciferase reaction) • CHO cells stably expressing luciferase (CHO-/UC cells)

Method 1. Harvest CHO-/UC cells, and resuspend at a density of 8 x 105 cells ml-1. 2. Place 0.25 ml of cell suspension in each well of a 48-well plate. 241

Maurice K. C. Ho and YungH. Wong Protocol 10.

Continued

3. After 24 h, wash the attached cells twice with serum-free DMEM. 4. Treat the cells with various drugs in 100 ul serum-free DMEM for 4 h in a CO2 incubator. 5. Add 100 (ul of Packard Luclite® reagent to each well. 6. Shake the 48-well plates for 5 min. 7. Transfer the cell lysates to each well of a white opaque 96-well plate. 8. Measure the luminescence on a Packard 9912V Counter in single photon counting mode for 0.02 min per well, following 10 min adaptation in the dark. Alternative protocolb Equipment and reagents • Light-proof 96-well microtiter plates (Microlite, Dynatech Laboratories Inc., Chantilly, VA) • 96-well luminometer (Model ML-1000 from Dynatech) . Lysis buffer: 1% Triton X-100; 25 mM glycylglycine, pH 7.8; 15 mM MgSO4; 4 mM EGTA; 1 mM DTT

(Reaction buffer: 25 mM glycylglycine; 15 mM MgSO4; 4 mM EGTA; 15 mM K3P04, pH 7.8; 1 mM DTT; 2 mM ATP » Start solution: 0.2 mM luciferin in 25 mM glycylglycine, 15 mM MgS04, 4 mM EGTA, 2 mM DTT

Method 1. Suspend the transfected CHO cells at a density of 1.25 x 105 cells ml"1, and transfer 200 |xl into each well of a 96-well plate. 2. After settling for 24 h, challenge the cells with appropriate drugs diluted in serum-free DMEM for 4 h in a CO2 incubator. 3. Wash the cells twice with PBS. 4. Lyse the cells by adding 200 (j-l of lysis buffer, and shake for 5 minutes. 5. Pipette and transfer the lysates to microcentrifuge tubes. Spin for 5 min at maximum speed at 4°C. Save the supernatant on ice. 6. Vortex the supernatant briefly. Mix 25 u.1 of supernatant with 90 pil of reaction buffer in the wells of a Microlite plate (Dynatech). 7. Start the reaction by adding 50 ul of start solution. 8. Measure the luminescence using a Dynatech Model ML-1000 96-well luminometer. •Adopted from ref. 67. 6 Modified from refs. 68 and 78.

Firefly luciferase is considered a better reporter enzyme than chloramphenicol acetyltransferase (CAT) in several aspects. Luciferase protein has a 242

9: Adenylyl cyclases and cyclic AMP shorter half-life within transfected cells (79, 80), and this is more suitable for an inducible transient assay such as signal transduction events. Several modifications of the luciferase assay protocol have enhanced the sensitivity of luciferase assay to 30-1000 times higher, in the subattomole range (80), than the CAT assay. Addition of coenzyme A enhances the intensity and the duration of the glow reaction, probably through the change of reaction kinetics including the formation of luciferyl-coenzyme A (81). Assay medium containing detergents such as Triton X-100, or polymers such as polyvinylpyrrolidone, can lyse the cells more efficiently, and these reagents can also remove the more hydrophobic oxidized luciferin from the reaction mixture, which inhibits the luciferase reaction (82, 83). Furthermore, microplate-scale assay for luciferase produces consistent data within minutes. The light intensity produced is linear for the amount of luciferase over eight orders of magnitude, which makes the assay much more stable and reliable. Dual reporter gene systems have also been developed, to include two different reporter enzymes in a single system under the control of constitutive and inducible promoters, respectively. They have the advantage that the constitutively expressed reporter acts as an internal control for normalization of the inducible reporter. The application of dual reporter gene systems in transient transfection can minimize the experimental discrepancies, such as the variations between different trials or transfection efficiencies. Firefly luciferase is usually used as the inducible reporter, while CAT, (3-galactosidase, or alkaline phosphatase serves as the control reporter (84, 85). However, the differences in the corresponding reaction chemistries, rates of expression, detection limits, and other characteristics between firefly luciferase and the control reporter may reduce the performance of the assay systems. The cloning of luciferase genes from other species (Renilla reniformis and Vargula hilgendorfii) has led to the development of more compatible dual-reporter gene systems. A commercially available dual-reporter gene system using Renilla luciferase as the constitutive reporter, and a special formulation of lysis buffer, enhances the extraction and the sensitivities for assaying the activities of both luciferases (Promega, Madison, WI; ref. 86).

4.2 PKA-responsive luciferase mutant The amino acids 217-220 of firefly luciferase, VRFS, is very similar to the consensus PKA phosphorylation site, RRFS. Mutation of Val-217 of luciferase to Arg (V217R mutant) creates a PKA substrate with lower specific activity, lower pH optimum, and a slight shift of emission spectrum (87). Most interestingly, the specific activity of the V217R mutant can be greatly reduced by PKA-mediated phosphorylation (up to 80%). This luciferase mutant can thus be applied as a bioluminescent indicator for PKA or even cAMP. However, luciferase is translocated into peroxisomes (88) due to the presence of the tripeptide signalling sequence, SKL, at the C-terminus of the protein (89). 243

Maurice K. C. Ho and Yung H. Wong Deletion of these three amino acids translocates luciferase to the cytosol (90), and allows the subsequent construction of a cytosol-soluble PKA-responsive luciferase mutant which can detect and quantify PKA activation in living cells (70). In a previous study (70), the cDNA encoding the PKA-inhibitable luciferase mutant has been transfected into COS-7 cells by the calcium phosphate method with a relatively low transfection efficiency (~5%). In spite of this limitation, the observed inhibition of luciferase activity by treatment with 100 u-M 8-(4-chlorophenylthio)-cAMP was about 5-10% within 4 min. By increasing the transfection efficiency or establishing stable cell lines, the luciferase mutant may facilitate in vivo monitoring of the dynamic alteration of intracellular cAMP level by various regulatory signals.

4.3 Indirect cAMP assay using luciferase Elevation of intracellular cAMP levels activates PKA, which in turn phosphorylates a variety of cellular proteins by consuming intracellular ATP. Since the oxidation of luciferin also consumes ATP, the amount of ATP in the cells can limit the luciferase reaction. Handa and Bressan (91) have developed an indirect method to monitor the increase of cAMP levels in various cell types by measuring the reduction of luciferase activity. It has been claimed that the ATP consumption rate is directly proportional to the amount of cAMP present in the reaction mixture under the defined experimental conditions. However, the method may only be applied for a very narrow range of cAMP levels. Moreover, ACs, G protein receptor kinases, and many other signalling molecules consume ATP during signal transduction processes. The measurement of ATP consumption may thus not give a good indication of cAMP accumulation.

5. Miscellaneous systems for monitoring AC activity 5.1 cAMP-responsive transcription of the chloramphenicol acetyltransferase gene Like luciferase, CAT is widely used as a reporter enzyme for monitoring the regulation of gene transcription (92). It has been applied to monitor (32adrenergic receptor- and dopamine D2 receptor-regulated AC activity (93, 94). However, this method has not been extensively used because of the complications of handling radioisotopes and setting up chromatographic procedures. Similarly to the luciferase assay (see Section 3.2.1), cAMP-responsive elements are conjugated with a strong promoter and CAT gene to form a DNA construct for transfection. The expression level of CAT is controlled by the amount of intracellular cAMP, which may be regulated by various receptors, G proteins, or drugs such as forskolin. Subsequent extraction of CAT from the cells treated in different ways can be assayed by the amount of 244

9: Adenylyl cyclases and cyclic AMP acetylchloramphenicol synthesized from [14C]-chloramphenicol. The substrate and product of the CAT-catalysed reaction are separated on a silica gel, and autoradiographs are obtained and quantified. A brief protocol of CAT assay based on Gorman et al. (95) and Herbomel et al. (96) is suggested below. Protocol 11. Assay of CAT activity extracted from transfected cells Equipment and reagents • • . • .

Microcentrifuge Ultrasonics sonicator with a cuphorn probe Silica gel thin-layer plate X-ray film 40 mM Tris-HCI, pH 7.4; 1 mM EDTA; 150 NaCI (resuspension buffer) . 250 mM Tris-HCI, pH 7.8; 5 mM DTT; 15% glycerol (storage buffer)

• 250 mM Tris-HCI, pH 7.8 • 10 mM acetyl coenzyme A (freshly prepared) • I14C]chloramphenicol (50 mCi mmor1, Dupont NEN) • Ethyl acetate • Chloroform-methanol (19:1)

Method 1. Wash each 100 mm plate of CAT gene-transfected cells with 5 ml of resuspension buffer. Scrape the cells off into 1 ml of resuspension buffer. 2. Spin down the cells briefly in a microcentrifuge tube (3000 g, 5 min), and resuspend in 150 uJ of storage buffer. Keep on ice. Add a step of freeze-thawing if necessary. 3. Sonicate the chilled cell suspension for 2-4 min (e.g. using Ultrasonics sonicator, with intensity of 6 and 50% duty cycle). 4. Spin the suspension for 10 min in a microcentrifuge, using maximum speed at 4°C. Save the supernatant in a new microcentrifuge tube. 5. Mix the following solution to start the reaction: 40-80 uJ of supernatant; 300 ul of 250 mM Tris-HCI, pH 7.8; 16 uJ of 10 mM acetyl coenzyme A; and 0.5 uCi of [14C]-chloramphenicol. 6. Incubate the reaction mixture for 30 min at 37°C. Replenish acetyl coenzyme A if the incubation time has to be prolonged beyond 45 min. 7. Stop the reaction by adding 1 ml of ethyl acetate to extract the substrate and products. 8. Let the solvent evaporate completely in a fume hood, and dissolve the residue in 20 ul ethyl acetate. 9. Spot on a silica gel thin-layer plate to separate the substrate and the acetylated products by solvent migration (chloroform-methanol, 19:1, 1.5 h).

245

Maurice K. C, Ho and Yung H. Wong Protocol 11.

Continued

10. Expose the silica gel to an X-ray film overnight to obtain an autoradiograph. 11. Quantify the spots corresponding to the 3-acetyl form of chloramphenicol and the substrate. Express the CAT activity as the percentage conversion of chloramphenicol into its 3-acetyl form.

5.2 Other methods of cAMP-induced enzymatic activity Reporter gene systems using different cloned enzymes have also been developed, including alkaline phosphatase, p-galactosidase, p-glucuronidase, and luciferases of the sea pansy Renilla reniformis and the marine ostracod crustacean Vargula hilgendorfii. Detailed descriptions and comparisons of the use of various reporter enzymes have been reported (97-100; see also Chapter 8).

5.3 Pigment movement in Xenopus laevis melanophores This interesting system makes use of the specialized property of amphibian melanophores to monitor the change of intracellular cAMP level upon various regulatory signals. Melanophores contain organelles called melanosomes which are dispersed throughout the cytosol when the intracellular cAMP or PIP2 turnover is increased by stimulating endogenous melanocytestimulating hormone receptor or shining light on them (101, 102). Treating the melanophores with melatonin diminishes intracellular cAMP level, and causes the aggregation of melanosomes towards the centre of the cell (101, 103). Introduction of various recombinant G protein-coupled receptors of choice into melanophores by electroporation facilitates the investigation of their signalling functions by monitoring the translocation of melanosomes. This technique has been applied to studying various types of G proteincoupled receptors (Table 3). The response of melanosome translocation to the change of intracellular second messengers occurs within several minutes, and it does not require any extraction or further manipulation. Transfected melanophores can be distributed into 96-well microtiter plates for examining huge numbers of reagents by measuring changes of light transmittance (104). Alternatively, incorporation of a digital imaging system allows semi-quantification of the colour changes after various treatments, for a large cell population or even within single cells (105).

5.4 cAMP-induced inward current in Xenopus oocytes The cystic fibrosis transmembrane regulator (CFTR) gene encodes a PKAactivated Cl- channel (112). Microinjection of complementary RNAs of the CFTR gene and the genes of various signalling components into Xenopus 246

9: Adenylyl cyclases and cyclic AMP Table 3. Studies of G protein-coupled receptors regulating adenylyl cyclase activity using Xenopus laevis melanophores Receptor

Effect on AC

Pigment movement

Reference

a2-Adrenergic fJ2-Adrenergic a-Calcitonin gene related peptide8 (3-Calcitonin gene related peptide8 ChemokineCXC4 Dopamine D2S and D3 Endothelin ETC8 Melanocyte-stimulating hormone8 Melatonina.b Oxytocin8 PACAPa Prostaglandin E,a Serotonina,c Vasoactive intestinal polypeptide8

Inhibitory Stimulatory Stimulatoryd Stimulatory Inhibitory Inhibitory Stimulatory Stimulatory Inhibitory Stimulatory Stimulatory Stimulatoryd Stimulatory Stimulatory

Aggregation Dispersion Dispersion Dispersion Aggregation Aggregation Dispersion Dispersion Aggregation Dispersion Dispersion Dispersion Dispersion Dispersion

106 107 108

108 109 110 108 101,104 101,111 108

108 108 110 108

aEndogenous receptors. 6 The endogenous Xenopus melatonin receptor is defined as Melic receptor. The pharmacological profile of the endogenous serotonin receptor is close to mammalian 5HT6 receptor. dWeak stimulation observed under the treatment of 1 (uM agonist.

oocytes allows the measurement of cAMP-induced inward current by the patch-clamp technique. Xenopus oocytes express endogenously type II/IV AC, and thus their cAMP levels can be increased by either as-GTP or (3-/ complex released from other G proteins (113). This technique has been adopted to investigate the regulation of cAMP production by various G protein-coupled receptors and AC subtypes (113-115), as well as to identify the receptor activity-modifying proteins (RAMPs 1 to 3) as critical components for the signalling specificity of the calcitonin receptor-like receptor (116). Patch-clamp measurement of inward Cl- current is a relatively fast response, and the sensitive read-outs allow the comparison of weak responses. If microinjection and patch-clamp techniques are available in the laboratory, this method provides a convenient real-time measurement of the fluctuations of cAMP levels within an intact cell environment.

5.5 Functional rescue of CYRl-defective Saccharomyces cerevisiae mutant Mutational analysis of the signalling components regulating AC can also be examined in a yeast expression system. CYRl gene encodes the yeast AC, and a CYRl -defective mutant TC41-1 has been isolated (117) and applied to screen Gas-insensitive mutants of mammalian type V AC (118). Application of the TC41-1 yeast mutant could be extended to the selection of functional 247

Maurice K. C. Ho and Yung H. Wong mutants of other signalling components that regulate particular subtypes of mammalian AC. Yeast culturing is more economical than mammalian cell cultures, and thus this yeast-based growth assay system allows screening of a large number of signalling events. The growth rates of the yeast transformants reflect the rate of cAMP synthesis, and hence the AC activities. Furthermore, this method avoids the use of radioactive isotopes, and no extraction or lysis procedure is required. Multiple intermolecular interactions can also be investigated in this system by controlling the expressions of various molecular candidates with the use of inducible promoters (118). This system provides simple but versatile ways to investigate various regulatory signals for AC.

5.6 Functional rescue of the Escherichia coli Acya TP2000 mutant Like the yeast mutant, an adenylyl cyclase-deficient mutant of E. coli has been found (119,120) and adopted for studying the regulation of a soluble form of mammalian adenylyl cyclase (32). The E. coli mutant TP2000 cannot grow if maltose is the sole carbon source, as in minimal medium (M63 medium with arginine and maltose), and colonies grown on McConkey agar are pale yellow instead of red. After transformation with soluble ACs containing the Q domain of type I AC and Q domain of type II AC, the TP2000 mutant can grow on minimal medium, and the colonies turn red on McConkey agar when forskolin is included in the medium or an active mutant of Gas is coexpressed (32). This system is an alternative to the yeast system for fast screening of functional mutants of AC (in soluble forms) as well as other regulatory components.

6. Perspectives Intracellular cAMP level is a critical parameter determining the proper functions of specific cell types. For a better understanding of the physiological roles of cAMP, a real-time molecular probe is therefore necessary to monitor the dynamic changes of cAMP level within living cells. PKA is directly stimulated by cAMP binding, and it is a key enzyme for cAMP-regulated signalling events such as alterations of the activities of various transcription factors. Several innovative techniques have been developed using PKA as a dynamic real-time cAMP sensor. Fluorescent dye-conjugated PKA subunits have been injected into Aplysia sensory neurons to study the cAMP-induced dissociation of the catalytic and regulatory subunits dynamically through confocal microscopy (122). The temporal and spatial parameters of the cAMP-induced dissociation event can be measured as the change of energy transfer rate between two different fluorescent dyes conjugated to the PKA subunits. Purified PKA subunits are required for fluorescent dye labelling, and this technique can only be applied to a limited number of cells, because of the requirement for 248

9: Adenylyl cyclases and cyclic AMP microinjection. Fluorescent cAMP analogues have been synthesized which can be delivered to a large number of cells by simple diffusion. The binding of these cAMP analogues to PKA regulatory subunits results in changes of fluorescence and anisotropy signals of the conjugated fluorescent dye. This technique facilitates the direct observation of the binding of cAMP to PKA in living cells (123). In another recent study, a fluorescent dye-labelled peptide was derived from the autophosphorylation site of the regulatory subunit of PKA, and it can permeate the plasma membrane (124). cAMP-induced phosphorylation of the fluorescent peptide reduced its fluorescence intensity. It has been applied for the localization of PKA in NG108-15 cells and primary hippocampal neuron cultures. The advances of X-ray crystallography and nuclear magnetic resonance techniques provide researchers with particular 'snapshots' of the actual molecular interactions between ACs and various regulatory components such as G protein subunits. Improvements in molecular modelling algorithms are also beneficial for the prediction of the dynamics of the inter- and intermolecular interactions involving AC. Computerized studies of macromolecules like AC will certainly become one of the major research fields in the future.

References 1. Krupinski, J., Coussen, F., Bakalyar, H. A., Tang, W.-J., Feinstein, P. G., Orth, K., Slaughter, C, Reed, R. R., and Gilman, A. G. (1989). Science, 244,1558. 2. Bakalyar, H. A., and Reed, R. R. (1990). Science, 250,1403. 3. Cali, J. J., Zwaagstra, J. C., Mons, N., Cooper, D. M., and Krupinski, J. (1994). .J Biol. Chem., 269,12190. 4. Defer, N., Marinx, O., Stengel, D., Danisova, A., lourgenko, V., Matsuoka, I., Caput, D., and Hanoune, J. (1994). FEBS Lett., 351,109. 5. Feinstein, P. G., Schrader, K. A., Bakalyar, H. A., Tang, W. J., Krupinski, J., Oilman, A. G., and Reed, R. R. (1991). Proc. Natl. Acad. Sci. USA, 88,10173. 6. Gao, B. N., and Gilman, A. G. (1991). Proc. Natl. Acad. Sci. USA, 88,10178. 7. Hellevuo, K., Yoshimura, M., Kao, M., Hoffman, P. L., Cooper, D. M., and Tabakoff, B. (1993). Biochem. Biophys. Res. Commun., 192, 311. 8. Hellevuo, K., Yoshimura, M., Mons, N., Hoffman, P. L., Cooper, D. M., and Tabakoff, B. (1995). J. Biol. Chem., 270,11581. 9. Ishikawa, Y., Katsushika, S., Chen, L., Halnon, N. J., Kawabe, J., and Homey, C. J. (1992). J. Biol. Chem., 267,13553. 10. Katsushika, S., Chen, L., Kawabe, J., Nilakantan, R., Halnon, N. J., Homey, C.J., and Ishikawa, Y. (1992). Proc. Natl. Acad. Sci. USA, 89, 8774. 11. Krupinski, J., Lehman, T. C., Frankenfield, C. D., Zwaagstra, J. C., and Watson, P. A. (1992). J. Biol. Chem., 267,24858. 12. Fremont, R. T., Chen, J., Ma, H. W., Ponnapalli, M., and lyengar, R. (1992). Proc. Natl. Acad. Sci. USA, 89, 9809. 13. Fremont, R. T., Matsuoka, L, Mattei, M. G., Pouille, Y., Defer, N., and Hanoune, J. (1996).J. Biol. Chem., 271,13900.

249

Maurice K. C. Ho and Yung H. Wong 14. Watson, P. A., Krupinski, J., Kempinski, A. M., and Frankenfield, C. D. (1994). J. Biol. Chem., 269,28893. 15. Yoshimura, M., and Cooper, D. M. (1992). Proc. Natl. Acad. Sci. USA, 89,6716. 16. Fremont, R. T. (1994). In Methods in enzymology (ed. R. lyengar). Vol. 238, p. 116. Academic Press, London. 17. lyengar, R. (1993) FASEB J., 7, 768. 18. Taussig, R., and Gilman, A. G. (1995). J. Biol. Chem., 270,1. 19. Sunahara, R. K., Dessauer, C. W., and Gilman, A. G. (1996). Annu. Rev. Pharmacol. Toxicol., 36, 461. 20. Bayewitch, M. L., Avidor-Reiss, T., Levy, R., Pfeuffer, T., Nevo, I., Simonds, W. F., and Vogel, Z. (1998). J. Biol. Chem., 273, 2273. 21. Yan, S.-Z., Hahn, D., Huang, Z.-H., and Tang, W.-J. (1996). J. Biol. Chem., 271, 10941. 22. Whisnant, R. E., Oilman, A. G., and Dessauer, C. W. (1996). Proc. Natl. Acad. Sci. USA,93,662l. 23. Harry, A., Chen, Y., Magnusson, R., lyengar, R., and Weng, G. (1997). J. Biol. Chem., 272,19017. 24. Scholich, K., Barbier, A. J, Mullenix, J. B., and Patel, T. B. (1997). Proc. Natl. Acad. Sci. USA, 94,2915. 25. Yan, S.-Z., Huang, Z.-H., Rao, V. D., Hurley, J. H., and Tang, W.-J. (1997). /. Biol. Chem., 272,18849. 26. Sunahara, R. K., Dessauer, C. W., Whisnant, R. E., Kleuss, C., and Oilman, A. G. (1997). J. Biol. Chem., 272, 22265. 27. Iwami, G., Kawabe, J., Ebina, T., Cannon, P. J. Homey, C. J., and Ishikawa, Y. (1995). J. Biol. Chem., 270,12481. 28. Chen, Y., Harry, A., Li, J., Smit, M. J., Bai, X., Magnusson, R., Pieroni, J. P., Weng, G., and lyengar, R. (1997). Proc. Natl. Acad. Sci. USA, 94,14100. 29. Ebina, T., Toya, Y., Oka, N., Kawabe, J., Schwencke, C., and Ishikawa, Y. (1997). FEBS Lett., 401,223. 30. Ebina, T., Toya, Y., Oka, N., Schwencke, C., Kawabe, J., and Ishikawa, Y. (1997). J. Mol. Cell. Cardiol., 29,1247. 31. Ebina, T., Toya, Y., Kawabe, J., and Ishikawa, Y. (1997). J. Cell. Biochem., 66, 450. 32. Tang, W.-J., and Gilman, A. G. (1995). Science, 268,1769. 33. Scholich, K., Wittpoth, C., Barbier, A. J., Mullenix, J. B, and Patel, T. B. (1997). Proc. Natl. Acad. Sci. USA, 94,9602. 34. Zhang, G., Liu, Y., Ruoho, A. E., and Hurley, J. H. (1997). Nature, 386, 247. 35. Tesmer, J. J., Sunahara, R. K., Gilman, A. G., and Sprang, S. R. (1997). Science, 278,1907. 36. Sambrook, J., Fritsch, E. F., and Maniatis, T. (ed.) (1989). Molecular cloning: a laboratory manual (2nd edn.), Ch. 16. p. 16.32-16.39. Cold Spring Harbour Laboratory Press, NY. 37. Wong, Y. H. (1994). In Methods in enzymology (ed. R. lyengar). Vol. 238, p. 81. Academic Press, London. 38. Vaheri, A., and Pagano, J. S. (1965). Virology, 27,434. 39. Al-Molish, M. L, and Dubes, G. R. (1973). J. Gen. Virol., 18,189. 40. Gonzalez, A. L., and Joly, E. (1995). Trends Genet., 11, 216. 41. Forsayeth, J. R., and Garcia, P. D. (1994). BioTechniques, 17,354. 250

9: Adenylyl cyclases and cyclic AMP 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72.

Jones, N., and Shenk, T. (1979). Cell, 17,683. Hearing, P., and Shenk, T. (1985). Mol. Cell. Biol, 5,3214. Graham, F. L., and van der Eb, A. J. (1973). Virology, 52,456. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K. (ed.) (1995). Current protocols in molecular biology, Ch. 9 9.1.1.-9.1.7. Greene Publishing Associates and John Wiley & Sons, New York. Fraley, R. Subramani, S., Berg, P., and Papahadjopoulos, D. (1980). J. Biol. Chem., 255,10431. Feigner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. W., Wenz, M., Northrop, J. P., Ringold, G. M., and Danielsen, M. (1987). Proc. Natl. Acad. Sci. USA, 84,7413. Duzgunes, N., and Feigner P. L. in Methods in Enzymology (ed. N. Duzgunes) Vol. 221, p. 303. Capaccioli, S., Di Pasquale, G., Mini, E., Mazzei, T., and Quattrone, A. (1993). Biochem. Biophys. Res. Commun., 197, 818. Lamb, B. T., and Gearhart, J. D. (1995). Curr. Opin. Genet. Devel., 5,342. Lee, J. T., and Jaenisch, R. (1996). Nucl. Acids Res., 24,5054. Feigner, P. L., Tsai, Y. J., Sukhu, L., Wheeler, C. J., Manthorpe, M., Marshall, J., and Cheng, S. H. (1995). Ann. NY Acad. Sci., 772,126. Wong, T. K., and Neumann, E. (1982). Biochem. Biophys. Res. Commun., 107,584. Shigekawa, K., and Dower, W. J. (1988). BioTechniques, 6, 742. Parent, C. A., and Devreotes, P. N. (1995). J. Biol. Chem., 270, 22693. Chen, M.-Y., Long, Y., and Devreotes, P. N. (1997). Genes Devel., 11,3218. Chen, J., Devivo, M., Dingus, J., Harry, A., Li, J., Sui, J., Carty, D. J., Blank, J. L., Exton, J. H., Stoffel, R. H., Inglese, J., Lefkowitz, R. J., Logothetis, D. E., Hildebrandt, J. D., and lyengar, R. (1995). Science, 268,1166. Fremont, R. T., Jacobowitz, O., and lyengar, R. (1992). Endocrinology, 131, 2774. Levin, L. R., and Reed, R. R. (1995). J. Biol. Chem., 270,7573. Tang, W.-J., Krupinski, J., and Gilman, A. G. (1991). J. Biol. Chem., 266,8595. Smigel, M. D. (1986). J. Biol. Chem., 261,1976. Farndale, R. W., Allan, L. M., and Martin, B. R. (1992). In Signal transduction: a practical approach (1st edn) (ed. G. Milligan), p. 75. IRL Press, New York. Steiner, A. L. (1974). In Methods in enzymology (ed. J. G. Hardman, and B. W. O'Malley). Vol. 38, p. 96. Academic Press, London. Salomon, Y. (1979). Adv. Cyclic Nudeotides Res., 10, 35. Johnson, R. A., and Salomon, Y. (1991). In Methods in enzymology (ed. R. A. Johnson, and J. D. Corbin). Vol. 195, p. 3. Academic Press, London. De Wet, J. R., Wood, K. V., DeLuca, M., Helinski, D. R., and Subramani, S. (1987). Mol. Cell. Biol., 1, 725. Himmler, A., Stratowa, C., and Czernilofsky, A. P. (1993). J. Receptor Res., 13,79. Castanon, M. J., and Spevak, W. (1994). Biochem. Biophys. Res. Commun., 198, 626. George, S. E., Bungay, P. J., and Naylor, L. H. (1997). J. Neurochem., 69,1278. Sala-Newby, G., and Campbell, A. K. (1992). FEBS Lett., 307, 241. Montiminy, M. R., Gonzalez, G. A., and Yamamoto, K. K. (1990). Trends Neurosci., 13,184. Berkowitz, L. A., and Gilman, M. Z. (1990). Proc. Natl. Acad. Sci. USA, 87, 5258. 251

Maurice K. C. Ho and Yung H. Wong 73. Meinkoth, J. L., Montiminy, M. R., Fink, J. S., and Feramisco, J. R. (1991). Mol Cell Biol., 11,1759. 74. Migeon, J. C., and Nathanson, N. M. (1994). J. Biol. Chem., 269, 9767. 75. Migeon, J. C., Thomas, S. L., and Nathanson, N. M. (1994). J. Biol. Chem., 269, 29146. 76. Albanese, C., Christin-Maitre, S., Sluss, P. M., Crowley, W. F., and Jameson, J. L. (1994). Mol. Cell. Endocrinol., 101, 211. 77. Spengler, D., Waeber, C., Pantaloni, C., Holsboer, F., Bockaert, J., Seeburg, P. H., and Journot, L. (1993). Nature, 365,170. 78. Brasier, A. R., Tate, J. E., and Habener, J. F. (1989). BioTechniques, 7,1116. 79. Thompson, J. F., Hayes, L. S., and Lloyd, D. B. (1991). Gene, 103,171. 80. Pazzagli, M., Devine, J. H., Peterson, D. O., and Baldwin, T. O. (1992). Anal Biochem.,204,315. 81. Wood, K. V. (1991). In Bioluminescence and chemiluminescence: current status (ed. P. E. Stanley and L. J. Kricka), p. 11. John Wiley & Sons, New York. 82. Kricka, L. J., and DeLuca, M. (1982). Arch. Biochem. Biophys., 217,674. 83. Williams, T. W., Burlein, J. E., Odgen, S., Kricka, L. J., and Kant, J. A. (1989). Anal. Biochem., 176, 28. 84. Brasier, A. R., and Ron, D. (1992). In Methods in enzymology (ed. Wu, R.). Vol. 216, p. 386. Academic Press, London. 85. Jain, V. K., Magrath, I. T., and Shimada, T. (1992). Biotechniques, 12, 681. 86. Hannah, R. R., Jennes-Clough M. L., and Wood, K. V. (1998). Promega Notes, 65, 9. 87. Sala-Newby, G., and Campbell, A. K. (1991). Biochem. J., 279, 727. 88. Gould, S. J., Keller, G. A., and Subramani, S. (1987). J. Cell. Biol., 105,292. 89. Gould, S. J., Keller, G. A., Hosken, N., Wilkinson, J., and Subramani, S. (1987). J. Cell. Biol., 108,1657. 90. Sala-Newby, G., Kalsheker, N., and Campbell, A. K. (1990). Biochem. Biophys. Res. Commun., 172, 477. 91. Handa, A. K., and Bressan, R. A. (1980). Anal Biochem., 102, 332. 92. Rosenthal, N. (1987). In Methods in enzymology (ed. S. L. Berger, and A. R. Kimmel). Vol. 152, p. 704. Academic Press, London. 93. Lefkowitz, R. J., and Caron, M. G. (1988). J. Biol. Chem., 263,4993. 94. Montmayeur, J. P., and Borrelli, E. (1991). Proc. Natl Acad. Sci. USA, 88,3135. 95. Gorman, C. M, Moffat, L. F., and Howard, B. H. (1982). Mol. Cell. Biol., 2,1044. 96. Herbomel, P., Bourachot, B., and Yaniv, M. (1984). Cell, 39, 653. 97. Bronstein, L, Fortin, J., Stanley, P. E., Stewart, G. S. A. B., and Kricka, L. J. (1994). Anal. Biochem., 219,169. 98. Wood, K. V. (1995). Curr. Opin. Biotech., 6, 50. 99. Groskreutz, D., and Schenborn, E. T. (1996). In Recombinant proteins: detection and isolation (ed. R. Tuan). Vol. 63, p. 11. Humana Press, New Jersey. 100. Groskreutz, D., and Schenborn, E. T. (1997). In Methods in molecular biology (ed. R. Tuan). Vol. 63, p. 11. Humana Press, New Jersey. 101. Abe, K., Robison, G. A., Liddle, G. W., Butcher, R. W., Nicholson, W. E., and Baird, C. E. (1969). Endocrinology, 85, 674. 102. Daniolos, A., Lerner, A. B., and Lerner, M. R. (1990). Pigment Cell Res., 3, 38. 103. Krause, D. N., and Dubocovich, M. L. (1991). Annu. Rev. Pharmacol. Toxicol., 31, 549. 252

9: Adenylyl cyclases and cyclic AMP 104. Potenza, M. N., and Lerner, M. R. (1992). Pigment Cell. Res., 5,372. 105. McClintock, T. S., Graminski, G. F., Potenza, M. N., Jayawickreme, C. K., RobyShemkovitz, A., and Lerner, M. R. (1993). Anal. Biochem., 209,298. 106. Lerner, M. R., Potenza, M. N., Graminski, G. F., McClintock, T., Jayawickreme, C. K., and Karne, S. (1993). In Ciba foundation symposium (ed. D. Chadwick, J. Marsh, and J. Goode). Vol. 179, p. 76. John Wiley & Sons, New York. The molecular basis of smell and taste transduction. 107. Potenza, M. N., Graminski, G. F., and Lerner, M. R. (1992). Anal. Biochem., 206, 315. 108. McClintock, T. S., Rising, J. P., and Lerner, M. R. (1996). J. Cell. Physiol, 167,1. 109. Chen, W., Shields, T. S., Stork, P. J. S., and Cone, R. D. (1995). Anal. Biochem., 226,349. 110. Potenza, M. N., and Lerner, M. R. (1994). Naunyn Schmiedebergs Arch. Pharmacol.,349,n. 111. Potenza, M. N., Graminski, G. F., Schmass, C., and Lerner, M. R. (1994). /. NeuroscL, 14,1463. 112. Gadsby, D. C., and Nairn, A. C. (1994). Trends Biochem. Scl, 19,513. 113. Uezono, Y., Bradley, J., Min, C., McCarty, N. A., Quick, M., Riordan, J. R., Chavkin, C., Zinn, K., Lester, H. A., and Davidson, N. (1993). Recept. Chan., 1, 233. 114. Uezono, Y., Ueda, Y., Ueno, S., Shibuya, I., Yanagihara, N., Toyohira, Y., Yamashita, H., and Izumi, F. (1997). Biochem. Biophys. Res. Commun., 241, 476. 115. Matsumoto, M., Kaibara, M., Uezono, Y., Izumi, F., Sumikawa, K., Sexton, P. M., and Taniyama, K. (1998). Biochem. Biophys. Res. Commun., 242,484. 116. McLatchie, L. M., Fraser, N. J., Main, M. J., Wise, A., Brown, J., Thompson, N., Solari, R., Lee, M. G., and Foord, S. M. (1998). Nature, 393,333. 117. Casperson, G. F., Walker, N., and Bourne, H. R. (1985). Proc. Natl. Acad. Sci. USA, 82, 5060. 118. Zimmermann, G., Zhou, D., and Taussig, R. (1998). J. Biol. Chem., 273, 6968. 119. Roy, A., and Danchin, A. (1982). Mol. Gen. Genet, 188, 465. 120. Beuve, A., Boesten, B., Crasnier, M., Danchin, A., and O'Gara, F. (1990). /. Bacteriol., 172,2614. 121. Stratowa, C., Himmler, A., and Czernilofsky, A. P. (1995). Curr. Opin. Biotechnol.,6,574. 122. Bacskai, B. J., Hochner, B., Mahut-Smith, M., Adams, S. R., Kaang, B. K., Kandel, E. R., and Tsien, R. Y. (1993). Science, 260, 222. 123. Mucignat-Caretta, C., and Caretta, A. (1997). Biochim. Biophys. Acta, 1357, 81. 124. Higashi, H., Sato, K., Ohtake, A., Omori, A., Yoshida, S., and Kudo, Y. (1997). FEBS Lett., 414,55.

253

This page intentionally left blank

10 Analysis of the polyphosphorylated inositol lipids of Saccharomyces cerevisiae STEPHEN K. DOVE and ROBERT H. MICHELL

1. Introduction The yeast Saccharomyces cerevisiae offers several advantages to the researcher interested in determining the functions of phosphoinositides in eukaryotes. The ease with which this yeast can be genetically manipulated, coupled with knowledge of the entire nucleotide sequence of its genome, means that certain approaches can be used which cannot yet be applied to other organisms. For instance, genes that encode proteins involved in the synthesis, degradation, or effector functions of phosphoinositides can be deleted or rendered partially defective, and the phenotypes of the mutants determined. In this way, the contribution of an enzyme or lipid to the operation of a signalling system can rapidly be gauged. By screening for suppression of the phenotypes of such mutants, genes whose actions lie upstream or downstream of a mutated gene can be identified and characterized. In addition, genetic complementation studies are used to test whether the function(s) of an inactivated yeast gene can be replaced by expression of an apparently orthologous gene from another organism (such as a mammal). This type of approach was central, for instance, to the identification of many elements of the eukaryotic cell cycle machinery, and of some components of the RAS signal transduction cascade. In the phosphoinositide arena, a combination of yeast genetic studies with determination of the amino-acid sequences of mammalian phosphoinositide 3-kinases first uncovered the involvement of phosphatidylinositol 3-phosphate (PtdlnsSP) in protein trafficking (1). Other genetic studies in yeast confirmed that an important function of PtdIns(4,5)P2 is to regulate the actin cytoskeleton (2,3). The latter notion, conceived over a decade ago, is only now becoming widely accepted. Pure genetics establishes functional relationships between the products of different genes, giving a broad picture of cellular events. Rigorous biochemical

Stephen K. Dove and Robert H. Michell investigations must follow, but some past analyses of yeast phosphoinositides have not been sufficiently precise to yield unambiguous conclusions. Mareover, the polyphosphoinositide complement of yeasts is more complex than was initially thought (4). This increases the difficulty of achieving entirely unambiguous analyses and makes it necessary to reassess some earlier work.

2. Structures and nomenclature of inositol glycerophospholipids Phosphatidylinositol[PtdIns;m-l,2-diacyl-glycero-3-phospho-D-l-mvoinositol] (Figure /) is generally much the most abundant inositol glycerophospholipid in eukaryotic cells. Seven other 'simple' glycerophosphoinositidcs, all phosphorylated derivatives of Ptdlns, were identified between 1949 and 1997 (Figure 1 see ref. 5), but Ptdlns(3,4,5)P'1, the most heavily phosphorylated member of this lipid family, is almost certainly absent from yeasts. The fatty acid profiles of mammalian phosphoinosilides are usually dominated by 1stearoyl, 2-arachidonyl lipids. However, the .sn-1-acyl residues of S. cerevisiae Ptdlns, and probably of all yeast phosphoinositides, are mostly paimitoyl (C160, 30-47%) or stearoyl (C180, 20-30%) (with oleoyl residues also abundant in cells grown in synthetic media), and their sn-2-acyl groups are mainly oleoyl (C[18:], 58-70%) or palmitoleoyl (C16;1 20-30%) (6). A large body of research on inositol and phosphoinositides in yeast is reviewed in refs. 7 and 8, and Figure 2 summarizes the likely pathways of glycerophosphoinositide interconversion in yeasts. In addition, Ptdlns glycans of varying complexity serve as the membranepenetrating lipid anchors of many eukaryotic cell-surface (glyco)proteins, including some in yeast (e.g. refs 9-11). Some eukaryotic cells, notably plants and yeasts, contain complex sphingolipids with inositol-containing headgroups (12-14). This article will deal only with methods for studying the structurally simple glycerophospholipids in yeast.

Figure 1. The eight 'simple' glycerophospholipids of eukaryotic cells, Ptdlns(3,4,5)P3 appears to be absent from yeasts.

256

10: Analysis of polyphosphorylated inositol lipids o/S. cerevisiae

Figure2. Likely pathways of inositol glycerophospholipid biosynthesis and interconversion in Sacchammyces cerevisiae. The S. cerevisiae gene products that serve as some of the enzymes in these pathways are indicated.

3. Analysis of inositol glycerophospholipids: general considerations Many of the major methods that are commonly used for studying the structures, metabolism and functions of polyphospboinositides and inositol polyphosphates in eukaryolic cells have been collected in two dedicated books (15, 16), which should be consulted for details of many methods. The Irvine text (15) is particularly strong on the (physico)chemislry of the phosphoinosLtidcs. on degradative methods for structure determination, and on mass assays for phosphoinositides, related lipids, and inositol phosphates. The strengths of the Shears book (16) include protocols for studying inositol phosphate and phosphoinositide turnover in intact cells, the phosphoinositidespecific phospholipases C. the Ins(l,4.5)p, receptor's multiple isoforms, and inositol phosphate-binding proteins. Until the last few years, we only recognized the 'original' three glycerophospholipids (Ptdlns. Ptdlns4p, and PtdIns(4,5)p 2 ), which were analysed 257

Stephen K. Dove and Robert H. Michell either intact or as the glycerophosphoinositol esters (abbreviated as GroPInsPnS) that are liberated following alkaline deacylation. PtdIns4P and PtdIns(4,5)P2 (and their deacylated derivatives GroPIns4P and GroPIns(4,5)P2) have large negative charges, mainly conferred by their monoester phosphate groups. These highly charged lipids are bound avidly by the usual silica gel matrices of thin layer plates, but special plate formulations that include oxalate and EDTA allow their resolution. The derived GroPInsPns bind tightly to strong anion exchange resins, so their elution demands much higher salt concentrations than are needed to release the deacylation products of most other membrane phospholipids, such as phosphatidylcholine and phosphatidylethanolamine. HPLC thus both separates the glycerophosphoester deacylation products of the more abundant membrane phospholipids and effectively resolves the GroPInsPns. The three 'original' inositol lipids (or their derived GroPInsPns) have no phosphomonoester groups (Ptdlns), one such group (PtdIns4P), or two (PtdIns(4,5)P2). This makes their resolution relatively easy, but the discovery in the last decade of PtdlnsSP, Ptdlns5P, PtdIns(3,4)P2, PtdIns(3,5)P2, and PtdIns(3,4,5)P3 has made the situation much more complex. Now we must routinely distinguish between at least three structural isomers each of PtdlnsP and of PtdInsP2, and our chosen extraction and analysis methods must accommodate PtdIns(3,4,5)P3, which has three highly charged phosphomonoesters. Fortunately, all of the known GroPInsPns can be resolved by careful anionexchange HPLC analysis of the deacylated polyphosphoinositides, and TLC on oxalate-impregnated plates resolves at least some of the intact PtdlnsP and PtdInsP2 isomers (see below). These separations require considerable care, and rigorous identification of the lipids must be an early element of any analysis of a cell's polyphosphorylated inositol lipid complement. Precise cochromatography with authentic standards during TLC or HPLC gives very strong clues to the identities of particular lipids in cell extracts, but firm assignment of identities to lipids should ideally include detailed structural analysis of their inositol polyphosphate headgroups. If this is not possible, it must at least be demonstrated that the chosen analytical method(s) can resolve all of the known phosphoinositide isomers (for example, all three PtdInsP2 isomers). The usual strategy for determining the structures and configurations of polyphosphoinositide headgroups relies on the selective oxidative cleavage of vicinal diols in these molecules. Descriptions of the application of this method to various inositol polyphosphates and polyphosphoinositides are available elsewhere (15, 17-21). Treatment of any myoinositol (poly)phosphate or GroPInsPn with sodium metaperiodate cleaves only those C-C bond(s) that link carbons that bear unsubstituted hydroxyl groups. The oxidation products are reduced and dephosphorylated, and the resulting polyols are identified, usually by HPLC. A final resolution between stereo-isomers - for example, to 258

10: Analysis of polyphosphorylated inositol lipids o/S. cerevisiae discriminate between PtdIns(3,4)P2-derived Ins(l,3,4)P3 and Ins(l,3,6)P3 (its enantiomer) - is sometimes needed. This is achieved by treating the oxidation-derived polyols with a selective polyol dehydrogenase that metabolizes only one enantiomer (see Stephens's contribution to ref. 15). Using such information, the position(s) of the phosphate (s) on the original headgroup can usually be assigned unambiguously. The only substantial modification we have made to this protocol is to use a preparative scale Aminex column to identify polyols after periodate oxidation, reduction, and dephosphorylation rather than a Polybore (or derivative) column: the Aminex column accepts a much greater sample volume without excessive peak spreading. For details of this column and the sequence in which the various polyols elute, see ref. 21. This approach was used to determine the structures of the headgroups of the PtdIns3P and PtdIns4P of yeast (16), and also the structure of the GroPIns(4,5)P2, presumably derived by phospholipase A action from PtdIns(4,5)P2, that is liberated when glucose-starved 5. cerevisiae are treated with glucose (22). We have used these techniques to define the headgroup structures of the PtdIns(3,5)P2 (4) and PtdIns(4,5)P2 of 5. cerevisiae (unpublished). This yeast also contains a small amount of a lipid whose deacylation product co-chromatographs with [32P]GroPIns(3,4)P2 (S. Dove, unpublished); the structure of this putative PtdIns(3,4)P2 remains to be confirmed.

