G Protein-Coupled Receptors: Essential Methods

G Protein-Coupled Receptors Essential Methods Editors David R. Poyner School of Life and Health Sciences, University o...

Author: David Poyner | Mark Wheatley

54 downloads 1803 Views 2MB 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

G Protein-Coupled Receptors Essential Methods Editors

David R. Poyner School of Life and Health Sciences, University of Aston, Birmingham, UK

Mark Wheatley School of Biosciences, Birmingham University, Birmingham, UK

A John Wiley & Sons, Ltd., Publication

G Protein-Coupled Receptors

G Protein-Coupled Receptors Essential Methods Editors

David R. Poyner School of Life and Health Sciences, University of Aston, Birmingham, UK

Mark Wheatley School of Biosciences, Birmingham University, Birmingham, UK

A John Wiley & Sons, Ltd., Publication

This edition ﬁrst published 2010,  2010 John Wiley & Sons, Ltd Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley’s global Scientiﬁc, Technical and Medical business with Blackwell Publishing. Registered ofﬁce: John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Other Editorial Ofﬁces: 9600 Garsington Road, Oxford, OX4 2DQ, UK 111 River Street, Hoboken, NJ 07030-5774, USA For details of our global editorial ofﬁces, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell The right of the author to be identiﬁed as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. 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, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. The contents of this work are intended to further general scientiﬁc research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting a speciﬁc method, diagnosis, or treatment by physicians for any particular patient. The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and speciﬁcally disclaim all warranties, including without limitation any implied warranties of ﬁtness for a particular purpose. In view of ongoing research, equipment modiﬁcations, changes in governmental regulations, and the constant ﬂow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. Readers should consult with a specialist where appropriate. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising herefrom. Library of Congress Cataloguing-in-Publication Data G protein-coupled receptors : essential methods / edited by David R. Poyner and Mark Wheatley. p. cm. Includes bibliographical references and index. ISBN 978-0-470-74914-2 (cloth : alk. paper) 1. Cell receptors. 2. Ion channels. I. Poyner, David. II. Wheatley, Mark. QH603.C43G67 2009 612 .01575 – dc22 2009031411 ISBN: 978-0-470-74914-2 A catalogue record for this book is available from the British Library. Typeset in 10/12 Times by Laserwords Private Limited, Chennai, India Printed in Singapore by Markono Pte. Ltd First impression– 2010

Contents Preface Contributors 1 Measurement of Ligand–G Protein-coupled Receptor Interactions Katie Leach, Celine Valant, Patrick M. Sexton and Arthur Christopoulos 1.1 Introduction 1.2 Methods and approaches References

2 Second Messenger Assays for G Protein-coupled Receptors: cAMP, Ca2+ , Inositol Phosphates, ERK1/2 Karen J. Gregory, Patrick M. Sexton, Arthur Christopoulos and Caroline A. Hick 2.1 Introduction 2.2 Methods and approaches 2.3 Troubleshooting References

3 Use of the [35 S]GTPγS Binding Assay to Determine Ligand Efﬁcacy at G Protein-coupled Receptors Elodie Kara and Philip G. Strange 3.1 Introduction 3.2 Methods and approaches 3.3 Troubleshooting Acknowledgements References

4 Quantitative Imaging of Receptor Trafﬁcking Andy R. James, Takeo Awaji, F. Anne Stephenson and Nicholas A. Hartell 4.1 Introduction 4.2 Methods and approaches 4.3 Troubleshooting References

ix xi 1 1 5 27

31

31 33 49 50

53 53 54 66 67 67

69 69 70 80 82

vi

CONTENTS

5 Production of Recombinant G Protein-coupled Receptor in Yeast for Structural and Functional Analysis Richard A.J. Darby, Mohammed Jamshad, Ljuban Grgic, William J. Holmes and Roslyn M. Bill 5.1 Introduction 5.2 Methods and approaches 5.3 Troubleshooting References

6 Monitoring GPCR–Protein Complexes Using Bioluminescence Resonance Energy Transfer Werner C. Jaeger, Kevin D.G. Pﬂeger and Karin A. Eidne 6.1 Introduction 6.2 Methods and approaches 6.3 Troubleshooting References

7 Using Intramolecular Fluorescence Resonance Energy Transfer to Study Receptor Conformation Cornelius Krasel and Carsten Hoffmann 7.1 Introduction 7.2 Methods and approaches 7.3 Troubleshooting References

8 A Disulﬁde Cross-linking Strategy Useful for Studying Ligand-induced Structural Changes in GPCRs Jian Hua Li, Stuart D.C. Ward, Sung-Jun Han, Fadi F. Hamdan and J¨urgen Wess 8.1 Introduction 8.2 Methods and approaches 8.3 General considerations, caveats and troubleshooting Acknowledgements References

9 Use of Fluorescence Correlation Spectroscopy to Study the Diffusion of G Protein-coupled Receptors Stephen J. Briddon, Jonathan A. Hern and Stephen J. Hill 9.1 Introduction 9.2 Methods and approaches 9.3 Troubleshooting References

85

85 86 106 107

111 111 114 128 128

133 133 136 143 144

147

147 149 162 164 164

169 169 170 190 191

CONTENTS

10 Identiﬁcation and Analysis of GPCR Phosphorylation Kok Choi Kong, Sharad C. Mistry and Andrew B. Tobin 10.1 Introduction 10.2 Methods Acknowledgements References

11 Measurement and Visualization of G Protein-coupled Receptor Trafﬁcking by Enzyme-linked Immunosorbent Assay and Immunoﬂuorescence Stuart J. Mundell, Shaista P. Nisar and Eamonn Kelly 11.1 Introduction 11.2 Methods and approaches 11.3 Troubleshooting References

12 Substituted Cysteine Accessibility Method (SCAM) George Liapakis and Jonathan A. Javitch 12.1 Introduction 12.2 Methods and approaches 12.3 Troubleshooting References

13 Homology Modelling of G Protein-coupled Receptors John Simms 13.1 Introduction 13.2 Methods and approaches 13.3 Troubleshooting 13.4 Automated methods for generating models of GPCRs References

Appendix Site-directed Mutagenesis and Chimeras

vii

197 197 198 213 213

215 215 217 226 227

229 229 230 247 248

251 251 252 269 269 270

275

Alex Conner, Mark Wheatley and David R. Poyner A.1 Introduction A.2 Why mutagenesis? A.3 Troubleshooting A.4 Conclusion References

Index

275 275 285 285 286

289

Preface

This book describes a number of techniques relating to research on G protein-coupled receptors (GPCRs), written by a number of leading international authorities. In line with the rest of the essential methods series, each chapter contains an overview of the method and this is followed by a series of detailed protocols, providing a bench-side guide. The literature on GPCRs is vast and the pace of investigation shows no sign of slackening. Given that these receptors are not only the largest protein family in the human genome but are also the single biggest target for therapeutic agents, this level of interest is not surprising. For any editor, this poses problems; the range of techniques that can be applied to GPCRs is vast and continues to grow, so it is impossible to cover them all in a single volume. In the case of this book, the techniques that are covered have been selected simply because we think that they will prove to be useful tools in future research and will contribute to increasing our understanding of GPCRs. In addition, they are of interest to the editors. Doubtless, readers wishing to ﬁnd descriptions of additional methods will be able to ﬁnd these elsewhere in the GPCR literature. Some of the techniques covered in this volume are very well known, such as mutagenesis or measurement of second messengers. We have included these because they are so fundamental to GPCR research and so might be of interest to the newcomer to the ﬁeld. Some of the other techniques presented might not be so widely employed, but they have proved their worth in numerous laboratories. Chapters 1 and 2, by Leach et al . and Gregory et al . respectively, review radioligand binding techniques (including detection of allosteric modulators) and measurements of a range of second messengers activated by GPCRs. Measurement of GTPγ S binding as an index of receptor activation is dealt with by Kara and Strange in Chapter 3. Quantitative GPCR imaging is considered in the context of receptor trafﬁcking by James et al . in Chapter 4. In Chapter 5, Darby et al . describe methods for overexpressing GPCRs, concentrating on Pichia pastoris as a host. In Chapter 6, Jaeger et al . consider how bioluminescence resonance energy transfer can be used to look at GPCR complexes, whereas Krasel and Hoffmann describe in Chapter 7 how ﬂuorescence resonance energy transfer can be used to measure receptor conformation. Chapters 8 and 12, by Li et al . and Liapakis and Javitich respectively, review the utility of engineered cysteines for artiﬁcial intramolecular disulﬁde bonds and cysteine scanning accessibility mutagenesis respectively. The new technique of ﬂuorescence correlation spectroscopy

x

PREFACE

is described by Briddon et al . in Chapter 9. Mundell et al . (Chapter 11) and Kong et al . (Chapter 10) consider receptor regulation; how this can be quantiﬁed and how GPCR phosphorylation can be measured. Simms reviews techniques for receptor modelling in Chapter 13 and the Appendix (by Conner et al .) considers methods for mutagenesis. We hope this volume will be useful to investigators in GPCR research. David R. Poyner Mark Wheatley

Contributors Takeo Awaji

Richard A.J. Darby

Department of Physiology, Tokyo Women’s Medical University School of Medicine, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162–8666, Japan

Pharmaceutical & Biological Sciences, School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK

Roslyn M. Bill Pharmaceutical & Biological Sciences, School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK

Karin A. Eidne Laboratory for Molecular Endocrinology – GPCRs, WAIMR and UWA Centre for Medical Research, QEII Medical Centre, Nedlands, Perth, Western Australia 6009, Australia

Karen J. Gregory Stephen J. Briddon Institute of Cell Signalling, School of Biomedical Sciences, University of Nottingham, Nottingham NG7 2UH, UK