4. Radioactive labelling of the phosphoinositides of yeast cells Because cells only contain small amounts of the polyphosphorylated glycerophospholipids, measurement of their steady-state concentrations often relies on isotopically labelling the lipids with precursors of known specific radioactivity. [32P]PO42- and [3H]inositol are the usual choices for radioactively labelling the phosphoinositides of animal cells, and both have been successfully used in yeast (e.g. ref. 4). [2-3H]inositol, which is commercially available at high specific activity, is the compound of choice for most studies. Glucose represses inositol catabolism in yeast (see ref. 7), so this process is very slow in typical growth media. Standard media for 5. cerevisiae usually contain substantial amounts of inositol, which isotopically dilutes added [3H]inositol and decreases the specific radioactivity of the biosynthetically labelled lipids, but many strains of S. cerevisiae grow well in medium lacking added inositol. In most situations, SD {synthetic deficient) medium (see Protocol 1), which is easily prepared inositol-free, is an excellent medium in which to label yeast isotopically. Only with mutant strains that grow poorly in SD medium is it necessary to resort to labelling in YPD (yeast peptone-dextrose) medium (see Protocol 1). 259

Stephen K. Dove and Robert H. Michell Protocol 1. Preparation of yeast growth media Recipe for SD medium (adapted from ref. 23) For many yeast strains, this recipe must be supplemented with aminoacids and/or adenine and uracil (100 mg I-1 each). . Nutrients (per litre): glucose, 4-20 g; . Vitamins (per litre): myo-inositol,a 10 mg; ammonium sulfate, 3.5 g; asparagine, 1.5 g calcium pantothenate, 2 mg; niacin, 0.4 mg; • Salts (per litre): *KH2P04, 1 g; MgSO4.7H2O, pyridoxine.HCI, 400 (x.g; thiamine.HCI, 0.5 g; NaCI, 0.5 g; CaCI2.6H20, 0.5 g 400 (ug; p-aminobenzoic acid, 200 (ug; . Trace elements (per litre): H3B03, 500 ug; riboflavin, 200 (ug; biotin, 20 ug; folic acid, ZnS04.7H20, 400 (ig; FeCI3.6H2O, 200 M.g; 2 (ug Na2MoO4.2H2O, 200 ug; Kl, 100 ug; CuSO4.5H2O, 40 ug

Method Prepare SD medium as follows: 1. Prepare a 5x stock solution of salts and minerals by weighing out the powders, and then dissolving a small amount at a time into a beaker of water; insoluble precipitates may be formed if the water is added to the powder. Autoclave and store at room temperature; a small amount of precipitate may form. Open under sterile conditions. 2. Prepare mixtures of the amino acids that are essential for the growth of the strain under analysis (minus any that are used for selection) as 'Drop-Out Powders' that can be added during medium preparation. Combine 10 g leucine with 2 g each of the other amino acids. Mix for 10 min end-over-end, and store at 4°C. 3. Weigh the ammonium sulfate, glucose, asparagine, and amino acids (as drop-out powders; final concentrations of approx. 50 mg l-1 for each amino acid, but 250 mg I-1 of leucine) into a 1 litre container. 4. Add 300 ml deionized water. Dissolve. 5. Add 200 ml of a 5x stock of salts and trace elements (sterilized by autoclaving, stored long-term at room temperature). 6. Add deionized water to 1 litre. Adjust to pH 5.7. Autoclave on liquid cycle for 15 min. 7. When cool, add 1 ml of a vitamin stock (1000X concentrated, pH 7.0, filter-sterilized, stored in aliquots at -20°C). Recipe for YPD medium (from ref. 23) • Glucose, 4-20 g I-1 (4 g for radiolabelling, 20 g for growth and storage). • Bactopeptone (Difco), 20 g I-1

« Yeast extract (Difco), 10 g I-1 • Adenine, 50 mg 1-1 • Uracil, 50 mg I-1

aTo make inositol-free SD medium, omit the inositol. *For a low-phosphate medium, reduce KH2P04 to 0.1088 g I-1 (800 (uM).

260

10: Analysis of polyphosphorylated inositol lipids of S. cerevisiae The concentration of inositol in the medium is not the sole determinant of the final specific radioactivity of yeast inositol lipids. A second major factor is the rate at which S. cerevisiae makes inositol by cyclizing glucose-6-phosphate to L-inositol 1-phosphate (synonymous with D-inositol 3-phosphate: Ins3P), using the inositol 3-phosphate synthase encoded by the INO1 gene (see ref. 7). Limiting the amount of available glucose limits inositol synthesis, thus increasing [2-3H]inositol incorporation into phospholipids. In practice, this means that yeast grown in [2-3H]inositol-labelled media such as SD, which contain only the minimum amounts of glucose, amino acids and vitamins needed for growth, contains phosphoinositides of higher specific radioactivity than cells grown in labelled media with a full amino-acid complement (e.g. Synthetic Complete (SC), see ref. 23). A second consideration is for how long cells are labelled. It is desirable, whenever possible, for labelled cells to get close to isotopic equilibrium, when changes in the relative amounts of radioactivity in the various inositol lipids will accurately reflect underlying changes in the relative concentrations of the phosphoinositides. To achieve this, cells should be grown through at least 5 or 6 divisions in a [3H]inositol-labelled medium: >95% of the total yeast mass is then synthesized from precursors in the [3H]inositol-labelled medium. For most yeast strains, this is conveniently achieved by growing yeast overnight, but we have sometimes had to radiolabel fab1::LEU2 disruptant strains for up to 48 h in inositol-free SD for them to complete 5 or 6 divisions. It should also be noted that cells grow more slowly in medium containing a high concentration of [3H]inositol than in unlabelled medium; the experimental design should allow for this. Small-scale pilot experiments are helpful in measuring growth rates under such circumstances, allowing larger-scale experiments to be better controlled. Another important issue is the growth status of the cells when they are harvested and exposed to a challenge that might modify phosphoinositide metabolism, since cells in exponential growth and in stationary phase often respond differently to stimuli. For example, the hyperosmotic enhancement of PtdIns(3,5)P2 synthesis is only seen in exponentially growing cells (unpublished observations). For each type of response, therefore, it is important initially to examine cells at different stages of growth, so as to do subsequent detailed studies under optimum conditions. Whenever it is appropriate, it is preferable to harvest and challenge cells when they are in early exponential phase, since they are easier to break open at that time (see later). For an explanation of growth phases, see Table 1. The numbers given will vary with medium composition, strain, temperature, and degree of aeration, so are meant only as a rough guide. S. cerevisiae can conveniently be labelled in inositol-free SD medium containing 10 uCi mr-1 [3H]inositol, for 5-6 divisions and to a density of 1-4 X 106 cells mr-1 (see Protocol 2). This procedure incorporates ~2 X 106 d.p.m. of [3H]inositol into the glycerolipids of each millilitre of culture. Ptdlns contains 261

Stephen K. Dove and Robert H. Michell Table 1. S. cerevisiae growth phases in SD medium with 0.4-2% glucose <5 X 106 cells ml-1 0.5-1 x 107 cells ml-1 1-5 x 107 cells ml-1 >5 X 107 cells ml-1

Early exponential (log) phase Mid-log phase Late log phase Increasing numbers of cells enter G0 (stationary phase)

—95% of this label, —0.5-2% are found in each of the three most abundant polyphosphoinositides (PtdIns3P, PtdIns4P and PtdIns(4,5)P2), and there are even smaller quantities (0.01-0.09%) in PtdIns(3,5)P2 and PtdIns(3,4)P2. If each starting cell sample for a planned experiment needs to contain —20 000 d.p.m. in [3H]PtdIns(4,5)P2, take the cells from —1 ml of such a culture. Protocol 2. Radiolabelling of yeast cells with [3H]inositol and measurement of changes Equipment and reagents • [3H]inositol (specific radioactivity 20-30 Ci mrnol-1, e.g. NET-114) • Wild-type S. cerevisiae, e.g. strain YHP1 • Phase contrast microscope and haemocytometer • Bunsen burner • Orbital shaker

• Inositol-free SD medium containing 0.4% (w/v) glucose • YPD medium » Borosilicate glass tubes, e.g. Corning 99445-13 (13 mm diameter, 100 mm long) • Liquid N2

Method 1. Sterilize an inoculating loop in a 'roaring' flame. Using it, pick a yeast colony from afresh (<1 week old) agar plate. Inoculate into 20 ml of YPD medium (or SD medium deficient in the appropriate amino acid if working with a strain containing a plasmid which requires selection), and incubate for 24 h (48 hforSD) at30°C in an orbital shaker at 150 r.p.m. 2. Using standard aseptic technique, remove a sample of the yeast suspension, and determine the cell number using a haemocytometer. 3. Inoculate inositol-free SD medium to a density of 6 X 104 cells ml-1. Add 10 uCi ml-1 [3H]inositol. Incubate at 30°C for 14-18 h (or to a cell density of 1-2 X 106 cells ml~1). 4. Harvest the cells by centrifugation (1000 g) for 5 min. Remove and retain the supernatant. Suspend the cells in this supernatant (50-100 uJ for each experimental sample), and incubate in a water bath at 30°C for 10 min. 5. If necessary, challenge the cells in a manner appropriate to the desired experiment. (For example, to hyperosmotically stress the cells and activate Ptdlns3P5-kinase to varying degrees, add an equal volume of 0.8-2.2 M NaCI in inositol-free SD medium, and incubate for 2-10 min.)

262

10: Analysis of pofyphosphorylated inositol lipids of S. cerevisiae 6. Transfer a 100 ul aliquot of the experimental incubation into a borosilicate glass tube, and immediately kill the cells by adding two volumes (200 nl) of ice-cold methanol. Vortex vigorously for 10 s, then freeze in liquid N2. Cells killed with methanol can be stored at -20 C for several weeks with no detectable change in their inositol glycerolipid composition, so long as the tube is sealed to prevent evaporation of methanol.

5. Extraction of inositol lipids from yeast Polyphosphorylated inositol lipids are usually extracted from cells using acidified variants of one of two basic extraction methods originated by Folch (24), and by Bligh and Dyer (25). In both methods, the first stage of lipid extraction employs a chloroform-methanol mixture that remains monophasic when it is mixed with the water present in an appropriate volume of tissue or cell suspension. These methods work well with easily disrupted animal cells, but must be modified for use with yeast because of the barrier presented by the cell wall. Attempts to extract yeast polyphosphoinositides with such acidified chloroform-methanol mixtures, without vigorous cell disruption, recover only a small percentage of yeast polyphosphoinositides. After cells have been killed with methanol (see Protocol 2, above), therefore, the yeast cells are thawed, glass beads are added to form a slurry, and the cooled cell-bead mixture is vortexed very vigorously to break the cells (see Protocol 3). At well-chosen bead-to-liquid ratios, this treatment effectively disrupts the cell walls. A powerful vortex mixer must be used, and it helps if the contact cushion on the vortex mixer is kept wet while vortexing. Given the variables involved, the effectiveness of any particular set of conditions must be empirically checked before adoption. Microscopic examination is the best way to determine when all cells are broken; killed cells in methanol appear almost normal, but only cell debris remains after effective disruption. Exponentially growing yeast break much more easily than cells in stationary phase. Once the cells are broken, the lipids can be extracted. The extraction process must separate lipids from macromolecular and water-soluble cell components. This initially involves solubilization of the cell contents in the correct monophasic mixture of chloroform, methanol, and water, acidified with HC1. For efficient extraction by Folch-based methods, the target chloroform:methanol:water (C:M:W) ratio is —14:7:1 by volume, which is achieved by adding 20 ml of an acidified 2:1 (v/v) chlorofornrmethanol mixture for every 1 ml of water present in the sample. A single clear solvent mixture should then be obtained. The samples are briefly incubated on ice, and then aqueous acid is added, with mixing, to achieve a solvent ratio of 8:4:3 (v/v/v) C:M:W: the sample should now split cleanly into two phases. About 263

Stephen K. Dove and Robert H. Michell 40% forms the methanol-water-rich upper phase (3:48:47 C:M:W (v/v/v)) and —60% constitutes the chloroform-rich lower phase (86:14:1 C:M:W (v/v/v)). Polyphosphorylated inositol lipids have a highly polar inositol grouping and a high negative charge. The acid in the extraction mixture protonates the monoester phosphates, so reducing the polarity of the lipids and preventing them from entering the top phase. It also prevents the lipids from binding tightly to denatured cell proteins, notably by preventing the formation of bivalent cation bridges. The HC1 concentration in the final aqueous phase should be —0.6 M for systems in which PtdIns(3,4,5)P3 is not present, which include yeast, or —1.0 M for cells in which Ptdlns(3,4,5)^3 may be present. Denatured proteins collect at the interface as a coagulated mass, and must be removed as completely as possible with the upper phase. The lower phase is then transferred to a glass vial, and dried in vacuo (for example, in a Speedvap evaporator). When large amounts of yeast are extracted, it is sometimes necessary to wash the lower, chloroform-rich, phase so as to remove residual protein and water-soluble cell constituents. For this purpose, 'synthetic upper phase' is made by mixing chloroform, methanol, and 0.6 M HC1 containing 1 mM EDTA in the proportions 3:48:47 (v/v/v). The volume of this 'washing solution' that is added should be the same as that of the removed upper phase, calculated from the fact that the volume of this upper phase is exactly ten times the volume of the original sample. After vortexing and centrifuging, the top phase is again removed, and the lower phase containing the lipids is recovered. Protocol 3 describes an extraction method we find to be effective for disruption and extraction of inositides from both S. cerevisiae and S. pombe. Protocol 3. Extraction of lipids from yeast Equipment and reagents Chloroform Methanohconcentrated HCI (100:1, v/v) Methanol:0.6M HCI-1 mM EDTA (1:1, v/v) 'Synthetic upper phase' (see text above) 0.6 M HCI, 1 mM EDTA Powerful vortex mixer

• Glass vials • Acid-washed glass beads (425-600 urn, e.g. Sigma G-8772) • Bovine brain polyphosphoinositides (10 mg ml-1 in chloroform:methanol 2:1), e.g. Sigma P6023

Method 1. Thaw a suspension of yeast cells that have been labelled, killed, and stored, as described in Protocol 2. To 0.3 ml of yeast suspension (0.1 ml sample killed with 0.2 ml methanol) in a Corning (13 mm x 100 mm) glass tube (No. 99445-13: see Protocol 2), add 0.8 g of glass beads.a Vortex the yeast-bead mixture for 4x1 min, with cooling on ice for 1 min between each period of vortexing. This procedure disrupts >90% of the yeast. 2. Add 1.33 ml of ice-cold chloroform and 0.47 ml of ice-cold methanol: concentrated HCI (100:1, v/v).b Vortex for 30 s. This should produce a

264

10: Analysis of polyphosphorylated inositol lipids of S. cerevisiae

3.

4.

5.

6.

7.

clear one-phase system. If cloudiness remains, add methanol dropwise, with re-mixing, until the mixture clarifies. Incubate on ice for 15 min. Add 10 (ug of bovine brain polyphosphoinositides in <10 uJ of chlorofornrmethanol (2:1, v/v). This serves as a carrier that will maximize recovery of the radioactive lipids. Add 0.4 ml ice-cold 0.6 M HCI, and vortex for 30 s.b The samples should now have split into two phases. Centrifuge for 1 min (only) at 100 g at 4°C to minimize cross-contamination between the phases; this is particularly important when samples have been labelled with 32P, or [32P]ATP of high specific radioactivity. (Prolonging the centrifugation can make the protein form a rigid interfacial layer that is difficult to remove.) Remove and discard the protein and the most of the upper phase without disturbing the lower phase. The lower phases of samples radiolabelled with [32P] should be 'washed' several times (the number depending upon the amount of radioactivity used) with synthetic upper phase. Transfer the lower phase to a glass vial, by placing a pipette under the surface of the lower phase and drawing it off. Remove the last drops by manipulating the pipette tip to the bottom of the tube (below the glass beads). Take care not to transfer any beads or upper phase. Washing the beads leads to better recoveries, but causes problems with the subsequent HPLC analysis, for unknown reasons. This may be due to the carry-over of some interfering non-radioactive substance from the beads. Dry the lower phases in vacua using an unheated Speedvap (or other vacuum evaporator) that is connected, through a cooled trap, to a water pump. Ensure that the samples are dry before removing them from the evaporator, remembering that HCI evaporates last and can become very concentrated. Add 100 jxl methanol to each sample and dry down again. Analyse the samples directly by TLC (Protocol 5) or deacylate them in preparation for HPLC analysisc (see Protocols 6-8).

"This quantity of beads is effective for 0.3 ml of a yeast suspension-methanol mixture, handled in the Corning glass tubes used in our laboratory. Appropriate conditions must be determined for other combinations of cell volume and vessel. "The volumes of chloroform and methanohconcentrated HCI (100:1, v/v) given are correct for a 0.1 ml sample of cells that has been killed with 0.2 ml methanol. For other situations, let the volume of the original cell sample be x ml. If the sample was killed with 2x ml of methanol, then add 13.3xml of chloroform and 4.7xml of methanol-concentrated HCI (100:1, v/v). At step 3, add 4x ml of 0.6 M HCI. For situations where it is necessary to wash with synthetic upper phase add 10x ml of this after removal of the first upper phase. These dried lipids cannot be safely stored. They should immediately be deacylated or analysed by TLC. Otherwise the small amount of residual acid that invariably remains even after repeated drying with methanol can hydrolyse the lipids or catalyse phosphate group migration around the inositol ring. If phosphate migration occurs, phosphate esters that are subsequently identified may not accurately represent those originally present in the samples.

265

Stephen K. Dove and Robert H. Michell

6. Resolution and identification of yeast phosphoinositides There are several methods for the resolution and analysis of inositol glycerolipids. The most universally useful methods are HPLC on strong anion-exchange columns and TLC on oxalate-impregnated plates (26). Polyphosphoinositides do not migrate off the origin of standard silicic acid plates in the solvents used for the analysis of the more abundant cellular phospholipids.

6.1 Analysis of polyphosphoinositides by TLC TLC requires only simple equipment, and it can be used to process many samples fairly quickly. It is best used for quantitative analyses, but only after the lipids being studied have been properly identified. For example: • Each spot on the TLC plates should be deacylated. Its component lipid(s) can then be provisionally identified by HPLC (see later), and possibly also subjected to structural analysis to confirm these provisional identities. These analyses should confirm that each spot that is to be quantified contains only a single glycerolipid species. For the HPLC analyses, it is more effective to methylamine-deacylate the lipids directly on the gel scraped from the TLC plates than to extract the lipid from the silica gel before deacylation. • Each spot should be checked for the presence of inositol-containing phosphosphingolipids (see Section 2) whose amide-linked fatty acid residues resist hydrolysis by methylamine. This check is done by scintillation counting a sample of the upper butan-l-ol:petroleum etherethyl formate phase that is obtained after deacylation by methylamine (see Protocol 7). This organic solvent phase contains the hydrolysis-resistant (i.e. non-saponifiable) lipids, and is usually discarded. Since yeast sphingolipids usually incorporate little radioactivity during short periods of labelling with 32Pi, this check is particularly important for samples that have been labelled with [3H]inositol for prolonged periods. • For quantitative experiments, care must be taken to ensure that all of the extracted lipid is spotted onto the TLC plate. After the first drying of the lipid extract, lipid adhering to the tube walls is washed to the bottom of the tube with three rounds of dissolution in 100 u1 of chloroformrmethanol (2:1, v/v) and re-drying. The concentrated lipid sample can then be dissolved in a small volume of solvent for transfer onto the TLC plate. For TLC to be used as a quantitative procedure, the position of each inositol lipid-containing 'spot' must be precisely defined, and its radioactivity determined. 32P is easily detected by autoradiography or with a Phosphorimager, but these methods are often too insensitive to allow direct detection of 3 H-labelled polyphosphoinositides, for which different strategies must be 266

10: Analysis of polyphosphorylated inositol lipids of S. cerevisiae adopted. One direct approach is to 'spike' some or all samples with a very small amount of mixed 32P-labelled lipids from yeast (isolated as in Protocol 3, from cells labelled as described in Protocol 4); lipid distribution on the chromatogram can then be visualized autoradiographically (for example, Figure 3). 32 P-labelled and 3H-labelled yeast lipids, which have the same composition and fatty acid profiles, will have the same chromatographic characteristics. Alternatively, mammalian inositol lipids can be used. These are commercially available unlabelled or 3H-labelled, and may also be prepared 32Pradiolabelled by modifying some of the methods found in ref. 27. They can either be detected through their radioactivity or else be added in sufficient quantity for direct detection. Mammalian inositides have polyunsaturated fatty acid profiles, so they may behave slightly differently during chromatography, and TLC can be a notoriously variable technique, especially between different laboratories. When the methods are being set up, therefore, the reference lipids being used should be checked for co-chromatography with 32 P-radiolabelled yeast lipids (made as described in Protocol 4). We find such mammalian lipids to be acceptable standards for routine analysis in the TLC system described below. Protocol 4.

32

P labelling of yeast lipids as standards

This method yields 32P-radiolabelled lipid standards for TLC and HPLC. For an explanation of how extraction solvent volumes are calculated, see Protocol 3. The yeast should not be handled at densities above 5-10 x 107 cells ml-1, when they start to enter stationary phase. Caution!! This protocol describes procedures using hazardous amounts of ionising radiation. Observe good laboratory and radiological practice. Samples containing 32P should be handled, preferably with manipulators, behind thick Perspex (Plexiglass) shielding. Safety goggles and gloves are obligatory. Regularly monitor the experimental area, samples, and gloves during and after the experiment, with a Geiger counter. Equipment and reagents • 1 mCi [32P]PO,2~, free of acid and carrier (e.g. PBS13, Amersham) • 250 ml low phosphate SD medium (see above). For most yeast strains, this should be prepared with a Pi concentration of -800 p.M (see Protocol 1) • Phase contrast microscope and haemocytometer

• 150 ml phosphate-free SD medium . 200 ml 2:1 (v/v) methanohwater • 20 ml chloroform:methanol:0.6 M HCI-1 mM H3PO4-1 mM EDTA (3:48:47, v/v/v) . Acid-washed 425-600 um diameter glass beads (e.g. Sigma G-8772) • Falcon 50 ml centrifuge tubes (cat. no. 352070)

Method 1. Using a flamed wire inoculating loop, inoculate 50 ml of YPD medium with a small amount of wild-type yeast strain (e.g. YHP1) from a fresh

267

Stephen K. Dove and Robert H. Michell Protocol 4.

Continued

agar plate. Grow overnight at 30°C and 130 r.p.m. Using standard aseptic technique, remove a sample of the yeast suspension, and determine the cell number using a haemocytometer. 2. Inoculate 250 ml of low phosphate SD medium to 5 x 104 cells ml-1. Incubate at 30°C and 130 r.p.m. to a density of 1.5 X 106 cells ml-1. 3. Harvest the cells by centrifugation (1000 g for 5 min), and suspend in 25 ml phosphate-free SD medium.3 Sediment the cells by centrifugation, and repeat this 'washing' procedure. 4. Suspend the cells in 3 ml phosphate-free SD medium in a 50 ml polypropylene Falcon centrifuge tube. Incubate in a water bath (with agitation) at 30°C for 10 min. 5. Add 1 mCi [32P]P042~and continue the incubation for 20-40 min. This labels the cells highly, but not to isotopic equilibrium. If necessary, the yeast may then be challenged to maximize labelling of some particular lipid. For example, addition of an equal volume of 2.2 M NaCI in phosphate-free SD medium for 5-10 min activates Ptdlns3P5kinase, and maximizes Ptdlns(3,5)P2 labelling (4). 6. Kill the cells by adding two volumes of ice-cold methanol, and centrifuge (1000 g, 5 min). Remove and discard the 32P-labelled supernatant (Caution!! These supernatants contain most of the 32P and are hazardous). Wash the cell pellet with 50 ml 2:1 (v/v) methanoliwater. Centrifuge again. Repeat this 'washing procedure' twice more. 7. Add 3 ml 2:1 (v/v) methanokwater and 12 g glass beads.3 Alternately, vortex four times for 1 min (or until all the yeast are disrupted) and cool on ice. 8. Add 13.3 ml of ice-cold chloroform, and 4.7 ml of ice-cold methanol: concentrated HCI (100:1, v/v). Vortex for 30 s and incubate on ice for 15 min. Add 4 ml ice-cold 0.6 M HCI-1 mM EDTA, vortex for 30 s, and centrifuge at 100 g for 2 min. 9. Remove and discard the upper phase and interfacial protein. Add 10 ml chloroform:methanol:0.6 M HCI (3:48:47, v/v/v) (inclusion of 1 mM H3P03 and 1 mM EDTA in the 0.6N HCI facilitates the washout of 32P, and chelation of bivalent cations) and vortex for at least 30 s. Centrifuge at 100 g and remove the upper phase. Repeat the washing procedure once more. 10. Transfer the lower phase (being careful not to transfer any glass beads) to a 20 ml glass scintillation vial or other glass vial. Dry in vacua in a Speedvap, connected through a cooled glass trap to a water pump. 268

10: Analysis of polyphosphorylated

inositol lipids of S. cerevisiae

11. Analyse directly by TLC (Protocol 5), or deacylate and analyse by HPLC (see Protocols 6-8). aThis ratio of glass beads to liquid should be empirically determined for each new situation (see Protocol 3). The quantity quoted here works with 7 x 108 cells in 3 ml of extraction medium in the specified 50 ml Falcon tube.

The best strategy is to spike each yeast sample with ~20 ug of brain polyphosphoinositides, or run polyphosphoinositide-spiked lipids in two or three reference lanes. Since other lipids present in samples influence the TLC migration of lipids, standards run in reference lanes should be mixed with unlabelled yeast lipids similar in composition to the test samples being studied. To visualize the standards, dried chromatograms are stained with iodine vapour. Protocol 5. Thin layer chromatography of polyphosphorylated inositol lipids (based on ref. 26) Equipment and reagents • LK6D (Whatman) or silica gel 60 (Merck or Sigma) TLC plates • Oxalate-EDTA solution. Add 1 g potassium oxalate and 59 mg EDTA (di-sodium) to 50 ml water. Stir until dissolved. Mix with 50 ml methanol . Oven (120°C) . TLC tank • Filter paper (Whatman 3MM) • Parafilm • BioRad gel loader tips, or 5 ul capillary tubes

Developing solvent mixture: CHCI3:CH3OH: 15 M NH4:H20 (90:70:4:16, by volume), made up in a glass bottle and mixed thoroughly Cylinder of 02-free N2 • Lipid standards: either [32P]-labelled yeast lipids, or unlabelled mammalian brain polyphosphoinositides (see text above) Radioactive ink (if plates are to be autoradiographed, or imaged in a Phosphorimager), or an 'iodine tank' (a TLC tank that has been equilibrated for at least 2 days with 3 g of solid I2)

Method 1. If lipids are to be detected by staining, prepare an 'iodine tank' at least 1-2 days before commencing the experiment. This is a well-sealed TLC tank containing a generous scattering of I2 crystals (~3 g), held at room temperature. 2. Wearing gloves, immerse LK6D or silica gel 60 TLC plates (5 x 20 cm, or 20 x 20 cm), one by one, into 100 ml of oxalate-EDTA. Gently agitate for 20 s. Remove and dry in a fume hood. Activate in a 120°C oven for 2 h. Cool over desiccant in a sealed container. 3. Place a piece of filter paper, cut to the same dimensions as the TLC plate, vertically in a clean TLC tank. Add developing solvent to a depth of 0.3-0.5 cm, running it over the filter paper so that the wetted filter sheet adheres to the tank wall, and its bottom edge remains immersed in the solvent in the upright tank. Replace the lid, and seal with Parafilm to limit solvent evaporation. 269

Stephen K. Dove and Robert H. Michell Protocol 5.

Continued

4. Especially for quantitative experiments, care must be taken to ensure that all of the extracted lipid is spotted onto the TLC plate. After the first drying of the lipid extract (Protocol 3 or 4), lipid adhering to the tube walls is washed to the bottom of the tube with three rounds of dissolution in 100 ul of chloroform:methanol (2:1, v/v), and redrying. 5. The concentrated lipid sample is then dissolved in 50 ul chloroform: methanol (2:1, v/v). When appropriate, add unlabelled or 32P-labelled lipid standards. Vortex gently. Incubate on ice for 10 min. Take an activated and cooled oxalate-impregnated TLC plate and, using capillary tubes or (better) BioRad gel loader tips, apply the lipid samples along a line at least 1.5 cm from the bottom edge of the plate. Gently load each sample onto as small an area as possible, drying off solvent in a stream of 02-free N2. Add a further 50 p.l of chlorofornrmethanol to each sample tube, vortex, and apply to the same spot. If standards are to be run in separate reference lanes (of which there should ideally be three, one in the centre of the plate and one at each edge), apply these to the plate. 6. Remove the lid of the TLC tank, gently place the loaded TLC plate into the solvent with the line of lipid samples across the bottom edge, and with the silica-oxalate surface facing the filter paper, and quickly replace the lid. Allow the plate to develop for ~2 h, or until the solvent front is within ~2 cm of the top edge of the plate. Remove the lid, and take out the plate and dry it in a fume hood for >1 h. Then move to either Step 6 or Step 7 below. 7. To make mass lipid standards visible. Quickly place the well-dried TLC plate in the 'iodine tank' (see Step 1 above), and replace the lid. Be quick, so as to limit escape of the iodine vapour that serves as the lipid staining agent. Take out the plate when it is 'over-stained' (—30 min after it becomes clearly stained), and allow excess I2 vapour to evaporate in a fume hood. Outline the relevant lipid spots with a needle. 8. For visualization of 32P-labelled lipid standards. Apply spots of radioactive ink to the four corners of the plate, outside the sample lanes and in an asymmetrical pattern. In a darkroom, under safelight conditions, place a sheet of X-ray film (e.g. Kodak MR) against the TLC plate, and put them into a in a film exposure cassette overnight (or longer, if necessary). Develop the film. Correctly orient the autoradiograph and, using it as a guide, outline the relevant spots with a pencil by pressing onto the autoradiograph, and so through to the TLC below. 270

10: Analysis of palyphosphorylated inositol lipids of S. cerevisiae 9. Excise each spot carefully using a clean razor blade. Transfer the excised silica gel to a piece of labelled aluminium foil, and thence into a scintillation vial. 10. Quantify radioactivity by liquid scintillation

spectrophotometry.

Figure 3. Auto radiogram of a TLC separation of the 32Prlabelled lipids from S. cerew's/ae (strain YHP1) that had been challenged with 1.1 M NaCI. Cells from 500 mi of an exponential phase culture were labelled for 5 min with 3 mCi (32P]PO42-; 1.1M NaCI was present for the final 3 min. Cells were killed and extracted as described above {Protocols 1 and 31 and lipids were resolved on a Silica Gel 60A TLC plate (Sigma).

6.2 Deacylation of inositol glycerolipids, and HPLC analysis of the resulting water-soluble GroPInsPns Anion-exchange HPLC is the most highly resolving of the chromatographic techniques available for analysing polyphosphorylated inositol derivatives. Using a standard two-pump gradient HPLC system, a single well-designed HFLC run can resolve all of the eight CjroPInsPNs that are liberated by alkaline mulhyamin-olysis or methanolysis from the currently known polyphosphoinositides (see figure 5). In addition, the glyecrol moiety of these glycerophosphoesters can be removed by a mild periodate oxidation-based procedure that was originally applied to Ptdlns by Brown and Stewart (28) and has more recently been used to determine the structures of several polyphosphoinositidcs (e.g. 4, 18, 20). The HPLC techniques described here are excellent both for resolving the inositol phosphates formed when GroPInsP n s are deglycerated, and also for separating many of the numerous inositol polyphosphates that are present in eukaryotic cells. Before phosphoinositide-derived and water-soluble Ciro/'Ins^s can be 271

Stephen K, Dove and Robert H. Micheil analysed by HPLC. the GroPIusPns must be obtained by cleaving the hydrophobic fatty acy! groups off the giycerophosphoinositol backbones of the diacylglyccrophospholipids. This is best achieved by alkaline mcthanolysis using monomethylamine (which is usually simply termed methylamtnc). This is simple, quantitative and, under controlled conditions, free of major side reactions. Bivalent metal ions in the deacylation reaction mixture, such as Oa 2 ~ and M g 2 , can promote lipid hydrolysis to free inositol phosphates, so they are chelated by inclusion in the 0.6 M HC1 (used to split the extraction into two phases) of 1 mM EDTA or 1 mM cyclohexancdiamine tetra-acc-tic acid (CDTA). Residual acid in dried lipid samples can also promote sidereactions. If acid is thought to be present, samples should be put through two or three cycles of methanol addition and drying in vacuo (see Protocol,?, step 6). Since methylamine deacylates ail of the diacylglycerophospholipids in a mixed lipid sample, of which the inositol lipids are usually a small minority, care must be taken not to 'overload" deacylation reactions with excess lipids. Approximately 200 u.g of mixed glycerophospholipids, whieh is roughly the amount present in ~108 yeast cells, can be quantitatively deacylatcd by 1 ml of methylamine reagent, prepared as described below. After the deacylation reaction is complete, unreacted methylamine is removed in vacua. The water-soluble glycerophosphoesters (including (he inositol lipid-derived GroPMnsPns) are then separated from the base-resistant lipids (sphingolipids, sterols, glycerides, etc.) by phase partition between water (which constitutes the lower phase, containing the GroPInsPns) and a butan-1-o1/pelroleum ether/ethyl formate mixture (the upper phase, containing the base-resistant lipids}. The ethyl formate neutralizes residual mclhylamine. The aqueous

Gas Regulator

Figure4. Apparatus for the preparation of methylamine solution 4to be used only in a fume hood!),

272

10: Analysis of polyphosphorylated inositol lipids of S. cerevisiae phase still contains some butan-1-ol, and is dried in vacuo. The recovered GroPInsPns are dissolved in water, neutralized with Hepes-KOH, pH 7.5, and stored frozen. Methylamine reagent is available commercially from several suppliers, but is somewhat variable in quality. It is easy to prepare from methylamine gas, which is available from several suppliers (including Aldrich, Fluka, BDH and Merck)). A small cylinder of gas yields enough reagent for several years of work: it needs a small flow regulator, and this is not interchangeable between cylinders from different manufacturers. The newly made reagent is immediately ready for use, and is stable at -20 °C for at least three years. Protocol 6. Preparation of methylamine reagent (modified from ref. 29) Equipment and reagents • Methylamine gas: 170-250 g cylinder and flow regulator (e.g. Aldrich anhydrous methylamine: cat. no. 29, 553-1) • 1 litre conical flask with a silicone rubber bung (with two holes of a size to accommodate disposable Pasteur pipettes) • Silicone rubber tubing

• A screw-top 500 ml glass bottle with volume graduations (e.g. Duran) • Solid C02 pellets (Dry Ice, Cardice) or wet ice • Mixture of 140 ml methanol + 105 ml water « 35 ml butan-1-ol

Method Caution! Methylamine gas is toxic and corrosive! The following procedure should be carried out in a fume hood. 1. Clamp the methylamine cylinder upright. Attach the flow regulator, ensuring that it is fully closed. 2. Mix the methanol and water in the 500 ml glass bottle. Immerse in an ice bucket filled with solid C02 pellets or wet ice, with the graduations above 250 ml visible. 3. Assemble the apparatus shown in Figure 4. Insert the rubber bung into the conical flask. Snugly fit two Pasteur pipettes through the holes in the bung. Connect one Pasteur pipette to the methylamine cylinder with silicone rubber tubing. Use another length of tubing to connect the other Pasteur pipette to a long-tipped Pasteur pipette, which is immersed to the bottom of the bottle containing the methanol-water mixture. 4. Open the main valve on the methylamine cylinder. Slowly open the flow regulator until methylamine gas bubbles gently through the methanol-water mixture. Continue until the volume of the solution has increased to 450 ml (~1-2 h). 5. Close the main valve of the cylinder. Allow the gas flow to cease, and then close the flow regulator and disconnect the cylinder.

273

Stephen K. Dove and Robert H. Michell Protocol 7.

Continued

6. Add 35 ml butan-1-ol to the methylamine reagent in the bottle, replace the lid securely and mix by gentle swirling. Store at -20°C. 7. Wash the conical flask and tubing thoroughly. Leave the tubing in the fume hood for at least a week to allow methylamine to evaporate; discard or keep only for use in making the methylamine reagent.

Protocol 7. Deacylation of diacylglycerolipids Equipment and reagents • . . «

Water bath at 53°C Methylamine reagent (see Protocol 6) Dried lipid extract Glass vials with polypropylene lids (e.g. Jencons 7 ml vials, cat no 694-008) . 500 mM Hepes-KOH, pH 7.5

• Speedvap (or similar) centrifugal vacuum evaporator, connected through a cooled trap to a water pump • Butan-1-ol:petroleum ether (b.p. 40-60°C): ethyl formate (20:4:1, v/v/v) . Ice

Method 1. Take a dried yeast lipid extract in a glass vial (both vial and lid must be labelled) with a tight-fitting polypropylene lid. Add 1 ml of monomethylamine reagent (to deacylate lipids from 50-100 million yeast cells plus 100 (ug of carrier brain phosphoinositides; adjust the reagent quantity for more or less cells, but never use less than 500 ul). Cap and put on ice. 2. Transfer to the water bath (53°C), and incubate for 5 rnin. Vortex to ensure complete dissolution of the lipids. Continue the incubation for 35 min, then transfer onto ice for 15 min. 3. Re-label the vials (labelling may be removed by the hot water), and remove their lids. Remove the methylamine in vacuo in a vacuum evaporator (2-3 h). The samples can be left overnight at this point. 4. Add 1 ml of the butan-1-ol:petroleum ethenethyl formate mixture, 0.95 ml water, and 50 ul 500 mM Hepes-KOH, pH 7.5. Vortex for 30 s. Centrifuge at 100 gr for 3 min. Transfer the lower phase, containing the GroPlnsPns, to a 2 ml Eppendorf tube. Add 1 ml of water to the upper phase, and vortex for 30 s. Centrifuge at 100 g for 3 min. Transfer the lower phase to the same Eppendorf tube with the same tip as was used to remove the first lower phase. Dry the combined lower phases in vacuo (2-3 h). 5. Dissolve the dried sample in 1.5 ml of water (or a smaller volume appropriate for your HPLC injection loop), and vortex. Test a small drop with universal indicator to check that its pH is close to neutral. Store at or below -20°C. We routinely aim to analyse all samples by HPLC within a couple of weeks, but samples can be stored frozen for several months without degradation. 274

10: Analysis of polyphosphorylated

inositol lipids of S. cerevisiae

7. Anion-exchange HPLC analysis of inositol lipidderived GroHnsPns from yeast Routine anion-exchange HPLC analysis of GroPInsPns is carried out on a two-pump gradient HPLC system with inert-metal (titanium) pistons and pipe-work. A 250 mm Partisphere 5 u.m SAX column (Whatman) gives excellent resolution (see Figures 5 and 6), but resolution that is adequate for most studies can be obtained with a shorter (125 mm) column and the gradients described below; GroPInsPnS will elute earlier from the 125 mm column than in the experiments shown in Figures 5 and 6. When new, these columns are supplied filled with methanol (and they should be so stored when not in use). To prevent irreversible damage, columns must be flushed with water for at least an hour before admitting any salt. In Protocol 7, a small amount of buffer (~12 mM) was added to the deacylated lipid samples to ensure that they were neutral for storage. This can slightly modify the HPLC elution characteristics. If parallel runs with standards are being undertaken, then the following two precautions ensure that the runs will line up. Pre-equilibrate the column with 80% 1.25 M (NH4)2HPO4, pH 3.8, for >1 h, and wash it with water for >1 h before analysing the first sample, otherwise the first and subsequent runs may show slightly different retention times for GroPInsPns. Secondly, adjust the samples to pH ~4 with HC1 just before injection. These precautions are unnecessary for many routine analyses, since retention times that differ by a fraction of a minute between runs will often go unnoticed, especially when fractions are being collected. Standards must be run regularly, since retention times of GroPInsPns vary slightly during the life of a column, and also with changes in back-pressure caused by variations in ambient temperature or the changing of guard cartridges. Whenever possible, the best practice is to include internal standards in important analytical runs. Anion-exchange HPLC columns are expensive, and several precautions can maximize their life and analytical resolution. The silica matrix is very slightly water-soluble, and routine use of a silica guard cartridge (also sold by Whatman) saturates the mobile phase with silicic acid and minimizes dissolution of the column matrix. The guard cartridge also traps particulates and impurities, keeping them out of the analytical column. Whatman sell an inline holder for their Partisphere pre-packed guard cartridges, which saves the effort of re-packing one's own guard cartridges. We change the guard cartridge when a guard cartridge in series with a 250 mm analytical column, and equilibrated with water, exerts a back-pressure of >1300 p.s.i. (values from Gilson system), or after two months use, or after 40 runs, whichever is reached first. A second precaution that prolongs column lifetime is always to change eluant flow rates slowly; rapid changes can compact or cause cavitation in the 275

Stephen K. Dove and Robert H. Michell

Figure5. Anion-exchange HPLC profiles of [3H]inositol-labelled and deacylated polyphosphoinositides from S. cerew's/aethat had been subjected to hyperosmotic stress. The trace in panel (a) comes from a sample separated using Gradient 1 (modified from ref. 30), which gives a fairly complete resolution of GroPlnsPns, whereas that in (b) was obtained using Gradient 2 (modified from ref. 18), which is quicker but is slightly less highly resolving.

276

10: Analysis of polyphosphorylated inositol lipids of S. cerevisiae column. We change flow rates at no more than 0.1 ml min-1 per min (e.g. >10 min to go from 1 ml min-1 to zero). An analytical column that is thus protected can yield more than 300 analyses, whereas a carelessly used column can lose resolution after less than 50 runs. In Protocol 8, we present two of the several good HPLC methods that are available for resolving the GroPInsPns derived from naturally occurring inositol lipids. Figure 5 shows HPLC chromatograms, using these two gradients, of the deacylated [3H]inositol-labelled lipids of one batch of hyperosmotically challenged S. cerevisiae, with the identities of the various GroPInsPn peaks indicated. The first gradient (Figure 5a) takes a long time, but resolves all known isomeric GroPInsPns except GroPIns4P and GroPInsSP. It is excellent for detailed analytical studies of complex lipid mixtures, and for isolating pure samples of individual glycerophosphoinositol esters for subsequent structural analysis or for use as standards. The second (Figure 5b) allows quicker sample turnaround, but does not fully resolve all isomers. It should be used only once it has been established, for instance by comparison with the previous method, that it adequately resolves all of the compounds of interest in a particular experimental situation. When setting up and calibrating these HPLC methods for the first time, authentic standards are used to establish the elution times of individual compounds. Artefacts can occasionally be formed during deacylation of lipids, and these may otherwise be mis-identified as GroPInsPns. For example, Ins(4,5)P2, which can be formed as a minor breakdown product of GroPIns(4,5)P2 during the methanolysis of yeast phosphoinositides, can be mistakenly identified as GroPIns(3,5)P2. Figures 5a and b indicate where Ins(4,5)P2 elutes in the two described HPLC systems. Ideally, separate samples of 32P-radiolabelled yeast lipids (prepared as described in Protocols 3 and 4) should each be spiked with a commercially available 3H-labelled lipid, and the mixture deacylated and resolved, so allowing the elution position of that [3H]GroPInsPn to be defined. This procedure also monitors the fidelity of lipid deacylation; if more than one 3Hlabelled peak is detected, there is a problem with deacylation (or the standard lipid). To ensure the continuing correct identification of the known GroPInsPns, the best method is for internal standards of one or two GroPInsPns to be included regularly in some samples in each experiment. Some sources for standards are listed below.

7.1 HPLC: sample injection and detection of radioactivity Delivery 'of a sample onto the column is achieved by injection into a coiled injection loop. The loop has a stated capacity when first purchased, but the volume of sample injected into a loop should not exceed half of this volume, otherwise some sample may be lost to the waste flow. There are two basic setups for injection, manual or automatic. Both rely upon a rheodyne valve, 277

Stephen K. Dove and Robert H. Michell which switches positions after the injection and delivers the sample onto the column. We use automatic injection whenever possible. When this is combined with computer control of the HPLC, and on-line flow detection of radioactivity (see below), the HPLC becomes automatic, and many runs can be processed in a 24 h period. Radioactive GroPInspns eluting from the column can be detected in two ways. The first is to connect the column outlet to an on-line flow detector (e.g. Model A-450 from Canberra-Packard). We use UltimaFlo AP (CanberraPackard) scintillation fluid with our A-450; this is both an excellent scintillation fluid and is expensive. A less costly option is to link the column, via a connector with a low dead volume (analytical size), to a fraction collector, and to collect fractions at 30 s intervals into 5 ml scintillation vials. Between 2 and 3 ml of a less costly scintillation fluid (such as Hiionic or FloScint V (Packard)) is then added to each vial, the samples are thoroughly mixed, and radioactivity is measured in a static liquid scintillation spectrophotometer. Static counting both yields lower backgrounds and allows many more counts to be accumulated from a given sample, because it is has a higher efficiency and it is possible to count individual samples for extended periods. Static counting is usually the preferred method for measuring the levels of GroPInsPns in yeast, because it greatly reduces the amount of radiolabel needed for a particular set of experiments. However, it is very labour-intensive, and it takes a long period to count the fractions from relatively few samples. In experiments in which the number of samples becomes excessive, a sensible option is to collect fractions only in those 'windows' in the elution profile that are relevant to the experiment. For example, in experiments in which it has previously been established that GroPInsPs will show little or no change, start the fraction collector just before the GroPInsP2s are due to elute. One danger of this approach is that one loses the useful correction factor that is afforded by measuring GroPIns labelling. The amount of radioactivity in GroPIns usually reflects the number of cells extracted, so it serves as a control on extraction efficiency; if there are major discrepancies in the Ptdlns radiolabel recovered across a group of samples, then data can be corrected to a standard amount of GroPIns. It can also be useful when comparing the labelling of different yeast strains, but interpretation of such data is more complex: for instance, because of differences in growth rates and therefore of rates of lipid labelling.

7.2 Commercially available standards for HPLC analysis of inositol lipid-derived GroPInsPns The number and variety of commercially available radiolabelled inositolcontaining lipids is now in decline. At the time of writing, [3H]PtdIns(4,5)P2 and [3H]PtdIns are available from Amersham International, NEN, and American Radiochemical Company (ARC), but only ARC sells [3H]PtdIns4P 278

10: Analysis of polyphosphorylated inositol lipids of S. cerevisiae and [3H]PtdIns(3,4,5)P3. ARC also sells [3H]GroPIns, [3H]GroPIns4P and [3H]GroPIns(4,5)P2 standards for HPLC; we have never purchased these, so cannot comment upon their quality. We know of no commercial sources of radiolabelled PtdIns3P, PtdIns(3,4)P2, or PtdIns(3,5)P2. Protocol 8.

HPLC analysis of GroPlnsPns

Equipment and reagents • Two-pump HPLC system with inert metal pumps (e.g. Gilson with 10Wti pump heads and titanium flow lines) • Partisphere 46 mm X 250 mm 5 (um SAX HPLC column (Whatman, cat. no. 4621 1505) • Guard cartridge holder (Whatman, cat. no. 4631 0003)

• Guard cartridges (Whatman, cat. no. 4641 0005) . Glass filter unit and 0.22 um filters (e.g. Sigma) . 1.25 M (NH4)2HPO4, adjusted to pH 3.8 with orthophosphoric acid

Method 1. Filter 2 litres of water and 2 litres of 1.25 M (NH4)2HPO4, pH 3.8, using a glass filter unit and a 0.22 um filter. 2. Start the flow of water to a 250 mm long 5 (xm SAX HPLC column at 0.1 ml min-1, using a two-pump HPLC system in which solvent A is filtered water, and solvent B is 1.25 M (NH4)2HPO4, pH 3.8. 3. Either load fraction collector racks with scintillation vials, or programme an on-line flow detector appropriately (checking that it has an adequate supply of scintillation fluid for the planned analytical run(s)). 4. Programme the HPLC controller with either Gradient 1 or Gradient 2 (see below). 5. Increase the flow rate to 1 ml min-1 over 10 min, and equilibrate with water for 20 min. The pressure should be around 900-1300 p.s.i. (250 mm column with guard cartridge on Gilson HPLC). 6. Centrifuge the sample at 15000 g for 5 min to remove particulates, and carefully transfer the supernatant to a fresh tube. 7. Turn the rheodyne valve to LOAD and fill the loop (with a stated capacity at least double the sample volume) with water using a clean blunt ended needle. Inject the sample into the loop. 8. Simultaneously turn the rheodyne valve to INJECT, and start the HPLC gradient controller, and the fraction collector or on-line flow detector. Gradient 1 (modified from ref. 30) • Reservoir A: water . Reservoir B: 1.25 M (NH4)2HP04, pH 3.8 • Flow rate 1ml min-1

• Pump programme: 0 min, 0% B; 5 min, 0% B; 65 min, 12% B; 85 min, 12% B; 110 min, 80% B; 120 min, 80% B; 125 min, 0% B

279

Stephen K. Dove and Robert H. Michell Protocol 8.

Continued

Gradient 2 (modified from ref. 18) • Reservoir A: water . Reservoir B: 1.25M (NH4)2HPO4 pH 3.8 • Flow rate: 1 ml min-1

• Pump programme: 0 min, 0% B; 5 min, 0% B; 45 min, 12% B; 52 min, 20% B; 64 min, 100% B; 80 min, 100% B; 95 min, 0% B.

References 1. Schu, P. V., Takegawa, K., Fry, M. J., Stack, J. H., Waterfield, M. D., and Emr, S. D. (1993). Science, 260, 88. 2. Homma, K., Terui, S., Minemura, M., Qadota, H., Anraku, Y., Kanaho, Y., and Ohya, Y.(1998). J. Biol. Chem., 273,15779. 3. Desrivieres, S., Cooke, F. T., Parker, P. J., and Hall., M. N. (1998). J. Biol. Chem., 273,15787. 4. Dove, S. K., Cooke, F., Douglas, M., Sayers, L., Parker, P., and Michell, R. H (1997). Nature, 390,187. 5. Michell, R. H. (1997). Essays in Biochem., 32, 31. 6. Wagner, S., and Paltauf, F. (1994). Yeast, 10,1429. 7. Paltauf, F., Kohlwein, S. D., and Henry, S. A. (1992). In The Molecular and cellular biology of the yeast Saccharomyces (eds. J. R. Pringle, J. R. Broach, and E. W. Jones) Vol 2, p. 415. Cold Spring Harbor Laboratory Press, NY. 8. Anderson, M. S., Kanipes, M. I., Jackson, J. C., Yates, J., Henry, S. A., and Lopes, J. M. (1995). Yeast, 11,187. 9. Muller, G., Schubert, K., Fiedler, F., and Bandlow, W. (1992). J. Biol. Chem., 267, 25337. 10. Muller, G., Gross, E., Wied, S., and Bandlow, W. (1996). Mol. Cell. Biol., 16,442. 11. Sipos, G., Reggiori, F., Vionnet, C., and Conzelmann, A. (1997). EMBO J., 16, 3494. 12. Hechtberger., P, Zinser, E., Saf, R., Hummel, K., Paltauf, F., and Daum, G. (1994). Eur. J. Biochem., 225,641. 13. Wells, G. B., Dickson, R. C., and Lester, R. L. (1998). J. Biol. Chem., 273,7235. 14. Dickson, R. C. (1998). Annu. Rev. Biochem., 67,28. 15. Irvine, R. F. (ed.) (1991). Methods in inositide research. Raven Press, New York. 16. Shears, S. B. (ed.) (1997). Signalling by inositides. IRL Press, Oxford. 17. Stephens, L. R., Hawkins, P. T., Carter, N., Chahwala, S. B., Morris, A., Whetton, A. D., and Downes, C. P. (1988). Biochem. J., 249, 271. 18. Stephens, L. R., Hawkins, P. T., and Downes, C. P. (1989). Biochem. J., 259, 267. 19. Stephens, L. R., Hawkins, P. T., Stanley, A. F., Moore, T., Poyner, D. R., Morris, P. J., Hanley, M. R., Kay, R. R., and Irvine, R. F. (1991). Biochem. J., 275, 485. 20. Stephens, L. R., Hughes, K. T., and Irvine, R. F. (1991). Nature, 351, 33. 21. Barker. C. J., French, P. J., Moore, A. J., Nilsson, T., Berggren, P. O., Bunce, C. M., Kirk, C. J., and Michell, R. H. (1995). Biochem. J., 306,557. 22. Hawkins, P. T., Stephens, L. R., and Piggott, J. R. (1993). J. Biol. Chem., 268,3374. 23. Sherman, F. (1991). In Methods in enzymology (ed. C. Guthrie, and G. Fink). Vol. 194, p. 3. Academic Press, London.

280

10: Analysis of polyphosphorylated inositol lipids o/S. cerevisiae 24. 25. 26. 27.

Folch, J., Lees, M., and Sloane-Stanley, G. H. (1957). J. Biol. Chem., 226,497. Bligh, E. G., and Dyer, W. J. (1959). Can. J. Biochem. Physiol., 37, 911. Gonzales-Sastri, F., and Folch-Pi, J. (1968). J. Lipid Res., 9,532. Letcher, A.J., Stephens, L. R. & Irvine, R.F. (1991). In Methods in inositide research, (ed. R. F. Irvine), p. 31. Raven Press, New York. 28. Brown, D. M., and Stewart, J. C. (1966). Biochim. Biophys. Acta, 125,413. 29. Clarke, N. G., and Dawson, R. M. C. (1981). Biochem. J., 195,301. 30. Auger K. R., Serunian, L. A., Soltoff, S. P., Libby, P. L, and Cantley, L. C. (1989). Cell, 57,167.

281

This page intentionally left blank

11 Phosphoinositide 3-kinases K. E. ANDERSON, L. R. STEPHENS, and P. T. HAWKINS

1. Introduction Phosphoinositide 3OH-kinases (PI3Ks) are enzymes which can phosphorylate one or more membrane inositol lipids in the 3-position of the inositol ring; in vitro they can make phosphoinositide(3)phosphate (PtdIns(3)P), phosphoinositide(3,4)bisphosphate (PtdIns(3,4)P2), and phosphoinositide(3,4,5)triphosphate (PtdIns(3,4,5)P3) from phosphoinositide (Ptdlns), phosphoinositide(4) phosphate (PtdIns(4)P), and phosphoinositide(4,5)bisphosphate (Ptdlns(4,5) P2), respectively. The substrate specificity of PI3Ks in the intact cell is a matter of current debate, but most evidence suggests that the preferred substrate of the sub-family of PI3Ks that are agonist sensitive (type 1 PISKs) is Ptdlns (4,5)P2, and also possibly PtdIns(4)P. Hence the rapid rise in the levels of PtdIns(3,4,5)P3 and PtdIns(3,4)P2 seen on stimulation is widely regarded as the major signal generated by this pathway. Type 1 PI3Ks play a key role in the acute signalling mechanisms used by a very large number of cell-surface receptors. All known receptors which possess intrinsic protein tyrosine kinases (e.g. receptors for growth factors, such as platelet-derived growth factor (PDGF), insulin, colony stimulating factor-1, and epidermal growth factor), and which have been studied appropriately, have been shown to activate PI3K. Furthermore, an ever-increasing number of receptors which utilize src-type protein tyrosine kinases to transduce their signals (e.g. receptors for cytokines such as interleukin (IL)-2, IL3, granulocyte macrophage colony stimulating factor (GM-CSF), or antigen receptors such as CDS and mlgM) are being found to activate this enzyme. A number of guanine nucleotide binding protein (G protein)-coupled receptors (e.g. receptors for formylated-met-leu-phe (fMLP), platelet activating factor, histamine, and ATP) present on cells of haematopoietic origin have also been found to activate a PI3K (for recent review on PISKs see ref. 1). The type 1 PISKs which are activated by protein tyrosine kinases are heterodimers that are composed of regulatory (p55-85) and catalytic (p110) subunits. The regulatory subunit has two SH2 (src-homology region two) domains, which have a very high affinity and specificity for binding of tyrosine

K. E. Anderson et al. phosphates contained within specific sequence motifs found in the cytoplasmic regions of receptors and receptor-associated proteins (these tyrosines are often sites of autophosphorylation in receptors with intrinsic protein tyrosine kinases). On appropriate cell stimulation, these tyrosine residues become phosphorylated, and PI3K becomes bound to them, thus driving the effective translocation of PI3K from the cytosol to the plasma membrane, a source of its lipid substrate. The docking of the SH2 domains of PI3Ks to specific tyrosine phosphate targets is also thought to enhance directly the intrinsic catalytic activity of the enzyme, resulting in a dual-level activation process. Although protein tyrosine kinase regulated PI3Ks are more widespread and well studied, the most intense burst of agonist-stimulated PtdIns(3,4,5)P3 synthesis is found in cells of myeloid origin, in response to agonists which use heterotrimeric G proteins. A novel form of PI3K has been purified, which can be activated by G protein (3-y subunits (2, 3). This G protein receptor controlled PI3K is a heterodimer of pllO-y catalytic (2, 4) and a novel plOl adaptor protein (5). Although G&y subunits of G proteins are known to activate the plOl/pllOy-PDK directly, it is not known where on the PI3K complex this interaction occurs. The functions of the PI3K signalling pathway are currently the subject of intense research. A number of strategies have been developed over the last few years that have enabled the identification of numerous agonist-stimulated cell responses that are regulated by this signalling pathway. These include membrane ruffling, cell movement, superoxide formation, secretion, cell division, cell survival, and glucose transport. For recent reviews on cellular responses regulated by PI3K, refer to Carpenter and Cantley (6), and Toker and Cantley (7). As yet we still know very little of the molecular mechanisms which allow PtdIns(3,4,5)P3/PtdIns(3,4)P2 to engage the regulatory systems which control these cellular events. Most is known about the activation of the key signalling intermediate protein kinase B (PKB), where PtdIns(3,4,5)P3 and PtdIns(3,4)P2 are thought to recruit both PKB and an activating phosphoinositide dependent kinase (PDK) to the plasma membrane, through the direct binding to the pleckstrin homology (PH) domains contained within the primary sequences of both of these proteins. PH domains are known to be capable of binding inositol phospholipids with varying degrees of specificity and affinity. A number of PH domains have been shown to bind PtdIns(3,4,5)P3 and/or PtdIns(3,4)P2 in vitro (e.g. those contained within Brutons tyrosine kinase (BTK), Vav, T lymphoma invasion and metastasis-1 (Tiam-1), son-of-sevenless (SOS), cytohesin, and centaurin). It is possible, therefore, that PtdIns(3,4,5)P3 and PtdIns(3,4)P2 are used as regulated signals that drive recruitment and assembly of numerous signalling complexes, which in turn serve to control downstream cellular events. 284

11: Phosphoinositide 3-kinases

2. Methods for identifying a role for the PI3K signalling pathway in cellular events Various strategies are available for inhibiting or increasing Type 1 PI3K activity in cells. They each have advantages and disadvantages, and are usually used in combination to build up a convincing case for the involvement ofPISK.

2.1 Inhibition of PI3K 2.1.1 Catalytic site inhibitors The most generally useful reagents for inhibiting PI3K are the structurally distinct catalytic-site inhibitors wortmannin and LY294002. The fungal metabolite wortmannin (available from Sigma) is a direct inhibitor of the catalytic subunit of p85/pllO-PI3K (8), and also inhibits G protein stimulated synthesis of PtdIns(3,4,5)P3 in intact cells (although reportedly with lower potency). We dissolve wortmannin at 10 mM in DMSO, and store in aliquots at -20°C (in the presence of dehydrated silica beads). Wortmannin is destroyed by prolonged exposure to aqueous solutions (e.g. >30 min) and light. As a result, wortmannin solutions are made immediately prior to use (used in the 20-200 nM range, either in vivo or in vitro), and if long term pretreatments are required, wortmannin must be readministered in fresh solution. LY294002 (also available from Sigma), which is soluble in ethanol or DMSO, acts as a competitive inhibitor for the ATP binding site of PI3K (9), and is generally used in the 5-50 jxM range on intact cells with a 5-15 min preincubation period. Both LY294002 and wortmannin are potent and cell permeable, allowing them to be used in a wide spectrum of assay formats. They are not fully characterized, however, with respect to their effects on all PI3K family members. Most agonist-stimulated PISKs are sensitive to wortmannin (although, as mentioned above, there is some evidence that those coupled to G proteins are less sensitive), and some agonist-insensitive PI3Ks are affected by wortmannin while others are not (i.e. addition of wortmannin to many cells will lower PtdIns(3)P levels; agonist-insensitive production of this lipid is thought to be involved in vesicular trafficking events from the Golgi apparatus). There is also question over the specificity of both wortmannin and LY294002 for the PI3K catalytic domain versus those of other lipid kinases and protein kinases, although it should be said that few direct, alternative targets have been found for wortmannin or LY294002 in the 20-100 nM range, or below 50 u,M, respectively. 2.1.2 Dominant negative PI3K alleles The most popular construct for this purpose is a p85 regulatory subunit of PI3K with a small deletion in the domain which binds the pi 10 catalytic 285

K. E. Anderson et al. subunit (Ap85). Thus expression of Ap85 in cells competes with endogenous PI3K for a limiting number of activating phosphotyrosines. This construct is usually used in transient transfection assays, co-transfecting with expression vectors encoding Ap85 and an output reporter (e.g. an epitope-tagged protein), or in conjunction with a single cell assay (e.g. ref 10). The major drawback is that it relies on a very high specificity for Ap85's SH2 domains for the 'correct' (i.e. PI3K specific) activating phosphotyrosine. Until the sites of interaction between G protein-regulated PI3K subunits plOl and pllO-y, and the sites where the G protein (3-y subunit interacts with one or both of the PI3K subunits have been identified, the above approach cannot be applied to inhibit these forms of PI3K. 2.1.3 Receptor mutants Receptor mutants have been created, in which the specific tyrosines known to be responsible, when phosphorylated, for docking PI3K have been changed to phenylalanines. These have been used successfully to provide evidence in a number of studies (particularly using the PDGF receptor for which a large number of 'control' constructs are available) for the involvement of the PI3K signalling pathway in a receptor's repertoire of cellular responses (e.g. refs 11, 12). This method requires a 'null' cell background in which to express the mutant constructs (usually stable cell lines are created expressing the mutant receptors), and relies on a lack of redundancy in the mechanism by which a receptor activates PI3K. It also relies on the assumption that deletion of the appropriate tyrosines only affects binding of PI3K, and not other signalling molecules. 2.1.4 Phosphorylated peptides Synthetic, phosphorylated peptides which mimic the binding sites of PI3K on receptors (e.g. the sequence surrounding Y740-Y751 of the PDGF (3receptor) have been used in a small number of studies to compete with, and hence inhibit, endogenous PI3K (e.g. ref 13). These studies utilize cell permeabilized cell systems or microinjection to administer the peptides, and require a number of control peptides to provide a convincing case for the specificity of action.

2.2 Constitutively active alleles of PI3K and downstream effectors It has become clear that one of the critical steps in activating many signalling cascades is the translocation of key intermediates to the plasma membrane. Furthermore, there are now a number of examples in which targeting these intermediates to the plasma membrane by artificial means has produced a 'constitutively active' protein, that is, where the activity of the protein is significantly above basal in the absence of the agonist (e.g. ref 14, 15). This 286

11: Phosphoinositide 3-kinases idea has been used successfully to generate activated alleles of PI3K (of both protein tyrosine kinase and G protein-regulated forms); the membrane targeting has been achieved in a number of ways, but the most popular is fusing the 'myristoylation consensus' (MGLCIKSKEDKSM) of a src-type tyrosine kinase to the N-terminus, or the 'CAAX motif and polybasic region of ras to the C-terminus. In a similar manner, very active alleles of two of PI3K targets, the protein kinases PKB and PDK, have been made using Nterminal myristoylation consensuses. These constitutively active alleles have been very useful in determining whether PI3K or PKB are sufficient to induce cellular responses (e.g. to provide a cell survival signal). Again, use of these alleles usually requires a single cell assay, co-transfection with reporter constructs, or the generation of stable cell lines. As controls for these experiments of constitutively active PI3K transfections, 'kinase dead' alleles of PDK can be used, containing mutations in the catalytic domain of the protein which destroy kinase activity. It must be appreciated that the signals generated by the 'constitutivelyactive' proteins may differ significantly from those generated by authentic receptors, for example, in the case of PI3Ks in the balance of the lipids generated (i.e. PtdIns(3,4,5)P3, PtdIns(3,4)P 2 and PtdIns(3)P), the timecourses of generation, and the precise subcellular locations in which they are made.

3. Measurement of the activation of PI3K Most workers quantify the activation of tyrosine kinase regulated PI3K by measuring agonist-stimulated association of PDK with receptors and/or tyrosine phosphorylated proteins. These techniques work well because the tyrosine phosphate-SH2 domain interaction is sufficiently strong to withstand most immunoprecipitation/washing protocols (the antibody does not necessarily need to bind to PDK directly, but can be directed to any component of the SH2-tyrosine phosphate co-ordinated signalling complex; simple antiphosphotyrosine antibodies are usually used). This interaction is revealed by lysing control and agonist-stimulated cells with detergent-based mixtures, preparing immunoprecipitates with an appropriate antibody, and measuring the agoniststimulated appearance of PDK activity (or protein) in the immunoprecipitates (see Protocol 1).

287

K. E. Anderson et al. Protocol 1. Antiphosphotyrosine antibody-directed immunoprecipitation of PI3K Equipment and reagents • Balanced salt solution (1.8 mM CaCI2; 5.37 • mM KCI; 0.81 mM MgSO4; 112.5 mM NaCI; 25.0 mM glucose; 25.0 mM N-[2-hydroxyethyl]piperazine-N'-[2-ethane-sulfonic acid] (Hepes)-NaOH, pH 7.4 at room temperature; 1.0 mM NaHCO3; 0.1% (w/v) fatty acid free-bovine serum albumin (BSA), pH 7.3 at room temperature; 0.2 nm filter, store at 4°C) . Lysis buffer (137 mM NaCI, 2.7 mM KCI, 1 mM MgCI2, 1 mM CaCI2, 1% (w/v) Nonidet P-40 (NP40), 10% (w/v) glycerol, 1 mg ml-1 • BSA, 20 mM 2-amino-2-(hydroxymethyl)1,3-propandiol (Tris); antiproteinases: 0.2 • mM phenylmethylsulfonyl fluoride (PMSF), 1 1 10 jj,g ml" leupeptin, 10 fig ml- antipain, 10 (ig ml-1 pepstatin A, and 10 mj ml-1 aprotinin; and antiphosphatase: 0.5 mM sodium orthovanadate" (dilute from 200 mM stock prepared by successive cycles of boiling and pH adjustment to 10.0), pH 8.0 at 4°C). • • Kinase assay buffer (20 mM p-glycerophosphate, 5 mM sodium pyrophosphate, 30 mM NaCI, 1 mM dithioerythritol (DTT), pH 7.2, 4°C) • ATP mix (3 (iM Na2ATP, 7.5 mM MgCI2, 0.25 mCi ml-1 [-y32P]ATP (Amersham PB10168))

Thin layer chromatography (TLC) plates (Merck Kieselgel 60 F-254, 20 cm X 20 cm) Prepare the plate prior to use by dipping briefly in a tank containing methanol:1% potassium oxalate, 2 mM ethylenedinitrilotetraacetic acid (EDTA) 1:1, (v/v). Air dry, bake in an oven for 2 h at 150°C. Cool to room temperature, and use quickly. Equilibrate the TLC tank in solvent of methanol:CHCI3:NH3(conc):H20, 300:210:45:75 (v/v/v/v). Antiphosphotyrosine monoclonal antibody (PY20, ICN Biomedicals Ltd) Protein A-Sepharose CL4B beads (Pharmacia). For each sample, pre-equilibrate 40 M.! of a 1:1 (v/v, packed beads:lysis buffer) suspension of beads for 2 h in 1 ml lysis buffer + 1% BSA, in 1.5 ml Eppendorf tubes. A positive displacement pipette should be used if possible to disperse beads from an even suspension. Ptdlns/cholate (3 mg ml-1 Ptdlns (Sigma P2517). Ptdlns is dried from a stock solution in CHCI3 (under a stream of N2 or argon or in vacuo), and then bath-sonicated at room temperature into kinase assay buffer with 1% cholate, until the cloudy solution goes clear.

A. Method 1. Wash cell suspensions with balanced salt solution (centrifuge at 1500 g for 4 s at room temperature, and carefully remove supernatant). Aliquot cells in balanced salt solution into 1.5 ml plastic microfuge tubes (EppendorfMe.g. 0.8 ml, 1.5 x 107 cells ml-1) and stimulate (with or without agonists) as required (1 mM sodium orthovanadate may be added to the salt solution to inhibit tyrosine phosphatases, and cells may be pre-treated with 1 mM dispropylfluorophosphate for 5 min at 20°C to inhibit proteases). If measurements are to be made using adherent cells (usually 6-10 cm tissue culture dishes), cellular quiescence is induced by serum starvation (incubation for 16-24 h, 37°C, 6% C02 in medium in the absence of fetal calf serum (PCS), supplemented with 0.1% fatty acid free-BSA). Following starvation, wash cells twice with balanced salt solution, and stimulate at 37°C as necessary. 2. Separate the cells from the medium. For non-adherent cells, separate by centrifugation in a microcentrifuge at 1500 g for 4 s at room temperature; for adherent cells, aspirate the medium. Lyse cells with 1 ml ice-cold lysis buffer, and incubate for 10 min on ice. For adherent

288

11: Phosphoinositide 3-kinases cells, incubate with lysis buffer for 10 min on ice whilst rocking, then harvest cells with a cell scraper and transfer to 1.5 ml Eppendorf tubes on ice. 3. Centrifuge samples at 4°C (10000 g, 5 min) to pellet insoluble material. Remove the supernatant (typically 0.5-1.0 mg ml-1 of cell-derived protein), and add to 8 Ul antiphosphotyrosine monoclonal antibody (PY20), plus 40 ul 1:1 (v/v, packed beads-lysis buffer) suspension of Protein A-Sepharose CL4B beads (pre-equilibrated in 1 ml lysis buffer + 1% BSA as described in Equipment and reagents). Mix samples by end-to-end rotation for 2 h at 0°C. 4. Microcentrifuge samples (200 g, 15 s, 4°C). Collect beads by carefully removing and discarding the supernatant. Wash beads with 1 ml solutions described below (wash at 4°C; mix sample by gentle vortexing or inversion, microcentrifuge (200 g, 15 s, 4°C), and carefully remove supernatant for each cycle), (a) 3x lysis buffer, 4°C; (b) 2x 0.5 M LiCI, 0.1 M Tris, pH 8.0, 4°C; (c) 1X 0.15 M NaCI, 10 mM Tris, 1 mM EOTA, pH 7.6, 4°C; (d) 1X 20 mM Hepes, 1 mM DTT, 5 mM MgCI2, pH 7.6, 4°C. B. Assay of immunoprecipitated PI3K activity 1. Carefully resuspend the final pellet of beads in 40 ul of assay buffer. Add 20 ul of Ptdlns-cholate (3 mg ml-1 Ptdlns, refer to Reagents for preparation). Alternatively, a mixture of lipid substrates can be used (in the presence of antiproteinases and antiphosphatases) in molar % of 26.9% phosphatidylethanolamine, 26.9% phosphatidylinositol, 25.5% phosphatidylserine, 2.8% sphingomyelin, 2.8% Ptdlns(4)P, 2.8% Ptdlns(4,5)P2 (to give a final concentration of 6 uM Ptdlns(4,5)P2 in the mixture added to the assay). 2. Mix lipids, assay buffer, and beads by gentle vortexing, and transfer the samples to a waterbath at 37°C for 5 min. 3. Add 40 ul [32P]ATP mix (yielding a total assay volume of approximately 120 ul), vortex gently, and incubate at 37°C for 15 min. 4. Terminate reactions by the addition of 0.45 ml CHCI3:methanol (1:2 v/v). Vortex mix to yield a homogenous extraction phase.6 Add 150 ul CHCI3 and 150 ul 0.1 M HCI. Mix by vortexing, then microcentrifuge (200 g, 10 min, room temperature). 5. Carefully remove the lower phase (use a siliconized Gilson tip, and equilibrate the tip in CHCI3 before use) into a fresh microcentrifuge tubes (think about the radiochemical hazard!). We do this by drawing a small quantity of the upper phase, with the final few microlitres of the lower phase, into the pipette tip; the small quantity of upper phase is held back in the pipette tip when the lower phase is expelled into a

289

K. E. Anderson et al. Protocol 1.

Continued

fresh tube. We support the microfuge tubes containing the radioactive assay mix inside a close-fitting but thick-walled clear glass test tube, both to open the microfuge tube and to hold it during the removal of the lower phase. By keeping the open neck of the tube pointed away, and using a small benchtop Perspex screen, it is possible to reduce substantially your exposure to 32P. 6. Add 600 (ul of 'synthetic' upper phase (methanol:1 M HCI:CHCI3, 48:47:3, (v/v/v)). Mix, and microcentrifuge as above. Carefully remove the lower phase to a fresh 1.5 ml tube and dry in vacuo. 7. Redissolve the dry lipid film by sonication and vortexing in 25ul CHCI3:methanol:HCI (200:100:1, (v/v/v). Spot onto a prepared TLC plate (e.g. as 10 ul aliquots, followed by washing the tube through with a further 25 uJ). Run for approximately 150 min (in solvent methanol: CHCI3:NH3(conc):H20, 300:210:45:75 (v/v/v/v)) until the solvent reaches the top. 8. Air dry the plate, and quantify the resolved labelled phosphoinositides by one of two methods: (a) visualize the labelled lipids by autoradiography. Scrape the phosphoinositide lipid spots into glass scintillation vials (lightly spray with water to avoid crumbling). Add scintillation fluid and determine the radioactivity by scintillation pcounting (e.g. Beckman LG/S3801 or Packard TR1600 scintillation counter), (b) Alternatively, dried TLC plates can be analysed by phosphorimaging, and quantified by integration analysis (we use the Molecular Analyst program from BioRad). a

A5' phosphatase is known to associate with p85/p110 PI3K (16). This phosphatase activity can be inhibited by the addition of 0.5 mM orthovanadate. bSamples can be stored at this point at -20°C.

4. Measurement of PI3K lipid products in the cell A more direct measure of the activation of PI3K, which makes no assumptions about the nature of any permanent physical interaction between PI3K and an identified protein, is to measure its primary product in cells, that is, PtdIns(3,4,5)P3 and PtdIns(3,4)P2; this is currently the only reliable method for measuring the activation of PI3K by G protein-coupled receptors. In performing these experiments, the low levels of 3-phosphoinositides relative to more conventional lipids (Ptdlns, PtdIns(4)P, and PtdIns(4,5)P2) must be taken into consideration. For example, some standard lipid measuring techniques, such as use of open columns of DOWEX-1 formate, to resolve the water-soluble products of deacylation of phospholipids, are not of sufficient resolving power or chromatographic efficiency to provide adequate sep290

11: Phosphoinositide 3-kinases aration of 3'-phosphoinositides from the more abundant Ptdlns, PtdIns(4)P, and PtdIns(4,5)P2. Detection of 3'-phosphoinositides is standardly achieved by extraction of phospholipids from radiolabelled cells, which can then be separated and quantified by either TLC (Protocol 3) or high performance/ pressure liquid chromatography (HPLC) (Section 4.4.2). Cellular phosphoinositides are most commonly labelled with [32P]-Pi, as it is relatively quick to reach quasi-steady state labelling, and is cost effective for a given level of sensitivity compared to use of [3H]-myo-inositol as a tracer (Protocol 2). A number of CHCl3-methanol based protocols to extract lipids from radiolabelled cells have been described, with varying degrees of recovery. Acidification of the extraction is essential to increase recovery of the most polar lipids, as is inclusion of ion-pairing reagents such as tetrabutylammonium sulfate (TEAS). Below is described a method used routinely in the laboratory to measure changes in lipids upon stimulation in non-adherent cells (Protocol 2). The extraction technique gives high, reproducible extraction of 3'-phosphoinositides. Modifications for [3H]-inositol labelling, and extraction for adherent cells are discussed following the protocol (Sections 4.2 and 4.3).

4.1

32

P-labelling

Protocol 2. Measurement of PI3K lipid products in intact cells Equipment and reagents • Balanced salt solutions (composition may vary slightly with cell type), e.g. 15 mM Hepes, 140 mM NaCI, 5 mM KCI, 2.8 mM NaHC03, 1.5 mM CaCI2, 0.06 mM MgS04, 5.6 mM glucose, 0.1% w/v fatty acid-free BSA, pH 7.4 at 37°C) • Folch Lipids (e.g. Sigma B1502) • Synthetic upper and lower extraction phases: prepare freshly by mixing 0.1 M HCI (containing 5 mM Na2EDTA and 5 mM TBAS):methanol:CHCI3 in the proportions 3:4:8. Centrifuge at 1000 g for 5 min to separate phases, and store in glass.

Methylamine reagent: prepared by slowly bubbling methylamine gas (Fluka cylinder and regulator) into a mixture of methanol:H2O:n-butanol 4:3:1 (v/v/v) on dry ice, until the final volume increases to 1.625X the original (Beware: the mixture heats up as methylamine is added). Aliquot reagents quickly into glass vials (e.g. 10 ml aliquots into glass scintillation vials) and store -80°C. When required for use, warm aliquots to room temperature and use immediately).c.d

Method 1. Wash the cells in balanced salt solutions (centrifuge at 1000 g, 5 min, then remove the supernatant). Label cells with [32P]-Pi in balanced salt solutions. Labelling variables are cell-dependent, but try 0.5 mCi ml-1, 70 min at 37°C, 2.5 x 107 cells ml-1. Wash cells in [32P]-free medium at room temperature (e.g. 3 times with 20 times the cell volume, being careful about radiochemical hazard) by centrifuging for 5 min at 1000 g at room temperature, and removal of the supernatant.

291

K. E. Anderson et al. Protocol 2.

Continued

2. Resuspend cells in balanced salts at approximately 2.5 x 107 cells ml-1. Stimulate 150 ul aliquots with or without 30 ul hormone as appropriate at 37°C. If possible, use glass bottles (e.g. Cam Lab, Microcap, 5 ml volume). Terminate incubations by addition of 750 ul CHCI3:methanol:H20 (32.6%:65.3%:2.1% v/v/v) to produce a homogenous primary extraction phase.3 3. Separate phases by addition of: (a) 725 ul CHCI3 containing 10 ug Folch lipids (2-5 ug of phosphorus of a Folch lipid fraction, ref.17) to act as carrier lipids for the extraction of [32P]-Pi; this can also contain [3H]-Ptdlns(4,5)P2 to act as an internal chromatographic and extraction marker, (b) 172 ul 2.4 M HCI, 5 mM TBAS (to give a final ratio of H2O:methanol:CHCI3 of 3:4:8). Vortex and microcentrifuge (1000 g, 5 min, room temperature) to separate the phases. 4. Remove the lower phase carefully into 2 ml Eppendorf tubes (use siliconized Gilson tips and equilibrate tips in CHCI3 before use) already containing 0.713 ml of synthetic upper phase. 5. Mix the phases and microcentrifuge (1000 g, 5 min, room temperature). Carefully remove the lower phase into clean 2 ml Eppendorf tubes. More efficient extraction of Ptdlns(3,4,5)P3 can be achieved by re-extracting both the initial upper phase (produced following Step 3b) and the synthetic upper phase (after Step 4), sequentially, with an aliquot (1.115 ml) of synthetic lower phase. Dry in vacua using a centrifugal dryerb (e.g. Aqua-vac, Uniscience). 6. Deacylate dry lipids by adding 200 ul of monomethylamine reagentd (prepared as described in Reagents).c Warm tubes to 53 °C for 5 min (place a heavy plate on the tubes to avoid caps blowing off). Vortex tubes vigorously, then return to the 53°C water bath for a further 25 min. Cool samples to room temperature, and dry in vacuo. 7. Resuspend the sample in 0.5 ml H20, and add 0.6 ml petroleum ether (b.p. 40-60°C):n-butanol:ethylformate, 4:20:1 (v/v/v). Vortex and microcentrifuge (7000 g, 5 min, room temperature). Carefully remove the upper organic phase and discard. Wash the lower, water-soluble phase with a further 0.6 ml of the petroleum ether:butanol:ethylformate mixture. Vortex, microcentrifuge, and discard the upper phase as above. Dry the lower phase and interphase in vacuo. 8. Lipids can then by quantified by TLC or HPLC analysis, as described in Section 4,4. "The sample can be stored at -20°C at this stage. bSamples can be stored at -80°C at this stage. c The methylamine reagent should be prepared as described, as the commercially available reagent can generate more complex mixtures of water-soluble deacylation products, probably as a result of some additional uncharacterized chemistry. d Great care must be taken when using methylamine, as it is extremely toxic and volatile.

292

11: Phosphoinositide 3-kinases

4.2 Adherent cells For measurement of changes in inositol phospholipids upon stimulation in adherent cells, several minor modifications to Protocol 2 are made. As serum in cultured cell media can activate or elevate PI3K activity, cellular quiescence is induced by serum starvation (incubation for 16-24 h in media in the absence of PCS supplemented with 0.1% fatty acid-free BSA). To avoid potential losses in transferring cells from tissue culture plates to extraction vessels, and to reduce your exposure to [32P], it is possible to seed cells in individual solvent-resistant small glass bottles (e.g. Wheaton ReactiTM vials 'shorty vials' 19 X 40 mm, Jencons Scientific Ltd). This allows the later extraction of lipids to be performed directly in the bottle, with no need to transfer material. As an example, we culture Swiss3T3 cells in individual glass bottles to a total volume of 1 ml at 1.6 X 104 cells ml-1 (18). Cells are grown to confluence, and serum-starved to induce quiescence. Cells are then washed in serum-free, phosphate-free medium, and labelled in the same medium containing [32P]-Pi. Again labelling conditions may vary with cell type (e.g. in a final volume of 0.4 ml, 0.5 mCi [32P]-Pi for 70 min). Following labelling, cells are washed several times with phosphate-free media, then stimulated as appropriate. Reactions are terminated by addition of 3.75 times the sample volume of CHCl3:methanol 1:2, v/v (e.g. 1.5 ml for 0.4 ml sample). A similar volume of CHC13 (containing 2-5 ug Folch lipid fraction as for Protocol 2, Step 3a) is added, followed by 0.875 times the sample volume (e.g. 0.35 ml for 0.4 ml sample) of 2.4 M HC1, 5 mM TEAS to give a final 3:4:8 ratio of aqueous:methanol:CHCI3 in each bottle, as described above (see ref. 18 for examples of extraction volumes). Vortex the samples vigorously for 2 min, then centrifuge (1000 g, 5 min) to separate the phases. Lipids are extracted as described in Protocol 2 from Step 4, taking into consideration the extraction volumes used. Alternatively, cells can be grown and processed in standard tissue culture plates (usually 35-60 mm dishes). Following stimulation, reactions are terminated by removal of medium by aspiration, and immediate addition of 0.5 ml of ice-cold 1 M HC1. Plates are immediately placed on ice, cells harvested with a cell scraper, and resultant cell debris plus HC1 transferred to small glass bottles. Plates are washed with 1.367 ml of HCl-methanol (0.3 ml 1M HC1, 5 mM TBAS, plus 1.067 ml methanol) and residual debris and solution combined with the first extract. CHCl3-methanol is added to cells in glass bottles to give a final ratio of 3:4:8. Lipids are then extracted as described in Protocol 2, with proportionally bigger volumes throughout (e.g. ref. 19).

4.3 [ 'H]-inositol labelling of cells PI3K lipid products can also be detected by [3H]inositol labelling of cells or tissues. To obtain equilibrium labelling, longer incubation periods are re293

K. E. Anderson et al. quired (24-48 h). It is difficult to maintain tissue viability for this length of time. Thus labelling of tissues (and some cell suspensions) is restricted to nonequilibrium conditions (1-5 h). It must also be noted that some cell types (such as platelets) do not take up [3H]inositol well. To obtain equilibrium labelling, cultured cells, or in vivo labelling of the whole animal, are used. For optimal incorporation, cultured cells are inositol-starved prior to labelling (e.g. 48 h in inositol-free medium (e.g. RPMI 1640 (Gibco)) supplemented with 10% v/v dialysed PCS). Cells are then washed and labelled in the same medium, but containing myo-[3H]inositol (Amersham). As with [32P]-labelling experiments, labelling conditions vary between cell types (try 2 uCi ml-1, 48 h). Cells are incubated for 24-48 h in this medium at 37 °C, 6% CO2. Cells are then serum-starved in inositol-free medium supplemented with 1% w/v fatty acid-free BSA prior to stimulation and lipid extraction (20). Some cells are extremely sensitive to myo-inositol starvation, and quickly detach and/or die. An effective compromise with these cells is to culture in an inositol-free medium supplemented with undialysed FCS.

4.4 Separation and quantitation of radiolabelled PI3K lipid products 4.4.1 Thin layer chromatography PtdIns(3,4,5)P3 was originally detected using TLC to separate extracted lipids of 32P-labelled N-formyl-norleu-leu-phe-norleu-tyr-lys-fluorescein (FLPEP) stimulated neutrophils (21). The TLC separation method employed was essentially that described in Protocol 1, Step 9 onwards. While sufficient to separate PtdIns(3,4,5)P3 from less polar lipids, this method was unable to separate phosphoinositide isomers such as PtdIns(3,4)P2 and PtdIns(4,5,)P2. For successful isomer separation, extracted lipids can be deacylated as described in Protocol 2, Steps 6 and 7, then separated using a modified TLC system described below (Protocol 3). Typical separation of deacylated lipids using this method is shown in Figure 1. Protocol 3. Separation of deacylated phosphoinositides by thin layer chromatography Equipment • Polyethyleneimine-cellulose (PEI) TLC plates, 20 cm x 20 cm (Macherey-Nagel, Germany Stock code MN/801053: 25 cm polygram CEL 300 PEI precoated plastic sheets for TLC).