Arthur Christopoulos Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton 3800, Australia

Alex Conner School of Medicine, Warwick University, Coventry CV4 7AL, UK

Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton 3800, Australia

Ljuban Grgic Pharmaceutical & Biological Sciences, School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK

Fadi F. Hamdan Molecular Signaling Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 20892, USA

xii

CONTRIBUTORS

Sung-Jun Han

Werner C. Jaeger

Molecular Signaling Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 20892, USA

Laboratory for Molecular Endocrinology – GPCRs, WAIMR and UWA Centre for Medical Research, QEII Medical Centre, Nedlands, Perth, Western Australia 6009, Australia

Nicholas A. Hartell Department of Cell Physiology and Pharmacology, University of Leicester, Leicester LE1 9HN, UK

Jonathan A. Hern Institute of Cell Signalling, School of Biomedical Sciences, University of Nottingham, Nottingham NG7 2UH, UK

Andy R. James Department of Pharmacology, The School of Pharmacy, University of London, London WC1N 1AX, UK

Mohammed Jamshad Pharmaceutical & Biological Sciences, School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK

Caroline A. Hick Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton 3800, Australia

Stephen J. Hill Institute of Cell Signalling, School of Biomedical Sciences, University of Nottingham, Nottingham NG7 2UH, UK

Jonathan A. Javitch Center for Molecular Recognition, Columbia University, P&S 11-401, Box 7, 630 West 168th Street, New York, NY 10032, USA

Elodie Kara

Department of Pharmacology, University of W¨urzburg, 97070 W¨urzburg, Germany

School of Pharmacy, University of Reading, PO Box 228, Whiteknights, Reading RG6 6AJ, UK

William J. Holmes

Eamonn Kelly

Pharmaceutical & Biological Sciences, School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, UK

Department of Physiology & Pharmacology, University of Bristol, Medical Sciences Building, University Walk, Bristol BS8 1TD, UK

Carsten Hoffmann

CONTRIBUTORS

Kok Choi Kong

Stuart J. Mundell

Department of Cell Physiology and Pharmacology, University of Leicester, Hodgkin Building, Lancaster Road, Leicester LE1 9HN, UK

Department of Physiology & Pharmacology, University of Bristol, Medical Sciences Building, University Walk, Bristol BS8 1TD, UK

Cornelius Krasel School of Pharmacy, University of Reading, Whiteknights, Reading, Berkshire RG6 6AH, UK

Katie Leach Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton 3800, Australia

Jian Hua Li Molecular Signaling Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 20892, USA

Shaista P. Nisar Department of Physiology & Pharmacology, University of Bristol, Medical Sciences Building, University Walk, Bristol BS8 1TD, UK

Kevin D.G. Pﬂeger Laboratory for Molecular Endocrinology – GPCRs, WAIMR and UWA Centre for Medical Research, QEII Medical Centre, Nedlands, Perth, Western Australia 6009, Australia

David R. Poyner Pharmaceutical and Biological Sciences, School of Life and Health Sciences, Aston University, Birmingham B4 7ET, UK

George Liapakis Faculty of Medicine, University of Crete, Voutes, Heraklion 71003, Crete, Greece

Patrick M. Sexton

Sharad C. Mistry

John Simms

Department of Cell Physiology and Pharmacology, University of Leicester, Hodgkin Building, Lancaster Road, Leicester LE1 9HN, UK

Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton 3800, Australia

Department of Pharmacology, University of Monash, Clayton, Victoria 3800, Australia

xiii

xiv

CONTRIBUTORS

F. Anne Stephenson

Stuart D.C. Ward

Department of Pharmaceutical and Biological Chemistry, The School of Pharmacy, University of London, London WC1N 1AX, UK

Molecular Signaling Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 20892, USA

Philip G. Strange School of Pharmacy, University of Reading, PO Box 228, Whiteknights, Reading RG6 6AJ, UK

Andrew B. Tobin Department of Cell Physiology and Pharmacology, University of Leicester, Hodgkin Building, Lancaster Road, Leicester LE1 9HN, UK

Celine Valant Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton 3800, Australia

Jurgen Wess ¨ Molecular Signaling Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD 20892, USA

Mark Wheatley School of Biosciences, Birmingham University, Birmingham B15 2TT, UK

1 Measurement of Ligand–G Protein-coupled Receptor Interactions Katie Leach, Celine Valant, Patrick M. Sexton and Arthur Christopoulos Drug Discovery Biology Laboratory, Monash Institute of Pharmaceutical Sciences and Department of Pharmacology, Monash University, Parkville, 3052, Australia

1.1 Introduction 1.1.1 Ligand–receptor interactions and the law of mass action Radioligand binding assays take advantage of the ability to detect the decay of radioactive material, which can be incorporated into a ligand of choice. The interaction of such a radioligand with a receptor preparation can subsequently be determined by capturing and measuring the amount of radioactivity present. Radioligand binding assays can be used to estimate molecular parameters, such as the density of receptors present in a tissue or cellular preparation or the afﬁnity of a ligand for binding to a receptor. The simplest scheme that describes the binding of a ligand to its receptor is based on the law of mass action: Kon −− −→ A+R− ← −− AR Koff

where the ligand A binds to the receptor R to form the ligand–receptor complex AR. The rate at which the ligand binds to the receptor, expressed as the number of binding events per unit of time, is dependent on the ligand concentration, the number of

G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

2

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

unoccupied receptors and the association rate constant Kon . In contrast to an enzymatic reaction, there is no degradation of the product AR; and if this reaction is reversible, then the ligand–receptor complexes can dissociate into free receptor and ligand, which is dependent on the concentration of ligand–receptor complexes and the dissociation rate constant Koff . Although the association and dissociation rates may differ, over time an equilibrium state will arise whereby the rate at which new ligand–receptor complexes are formed will equal the rate at which ligand–receptor complexes break down. At equilibrium, the ratio of the dissociation and association rate constants provides a useful measure of the overall strength with which a ligand interacts with a receptor, the equilibrium dissociation constant Ka (although sometimes referred to as Kd ) and expressed in moles per litre. Ka is also the concentration of ligand that binds half the receptors present. Therefore, at equilibrium, the concentration of ligand–receptor complexes is governed by the total receptor density [RT ], the ligand concentration [A] and the equilibrium dissociation constant of the ligand: [AR] =

[RT ] × [A] [A] + Ka

(1.1)

where [RT ] = [R] + [AR] and Ka = Koff /Kon . Equation 1.1 is often referred to as the Hill–Langmuir binding isotherm and describes equilibrium binding of a ligand to a receptor under the law of mass action, such that a hyperbolic curve will result when binding is plotted against the molar ligand concentration. This was ﬁrst used by A.V. Hill to describe the binding of oxygen to haemoglobin [1, 2]. If ligand concentrations are expressed in logarithmic space, then a sigmoidal concentration–occupancy curve will be apparent. If binding of a ligand to a receptor at equilibrium follows a simple mechanism, where the binding of one ligand molecule is unaffected by concomitant binding events and where the ligand binds to only a single class of receptor sites, then the concentration–occupancy relationship plotted on a logarithmic scale will follow a sigmoidal curve that approximately spans from 10 to 90% occupancy over a 100-fold, or two log-unit, concentration range. The steepness of the slope of such a curve in linear space, the Hill coefﬁcient, will equal unity.

1.1.2 Competitive interactions at G protein-coupled receptors 1.1.2.1 Antagonist binding Although ligand–G protein-coupled receptor (GPCR) interactions can be quantiﬁed by observing the binding of a radiolabelled ligand to a receptor, it is sometimes more practical to measure the ability of a ﬁxed concentration of radioligand to bind to the receptor in the presence of increasing concentrations of an unlabelled ligand, to indirectly determine the interaction of the unlabelled ligand with the receptor. If we consider the binding of ligand A in the presence of a competitor B at equilibrium: Kon B

Kon A

Koff B

Koff A

← −− A + R + B − −− −→ A + BR − −− −→ ← −− AR + B

1.1 INTRODUCTION

3

then we can deﬁne the concentration of receptors bound to the radiolabelled ligand A in the presence of an unlabelled competitive ligand B, which was ﬁrst derived by Gaddum [3, 4]. A form of this relationship can be described by [AR] =

[RT ] × [A] [B] [A] + Ka 1 + Kb

(1.2)

Competitive binding data are commonly expressed as the fractional inhibition of radioligand binding in the presence of the competitor. However, the binding of some ligands does not follow the simple law of mass action and receptor occupancy is not always directly proportional to ligand concentration (see Section 1.1.2.2). Under these circumstances, the Hill slope may vary from unity and must, therefore, be empirically incorporated into any ligand binding equation to derive the steepness of the slope describing the concentration–occupancy relationship, as shown in Equation 1.3: Y =

Top − Bottom + Bottom [B]n 1+ IC50

(1.3)

where Y is radioligand binding, Top is the top asymptote of the curve equal to total binding of the radioligand in the absence of competitor B, Bottom is the bottom asymptote of the curve equal to nonspeciﬁc binding, n is the Hill coefﬁcient and IC50 is the concentration of B that inhibits 50% of radioligand binding. If the Hill coefﬁcient equals unity, then the equilibrium dissociation constant of the unlabelled ligand Kb can be determined using the Cheng–Prusoff equation [5]: Kb =

IC50 [A] 1+ Ka

(1.4)