Method 1. Extract and deacylate lipid products from radiolabelled cells or tissues as described in Protocol 2, Dry the final sample in vacuo. 2. Redissolve the dried extracted samples in 2 ul 20 mM HCI-50 mM

294

11: Phosphoinositide 3-kinases NaH2P04 mix by brisk tapping, vortex briefly, then microcentrifuge (10000 g, 20 s, room temperature). 3. Spot the sample onto the PEI-TLC plate. 4. Wash through the original tube by adding a further 2 ul HCI-NaH2PO4, vortex mix, and microcentrifuge (10000 g, 20 s, room temperature). Spot out onto the dry initial sample spot, using the same tip. 5. Develop the TLC plate in a tank equilibrated with 0.5 M HCI, until the solvent reaches the top (around 90 min). Remove the plate, and air dry it. 6. Quantify the resolved labelled phosphoinositides by either autoradiography (the plastic backing of these plates can be cut easily with scissors, making scraping unnecessary), or phosphorimaging, as described in Protocol 7B, Step 6.

This method of TLC can provide an effective means of separation of labelled PI3K lipid products. However, background levels between lipid spots can be high, and sensitivity of detection fairly low. As a result, detection of 3'phosphoinositides using this method is usually restricted to cell types which display large increases in 3'-phosphoinositides, such as those observed in stimulated neutrophils (21, 22). 4.4.2 High performance/pressure liquid chromatography The most accurate and powerful means of separation and quantitation of deacylated PI3K lipid products in cells is by HPLC. Such analysis provides very low background levels between peaks, and greater sensitivity of detection, resulting in clear baseline separation of isomers of phosphoinositide lipids such as PtdIns(3,4)P2 and PtdIns(4,5)P2. Standard gradients have been developed to separate 3'-phosphorylated lipids, and identification of lipid peaks by retention time is well documented. It is often useful in initial runs to include deacylated lipid standards with the samples. Such phosphoinositide standards are labelled with a different isotope from that of the sample, and deacylated as described in Protocol 2, Steps 6 and 7 (i.e. [32P]-lipid extracts with [3H]inositol labelled standards). Standards can be easily made by immunoprecipitating PI3K activity as described in Protocol 1, using either [32P]-ATP and commercial exogenous lipids, or [3H]inositol labelled lipids, and non-radioactive ATP. In addition to use of an antiphosphotyrosine antibody for this immunoprecipitation, a p85 antibody against the PI3K regulatory subunit (mouse monoclonal IgGl, Santa Cruz Biotechnology Inc.) can also effectively immunoprecipitate PI3K activity using a similar protocol to that in Protocol 1, but using the antibody in a suspension of Protein GSepharose. Production of deacylated [32P]-PtdIns(4,5)P2 for an HPLC standard has been described previously (23). 295

K. E. Anderson et al.

Figure 1. Thin layer chromatography separation of deacylated lipid products from [32P]labelled permeabilized neutrophils. Lipids were extracted and deacylated as described in Protocol 2, and deacylated products separated by TLC as described in Protocol 3. Samples are from non-stimulated basal neutrophils (Lane 1), and from permeabilized neutrophils challenged with GTPyS and fMLP (Lane 2). Lane 3 shows separation of 3'phosphoinositide standards produced by PI3K activity using exogenous Ptdlns(4)P and Ptdlns(4,5)P2 as substrates. GroPlnsP, deacylated phosphoinositide; GroPlns(3,4)P2, deacylated phosphoinositide(3,4)bisphosphate; GroPlns(4,5)P2, deacylated phosphoinositide(4,5)bisphosphate; GroPlns(3,4,5)P3, deacylated phosphoinositide(3,4,5)triphosphate.

Standards may be co-injected with the sample, or alternatively may be added to the sample prior to lipid extraction as a marker of extraction recovery. Some workers also routinely include nucleotide standards, adenosine mono-, di-, and tri- phosphate (AMP, ADP and ATP respectively), in their samples, with the column output attached to an ultraviolet detector, as an assessment of column performance. Lipids are extracted from radiolabelled cells, and deacylated as described in 296

11: Phosphoinositide 3-kinases Table 1. Gradient program for high performance liquid chromatography separation of deacylated products derived from lipid extracts of radiolabelled cell and tissues

Time (min)

% H2O (A)

% NaH2P04(B)

0 1 30 31 60 61 80 81 91 110

100 99 94 85 75 67 40 0 0 100

0 1 6 15 25 33 60 100 100 0

Protocol 2, and dried in vacua. Pellets are redissolved in 2.0 ml water by briefly bath-sonicating and vortexing, then 0.45 (um filtered. Deacylated lipids are separated on an anion-exchange column (we use a Whatman Partisphere 5 SAX column 12.5 cm, ref. 24). Dual pumps are employed at a flow rate of 1.0 ml min-1, and a gradient of filtered and degassed H2O and 1.25 M NaH2PO4 is used, as outlined in Table 1. Lipid products can be quantified by one of two means: (a) The column output can be connected to a fraction collector. Collect 0.5 ml fractions directly to scintillation vials (to avoid transfer of samples), add scintillant (check it will dissolve salt in the fractions; we add 0.5 ml methanol:H20, 1:1 v/v, for fractions 1-130, and 0.5 ml H2O for fractions 130-180, plus 3.5 ml 'Packard 299' scintillant), and count on a B-counter. (b) On-line p-counters and auto-integrators are available with some HPLC systems, that can quantify the amount of 32P/3H-c.p.m. in each peak, as well as a percentage of total c.p.m. Both strategies can be set up to allow quantitation of two independent, appropriately distinct, isotopes in a single HPLC run. A typical HPLC profile and peaks is shown in Figure 2.

4.5 Mass analysis of PtdIns(3,4,5)P3 In addition to radiolabelled detection of PtdIns(3,4,5)P3, it is possible to measure the mass of the lipid product in the cell. This is very advantageous when non-equilibrium labelling is used. Recently a novel method for the measurement of PtdIns(3,4,5)P3 has been described. This method involves the alkaline hydrolysis of PtdIns(3,4,5)P3 to produce Ins(l,3,4,5)P4, which is then used to displace 32P-InsP4 from a highly specific binding protein from cerebellar membrane preparations. The mass assay is reported to be 297

K. E. Anderson et al.

Figure 2. Anion-exchange high-performance liquid chromatography separation of the deacylated products derived from lipid extracts of either control (open circles) or fMLPstimulated (filled circles) 32P-labelled human neutrophils. Inset figure shows the entire elution profile. Numbered peaks identified as deacylated products of Ptdlns(3,4)P2 (1), Ptdlns(4,5)P2 (2), and Ptdlns(3,4,5)P3 (3). The elution times of internal 3H-labelled standards are shown. Reprinted with permission from Stephens ef. al. (1991) Nature, 351, 33. Copyright (1991) Macmillan Magazines Limited.

simple, reproducible, and highly sensitive, detecting PtdIns(3,4,5)P3 at subpicomolar levels (25). This assay, however, has yet to become commercially available.

5. Summary In this chapter we have attempted to outline some of the central methods currently available for measurement of PI3K activity in cells and tissues. A combination of strategies of inhibition and activation are often required to construct a clear picture of the involvement of PI3K in cellular events. The most direct measurement of cellular PI3K activity is the quantification of PI3K primary lipid products (namely PtdIns(3,4)P2 and PtdIns(3,4,5)P3) under both basal and stimulated conditions. This is most often achieved by 298

11: Phosphoinositide 3-kinases extracting lipids from radiolabelled cells or tissues, and separating deacylated lipid products by either TLC or, more accurately, HPLC.

References 1. Equinoa, A., Krugmann, S., Coadwell, J., Stephens, L., and Hawkins, P. (1997). In Molecular mechanisms of signalling and membrane transport (ed. K. W. A. Wirtz). p. 171. Springer-Verlag, Berlin. 2. Stephens, L., Smrcka, A., Cooke, F. T., Jackson, T. R., Sternweis, P. C, and Hawkins, P. T. (1994). Cell, 77,83. 3. Thomason, P. A., James, S. R., Casey, P. J., and Downes, C. P. (1994). J. Biol. Chem., 269,16525. 4. Stoyanov, B., Volinia, S., Hanck, T., Rubio, I., Loubtchenkov, M., Malek, D., Stoyanova, S., Vanhaesebroeck, B., Dhand, R., NUrnberg, B., Gierschik, P., Seedorf, K., Hsuan, J. J., Waterfield, M. D., and Wetzker, R. (1995). Science 269, 690. 5. Stephens, L. R., Eguinoa, A., Erdjument-Bromage, H., Lui, M., Cooke, F., Coadwell, J., Smrcka, A. S., Thelen, M., Cadwallader, K., Tempst, P., and Hawkins, P. T. (1997). Cell, 89,105. 6. Carpenter, C. L., and Cantley, L. C. (1996). Curr. Opin. Cell. Biol, 8,153. 7. Toker, A., and Cantley, L. C. (1997). Nature, 387,673. 8. Arcaro, A., and Wymann, M. P. (1993). Biochem. J., 296,297. 9. Vlahos, C. J., Matter, W. F., Hui, K. Y., and Brown, R. F. (1994). J. Biol. Chem., 269(7), 5241. 10. Wennstrom, S., Hawkins, P., Cooke, F., Hara, K., Yonezawa, K., Kasuga, M., Jackson, T., Claesson-Welsh, L., and Stephens, L. (1994). Curr. Biol., 4,385. 11. Fantl, W., Escobedo, J. A., Martin, G. A., Turck, C. W., del Rosario, M., McCormick, F., Williams, L. T. (1992). Cell, 69, 413. 12. Kashishian, A., Kazlauskas, A., Fay, F. S., and Corvera, S. (1992). EMBO J., 11, 1373. 13. Kotani, K., Yonezawa, K., Hara, K., Ueda, H., Kitamura, Y., Sakaune, H., Ando, A., Chavanieu, A., Calas, B., Grigorescu, F., Nishiyama, M., Waterfield, M. D., Kasuga, M. (1994). EMBO J., 13,2313. 14. Klippel, A., Reinhard, C., Kavanaough, M., Apell, G., Escobedo, M.-A., and Williams, L. T. (1996). Mol. Cell. Biol., 16,4117. 15. Anderson, K. E., Coadwell, J., Stephens, L. R., and Hawkins, P. T. (1998). Curr. Biol., 8,684. 16. Jackson, S. P., Schoenwalder, S. M., Matzaris, M., Brown, S., and Mitchell, C. A. (1995). EMBO J., 14,4490. 17. Folch, J. (1949). J. Biol. Chem., 177,497. 18. Jackson, T. R., Stephens, L. R., and Hawkins, P. T (1992). J. Biol. Chem., 267, 16627. 19. Nobes, C. D., Hawkins, P., Stephens, L., and Hall, A. (1995). J. Cell. Sci., 108,225. 20. Corey, S., Eguinoa, A., Puyana-Theall, K., Bolen, J. B., Cantley, L., Mollinedo, F., Jackson, T. R., Hawkins, P. T., and Stephens, L. R. (1993). EMBO J., 12,2681. 21. Traynor-Kaplan, A. E., Harris, A., Thompson, B., Taylor, P., and Sklar, L. A. (1989). Nature, 334,353. 299

K. E. Anderson et al. 22. Stephens, L., Jackson, T., and Hawkins, P. T. (1993). J. Biol. Chem., 268,17162. 23. Letcher, A. J., Stephens, L. R., and Irvine, R. F (1990). in Methods in inositide research (ed. R. F. Irvine), p. 31. Raven Press, New York. 24. Stephens, L. R., Hughes, K. T., and Irvine, R. F. (1991). Nature, 351, 33. 25. Van der Kaay, J., Batty, I. H., Cross, D. A. E., Watt, P. W., and Downes, C. P. (1997). J. Biol. Chem., 272,5477.

300

12 Phospholipase D and phosphatidylcholine metabolism KATHRYN E. MEIER and TERRA C. GIBBS

1. Introduction Phospholipase D (PLD) catalyses the hydrolysis of phospholipids to phosphatidic acid (PA). The preferred substrate for most PLDs is phosphatidylcholine, although utilization of other substrates (e.g. phosphatidylethanolamine) has been reported. PLDs are expressed in mammals, insects, plants, yeast, and certain bacteria. In mammalian cells, PLD is activated in response to a broad range of agonists (reviewed in refs 1-4). In general, PLD activity is increased by agonists that bind to G protein-coupled receptors (GPCRs) inducing phospholipase C (PLC) activation. Various growth factors, hormones, and cytokines can activate PLD. Phorbol-12-myristate-13-acetate (PMA), a tumour promoter that activates protein kinase C (PKC) isoforms, induces PLD activation in most mammalian cell types. The roles of PLDs in signal transduction remain of great interest, but poorly defined. The recent molecular cloning of PLD isoforms has greatly accelerated work in this research area. A PLD sequence was first obtained from plants (5), with subsequent identification of yPLDl/SPO14, the sequence encoding a yeast PLD (6-8). Two mammalian isoforms, PLD1 and PLD2, have been identified (9-11). Plant PLDs are soluble enzymes, whereas the yeast (12) and mammalian (13) enzymes are membrane-localized, despite the lack of a transmembrane domain. There are four regions of conserved homology between plant, yeast, and mammalian PLDs, two of which are known to be required for catalytic activity. The enzymes range in size from 90 kDa (plant) to 195 kDa (yeast). Mammalian PLD1 (120kDa) is activated by small GTP-binding proteins, such as ARF and Rho (reviewed in 1-4, 14). PLD1 is also activated by phosphatidylinositol-4,5-bisphosphate (PIP2), and by protein kinase C (PKC), via a mechanism that does not require protein phosphorylation. Although PLD2 (105kDa) is stimulated by PIP2, other factors regulating its activity are not yet well defined. PLD1 and PLD2, when over-expressed in COS cells, are both activated in response to PMA, but show distinct subcellular localizations (10). A distinguishing feature of the PLD-mediated reaction is the ability of the

Kathryne E. Meier and Terra C. Gibbs enzyme to utilize either water or a primary alcohol as substrate. This is termed 'transphosphatidylation'. This feature has been extensively utilized to analyse PLD activation in intact cells, in which PA is rapidly converted to diglyceride by PA phosphohydrolase, and/or to other products. The phosphatidylalcohol products are more stable, though subject to metabolism by phospholipase A2 (PLA2) (13). Another advantage is that phosphatidylalcohols are generated only subsequent to addition of an exogenous alcohol to the cells. Use of the transphosphatidylation reaction has also proved useful for in vitro assays for PLD activity. It has been suggested that endogenous alcohols may serve as substrates for transphosphatidylation (15). The extent to which this reaction occurs is not clear. There is some evidence to suggest that alcohols, in addition to serving as substrates, stimulate PLD activity (16). The reaction generally proceeds to the greatest extent with butanol or pentanol, depending on the specific PLD (12,16). Due to the ready commercial availability of phosphatidylethanol and phosphatidylbutanol as chromatography standards, ethanol or butanol is usually employed in intact cell assays. An advantage of ethanol is the vast literature concerning effects of this alcohol on intact cells. It should be mentioned that forms of PLD that do not utilize alcohols may exist (17). However, the effects of agonists on PLD activity have been largely characterized through use of the transphosphatidylation reaction.

2. PLD assays in intact mammalian cells The standard method used to measure PLD activation in intact cells involves the use of metabolic labelling to incorporate a tritiated fatty acid into the phospholipid pool. The cells are then incubated with and without agonists in the presence of an alcohol (ethanol or butanol). Radioactivity in PA and phosphatidylalcohol (e.g. phosphatidylethanol [PEt]) is then determined following extraction of lipids, and separation by thin-layer chromatography. Alternative approaches in intact cells have utilized a [3H]-alkyl-lyso-phospholipid (l-O-alkyl-2-acyl-sn-glycero-3-phosphocholine) as precursor (18), or [14C]-butanol as substrate (19). The former method has been used successfully by many investigators, but may preferentially assay for PLD isoforms that preferentially utilize alkyl-phospholipids (10, 21). For further discussion of the substrate specificity of PLD, see ref. 22. The latter approach using [14C]butanol is rather hazardous, since the radioisotope is volatile. Moreover, production of PA cannot be readily assessed when this approach is used.

2.1 Metabolic labelling of cells with [3H]-fatty acids A variety of fatty acids have been used for this purpose. [3H]-palmitic and -oleic acids are most commonly used. These fatty acids are relatively inexpensive, and are taken up into a wide spectrum of phospholipid species. [3H]-myristic and -arachidonic acids have also been used. The latter should be used with 302

12: Phospholipase D and phosphatidylcholine metabolism caution, in view of a report that phospholipids incorporating arachidonate may be preferentially utilized by one form of PLD over another (21). [3H]palmitate and -oleate acid stock solutions, supplied in ethanol and stored at —20°C, are relatively stable, and can be used for at least one year after purchase. If the concentration of the stock solution is at least 5 mCi ml"1, then the isotope can be added directly to the culture medium (giving a final concentration of 0.2% ethanol or less). Addition of higher amounts of ethanol to the culture medium should be avoided, since this could cause toxic effects and/or increase basal production of phosphatidylethanol. Excellent results can be obtained using adherent cells grown in 35 mm dishes (50-100% confluence). Suspension-growing cells can also be used; in this case, all washing steps will require microcentrifugation. Uptake of the fatty acids into cells ensues immediately, with an approximately linear rate of incorporation of isotope over a 24 h period (Bradshaw and Meier, unpublished data). An overnight incubation (18-24 h) is typically used to maximize incorporation of the isotope, and its distribution between phospholipid pools. Protocol 1. Incubation of cells with radiolabelled precursor Equipment and reagents • Cells, growing either in 35 mm culture dishes, or in suspension • [3H]-fatty acid (e.g. palmitic acid) (Amersham or Dupont-NEN) • Cell culture medium

• Cell culture facility with laminar flow hood and incubator • Radioactive isotope usage and disposal facilities

Method 1. Prepare the isotope at a concentration of 10 n-Ci mM in the desired culture medium.a 2. Remove the culture medium, using sterile technique; replace with isotope-containing medium. 3. Return cells to the incubator for the duration of the labelling period. ° Serum-containing medium is usually used. A total volume of 2 ml is suggested to conserve isotope while providing adequate coverage of a 35 mm dish. It is not necessary to replace all of the medium when adding the isotope, although this is convenient.

2.2 Incubation of cells with agonists It is advantageous to rinse the cells free of the isotope-containing medium prior to incubation with agonists. This will reduce the assay background, as well as making the subsequent procedures less hazardous for laboratory personnel. It is convenient to process replicate samples of cells at the same time. The assay should be performed with triplicate or quadruplicate samples. 303

Kathryne E. Meier and Terra C. Gibbs Note that it takes approximately 2 min to wash and add fresh medium to 4 dishes of cells, but about 5-10 min to harvest and extract these dishes. The incubations should be staggered accordingly. We routinely assay 40 dishes per experiment. Some investigators use concentrations of ethanol higher than 0.5%, or use butanol, in the intact cell assay. We recommend using 0.5% ethanol because: (a) The radioactivity recovered in PEt is more than adequate, using the method described below. (b) The side-effects of ethanol on intact cells have been well-established. Note that even 0.5% ethanol is a very high dose with respect to human toxicity. Although PEt is a relatively stable product, as compared to PA, its gradual loss is detectable in intact cells. PLD activation in response to physiological agonists is often transient (5-10 min). Therefore, a time course experiment should be performed initially to establish the optimal time for measurement of agonist-induced PLD activation. An additional approach to determining the kinetics is to add ethanol at various times after the agonist (23). Protocol 2. Assay incubation Equipment and reagents • Cells metabolically labelled with the radioactive precursor • Incubation medium, pre-warmed to 37°C • Cell culture incubator

• Stock solutions of agonists/antagonists (e.g. at 1000X the desired final concentration) • Radioactive isotope usage and disposal facilities

Method 1. Prepare Dulbecco's modified Eagle's medium (or other compatible medium), supplemented with 10 mM Hepes (pH 7.5) (DME/H); warm this solution to 37°C. 2. For attached cells,a use a disposable pipette to remove the medium (transferring to a radioactive waste container). Wash each dish twice by gently adding 2 ml DME/H, agitating the dish manually for a few seconds, then withdrawing the medium using a pipette (transferring to a radioactive waste container). Finally, add 2 ml DME/H as incubation medium. 3. Add ethanol (95%) to a final concentration of 0.5% (v/v) a few seconds before the agonist is added. Next add the desired concentration of agonist or vehicle to each dish. Swirl dishes to mix after each addition. Return cells to the incubator for the duration of the incubation. •Suspended cells are handled similarly, except that washes require brief microcentrifugation.

304

12: Phospholipase D and phosphatidylcholine metabolism

2.3 Extraction of phospholipids from intact cells The following instructions apply for attached cells. For suspended cells, washing steps will need to be carried out in a microfuge tube, using centrifugation to collect cells before removal of medium. The lipid extraction should be carried out in a fume hood, since chloroform is used. The procedure for drying samples under nitrogen deserves special attention. This step can be time-consuming, depending on the number of samples, since each sample usually requires approximately 15-30 min of drying time. Use of a multi-port evaporating system greatly facilitates this process. The Meyer N-Evap (Organomation) is a stable and reliable system for this purpose, though adequate systems can be constructed in the laboratory. The dried samples can be covered and stored overnight at — 20 °C prior to chromatography. Protocol 3. Lipid extraction Equipment and reagents • Cells, incubated with the desired agonists • Vacuum aspirator with in-line trap for radioactive waste • Dulbecco's phosphate-buffered saline solution (PBS) (Sigma Chemical Co.), cooled to 4°C . Methanol:6N HCI (50:2, v/v)

• Glass test tubes (i.e., 12 x 75 rnm), numbered • Chloroform . 1 M NaCI • Nitrogen gas with multiport evaporating system

Method 1. Remove the incubation medium from each dish by vacuum aspiration. 2. Add ~4 ml ice-cold PBS, swirl briefly, and then remove the wash using a vacuum aspirator. 3. Add 1 ml methanolic HCI to each dish. Scrape the cells from the culture dish using a disposable cell scraper. Transfer the resulting suspension to a glass tube. 4. Add 0.5 ml chloroform to each sample, followed by 0.28 ml 1M NaCI. Mix each sample well using a vortex mixer. 5. To accomplish phase separation, the samples are left on ice for 15-20 min and/or subjected to low-speed centrifugation for 3-5 min. 6. Collect the lower organic phase, using a disposable glass pipette. Transfer this phase to a fresh 12 x 75 mm glass tube. Discard the remaining upper phase as liquid radioactive waste. 7. Evaporate the lower organic phase under a stream of nitrogen gas at room temperature. 8. Store the dried samples at -20°C after covering the tops with Parafilm. 305

Kathryne E. Meier and Terra C. Gibbs

2.4 Thin-layer chromatography The TLC step is straightforward if performed with the reagents described below. Note that PA and PEt are not particularly easy to resolve. Use of alternative solvent systems or TLC plates can result in insufficient separation of products, and is not recommended. Since PA and PEt represent minor fractions of the total phospholipid, they must be well separated from other products to achieve accurate quantification. The TLC solvent should be freshly prepared. Perform the separation in a fume hood. Protocol 4.

Separation of PLD reaction products

Equipment and reagents • Phosphatidic acid (Avanti) . Phosphatidylethanol (Avanti) .Silica gel chromatography plates: Whatman #4865-821, 60 A, 20 x 20 cm, 250 (xm thickness, 19 channels, with pre-adsorbent strip, without fluorescent indicator (Fisher Scientific Co.)

« Chromatography solvents: ethyl acetate, trimethylpentane, acetic acid, water • Glass separatory funnel (500 ml) • TLC tank (glass) containing solvent . TLC tank (glass) containing iodine crystals . Fume hood

Method 1. Prepare the TLC solvent mixture (ethyl acetate:trimethylpentane:acetic acidiwater; 90:50:20:100, v/v) in a separatory funnel/ Mix the solvent, venting the funnel according to standard chemical laboratory practice. Allow the phases to separate for 20 min at room temperature. Discard the lower phase, and retain the upper phase as the TLC solvent. Transfer the solvent to the TLC tank, and cover the tank. 2. Add 30 uJ chloroform:methanol (90:10, v/v) to each dried sample, along with 1 ul phosphatidylethanol (5 mg ml-1 in chloroform:methanol) and 1 ul phosphatidic acid (5 mg ml-1 in chloroform:methanol) as chromatography standards. Use a vortex mixer to dissolve the samples. Keep the samples cold to avoid evaporation of solvent. 3. Mark the TLC plate with a pencil to show where the samples should be applied (2 cm from the bottom of the plate, in the centre of the lane). Use a manual pipettor with a disposable tip (e.g. Pipetman) to apply the entire sample to the TLC plate. Apply each sample to a separate lane. On each plate, use two lanes at one end of the plate to load each of the standards alone (PA and PEt) for purposes of calibration. 4. Using a heat gun, apply warm air to the plate until the applied spots have dried. 5. Place the TLC plate in the solvent tank at a 30° angle. The sample origin should be above the level of the solvent. Cover the tank. Develop the plate until the solvent has migrated within ~1 cm of the top of the plated.5-2 h). 306

12: Phospholipase D and phosphatidylcholine metabolism 6. Remove the plate from the tank. Allow the solvent to evaporate in the fume hood. 7. Place the dried plate in the iodine tank. Cover the tank. Observe the staining of the PA and PEt bands as it progresses (~5 min). The RF values for PA (lower band) and PEt (upper band) are approximately 0.2 and 0.3, respectively.b When the yellow-brown bands are clearly visible, remove the plate from the tank. Immediately mark the positions of the bands, in each lane, using a pencil. Leave the plate in the fume hood until the iodine staining disappears. 8. Prepare three labelled scintillation vials for each samplec. One vial is for PA, one for PEt, and one for the remainder of the lane. Using a single-edged razor blade, scrape the PEt band for each sample into a separate vial. Repeat the procedure for the PA bands. Finally, scrape and collect the remainder of each lane.b 'If the indicated proportions are added as millilitres, the necessary volume of solvent will be obtained. 6 The bands must run clear of the loading zone area of the plate, or significant background will be observed. If the standards do not run far enough up the plate, or are not adequately separated from each other, the plate can be re-developed in the same solvent to improve the separation. 0 Mini-vials (~7 ml capacity) will suffice. Folded weighing paper can be used to collect the scraped silica. Use of a particle filter face mask or other physical barrier is advised, to prevent inhalation of silica dust containing isotope.

2.5 Quantification of PLD activity Since the specific activity of the metabolically-labelled phospholipid precursor is unknown, and because the above procedure is likely to yield somewhat variable recovery of cellular phospholipid, the results of this assay are usually normalized to the total radioactivity recovered. In other words, production of PEt and PA is expressed as a percent of the total radioactivity recovered from each cell sample (described below). This calculation corrects for small variations in the number of cells per sample, the amount of isotope added, and the recovery of cells and phospholipid. In our experience, this normalization results in standard error values of less than 10% between assay replicates. Samples in which recovery of radioactivity is extremely poor (i.e. less than 50% of the total obtained from other dishes) should be excluded from the final calculations. In unstimulated cells, the amount of radioactivity recovered in PA is usually more than is recovered in PEt, both being of the order of several thousand c.p.m.. The percentage of total radioactivity recovered as PEt is approximately 0.3-3%, and that recovered as PA is 1-5%, depending on the cell line and agonist used. The total radioactivity recovered varies according to the number of cells and duration of labelling, but is of the order of several hundred thousand c.p.m.. 307

Kathryne E. Meier and Terra C. Gibbs It can be difficult to assess 'basal' PLD activity accurately. We recommend that each experiment include cells incubated without ethanol, with ethanol, and with both ethanol and agonist. In theory, the difference in the PEt radioactivity between unstimulated cells incubated with and without ethanol represents 'basal' PLD activity. However, we find that the radioactivity recovered as PEt is usually similar in cells incubated in the absence or presence of ethanol, suggesting that basal activity is very low. Hence, it is much easier to analyse the extent of PLD activation than to measure changes in basal activity. The radioactivity recovered as PA is somewhat variable, and probably includes products other than PA. Therefore, this value is often not presented in published papers. However, when production of PA is fairly consistent, it can be used as one indicator of PLD activity. Protocol 5. Assay quantification Equipment and reagents • Scintillation vials • Liquid scintillation cocktail

• Liquid scintillation counter • Calculator or computer

Method 1. Add scintillation fluid to each vial.8 Cap each vial and then shake to ensure adequate mixing. 2. Quantify radioactivity, using a liquid scintillation counter. 3. Calculate the sum of the c.p.m. recovered in the three vials (PEt, PA, and remainder) for each sample. 4. Divide the PEt and PA c.p.m. by the total c.p.m.. These values represent PEt and PA as a percentage of the total radioactivity recovered. 5. Perform further statistical analysis (e.g. mean ± error) of the replicate samples as desired. • Non-toxic scintillation fluid that is water-compatible is acceptable. We add 2 ml of fluid to the PA and PEt vials, and 4 ml to the vials containing the other products.

2.6. Interpretation of results An example of the results of an intact PLD assay is shown in Figure 1. The time course of PMA-induced production of PEt and PA is shown for PC12, a rat neuronal cell line. PLD activity in these cells has been characterized previously (24). In the left panel, the '0' time point shows radioactivity recovered as PEt in unstimulated cells incubated without ethanol. Note that this value is similar to those obtained after addition of ethanol (control, 15, and 30 min), suggesting minimal 'basal' PLD activity. In contrast, the effects 308

12: Phospholipase D and phosphatidylcholine metabolism

Figure 1. PLD assay in intact mammalian cells. PC12 cells were metabolically labelled with [3H]palmitic acid. Washed cells were then incubated with 0.5% ethanol in the absence and presence of 100 nM PMA for the indicated times. Production of phosphatidylethanol (PEt) and phosphatidic acid (PA) was assessed following extraction of cellular lipid and TLC separation. Data are expressed as the percentage of total radioactivity recovered in each product. Each point represents the mean ± SD of values obtained from quadruplicate dishes of cells.

of agonist on PLD activity are clear. PMA causes a time-dependent increase in PEt levels. An increase in PA production is also evident in this experiment. The response to PMA appears to cease after 30 min, since PEt levels do not increase further, and PA levels decline, at 60 min. The data also suggest that PEt is more stable than PA, as expected on the basis of published literature.

3. PLD assays with broken-cell preparations Assays of this nature have been developed more recently, and vary between laboratories. The protocol described herein utilizes a fluorescent substrate, and is applicable to plant, yeast, and mammalian PLD enzymes (12, 13, 16). Other investigators have used alternative fluorescent substrates (8), or radioactive substrates such as [3H]-phosphatidylcholine (25). Another approach is to label metabolically intact cells, and then subsequently measure PLD activity in membrane preparations (26). This approach has the advantage that endogenous substrates are utilized, while the former approaches have the advantage of using a chemically defined substrate. Particular advantages of the fluorescent in vitro assay include very high sensitivity, rapidity, low cost, and low hazard to personnel. A disadvantage of the assay is the need for equipment that can image and quantify fluorescent signals. While such equipment is becoming more commonly available, quantification is always more indirect than with a radioactive assay. The fluorescent assay is particularly well suited to more non-quantitative applications, such as 309

Kathryne E. Meier and Terra C. Gibbs

Figure2, Fluorescent in vitro assay for PLD activity. Two samples were incubated with BODIPY-phosphatidylcholine (BPO by the method described in the text: membranes prepared from PC12 cells, and a commercially available preparation of PLD from peanut (Sigma Chemical Co.). The reaction mixtures were separated by TLC, and imaged using a Molecular Dynamics Fluorlmager, One lane contains BPC alone as a control. The fluorescent products are indicated. For a key to the abbreviations, see Table 1.

testing column fractions for PLD activity, or high-throughput screening tor PLD activators or inhibitors. Another advantage of this method is that all the major products derived from the fluorescent substrate are visible when the reaction is analysed (Figure 2). This feature lends assurance that changes in product levels are not due to an increase or decrease in the activity of other enzymes that utilize the substrate (e.g. PLA2). The assay can potentially he adapted tor use in assays for the activities of these enzymes (e.g. PA phosphohydrolase [27]). An additional advantage to the method reported here is its ability to detect the activated state of PLD (13, 23, 24, 28). In other words, when membranes arc prepared from cells incubated in the presence of agonists that activate PLD, enhanced PLD activity is observed in the membrane preparation. It should be noted that the magnitude of this effect is not as large as that seen in intact cells. Although the mechanistic basis for the increase has not yet been determined, the presence of translocated or activated effectors (e.g. protein kinase C) in the membrane preparation is likely responsible. Our laboratory has usually examined guanine nucleotide-independent PLD activity in mammalian membranes. This activity probably represents PLD2, which is appar310

12: Phospholipase D and phosphatidylcholine metabolism ently independent of small GTP-binding proteins such as ARF and rho. The method can be used to assess PLD1 activity if the appropriate effectors (e.g. guanine nucleotide and GTPases) are added to the membrane assay (Meier et al., unpublished data). It should be noted that the substrate, BPC, has an ether linkage at the 1position, and that the BODIPY fluorophore is also conjugated at this position. One advantage of this substrate is that the products generated by PLA2, PLC, and PLD all retain the BODIPY fluorophore, and are thus fluorescent. The ether linkage makes the substrate relatively resistant to degradation by other lipases. A potential disadvantage is that some phospholipases may not utilize the alkyl lipid efficiently. We have not yet found this to be a limitation of the method.

3.1 Prep aration of cell membranes Mammalian PLD isoforms are predominantly localized to the cell membrane. However, since cytosolic activities have been described in some reports, it is important to establish whether the activity of interest is membrane-bound or cytosolic. Whole-cell extracts are not used because: (a) The high concentrations of detergents (e.g. Triton X-100) needed to make a homogenous extract are inhibitory to PLD activity. (b) Cytosolic components (e.g. cytosolic PA phosphohydrolase) can interfere with the assay. The procedure described here is rapid, simple, and applicable to a variety of cell lines. We recover all of the starting material as either 'cytosol' or 'membranes', so that no portion of the cell is discarded. This type of membrane preparation includes all cellular membranes, rather than only plasma membrane, and probably includes nuclear material (including DNA) as well. The fractions obtained by this protocol can be used for additional purposes (e.g. assay for MAPK activation (see Chapter 14), or PKC translocation [29]). Protocol 6.

Preparation of membrane and cytosol

Equipment and reagents • Lysis buffer: 10 mM Hepes, pH 7.5; 80 mM 3-glycerophosphate; 10 mM EGTA; 2 mM EDTA; 2 mM dithiothreitol; cooled to 4"C . Cells incubated with the desired agonists or antagonists « Dulbecco's PBS

• Disposable cell scrapers . probe sonicator (e.g. Branson Model 250 Sonifier with microtip; Fisher Scientific Co.) .Refrigerated ultracentrifuge with smallcapacity rotor (e.g. Beckman Instruments Ultima)

Methods 1. Incubate cells with the desired agonists or antagonists/ 2. Remove the incubation medium from the cells. Wash the cells twice 311

Kathryne E. Meier and Terra C. Gibbs Protocol 6.

Continued

with ice-cold PBS (e.g. 2-6 ml washes for a 100 mm dish). Swirl the dish briefly, then discard the rinse solution. 3. Add 1 ml of ice-cold PBS. Scrape the cells from the dish, using a disposable cell scraper. Transfer the cells to a 1.5-ml plastic microfuge tube. Keep the samples ice-cold. 4. Sediment cells by centrifugation for 30 s using a microfuge. Discard the supernatant. 5. Resuspend the cell pellet in 0.25-1 ml ice-cold lysis buffer.6 6. Disrupt the cells by sonication.C 7. Separate membranes and cytosol by centrifugation at 100000 g for 20 min. 8. Transfer the supernatant, or 'cytosol', to a separate tube if it is to be retained. Leave the pellet, or 'membranes', in the centrifuge tube. Freeze samples at -80°C until assay. "The incubation medium may contain serum, if desired. b The volume of buffer is not critical if only membranes are to be utilized. As a general rule, a 100 mm dish of confluent cells should be resuspended in ~0.5 ml buffer. Use the same volume for all samples in the experiment. °A probe sonicator, not a bath sonicator, must be used. The probe should fit within the microfuge tube. Rinse the probe with water and wipe clean between samples. The power output on the sonicator should initially be adjusted so that the majority of the cells are disrupted within approximately 2 s. Longer sonications result in heating of the sample. The extent of disruption can be judged by microscopical examination.

3.2 Fluorescent PLD assay This assay is rapid and simple, but its successful application requires precision and practice. Use of alternative TLC plates or solvent systems is not recommended, as the separation of products is very sensitive to any changes in the chromatography conditions. Protocol 7. Preparation of fluorescent substrate Equipment and reagents • BODIPY-PC (Molecular Probes, catalogue #03771) • Resuspension buffer (400 mM NaCI, 66 mM Hepes, pH 7.5)

• Octylglucoside (10 mM in water) • Probe sonicator

Method 1. Dissolve 1 mg BODIPY-PC (BPC) in 1 ml ethanol. Store this stock solution at -20°C.

312

12: Phospholipase D and phosphatidylcholine metabolism 2. Dry 100 ul BPC stock solution under nitrogen.3 3. Add 20 ul 10 mM octylglucoside. Resuspend BPC by vigorous vortex mixing. 4. Add 480 ul resuspension buffer. Mix briefly, using a vortex mixer. 5. Sonicate the sample for —30 s,b Maintain the sample on ice throughout the procedure. 6. Store the solubilized substrate at -20°C. During use, maintain the solution at 4°C. 'Before drying, the solution is fluorescent green. After complete drying, the substrate is a dark red-orange colour. "For the Branson Sonifier 250, use output setting of 2 with a 1/4 inch microtip. The sonication is complete when the red-orange substrate is fully solubilized.

Protocol 8.

Assay incubation and chromatography

Equipment and reagents • Protein assay system, compatible with reducing agent (e.g. Bradford assay), and capable of measuring protein in small volumes of sample (e.g. 1-2 (ul • Cell membrane preparation • Lysis buffer (described in Section 3.1) • Solubilized BPC (1 mg ml-1 see Protocol 7] • Butanol (9%, v/v, in water)

• Plastic-backed silica gel TLC plates: EM Science #5748/7, 60 A, 20 x 20 cm, without fluorescent indicator, Fisher Scientific .TLC solvent (methanol:chloroform:water: acetic acid, 45:45:10:2, v/v) • TLC solvent tank, equilibrated with solvent • UV lamp, preferably long-wave

Method 1. Resuspend the membrane sample in lysis buffer for measurement of protein concentration. In our laboratory, we usually resuspend the membranes prepared from one 100 mm dish of attached cells in approximately 50uJ lysis buffer. The resulting suspension will be extremely concentrated (~1 mg mr-1 protein) and viscous, and should be homogenized thoroughly by repetitive pipetting through a smallbore disposable pipette tip. Maintain the preparation at 4°C. 2. Determine the protein concentration, using 1-2 uJ of the membrane preparation. 3. Calculate the volume of each membrane preparation that needs to be used to provide an equal amount of protein for all the samples to be assayed (e.g. use 5 or 10 ug protein for each sample).a Add the desired volume to numbered microfuge tubes; keep cold. 4. Add 5uJ of the 1 mg ml-1 BPC suspension to each assay tube. Then add 1.5 ul of butanol (9%) to samples in which transphosphatidylation is

313

Kathryne E. Meier and Terra C. Gibbs Protocol 8. Continued desired. Add other molecules of interest (e.g. drugs, nucleotides, GTPases); these should be diluted in lysis buffer. Finally, add lysis buffer to bring the total volume of all samples up to 12.5 (J. 5. To initiate the assay, place tubes in a water bath set at 30°C. Incubate for 60 min. 6. While the assay is in progress, prepare the TLC plate.b Draw a straight line in pencil across the plate ~1 cm from the bottom. Next, draw tick marks ~1 cm apart along the line to indicate the positions for sample application. 7. Remove the samples from the water bath. Remove any insoluble debris from the samples by very brief (~5 s) microcentrifugation. 8. Apply 5 ul of each sample to the TLC plate, using a pipettor with a disposable tip.c 9. Dry the applied samples, using a brief application of warm air with a heat gun. Dry only until the solvent has evaporated (~ 2 min). 10. Place the plate in the TLC tank at a 30° angle. Cover the tank and develop the plate for ~15 min, or until the solvent has migrated within ~1 cm of the top of the plate.d 11. Remove the plate from the tank. Allow the solvent to evaporate in a fume hood. 12. Assess the results by using a hand-held UV lamp to illuminate the dried plate. aThe PLD assay requires 3-10 ug protein per sample, in a maximum volume of 5.1 ul. Thus, the protein concentration of the suspension needs to be in the range of 0.5-2.0 mg ml-1. bCut a 20 x 20 cm plate in half. One half is sufficient for 20 samples. The size can be further reduced if fewer samples are to be run. cThe sample must be applied precisely and as a single spot. Errors in sample application will make the results difficult to interpret. d lf desired, the progress of the chromatography can be monitored by briefly illuminating the TLC tank with a hand-held UV light in a darkened room.

3.2.1 Analysis of results Assay results can be imaged in several ways. First, the developed plate can be photographed on Polaroid-type film under UV light, as is commonly done to photograph ethidium bromide-stained agarose gels. Second, the plate can be imaged under UV light using a digital camera system. The advantage of this approach is that the digital image can be captured, labelled, and quantified using various software packages. Third, an instrument designed for imaging and quantification of fluorescent samples (e.g. Molecular Dynamics STORM system) can be used. The advantages to this last approach include very high sensitivity, ability to adjust the sensitivity and output, and availability of soft314

12: Phospholipase D and phosphatidylcholine metabolism ware dedicated to image quantification. With respect to sensitivity, these systems can provide excellent documentation of fluorescent bands that are undetectable by the human eye. The disadvantage is the substantial cost of the system. Results can be quantified in two ways. First, the plate can be scanned using a scanning fluorimeter (e.g. from Helena Laboratories). The advantage of this approach is that it is rapid, and provides simple and accurate quantification of major peaks. The disadvantage is that the instrument may not be able to discriminate minor product bands. Second, the quantification can be done using imaging software such as that available on the Molecular Dynamics STORM system. The advantage of this approach is the ability to define and quantify each product precisely. The disadvantage is that the analysis can be timeconsuming, since it is usually necessary to define individual product bands. With either of the quantification methods mentioned above, the products should be normalized to the total fluorescence recovered. This is similar to the approach used for the intact cell PLD assay, and the reasons for normalization are similar. The fluorescence of all products should be summed (in arbitrary units), and the individual products divided by this sum to give the percentage of total fluorescence. It is possible to convert these values to mols of product formed, since the amount of substrate present in the reaction is known. 3.2.2 Interpretation of data The separation of products from a PLD assay utilizing mammalian cell membranes and a plant PLD preparation is shown in Figure 2. Both samples were incubated in the presence of butanol. The transphosphatidylation product, BODIPY-phosphatidylbutanol (B-PBt), is generated by both samples. Note that BODIPY-phosphatidic acid (B-PA), the hydrolysis product, is seen only with the plant enzyme. B-PA is also generated by yeast membranes. BODIPY-lyso-phosphatidylbutanol (L-BPBt) is observed in the reaction with mammalian membranes. L-BPBt is generated as a result of the action of a calcium-independent PLA2 on B-PBt. This enzyme activity is present in membranes prepared from yeast and mammalian cells. If L-BPBt is a major product, as is often the case, it may be necessary to use the sum of B-PBt and L-BPBt as the value for PLD activity. Similarly, when both B-PBt and B-PA are generated, the sum of these products represents PLD activity. Other products detected in the assay with mammalian membranes include BODIPY-lysophosphatidic acid (B-LPA), BODIPY-diglyceride (B-DG), and BODIPY-monoglyceride (B-MG). The migration positions (Rf values) of these products are listed in Table 1, along with a summary of the major routes for their generation. A scheme showing pathways for BPC metabolism is presented in Figure 3. An experiment showing the effects of PMA on membrane PLD activity is presented in Figure 4. Membranes were prepared from PC12 cells that had 315

Kathryne E. Meier and Terra C. Gibbs Table 1. HF values for BPC reaction products BODIPY-product

Abbreviation

«F

Primary routes for generation

Lysophosphatidylcholine Phosphatidylcholine Lyso-phosphatidic acid Phosphatidic acid Lyso-phosphatidylbutanol Phosphatidylbutanol Monoglyceride Diglyceride

L-BPC B-PC L-BPA B-PA L-BPBt B-PBt B-MG B-DG

0.31' 0.51 0.57 0.65 0.72 0.80 0.92 0.98

PLA2" (Substrate) PLD, then PLA2; PLA2, then lyso-PLDc PLD PLD (+ butanol),then PLA2 PLD (+ butanol) PLD, then PLA2, then PAP" PLD, then PAP; PLC

"The Rf values shown are from a single representative example. Actual values can vary considerably depending on the chromatography conditions. However, the relative migration of the products are reproducible using the protocol described herein. 6 Abbreviations used: B, BODIPY; PAP, phosphatidate phosphohydrolase; PLA2, phospholipase A2; PLC, phospholipase C; PLD, phospholipase D. c L-BPA often co-migrates with the substrate, BPC. d B-MG can potentially be generated through several pathways (see Figure 4). B-MG and B-DG are frequently not resolved from each other, since both migrate near the solvent front.

Figures. Scheme showing products generated in the in vitro PLD assay. The abbreviations used are defined in Table 7. This scheme does not show all possible reactions involving the BPC substrate, but rather summarizes the reactions that we have observed to occur in mammalian membranes.

been incubated for the indicated times with 100 nM PMA. Activation of PLD is detected by the increase in B-PBt production. In this experiment, the B-PBt levels were quantified using NIH Image software. PMA increased PLD activity by 45% at 5 min. The activity returned to nearly basal levels (10% increase) by 20 min. B-PA was not produced by PC12 membranes incubated with or without butanol. With peanut PLD, B-PA was generated both in the absence and presence of butanol. B-LPBt was not observed as a major product in PC12 membranes in this particular experiment. 316

12: Phospholipase D and phosphatidylcholine metabolism

Figure 4. Phospholipase D assay in mammalian membrane preparations, PC12 cells were incubated with 100 nM PMA for the indicated times. Membranes were then prepared and assayed for PLO activity using the in vitro fluorescent assay. Controls include peanut PLD incubated with and without 1.15% butanol (right), PC12 membranes incubated without butanol (right), and substrate alone (left). Only the most relevant portion of the TLC plate is shown. The positions of the fluorescent products are indicated on the right. The TLC plate was imaged using a Molecular Dynamics Fluorlmager.

3,2.3 Analysis of other types of samples The information provided above refers to assays conducted with mammalian cell membranes. Assays with soluble plant or bacterial enzymes can be performed similarly. This assay has been used to characterize PLD activity in yeast membranes (7,12). In this case, the assay is the same, but the membrane preparation differs. In brief, yeast membranes are prepared by breaking the cells with glass beads. This procedure is detailed in ref, 12 (sec also Chapter 10). The assay has also been applied to samples of mammalian tissue (13). In this case, there are several additional considerations. It is advisable to include protease inhibitors in the homogenization buffer. It is necessary to remove insoluble debris by a low-speed centrifugation after tissue homogcnization. Finally, we found it necessary to extract the final reaction mixture with chlorofornrmethanol (9:1) prior to spotting (he reaction on the TLC plate. This was done because tissue components interfered with the chromalography if the entire reaction mixture was applied to the plate. This procedure is detailed in ref. 13.

Acknowledgements The authors thank Dr. G. Patrick Meier for helpful suggestions. The work of Drs. Krishna M. Ella, Linda G. Jones, and Cynthia D. Bradshaw was critical to the development of the assays described herein. Support from the United 317

Kathryne E. Meier and Terra C. Gibbs States Department of Defense (DAMD 17-98-8524) and the National Science Foundation (EPS-9630167) is gratefully acknowledged.

References 1. Exton, J. H. (1997). PhysioL Rev., 77, 303. 2. Singer, W. D., Brown, H. A., and Sternweis, P. C. (1997). Annu. Rev. Biochem., 66,475. 3. Morris, A. J., Hammond, S. M., Colley, C., Sung, T. C., Jenco, J. M., Sciorra, V. A., Rudge, S. A., and Frohman, M. A. (1997). Biochem. Soc. Trans., 25,1151. 4. Martin, A., Saqib, K. M., Hodgkin, M. N., Brown, F. D., Pettit, T. R., Armstrong, S., and Wakelam, M. J. (1997). Biochem. Soc. Trans., 25,1157. 5. Wang, X., Xu, L., and Zheng, L. (1994). J. Biol. Chem., 269,20312. 6. Rose, K., Rudge, S. A., Frohman, M. A., Morris, A. J., and Engebrecht, J. A. (1995). Proc. Natl. Acad. Sci. USA, 92,12151. 7. Ella, K. M., Dolan, J. W., Qi, C., and Meier, K. E. (1996). Biochem. J., 314,15. 8. Waksman, M., Eli, Y., Liscovitch, M., and Gerst, J. E. (1996). J. Biol. Chem., 271, 2361. 9. Hammond, S. M., Altshuller, Y. M., Sung, T.-C, Rudge, S. A., Rose, K., Engebrecht, J., Morris, A. J., and Frohman, M. A. (1995). J. Biol. Chem., 270, 29640. 10. Colley, W. C., Sung, T.-C., Roll, R., Jenco, J., Hammond, S. M., Altshuller, Y., Bar-Sagi, D., Morris, A. J., and Frohman, M. A. (1997). Curr. Biol, 7,191. 11. Kodaki, T., and Yamashita, S. (1997). J. Biol. Chem., 272,11408. 12. Ella, K. M., Dolan, J. W., and Meier, K. E. (1995). Biochem. J., 307,799. 13. Ella, K. M., Meier, G. P., Bradshaw, C. D., Huffman, K. M., Spivey, E. C., and Meier, K. E. (1994). Anal. Biochem., 218,136. 14. Kanaho, Y., Yokozeki, T., and Kuribara, H. (1996). J. Lipid Med. Cell Sig., 14, 223. 15. van Blitterswijk, W. J., and Hilkmann, H. (1993). EMBO J., 12,1655. 16. Ella, K. M., Meier, K. E., Kumar, A., Zhang, A., and Meier, G. P. (1997). Biochem. Mol. Biol. Intl., 41,715. 17. Waksman, M., Tang, X., Eli, Y., Gerst, J. E., and Liscovitch, M. (1997). J. Biol. Chem., 272,36. 18. Pai, J.-K., Siegel, M. I., Egan, R. W., and Billah, M. M. (1988). Biochem. Biophys. Res. Commun., 150,355. 19. Randall, R. W., Bonser, R. W., Thompson, N. T., and Garland, L. G. (1990) FEBS Lett., 264,87. 20. Daniel, L. W., Huang, C., Strum, J. C., Smitherman, P. K., Greene, D., and Wykle, R. L. (1993).J. Biol. Chem., 268,21529. 21. Song, J., and Foster, D. A. (1993). Biochem. J., 294,711. 22. Liscovitch, M., and Chalifa-Caspi, V. (1996). Chem. Phys. Lipids, 80,37. 23. Jones, L. G., Ella, K. M., Bradshaw, C. D., Gause, K. C., Dey, M., WisehartJohnson, A., Spivey, E. C., and Meier, K. E. (1994). J. Biol. Chem., 269,23790. 24. Ella, K. M., Qi, C., McNair, A. F., Park, J.-H., Wisehart-Johnson, A. E., and Meier, K. E. (1997). J. Biol. Chem., 272,12909. 25. Morris, A. J., Frohman, M. A., and Engebrecht, J. (1997). Anal. Biochem., 252,1. 318

12: Phospholipase D and phosphatidylcholine metabolism 26. Olson, S. C., Bowman, E. P., and Lambeth, J. D. (1991). J. Biol. Chem., 266,17236. 27. Meier, K. E., Gause, K. C., Wisehart-Johnson, A. E., Gore, A. C. S., Finley, E. L., Jones, L. G., Bradshaw, C. D., and Ella, K. M. (1998). Cell. Signal., 10,415. 28. Qi, C, Park, J.-H., Shirley, D. W., Bradshaw, C. D., Ella, K. M., and Meier, K. E. (1997)./. Cell. Physlol., 174,261. 29. Sansbury, H. M., Fulwood, S., Wisehart-Johnson, A. E., Qi, C., and Meier, K. E. (1997). Carcinogenesis, 18,1817.

319

This page intentionally left blank

Is] Signal transduction by sphingosine kinase DAGMAR MEYER ZU HERINGDORF, CHRIS J. VAN KOPPEN and KARL H. JAKOBS

1. Introduction Several new findings in the last few years have drastically challenged our view of the role of sphingolipids in cellular function. It is now evident that not only the complex sphingolipids of the plasma membrane, but also small molecules of sphingolipid metabolism, play an essential role in cellular regulation. Analogous to the breakdown products of glycerophospholipids, certain small sphingolipids are messenger molecules which transduce information both between cells and within cells. Of these, ceramide and sphingosine-1phosphate (SPP) have gained major attention (1), although others may also have a signalling function. Ceramide is cleaved from sphingomyelin by sphingomyelinase(s) (see Figure 1), and apparently plays an important role in apoptosis. SPP was originally considered as an intermediate product in the breakdown of sphingosine, but now it is recognized that it controls a variety of cellular functions. SPP is produced from sphingosine by sphingosine kinase (2), and cleaved either by a specific lyase (3) or phosphatase(s) (4), see Figure 1. Sphingosine kinase has recently been purified from rat kidney (5). Yeast analogues of the lyase and phosphatase enzymes have been cloned (6, 7). Considering the wide spectrum of cellular events which are controlled by SPP (see below), it is evident that investigation of the enzymes that regulate SPP levels is of great and challenging importance. In the last few years, our knowledge about SPP as a signalling molecule has increased in a nearly exponential manner. While until recently it was a matter of debate whether SPP is an intracellular second messenger or a ligand at a plasma membrane receptor, it is now proven that SPP can have both types of function (see Figure 2).

1.1 G protein-coupled sphingolipid receptors Evidence that extracellular SPP acts by activating a specific G protein-coupled receptor (GPCR), rather than an intracellular target, was primarily gained

Dagmar Meyer zu Heringdorfet al.

Figure 1. Overview of the metabolism of sphingosine-1-phosphate (SPP). The numbers indicate catalysis by the following enzymes: (1) sphingomyelinase, (2) sphingomyelin synthase, (3) ceramidase, (4) sphingosine N-acyltransferase, (5) sphingosine kinase, (6) sphingosine-1-phosphate phosphatase, (7) sphingosine-1-phosphate lyase, (8) serine palmitoyltransferase.

from experiments on atrial myocytes. In excised membrane patches from these cells, SPP activated the acetylcholine-dependent K+ current (IK(ACh)) in a Gi protein-dependent manner, but only from the extracellular face of the membrane (8). Similarly, SPP caused neurite retraction in neuroblastoma cells only when applied extracellularly, and not on microinjection (9); this action, however, was dependent on pertussis toxin (PTX)-insensitive G proteins. Finally, using SPP fixed to glass beads, it could be demonstrated that SPP acts from the cell exterior at a plasma membrane receptor to regulate cell motility, and does not require uptake into the cells (10). SPP receptors regulate a variety of responses in many different cell types. Moreover, not only SPP but also other lysosphingolipids, such as sphingosylphosphorylcholine or glucopsychosine, can apparently act via GPCRs. Comparing the activities of these lipid mediators in different cell types, the existence of several sphingolipid receptor subtypes is highly likely. This subject is more extensively discussed in ref. 11. Very recently, the concept of specific sphingolipid receptors has been vali322

13: Signal transduction by sphingosine kinase

Figure 2. Roles of sphingosine-1-phosphate (SPP) in signal transduction. SPP, the product of sphingosine kinase (SPK), is both a ligand at a plasma membrane receptor and an intracellular second messenger. Intraceltular production of SPP is triggered by activation of SPK by various G protein-coupled receptors (GPCRs), e.g. muscarinic acetylcholine receptors, as well as by PDGF and antigen (FcRI receptors. Extracellular SPP may be derived from activated platelets or from as yet unknown sources. Coupling of SPP plasma membrane receptors may occur via pertussis toxin-sensitive and -insensitive G proteins (Gi Gx).

dated by studies on the orphan GPCRs of the EDG family. Certain members of this receptor family mediate cellular responses to lysophosplialidic add, and others to SPP or sphingosylphosphorylcholine. Thus, genes of the EDG family appear to be the molecular correlates to the above mentioned receptors (12-14).

1.2 Receptor regulation of sphingosine kinase and the role of intracellular SPP Sphingosine kinase provides SPP. which acts either directly inside the cell, or which is released to activate SPP receptors in an autoerine or paracrine fashion. There is ample evidence that SPP can act independently of GPCRs. For example, microinjcetion of SPP into HEK-293 cells leads to rapid release of stored C a - , even when extracellular SPP is completely inactive because signalling via GPCRs has been knocked out by PTX (15). Furthermore, muscarinic aeetylcholinc receptor (mAChR)-emediated Ca2. signalling, which involves sphingosine kinase activation, as outlined below, is not influenced by desensitization of SPP receptors (see Figure 3). Until now, only a few studies deal directly with receptor regulation of 323

Dagmar Meyer zu Heringdorfet al.

Figure3. Desensitization of SPP receptors leaves sphingosine kinase-mediated signalling intact. Intracellular free Ca2+ concentration ([Ca2+]i) was measured in cells loaded with the Ca2+-sensitive fluorescence indicator fura-2, as described, for example, in ref. 15. HEK-293 cells expressing the m2 muscarinic acetylcholine receptor (which uses sphingosine kinase to mobilize intracellular Ca2+, see ref. 15) were challenged with BSA solution (see Protocol 7), SPP (1 y.M in BSA solution) or carbachol (1 mM). Additions were made every 60 s in the indicated order. Note that after a challenge with SPP, a second addition of this agonist no longer increases [Ca2+],, indicating desensitization of SPP receptors. However, Ca2+ mobilization by the mAChR agonist carbachol remains unaffected. Thus, SPP which is produced intracellularly after stimulation of the m2 receptor acts independently of plasma membrane SPP receptors. Values are mean ± SD from triplicates of a representative experiment.

sphingosine kinase and the mechanisms involved in this process. Stimulation of intracellular SPP production has been observed upon activation of tyrosine kinase-linked receptors, such as platelet-derived growth factor (PDGF) receptors (16, 17) and antigen receptors (18, 19). Recently, it was recognized that members of the large family of GPCRs as well are able to stimulate SPP production by sphingosine kinase. This was demonstrated for mAChRs overexpressed in HEK-293 cells, and for formyl peptide receptors endogenously expressed in HL-60 cells (15, 20). Inhibition of sphingosine kinase was used to determine the cellular events downstream of the sphingosine kinase reaction. For example, it has been reported that PDGF-stimulated mitogenesis in NIH3T3 cells, antigen-induced Ca2+ mobilization and vesicle trafficking in monocytes, as well as activation of extracellular signal-regulated kinase by various receptors in airway smooth muscle cells, are dependent on sphingosine kinase (16, 18, 19, 21). Furthermore, Ca2+ mobilization by various GPCRs in different cell types was distinctly sensitive to sphingosine kinase inhibition (15,20). One intracellular target of SPP appear to be Ca2+ stores (15,22), although the cloned 'sphingolipid Ca2+ release-mediating protein from endoplasmic reticulum' is gated by sphingosylphosphorylcholine, and not by SPP (23). Thus, it may be speculated that sphingosine kinase is the long searched-for phospholipase C-independent Ca2+ mobilization pathway of several receptors. 324

13: Signal transduction by sphingosine kinase

2. How to study involvement of sphingosine kinase in a certain signalling pathway 2.1 'Positively': testing the intracellular activity of SPP We would like to stress that it is not appropriate to study the effects of intracellular SPP by simply adding SPP to intact cells. This was done in earlier work on SPP, before discovery of the above-mentioned specific SPP receptors. In the meantime, the widespread occurrence of SPP plasma membrane receptors has been recognized. Moreover, it is not clear whether extracellular SPP can cross plasma membranes at all. PTX treatment of cells may be used to check whether Grtype G proteins are involved in the action of SPP. At present, intracellular microinjection and caged SPP are the two methods of choice to study intracellular actions of SPP. 2.1.1 PTX treatment PTX, by ADP-ribosylating certain G-protein a-subunits, leads to uncoupling of GPCRs from the respective G proteins (see also Chapter 5). As a consequence, the signalling pathways depending on these PTX-sensitive G proteins are disrupted. Usually, treatment of intact cells with 100 ng ml-1 of the toxin for 24 h will suffice to achieve more than 90% inhibition. Be aware of the following: (a) inhibition of the action of SPP by PTX strongly points towards involvement of SPP receptors. (b) no influence of PTX may either imply a G protein-independent pathway, or involvement of PTX-insensitive G proteins. 2.1.2 Microinjection SPP may be microinjected into intact cells to circumvent the plasma membrane barrier. However, microinjection techniques are hampered by highpressure application of the injection solution at the cell surface. Since SPP receptors are expressed by many cell types, and are often coupled to Gitype G proteins, we recommend blocking the function of these SPP receptors by pretreatment of the cells with PTX. 2.1.3 Caged SPP Caged SPP has recently been developed to get SPP into intact cells. For this, cells are preloaded with caged SPP and washed, and then the active lipid is released from the caged compound by a flash of UV light. The reader is referred to ref. 24. However, since caged SPP is a lipophilic compound, it may flow out of the preloaded cell before it is activated by the flash of light. As a consequence, a small amount of extracellular SPP may be generated, which could be sufficient to activate the high-affinity SPP receptors, which are usually activated by nanomolar concentrations of SPP. 325

Dagmar Meyer zu Heringdorfet al.

Figure 4. Specificity of sphingosine kinase inhibitors in HEK-293 cells. HEK-293 cells expressing the m2 muscarinic acetylcholine receptor were treated without and with 100 uM carbachol. Phospholipase C (PLC), sphingosine kinase (SPK), and increase in intracellular free Ca2+ concentration ([Ca2+]i) were measured as described in ref. 15. DHS and DMS were dissolved in BSA solution as described in Protocol 1. Pretreatment was for 10 min with 30 uM DHS and 15 uM DMS, or the indicated concentrations of sphingosine kinase inhibitors. Basal and carbachol-stimulated SPP production, as well as carbacholstimulated Ca2+ mobilization, were largely inhibited. In contrast, basal and carbacholstimulated phospholipase C activity remained unaffected. Data are mean ± SEM (n = 4) in (A) and (B), and mean ± SEM (n = 3) in (C). (A) and (C) are adapted from ref. 15.

2.2 'Negatively': inhibiting sphingosine kinase Potent and specific inhibitors of sphingosine kinase comprise DL-threodihydrosphingosine (DHS; ref. 25) and N, W-dimethylsphingosine (DMS; ref. 26). They have been widely used to demonstrate involvement of sphingosine 326

13: Signal transduction by sphingosine kinase kinase in receptor-controlled cellular events, such as Ca2+ mobilization and DNA synthesis (see Introduction). IC50 values reported are 5-50 uM for DHS and 5-10 uM for DMS. Specificity of these inhibitors is underlined by the observation that they inhibit only certain signalling events. For example, phospholipase C stimulation is generally not affected (see Figure 4). Furthermore, in HL-60 leukaemia cells, only receptor-stimulated Ca2+ mobilization and p-glucuronidase release was inhibited by DHS and DMS, but not phospholipase C stimulation and production of superoxide anions (20). Since the above-mentioned sphingosine kinase inhibitors are not readily soluble, we provide a protocol how to handle them. Protocol 1. Application of sphingosine kinase inhibitors Reagents • For dilution of the lipids, prepare a solution of 1 mg ml~1 BSAa in water (instead of water, an aqueous buffer solution may also be used)

Method 1. Dissolve DHS or DMS (both from Biomolb) at 10 mM in methanol, and store in aliquots at -20°C. 2. Prewarm aliquots of the stock solutions to 37°C. Add prewarmed BSA solution to a final concentration of 600 uM of the lipids. Mix immediately. 4. Do not put the lipids on ice. Instead, use only freshly prepared sphingosine kinase inhibitors. They are stable at room temperature for some hours. Decline in activity may be due to the lipids sticking to the tube wall. To minimize sticking, use polypropylene tubes. 5. Preincubate cells or subcellular preparations with BSA complexes of DHS or DMS for 1-10 minc before addition of receptor agonists or other stimulatory agents. Check different concentrations of the respective receptor agonists, since the contribution of sphingosine kinase might be dependent on the level of receptor activation. a

Use fatty acid-free BSA, and always check whether there is an effect of the BSA solution itself. In our experience, DHS from Sigma has some impurities, e.g. it causes an increase in intracellular Ca2+ in HEK-293 cells, whereas DHS from Biomol does not. c Sphingosine kinase inhibitors are specifically useful to study the contribution of the enzyme to fairly immediate receptor reactions. Long-term treatment with DHS or DMS at high concentrations may be cytotoxic. If your total incubation time with DHS or DMS exceeds 1 h, check cell viability. b

3. How to study regulation of sphingosine kinase Only a few reports exist in the literature dealing directly with regulation of sphingosine kinase (see Introduction). Nevertheless, a number of different 327

Dagmar Meyer zu Heringdorf et al. methods have been used. This reflects the fact that this field of research has just started, and that there are still some problems with the assay. One major difficulty is the solubility of the sphingosine kinase product, SPP, which is generally poor, but not restricted to either polar or nonpolar solvents. Other problems are the rapid degradation of the SPP formed, and the rapid transformation of sphingolipids into each other (see Figure 7). Sphingosine kinase activity can be analysed in intact cells, as well as in subcellular fractions. Receptor regulation of the enzyme may be better studied in intact cells. However, after stimulation of intact cells with PDGF or antigen, an increased Vmax of sphingosine kinase compared to control cells was observed in cytosolic preparations (16, 18). Although SPP production by sphingosine kinase is most often studied by measuring conversion of labelled precursors to labelled SPP (see below), quantification of SPP using acetylation with [3H] acetic anhydride is also possible (27). This method has been successfully applied to analyse levels of SPP in human body fluids (28) and rat tissues (29). For this, SPP is extracted from tissues and cells in two steps, first under alkaline conditions into the aqueous phase of a Folch partition (to remove sphingosine and other lipids which are also N-acetylated), then under acidic conditions back into the chloroform phase, which is then used for the acetic anhydride reaction. The product is washed and separated by TLC, and can be quantified by scraping and counting of the spots. The authors report a recovery of 37 ± 6%, and a sensitivity between 30 pmol and 10 nmol of SPP. For details, refer to ref. 27.

3.1 Measurement of sphingosine kinase activity in intact cells 3.1.1 Overview Assaying sphingosine kinase involves the following steps: the kinase reaction itself performed in the presence and absence of putative stimuli, extraction of the lipids, and separation and quantification of the product. Table 1 gives an overview on the methods used by different groups. 3.1.2 Labelling of cells versus direct application of labelled sphingosine Labelling cells with 32P; not only requires high amounts of radioactivity, but also leads to a large number of labelled molecular species. The same is true for labelling cells with [3H] or [14C]serine, which is incorporated into the cellular sphingolipids by the serine palmitoyltransferase reaction (see Figure 1). As a consequence, a two-dimensional TLC is generally required to separate and identify SPP. Another possibility is to offer [3H]sphingosine directly to cells. Since most biological activities of extracellularly applied sphingosine (e.g. increase in [Ca2+]i, and mitogenesis) are observed at sphingosine concentrations > 100 nM, it is possible to add a sufficient amount of radioactivity 328

13: Signal transduction by sphingosine kinase Table 1. Sphingosine kinase assays in intact cells Reference

Assay principle

Lipid extraction

30

Labelling of cells 32 with Pi

Chloroform:methanol:HCI Two-dimensional TLC on (100:200:1 by volume), silica gel 60 G (lipids of the phase separation chloroform phase)

25

Labelling of Methanol:chloroform platelets with 32Pi (2:1 by volume) + 5% triethylamine

16

Labelling of cells with [3H]serine Incubation with [3H]sphingosine

31

18

Labelling of cells with 32Pi

Details not reported

Separation/Quantification of SPP

Conversion of SPP to Ncaproyl-SPP, alkaline hydrolysis, phase separation, one-dimensional TLC on silica gel 60 G Two-dimensional TLC

Chloroform:methanol:HCI One-dimensional TLC on silica gel 60 G (100:200:1 by volume), phase separation Details not reported Two-dimensional TLC

without directly stimulating the cells. As reported for different cell types, the applied [3H]sphingosine is rapidly converted to [3H]SPP, whether or not a stimulus is present (15, 20, 30, 31; see also Figure 5). Stimulation of receptors rapidly and transiently enhances formation of [3H]SPP. The time course and magnitude of receptor-stimulated [3H]SPP formation from directly offered [3H]sphingosine were generally very similar in different cell types, and closely resembled those of receptor-stimulated SPP formation in prelabelled cells (16, 18). 3.1.3 Advantages and disadvantages of phase separation At alkaline pH, SPP largely distributes into the aqueous phase, and sphingosine into the organic phase. At acidic pH, SPP is evenly distributed between both phases (31). Most protocols for sphingosine kinase assays include a phase separation. However, in our experience, phase separation can lead to large intra- and inter-assay variations (see Table 2). Thus, it may be better to avoid phase separation at all when extracting SPP from cells. Instead, cells can be extracted with methanol:chloroform (2:1, v/v) in the absence of water. By this procedure, proteins are denatured, and can be removed by centrifugation (see Protocol 2). 3.1.4 TLC analysis of SPP SPP migrates and separates from sphingosine, e.g. in the solvent systems of Table 3. 329

Dagmar Meyer zu Heringdorf et al. Table 2. Phase distribution of SPP produced by sphingosine kinase in intact cells Total activity

Organic phase

Aqueous phase (calculated) SPP Sphingosine Other

c.p.m.

%

c.p.m.

%

c.p.m.

%

1100 ± 540 200 ± 30 900 ± 460

76

350 ± 240 8000 ± 3600 4800 ± 3000

24 98 84

1450 8200 5700

10 53 37

2 16

HEK-293 cells suspended in HBSS were incubated with [3H]sphingosine for 10 min at 37 °C in a total volume of 200 ul. The reaction was stopped by addition of 2 ml chlorofom:methanol (2:1, v/v) and 400 ul of 10 mM NaOH. pH was > 12. Tubes were vortexed and centrifuged, and phases were separated. The aqueous phase was neutralized with HCI, and both phases were dried down. Lipids were separated by TLC, and quantified as described in Protocol 2. Data are mean ± SD(n = 4) from a typical experiment.

Table 3. Mobilities of SPP and sphingosine in selected solvent systems on silica gel 60 G TLC plates Solvent system (v/v) n-butanol:acetic acid:water (3:1:1) Chloroform:methanol:water (60:35:8)

RF value SPP

Sphingosine

0.51 0.25

0.67 0.39

RF values of several other lipids in these and other solvent systems can be found in ref. 4.

Spots of standard lipids can be visualized with (a) iodine vapour: (i) non-selective staining of lipids (ii) least sensitivity (b) ninhydrin: (i) detects free amino groups (ii) stains e.g. sphingosine and SPP, but not ceramide (c) molybdenum blue: (i) detects phosphate 3.1.5 Inhibition of SPP degradation SPP lyase is a pyridoxal phosphate-dependent enzyme, and thus can be inhibited by 4-deoxypyridoxine (500 uM). L-canaline (Sigma, 100 |uM) is also inhibitory. Although both compounds have been applied, there is no consistent information about their effectiveness. There is no specific inhibitor of SPP phosphatases. 3.1.6 Method At the present time, it is obvious that we can only suggest how to perform a sphingosine kinase assay in intact cells. The assay procedure of Protocol 2 330

13: Signal transduction by sphingosine kinase

Figure 5. Formation of [3H]SPP from [3H]sphingosine in HEK-293 cells. Sphingosine kinase activity was measured in intact HEK-293 cells according to Protocol 2. (A) Time course of basal [3H]SPP production from [3H]sphingosine. [3H]SPP levels reach a plateau after 2-3 min. (B) Quantification of intracellular [3H]SPP, [3H]sphingosine, and other [3H]labelled molecular species 30 sec, 1 min, and 60 min after [3H]sphingosine addition. Theoretically, sphingosine can be metabolized by cells to SPP, ceramide, or W-methylsphingosine (see Figure 1). (C) Time course of carbachol-stimulated [3H]SPP formation, expressed as percentage above the basal level. Activation of the m3 muscarinic acetylcholine receptor leads to a rapid and transient elevation of [3H]SPP levels. Note that stimulation is maximal at 30 sec, about 100% above basal, and very transient. Data in (A) and (B) are mean ± SD (n = 4) from representative experiments. (C) is adapted from ref. 15.

was successful for detecting stimulation of the enzyme by various GPCRs (15, 20). It is based on addition of [3H]sphingosine together with the stimulus, avoids a phase separation, and identifies [3H]SPP after onedimensional TLC by scraping of the spots followed by liquid scintillation counting. 331

Dagmar Meyer zu Heringdorfet al. Protocol 2. Sphingosine kinase assay in intact cells Equipment and reagents • Poly-L-lysine, fatty acid-free BSA, ninhydrin, and molybdenum blue spray reagents (Sigma Chemical Company) • [3H]sphingosine (15-20 Ci mmol-1, American Radiolabeled Chemicals Inc., or NEN Life Science Products), unlabelled sphingosine, and sphingosine-1-phosphate (both Biomol) • Speedvac centrifuge (Savant)

• Chloroform, methanol, 1-butanol, acetic acid (Merck) . Silica gel 60 G TLC plates, channelled (Merck or Whatman) . Hanks' balanced salt solution (HBSS): 118 mM NaCI, 5 mM KCI, 1 mM CaCI2, 1 mM MgCI2, 5 mM D-glucose, 15 mM Hepes, pH 7.4

Method 1. Plate cells on poly-L-lysine-coated 60 mm diameter plates, and allow to grow overnight until about 80% confluent. 2. Wash plates with HBSS or 25 mM Hepes-buffered cell culture medium, and equilibrate for 60 min at 37°C. Use three extra plates for protein measurement. 3. Remove the buffer, and add 2 ml of fresh buffer containing 1 mg ml-1 fatty acid-free BSA and [3H]sphingosine (100 000-200000 c.p.m, final concentration 2-10 nM), with or without receptor agonist or other stimulus. 4. Incubate plates for 10 s-10 min at 37°C. 5. Stop incubation by rapid removal of the medium, and addition of 1 ml ice-cold methanol. Put the plates on ice. 6. Scrape the cells from the plates, and transfer to polypropylene tubes. Scrape a second time with 1 ml of ice-cold methanol. Add 1 ml chloroform to the tubes. 7. Vigorously vortex the tubes for 10 s. Centrifuge tubes for 10 min at 2000 g, and pour the supernatant into a new tube. 8. Re-extract the pellet with 3 ml methanol:chloroform (1:2). Combine the supernatants, and evaporate them to dryness in a Speedvac centrifuge. 9. Dissolve the samples in 25 ul methanol, and spot onto Silica gel 60 G TLC plates. Rinse the tubes with another 25 ul of methanol. Apply standard lipids, using 5 ul of 10 mM sphingosine and 20 ul of 1 mM sphingosine-1-phosphate for identification of [3H]sphingosine and [3H]sphingosine-1-phosphate, respectively. 10. Perform TLC in 1-butanol:acetic acid:water (3:1:1). 11. Visualize sphingosine-1-phosphate (RF value of 0.51) and sphingosine (RF value of 0.67) on the plates by spraying with ninhydrin reagent (at room temperature), and subsequent development for 15 min at 80°C.

332

13: Signal transduction by sphingosine kinase If necessary, sphingosine-1-phosphate (but not sphingosine) spots can be stained more strongly with molybdenum blue at room temperature. 12. Mark the sphingosine-1-phosphate and sphingosine spots with a pencil. Spray the plates with distilled water (to prevent dust formation), and scrape identical surface areas into scintillation vials. Add 3.5 ml of scintillation fluid, vortex, and count.

Sphingosine binds extremely strongly to plastic tubes and pipette tips. It is therefore advisable to use pipette tips only once, and to include BSA as a [3H]sphingosine carrier in aqueous solutions. As serum lipids (like lysophosphatidic acid) are able to activate sphingosine kinase, it is necessary to use delipidated (fatty acid-free) BSA. One should verify that a given fatty acid-free BSA batch is indeed ineffective in the cells under study. Typically, peak increases in receptor-stimulated [3H]SPP formation are no larger than 50-100% above basal [3H]SPP formation, and are very transient. Consequently, the assays have to be performed at least in triplicate, and timecourse studies are usually required to detect the receptor effects. To ensure that sphingosine kinase activity is indeed measured, basal and receptorstimulated SPP production should be efficiently inhibited by DHS or DMS (see above). Sphingosine kinase activity can also be assayed on cells in suspension. Since some cell types release [3H]SPP upon addition of [3H]sphingosine (10), we recommend removing the extracellular medium from the cells before extraction of the lipids. Since the time course of receptor-stimulated SPP production by sphingosine kinase is rapid, we prefer filtration of the cells to centrifugation. Protocol 3. Modification of sphingosine kinase assay (Protocol 2) for cells in suspension Additional equipment • Whatman GF/C filters

• Filtration unit

Method 1. Wash cells twice with ice-cold HBSS, and resuspend at a density of 5-10 x 106 cells ml-1. 2. Prepare a reaction mix containing 2 mg ml-1 fatty-acid free BSA, and 1 uCi ml-1 [3H]sphingosine in HBSS. 3. Prewarm tubes containing 100 ul reaction mix, with and without the chosen stimulus, for 5 min at 37°C. In parallel, prewarm the cells to 37°C. 333

Dagmar Meyer zu Heringdorf et al. Protocol 3. Continued 4. Start the reaction by adding 100 ul of cell suspension to each tube. 5. Stop the reaction after the intended incubation time by addition of 1 ml ice-cold HBSS, and rapid filtration through GF/C filters.a Wash once with 1 ml of ice-cold HBSS. 6. Put the filters into polypropylene tubes, and add 2 ml methanol, 1 ml chloroform to extract the lipids. 7. Continue with step 7 of Protocol 2. a

If you are sure that your cells do not release SPP, you can stop the reaction by simply adding 2 ml methanol, 1 ml chloroform, and continue with Step 7 of Protocol 2. The small amount of water from the reaction resolves completely in the excess of methanol:chloroform, and does not cause phase separation.

3.2 Measurement of sphingosine kinase activity in subcellular fractions Measuring sphingosine kinase activity in subcellular fractions can be performed using [y-32P]ATP, or [3H]sphingosine as substrates. Most often, sphingosine (labelled and unlabelled) is provided as a BSA complex; however, it can also be dissolved in Triton X-100 (5) or provided in mixed micelles with octyl-(p-Dglucopyranoside (25). Detergents or glycerol, etc., may interfere with the TLC separation, and probably have to be removed employing a phase separation. The discussion in Section 3.1. about lipid extraction, phase separation, and TLC analysis may also be applied to the assay in subcellular fractions. Here, an assay currently used in our laboratory is described. Protocol 4. Sphingosine kinase assay in subcellular fractions Equipment and reagents • Equipment and reagents of Protocol2 . Lysis buffer: 100 mM potassium phosphate, pH 7.4; 1 mM EDTA; 10 mM MgCI2; 1 mM 2mercaptoethanol

• Reaction mix: 100 mM potassium phosPhate, pH 7.4; 2 mM ATP; 10 mM MgCI2; 1 mM EDTA; 1 mM 2-mercaptoethanol; 2 mg ml-1 fatty acid-free BSA; 20 uM sphingosine; 0.5 uCi ml-1 [3H]sphingosine

A. Preparation of cytosola 1. Wash cells in HBSS, and resuspend in ice-cold lysis buffer at a density of 1-10 x 106 cells ml-1. 2. Disrupt cells on ice by Dounce homogenization.b 3. Centrifuge for 20 min at 50 000 g and 4 °C. 4. Re-centrifuge the supernatant for 60 min at 100000 g and 4°C. 5. Determine protein concentration of the supernatant. 334

13: Signal transduction by sphingosine kinase B. Assaying sphingosine kinase activity 1. Prepare polypropylene tubes containing 100 ul of reaction mix, together with putative stimuli or inhibitors.c Prewarm to 37°C for 5 min before the start of the reaction. 2. Start the reaction by addition of 100 ul of cytosol (100-200 ug protein) to the tubes. 3. Incubate at 37°C for 30 s-20 min. 4. Stop the reaction by addition of 2 ml methanol. Add 1 ml chloroform. 5. Continue with Step 7 of Protocol 2. * You can use either the crude cell-lysate, the 50 000 g supernatant, or the 100 000 g supernatant. Membrane fractions can also be employed. Feel free to fractionate your cells as you like; adjust the buffer conditions to the above described lysis buffer. b Nitrogen cavitation or freeze-thawing are also suitable. c Some inhibitors may require a pretreatment of the cellular preparation.

3.3 Final remarks The sphingosine kinase/SPP pathway is a signal transduction system the significance of which has only very recently been recognized. Most likely, for many G protein-coupled and tyrosine-kinase-linked receptors, it represents a widespread signalling pathway of great importance. Thus, more sphingosine kinase measurements are urgently needed to improve our knowledge of this pathway. The more sphingosine kinase assays are performed and the more we know about the mechanisms involved in the sphingosine kinase regulation, the more the assays and methodologies used to monitor this pathway, which are currently still in their trial-and-error phases, will be improved.

References 1. 2. 3. 4. 5. 6. 7. 8. 9.

Spiegel, S., and Milstien, S. (1995). J. Membrane Biol, 146, 225. Buehrer, B.M., and Bell, R.M. (1993). Adv. Lipid Res., 26, 59. van Veldhoven, P. P., and Mannaerts, G. P. (1993). Adv. Lipid Res., 26, 69. van Veldhoven, P. P., and Mannaerts, G. P. (1994). Biochem. J., 299, 597. Olivera, A., Kohama, T., Tu, Z., Milstien, S., and Spiegel, S. (1998). J. Biol. Chem., 273, 12576. Saba, J. D., Nara, F., Bielawska, A., Garrett, S., and Hannun, Y. A. (1997). J. Biol. Chem., 272, 26087. Mao, C., Wadleigh, M., Jenkins, G. M., Hannun, Y. A., and Obeid, L. M. (1997). J. Biol. Chem., 272, 28690. van Koppen, C. J., Meyer zu Heringdorf, D., Laser, K. T., Zhang, C., Jakobs, K. H., Bunemann, M., and Pott, L. (1996). J. Biol. Chem., 271, 2082. Postma, F. R., Jalink, K., Hengeveld, T., and Moolenaar, W. H. (1996). EMBO J., 15, 2388. 335

Dagmar Meyer zu Heringdorf et al. 10. Yamamura, S., Yatomi, Y., Ruan, F., Sweeney, E. A., Hakomori, S., and Igarashi, Y. (1997). Biochemistry, 36, 10751. 11. Meyer zu Heringdorf, D., van Koppen, C. J., and Jakobs, K. H. (1997). FEBS Lett., 410, 34. 12. An, S., Bleu, T., Huang, W., Hallmark, O. G., Coughlin, S. R., and Goetzl, E. J. (1997). FEBS Lett., 417, 279. 13. Lee, M.-J., Van Brooklyn, J. R., Thangada, S., Liu, C. H., Hand, A. R., Menzeleev, R., Spiegel, S, and Hla, T. (1998). Science, 279, 1552. 14. Zondag, G. C. M., Postma, F. R., van Etten, I., Verlaan, I., and Moolenaar, W. H. (1998). Biochem. J., 330, 605. 15. Meyer zu Heringdorf, D., Lass, H., Alemany, R., Laser, K. T., Neumann, E., Zhang, C., Schmidt, M., Rauen, U., Jakobs, K. H., and van Koppen, C. J. (1998). EMBOJ., 17, 2830. 16. Olivera, A., and Spiegel, S. (1993). Nature, 365,557. 17. Pyne, S., Chapman, J., Steele, L., and Pyne, N. (1996). Eur. J. Biochem., 237, 819. 18. Choi, O. H., Kim, J., and Kinet, J. (1996). Nature, 380, 634. 19. Melendez, A., Floto, R. A., Gilloly, D. J., Harnett, M. M., and Allen, J. M. (1998). J. Biol. Chem., 273, 9393. 20. Alemany, R., Meyer zu Heringdorf, D., van Koppen, C. J., and Jakobs, K. H. (1999). J. Biol. Chem., 274, 3994. 21. Tolan, D., Conway, A.-M., Steele, L., Pyne, S., and Pyne, N. (1996). Br. J. Pharmacol., 119, 185. 22. Ghosh, T. K., Bian, J., and Gill, D. L. (1994). J. Biol. Chem., 269, 22628. 23. Mao, C., Kim, S. H., Almenoff, J. S., Rudner, X. L., Kearney, D. M., and Kindman, L. A. (1996). Proc. Natl. Acad. Sci. USA, 93, 1993. 24. Qiao, L., Kozikowski, A. P., Olivera, A., and Spiegel, S. (1998). Bioorg. Med. Chem. Lett., 8, 711. 25. Buehrer, B. M., and Bell, R. M. (1992). J. Biol. Chem., 267, 3154. 26. Yatomi, Y., Ruan, F., Megidish, T., Toyokuni, T., Hakomori, S., and Igarashi, Y. (1996). Biochemistry, 35, 626. 27. Yatomi, Y., Ruan, F., Ohta, H., Welch, R. J., Hakomori, S., and Igarashi, Y. (1995). Anal. Biochem., 230, 315. 28. Yatomi, Y., Igarashi, Y., Yang, L., Hisano, N., Qi, R., Asazuma, N., Satoh, K., Ozaki, Y., and Kume, S. (1997). J. Biochem., 121, 969. 29. Yatomi, Y., Welch, R. J., and Igarashi, Y. (1997). FEBS Lett., 404, 173. 30. Zhang, H., Desai, N. N., Olivera, A., Seki, T., Brooker, G., and Spiegel, S. (1991). J. Cell Biol., 114, 155. 31. Yatomi, Y., Ruan, F., Hakomori, S., and Igarashi, Y. (1995). Blood, 86, 193.

336

14 Detection, isolation, and quantitative assay of mitogenactivated protein kinases in intact cells and tissues NEIL G. ANDERSON

1. Introduction The landmark discovery of the insulin-stimulated protein kinase, originally termed microtubule-associated protein 2 (MAP-2) kinase, by Sturgill and colleagues in 1987 (1), was of fundamental importance in initiating the huge advances since made in our understanding of how messages received by receptors on the cell surface are transduced to the cell interior to produce specific responses. Although originally isolated from insulin-stimulated 3T3Ll adipocytes, MAP-2 kinase was subsequently shown to be present in all cells, to be activated in response to an enormous variety of extracellular stimuli, and to have been evolutionarily conserved from yeast to man (see ref. 2 for a review). Appreciation of the ubiquitous nature of MAP-2 kinase led to it being renamed mitogen-activated protein (MAP) kinase. Since then, a number of related protein kinases have been isolated. These all have similar structures and mechanisms of activation, but co-exist in cells as members of parallel signalling pathways stimulated by distinct types of extracellular stimulus. Three major classes of MAP kinase have, so far, been characterized in detail. The first class, activated in cells stimulated by many hormones and growth factors, are generally known as extracellular signal regulated kinases ERK-1 and ERK-2, the latter being the original MAP-2 kinase. The other two classes of MAP kinase, generally activated in cells subjected to various stresses such as irradiation or osmotic shock, are the c-jun kinases (JNKs) and the p38 MAP kinases. The major properties of the three classes of mammalian MAP kinase are summarized in Table 1. This chapter details methods available for isolating MAP kinases from stimulated cells or tissues, and for determining their activation status, both

Table 1. Major properties of mammalian mitogen-activated protein kinases MAPK class

Enzymes (Approx. Mr)

Phosphorylation sequence

Activating kinasesa

Extracellular stimuli

ERK

ERK1 (44 kDa) ERK2 (42 kDa)

TEY

MKK1 (MEK1) MKK2 (MEK2)

JNK

SAPK1a/JNK2(54kDa) SAPK1b/JNK3 (54kDa) SAPK1c/JNK1(46kDa) SAPK2a/p38a (38 kDa) SAPK2b/p38p (38 kDa) SAPK3/p38y (38 kDa) SAPK4 (38 kDa)

TPY

SKK1 (MEK4) SKK4 MKK7 SKK2 (MEK3) SKK3 (MEK6)

hormones growth factors phorbol esters UV light ionizing radiation proinflammatory cytokines osmotic shock oxidative stress heat shock

p38

TGY

Abbreviations: ERK, extracellular signal-regulated kinase; MKK, MAP kinase kinase; MEK, mitogen-activated or ERK kinase; SAPK, stress-activated protein kinase; SKK, SAPK kinase. a The activating kinases, in most cases, phosphorylate all members of the class of MAP kinase to which they are ascribed in the Table.