1.1.2.2 Agonist binding Although binding of an antagonist to a receptor will often display a concentration– occupancy relationship that has a Hill coefﬁcient of unity, agonist binding to GPCRs is usually more complex. This has been well characterized in competition assays between agonists and radiolabelled antagonists, which often yield shallow curves, with Hill coefﬁcients less than unity. These shallow curves reﬂect different receptor states, for which the agonist has different afﬁnities and the (often) biphasic nature of the curves gives rise to a competition curve that spans greater than a twofold concentration range of competitor (see Figure 1.1). The addition of guanine nucleotides such as guanosine diphosphate, guanosine triphosphate (GTP), guanylylimidodiphosphate (GppNHp) and guanosine 5 -O(3-thiotriphosphate) (GTPγ S) often alters the proportion and afﬁnity of the two binding sites, demonstrating that the dispersion of agonist afﬁnity states reﬂects the formation of a ternary complex consisting of an agonist, a GPCR and a guanine

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS % Specific radioligand binding

4

100

One-site binding Two-site binding

80 60 40 20 0 −12

−10

−8 −6 −4 Log [competitor] (M)

−2

Figure 1.1 Theoretical competition binding curve simulated using a one-site binding ﬁt that follows a Hill slope of unity (solid line) in comparison with a two-site binding ﬁt that follows a shallow slope (dashed line).

nucleotide binding protein (G protein). A ternary complex model (TCM) has been proposed to explain the shallow curves observed with agonists versus radiolabelled antagonists in competition binding assays, but requires a number of assumptions that are not often met [6]. For instance, it must be assumed that the G protein is limiting so that not all of the receptors can form a complex with the G protein, enabling both G protein-coupled (high-afﬁnity) and -uncoupled (low-afﬁnity) receptor species to be observed. This is rarely observed in cellular systems used to study many ligand–GPCR interactions, where G protein levels often exceed those of receptor expression levels. More sophisticated extensions of the TCM have been developed to account for the ability of agonists to bind with higher afﬁnity to receptors that have been mutated to exert constitutive activity than to their wild-type counterparts, even in the absence of G proteins [7]. However, in general there is little advantage to using the extended TCM for routine data analysis, as the simpler TCM can adequately approximate the binding of agonists to a receptor.

1.1.3 Allosteric ligands For a number of GPCRs, there are ligands that can bind to the receptor at a site that is topographically distinct from the endogenous, or orthosteric, ligand binding site [8]. These binding sites and the ligands that bind to them are referred to as ‘allosteric’. Since allosteric ligands do not directly compete for binding with the orthosteric ligand, they have the ability to form a ternary complex in which both the orthosteric and the allosteric ligand occupy the receptor (see Figure 1.2). Binding of an allosteric ligand to a receptor may alter the receptor conformation such that binding of the orthosteric ligand is altered, and vice versa. These changes in binding are termed ‘cooperative effects’ [9]. In terms of binding, the allosteric TCM predicts that an allosteric modulator may inhibit (0 < α < 1), enhance (α > 1) or have no effect (α = 1) on the binding of an

1.2 METHODS AND APPROACHES

AR

Kb /α

Ka

5

ARB Ka /α

R

Kb

RB

Figure 1.2 A ternary complex model describing the binding of an orthosteric ligand A and an allosteric modulator B to a receptor R. Ka and Kb are the equilibrium dissociation constants of R for A and B respectively, α is the binding cooperativity between A and B (the selectivity of A for R and RB or of B for R and AR) and, therefore, determines the effect of B on the binding of A and vice versa.

orthosteric ligand. The formation of a ternary complex between the receptor, the orthosteric ligand and the allosteric modulator can be described by

[AR] + [ARB] = [A] + Ka

[RT ] × [A] [B] α[B] 1+ 1+ Kb Kb

(1.5)

Several recent reviews provide detailed information regarding our understanding of such compounds and the methods that can be used to detect and analyse allosteric interactions [8, 10, 11].

1.2 Methods and approaches 1.2.1 General considerations for radioligand binding assays 1.2.1.1 Buffers Acidic, basic or neutral buffers, consisting of pH buffering agents such as trishydroxymethylaminomethane hydrochloride (tris acid), trishydroxymethylaminomethane (tris base) and 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES), are all commonly used for radioligand binding assays, usually at concentrations of 10–50 mm. HEPES-based buffers are popular as HEPES is relatively heat stable. For some receptors, ligand binding and receptor activity are highly dependent upon pH, particularly as both the receptor and the ligand may be protonated (acidiﬁed) or deprotonated depending on the pH. Many researchers chose to perform radioligand binding assays at pH 7.4, which resembles a physiological environment. The composition of the buffer is particularly important, as certain trace metals can directly interact with GPCRs and alter their behaviour. For instance, 100 mm NaCl is commonly used to maintain a high ionic strength, yet sodium can destabilize receptor–G protein formation by interacting with a highly conserved aspartate residue in transmembrane domain 2 of Family A GPCRs [12], which can have implications for the binding of agonists and inverse agonists. Sodium may also have less speciﬁc effects that are due to the ionic strength of the buffer, and comparing the effects of sodium

6

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

and potassium may be useful to determine the potential impact of this phenomenon. Alternatively, 100 mm N -methyl-d-glucamine may be substituted for sodium to alter ionic strength whilst having no effects on receptor–G protein coupling. Additional ions may also have direct effects on ligand binding or receptor activity. For instance, magnesium is required for efﬁcient coupling between the receptor and G protein and may, therefore, enhance the proportion of high-afﬁnity agonist binding sites. Magnesium is usually present in the buffer at concentrations between 1 and 10 mm, a concentration that can also be used in functional membrane-based assays such as [35 S]GTPγ S binding assays. Calcium can directly bind to members of the Family C GPCRs to alter the binding of ligands [13] and can also activate certain proteases that may enhance degradation of the receptor. Ethyleneglycoltetraacetate (EGTA) may be added to the assay buffer to chelate calcium ions and act as a protease inhibitor, whilst ethylenediaminetetraacetate (EDTA) is useful for chelating additional trace metals, such as magnesium, to prevent ligand oxidation. For these reasons, EDTA is a useful component in the buffer used to prepare membranes for radioligand binding (see Section 1.2.2.1). Additional protease inhibitors may also be added, such as serine-, cysteine- and metallo-protease inhibitors. Finally, the buffer must also be optimal for the ligands to be used in the radioligand binding assay. Some ligands, particularly 5-hydroxytryptamine and catecholamines such as dopamine, epinephrine and norepinephrine, are particularly susceptible to oxidation, so an antioxidant such as ascorbic acid can be added to the buffer to prevent this. Other ligands, such as proteins and peptides, may stick to plastic and glass, and a protein such as bovine serum albumin (0.001–0.1%) may be required to coat these surfaces to reduce adsorption of the ligand.

1.2.1.2 Temperature and incubation time Radioligand binding assays are usually performed at 20–37 ◦ C. Although 37 ◦ C is physiologically relevant, receptors studied in membrane systems may become unstable at high temperatures. If lower temperatures are used, then care must be taken to ensure that equilibrium binding is reached. Therefore, the temperature at which the experiment is performed depends upon the stability of the receptor and ligands and the binding kinetics of the ligand. The law of thermodynamics predicts that equilibrium will be reached faster at higher temperatures, with an approximate doubling in the reaction rate with a 10 ◦ C increase in temperature. With regard to incubation time, this may vary between receptors. The binding of a ﬁxed concentration of radioligand should initially be measured at different time points to determine when equilibrium is attained (see Section 1.2.2.5). The rate at which the ligand–receptor interaction approaches equilibrium is often termed the Kobserved or Kobs , measured in units of inverse time. Kobs is dependent on the association and dissociation rate constants of the ligand and the ligand concentration, with lower ligand concentrations taking longer to reach equilibrium. Thus, binding of a radioligand to a receptor starting from time point 0 will follow Y = [AR]Eq × (1 − e−Kobs t )

(1.6)

1.2 METHODS AND APPROACHES

7

where Kobs = Kon [A] + Koff and [AR]Eq is binding once equilibrium has been reached (expressed in units of the Y axis, such as disintegrations per minute (dpm)). Low ligand concentrations should, therefore, be used when testing equilibration time. It is recommended that ﬁve times the dissociation half-life of any ligand should be allowed in order to reach approximately 97% equilibrium binding with the receptor, which is considered to be sufﬁcient.

1.2.1.3 Ligand depletion Under ideal experimental conditions only a small fraction of the total ligand added will bind to the receptor or to nonspeciﬁc sites. Thus, the free-ligand concentration throughout the assay is generally close to the total concentration of ligand added to the assay. However, if a large proportion of the ligand added is bound, either speciﬁcally or nonspeciﬁcally, then the concentration of free ligand in solution will deviate significantly from the concentration added to the assay. Generally, if less than 10% of the ligand is bound at each given ligand concentration, then ligand depletion is minimal and, therefore, is not a concern. Initial experiments can be carried out to optimize the protein content of the assay and ensure that excess radioligand does not bind to the protein added. This can be achieved by measuring the binding of a ﬁxed radioligand concentration in the presence of different concentrations of the receptor preparation and calculating the percentage of radioligand bound at each receptor concentration. If ligand depletion does appear to be a problem, then the assay format may be altered to overcome this. The assay volume may be increased, but a greater amount of ligand will also be required to obtain the same concentration whilst receptor numbers will remain constant. Alternatively, the free radioligand can be measured in each tube if a centrifugation assay is employed (see Section 1.2.1.4). Otherwise, analysis techniques that account for the differences between the added and free ligand concentration can be used (see Section 1.2.2.2).