14: Detection, isolation and quantitative assay of mitogen activated quantitatively and semi-quantitatively. The chapter will focus on the ERK class, with reference to the other classes when appropriate.

2. General mechanism of activation of MAP kinases Early studies revealed that MAP kinase was a phosphoprotein, and that its activation required phosphorylation on both threonine and tyrosine residues. This suggested that two upstream protein kinases were involved in the activation mechanism (3). However, it was shown subsequently that a single, dualspecificity kinase catalysed phosphorylation of both residues, and was sufficient to activate MAP kinase in vitro (4). The activating kinase was named MAP kinase kinase, or MAP kinase/ERK kinase (MEK). Phosphorylation of MAP kinase by MEK occurs on single threonyl and tyrosyl residues separated by a glutamic acid. This TEY sequence is conserved in ERK MAP kinase homologues from yeast to man. More recent work has shown that JNK and p38 MAP kinases are activated in a very similar fashion. Each has the requirement for phosphorylation of a threonine and tyrosine residue, although the intervening residue is proline, in the case of JNKs, and glycine, in the case of p38 MAP kinases. In addition, each class has its own specific family of activating kinases. The tight correlation between MAP kinase phosphorylation and activity has been exploited in many of the methods used to assess MAP kinase activation status in living cells or tissues. Indeed, early work relied exclusively on the partial purification of phosphorylated MAP kinase from cells metabolically labelled with [32P]inorganic phosphate. Antibodies capable of recognising proteins phosphorylated on tyrosine residues (anti-PTyr antibodies) were also useful in allowing a preliminary assessment of MAP kinase activation (see Section 5.1). More recently, it has been possible to synthesize peptides containing the phosphorylated T-x-Y sequence, and to use these as immunogens to generate antibodies which recognize specifically the phosphorylated, active forms of MAP kinases (see Section 5.3).

3. MAP kinase substrates Early work on MAP kinases employed MAP-2 as a relatively specific in vitro substrate for assays in crude or partially purified cell extracts. However, purification of MAP-2 from bovine brain is a difficult and lengthy procedure. The discovery that another neural protein, myelin basic protein (MBP), was also an excellent MAP kinase substrate overcame this problem, since MBP is relatively easy to obtain and is available commercially. Indeed, MBP was used to obtain the first information on the consensus primary amino acid sequence surrounding the serine or threonine residue required for phosphorylation by MAP kinase (5). Such work showed that the threonine residue predominantly phosphorylated by MAP kinase was followed on its carboxyl side by three 339

Neil G. Anderson Table 2. Examples of known physiological substrates of MAP kinases Substrate type

Examples

Probable specificity in vivo

Protein kinases

MAPKAP kinase 1 (p90rsk) MAPKAP kinase 2 MAPKAP kinase 3 Mnk 1/Mnk 2

ERKs p38 p38 p38, ERKs

Transcription factors

c-Jun TCF P62 ATF2 Max CHOP PPARy STAT3

JNKs ERKs p38, JNKs p38 p38 ERKs? ERKs

Others

PhospholipaseA2 Oestrogen receptor SOS Myelin basic protein

ERKs, p38 ERKs ERKs, JNKs ERKs

prolyl residues. Subsequent studies on both in vitro and in vivo substrates for all classes of MAP kinase has revealed a preference for a prolyl residue immediately carboxy-terminal to the target serine or threonine. MAP kinases are thus members of the class of kinases sometimes referred to as prolinedirected protein kinases, which also includes cyclin-dependent kinases. Most MAP kinase assays now employ MBP or short peptides mimicking a known phosphorylation site derived from an in vivo substrate. Examples of well characterized MAP kinase substrates are listed in Table 2.

4. Testing for activation of known MAP kinases 4.1 Cultured cells Activation of MAP kinases occurs in most cell types in response to stimulation of the cells with a wide variety of different agents. Activation of ERK MAP kinases by mitogens occurs during the re-entry of quiescent cells to the cell cycle. Generally, ERK activity is very low in GO arrested cells, but would be easily detectable in asynchronous growing cultures. To determine whether a particular agent causes ERK activation, it is usually necessary to switch cells from their normal growing medium to one in which they will enter G0. Activation of ERKs is generally rapid, and is detectable within 2-5 min of adding the extracellular stimulus. The activation may be transient, peaking at around 5-10 min before declining to basal levels, or more sustained, with activity remaining above basal levels for several hours. For initial studies, a time course from 2-60 min should be sufficient to determine whether a particular MAP kinase is activated under the test conditions. The main practical 340

14: Detection, isolation and quantitative assay of mitogen activated consideration is preserving the activation of the enzyme between preparation of cell extract and assay. Loss of MAP kinase activation in cell lysates, due primarily to phosphatase action, is rapid. But this can be prevented by maintaining cell lysates at 4°C, and adding inhibitors of both proteases and phosphatases to the lysis buffer. Protocol 1 details the procedures for preparation of lysates from growth factor-stimulated adherent cells, such as 3T3 fibroblasts. Protocol 1. Preparation of extracts from growth factor-stimulated adherent cells Reagents • Lysis buffer: 25 mM Hepes, pH 7.5 (4°C); 5 mM EDTA; 50 mM sodium chloride; 50 mM sodium fluoride; 30 mM sodium pyrophosphate; 10% glycerol; 1% NP-40. Store at 4°C. Stable for 1-2 weeks. Immediately prior to the experiment, add the following cocktail of inhibitors (all obtained from Sigma), all of which can be stored in aliquots at -20°C for several months: • Sodium orthovanadatea: stock solution, 100 mM (in water); final concentration, 1 mM

• Phenylmethylsulfonyl fluoride: stock solution, 200 mM (in isopropanol), final concentration, 1 mM • Leupeptin: stock solution, 2 mg ml-1 (in water); final concentration, 2 ug ml-1 • Aprotinin: stock solution, 2 mg ml-1 (in water); final concentration, 2 ug ml-1 • Pepstatin A: stock solution, 1 mg ml-1 (in ethanol); final concentration, 2 ug ml-1

Method 1. Switch confluent cell cultures to low serum or serum-free medium for 12-24 h. The use of Hepes (25 mM final concentration) in this medium is recommended, to help maintain physiological pH during subsequent manipulation of the cultures. 2. Add the growth factor in a small volume (less than 1% of total medium volume). 3. Return cells to the incubator. 4. At the end of the treatment, place the dishes of cells on ice, and immediately remove the medium by aspiration. 5. Carefully dispense an appropriate volume of ice-cold phosphatebuffered saline (PBS) onto the cell monolayer (e.g. use 10 ml for cells grown on a 100 mm plate). This and all subsequent steps should be conducted on ice or in a cold room (4°C) 6. Remove the PBS, and repeat the wash once more. 7. Tilt the dishes so that all of the remaining PBS can be removed. 8. Add lysis buffer (at 4°C) to the monolayer. Use 1.0 ml for a 100 mm dish of cells. 9. Gently scrape the cells into the lysis buffer, using a plastic stirring rod (Nalgene). 341

Neil G. Anderson Protocol 1. Continued 10. Transfer the cell suspension to microcentrifuge tubes, and rotate them on a wheel for 20 min. 11. Centrifuge the extracts at 15000 g for 10 min at 4°C. 12. Transfer the resulting supernatant solution to fresh tubes. 13. Retain an aliquot of the extract for assay of total protein. The remaining extract can then be treated as follows: (a) snap-freeze in liquid nitrogen for assay at a later date. (b) boil in the presence of SDS-sample (Laemmli) buffer for gel analysis (see Protocol 2). (c) use immediately for further purification, immunoprecipitation, or direct assay (see below). a

To prepare sodium orthovanadate, follow the method detailed in ref. 6.

4.2 Other sources As illustrated above, extraction of active MAP kinases from adherent monolayers of cultured cells is relatively straightforward. The ability to chill the cells rapidly and lyse them in situ minimizes the degree of post-lysis inactivation of the enzyme. For cells grown in suspension culture or for tissue sources, the additional processing steps increase the possibility of losing all or part of the original activation. However, provided that the cells or tissue can be rapidly chilled after stimulation, there should be few practical difficulties in measuring MAP kinase activation. 4.2.1 Suspension cells Procedures are generally identical to those given in Protocol 1, with the following exceptions. After the period of stimulation with growth factor or other agent, the cells should be rapidly sedimented at room temperature. After removing the medium, the cell pellet should then be washed once in icecold PBS, and then solubilized in lysis buffer. Thereafter, proceed as in Protocol 1.

4.2.2 Tissues Methods are available for extracting protein kinases, including MAP kinases, from whole tissues such as skeletal muscle (7) although, for obvious practical reasons, detailed kinetic studies are not feasible. It is possible, for example, to measure the acute activation of ERK MAP kinases in rabbit skeletal muscle following administration of an intravenous dose of insulin to the animal. More commonly, the investigator wishes to measure the level of activation of MAP 342

14: Detection, isolation and quantitative assay of mitogen activated kinase in a series of clinical tissue biopsies, to compare enzyme activity between, for example, tumour samples and their normal counterparts. It is not possible to provide a method suitable for extraction of active enzyme from all types of tissue. However, the same principles apply: the tissue must be rapidly frozen in liquid nitrogen immediately after removal from the donor. Thereafter, enzyme activities should be relatively stable if the tissue is stored at or below — 70°C. For analysis, the frozen material should be ground to a fine powder in the presence of liquid nitrogen, and then homogenized at 4°C in a buffer containing protease and phosphatase inhibitors (see Protocol 1). Following centrifugation (15 000 g, 10 min) to remove insoluble material, the supernatant solution can be used for further purification or for immunoprecipitation.

5. Assessment of MAP kinase activity by gel electrophoresis and immunoblotting In general, the procedures outlined in this section are useful as an indicator of MAP kinase activation, and are not suitable for accurate quantification of enzyme activity. They are therefore extremely useful for initial characterization of the involvement of a MAP kinase in a particular process, or for determining whether a known MAP kinase is activated by a given cellular stimulus.

5.1 Detection of phosphotyrosine in whole cell extracts Since all species of MAP kinase are phosphorylated on tyrosyl residues when active, detection of a phosphotyrosyl protein of the appropriate molecular mass by immunoblotting with anti-phosphotyrosine antibodies has been used extensively as an indicator that the MAP kinase in question is activated. The obvious advantage of this method is its simplicity. Unpurified cell extracts can be denatured, electrophoresed on SDS-PAGE, and finally immunoblotted with a widely available class of antibody. Another advantage is that MAP kinase activation can be assessed in relation to the tyrosyl phosphorylation of other proteins (e.g. receptor tyrosine kinases and their substrates) on the same immunoblot (Figure 1). It should be remembered, however, that not all commercially available anti-phosphotyrosine antibodies recognize the phosphotyrosyl residue present in active MAP kinases. In this respect, the monoclonal antibody (clone 4G10, Upstate Biotech.) strongly recognizes activated ERKs, whereas another monoclonal, PY20 (Transduction Labs), reacts with the protein only weakly on immunoblots. In any case, caution should be exercised in assigning phosphotyrosyl proteins of the appropriate size on gels as MAP kinases without corroborating evidence. Showing that the phosphotyrosyl protein co-localizes with the MAP kinase of interest by re-probing the 343

Neil G. Anderson

Figure 1. An example of an anti-phosphotyrosine Western blot of lysates from 3T3-F442A fibrotalasts stimulated with various agents. Cells were starved in serum-free medium for 16 h prior to addition of agents for 10 min. Lane 1: no addition; lane 2: growth hormone (10 nM); lane 3: 12-tetradecanoylphorboi-13-acetate (10 nM); lane 4: epidermal growth factor (10 nM); lane 5: insulin-like growth factor (10 nM); lane 6: insulin (10 nM). The apparent molecular masses of proteins undergoing an increase in phosphotyrosyl content following stimulation of the cells are indicated to the right of the blot. The p44 and p42 phosphotyrosyl proteins were tentatively identified as ERK-1 and ERK-2 respectively.

immunoblol with an antibody to the MAP kinase would provide supporting, although not definitive, evidence. This procedure is based on the premise that activated MAP kinases arc tyrosyl phosphorylated. Whilst this is certainly true, the converse assumption may not always be. In other words, MAP kinases phosphorylated on lyrosine hut not on threonine will be inactive (3). Therefore, detection of tyrosyl phosphorylated MAP kinase does not necessarily indicate that the MAP kinase is active. Indeed, singly phosphorylatcd species of ERK.2 have been detected in growth factor-stimulated cells (8). It is thus inadvisable to use this technique as the sole indicator of MAP kinase activation until the system in question has been thoroughly characterized.

5.2 Mobility shifts The presence of phosphate on proteins, especially seryl- or threonyl phosphate, often slightly retards their relative mobility on SDS-PAGE. This phenomenon permits separation of the inactive and active forms of MAP kinase on gels under specified conditions, as outlined in Protocol 2. 344

14: Detection, isolation and quantitative assay of mitogen activated Protocol 2. Separation of dephospho- (inactive) and phospho(active) MAP kinases on polyacrylamide gels, and immunodetection Equipment and reagents • Large format vertical gel electrophoresis system (e.g. Protean II, BioRad Laboratories) • Compatible system for Western transfer (e.g. Transblot, Pharmacia) . Acrylamide:bisacrylamide solution A (30% (w/v) acrvlamide:0.32% (w/v) bisacrylamide) • Acrylamide:bisacrylamide solution B (30%:0.8%) . 1.5 M Tris-HCI, pH 8.8

• 0.5 M Tris-HCI, pH 6.7 . 10% SDS /v,/v,/v',/v'-tetrametriylethylenediamine (TEMED) . 10% ammonium persulfate (prepare fresh) .PonceauS(0.5%,prepared in 1% acetic acid) . Tween Tris-buffered saline (TTBS): 10 mM Tris-HCI, pH 7.4; 0.15 M NaCI; 0.1% Tween-20

Method 1. Prepare 40 ml of separating gel solution from the above stock solutions, to give 0.375 M Tris-HCI, pH 8.8, 12% acrylamide (from solution A), and 0.1% SDS. 2. De-gas the solution, then add ammonium persulfate (0.05% final concentration) and TEMED (0.0375% final concentration). 3. Pour the gel, using 12 x 16 cm plates and 1.5 mm spacers, and overlay with water-saturated butanol. 4. Allow to set for at least 2 h. 5. Remove and rinse away the butanol layer, then add the stacking gel (0.125 M Tris-HCI, pH 6.7; 4% acrylamide (from solution B); 0.1% SDS; 0.1% ammonium persulfate; 0.05% TEMED) with a multi-well comb in place. 6. Allow to set for 30-60 min. 7. Remove the comb, rinse the wells with tank buffer, and apply samples equalized for protein content (10-100 ug per sample). 8. Run the gel at a constant current setting of 30 mA until the dye front reaches the bottom of the plates (about 4 h). 9. Transfer the proteins to nitrocellulose (Hybond C [Amersham] or similar) for 2 h at 400 mA. 10. Stain the nitrocellulose with Ponceau S, to confirm transfer and equal protein loading. 11. Block with 3% BSA solution for 3-24 h. 12. Rinse with TTBS for 15 min. 13. Probe with MAP kinase antibody for 1-3 h. 14. Rinse with TTBS. 345

Neil G. Anderson Protoco 1. Continued 15. Probe with horseradish peroxidase-conjugated secondary antibody for 1 h. 16. Rinse and apply agents for detection of chemiluminescence (e.g. Supersignal, Pierce-Warriner).

There are now available commercially a wide range of both monoclonal and polyclonal antibodies which are suitable for Western blotting and detection of mobility shifts (see Table 3). An example of the technique is illustrated in Figure 2. The advantage of this method is that it gives an indication of the proportion of the total cellular MAP kinase protein which is in the active state. Note that some cell stimuli, such as platelet-derived growth factor, cause all of the immunodetectable ERK2 to shift to the slower migrating (active) form, indicating that this growth factor induces activation of close to 100% of the cellular MAP kinase protein (Figure 2). Other stimuli are less potent, inducing only a proportion of the total ERK protein to shift. This implies that the relative amount of retarded protein should correlate with the MAP kinase catalytic activity originally present in the sample. Thus by estimating the ratio of protein in the active form (upper band) to the total (upper plus lower bands), relative activation status can be determined. Band densities may be estimated from an autoradiogram by densitometry, or by direct measurement of chemiluminescence. Due to the inherent non-linear characteristics of X-ray film, the latter is preferable. Although this method Table 3. Some of the commercial antibodies suitable for MAP kinase studies Enzyme recognized

Company

Catalogue No.

Techniques

ERK1 ERK1 ERK2 ERK2 ERK2 ERK2 JNK1 JNK3 p38(pan) p38(SAPK2a) phospho-ERK1, 2 phospho-JNK1, 2 phospho-p38

Santa Cruz Biotech. Upstate Biotech. Santa Cruz Biotech. Zymed Labs. Inc. Transduction Labs. Upstate Biotech. Santa Cruz Biotech. Upstate Biotech. Santa Cruz Biotech. Upstate Biotech. New England Bio. New England Bio. New England Bio.

sc-93 06-182 sc-154 Z033 E16220 06-333 sc-1648 06-425 sc-535 06-620 9101 9251 9211

WB, IPK WB WB, IPK, 1C WB WB, 1C WB, IPK WB, 1C WB, IPK WB, IPK, lC WB WB WB WB

This is a list of some of the antibodies suitable for the techniques outlined in this Chapter. It is not intended to be a recommended list, only one covering antibodies with which the author has had some experience. Many other suitable antibodies are available from several other companies. Abbreviations: WB, Western blotting; IPK, kinase assay after immunoprecipitation; lC, immunocytochemistry.

346

14: Detection, isolation and quantitative assay of mitogen activated

Figure 2. Detection of MAP kinase activation by mobility shift. Rat-1 fibroblasts were stimulated with the indicated agents for 5 min, and cell lysates prepared as described in Protocol 1. The lysates were then denatured, and run on a 12% low bis-acrylamide gel, prior to imrnunoblotting with a monoclonal ERK-2 antibody as described in Protocol 2. The positions of the non-phosphorylated (inactive) and phosphorylated (active) ERK-2 are indicated.

Figure3. Comparison of the sensitivities of the mobility shift method and direct enzymic assay for detection of MAP kinase activation. 3T3-F442A fibroblasts were left untreated (C), or stimulated with 10 nM insulin (Ins), or 10 nM growth hormone (GH), for 10 min. Cells were lysed, and the lysates were then divided and used for assessment of MAP kinase mobility shift according to Protocol 2 (top panel), or ERK-2 kinase activity by immunoprecipitation according to Protocol 7 [bottom panel).

can be extended to obtain data on relative MAP kinase activation, it may not be sensitive enough to detect subtle changes in MAP kinase activation (Figure 3). It also makes the assumption that mobiiity shift always indicates activation. Whilst this is almost certainly always true, singly phosphorylated MAP kinase may also exhibit retarded mobility on SDS-PAGE, and the technique should not be the sole indicator of MAP kinase activation without corroborative evidence from direct enzyme assays. 347

Neil G.

5.3 Phospho-specific antibodies Recent advances in the chemistry and technology associated with the synthesis of peptides has enabled routine, cost-effective production of peptides phosphorylated on one or more amino acid residues. These peptides are now being used extensively to produce antibodies capable of distinguishing proteins, phosphorylated on specified residues, from their non-phosphoryiated counterparts. Given the universal role of phosphorylation in regulating enzymc activity, these 'phospho-specific' antibodies represent a valuable new tool for detecting activation of enzymes in cell or tissue extracts by immunoblotting. In addition, the use of these antibodies in immunocytochemistry applications should permit simultaneous analysis of an enzyme's cellular or subcellular localization in conjunction with its activation status (see Section 7 and Protocol 8). Using peptides modelled on the amino acid sequence of MAP kinases containing the phosphorylatable threonyl and tyrosyl residues, a number of phospho-specific MAP kinase antibodies have been produced, some of which are available commercially. An example of a Western blot using an anti-active ERK antibody (New England Bio-Labs) is shown in Figure 4,

Figure 4. Use of phosphorylation state-specific antibodies to detect ERK activation. Rat-1 fibroblasts were pretreated or not with 1 mM dibutyryl cyclic AMP for 10 min prior to the addition of nothing (C), platelet-derived growth factor (10 ng ml-1; P), or [D-Ala, D-Leu]enkephalin (1 uM; D) for 5 min. Cell lysates were then prepared and immunoblotted with phospho-specific ERK antibodies (New England Bio-Labsl as described in Section 5.2, The positions of ERK-1 and ERK-2 are indicated.

Protocol 3. Purification and characterization of phospho-specific ERK antibodies from immunized rabbits Equipment and reagents • Protein A-agarose (Pierce-Warrinerl • Disposable mini columns (BioRad) • AffiGel-10 (BioRad) .

. MAP kinase peptides [phosphorylated and non-phosphoryfated) Column buffer: 100 mM Tris-HCI, pH 8.0

A. Preparation of affinity columns 1. Couple 25 mg of each peptide to two separate batches of AffiGel-10 matrix (1 ml packed volume), according to the manufacturer's instructions. 348

14: Detection, isolation and quantitative assay of mitogen activated 2. Pour into 5 ml disposable columns, and store at 4°C in Column buffer containing 0.02% sodium azide. B. Immunoglobulin purification 1. Prepare a column containing 2 ml Protein A-agarose 2. Apply 5 ml of antiserum to the column at a flow rate of 1 ml min-1 3. Wash the column extensively with Column buffer (at least ten volumes), until A280 of the eluted material is zero. 4. Elute IgG with 100 mM glycine (pH 3.0), collecting 0.9 ml fractions into 0.1 ml of 1 M Tris-HCI, pH 8.0. 5. Measure the A280, and pool the fractions containing protein (OD > 0.1). C. Isolation of specific antibodies 1. Apply pooled IgG fractions to the column (column 1) containing the non-phosphorylated peptide affinity matrix, and allow the unbound material to flow directly on to a tandemly-linked second column (column 2) containing the phosphorylated peptide affinity matrix. 2. Detach column 1. 3. Wash column 2 extensively with Column buffer. 4. Elute the affinity purified antibody as in B., Step 4. 5. Pool the fractions containing protein, and concentrate using a Centricon filter (Amicon).

6. Measurement of MAP kinase enzyme activity Whilst in many cases it may be sufficient to use one of the procedures outlined in Section 5 to assess relative activation of MAP kinases, often it is desirable to obtain a more precise measurement of the activity of the enzyme. Direct enzyme assays are also appropriate when attempting to chromatograph several MAP kinase activities from a single extract. MAP kinase assays are generally performed either directly in partially purified cell or tissue extracts, or following immunoprecipitation with an antibody which preserves the enzyme in its activated state. Although each class of MAP kinase possesses its own set of known physiological substrates, there is considerable overlap, and in vitro most of the commonly used protein or peptide substrates are phosphorylated by all types of MAP kinase. The presence of a prolyl residue on the carboxy terminal side of the target serine or threonine often appears to be sufficient to render the peptide a MAP kinase substrate in vitro. For assays of MAP kinase following immunoprecipitation, MBP is relatively inexpensive and easily obtained. On the other hand, for assaying MAP kinase in an 349

Neil G. Anderson unpurifled or partially purified extract, it would be preferable to use an appropriate peptide substrate to reduce phosphorylation of a protein substrate by other protein kinases.

6.1 Assay of MAP kinase in partially purified cell extracts The number of protein kinases present in a typical cell extract makes it inappropriate to assay MAP kinase directly in unpurified extracts. A simple procedure is available for partial purification of ERK MAP kinases from cell or tissue extracts by batch adsorption to phenyl-Sepharose (9). This is outlined in Protocol 4. In early studies on MAP kinases, it was realized that ERKs bind tightly to this matrix, only eluting with relatively high concentrations of ethylene glycol. Consequently, phenyl-Sepharose chromatography represents a simple but powerful method for rapid removal of > 90% of cellular protein from an extract in a single step. The final eluate contains ERK MAP kinases free from the majority of other contaminating kinases. The hydrophobic nature of the interaction between MAP kinase and the phenyl group means that detergent-solubilized cell or tissue extracts cannot be used, however. Protocol 4. Partial purification of extracts by batch adsorption to phenyl-Sepharose Equipment and reagents • Cell lysis buffer: 25 mM Tris-HCI, pH 7.5 (4°C); 40 mM 4-nitrophenyl phosphate; 25 mM NaCI; 10 uM dithiothreitol. Store at -20°C in the dark. Prior to use add the cocktail of protease inhibitors and sodium orthovanadate, exactly as described in Protocol 1.

• Phenyl-Sepharose (Pharmacia) . Ethylene glycol .Refrigeratedmicrocentrifuge

Method 1. Dispense 0.2 ml aliquots (packed volume) of phenyl-Sepharose into the required number of microcentrifuge tubes on ice. Wash twice with lysis buffer (0.5 ml each wash). 2. Lyse tissue or cells by sonication or homogenization in lysis buffer 3. Centrifuge homogenate at 20000 g max (4°C) to remove unbroken cells. 4. Measure the protein concentration of the supernatant, and apply 0.2-0.5 mg of protein to 0.2 ml (packed volume) of phenyl-Sepharose in a microcentrifuge tube. 5. Gently invert the tube to allow the cell extract to mix with the matrix. 6. After 5 min spin tubes at 10 000 g for 10 s.

350

14: Detection, isolation and quantitative assay of mitogen activated 7. Using a glass Pasteur pipette linked to an aspirator, carefully remove the unbound material. 8. Replace with 0.5 ml of lysis buffer, mix gently, and re-centrifuge. 9. Repeat Step 7, then wash twice with lysis buffer containing 35% ethylene glycol. 10. Add 0.2 ml lysis buffer containing 60% ethylene glycol. 11. Gently mix, and place on ice for 5 min. 12. Centrifuge as in Step 5. 13. Remove the 60% eluate, and retain for assay.

To obtain information on which MAP kinases are activated in a particular cell type in response to a novel ligand or other stimulus, a useful initial experiment is to conduct a broad chromatographic analysis of the cell extract to separate the different MAP kinase activities. Anion exchange chromatography using diethylaminoethyl (DEAE)-cellulose or DEAE-Sepharose can be used. However, the greater resolution and speed offered by chromatography on matrices such as Mono-Q using an FPLC system (Pharmacia) make such methods preferable. Comparing the profiles of enzyme activities from control and treated cells gives an overall picture of the MAP kinases activated (Figure 5). To determine which of the peaks of activity contain known MAP kinases, the fractions can be immunoblotted in parallel with the appropriate antibodies (see Protocol 5). The initial analysis over a wide salt gradient can subsequently be refined to detect 'peaks within peaks', using shallower NaCl gradients in the desired range. The assays are conducted in the presence of a range of protein kinase inhibitors (see Protocol 6), to minimize background phosphorylation of substrate. This is especially important if myelin basic protein is used as substrate. Protocol 5. Anion exchange chromatography of cell extracts using Mono-Q columns Equipment and reagents • FPLC system (Pharmacia) • Mono Q HR 5/5 FPLC column (Pharmacia) • Cell lysis buffer (see Protocol 7)

• 100% trichloroacetic acid • 0.25% sodium deoxycholate

A. Fractionation of extract 1. Prepare lysates from untreated and stimulated cells (1-5 x 107) according to Protocol 1. 2. Apply the cell lysate to a Mono Q column previously equilibrated in equilibration buffer (lysis buffer without Triton X-100 and protease inhibitors).

351

Neil G. Anderson Protocol 5. Continued 3. Wash the column with equilibration buffer until A280 is zero. 4. Elute the bound material with a 0-0.8 M NaCI linear gradient, collecting forty 1 ml fractions on ice. 5. Assay each fraction immediately according to Protocol 6. B. Precipitation of fractions for immunoblotting 1. Precipitate the remaining protein in each fraction by adding the appropriate volume of trichloracetic acid and sodium deoxycholate (to give final concentrations of 5% and 0.025% respectively). 2. Leave on ice for 10 min. 3. Centrifuge at 12000 g for 10 min. 4. Remove the supernatant, and resuspend the pellet in SDS-sample buffer. 5. Boil for 5 min, and then store the samples at -20°C.

Figure 5. Use of anion exchange chromatography to separate MAP kinase activities from cell extracts. Lysates were prepared, according to Protocol 4, from untreated (open circles) and growth hormone-treated (filled circles) 3T3-F442A fibroblasts. The lysates were then applied to a 1 ml EconoPac Q column (BioRad laboratories). After washing, the column was chromatographed using a linear sodium chloride gradient (0-0.4 M), indicated by the dashed line. Fractions were collected and assayed for MAP kinase activity, using myelin basic protein as substrate.

352

14: Detection, isolation and quantitative assay of mitogen activated Protocol 6. MAP kinase assay Reagents • MBP peptide [APRTPGGRR] (Upstate Biotech.): prepare stock solution at 0.8 mg ml-1 in 50% glycerol, store at -20oC • [y-32P]-ATP (specific activity > 3000 Ci mmor-1) • P81 phosphocellulose paper (Whatman), cut into 2 x 2 cm squares . Phosphoric acid (180 mM) • ATP stock solution (10 mM)a

• MgCI2 stock solution (1 M) . Kinase assay buffer: 10 mM MOPS, pH 7.2 (30°C); 10 mM p-glycerophosphate; 2 mM EGTA; 25 uM dithiothreitol; 0.1 mM sodium orthovanadate; 2 uM protein kinase inhibitor (PKI) peptide (Sigma); 10 uM calmidazolium (Calbiochem); 2 uM PKC inhibitor peptide 19-31 (Calbiochem)

Method 1. Prepare 8x concentration of ATP-Mg mix containing 400 uM ATP, 80 mM MgCI2, and [y-32P]ATP (0.5 uCi ul-1). 2. On ice add 5 ul sample to 5 ul substrate peptide and 25 ul kinase buffer. 3. Pre-incubate tubes for 2 min at 30°C, then add 5 ul ATP mix. Stagger the starts at 20 s intervals. 4. Briefly vortex, then return to the 30°C water bath. 5. Assay for the appropriate time (see Section 6.3). 6. Remove 30 ul of the reaction mix, and spot onto a P81 paper square numbered in pencil. 7. After about 10 s, drop the paper into a beaker containing 180 mM phosphoric acid (use —20 ml per assay). 8. Once all assay papers are in phosphoric acid, stir for 5 min. 9. Decant the phosphoric acid, and do four further washes of the papers (5 min each wash). 10. Finally, rinse the papers in ethanol, and place on absorbent paper to dry. 11. Place P81 papers in scintillation fluid, and count the radioactivity in a scintillation counter. * Dissolve ATP (disodium salt) in water to give approximately 10 mM. Adjust pH to 7. Measure absorbance at 260 nm. Calculate the exact concentration of ATP using e260 = 15 400.

6.2 Assay of MAP kinases following immunoprecipitation A number of antibodies are now available commercially which allow immunoprecipitation of several of the MAP kinases in an active form which can be assayed on the immunocomplex. The major advantage of this method is its ability to purify, rapidly and selectively, the MAP kinase of interest from 353

Neil G. Anderson a crude cell extract, to enable kinase activity to be determined in the absence of other protein kinases. However, it is important to consider potential limitations of this method, which could result in underestimation of the amount of kinase activity present in the original cell extract. Firstly, when complexed with an antibody, the kinase may lose intrinsic activity. This possibility can be examined by measuring the activity of a known amount of the purified enzyme in the presence or absence of an immunoprecipitating antibody. Secondly, co-immunoprecipitation of other kinases cannot usually be ruled out. The use of narrow specificity MAP kinase substrates for the assay will minimize the contribution of contaminating kinases to the activity measured. Finally, for quantitative measurements, it is necessary to determine empirically the quantity of antibody required to precipitate the entire MAP kinase cellular protein from a given cell extract. The efficiency of the antibody can be determined by immunoblotting the cell extract for the relevant MAP kinase before and after immunoprecipitation. Protocol 7. Assay of ERK MAP kinase by immune complex kinase assay Equipment and reagents • • . •

Rotating wheel for microcentrifuge tubes Microcentrifuge Aspirator Protein A-agarose or protein G-agarose (depending on the class of immunoprecipitating antibody used)

• Cell lysis buffer (see Protocol 7) • High salt buffer (10 mM Tris-HCI, pH 8.0; 0.5 M LiCI) • Kinase buffer (see Protocol 6); the inhibitors PKI, calmidazolium, and PKC inhibitor peptide may be omitted

Method 1. Prepare cell or tissue lysates (see Protocol 1). 2. Pre-clear lysates by adding 20 ul (packed volume) protein A- or protein G-agarose. 3. Rotate on the wheel for 1 h. Meanwhile conduct a protein assay on an aliquot of each sample. 4. Microcentrifuge for 10 s, and transfer the supernatant to a fresh tube. 5. To equal quantities of cellular protein, add an appropriate amount (previously determined) of precipitating antibody. 6. Incubate from 2 h to overnight at 4°C. 7. Transfer samples to fresh tubes containing 20 ul (packed volume) of protein A- or protein G-agarose. 8. Rotate on the wheel for 1 h. 9. Wash immunoprecipitates by centrifugation and careful aspiration of the resulting supernatant with 0.5 ml lysis buffer (twice), then with 0.5 ml high salt buffer. 354

14: Detection, isolation and quantitative assay of mitogen activated 10. Finally wash immunoprecipitates with 0.5 ml kinase buffer. 11. Remove all remaining buffer from the immunoprecipitates, using a fine pipette. 12. Resuspend beads in 25 ul kinase buffer plus 5 ul MBP peptide or protein.a 13. Preincubate at 30°C for 2 min. 14. Add 5 ul [32P]ATP-Mg mix as in Protocol 6. 15. Incubate for an appropriate time, then centrifuge reaction tubes. 16. Carefully remove 25 ul of reaction mixture, and spot onto P81 paper. 17. Proceed as in Protocol 6. a

If MBP protein is to be used as substrate for the kinase assay, terminate the reaction by adding 10 ul of 4X concentrated Laemmli buffer. Boil the samples for 2 min, then centrifuge the tubes for 5 min in a microcentrifuge. Apply 20 ul of the supernatant to a 13% mini-gel. After running the gel, stain with Coomassie Blue to locate the MBP band. Excise the bands and count the associated radioactivity.

6.3 Assay conditions Whichever method is used to assay MAP kinase activity, it is essential to establish optimal standardized assay conditions to allow valid comparisons to be made between experiments. MAP kinase phosphotransferase activities are usually quoted as picomoles of phosphate incorporated into substrate per minute under standard assay conditions. It should first be established that the measured incorporation of 32P into substrate is linear with respect to the assay time. Non-linearity will result in under- or overestimation of true phosphotransferase activity. Pre-incubation of the reaction mixture at 30°C, prior to addition of ATP, is essential to eliminate the lag in reaction rate as the mixture reaches 30°C. Depletion of substrate (peptide or ATP) during the course of the assay can lead to underestimation of enzyme activity. This is more likely to be a problem when conducting assays in unpurified extracts. The concentration of ATP should be at least 50 uM, and that of substrate at least 25 uM. To detect enzyme activity with reasonable sensitivity, the use of a specific radioactivity of at least 2500 c.p.m. per picomole of ATP is suggested. This entails the use of 2.25 uCi of radioactivity per assay under the conditions quoted in Protocol 6. When performing several assays under these conditions, the accompanying radiation hazard is a serious consideration. Assays of purified or immunoprecipitated kinases could, almost certainly, be conducted satisfactorily at lower specific activities. Finally, when conducting assays in unpurified cell extracts, the potential influence of endogenous diffusible inhibitors should be investigated. This is easily accomplished by checking the linearity of the relationship between amount of sample and phosphotransferase activity measured. If inhibitors are present, this relationship breaks down at high concentrations of sample. 355

Neil G. Anderson

6.4 Other MAP kinase assays 6.4.1 Assay of JNK and p38 MAP kinases For assay of specific MAP kinase family members, several immunoprecipitating antibodies are now available. Following the general procedure given in Protocol 7 allows determination of the activities of these kinases in cell or tissue extracts. As stated previously, all MAP kinases will phosphorylate myelin basic protein or peptides encompassing the relevant phosphorylated sites in vitro. Obviously there is a much greater degree of specificity in vivo, and this can be used for the more selective assay of individual family members. For example, although ERK MAP kinases can phosphorylate c-jun in vitro, it is now well established that JNKs are primarily responsible for c-jun phosphorylation in intact cells. Based on this, an assay for JNKs has been established, involving the use of recombinant c-jun fragments fused to glutathione S-transferase to precipitate JNKs from cell extracts (10). A solid state assay is then conducted on glutathione-Sepharose beads, and the phosphorylated c-jun resolved by electrophoresis. Immune-complex assays for p38 MAP kinases have employed a fragment of the transcription factor ATF-2 as a relatively selective substrate for this class (11). 6.4.2 Non-radioactive MAP kinase assay procedures A number of novel non-radioactive methods for detection of MAP kinase activities have recently been developed. These procedures generally rely on the specific detection of a phosphorylated peptide assay product, either immunochemically or colorimetrically. For example, antibodies which recognize the phosphorylated residue in myelin basic protein or c-jun have been developed (now available from Upstate Biotech.) which can be used to detect the reaction product by Western blotting or ELISA. In this way, the MAP kinase of interest can be immunoprecipitated and assayed using the method given in Protocol 7. The reaction mixture can then be analysed by gel electrophoresis, and Western blotting using the phospho-specific antibody. As discussed above, quantification of the immunoreactive phospho-protein is the potential limitation of such a procedure. The high cost of such a procedure is also likely to be a factor for most investigators. However, the linkage of this principle to an ELISA detection method is clearly worth considering for high throughput analysis of multiple samples.

7. Determination of MAP kinase subcellular localization and activation by immunocytochemistry Downstream biological consequences of MAP kinase activation are often a result of translocation of the activated protein to another part of the cell 356

14: Detection, isolation and quantitative assay of mitogen activated where the relevant substrate resides. For example, it has been demonstrated that prolonged activation of ERK MAP kinases by some mitogens or differentiating agents can lead to their translocation to the nucleus. This is hardly surprising, given the fact that many of the true physiological substrates of MAP kinases are transcription factors, e.g. c-jun, p62TCF, and ATF2 (Table 2). However, nuclear translocation does not always occur, and determining the movement of a particular MAP kinase in response to a novel ligand may be of fundamental importance in describing the biological consequences of MAP kinase activation. Many of the antibodies now available are sufficiently specific to allow immunostaining of fixed cells or frozen tissue sections. In addition, the recent development of phospho-specific MAP kinase antibodies should allow the selective localization of the active forms of the protein. A generalized procedure for immunocytochemistry of ERK MAP kinases in fibroblasts is given in Protocol 8. However, it is likely that these procedures would require some optimization for other cell types and for frozen sections. Protocol 8. Detection of ERK MAP kinase by immunocytochemistry Equipment and reagents • Glass coverslips • Fluorescence microscope • 3.7% formaldehyde solution (prepared in PBS containing 0.2% Triton X-100)

• Blocking buffer (1% BSA in PBS) • Mounting medium (50% glycerol, 0.1% pphenylenediamine dihydrochloride, in PBS) « Transparent nail polish

Method 1. Grow cells on glass coverslips in 12-well cell-culture plates. 2. Treat cells with growth factor or other agent. 3. Rinse cells twice with ice-cold PBS. 4. Drain off PBS, and add 0.5 ml formaldehyde solution. 5. After 10 min, drain off formaldehyde solution, and air-dry coverslips. 6. Place coverslips on filter paper in small Petri dishes. 7. Apply sufficient blocking buffer (~50ul) to cover cells, and leave for 30 min. 8. Drain off blocking buffer onto filter paper. 9. Apply anti-ERK antibody (1:50 dilution), and leave for 1h. 10. Remove antibody solution, and rinse coverslips three times with PBS. 11. Apply an appropriate dilution of FITC-conjugated secondary antibody for 1 h. 12. Remove the antibody solution, and rinse three times with PBS. 357

Neil G. Anderson Protocol 8.

Continued

13. Invert the coverslip onto a microscope slide bearing 10 ul mounting medium. 14. Seal the edges of the coverslip with nail polish. 15. View under a fluorescence microscope.

8. Assay of upstream activators of MAP kinases All MAP kinases are activated by at least one class of MAP kinase kinases (MAPKK), which in turn are activated by MAP kinase kinase kinases (MAPKKKs). For example, activation of ERKs involves the activation of Raf kinase (the MAPKKK), which then activates MEK (the MAPKK), which then activates ERK. However, for each of the three major classes of MAP kinase, several kinases are known to exist at each level of the three-member signalling module. Thus the two ERKs can be activated by two different MAPKKs (MEK 1 and MEK 2), which in turn can be activated by three different Raf kinases (A-Raf, B-Raf, and Raf-1). The upstream regulation of the SAPKs is even more complex. For this reason, it is beyond the scope of this chapter to describe individual methods for assaying all of these enzymes. Most of these kinases can be assayed routinely using immunoprecipitating antibodies and procedures based on the ERK immunocomplex assay given in Protocol 7. The choice of substrate is critical, since the MAPKKKs and MAPKKs generally have a much narrower substrate specificity than do MAP kinases. For example, ERKs are the only known physiological substrate of MEK1 or MEK2. Consequently, to measure MEK catalytic activity, it is necessary to use ERK protein as substrate. Similarly, Raf kinase activity should be measured using MEK as substrate. It is now possible to generate recombinant versions of both ERK and MEK, and their use in these assays has been described in detail previously (12, 13).

9. Chemical inhibitors of MAP kinase pathways Analysis of the role of MAP kinases in various biological processes has been aided with the recent identification of two classes of chemical inhibitor. The Parke-Davis compound PD098059 shows a high degree of selectivity towards MEK1 and MEK2 (14). This compound inhibits the activation of MEK by Raf, and therefore the activation of ERKs and responses downstream of ERKs are abrogated. However, the non-competitive nature of the inhibition means that previously activated MEK is not affected by the inhibitor. In addition, the effectiveness of PD098059 in blocking ERK activation appears to be dependent on the level of Raf activation. In Swiss 3T3 fibroblasts, the 358

14: Detection, isolation and quantitative assay of mitogen activated inhibitor almost completely blocks ERK2 activation by insulin, which is a weak activator of Raf-1, but is only partially effective in blocking ERK2 activated by epidermal growth factor, which is a strong activator of Raf-1 (14). PD098059 generally shows maximal effectiveness in cultured cells at concentrations of 30-100 uM. The SmithKline-Beecham compounds SB202190 and SB203580 potently inhibit p38 MAP kinases without affecting ERKs, JNKs, or a wide range of other kinases. However, it has been shown that chronic inhibition of the p38 MAP kinases can lead to the activation of ERK kinases by unknown mechanisms (15). Such observations demonstrate that caution should be exercised in the use of MAP kinase inhibitors, especially when measuring long-term cell responses, such as proliferation or apoptosis. In addition, a recently identified novel isoform of p38, termed p38-8, was shown to be insensitive to SB203580 (16).

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

Ray, L. B., and Sturgill, T. W. (1987). Proc. Natl. Acad. Sci. USA., 84, 1502. Minden, A., and Karin, M. (1997). Biochim. Biophys. Acta, 1333, F85 Ferrell, J. E., Jr. (1996). Curr. Top. Dev. Biol., 33, 1. Anderson, N. G., Mailer, J. L., Tonks, N. K., and Sturgill, T. W. (1990). Nature, 343, 651. Nakielny, S., Cohen, P., Wu, J., and Sturgill, T. W. (1992). EMBO J., 11, 2123. Erickson, A. K., Payne, D. M., Martino, P. A., Rossomando, A. J., Shabanowitz, J., Weber, M.J., Hunt, D. F., and Sturgill, T. W. (1990). J. Biol. Chem., 265, 19728. Gordon, J. A. (1991). In Methods in enzymology (ed Hunter, T. and Sefton, B. M.). Vol. 201, p. 477. Academic Press, London. Anderson, N. G., Wolf, B. B., and Sturgill, T. W. (1991). Adv. Prot. Phosphatases, 6, 119. Anderson, N. G., Kilgour, E., and Sturgill, T. W. (1991). J. Biol. Chem., 266, 10131. Westwick, J. K., and Brenner, D. A. (1995). In Methods in enzymology (ed Balch, W. E., Der, C. J. and Hall, A.). Vol. 255, p. 342. Academic Press, London. Kummer, J. L., Rao, P. K., and Heidenreich, K. A. (1997). J. Biol. Chem., 272, 20490. Reuter, C. W. M., Catling, A. D., and Weber, M. J. (1995). In Methods in enzymology (ed Balch, W. E., Der, C. J. and Hall, A.). Vol. 255, p. 245. Academic Press, London. Alessi, D. R., Cohen, P., Ashworth, A., Cowley, S., Leevers, S. J., and Marshall, C. J. (1995). In Methods in enzymology (ed Balch, W. E., Der, C. J. and Hall, A.). Vol. 255, p. 279. Academic Press, London. Alessi, D. R., Cuenda, A., Cohen, P., Dudley, D. T., and Saltiel, A. R. (1995). /. Biol. Chem., 270, 27489. Jarvis, W. D., Fornari, Jr., F. A., Auer, K. L., Freemerman, A. J., Szabo, E., Birrer, M. J., Johnson, C. R., Barbour, S. E., Dent, P., and Grant, S. (1997). Mol. Pharmacol., 52, 935. Jiang, Y., Gram, H., Zhao, M., New, L., Gu, J., Feng, L., Di, P-F., Ulevitch, R. J., and Han, J. (1997). J. Biol. Chem., 272, 30122. 359

This page intentionally left blank

15 Measuring inositol 1,4,5-trisphosphate-evoked 45 Ca2+ release from intracellular Ca2+ stores COLIN W. TAYLOR and JONATHAN S. MARCHANT

1. Introduction The receptors for inositol 1,4,5-trisphosphate (Ins(l,4,5)P3) are intracellular Ca2+ channels which open in response to the concerted effects of cytosolic Ca2+ and Ins(l,4,5)P3, and thereby allow Ca2+ stored within the lumen of the endoplasmic reticulum to leak into the cytosol. The resulting increase in cytosolic [Ca2+] regulates many aspects of cellular behaviour. In the twentyfive years since Ins(l,4,5)P3 was shown to provide the link between receptors in the plasma membrane that stimulate polyphosphoinositide hydrolysis and the release of Ca2+ from intracellular stores (1), novel experimental approaches have often provided the impetus for the steps towards our current understanding. Simple methods to separate [3H]-inositol phosphates, and the use of Li+ to block the final step in their degradation (2), allowed the effects of numerous stimuli on polyphosphoinositide hydrolysis to be established in many cells. Methods that allowed selective permeabilization of the plasma membrane, without destroying the integrity of intracellular organelles (3), were instrumental in allowing the effects of Ins(l,4,5)P3 on intracellular Ca2+ stores to be demonstrated (1). The introduction of fluorescent Ca2+ indicators (4), and their subsequent use with caged compounds (5, 6) and confocal microscopy methods (7), revealed the complex spatial organization of intracellular Ca2+ signals, and established that the signals evoked by Ins(l,4,5)P3 are often regenerative. Small, very localized Ca2+ release events, probably reflecting the opening of the channels of relatively few Ins(l,4,5)P3 receptors, can, if the stimulus exceeds a critical threshold, lead to recruitment of additional Ca2+ release events and thereby to the propagation of a regenerative Ca2+ wave across the cell (8). The luminescent Ca2+ indicator protein, aequorin, has recently enjoyed a revival of interest because it can be targeted to specific

Colin W. Taylor and Jonathan S. Marchant intracellular compartments, and so provide further opportunities to resolve the spatial organization of Ca2+ signals (9). The families of Ca2+ indicators derived from green fluorescent protein look set to provide even more versatile means of visualizing the subcellular organization of intracellular Ca2+ signals (10). Preparation of [32P]-Ins(l,4,5)P3 (11), and then the commercial availability of [3H]-Ins(l,4,5)P3 with high specific activity, allowed Ins(l,4,5)P3 receptors to be identified using radioligand binding methods, and paved the way for their purification by conventional affinity purification methods (12), and subsequent functional reconstitution. Molecular cloning then provided the first structural information, allowed receptor subtypes and splice variants to be identified, and has begun to establish the relationships between Ins(l,4,5)P3 receptor structure and function (13). Throughout these developments, both synthetic chemistry (14) and natural sources (15, 16) have provided ligands with which to address structure-activity relationships and to manipulate Ins(l,4,5)P3 receptor behaviour. Because Ins(l,4,5)P3 receptors can respond very rapidly to Ins(l,4,5)P3, it is important that assays of their function have the appropriate sub-second temporal resolution, and that the methods used minimize the likelihood of the Ca2+ released exerting feedback effects on the behaviour of the receptor (17). From a practical standpoint, the latter is a very significant problem. The biphasic effects of cytosolic Ca2+ on Ins(l,4,5)P3 receptor behaviour are kinetically complex (17), and while fluorescent Ca2+ indicators used with conventional stopped-flow methods (18, 19) provide the simplest and least timeconsuming means of examining the rapid kinetics of Ins(l,4,5)P3-evoked Ca2+ release, they are inherently prone to the difficulty of disentangling direct effects of Ins(l,4,5)P3 from feedback regulation by Ca2+. This chapter describes an alternative method, rapid superfusion of permeabilized cells, and while it focuses specifically on analysis of the responses of rat hepatocytes to Ins(l,4,5)P3, superfusion methods are also readily applicable to analyses of other ion channels. Two other volumes in this series describe methods specifically related to Ca2+ signalling (20), and to inositol phosphate actions and metabolism (21), and additional volumes include relevant chapters on bilayer recording (22) and radioligand binding (23). Detailed descriptions of various methods used in Ca2+ signalling research are also provided in other books (24-26).

2. Preparation, permeabilization, and 45Ca2+ loading of hepatocytes 2.1 Isolation of rat hepatocytes In the thirty years or so since suspensions of viable hepatocytes were first isolated in high yield from intact liver, the methods used to isolate hepatocytes have evolved in many laboratories. The essential principles underlying the 362

15: Measuring Ca2+ release method, and many of the technical details, are comprehensively reviewed in an excellent book (27). Protocol 1 describes the method currently used in our laboratory. Protocol 1. Isolation of viable rat hepatocytes Equipment and reagents • Fine and coarse scissors and forceps • 50 ml plastic syringe • Thin-walled (1 mm inside diameter, 2 mm outside diameter) PVC cannulation tubing (Portex) • • • • • •

.

• 50 ml of MEM, supplemented with additional Ca2+ (final concentration 5 mM) and collagenase (20 mg per 50 ml type 1 from Clostridium histolyticum, ~250 units mg-1 Worthington) • 100ml of MEM supplemented with bovine Silk sutures serum albumin (BSA, Sigma, type V; 2 g Gauze per 100 ml) Petri dish . Eagle's medium (Sigma) supplemented 25 ml conical centrifuge tubes with NaHCO3 (26 mM) and BSA (2 g per 100 Liver perfusion chamber (27) ml) at 4°C Ca2+-free minimal essential medium, MEM • Trypan Blue solution (0.4 mg Trypan Blue (500ml: 116 mM NaCI, 5.4 mM KCI, 0.8 mM in 100 ml of medium containing 150 mM MgSO4, 1 mM NaH2P04, 25 mM NaHCO3, Nacl, 10 mM Hepes, pH 7.4) 11 mM glucose, 500 uM EGTA, gassed with • Gyratory waterbath (New Brunswick 95% O2-5% C02, pH 7.4 at 37°C) Scientific, model G76) MEM (500 ml: as above, but with the EGTA • Bench-top centrifuge omitted and replaced by 1.8 mM CaCI2) • Haemocytometer and microscope

Method 1. Rapidly open the abdomen of an adult male rat (-150 g), killed by cervical dislocation, expose the hepatic portal vein, and tie a loose silk ligature around the vein. 2. Cannulate the vein, using tubing attached to a 50 ml syringe containing Ca2+-free MEM (gassed with 95% O2-5% C02, pH 7.4 at 37°C), secure the cannula with the silk ligature, sever the vena cava, and then slowly manually perfuse the liver with the contents of the syringe. The entire liver should be uniformly cleared of blood. 3. Remove the liver, and secure the cannula to a gravity-fed perfusion system (—10 ml min-1) within a chamber held at 37°C. Continue the perfusion with Ca2+-free MEM for 10 min. Perfuse for 1 min with MEM, and then with MEM supplemented with collagenase (recirculate this medium if necessary). The medium must be gassed with 95% 02-5% C02 throughout. 4. After 10-15 min, the liver will be substantially digested, and it can be removed for gentle mechanical disruption in a Petri dish. If necessary, transfer the dispersed cells to the MEM containing collagenase, and incubate at 37°C in a gyratory waterbath (80 r.p.m.) for 10 min to complete the collagenase digestion. 5. Filter the cells through gauze into 25 ml centrifuge tubes, and recover the cells by centrifugation (40 g, 2 min). 363

Colin W. Taylor and Jonathan S. Marchant Protocol 1. Continued 6. Wash the cells by aspirating the supernatant, gently resuspending the cells in MEM containing 2% BSA, and re-centrifuging. Repeat. 7. Finally, resuspend the cell pellets in 100 ml of the cold Eagle's medium, and leave for about 15 min before assessing cell viability. The viability of the cells, assessed by their ability to exclude Trypan Blue, should exceed 95%. Count the cells using a haemocytometer; a single liver typically provides >3 x 108 cells. 8. The cells can be stored in Eagle's medium at 4°C for 1-2 days without their viability falling below 90%.

2.2 Cell permeabilization, loading of intracellular stores with45Ca2+, and the effects of Ins(l,4,5)P3 In order that Ins(l,4,5)P3, a charged and therefore membrane-impermeant messenger, can gain access to its receptors in the membranes of the intracellular Ca2+ stores, the plasma membrane must be selectively permeabilized while minimizing damage to intracellular organelles. A variety of approaches (3) have been used to permeabilize the plasma membrane, including electroporation, bacterial toxins (streptolysin O and a-toxin), activation of P2X7 receptors by ATP to open large membrane pores, and even removal of bivalent cations from the extracellular medium. The most widely used and simplest means of achieving selective permeabilization of the plasma membrane is provided by detergents (notably saponin and digitonin) that preferentially interact with cholesterol. Careful use of these detergents introduces pores in the plasma membrane that allow the passage of molecules with relative molecular masses < 200 000 (28), while sparing the cholesterol-deficient membranes of intracellular organelles. Protocol 2 describes a typical procedure. It is, however, important to stress that while it is possible with meticulous attention to the permeabilization conditions to minimize disruption of the intracellular morphology (29), under less stringent conditions the intracellular stores retain their ability to sequester and then release Ca2+, but the intracellular architecture may be severely perturbed. The continuity of the endoplasmic reticulum may be disrupted (30). Even very large proteins may be lost from the cytosol: under the conditions (Protocol 2) we use to permeabilize hepatocytes; for example, most lactate dehydrogenase (Mr 130 000) is lost within 3 min (31). Finally, when permeabilized hepatocytes are pelleted by centrifugation, a fraction of the original Ins(l,4,5)P3-sensitive stores are recovered from the supernatant, indicating severe disruption of intracellular architecture (18, 31). Despite these limitations, permeabilized cells afford a simple means of studying Ins(l,4,5)P3 receptor behaviour under precisely controlled conditions; indeed, permeabilized hepatocytes have recently been shown to be capable of generating Ca2+ oscillations (32). 364

15: Measuring Ca2+ release With the plasma membrane no longer protecting the cell interior from the incubation medium, it is important to ensure that the composition of the latter approximates to that of the cytosol, and that the ATP required for active Ca2+ uptake into intracellular stores is maintained at an appropriate level throughout the experiment. The ATP that is hydrolysed to ADP during the incubation can be enzymatically regenerated by adding an ATP-regenerating system, typically creatine phosphate and creatine phosphokinase, to the incubation medium. Unless it is important to be able to remove ATP during the experiment, such an ATP-regenerating system is therefore usually added to the cytosol-like medium. 45 Ca2+ provides the most common and straightforward means of observing ATP-dependent Ca2+ uptake into the intracellular stores of permeabilized cells (Protocol 2). Addition of a fluorescent indicator to a cytosol-like medium with no other Ca2+ buffers provides an alternative approach; as the stores actively sequester Ca2+, the free [Ca2+] in the medium falls, and the fall can be readily recorded using a conventional fluorescence spectrofluorimeter. For experiments using 45Ca2+, either filtration through glass fibre filters or centrifugation can be used to separate the 45Ca2+ trapped within intracellular stores from the much larger amount of 45Ca2+ present in the incubation medium; the former is simpler and most easily adapted to large numbers of samples. In a typical experiment, permeabilized cells would be allowed to load to steady state with 45Ca2+ in a cytosol-like medium (Figure 1), the appropriate additions would be made to the cells (e.g. Ins(l,4,5)P3), and soon afterwards the effects of the additions on the 45Ca2+ content of the stores would be assessed by stopping the incubations, and then rapidly separating the 45Ca2+ trapped within the stores from that in the incubation medium. Because some 45Ca2+ binds non-specifically to both the cells and filters, appropriate controls are required to identify the active ATP-dependent 45Ca2+ uptake into the intracellular stores (i.e. the 45Ca2+ content determined before addition of ATP, or after addition of the ionophore ionomycin to prevent active Ca2+ uptake). Protocol 2 describes a typical method for loading the intracellular stores of permeabilized hepatocytes with 45Ca2+, and then assessing the effects of Ins(l,4,5)P3. Protocol 2.

Loading intracellular Ca2+ stores with 45Ca2+

Equipment and reagents • Rat hepatocytes, isolated according to • ATP-regenerating system containing a 3:5:5 Protocol 1 mixture of ATP (500 mM), phosphocreatine • Cytosol-like medium (CLM: 140 mM KCI; 20 (1 M), and creatine phosphokinase (1 mM NaCI; 1 mM MgCI2; 1 mM EGTA; 20 unit ul-1); each of the reagents is from mM Pipes; CaCI2, 300 uM, to give a free Boehringer [Ca2+] of 200 nM; pH 7.0 at 37°C) • Trypan Blue solution (Protocol 1) 2+ • Ca -free CLM, in which the CaCI2 is • Carbonyl cyanide p-trifluoromethoxyomitted from CLM phenylhydrazone (FCCP: 10 mM in ethanol) • Saponin solution (Sigma, 1 mg 20ul-1 in • lns(1,4,5)P3 (American Radiolabeled Ca2+-free CLM) Chemicals)

365

Colin W. Taylor and Jonathan S. Marchant Protocol 2. 45

Continued 1

• CaCI2 (~8 Ci g- , Dupont) . Glass fibre filters (Whatman, GF/C 25 mm • For some variations of the protocol, the diameter) following may also be required: thapsigar• Ten-place vacuum filtration manifold gin (Alamone Laboratories), BAPTA or (Hoeffer Scientific Instruments) EGTA (Sigma), and a solution of apyrase . Scintillation vials (5 ml) and hexokinase in glucose (Sigma; final • Ecoscint A scintillation fluid (National concentrations, 50 units ml-1 for each Diagnostics) enzyme, and 10 mM for glucose) •Gyratory waterbath • Wash medium (310 mM sucrose; 1 mM trisodium citrate, PH 7) • Bench-top centrifuge • Disposable plastic tubes (75 x 12 mm)

Method 1. Confirm the viability of the hepatocytes using Trypan Blue. Use only preparations with >95% plasma membrane integrity. 2. Remove a volume of Eagle's medium containing ~107 cells, pellet them by centrifugation (40 g, 2 min), and gently resuspend the cells in 25 ml of Ca2+-free CLM supplemented with saponin (5 ul, final saponin concentration 10 ug ml-1) at 37°C. 3. Incubate the cells in a conical flask in a gyratory waterbath (80 r.p.m.) at 37 °C, and periodically sample the cells to monitor the integrity of the plasma membrane using Trypan Blue. Permeabilization of >90% of the cells typically takes 5-10 min, but add more saponin if it takes longer. 4. When >90% of the cells are permeabilized, remove the cells from the medium by centrifugation (650 g, 2 min), resuspend them in 6 ml of CLM at 37°C, and add FCCP (6 ul, final concentration 10 uM) and 45 Ca2+ (final concentration ~2 uCi ml-1). 5. Incubate the cells in a flat-bottomed tube in a gyratory waterbath at 37°C, to keep the cells in suspension. 6. Remove duplicate 200 ul samples to disposable plastic tubes containing ice-cold wash medium (5 ml), rapidly filter the contents through Whatman GF/C filters, and then rinse the tube with a further 5 ml of wash medium. These tubes provide a measurement of the ATPindependent (i.e. non-specific) binding of 45Ca2+, which will be subtracted from all subsequent measurements. 7. Add the ATP-regenerating system (13 ul ml-1, giving final concentrations of 1.5 mM ATP, 5 mM creatine phosphate, and 5 units ml-1 creatine phosphokinase) to the cells to initiate active 45Ca2+ uptake. For hepatocytes, the steady-state 45Ca2+ content is attained within 5 min (Figure 1), but the timing needs to be independently determined for each cell type and incubation condition. The ATP concentration should be increased (to 7.5 mM) and the creatine phosphokinase omitted from experiments in which the ATP is to be enzymatically removed during the experiment. 366

15: Measuring Ca2+ release 8. At appropriate times, remove samples of cells (200 ul) to add to appropriate stimuli (e.g. lns(1,4,5)P3, 4 ul at fifty times its final concentration). lns(1,4,5)P3 typically causes a half-maximal response (EC50) when its concentration is about 200 nM. 9. Stop the reactions after a suitable interval (typically 30s- 5 min), using the method described under Step 6. 10. Add each filter to a scintillation vial, add Ecoscint A scintillation fluid (5 ml), leave overnight for the filters to dissolve, and then determine the 45Ca2+ activity using a liquid scintillation counter. Determine the 45 Ca2+ activity of the CLM by counting a 50 ul aliquot, and from that and its total Ca2+ concentration (300 uM), determine the specific activity of the 45Ca2+ in the incubation medium. Calculate the ATP-dependent 45Ca2+ content of each of the samples of permeabilized cells, and from the cell density and specific activity of the 45Ca2+ compute the steady-state ATP-dependent Ca2+ content of the stores (typically 1-2 nmol per 106 cells for hepatocytes; Figure 1). The effects of the experimental manipulations can then be expressed relative to the steady state Ca2+ content.

Methods similar to those described in Protocol 2 have been successfully applied, with appropriate modifications, to a variety of cell types and

Figure 1. ATP-dependent accumulation of 45Ca2+ into intracellular stores. Permeabilized rat hepatocytes (107 cells ml-1) were incubated in CLM (free [Ca2+] = 200 nM) supplemented with 45 Ca 2+ , ATP, and FCCP for the indicated periods. Cells were then filtered through Whatman GF/C filters using a Brandel receptor-binding harvester, and the amount of radioactivity trapped on the filters was determined by liquid scintillation counting. The results are typical of five independent experiments.

367

Colin W. Taylor and Jonathan S. Marchant experimental requirements. Some of the more useful variations are described below: • The method can be easily adapted to larger numbers of samples by using a Brandel 24-place receptor binding harvester (SEMAT) in place of the filter manifold to allow simultaneous filtration of 24 samples. • Unidirectional 45Ca2+ efflux measurements, rather than the steady-state measurements described in Protocol 2, are straightforward if further active 45 Ca2+ uptake is inhibited before addition of the experimental stimulus. Thapsigargin (to selectively inhibit the Ca2+ pump of the endoplasmic reticulum), removal of ATP (using apyrase, glucose, and hexokinase), or rapid chelation of all free Ca2+ by addition of excess BAPTA or EGTA, alone or in combination (33), provide effective means of rapidly inhibiting further 45Ca2+ uptake. • The composition or temperature of the medium in which the cells are stimulated with Ins(l,4,5)P3 may be varied independently of that used to load the intracellular stores (typically CLM because it most closely mimics the cytoplasmic conditions within an unstimulated cell). This allows the effects of modulators (e.g. Ca2+ or other bivalent cations) (34) on Ins(l,4,5)P3 receptor behaviour to be characterized. The computer programs and fluorescent indicator methods described in a previous volume (20) provide the means of estimating and then reliably determining the free [Ca2+] of incubation media.

3. Rapid kinetic measurements of 45Ca2+ release from intracellular stores 3.1 Rapid superfusion methods In cell populations, initial rates of Ins(l,4,5)P3-stimulated Ca2+ release provide the most direct measure of the activity of the Ca2+ channels, but such rates can be determined only if Ca2+ release can be measured with sufficient temporal resolution. As the temperature is reduced, rates of Ca2+ mobilization are substantially slowed (18, 35), but they are still too fast for initial rates to be resolved with conventional methods. There is also the additional problem that changes in temperature may themselves significantly influence important characteristics of the response. To overcome these problems, rapid filtration (36-38) and rapid superfusion techniques (31, 39-41) have been used to resolve the kinetics of responses to Ins(l,4,5)P3. Both techniques are based upon the pressure-driven movement of medium (typically a cytosol-like medium) around a biological sample immobilized on a porous support. Rapid filtration systems, such as that manufactured by Bio-Logic (36, 37, 42), are designed to allow measurement of the amount of 45Ca2+ left within the cells after a period of superfusion with medium containing the appropriate stimulus 368

15: Measuring Ca2+ release (e.g. Ins(l,4,5)P3). Each immobilized sample of cells therefore provides only a single determination of 45Ca2+; the effects of stimuli must then be assessed by comparison with parallel experiments. Rapid supervision techniques aim to record 45Ca2+ release continuously from a sample of immobilized cells, and are thereby able to detect effects of stimuli within a single run, avoiding the variability introduced by comparisons that depend wholly on parallel runs. Furthermore, the amount of 45Ca2+ released (the signal in superfusion experiments) is typically small relative to that remaining within the cells (the signal recorded in rapid filtration experiments); rapid superfusion therefore offers a much improved signal-to-noise ratio. Both approaches, however, suffer from the drawback that the high temporal resolution of the techniques requires that media are superfused at high flow rates (typically several ml s-1; Protocol 3) throughout the experiment, and that can prove prohibitively expensive for some reagents. Several groups have used rapid superfusion to study the kinetics of intracellular Ca2+ release (31,39-41,43), and the remainder of this chapter will describe the methods in detail. It is, however, worthwhile to consider first the advantages offered by rapid superfusion, because it is undoubtedly more costly, more technically demanding, and more time-consuming than alternative methods. The first major advantage is the ability to measure unidirectional 45Ca2+ release from the intracellular stores, under conditions where local cytosolic Ca2+ gradients (44) are effectively minimized. To resolve the Ins(l,4,5)P3 concentration-dependence of the opening of the Ins(l,4,5)P3 receptor, for example, initial rates of Ca2+ release must be measured under conditions that prevent the feedback regulation by Ca2+ that would otherwise contribute to the response (45). Unless such feedback is eliminated, it becomes very difficult to resolve whether a steep Ins(l,4,5)P3 concentration-effect relationship results from positive feedback by cytosolic Ca2+, or from a genuine need for several subunits of the receptor to bind Ins(l,4,5)P3 before the channel opens (31). The combination of a high concentration of a Ca2+ buffer (EGTA or BAPTA) in the superfusion medium, together with the very rapid flow of medium through the sample, provides the best means of dispersing microdomains of high [Ca2+] in the vicinity of active receptors (36). Even this method may not, of course, prevent the mouth of an open channel from being exposed to a substantial increase in [Ca2+]. When fluorescent indicators are used to record Ca2+ mobilization, the medium [Ca2+] must increase (there is otherwise no signal to detect) (45). Furthermore, with high affinity indicators like Fura-2 and Fluo-3, both the speed and amplitude of local cytoplasmic Ca2+ changes can be severely underestimated, especially during the early stages of Ca2+ mobilization, when spatial gradients of Ca2+ are high (46). Secondly, rapid superfusion can effect rapid changes in ligand concentration on a millisecond timescale (see below). Unless Ins(l,4,5)P3 receptors are exposed to such rapid changes in Ins(l,4,5)P3 concentration, channel activity will be desynchronized by the slowly rising concentration of agonist: 369

Colin W. Taylor and Jonathan S. Marchant some channels will be opening, some closing, and some still waiting to bind agonist. It then becomes very difficult to resolve the kinetics of the various aspects of receptor behaviour. Thirdly, rapid superfusion affords the opportunity to add and remove components of the superfusion medium rapidly: no other method is at present capable of allowing such sequential changes in the composition of the medium, while retaining the temporal resolution required to record the activity of Ins(l,4,5)P3 receptors. Ultimately, of course, single channel methods may provide an alternative means of achieving these aims (47), but the electrophysiological methods so far used to examine Ins(l,4,5)P3 receptor behaviour are incompatible with very rapid changes of medium (48, 49).

3.2 Rapid superfusion apparatus 3.2.1 Overview In this section, we describe the construction of a rapid superfusion apparatus which we have used to investigate the behaviour of Ins(l,4,5)P3 receptors in permeabilized rat hepatocytes (31, 41). The equipment could, however, easily be adapted to other applications, because similar methods have previously been used to examine neurotransmitter release from synaptosomes (50-54), the kinetics of cGMP-stimulated cation efflux in bovine retinal rods (55), the activation of neuromuscular nicotinic acetylcholine receptors (31), and the enzymology of the Na+ pump (56). Various designs of superfusion system have been reported (43, 53, 54, 57); that described here has evolved from its original description (56) and through two subsequent versions (55, 58). Figure 2 shows both a schematic representation and a photograph of our rapid superfusion apparatus. The components needed for its assembly are all available commercially, and are referred to in each of the subsequent sections by a component number (C1-C33, Table 1). The total cost of constructing the apparatus is about GBP4000. Briefly, immobilized cells are mounted in a chamber linked by four solenoid valves to pressurized stainless steel vessels containing the superfusion media. Fluid flows continuously from one of the pressurized reservoirs over the sample, and is partitioned into discrete samples during a single rotation of a fraction collector. Each fraction contains the 45Ca2+ released within a particular collection interval, and therefore represents the average rate of Ca2+ mobilization during that period. More detailed descriptions of the main components of the superfusion apparatus follow in the next four sections. 3.2.2 The superfusion chamber The biological sample is housed in the exit port of a stainless-steel manifold (Cl), fixed in position by an x-y adjustable support arm attached to the turntable casing. Four solenoid valves (C2) gate independent inputs to the 370

15: Measuring Ca2- release

Figure 2. The rapid superfusion apparatus. (a) Four fluid reservoirs are independently connected to a common superfusion chamber housed within the solenoid manifold. The superfusion chamber is accessed by four coaxial ports, each gated by a solenoid-driven valve. Fluid flows continuously from one of the pressure reservoirs over the sample, and is captured by the fraction collector during a single revolution, (b) Photograph of the rapid superfusion apparatus.

371

Colin W. Taylor and Jonathan S. Marchant Table 1. Components of rapid superfusion apparatus

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15 C16 C17 C18 C19 C20 C21 C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 C33

Component

Supplier

Part number

solenoid manifold assembly solenoid (series-9) Teflon poppet armature Kalrez O-ring buffer spring armature spring Teflon bulkhead connector stainless steel washer 1/4-28" UNF outlet fitting PTFE tubing (1/8" by 1/16") pressure filtration vessels stainless steel support screen 1/8" NPT-1/4-28" UNF fitting 3-way valves tube end fittings gripper fittings PTFE cones air thermostat Treff 1.5 ml microfuge tubes optical tachometer stepper motor pulley belt motor indexer X-ware I/O board 8-channel valve driver Reed switch master switch cellulose nitrate SC filter (8 um) pre-filter post-filter Teflon tube-to-tube connector

General Valves General Valves General Valves General Valves General Valves General Valves General Valves General Valves Anachem General Valves Omnifit Gelman Gelman Custom built Omnifit Omnifit Omnifit Omnifit Merck Anachem RS MicroDrives Contitech Parker Hannifin Parker Hannifin Amplicon General Valves RS RS Schleicher & Schuell Schleicher & Schuell Schleicher & Schuell General Valves

9396901+ 93130502 3230502 91810312 9701001 720011 7206401 1310622 P407X 511252 3011 4280,4287 4235 1109 2210 2312 1524 ST60 968160902 CT6-205665 HY2003450350A8 140XL037 Digiplan, PDX15

PC36AT 9031100 338169 333669 AE99 GF51 GF52 1320622

superfusion chamber. Each solenoid is constructed from non-passivated stainless steel, and only inert materials line the path of the superfusion media. Valve (C2) leakage has been estimated to be less than 1 X 10-8 He per second per atmosphere, such that pressures up to 85 atm can be withstood with response times as low as 2 ms (General Valves Corporation, product literature). However, we find that erosion of a Teflon poppet (C3) within the solenoid valve occurs with time, resulting in valve leakage. This component and others within the valves assembly (C4—C7) are therefore routinely replaced. The volume of the exit port of the solenoid manifold is reduced with an adapted Teflon connector (C8) to accommodate the filter array (see later and Figure 4). A stainless steel washer (C9) is immobilized immediately below the common outlet port. The filter array (Figure 4) is secured within the superfusion 372

15: Measuring Ca

release

chamber using a modified fitting (C10), with the internal bore narrowed by PTFE tubing (Cll) to maintain laminar flow. The superfusate leaves through this outlet as a continuous uninterrupted stream. When the outlet fitting is not narrowed, the data points show substantial scatter (not shown). Superfusion media are held under nitrogen pressure in 200 ml stainless steel filtration vessels (C12). A narrow-gauge filter support (C13) is inserted at the base of each pressure vessel, to avoid solid contaminants reaching the solenoid valves. Connecting valves and tubing were selected to present negligible flow resistance relative to the filter. The inlet and outlet ports of each reservoir are connected to PTFE tubing (Cll) by way of machined stainless steel connectors (C14) and 3-way valves to allow pressure release (C15). Connection to the solenoid entry ports is made with tube end fittings (C16) and stainless steel gripper fittings (C17) to provide flangeless, inert, leak-free connections. All valved connections are made with pressure-resistant, inert PTFE cones (C18). The temperature of the entire apparatus can be regulated by housing it within a heated Perspex chamber (C19), or in a refrigerator. Rapid changes in the composition of the superfusion medium are effected by virtue of rapid solution flow through minimal dead volume. Exchange kinetics are routinely followed in every experiment, by supplementing one of the media with a biologically inert radiotracer ([3H]-inulin), which serves the additional role of an internal time marker, permitting changes in the rate of 45 Ca2+ efflux from the intracellular stores to be directly referenced to the time of solution exchange. The half-time for solution exchange under experimental conditions with the filter array present is measured at different superfusion pressures, by introducing a variety of radiolabels ([3H]inulin, [3H]glucose, and 45 Ca2+) into one of two superfusion media, and measuring the rate of appearance and disappearance of radioactivity in the superfusion effluent. Figure 3 shows two switching events between CLM, and CLM supplemented with trace amounts of [3H]-inulin. The kinetics of both solution influx and efflux are described by mono-exponential functions, with a half-time of 30 ± 4 ms at 30 p.s.i. (n = 10). This corresponds to a 5-95% rise time of -130 ms. The halftimes for solution exchange at 20 p.s.i. and 60 p.s.i. are 46 ± 6 ms (n = 10) and 20 ± 1 ms (n = 6) respectively. Each vial on the fraction collector collects 97 ± 3% of the activity expected, calculated from the known dilution of radiolabel, and a measured flow rate of 2 ml s-1. 3.2.3 Variable speed fraction collection The effluent from the superfusion chamber is collected into 100 fractions by cuvettes rotating at the perimeter of a turntable disc (diameter ~42 cm) mounted on a variable-speed fraction collector. The system provides a continuously variable sample interval of between 9 ms and 3000 ms. The turntable platter is made from lightweight material (total mass ~300 g) to minimize torque and inertia, and thereby maximize rotational velocity. Because the length of an experiment is limited by the time taken for a single 373

Colin W. Taylor and Jonathan S. Marchant

Figure3. Kinetics of solution exchange in the super-fusion chamber. The rate of appearance and disappearance of [3H]inulin in the superfusate was measured during two switching events. Each point denotes the amount of [3H]inulin detected in a 20 ms sample collected during superfusion of the chamber with CLM at 2 ml s-1 and 30 p.s.i. Data are expressed as percentages of the expected value, calculated from the dilution of radiolabel and a calculated collection volume of 40 ul (20 ms at 2 ml s-1). The kinetics of solution exchange were fitted with mono-exponential functions (solid line).

rotation of the turntable, a variable-speed fraction collector is essential: it allows the samples that are important for the kinetic analysis to be collected at high temporal resolution, while the remaining samples, which are essential to establish the size of the total 45Ca2+ pool (Protocol 3), can be collected with much lesser temporal resolution into the minimal number of vials. The versatility provided by the variable-speed fraction collection is also advantageous in experiments involving long preincubation periods, where the effluent from the preincubation can be collected into very few vials at a slow collection speed, leaving most of the 100 cuvettes available for collection of the most important samples, with maximal temporal resolution. The turntable is driven from a stepper motor (C22), indirectly geared (3:1, C23) to the turntable platter to generate a wide range of collection intervals. All connections are tightly welded to prevent slippage. The stepper motor is driven by a motor indexer (C24), programmed independently to the superfusion protocol using X-Ware software (C25). The stepper motor (0.03° increments) can accelerate extremely rapidly (±999 rev s-2) over the practicable range of velocities (5-0.01 rev s-1), and can generate a broad range of collection intervals (continuously variable between 9 ms and 3000 ms). The time taken to change velocity within a revolution is verified by sampling the volume of fluid collected by successive vials during a range of velocity steps. In practice, acceleration and deceleration are essentially instantaneous over 374

15: Measuring Ca2+ release the experimental range of collection intervals (9-3000 ms), although to avoid any possible problems, large steps in velocity are effected as gradual ramps and completed before switchings between superfusion media. Rotational speed is verified with an optical tachometer (C21). Cuvettes are closely juxtaposed on a single strip of masking tape, and screwed into the turntable platter. The lumina are bridged using microfuge tubes (C20) to minimize splatter. 3.2.4 Electronic control system Superfusion is controlled from a 24-line programmable digital I/O board (C26) inside a Viglen III/LS microcomputer. The I/O board communicates with both the valve driver (C27) and motor indexer (C24). The sequence and duration of the superfusion protocol is specified using a custom-written GWBASIC program. Motion protocols for the turntable are programmed in XWare (C25) via the RS232 interface. Two sequential signals are required to initiate superfusion: first, a signal from the operator via the computer, to indicate that the superfusion chamber is ready; and second, an automated input from the motor indexer, triggered by closure of the magnetic reed switch (C28), to indicate correct alignment of the turntable. A switch (C29) triggers indefinite turntable rotation until the motor indexer receives this home-position input. The valve driver then energizes the solenoids in the desired sequence, and this is registered on status LEDs and by logic output signals reported through the program interface to confirm independently the switching protocol. The motor driver will terminate superfusion if a programmable end-of-travel limit is reached. These two sequential signals protect the system against current fluctuations, and permit turntable rotation independent of superfusion. This arrangement also permits 'signal averaging', the collection of sequential identical experiments into the same sample holders. 3.2.5 The filter array Selection of the filter system for immobilization of the biological sample is a primary concern, as the overall temporal resolution of the apparatus is very much determined by it. Ideally, the filter array should retain the maximal amount of the biological sample without clogging, provide mechanical stability to prevent damage from shear forces and turbulence at high flow rates, yet occupy minimal volume and provide minimal resistance to flow. Although high flow rates increase the temporal resolution, an increased pressure gradient across the sample decreases the amount of intact material retained, and also leads to artefactual release of radiolabel from the biological sample when the fluid flow is perturbed on switching. Temporal resolution is inevitably therefore a compromise with the need to maintain mechanical stability of the sample: the exact compromise will depend upon the nature of the biological specimen. We have found that the optimal filter composition for stable immobilization of permeabilized rat hepatocytes consists of a cellulose nitrate filter triplet 375

Colin W. Taylor and Jonathan S. Marchant (8 uM, C30), protectcd by a pre-filter to dampen pressure surges on switching (C31), and a post-filter (C32) to trap dislodged material (Figure 4). The filter array is supported by a 5 mm diameter stainless steel mesh (C33) and a stainless steel washer (C9), The use of permeabili/cd cells has obviated the need for low porosity filters, thereby permitting a high flow rate, and hence high temporal resolution without sacrificing the stability of the sample. During an experimental run (Protocol 3), the filter array is assembled externally to the supcrfusion apparatus in one end of a connector fitting (C33). Permeabilized cells loaded to steady state with 45Ca2+ (Protocol 2) are immobilized onto the filter triplet by filtration, and the filter array is then rapidly transferred to the superfusion apparatus. The steps between transferring the cells to the filter and initiating the experimental protocol are completed within 20 s. Since the half-time for passive 45Ca2 efflux from the intracellular stores is 162 ± 13 s at 20°C (n = 3), the Ca2+ contents of the intraeellular stores at the onset of the experiment are no less than -92% of the steady-state level. Using this filter array, the rate of 45 Ca 2+ release from permeabilized hepatocytes is unaffected by switchings between control media, confirming that the apparatus allows rapid changes in superfusion medium without dislodging or damaging the intraeellular stores (Figure 5). The only artefactual release of

Figure 4. Filter array used for immobilization of permeabilized rat hepatocytes. The intraeellular stores of permeabilized rat hepatocytes were loaded with 45Ca?' and immobilized on a cellulose ester filter triplet, sandwiched between a pre-filter and post-filter baffle and supported by stainless steel gauze and washers. An outlet fitting (C10), narrowed with PTFE tubing, secured the assembly into the superfusion chamber shown on the left. 376

15: Measuring Ca2+ release

Figure 5. Stable immobilization of permeabilized rat hepatocytes on the filter array. The intracellular stores of permeabilized hepatocytes were loaded to steady state with 45Ca2+, and immobilized on the filter array (Protocol 3). After two 5 s prewashes, cells were superfused with CLM for 8.5 s, interrupted by control switchings at 3.25 s and 5.875 s. The timecourse of the switching event is reported with the inert volume marker [3H]inulin (dotted line, right axis). After 8.5 s, cells were superfused with CLM supplemented with Triton X-100 (T, 50 mg per 100 ml). 46Ca2+ release (black circles) collected during each 125 ms interval is expressed as a percentage of the total amount of 45Ca2+ within the stores trapped on the filters (left axis). The trace has been corrected for the unstimulated rate of 46 Ca2+ efflux.

45

Ca2+ occurs when the fluid flow is first initiated, and this probably reflects the washout of 45Ca2+ from cells damaged during setting-up. However, subsequent superfusion of the detergent, Triton X-100 (50 mg per 100 ml), causes a rapid increase in the rate of release of 45Ca2+, such that -90% of the 45Ca2+ remaining on the filter is released within 250 ms. Similar results, although with a slower time course, are obtained with the Ca2+ ionophore ionomycin (not shown). Integration of the area beneath the data points can be used to estimate the total amount of 45Ca2+ initially immobilized on the filter. Because the total amount of sequestered 45Ca2+ immobilized on the filters varies between independent runs, cells are routinely superfused with CLM 377

Colin W. Taylor and Jonathan S. Marchant containing Triton X-100 (50 mg per 100 ml) at the end of every experiment, to quantify the amount of stored 45Ca2+ immobilized on the filters.

3.3 Rapid responses to Ins(l,4,5)P3 Protocol 3 describes a typical method using the rapid superfusion apparatus to examine the rapid kinetics of 45Ca2+ evoked by Ins(l,4,5)P3. Protocol 3.

Rapid kinetics of lns(1,4,5)P3-stimulated 45Ca2+ efflux

Equipment and reagents • Rapid superfusion apparatus • PVC tubing (Portex, 1 mm inside diameter, 2 mm outside diameter) • Permeabilized rat hepatocytes loaded to steady state with 45Ca2+ in CLM (Protocol 2} • lns(1,4,5)P3 • Triton X-100 (50 mg per 100 ml)

CLM [3H)-inulin (as an inert volume marker) Treff 1.5 ml microfuge tubes (Scotlab) Scintillation vials Ecoscint A scintillation fluid (National Diagnostics)

Method 1. Connect three cylinders to the superfusion chamber, one containing CLM (Protocol 2), the second containing CLM supplemented with lns(1,4,5)P3 (10 uM) and trace amounts of [3H]-inulin, and the third containing CLM with Triton X-100 (50 mg per 100 ml). Block all unused solenoid valve ports with an occluded screw (C16), and pressurize the cylinders with nitrogen. 2. Load the intracellular Ca2+ stores of permeabilized rat hepatocytes (107 cells ml-1 to steady state with 45Ca2+ by incubation for 5 min at 37°C (Protocol 2). During this time, prime all the tubing leading to the superfusion chamber by sequentially energizing the connected valves for 500 ms. Wash out residual traces of medium from the superfusion chamber by perfusing for 5 s with CLM. 3. Immobilize the cells on the filter array by first transferring them to PVC tubing connected to a 1 ml syringe, and then gently forcing the suspension into the filter triplet, prerinsed with CLM, and previously assembled in one end of a Teflon connector (C33). Rapidly transfer the filter array into the superfusion chamber and secure it by adding the stainless steel mesh and washer (C9) and then tighten with a Teflon screw (C10). 4. Adjust the flow rate of the superfusate through the sample to 2 ml s-1 during two prewashes (5 s), by tightening or loosening the Teflon screw (C10). 5. Secure the solenoid manifold in position above the fraction collector, and initiate a superfusion protocol. A typical sequence would com378

15: Measuring Ca2+ release prise 2 s exposure to CLM, followed by a 5 s exposure to lns(1,4,5)P3, and finally a 4 s exposure to Triton X-100 (50 mg per 100 ml) to release all 45Ca2+ remaining within the intracellular stores. The exact timings are modified to meet the specific aims of the experiment. 6. Determine the radioactivity released into each sample (3H and 45Ca2+) by dual-label liquid scintillation counting in EcoScint-A scintillation cocktail. 7. Most experiments are likely to be analysed by expressing the amount of 45Ca2+ released by lns(1,4,5)P3 as a fraction of the entire 45Ca2+ content of the immobilized cells (i.e. that released during the entire experiment, together with that released by Triton X-100 at the end (Figures 6-8). Alternative forms of analysis may be more appropriate for some experiments (59) (e.g. 'fractional release rate', i.e. rates expressed as fractions of the Ca2+ content of the stores at the beginning of each interval during which the 45Ca2+ release is recorded). Whichever form of analysis is adopted, simple spreadsheets (e.g. Microsoft Excel) provide the most convenient means of processing the large amounts of data generated.