1.2.1.4 Separation of bound from free radioligand In order to measure the amount of radioligand bound to the receptor preparation, bound radioligand must usually be separated from free radioligand. The most common separation method is vacuum ﬁltration, whereby the samples are rapidly ﬁltered, generally onto glass-ﬁbre ﬁlter paper, and washed to remove radioligand that is weakly bound to the ﬁlter or to the receptor preparation. Be aware that high concentrations of membrane can clog the ﬁlter pores, leading to slower ﬁltration and washing rates. The wash buffer should be ice-cold and washing should be rapid so as to minimize dissociation of the radioligand from the receptors. Filtration of samples can, however, lead to loss of receptors, as some can pass through the glass-ﬁbre ﬁlters. High-speed centrifugation assays may be used to minimize loss of receptor and ensure that a greater proportion of the ligand-bound receptor is collected. However, these assays may require more protein in order to pellet the membrane efﬁciently. The supernatant can then be removed and the pellet rapidly washed. Centrifugation assays are also useful when ligand depletion is a problem, as the free radioligand concentration can be determined following termination of the assay.

8

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

1.2.2 General assay protocols 1.2.2.1 Membrane preparations Radioligand binding may be studied in a cell line or tissue that endogenously expresses the receptor of interest, or a recombinant cell system in which receptor expression is induced. These assays may be performed on whole cells, on tissues or on soluble and puriﬁed receptors. However, a common approach is to prepare membrane preparations from the cell lines or tissues expressing the receptor (Protocol 1.1). Although the generation of membranes can be more expensive than using live cells, the use of membrane preparations is particularly convenient because the membranes can be frozen and stored for several months and defrosted when required. It is recommended to perform each repeat of a radioligand binding assay on a new membrane preparation made from a new tissue preparation or cell passage number.

PROTOCOL 1.1 Preparation of Membranes from Adherent Mammalian Cells Equipment and Reagents • Lifting buffer; for example, 2 mM EDTA in a phosphate-buffered saline (PBS) solution (150 mM NaCl, 16 mM Na2 HPO4 , 4 mM NaH2 PO4 ) for Chinese hamster ovary (CHO) cells, or PBS alone for less adherent cell lines, such as HEK 293 cells • HEPES-based buffer; for example, 20 mM HEPES, 10 mM EDTA, pH 7.4 • Low-EDTA HEPES-based buffer; for example, 20 mM HEPES, 0.1–1 mM EDTA, pH 7.4 • Low- and high-speed chilled centrifuges • Homogenizer.

Method 1 Grow cells to 80–90% conﬂuence in T175 ﬂasks (175 cm2 ). Generally, one T175 ﬂask will provide sufﬁcient membrane for approximately 50 assay tubes using 15 µg of protein per tube. Remove media and add 10 ml warm lifting buffera to each ﬂask. For cell lines that are particularly adherent, an initial wash with calcium-free PBS may be required to remove any remaining media. For cell lines that require lifting buffer containing EDTA, incubate for 2–5 min at 37 ◦ C. Gently tap the ﬂasks to detach cells and collect the cell suspension in appropriate tubes. Wash each ﬂask with 10 ml PBS and collect the wash. 2 Centrifuge the cell suspension at approximately 200 g (approximately 1200 rpm in most bench-top centrifuges) for 10 min and resuspend the cell pellet in 2–3 ml HEPES- or tris-based bufferb per T175 ﬂask, or an appropriate volume to ensure efﬁcient homogenization (step 3). 3 All subsequent steps should be performed at 4 ◦ C to reduce activation of proteases. Homogenize the cell solution using a homogenizer. Perform three 5–10 s bursts at top speed (∼20 000 rpm) with cooling on ice in between each burst.

1.2 METHODS AND APPROACHES

9

4 Centrifuge the resulting cell lysate at 600 g (approximately 1700 rpm in most bench-top centrifuges) for 10 min to separate the nuclear fraction and additional cell debris. If a large pellet is obtained, resuspend the pellet and repeat steps 3 and 4, combining the supernatant obtained following each centrifugation step. 5 Transfer the remaining suspension to new tubes and centrifuge at approximately 40 000 g or higher for 1 h. Resuspend the resulting protein pellet in a HEPES- or tris-based buffer,c which can be the buffer that will be used for subsequent radioligand binding assays, or a predominantly HEPES or tris-based buffer with a low concentration (0.1–1 mM) of a chelating agent such as EDTA or EGTA. Resuspend in approximately 0.5–2 ml buffer per T175 ﬂask, depending on the expression level of the receptor and subsequently the amount of protein required for each assay. 6 Homogenize the membrane suspension brieﬂy and dispense into aliquots of a suitable volume. Membranes can be stored at −80 ◦ C, generally for up to 12 months, and the protein content of the preparation can be determined using an appropriate method such as that of Lowry [14] or Bradford [15].

Notes a It

is not recommended to use trypsin for harvesting cells, as receptors may be hydrolysed.

b The

HEPES- or tris-based buffer should contain 1–10 mM EDTA, EGTA or both to reduce proteolysis of the receptor of interest following homogenization steps. c Ten

conﬂuent T175 ﬂasks should generate approximately 10 ml of protein at a concentration of 1–2 mg/ml.

1.2.2.2 Saturation binding assays Saturation binding assays are used to determine the binding of different radioligand concentrations to a receptor at equilibrium directly. These assays can be used to derive direct measurements of the total receptor number or density present in the system under investigation and to determine the afﬁnity of the radioligand for the receptor. However, the radioligand will not only bind speciﬁcally to the receptor of interest, but will also bind to additional sites within a membrane or cell preparation or to the tubes used to perform the radioligand binding assay. Therefore, nonspeciﬁc binding must be determined in parallel using a high concentration of a competing ligand to displace each radioligand concentration from the receptor. Where possible, the competitive ligand used to deﬁne nonspeciﬁc binding should not be an unlabelled form of the radioligand, as both compounds will compete for the same nonspeciﬁc binding sites. Ideally, at least 100–1000 times the Ka of the competitive ligand used to deﬁne nonspeciﬁc binding should be used to ensure full receptor occupancy. Protocols 1.2–1.6 for radioligand binding assays describe assays performed in 0.5–1 ml volumes, but are applicable to smaller volume assays that can be scaled down. Traditionally, speciﬁc binding is determined by subtracting nonspeciﬁc binding from total binding and the data are analysed using Equation 1.7 (see Figure 1.3). However,

10

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS 6000

Total Binding Nonspecific Binding Specific binding

DPM

4000

2000

0 0

2

4

6

8

10

[3H]NMS (nM)

Figure 1.3 Saturation radioligand binding. Binding of the radioligand, [3 H]N-methylscopolamine ([3 H]NMS), to the M4 muscarinic acetylcholine (ACh) receptor stably expressed in FlpIn CHO cells, where nonspeciﬁc binding was determined in the presence of 10 µm atropine. Membranes were incubated with [3 H]NMS for 1 h at 37 ◦ C before the assay was terminated as described in the methods section.

with the development of more sophisticated computer software, it is now possible to ﬁt the data globally to a model of receptor and nonspeciﬁc site occupancy, shown in Equation 1.8. If data are ﬁtted globally, then this means that a family of curves is ﬁtted to a particular model, rather than just a single curve. Parameters common to both datasets can be shared between them, enabling their determination from the relationship between all curves (a comprehensive explanation of global ﬁtting is provided in [16]). The radioligand–receptor complex concentration [AR] that forms at equilibrium at each radioligand concentration is reﬂected by the speciﬁc binding of the radioligand, expressed on the y-axis. We usually call the total number of receptors that are deﬁned by our radioligand as the Bmax , so speciﬁc binding of each radioligand concentration in a saturation binding assay is deﬁned by the Hill–Langmuir occupancy equation using the terms shown in Equation 1.7. Note the relationship to Equation 1.1. Y =

Bmax × [A] [A] + Ka

(1.7)

Nonspeciﬁc binding is generally linearly proportional to the radioligand concentration and, therefore, is described by the equation for a straight line. As the total binding TB measured in the experiments represents both speciﬁc and nonspeciﬁc binding, radioligand binding, expressed on the y-axis, to one class of binding sites is deﬁned by TB =

Bmax × [A] + NS × [A] [A] + Ka

(1.8)

where NS is the nonspeciﬁc binding. In the presence of ligand depletion, binding is deﬁned by TB =

Bmax × ([A]T − TB) + ([A]T − [TB]) × NS ([A]T − TB) + Ka

(1.9)

1.2 METHODS AND APPROACHES

11

where [A]T is the total radioligand concentration added. As TB appears on both sides of Equation 1.9 (implicit equation), it cannot be entered into most nonlinear regression programs. The equation can, however, be rearranged into a quadratic equation: √ −b + b2 − 4ac (1.10) TB = 2a where a = −1 − NS b = [A]T [2(NS + 1)] + Ka (NS + 1) + Bmax and c = −[A]T [NS(Ka + [A]T )] + Bmax If total and nonspeciﬁc binding data are ﬁtted globally, then nonspeciﬁc binding should be shared between datasets so that the fraction of total binding that is nonspeciﬁc at each radioligand concentration can be determined. This type of analysis will derive the Ka and the maximum level of binding of the radioligand without the need to subtract nonspeciﬁc from total binding. However, the raw data values that are derived from a saturation radioligand binding assay are usually more useful if converted to the amount of radioligand bound to our receptor preparation. We usually deﬁne the total number of binding sites in reference to the amount of protein or cells. For instance, 15 µg protein were added into each tube of the saturation assay shown in Figure 1.3; therefore, 1.7 pmol of receptor is expressed per milligram of protein.

PROTOCOL 1.2 Saturation Binding Assays Equipment and Reagents • Membrane preparation (Protocol 1.1) • Binding buffera • 5 ml polypropyleneb assay tubes (Techno-plas) • Radioligand • Competitive antagonist • Wash bufferc • Water bath • Vacuum harvester (Brandel) • Glass-ﬁbre ﬁlter paperd (Whatman) • Liquid scintillation cocktaile (PerkinElmer) • Liquid scintillation counter (e.g. Packard Tri-Carb LS counter).

Method 1 Dilute the receptor preparation, radioligand and competitive antagonist in assay buffer to 10× the ﬁnal concentrations required in the assay, so that each can be diluted by this factor when added to the ﬁnal assay mix.

12

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

2 In an appropriate volume of binding buffer (usually between 0.5 and 1 ml), prepare one set of assay tubes containing increasing concentrations of radioligand and one identical set of tubes that also contain a saturating concentration of a competitive ligand to deﬁne nonspeciﬁc binding (see Table 1.1).

Table 1.1 Tube set-up used to determine saturation binding of [3 H]NMS to the M4 muscarinic ACh receptor using a 500 µl ﬁnal assay volume, where nonspeciﬁc binding is determined in the presence of 10 µM atropine. Final [[3 H]NMS] (M)

[3 H]NMS (µl)

100 µM atropine (10 µM ﬁnal concentration) (µl)

Buffer (µl)

Membrane (µl)

0

450

50

0

400

50

0

290

50

0

400

50

0

290

50

0

400

50

0

290

50

0

400

50

50

400

50

50

350

50

50

240

50

50

350

50

50

240

50

50

350

50

50

240

50

50

350

50

Total binding 0 1 × 10

0 −11 −11

3.16 × 10

−10

50 (of 1 × 10

−10

160 (of 1 × 10

M)

−10

−9

M)

1 × 10

50 (of 1 × 10

3.16 × 10−10

160 (of 1 × 10−9 M)

−9

50 (of 1 × 10

1 × 10

3.16 × 10

−9

−8

160 (of 1 × 10 50 (of 1 × 10

1 × 10

−8

M)

M)

−8

−7

M)

M)

Nonspeciﬁc binding 0

0 −11

1 × 10

−11

50 (of 1 × 10

−10

M)

−10

3.16 × 10

160 (of 1 × 10

1 × 10−10

50 (of 1 × 10−9 M)

3.16 × 10

−10

−9

50 (of 1 × 10

1 × 10

3.16 × 10 −8

1 × 10

160 (of 1 × 10

−9

−8

160 (of 1 × 10 50 (of 1 × 10

−9

M)

M)

−8

−7

M)

M)

M)

3 Start the binding reaction by addition of membrane protein,f bringing the assay to the desired volume and incubate the reaction for sufﬁcient time so as to reach equilibrium binding (at least ﬁve times the dissociation half-life of the ligand). 4 At the appropriate time, terminate the reaction by rapid vacuum ﬁltration through glass-ﬁbre ﬁlter paper followed by three to four 4 ml washes with ice-cold wash buffer to separate bound from free radioligand. 5 Determine radioactivity by liquid scintillation counting.

1.2 METHODS AND APPROACHES

13

Notes a For

any radioligand binding assay, the buffer may be a simple buffer containing no or low concentrations of ions with chelating agents such as EDTA and EGTA (e.g. 20 mM HEPES, 1 mM EDTA, 1 mM EGTA, pH 7.4) or the buffer may resemble a more physiological environment or a buffer used for functional assays such as [35 S]GTPγ S binding assays (e.g. 20 mM HEPES, 100 mM NaCl, 10 mM MgCl2 , pH 7.4). Care should be taken when choosing the appropriate acid or base to adjust the pH of the buffer. Sodium hydroxide (NaOH), for instance, will alter the concentration of sodium ions in the solution, which may have effects on the binding of ligands. 1–2 M KOH is more appropriate for adjusting the pH of radioligand binding buffers. b Proteins and peptides are less likely to stick to polypropylene than to polystyrene-based plastics. c PBS

or other simple saline solution, such as 0.9% NaCl.

d

GF/B-grade glass-ﬁbre ﬁlter paper has larger pores and, therefore, is generally better suited to cell-based binding assays, whilst GF/C-grade glass-ﬁbre ﬁlter paper is better suited for membrane-based assays. However, higher nonspeciﬁc binding may be apparent when using GF/C-grade glass-ﬁbre ﬁlter paper. e Certain

scintillation cocktails may be better suited to speciﬁc applications. For instance, whilst Perkin Elmer’s Ultima Gold is suitable for both aqueous and nonaqueous samples, Irga Safe Plus is more suited to aqueous samples. f The

appropriate protein concentration should be optimized prior to the performance of any ligand binding assay to ensure that ligand depletion does not occur.

1.2.2.3 Heterologous and homologous competition binding experiments The use of high radioligand concentrations in saturation binding experiments can be expensive, whilst many ligands are unavailable in radioactive form, meaning that their afﬁnity for a receptor cannot always be directly determined. Competition binding assays, however, are an alternative means to study interactions between the receptor and an unlabelled ligand under equilibrium conditions, by measuring the binding pattern of a single radioligand concentration in the presence of increasing concentrations of an unlabelled competitor (Protocol 1.3). The concentration of the unlabelled ligand that causes 50% inhibition of radioligand binding is termed the IC50 . However, this does not necessarily mean that this concentration of ligand will bind half the receptors, as the IC50 of the unlabelled ligand will depend upon the ability of the radioligand to bind to the receptor (e.g. its equilibrium dissociation constant), the ability of the unlabelled ligand to bind to the receptor and the concentration of the radiolabelled ligand. Thus, the IC50 will normally be greater than the equilibrium dissociation constant Kb of the unlabelled ligand. However, the Cheng–Prusoff equation (Equation 1.4) can be used to determine the Kb of the competitor in a competition binding experiment. If one of the ligands being used for the assay is an agonist, then it may be desirable to eliminate, or at least minimize, the ability of the G protein to couple to the receptor, so that the interaction between the ligand and the receptor is largely independent of G protein coupling. For this reason, many researchers choose to perform agonist

14

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

competition binding assays in the presence of guanine nucleotides and, hence, enable the data to be ﬁtted to a simple one-site binding model. However, if the Hill coefﬁcient of the inhibition curve is signiﬁcantly different from unity, then this may indicate that the agonist binds to two different receptor species, which are usually described as a high- and low-afﬁnity state. Under these circumstances, binding of the competitor to two sites may be the preferred model (see Equation 1.14). An alternative approach to competition binding assays measures the competition between the radioligand and a non-radiolabelled version of the radioligand. This type of competition assay is called homologous competition binding [17]. It can be utilized to determine the afﬁnity of the radioligand for a receptor in addition to the number of binding sites present, if the radioligand, A, and unlabelled competitor, B, share an identical afﬁnity for the receptor. Under these circumstances, binding of the radioligand will be a fraction of the total ligand bound ([A] + [B]) and will be deﬁned by [AR] = where

[A] × [Rt ] [A] + [B] + Ka

(1.11)

Ka = IC50 − [A]

(1.12)

% Specific [3H]CCPA binding

If we consider a heterologous competition binding assay, binding of a radioligand, expressed on the y-axis, in the presence of increasing concentrations of a competitor, expressed on the x-axis, will follow a sigmoidal curve that can be described by Equation 1.3 (see Figure 1.4). Depending on the nonlinear regression program used to analyse the data, the program must be told to deﬁne parameters in a logarithmic

100 80 60 40 20 0 −12

−11

−10 −9 −8 −7 Log [R-PIA] (M)

−6

Figure 1.4 Competition binding of the radiolabelled agonist, [3 H]2-chloro-N6 -cyclopentyladenosine ([3 H]CCPA), at the A1 adenosine receptor by increasing concentrations of the unlabelled agonist, N6 -((R)-2-phenylisopropyl)adenosine (R-PIA). Membranes were incubated with [3 H]CCPA and R-PIA for 1 h at 30 ◦ C before termination of the assay as described in the methods section. Data were transformed to the percentage of speciﬁc [3 H]CCPA binding and ﬁtted closely to a model that described one-site binding with a Hill coefﬁcient of unity.

15

1.2 METHODS AND APPROACHES

form if this is how they are to be reported. Alternatively, Equation 1.3 can be recast to reﬂect this, as shown by the following Y =

Top − Bottom + Bottom 1 + 10(log[B]−log IC50 )n

(1.13)

If the Hill coefﬁcient n is not signiﬁcantly different from unity, then this may be constrained as such in the analysis. If, however, it appears to vary from unity, then the data may sometimes be ﬁtted best by a two-site binding model: fraction_1 1 − fraction_1 + Bottom Y = (Top − Bottom) + 1 + 10log[B]−log IC50_1 1 + 10log[B]−log IC50_2 (1.14) For analysis of homologous binding data we can use Equation 1.15 to deﬁne binding: Y =

PROTOCOL 1.3

Top − Bottom + Bottom 1 + 10log[B]−log[A] + 10log Ka −log[A]

(1.15)

Competition Binding Assays

Equipment and Reagents • Membrane preparation (Protocol 1.1) • Binding buffera • 5 ml polypropylene assay tubes (Techno-plas) • Radioligandb • Unlabelled ligandc • Guanine nucleotided (Sigma) • Wash buffere • Water bath • Vacuum harvester (Brandel) • Glass-ﬁbre ﬁlter paper (Whatman) • Liquid scintillation cocktail (PerkinElmer) • Liquid scintillation counter (e.g. Packard Tri-Carb LS counter).

Method 1 Dilute the receptor preparation, radioligand, unlabelled ligand and guanine nucleotides (if required) in assay buffer to 10× the ﬁnal concentration desired in the assay. 2 In an appropriate volume of binding buffer, add a single concentration of radioligand to all assay tubes and add increasing concentrations of the unlabelled ligand into different tubes. Ensure that you prepare tubes in which no unlabelled ligand is added to deﬁne

16

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

total binding, in addition to tubes that contain no unlabelled ligand but a high concentration of an alternative competitive ligand to deﬁne nonspeciﬁc binding. If desired, add an appropriate concentration of guanine nucleotides to each tube to minimize G protein coupling to the receptor (see Table 1.2).