Results from a typical rapid superfusion experiment are shown in Figure 6a. A maximal concentration of Ins(l,4,5)P3 (10 uM) caused a transient release of 45Ca2+ from the intracellular stores: the rate of 45Ca2+ release rapidly accelerated towards a peak, which was abruptly followed by a protracted decline in the rate of 45Ca2+ mobilization (59). A saturating concentration of Ins(l,4,5)P3 released -30% of the intracellular Ca2+ stores (Figure 6b), after which no further 45Ca2+ release could be evoked by a further addition of Ins(l,4,5)P3. The size of the Ins(l,4,5)P3-sensitive Ca2+ stores detected by the superfusion protocol is similar to that observed with conventional methods (35), and substantially larger than that found in brain synaptosomes (—6%) (39) or hepatic microsomes (~3%) (40). The fraction of the intracellular Ca2+ stores released by prolonged superfusion with Ins(l,4,5)P3 increased with Ins(l,4,5)P3 concentration, indicating that the unusual 'quantal' pattern of Ins(l,4,5)P3-evoked Ca2+ release was retained (60). The concentrationdependence of the ultimate extent of Ins(l,4,5)P3-stimulated 45Ca2+ release (half-maximally effective concentration, EC50 477 ± 21 nM), and of the maximal rate of 45Ca2+ mobilization differed by only twofold (EC50 941 ± 21 nM) (31, 41). This observation, that rates of Ca2+ release are more sensitive than their eventual extents to Ins(l,4,5)P3 (59), further underscores the necessity for rapid measurements, if the characteristics of the initial states of the Ins(l,4,5)P3 receptor are to be examined. The rank order of potency of several different agonists (Ins(l,4,5)P3 > 3-deoxy-Ins(l,4,5)P3 > 2,3-dideoxyIns(l,4,5)P3 > Ins(2,4,5)P3) was the same as previously reported with conventional methods (31, 61) (Figure 6c). In summary, therefore, the key 379

Colin W. Taylor and Jonathan S. Marchant

Figure 6. Rapid superfusion with lns (1,4,5)P3 evokes responses with similar characteristics to those recorded using conventional methods. (A) The time course of 45Ca2 mobilization in response to two sequential superfusions with 10 uM lns(1,4,5)P3 in CLM (solid bar) is shown. The shaded area indicates the kinetics of delivery of lns(1,4,5)P3 to the superfusion chamber, detected using [3H]inulin. Each point represents the average rate of 45Ca21 efflux during each 80 ms collection interval. (B) The cumulative amount of 45 Ca 2 1 released with time, obtained by integration of the results shown in (a), and expressed relative to the size of the intracellular Caz store, (C) Concentration-dependence of the eventual extent of 45Ca21 release in response to lns(1,4,5)P3 (black circles), 3-deoxylnsl1,4,5)P3 (open squares), 2,3-dideoxy-lns(1,4r5)P3 (black squares), and lns(2,4,5)P3 (open circles). Results, which are means ± SEM of at (east three independent experiments, are expressed as percentages of the maximal response.

380

15: Measuring Ca2+ release characteristics of Ins(l,4,5)P3-evoked Ca2+ release are preserved under the conditions required for their analysis by rapid superfusion methods. An additional advantage of these rapid superfusion methods is shown by the results in Figure 7, in which the kinetics of the initial activation of the Ins(l,4,5)P3 receptor are resolved (31). Our previous detailed analysis of the delay between exposure of Ins(l,4,5)P3 receptors to Ins(l,4,5)P3 and channel opening led us to propose a model in which binding of Ins(l,4,5)P3 causes a Ca2+-binding site to become exposed, and only when all four subunits of the tetrameric receptor have bound both Ins(l,4,5)P3 and Ca2+ does the Ca2+ channel open (31). A unique advantage of the superfusion system is evident from Figure 8, in which the kinetics of the termination of 45Ca2+ release after brief pulses of Ins(l,4,5)P3 are examined (59). By rapidly removing Ins(l,4,5)P3 at different phases of the response, the superfusion apparatus allows the rate of termination of 45Ca2+ release, reflecting the rate of dissociation of Ins(l,4,5)P3 from its receptor, to be measured. The kinetics of the decay in the rate of 45Ca2+ release are faster when Ins(l,4,5)P3 is removed after 380 ms (the half-time for recovery is 195 ± 20 ms, n = 10) than when it was removed after 640 ms (when the half-time for recovery is 360 ± 35 ms, n = 4). The results are consistent with Ins(l,4,5)P3 first binding to a lowaffinity active receptor conformation, after which the receptor rapidly (with a

Figure 7. Delayed Ca2+ release after maximal stimulation of lns(1,4,5)P3 receptors with lns(1,4,5)P3. Permeabilized hepatocytes loaded with 45Ca2+ were stimulated with a supramaximal concentration of lns(1,4,5)P3 (300 uM), and its arrival in the effluent was monitored by inclusion of [3H]-inulin (black circle). 45Ca2+ release during each 9 ms interval of the superfusion was recorded (open circle). Even though the concentration of lns(1,4,5)P3 to which the cells are exposed exceeds that required for maximal stimulation within 9 ms, there is no detectable release of 45Ca2+ for 31 ± 2 ms. These results, together with those from additional rapid superfusion experiments, provide the evidence from which we suggest that both lns(1,4,5)P3 and Ca2+ must sequentially bind to the lns(1,4,5)P3 receptor before its channel can open. Reproduced with permission from ref. 31.

381

Colin W. Taylor and Jonathan S. Marchant

Figure 8. Kinetics of channel closure after removal of lns(1,4,5)P3. Cells were exposed to lns(1,4,5)P3 for the periods indicated, and the rate of decay in the rate of 45Ca2+ release was measured in response to a 380 ms pulse (black triangle), a 640 ms pulse (open triangle), or a 4 s challenge with 10 uM lns(1,4,5)P3 (open circle). For clarity, only the latter parts of the traces that employed short pulses of lns(1,4,5)P3 are shown. Results are typical of at least four independent experiments. The inset shows normalized monophasic curve fits to the kinetics of the decay in the rate of 45Ca2+ release following the removal of lns(1,4,5)P3, after 380 ms (solid curve), and 640 ms (dashed curve). Reproduced with permission from ref. 59.

half-time of 250 ms) switches to a higher affinity, less active conformation (59). In summary, rapid superfusion methods are beginning to reveal some of the complex changes in Ins(l,4,5)P3 receptors that follow their initial activation (31, 59), and it is particularly satisfying that the results obtained with these methods, which allow rigorous control of experimental conditions, are in close agreement with those obtained using flash photolysis of caged Ins(l,4,5)P3 in intact cells (62).

Acknowledgements Supported by The Wellcome Trust and BBSRC.

References 1. Streb, H., Irvine, R. K, Berridge, M. J., and Schulz, I. (1983). Nature, 306, 67. 2. Berridge, M. J., Downes, C. P., and Hanley, M. R. (1982). Biochem. J., 206, 587. 382

15: Measuring Ca2+ release 3. Schulz, I. (1990). In Methods in enzymology (ed. S. Fleischer and B. Fleischer) Vol. 192, p. 280. Academic Press, London. 4. Tsien, R. Y. (1989). Anna. Rev. Neurosci., 12, 227. 5. Adams, S. R., and Tsien, R. Y. (1993). Annu. Rev. Physiol, 55, 755. 6. Callamaras, N., and Parker, I. (1998). In Methods in enzymology. 291, (ed. G. Marriott.), p. 380, Academic Press, London. (In Press) 7. Parker, I., Callamaras, N., and Wier, W. G. (1997). Cell Calcium, 21, 441. 8. Berridge, M. J. (1997). J. Physiol., 499.2, 291. 9. Brini, M., Marsault, R., Bastianutto, C., Alvarez, J., Pozzan, T., and Rizzuto, R. (1995). J. Biol. Chem., 270, 9896. 10. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Nature, 388, 882. 11. Spat, A., Bradford, P. G., McKinney, J. S., Rubin, R. P., and Putney, J. W., Jr. (1986). Nature, 319, 514. 12. Supattapone, S., Worley, P. F., Baraban, J. M., and Snyder, S. H. (1988). J. Biol. Chem., 263, 1530. 13. Mikoshiba, K. (1997). Curr. Opin. Neurobiol., 7, 339. 14. Potter, B. V. L., and Lampe, D. (1995). Angew. Chem. Int. E. Engl., 34, 1933. 15. Takahashi, M., Tanzawa, K., and Takahashi, S. (1994). J. Biol. Chem., 269, 369. 16. Gafni, J., Munsch, J. A., Lam, T. H., Catlin, M. C., Costa, L. G., Molinski, T. F., and Pessah, I. N. (1997). Neuron, 19, 723. 17. Taylor, C. W. (1998). Biochim. Biophys. Acta., 1436, 19. 18. Champeil, P., Combettes, L., Berthon, B., Doucet, E., Orlowski, S., and Claret, M. (1989). J. Biol. Chem., 264, 17665. 19. Meyer, T., Holowka, D., and Stryer, L. (1988). Science, 240, 653. 20. McCormack, J. G., and Cobbold, P. H. (ed.) (1991) Cellular calcium: a practical approach. IRL Press, Oxford. 21. Shears, S. B. (ed.) (1997) Signalling by inositides: a practical approach, IRL Press, Oxford. 22. Williams, A. J. (1995) In Ion channels: a practical approach (ed. R. H. Ashley), p. 43. IRL Press, Oxford. 23. Hulme, E. C. (ed.) (1992) Receptor-ligand interactions: a practical approach, IRL Press, Oxford. 24. Nucitelli, R. (ed.) (1994) Methods in cell biology, Vol. 40, Academic Press, San Diego. 25. Boulton, A. A., Baker, G. B., and Taylor, C. W. (ed.) (1992) Neuromethods, Vol. 20, p.. Humana Press, Totowa, New Jersey. 26. Mason, W. T. (ed.) (1993) Fluorescent and luminescent probes for biological activity. A practical guide to technology for quantitative real-time analysis. Academic Press, London. 27. Berry, M. N., Edwards, A. M., and Barritt, G. J. (1991) Isolated hepatocytes. Preparation, properties and applications, Elsevier, Amsterdam. 28. Matter, N., Ritz, M-F., Reyermuth, S., Rogue, P., and Malviya, A. N. (1993). J. Biol. Chem., 268, 732. 29. Burgess, G. M., McKinney, J. S., Fabiato, A., Leslie, B. A., and Putney, J. W., Jr. (1983). J. Biol. Chem., 258, 15336. 30. Renard-Rooney, D. C., Hajnoczky, G., Seitz, M. B., Schneider, T. G., and Thomas, A. P. (1993). J. Biol. Chem., 268, 23601. 383

Colin W. Taylor and Jonathan S. Marchant 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.

Marchant, J. S., and Taylor, C. W. (1997). Curr. Biol., 7, 510. Hajnoczky, G., and Thomas, A. P. (1997). EMBO J., 16, 3533. Beecroft, M. D., and Taylor, C. W. (1998). Biochem. J., 334, 431 Marshall, I. C. B., and Taylor, C. W. (1994). Biochem. J., 301, 591. Beecroft, M. D., and Taylor, C. W. (1997). Biochem. J., 326, 215. Combettes, L., Hannaert-Merah, Z., Coquil, J.-F., Rousseau, C., Claret, M., Swillens, S., and Champeil, P. (1994). J. Biol. Chem., 269, 17561. Hannaert-Merah, Z., Combettes, L., Coquil, J-F., Swillens, S., Mauger, J-P., Claret, M., and Champeil, P. (1995). Cell Calcium, 18, 390. Moulin, M-J., and Dupont, Y. (1988). J. Biol. Chem., 263, 4228. Finch, E. A., Turner, T. J., and Goldin, S. M. (1991). Science, 252, 443. Dufour, J-F., Arias, I. M., and Turner, T. J. (1997). J. Biol. Chem., 272, 2675. Marchant, J. S., Chang, Y-T., Chung, S.-K., Irvine, R. F., and Taylor, C. W. (1997). Biochem. J., 321, 573. Dupont, Y., and Moulin, M.-J. (1987). In Methods in enzymology (ed. L. Packer, and R. Douce) Vol. 148, p. 675. Academic Press, London. Home, J. H., and Meyer, T. (1995). Biochemistry, 34, 12738. Stern, M. D. (1992). Cell Calcium, 13, 183. Taylor, C. W., and Traynor, D. (1995). J. Membrane Biol., 145, 109. Ogden, D., Khodakhah, K., Carter, L, Thomas, M., and Capiod, T. (1995) Pflug. Archiv., 429, 587. Jonas, E. A., Knox, R. J., and Kaczmarek, L. (1997). Neuron, 19, 7. Bezprozvanny, L, and Ehrlich, B. E. (1994). J. Gen. Physiol., 104, 821. Mak, D.-O. D., and Foskett, J. K. (1997). J. Gen. Physiol., 109, 571. Pearce, L. B., Buck, T., and Adamec, E. (1991). J. Neurochem., 75, 636. Turner, T. J., and Goldin, S. M. (1989). Biochemistry, 28, 586. Turner, T. J., Adams, M. E., and Dunlap, K. (1993). Proc. Natl. Acad. Sci. USA, 90, 9518. Raiteri, M., Angelini, F. and Levi, G. (1974). Eur. J. Pharmacol., 25, 411. Minnema, D., and Michaelson, I. A. (1985). J. Neurosci. Methods, 14, 193. Pearce, L. B., Calhoon, R. D., Burns, P. R., Vincent, A., and Goldin, S. M. (1988). Biochemistry, 27, 4396. Forbush, B. (1984). Anal. Biochem., 140, 495. Mulder, A. H., Van der Berg, W. B., and Stoff, J. C. (1975). Brain Res., 99, 419. Turner, T. J., Pearce, L. B., and Goldin, S. M. (1989). Anal. Biochem., 178, 8. Marchant, J. S., and Taylor, C. W. (1998). Biochemistry 37, 11524. Muallem, S., Pandol, S. J, and Becker, T. G. (1989). J. Biol. Chem., 264, 205. Safrany, S. T., Wilcox, R. A., Liu, C., Dubreuil, D., Potter, B. V. L., and Nahorski, S. R. (1993). Mol. Pharmacol., 43, 499. Parker, I., Yao, Y., and Ilyin, V. (1996). Biophys. J., 70, 222.

384

A1 List of suppliers Alamone Labs Ltd., Shatner Center 3, PO Box 4287, Jerusalem, 91042, Israel. American Radiolabeled Chemicals Inc., 11624 Bowling Green Drive, St Louis, MO 63146, USA. Amersham Pharmacia Biotech, Amersham Place, Little Chalfont, HP7 9NA, UK. Amicon Amicon, Inc., 72 Cherry Hill Drive, Beverly, MA 01915, USA. Amicon Ltd., Upper Mill, Stonehouse, Gloucestershire, GL10 2BJ, UK. Amplicon, Centenary Industrial Estate, Hollingdean Road, Brighton, BN2 4AW, UK. Anachem, Anachem House, 20 Charles Street, Luton, Bedfordshire, LU2 OEB, UK. Anatrace, Inc., 434 West Dussel Drive, Maumee, OH 43537, USA. Bellco Bellco Glass Inc, PO Box B, 340 Edrudo Road, Vineland, NJ 08360, USA. Philip Harris Scientific, Lynn Lane, Shenstone, Lichfleld, Staffordshire, WS14 OEE, UK. Bio-Rad Laboratories Ltd, Bio-Rad House, Maylands Avenue, Hemel Hempstead, Hertfordshire, HP2 7TD. Calbiochem-Novabiochem Calbiochem-Novabiochem Corporation, 10394 Pacific Center Court, San Diego, CA 92121, USA. Calbiochem-Novabiochem (UK) Ltd., Boulevard Industrial Park, Padge Road, Beeston, Nottingham, NG9 2JR, UK. Clontech Laboratories Clontech UK, Unit 2, Intec 2, Wade Rd., Basingstoke, RG24 8NE, UK. Clontech Laboratories, Inc., 1020 East Meadow Circle, Palo Alto, CA 94303, USA. Contitech Power Transmissions Systems Ltd., PO Box 26, Wigan, WN2 4WZ, UK. Dynatech Laboratories Inc., 14340 Sullyfield Circle, Chantilly, VA 22021, USA. Gelman Sciences Ltd., Brackmills Business Park, Caswell Road, Northampton, NN4 0EZ, UK.

List of suppliers Genera] Valves, 19 Gloria Lane, PO Box 1333, Fairfield, NJ 07004, USA. Gibco BRL, 3175 Staley Road, Grand Island, NY 14072, USA. Hoeffer Scientific Instruments, Newcastle, Staffs, ST5 0TW, UK. ICN Biomedicals Ltd., Lincoln Road, Cressex Industrial Estate, High Wycombe, Bucks, HP12 3XJ. Invitrogen BV, De Schelp 12, 9351 NV Leek, The Netherlands. Jencons Scientific Ltd., Cherrycourt Way Industrial Estate, Stanbridge Road, Leighton Buzzard, LU7 8UA. JRH Biosciences JRH Biosciences Inc., 13804 W. 107th Street, Lenexa, KS 66215, USA. AMS Biotechnology, 12 Thorney Leys Park, Witney, Oxon, OX8 7GE, UK. Life Technologies Ltd, 3 Fountain Drive, Inchinnan Business Park, Paisley, PA2 9RF, UK. Macherey-Nagel, Postfach 101352 D-52313, Diiren, Germany. MicroDrives, Aercon House, Alfred Road, Gravesend, Kent, DA11 7QF, UK. Molecular Devices Molecular Devices Ltd., Unit 6, Raleigh Court, Rutherford Way, Crawley, West Sussex, RH10 2PD. Molecular Devices Corporation, 1311 Orleans Drive, Sunnyvale, CA 94089. Molecular Probes, Inc., 4849 Pitchford Avenue, Eugene, OR, USA. National Diagnostics, Unit 3, Chamberlain Road, Aylesbury, Bucks, HP19 3DY, UK. New Brunswick Scientific Co. Ltd., Edison, NJ, USA. Novagen Novagen Inc., 601 Science Drive, Madison, WI 53711, USA. Cambridge Bioscience, 24-25 Signet Court, Newmarket Road, Cambridge, CB5 8LA, UK. NUNC, distributed by Life Technologies. Omnifit, 2 College Park, Coldhams Lane, Cambridge, CB1 3HD, UK. Packard Bioscience Packard Instrument Company, 800 Research Parkway, Meriden, CT 06450, USA. Canberra Packard Ltd., Brook House, 14 Station Road, Pangbourne, Berkshire RG8 7AN, England. Parker Hannifin, 21 Balena Close, Poole, Dorset, BH17 7DX, UK. Parr Parr Instrument Company, 211 Fifty-third Street, Moline, Illinois 61265-1770, USA. Scientific & Medical Products Ltd., Shirley Institute, 856 Wilmslow Road, Didsbury, Manchester, M20 2SA, UK. Pharmacia, 800 Central Avenue, Piscataway, NJ 08854, USA. Pharmingen Pharmingen, 10975 Torreyana Road, San Diego, CA 92121, USA. 386

List of suppliers Becton Dickinson UK Ltd., Between Towns Road, Cowley, Oxford, OX4 3LY, UK. Pierce-Warriner Pierce Chemical Co., 3747 N. Meridian Rd., PO Box 117, Rockford, IL 61105, USA. Pierce & Warriner (UK) Ltd., 44 Upper Northgate Street, Chester CHI 4EF, UK. Portex Ltd., Hythe, Kent, CT21 6JL, UK. Qiagen Ltd, Boundary Court, Gatwick, Road, Crawley, RH10 2AX, UK. Research Biochemicals International, 1 Strathmore Road, Natick, Massachusetts 01760-2447, USA. RS, Venture Close, Lammas Road, Corby, Northants, NN17 1UB, UK. Santa Cruz Biotechnology Santa Cruz Biotechnology Inc., 2161 Delaware Avenue, Santa Cruz, California 95060. Calbiochem-Novabiochem (UK) Ltd, Boulevard Industrial Park, Padge Road, Beeston, Nottingham, NG9 2JR. Insight Biotechnology Ltd., PO Box 520, Wembley, Middlesex, HA9 7YN, UK. SEMAT Technical (UK) Ltd, 1 Executive Park, Hatfield Road, St Albans, Herts, AL1 4TA, UK. Sigma-Aldrich. PO Box 14508, St. Louis, MO 63178, USA. Transduction Laboratories Transduction Laboratories, 133 Venture Court, Lexington, KY 40511-2624, USA. Affiniti Research Products Ltd., Mamhead Castle, Mamhead, Exeter, EX6 8HD, UK. Tropix, 47 Wiggins Ave., Bedford, Massachusetts 01730, USA. Upstate Biotechnology Upstate Biotechnology, 199 Saranac Avenue, Lake Placid, NY 12946, USA. TCS Biologicals Ltd., Botulph Claydon, Buckingham, MK18 2LR, UK. Wallac Wallac Inc., 9238 Gaither Road, Gaithersburg, Maryland 20877, USA. EG&G Ltd., Milton Keynes, 20 Vincent Avenue, Crownhill Business Centre, Crownhill, Milton Keynes, MK8 0AB, UK. Whatman International Ltd., St Leonard's Road, 20/20 Maidstone, Kent, ME16 0LS. UK. Worthington Biochemical Corporation, Halls Mill Road, Freehold, NJ 07728, USA. Zymed Zymed Laboratories Inc., 458 Carlton Court, South San Francisco, CA 940802012, USA. Cambridge Bioscience, 24-25 Signet Court, Newmarket Road, Cambridge CBS SLA, UK. 387

This page intentionally left blank

Index AcMNPV see baculovirus A1 adenosine receptor, interactions with G proteins 128-32 adenovirus culture 228-30 replication-defective 228-9 transfection into mammalian cells 228—32 adenylyl cyclase assay 233–1,240,244-8 see also cyclic AMP, assay and hormonal stimulation 73 purification 224–5 recombinant 226 regulation 223–4,240-1 structure 223, 225 transfection into mammalian cells 226–33 ADP-ribosylation by bacterial toxins 77-81 G proteins 73, 77-81,325 «2 adrenergic receptor assay 117-19 specific antagonist binding 117-19 a2A adrenergic receptor, agonists 121-3 (padrenergic receptor, agonist stimulation and phosphorylation 59 (p2 adrenergic receptor (p2AR) agonist binding and conformational change 8–9,32-3 assay 6 expression in baculovirus system 156 expression in mammalian cells 39 fluorescence labelling 2, 6–16 palmitoylation 41-3 purification 2–6,44-5 site-selective labelling 11-16 adrenoceptor see adrenergic receptor aequorin, measurement of intracellular calcium 173, 177-8,182-4,214-15, 361-2 affinity labelling, identification of binding sites 20 affinity purification, G protein-coupled receptors (GPCRs) 40–3 agonists activation of phospholipase D 303–1 for a2A adrenergic receptor 121-3 incubation with cells 303—4 inverse, assays 200-1 measurement of efficacy 121-3 reporter-gene assays 198–201 agonist trafficking 130 alcohols, substrates for phospholipase D 302 alkaline phosphatase see secreted placental alkaline phosphatase (SEAP)

antagonists, reporter-gene assays 199 antibodies G-protein a subunits 87-90 mitogen-activated protein (MAP) kinases 346 phospho-specific 348-9 phosphotyrosine 343 production 87-90, 348-9 specificity 87-8,90,343 antigenicity bacterial fusion proteins 60–1 G protein-coupled receptors (GPCRs) 61 apoptosis, involvement of ceramide 321 ATP, metabolic labelling for cyclic AMP assay 234–5 Autographa californica multiple nuclear polyhedrosis virus see baculovirus autoradiography, SDS-PAGE gels 86 baculovirus amplification 155–6 cell lines for culture 140–4,166 construction of recombinant 145, 147-55 expression of G protein-coupled receptors (GPCRs) 37-8 infection of Sf9 cells 4, 157-60 life cycle 144–5 promoters for recombinant protein expression 145–6,167 protease-null strain 158 recombinant protein expression levels 156–8 baculovirus expression vector system (BEVS) 139–69 see also baculovirus binding sites, identification affinity labelling 20 site-directed mutagenesis 20, 24 substituted-cysteine accessibility method (SCAM) 20–4, 30–1 biotinylation, cell-surface receptors 52 blue fluorescent protein (BFP) 216 Ca2+ channels, inositol 1,4,5-triphosphate (Ins(l,4,5)P3) receptors 361-2 calcium ions fluorescent indicators 214-15, 361-2, 365, 369 intracellular measurement 173, 177-8, 182-4,205,214-17,361-2 loading intracellular stores 365-8 mobilization by sphingosine kinase 324

Index calcium ions (continued) regulation of cellular behaviour 361-2 release from intracellular stores 368-70, 378-82

response to inositol 1,4,5-triphosphate (Ins(l,4,5)P3) 361-2, 365, 367-8, 378-82 second messenger 105, 361-2 cAMP see cyclic AMP cell culture oxygen air-lift fermenters 167 serum-free media 143–1, 168 Sf9 cells 4, 38, 140-4,155-6,158-60,167-8 see also culture media cell membranes see plasma membranes cells, viability testing 366 cell-surface receptors biotinylation 52 palmitoylation 51–t and phosphoinositide 3-kinases (PI3Ks) 283 ceramide and apoptosis 321 signal transduction 321-2 chloramphenicol acetyltransferase (CAT) assay 207-10 reporter gene 195 cholera toxin, action on G proteins 73-4, 77-81, 87, 105-6 chromatography inositol glycerophospholipids 258-9 separation of mitogen-activated protein (MAP) kinases 351-2 see also HPLC; thin-layer chromatography (TLC) c-jun kinases (JNKs) antibodies 346 assay 356 properties 337-8 see also mitogen-activated protein (MAP) kinase conformational change, and receptor activation 32-3 constitutively active mutants, and receptor activation 32-3 constitutive receptor signalling, reporter-gene assays 200–1 cotransfection, construction of recombinant baculovirus 145, 147-52 CRE (cAMP response element), component of promoters 172, 174-80, 240-1 CRE-binding element (CREB), transcription factor 174-5, 178-9,240 culture media mammalian cells 39 for reporter-gene assays 192-5 Saccharomyces cerevisiae 259-61 serum-free 143–4, 168

Sf9 cells 38, 141,143-4,168 cyclic adenosine monophosphate see cyclic AMP cyclic AMP (cAMP) assay 233–4 fluorescent analogues 249 isolation of tritiated 236–9 second messenger 105 stimulation of protein kinase A 248-9 cysteine fluorescent labelling 6-8,12-15 prenylation 35 reaction with methane thiosulfonate (MTS) derivatives 25-7 substitution in receptors 21-2 cytosol, preparation 311-12 deacylation, inositol glycerolipids 271–1 dopamine D2 receptor, mapping by substituted-cysteine accessibility method (SCAM) 20, 30-2 ecdysone-inducible expression system 195 effectors, regulation by receptor-G protein fusion proteins 132–4 electroblotting see immunoblotting electroporation permeabilization of membranes 364 transfection method 233 ELISA chloramphenicol acetyltransferase (CAT) assay 207-10 mitogen-activated protein (MAP) kinase assay 356 testing antiserum specificity 87-8, 90 epitope tagging, G protein-coupled receptors (GPCRs) 60-1 extracellular signal-regulated kinases (ERKs) activation 339-40,342 antibodies 346, 348-9 assay of upstream activators 358 chemical inhibition of signalling pathways 358-9 immunocytochemical localization 357-8 immunoprecipitation 354-5 phosphorylation 339 properties 337-8 purification with phenyl-Sepharose 350–1 substrates 339–40 see also mitogen-activated protein (MAP) kinases 'FLAG' epitope purification of P2 adrenergic receptor (p32AR) 3 390

Index purification of G protein-coupled receptors (GPCRs) 43-5, 61 fluorescence resonance energy transfer (FRET) measurement of calcium ions 216-17 study of protease activity 216-17 using green fluorescent protein (GFP) 173, 216-17 Fluorometric Imaging Plate Reader (FLIPR) calcium-ion measurement 214 fluorimetric assay 205 fusion proteins construction 108-10 expression 110-11,119-21 functional assay 113-17 interactions with A1 adenosine receptor 128-32 measurement of agonist efficacy 121-3 measurement of GTPase activity 117, 119-21 purification 163-4 receptor-G protein 103-37 regulation of effectors 132-4 study of role of N-terminal acylation 124-8 G0 cells, extracellular signal-regulated kinase (ERKs) activity 340 p-galactosidase adenylyl cyclase assay 246 assay 210-12 reporter gene 173,195 sensitivity 210-11 gel electrophoresis see SDS-PAGE p-glucuronidase, adenylyl cyclase assay 246 glutathione S-transferase (GST) fusion proteins, purification 163-4 glycine, myristoylation 35 GPCR phosphorylation determination by immunoprecipitation 60-4 in membranes 64-6 GPCR signal transduction, use of reporter genes 178-89 G-protein a subunits antisera 87-90 classification 103—4 downregulation 92-3 fusion proteins 103-37 immunoprecipitation 92-5 N-terminal acylation 124-8 quantification 95-8 see also G proteins G protein-coupled receptor kinases (GRKs) 59 G protein-coupled receptors (GPCRs) activation 1-2,32,321-3 activation of MAP kinases 184-9

activation of sphingosine kinase 324 affinity purification 40-3 antigenicity 61 desensitization 59 EDG family 322-3 effects of agonist binding 1,32,59 epitope tagging 60-1 fusion proteins 103-37 immunoprecipitation 40, 43-5,60-4 palmitoylation 35-56 and phosphoinositide 3-kinases (PI3Ks) 283, 290 phosphorylation 59-72 purification 2-6, 40-5, 60-4 regulation of adenylyl cyclase 241 solubilization 62-3 structure 1 study of signal transduction 171-221 see also GPCR phosphorylation; GPCR signal transduction G proteins activation 100-1, 103, 105 activation of phosphoinositide 3-kinases (PI3Ks) 284 ADP-ribosylation 73-4, 77-81, 87,105-6, 325 expression in baculovirus-infected cells 166-7 extraction from cells 75-7 fusion proteins 103-37 GTPase activity 98-100, 103, 113-14, 117, 119-21 identification 73-101 interactions with A1 adenosine receptor 128-32 mode of action 103 myristoylation 124-5 receptor-stimulated binding of GTP 100-1, 103 regulation of adenylyl cyclase 223–1,241 SDS-PAGE 81-6 structure 73-4, 103 substrate for toxins 73–1, 87, 99, 105-6, 124-6 toxin-insensitive 74, 78, 105-6, 124-6 see also G-protein a subunits green fluorescent protein (GFP) applications 215-17 assay 215-17 localization of intracellular calcium 362 long half-life 216 see also blue fluorescent protein (BFP) GTP, receptor-stimulated binding to G proteins 100–1 GTPase activity of G proteins 98-100, 103, 113-14, 117,119-21 measurement of activity 115-17 391

Index guanine nucleotide-binding proteins see G proteins

isoproterenol, binding to (32 adrenergic receptor (P2AR) 8-9, 32

hepatocytes, isolation from rat 362-4 hexahistidine see polyhistidine high-pressure liquid chromatography see HPLC hormone response elements see transcriptionfactor binding sites HPLC analysis of phosphoinositides 291, 295-8 anion-exchange columns 275, 277 calibration 277-9 inositol phospholipids 271-80 sample injection 277-8

p-lactamase, assay 212-13 lipids, extraction from Saccharomyces cerevisiae 263-5 Lipofectamine, transfection of cells 111-12, 191-4 lipofection, methods 111-12, 191-4, 233 luciferase applications 233-48 assay 201-4 dual assays 203-5, 243 firefly 201, 239-43 LuFLIPRase dual assay 205 Renilla 203, 246 reporter gene 175-7, 182, 186-8 ,195, 197, 239-3 lysine, fluorescent labelling 15-16 lysosphingolipids, activation of G proteincoupled receptors (GPCRs) 322

immunoblotting G proteins 91-2, 96-8 mitogen-activated protein (MAP) kinases 343-4, 351-2 immunoprecipitation determination of GPCR phosphorylation 60-4 extracellular signal-regulated kinases (ERKs) 354-5 G-protein a subunits 92-5 G protein-coupled receptors (GPCRs) 40, 43-5, 60-4 mitogen-activated protein (MAP) kinases 353-6 phosphoinositide 3-kinases (PI3Ks) 287-90 inositol 1,4,5-triphosphate (Ins(l,4,5)P3) and desensitization of PLC-coupled receptors 66-71 effects on intracellular calcium 361-2, 365, 367-8, 378-82 extraction from cells 68 preparation of binding protein 69-70 radio-receptor assay 70-1 receptor activation 381-2 receptors 361-2 second messenger 105 inositol glycerolipids, deacylation 271-4 inositol glycerophospholipids analytical methods 257-9 biosynthesis 256-7 nomenclature 256 radioactive labelling 259-63 structures 256, 258-9 inositol lipids extraction from yeasts 263-5 standards for HPLC 278-9 inositol phospholipids, identification 266-71 ion-exchange chromatography isolation of tritiated cyclic AMP 236-9 purification of recombinant proteins 162-3

mammalian cells, expression of G proteincoupled receptors (GPCRs) 38-9 mammalian tissue, phospholipase D assay 317 MAP kinase kinase (MEK), activation of MAP kinase 339 MEK see MAP kinase kinase membranes see plasma membranes methane thiosulfonate (MTS) derivatives, use in substituted-cysteine accessibility method (SCAM) 24-33 methylamine, deacylation of inositol glycerolipids 272-4, 292 microtubule-associated protein 2 (MAP-2) kinase see mitogen-activated protein (MAP) kinase mitogen-activated protein (MAP) kinases activation 184-9, 339-43 antibodies 346 assay 343-56 assay of upstream activators 358 chemical inhibition of signalling pathways 358-9 chromatographic separation 351-2 extraction from cells 340-2 extraction from tissues 342-3 gel electrophoresis 356 immunoblotting 351-2 immunocytochemical localization 356-8 immunoprecipitation 353-6 phosphorylation 339 phosphorylation of transcription factors 185-6 phospho-specific antibodies 348-9

392

Index properties 337-8 substrates 339-40 mitogens, activation of extracellular signalregulated kinases (ERKs) 340 mobility-shift assay, activation of mitogenactivated protein (MAP) kinases 344-7 myelin basic protein (MBP), substrate for mitogen-activated protein (MAP) kinases 339-10, 349, 355 myristoylation of glycine residues 35 G proteins 124-5 nickel-agarose, binding of polyhistidine 43-5, 164-6 nitrobenzdioxazol (IANBD) cysteine specificity 6-8,12-15 fluorescence and solvent polarity 6-8 labelling p2 adrenergic receptor (p2AR) 6-15 nitrogen cavitation, lysis of cells 161-2 S-opioid receptor, palmitoylation 51-4 p38 MAP kinases activation 339 antibodies 346 assay 356 chemical inhibition of signalling pathways 358-9 phosphorylation 339 properties 337-8 substrates 339-40 see also mitogen-activated protein (MAP) kinase palmitoylation P2 adrenergic receptor (p2AR) 41-3 autocatalytic 54-6 cell-surface receptors 51-4 chromatographic analysis 47 detection 36-47 determination of half-life 49-50 G protein-coupled receptors (GPCRs) 35-56 kinetics 47-51 mechanism 54 nature of chemical linkage 45-7 8-opioid receptor 51-4 regulation by agonists 50-1 rhodopsin 35-6, 54 synthetic peptides 54-6 palmitoyl-CoA, synthesis 54-5 pertussis toxin (PTX) ADP-ribosylation of G proteins 73—4, 77-81, 87, 105-6, 325

disruption of signalling pathways 325 inhibition of sphingosine-1-phosphate (SPP) 325 use in reporter-gene assays 196-7 phosphoinositide 3-kinases (PI3Ks) activation 283-4, 287-98 activation of protein kinase B 284 analysis of lipid products 290-8 catalytic site inhibitors 285 and cell-surface receptors 283 constitutively active alleles 286-7 dominant negative alleles 285-6 functions 283-1 and G protein-coupled receptors (GPCRs) 283, 290 identifying a role in signalling pathways 285-7 immunoprecipitation 287-90 inhibition 285 receptor mutants 286 phosphoinositides [3H]inositol labelling 293-4 32 P-labelling 291-3 analysis 291, 294-8 deacylation 292 mass analysis 297-8 phospholipase C-coupled receptors, desensitization 66-71 phospholipase D activation 301, 303-4 assay 302-17 fluorescent substrate 311-13 mammalian 301, 309-12, 315, 317 molecular structure 301 plants 301, 315, 317 and signal transduction 301 substrates 301-2 transphosphatidylation 301-2 yeast 301, 315, 317 phospholipids extraction from cells 305 labelling with tritiated fatty acids 302-3 thin-layer chromatography 306-7 phosphorimaging, of SDS-PAGE gels 86 phosphorylation and agonist stimulation of p-adrenergic receptor 59 desensitization of receptors 59 G protein-coupled receptors (GPCRs) 59-72 mitogen-activated protein (MAP) kinases 339 phosphotyrosine, antibodies 343 plants, phospholipase D 301, 315, 317 plasma membranes GPCR phosphorylation 64-6 permeabilization 364 preparation 65, 112-13, 311-12

393

Index plasma membranes (continued) source of G proteins 75-7 stimulation of receptor phosphorylation 65-6 plasmids, transfer 145-6 polyhistidine binding to nickel-coupled resins 43-5 purification of G protein-coupled receptors (GPCRs) 43-5 purification of recombinant proteins 164-7 proline-directed protein kinases see mitogenactivated protein (MAP) kinases promoters CRE 172, 174-80 natural 174 reporter genes 173-7 synthetic 174-7 protease activity, study using fluorescence resonance energy transfer 216-17 protein kinase A assay 243-4 stimulation by cyclic AMP 248-9 protein kinase B, activation by phosphoinositide 3-kinases 284 protein kinase C, activation by agonist binding to GPCR 182 protein tyrosine kinases, activation of phosphoinositide 3-kinases 283-4 quantification densitometric 97-8 G-protein a subunits 95-8 immunoblotting 96-8, 346 mitogen-activated protein (MAP) kinases 346-7, 356 phospholipase D activity 307-9, 314-16 radioactive labelling 98 receptor activation conformational change 32-3 constitutively active mutants 32-3 receptor-binding sites, probing with substituted-cysteine accessibility method (SCAM) 19-33 receptors, regulation of palmitoylation 50-1 recombinant proteins purification 162-6 release from cells 160-2 reporter-enzyme assays chloramphenicol acetyltransferase (CAT) 207-10, 244-6 dual luciferase 203-4, 243 firefly luciferase 201-2, 240-3 p-galactosidase 210-12 p-lactamase 212-13 LuFLIPRase 205

Renilla luciferase 203-4 secreted placental alkaline phosphatase (SEAP) 205-7 reporter-gene assays agonists 198-201 antagonists 199 cell preparation 197-201 choice of cell line 189-90 choice of reporter enzyme 195 constitutive receptor signalling 200-1 culture media 192-5 drug exposure 196 dual 243 dynamic range 202, 243 expression protocols 190-5 optimization 195-7 sensitivity 202, 211, 243 signal-to-noise ratio 196 use of inhibitors 196-7 reporter genes activation of MAP kinase 185-9 aequorin 173, 182-1 alkaline phosphatase (SEAP) 173, 180-1, 195, 197, 205-7, 246 assay of cyclic AMP 239-13 chloramphenicol acetyltransferase (CAT) 195, 244-6 choice of 195 construction 173-7 definition 171-3 (p-galactosidase 173, 195 (p-glucuronidase 246 green fluorescent protein (GFP) 173 p-lactamase 195, 197 luciferase 175-7, 182, 186-8, 195, 197, 239-43 promoters 173-7 study of signal transduction 171-221 transcription-factor binding sites 174-6 use for screening 173 reporter proteins 177-8 aequorin 173, 177-8, 182-1, 214-15, 361-2 green fluorescent protein (GFP) 173, 178, 215-17, 362 rhodopsin palmitoylation 35-6, 54 photoactivation 2 Saccharomyces cerevisiae advantages in phosphoinositide research 255-6 culture media 259-61 extraction of lipids 263-5 saponin, permeabilization of membranes 364-7 Schizosaccharomyces pombe, extraction of inositol lipids 264-5

394

Index scintillation proximity assay (SPA) 114-15 SDS-PAGE autoradiography of gels 86 gel composition 83 of G proteins 80-6 phosphorimaging of gels 86 phosphorylation of MAP kinases 344-7 see also gel electrophoresis SEAP see secreted placental alkaline phosphatase (SEAP) secondary structure, and substituted-cysteine accessibility method (SCAM) 31-2 second messengers 105, 361-2 secreted placental alkaline phosphatase (SEAP) colorimetric assay 205-7 heat stability 206 reporter gene 173, 180-1, 195, 197 Sf9 cells cell culture 4,38, 140-4, 155-6, 158-60, 167-8 culture media 38, 141, 143-4, 168 expression of G protein-coupled receptors (GPCRs) 37-8 expression of multimeric proteins 166-7 freezing 143 infection with baculovirus 4, 157-60 lysis 160-2 thawing 141 shuttle vector, construction of recombinant baculovirus 152-5 signalling pathways disruption by pertussis toxin 325 sphingosine kinase 325-7, 335 signal transduction ceramide 321-2 G protein-coupled receptors (GPCRs) 171-221 phospholipase D 301 sphingosine kinase 321-36 sphingosine-1-phosphate (SPP) 321, 323, 335 study using reporter genes 171-221 site-directed mutagenesis, identification of binding sites 20, 24 sphingosine kinase assay 328-35 calcium ion mobilization 324 inhibition 326-7 production of sphingosine-1-phosphate (SPP) 322-4 regulation 327-8 signal transduction 321-36 stimulation by G protein-coupled receptors (GPCRs) 324 sphingosine-1-phosphate (SPP) activation of G protein-coupled receptors (GPCRs) 321-3

caged 325 GPCR-independent action inhibition by pertussis toxin 325 inhibition of degradation 330 intracellular activity 323—4 metabolism 321-2 microinjection into cells 325 receptors 321-3, 325 signal transduction 321, 323, 335 stimulation by G protein-coupled receptors (GPCRs) 324 thin-layer chromatography (TLC) 329-30 sphingosine-1-phosphate lyase, inhibition 330 stress-activated protein kinases (SAPKs) see mitogen-activated protein (MAP) kinases substituted-cysteine accessibility method (SCAM), probing receptor-binding sites 19-33 superfusion, rapid advantages 369-70 apparatus 370-8 calcium release from intracellular stores 368-70, 378-82 cellular response to inositol 1,4,5triphosphate (Ins(l,4,5)P3) 368-70, 378-S2 kinetics 373-1, 377-82 signal-to-noise ratio 369 thin-layer chromatography (TLC) chloramphenicol acetyltransferase (CAT) assay 207-10 inositol phospholipids 266-71 lipids bound to receptors 47 phosphoinositides 291, 294-5 phospholipase D assay 312-14 phospholipids 306-7 sphingosine-1-phosphate (SPP) 329-30 transcription-factor binding sites, reporter genes 174-6 transcription factors CREB (CRE-binding protein) 174-5, 178-9 phosphorylation by MAP kinases 185-6 transfection adenovirus-mediated 228-32 calcium phosphate-mediated 233 choice of vector 111 DEAE-dextran/chloroquine mediated 226-8 efficiency 228 electroporation 233 using Lipofectamine 111-12, 191-4 liposome-mediated 233 methods 22-3, 111-12, 191-3, 226-33

395

Index transphosphatidylation, by phospholipase 301-2 Trypan blue, testing viability of cells 366 viability of cells, testing 366 viral gene activity, time course 144—5

D

wortmannin, use in reporter-gene assays 196-7 yeast phospholipase D 301, 315, 317 see also Saccharomyces cerevisiae; Schizosaccharomyces pombe

Western blotting, mitogen-activated protein (MAP) kinases 343-4, 356

396