Table 1.2 Tube set-up used to determine binding of [3 H]CCPA in the presence of increasing concentrations of R-PIA using a 500 µl ﬁnal assay volume, where nonspeciﬁc binding is determined in the presence of 100 µM R-PIA. 1 mM GppNHp Buffer 20 nM [3 H]CCPA Membrane (100 µM ﬁnal (µl) (2 nM ﬁnal (µl) concentration) concentration) (µl) (µl)

Final [R-PIA] R-PIA (M)

0

0

3.16 × 1×

10−10

3.16 × 1×

10−11 10−10

10−9

3.16 ×

10−9

1 × 10−8 3.16 × 1×

10−7

3.16 × 1×

10−8 10−7

10−6

Nonspeciﬁc

160 (of 1 × 50 (of 1 ×

10−9 M)

160 (of 1 × 50 (of 1 ×

10−10 M) 10−9 M)

10−8 M)

160 (of 1 ×

10−8 M)

50 (of 1 × 10−7 M) 160 (of 1 × 50 (of 1 ×

10−7 M)

10−6 M)

160 (of 1 ×

10−6 M)

50

350

50

50

50

190

50

50

50

300

50

50

50

190

50

50

50

300

50

50

50

190

50

50

50

300

50

50

50

190

50

50

50

300

50

50

50

190

50

50

50 (of 1 ×

10−5 M)

50

300

50

50

50 (of 1 ×

10−3 M)

50

300

50

50

binding

3 Start the reaction by the addition of membrane and incubate the assay for the desired time so as to reach equilibrium binding. 4 Terminate the reaction and determine radioactivity as described in Protocol 1.2.

Notes a As

described in Protocol 1.2.

b The

radioligand concentration should be approximately equal to its Ka or lower.

c This

will be an unlabelled form of the radioligand for homologous competition binding assays, or an alternative unlabelled ligand for heterologous competition binding assays. d If guanine nucleotides are to be added to the assay, 100 µM GppNHp or 1 mM GTP is generally sufﬁcient to uncouple receptor–G protein complexes. e As

described in Protocol 1.2.

1.2 METHODS AND APPROACHES

17

1.2.2.4 Equilibrium binding experiments with an allosteric modulator Saturation assays performed in the absence and presence of a putative allosteric modulator are a useful means to determine whether a shift in radioligand afﬁnity in the presence of the modulator is in agreement with an allosteric mode of action (Protocol 1.4). For a competitive ligand, if the ratio of the afﬁnity of the radioligand in the presence (Ka ) and absence (Ka ) of this competitor, otherwise known as the ‘dose ratio’, is determined, then a Schild plot of log[(Ka /Ka ) − 1] against the log competitor concentration should yield a straight line with a slope of unity [18]. If, however, there is negative cooperativity between the putative modulator and the orthosteric ligand, then the slope may differ from unity and a curvilinear plot may result. The curvature in the Schild plot will represent the cooperativity between the two ligands, reﬂecting a limit in the shift in the radioligand afﬁnity. Positive allosteric interactions, in contrast, can be characterized by an enhancement in the afﬁnity of the radioligand, causing a leftward shift in the radioligand binding curve. Under these circumstances, the dose ratio can be determined as the ratio of radioligand afﬁnities in the absence and presence of the allosteric modulator and a Schild plot of log(Ka /Ka ) against the log modulator concentration may reﬂect the positive cooperativity [8]. Competition binding assays may also be used to study the direct effects of an allosteric modulator on the binding of a radioligand at the orthosteric site; although the allosteric modulator will not directly compete for binding with the orthosteric radioligand, the cooperativity will mediate alterations in the binding of the radioligand. If we consider the scheme shown in Figure 1.2, then fractional occupancy of a receptor population by the orthosteric ligand [A] in the presence of an allosteric modulator [B] can be described by

ρAR+ARB = [A] + Ka

Bmax × [A] [B] α[B] 1+ 1+ Kb Kb

(1.16)

Note that the apparent afﬁnity Kapp of the radioligand in the presence of the allosteric modulator is deﬁned by [B] 1+ Kb Kapp = Ka (1.17) α[B] 1+ Kb Care must be taken when choosing the radioligand concentration for such assays. If the allosteric modulator has very high negative cooperativity with the radioligand, it may appear competitive if a low radioligand concentration is employed giving rise to low radioligand receptor occupancy. As the radioligand concentration is increased, the ability of the allosteric modulator to inhibit radioligand binding will be reduced, which is seen by an increase in the bottom asymptote of the inhibition curve. Subsequently, full inhibition by the allosteric modulator may not be achieved using high radioligand concentrations, preventing its actions from being distinguished from a competitive antagonist (see Figure 1.5). However, nonequilibrium binding artefacts may greatly

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS % Specific [3H]NMS Binding

18

100

[3H]NMS (nM)

80

2 0.2

60 40 20 0 −9

−8

−7

−6

−5

−4

−3

Log [C7/3-phth] (M)

Figure 1.5 Normalized binding of two different concentrations of the radioligand, [3 H]NMS, at the M4 muscarinic ACh receptor in the presence of increasing concentrations of the muscarinic receptor allosteric modulator, heptane-1,7-bis(dimethyl-3 -phthalimidopropyl)ammonium bromide (C7 /3-phth). Membranes were incubated with C7 /3-phth for 1 h at 37 ◦ C before the assay was terminated as described in Section 1.2.

inﬂuence the results obtained in such an experiment and care must be taken to interpret the observations correctly [19]. Several alternative approaches have been described to determine the effects that the allosteric modulator can exert [10, 20]. One approach is to compete radioligand binding with an unlabelled competitive ligand in the presence of different concentrations of allosteric modulator [20], which can give detailed information regarding the interaction of the allosteric modulator with both the radioligand and the orthosteric competitor (Protocol 1.4). The assay described in Protocol 1.4 can be used to determine the cooperativity between the allosteric modulator and both the radioligand and unlabelled competitor, where binding of the radioligand [A] in the presence of the competitor [B] and the allosteric modulator [X] is described by

Y = [A] +

Ka KX α[X] + KX

Bmax × [A] [B] [X] β[B] × [X] 1+ + + Kb Kb Kb KX

(1.18)

where Ka , Kb and KX are the equilibrium dissociation constants of the radioligand, competitor and allosteric modulator respectively. Figure 1.6 shows competition binding at the M4 muscarinic ACh receptor in the absence and presence of an allosteric modulator, which exerts weak negative cooperativity with the radioligand, [3 H]NMS. In contrast, however, the same modulator exerts positive cooperativity with ACh, demonstrated by the leftward shift in the apparent ability of ACh to compete for [3 H]NMS binding.

19

% Specific [3H]NMS binding

1.2 METHODS AND APPROACHES [Modulator] (mM) 100

0 0.3

80

1 60

3 10

40 20 0 −10

−9

−8

−7 −6 −5 Log [ACh] (M)

−4

−3

Figure 1.6 Competition of binding of the radioligand, [3 H]NMS, at the M4 muscarinic ACh receptor by increasing concentrations of the agonist, ACh, in the absence and presence of increasing concentrations of an allosteric potentiator of ACh binding. Membranes were equilibrated with ACh and the allosteric modulator for 3 h at 37 ◦ C before termination of the assay as described in Section 1.2.

PROTOCOL 1.4 Orthosteric Radioligand Binding in the Presence of an Allosteric Modulator Equipment and Reagents • Membrane preparation (Protocol 1.1) • Binding buffera • 5 ml polypropylene assay tubes (Techno-plas) • Radioligandb • Unlabelled competitor • Allosteric modulator • Guanine nucleotidesc (Sigma) • Wash bufferd • Water bath • Vacuum harvester (Brandel) • Glass-ﬁbre ﬁlter paper (Whatman) • Liquid scintillation cocktail (PerkinElmer) • Liquid scintillation counter (e.g. Packard Tri-Carb LS counter).

20

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

Method 1 Dilute the receptor preparation, radioligand, unlabelled competitor, allosteric modulator and guanine nucleotides (if required) in assay buffer to 10× the ﬁnal concentration desired in the assay. 2 In an appropriate volume of binding buffer, add a single concentration of radioligand to all tubes. 3 For a control curve in the absence of allosteric modulator, add increasing concentrations of the unlabelled competitive ligand into different tubes. Remember to prepare tubes in which no unlabelled ligand is added to deﬁne total binding, in addition to tubes that contain no unlabelled ligand but a high concentration of an alternative competitive ligand to deﬁne nonspeciﬁc binding. 4 To a second set of tubes, add increasing concentrations of the unlabelled ligand into different tubes in addition to the allosteric modulator. Several concentrations of modulator should be tested. If each concentration–response curve will be harvested separately, for every curve derived in the presence of the allosteric modulator, prepare tubes in which no unlabelled ligand is added but the allosteric modulator is present, to deﬁne total binding in the presence of the modulator, in addition to tubes that contain no unlabelled ligand and no allosteric modulator. This will account for variation between total binding counts on different pieces of ﬁlter paper. The total binding in the absence of allosteric modulator can be used to transform data if the allosteric modulator shows negative or positive cooperativity with the radioligand and, therefore, reduces or increases its binding. Also ensure that the vehicle in which the allosteric modulator is prepared exerts no effect on radioligand binding. In addition, prepare tubes that contain no unlabelled ligand but a high concentration of a different competitive ligand to deﬁne nonspeciﬁc binding (see Table 1.3).

Table 1.3 Tube set-up used to determine the binding of [3 H]NMS to the M4 muscarinic ACh receptor in the presence of increasing concentrations of ACh and an allosteric modulator, using a 500 µl ﬁnal assay volume, where nonspeciﬁc binding is determined in the presence of atropine. Final [ACh] (M)

Allosteric 1 mM GppNHp Buffer 2 nM [3 H]NMS Membrane modulator (100 µM ﬁnal (µl) (0.2 nM ﬁnal (µl) (µl) concentration) concentration) (µl) (µl)

ACh (µl)

Control curve in the absence of allosteric modulator 0

0

0

50

350

50

50

1 × 10−9

50 (of 1 × 10−8 M)

0

50

300

50

50

3.16 × 10

−9

160 (of 1 × 10

−8

M)

0

50

190

50

50

0

50

300

50

50

1 × 10−8

50 (of 1 × 10−7 M)

3.16 × 10−8

160 (of 1 × 10−7 M)

0

50

190

50

50

10−7

50 (of 1 × 10 –6 M)

0

50

300

50

50

3.16 × 10 –7

160 (of 1 × 10−6 M)

0

50

190

50

50

0

50

300

50

50

1× 1×

10−6

50 (of 1 ×

10−5 M)

(continued overleaf )

21

1.2 METHODS AND APPROACHES

Table 1.3 (continued) Allosteric 1 mM GppNHp Buffer 2 nM [3 H]NMS Membrane modulator (100 µM ﬁnal (µl) (0.2 nM ﬁnal (µl) (µl) concentration) concentration) (µl) (µl)

Final [ACh] (M)

ACh (µl)

3.16 × 10−6

160 (of 1 × 10−5 M)

0

50

190

50

50

10−5

50 (of 1 × 10−4 M)

0

50

300

50

50

3.16 × 10−5

160 (of 1 × 10−4 M)

0

50

190

50

50

0

50

300

50

50

0

50

190

50

50

0

50

300

50

50

0

50

300

50

50

0

50

300

50

50

1×

1 × 10

−4

3.16 × 10−4 1×

10−3

Nonspeciﬁc binding

50 (of 1 × 10

−3

M)

160 (of 1 × 10−3 M) 50 (of 1 ×

10−2 M)

50 (of 100 µM) atropine

Curve in the presence of allosteric modulator Vehicle control

50 vehicle

Total binding

0

0

50

350

50

50

0

0

50

50

300

50

50

50

50

250

50

50

1 × 10−9 3.16 ×

10−9

50 (of 1 × 10−8 M) 160 (of 1 ×

10−8 M)

50

50

140

50

50

50

50

250

50

50

1 × 10−8

50 (of 1 × 10−7 M)

3.16 × 10−8

160 (of 1 × 10−7 M)

50

50

140

50

50

10−7

50 (of 1 × 10−6 M)

50

50

250

50

50

3.16 × 10−7

160 (of 1 × 10−6 M)

50

50

140

50

50

50

50

250

50

50

50

50

140

50

50 50

1× 1×

10−6

3.16 × 10−6 1 × 10

−5

50 (of 1 ×

10−5 M)

160 (of 1 × 10−5 M) 50 (of 1 × 10

−4

50

50

250

50

3.16 × 10−5

160 (of 1 × 10−4 M)

50

50

140

50

50

1 × 10−4

50 (of 1 × 10−3 M)

50

50

250

50

50

3.16 × 10−4

160 (of 1 × 10−3 M)

50

50

140

50

50

10−3

50 (of 1 × 10−2 M)

50

50

250

50

50

50 DMSO

50

250

50

50

1×

Nonspeciﬁc binding

50 (of 100 µM) atropine

M)

5 Start the reaction by the addition of membrane and incubate the assay for the desired time so as to reach equilibrium binding. 6 Terminate the reaction and determine radioactivity as described in Protocol 1.2.

Notes a As b

described in Protocol 1.2.

The radioligand concentration should approximately equal its Ka .

c See d

note d in Protocol 1.3.

As described in Protocol 1.2.

22

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

1.2.2.5 Kinetic radioligand binding assays Kinetic assays can be used to determine the association or dissociation rate constants of a radioligand (Protocol 1.5). This information is useful, for instance, for deciding the correct incubation time for many radioligands to reach equilibrium conditions so that equilibrium dissociation and association constants can be accurately measured in subsequent assays [21]. Kinetic assays can additionally be used to reveal cooperativity between two binding sites, where the binding kinetics of the orthosteric ligand may be modulated. Method 1 in Protocol 1.5 describes measurement of the time taken to reach equilibrium binding, which will be represented by a plateau in the binding curve over time. The rate at which equilibrium is reached is dependent on the radioligand concentration and the rate at which the ligand associates with and dissociates from the receptor. In the presence of a single radioligand concentration, this information can be used to calculate the association rate constant Kon only if the rate at which the ligand dissociates from the receptor Koff is known. If Koff is known, then we can use Equation 1.19 to calculate Kon : Kobs − Koff Kon = (1.19) [A] However, if the Koff of the ligand is unknown, then both Kon and Koff can be determined by measuring the Kobs of multiple concentrations of radioligand. Traditionally, Kobs (in inverse time) is plotted against the inverse radioligand concentration to derive a straight line with a gradient equal to Kon and a y-intercept equal to Koff . Alternatively, the Kobs of at least two different concentrations of radioligand can be ﬁtted globally to the kinetic model shown in Equation 1.6. Otherwise, the rate of radioligand association can be determined by measuring both the association and dissociation of the radioligand in one experiment. A kinetic assay of this kind allows us to measure the dissociation rate of the radioligand directly in inverse time and will yield the Kobs , as described for association experiments; so, if we know the radioligand concentration, then we can deﬁne occupancy over time using Equation 1.6.

PROTOCOL 1.5 Measurement of Radioligand Binding Kinetics Equipment and Reagents • Membrane preparation (Protocol 1.1) • Binding buffera • 5 ml polypropylene assay tubes (Techno-plas) • Radioligand • Competitive ligand • Wash bufferb • Water bath • Vacuum harvester (Brandel)

23

1.2 METHODS AND APPROACHES

• Glass-ﬁbre ﬁlter paper (Whatman) • Liquid scintillation cocktail (PerkinElmer) • Liquid scintillation counter (e.g. Packard Tri-Carb LS counter).

Method 1: Association Binding Kinetics 1 Dilute the receptor preparation, radioligand and competitive ligand in assay buffer to 10× the ﬁnal concentration desired in the assay. 2 Prepare tubes containing the radioligand and identical tubes also containing a saturating concentration of a competitive ligand to deﬁne nonspeciﬁc binding.c Tubes that contain no radioligand must also be prepared to represent binding of the radioligand at the zero time point (i.e. no radioligand binding). Membrane can be added to these tubes before the start of the assay, as the values obtained from these tubes will purely be representative of background counts that arise from sources other than the radioligand. 3 Stagger the addition of membrane, bringing the assay to the desired ﬁnal volume, so that binding of the radioligand can be determined at various time points. For instance, add membrane to the ﬁrst tube at time point 0 and make subsequent membrane additions to successive tubes at 5, 10, 15 and 18 min. Separate bound from free radioligand (as described in Protocol 1.2) at 20 min. Thus, tubes containing membrane added at time points 0, 5, 10, 15 and 18 min will represent 20, 15, 10, 5 and 2 min incubation between the membrane and radioligand respectively, whilst tubes containing membrane but no radioligand will represent time point 0 (see Table 1.4).

Table 1.4 Tube set-up used to determine the binding of [3 H]NMS to the M4 muscarinic ACh receptor over time, using a 1 ml ﬁnal assay volume, where nonspeciﬁc binding is determined in the presence of atropine. Time of [3 H]NMS addition (min)

Association time (min)

2 nM [3 H]NMS (0.2 nM ﬁnal concentration) (µl)

Buffer (µl)

100 µM Atropine (10 µM ﬁnal concentration) (µl)

Membrane (µl)

Speciﬁc binding 0

20

100

800

0

100

5

15

100

800

0

100

10

10

100

800

0

100

15

5

100

800

0

100

18

2

100

800

0

100

–

0

0

900

0

100 (continued overleaf )

24

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

Table 1.4 (continued) Time of [3 H]NMS addition (min)

Association time (min)

2 nM [3 H]NMS (0.2 nM ﬁnal concentration) (µl)

Buffer (µl)

100 µM Atropine (10 µM ﬁnal concentration) (µl)

Membrane (µl)

100

700

100

100

Nonspeciﬁc binding 0

20

5

15

100

700

100

100

10

10

100

700

100

100

15

5

100

700

100

100

18

2

100

700

100

100

–

0

0

800

100

100

4 Determine radioactivity as described in Protocol 1.2.

Method 2: Dissociation Binding Kinetics 1 Dilute the receptor preparation and radioligand in assay buffer to 10× the ﬁnal concentration desired in the assay. Dilute the competitive ligand to 100 000× its Ka value, as this will be diluted 100× into the assay, so the ﬁnal concentration of competitive ligand will be 1000× its Ka . 2 Prepare tubes containing the radioligand. Also prepare one set of tubes containing a saturating concentration of a competitive ligand to deﬁne nonspeciﬁc binding at each time point. 3 Stagger the addition of membrane to each tube so that the membrane is incubated with the radioligand for the same amount of time in each tube prior to the addition of competitor (see Table 1.5).

Table 1.5 Tube set-up to determine the dissociation kinetics of [3 H]NMS from the M4 muscarinic ACh receptor over time, using a 1 ml ﬁnal assay volume. Time of Time of Dissociation 2 nM [3 H]NMS Buffer Membrane 1 mM Atropine membrane atropine time (0.2 nM ﬁnal (10 µM addition addition (min) concentration) ﬁnal (min) (min) (µl) concentration) 0

60

20

100

800

100

10

5

65

15

100

800

100

10

10

70

10

100

800

100

10 (continued overleaf )

25

1.2 METHODS AND APPROACHES

Table 1.5 (continued) Time of Time of Dissociation 2 nM [3 H]NMS Buffer Membrane 1 mM Atropine (10 µM membrane atropine time (0.2 nM ﬁnal addition addition (min) concentration) ﬁnal (min) (min) (µl) concentration) 15

75

5

100

800

100

10

18

78

2

100

800

100

10

20

–

0

100

800

100

0

Note that this tube set-up can be used for both speciﬁc and nonspeciﬁc binding; however, for nonspeciﬁc binding, atropine must be present throughout the entire incubation period that the receptor is incubated with [3 H]NMS, rather than staggering its addition following equilibrium binding of [3 H]NMS.

4 Following equilibrium binding of the radioligand with the receptor,d stagger the addition of a small volume (∼10 µl to minimize adjustment of the assay volume and, thus, prevent changes in the equilibrium binding of the radioligand) of a saturating concentration of a competitive ligand, which will bind to the unoccupied receptors and prevent reassociation of the radioligand with the receptor, or by diluting the sample by at least 100-fold to reduce the free concentration of radioligand by this factor (so that its concentration is far lower than its Ka and binding will subsequently be negligible). 5 At the appropriate time, terminate the reaction to separate bound from free radioligand and determine radioactivity as described in Protocol 1.2.

Notes a As b

described in Protocol 1.2.

As described in Protocol 1.2.

c For

nonspeciﬁc binding, ensure that the competitive ligand is present in the assay for the entire incubation period with radioligand. d Be

aware that low ligand concentrations will take longer to equilibrate.

1.2.2.6 Kinetic radioligand binding assays in the presence of an allosteric modulator using isotopic dilution A competitive ligand cannot change the dissociation binding kinetics of an orthosteric ligand as the two cannot occupy the receptor at the same time. However, when an allosteric modulator binds to a receptor, it may alter the receptor conformation such that it forms a ‘new’ receptor species that has a new set of afﬁnities for its orthosteric ligands. This is often reﬂected in alterations in the association and/or dissociation kinetics of the orthosteric ligand. Kinetic radioligand binding assays can, therefore, be used to reveal important information regarding the binding mechanism of a ligand and are a sensitive way to determine whether two ligands can bind to a receptor

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

% Specific [3H]NMS binding

26

120

[3H]NMS dissociation

100

[3H]NMS dissociation± C7/3-phth

80 60 40 20 0 0

25

50 75 Time (min)

100

125

Figure 1.7 Dissociation of the radioligand, [3 H]NMS, from the M4 muscarinic ACh receptor in the absence and presence of the muscarinic receptor allosteric modulator, C7 /3-phth. The radioligand was allowed to equilibrate with the receptor for 1 h at 37 ◦ C before its dissociation was observed by the addition of a high concentration (100× the K a ) of atropine in the absence and presence of C7 /3-phth and termination of the assay as described in Section 1.2.

simultaneously (Protocol 1.6). Determination of the dissociation rate of a ligand in the absence and presence of another is the simplest way to measure kinetic effects (see Figure 1.7). The radioligand is usually allowed to reach equilibrium binding and its association with the receptor is prevented, as described previously, in the absence or presence of a putative allosteric modulator (for example, see [20] for protocols).

PROTOCOL 1.6 Measurement of Radioligand Dissociation Kinetics in the Presence of an Allosteric Modulator Equipment and Reagents • Membrane preparation • Binding buffera • 5 ml polypropylene assay tubes (Techno-plas) • Radioligand • Allosteric modulator • Competitive antagonist • Wash bufferb • Water bath • Vacuum harvester (Brandel) • Glass-ﬁbre ﬁlter paper (Whatman) • Liquid scintillation cocktail (PerkinElmer) • Liquid scintillation counter (e.g. Packard Tri-Carb LS counter).

REFERENCES

27

Method 1 Dilute the receptor preparation in assay buffer to 10× the ﬁnal desired assay concentration and dilute the radioligand in assay buffer to approximately 10× its Ka .c Dilute both the competitive ligand and the allosteric modulator in assay buffer to 200× the ﬁnal desired assay concentration. Further dilute the competitive ligand at a 1 : 1 ratio with either the allosteric modulator or with assay buffer so that each ligand is 100× concentrated. 2 Prepare two sets of tubes containing radioligand, so that radioligand dissociation can be measured in one set and radioligand dissociation in the presence of the allosteric modulator can be measured in the other set. Also prepare tubes containing radioligand in which a saturating concentration of a competitive ligand will be incubated with the receptor throughout the experiment to deﬁne nonspeciﬁc binding. 3 Stagger the addition of membrane to tubes containing the radioligand and incubate for sufﬁcient time so as to reach equilibrium binding. 4 Stagger the addition of the competitive ligand with and without the putative allosteric modulator. The tube set-up outlined in Table 1.5 can be used for this assay, with the allosteric modulator–competitive ligand mix being added in the appropriate tubes in place of the competitive ligand alone. 5 At the appropriate time, separate bound from free radioligand and determine radioactivity as described in Protocol 1.2.

Notes a As

described in Protocol 1.2.

b As

described in Protocol 1.2.

c Generally,

to observe kinetic effects, higher concentrations of an allosteric modulator will be required than those required to observe cooperative effects in equilibrium binding assays; therefore, reducing receptor occupancy by the radioligand will increase the ability to detect alterations in the binding kinetics of the radioligand.

References 1. Barcroft, J. and Hill, A.V. (1910) The nature of oxyhæmoglobin, with a note on its molecular weight. J. Physiol., 39, 411–428. 2. Hill, A.V. (1910) The possible effects of the aggregation of the molecules of haemoglobin on its dissociation curves. Proc. Physiol. Soc., 40, iv–vii. 3. Gaddum, J.H. (1937) The quantitative effects of antagonistic drugs. J. Physiol., 89, 7P–9P. One of the original publications by Gaddum describing the concept of competitive antagonism. 4. Gaddum, J.H. (1943) Introductory address. Part I. Biological aspects: the antagonism of drugs. Trans. Faraday Soc., 39, 323–332.

28

CH 1 MEASUREMENT OF LIGAND–G PROTEIN-COUPLED RECEPTOR INTERACTIONS

5. Cheng, Y. and Prusoff, W.H. (1973) Relationship between the inhibition constant (Ki ) and the concentration of inhibitor which causes 50 per cent inhibition (I50 ) of an enzymatic reaction. Biochem. Pharmacol., 22, 3099–3108. 6. De Lean, A., Stadel, J.M. and Lefkowitz, R.J. (1980) A ternary complex model explains the agonist-speciﬁc binding properties of the adenylate cyclase-coupled beta-adrenergic receptor. J. Biol. Chem., 255, 7108–7117. The original publication describing a ternary complex model for GPCR activation. 7. Samama, P., Cotecchia, S., Costa, T. and Lefkowitz, R.J. (1993) A mutation-induced activated state of the beta 2-adrenergic receptor. Extending the ternary complex model. J. Biol. Chem., 268, 4625–4636. The original publication describing an extended ternary complex model for GPCR activation. 8. May, L.T., Leach, K., Sexton, P.M. and Christopoulos, A. (2007) Allosteric modulation of G protein-coupled receptors. Annu. Rev. Pharmacol. Toxicol., 47, 1–51. A good review covering the concepts of allosterism and the detection and quantiﬁcation of allosteric effects. 9. Ehlert, F.J. (1988) Estimation of the afﬁnities of allosteric ligands using radioligand binding and pharmacological null methods. Mol. Pharmacol., 33, 187–194. The original publication describing an allosteric ternary complex model. 10. Christopoulos, A. and Kenakin, T. (2002) G protein-coupled receptor allosterism and complexing. Pharmacol. Rev., 54, 323–374. 11. Langmead, C.J. and Christopoulos, A. (2006) Allosteric agonists of 7TM receptors: expanding the pharmacological toolbox. Trends Pharmacol. Sci., 27, 475–481. 12. Horstman, D.A., Brandon, S., Wilson, A.L. et al. (1990) An aspartate conserved among G-protein receptors confers allosteric regulation of alpha 2-adrenergic receptors by sodium. J. Biol. Chem., 265, 21590–21595. 13. Galvez, T., Urwyler, S., Pr´ezeau, L. et al. (2000) Ca2+ requirement for high-afﬁnity γ -aminobutyric acid (GABA) binding at GABAB receptors: involvement of serine 269 of the GABABR1 subunit. Mol. Pharmacol., 57, 419–426. 14. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem., 193, 265–275. 15. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem., 72, 248–254. 16. Motulsky, H. and Christopoulos, A. (2004) Fitting Models to Biological Data Using Linear and Nonlinear Regression. A Practical Guide to Curve Fitting, Oxford University Press, Oxford. 17. DeBlasi, A., O’Reilly, K. and Motulsky, H.J. (1989) Calculating receptor number from binding experiments using same compound as radioligand and competitor. Trends Pharmacol. Sci., 10, 227–229. 18. Arunlakshana, O. and Schild, H.O. (1959) Some quantitative uses of drug antagonists. Br. J. Pharmacol. Chemother., 14, 48–58. One of the original publications describing concepts of competitive and noncompetitive antagonism and the use of the ‘Schild plot’.

REFERENCES

29

19. Avlani, V., May, L.T., Sexton, P.M. and Christopoulos, A. (2004) Application of a kinetic model to the apparently complex behavior of negative and positive allosteric modulators of muscarinic acetylcholine receptors. J. Pharmacol. Exp. Ther., 308, 1062–1072. 20. Lazareno, S. and Birdsall, N.J. (1995) Detection, quantitation, and veriﬁcation of allosteric interactions of agents with labeled and unlabeled ligands at G protein-coupled receptors: interactions of strychnine and acetylcholine at muscarinic receptors. Mol. Pharmacol., 48, 362–378. 21. Motulsky, H.J. and Mahan, L.C. (1984) The kinetics of competitive radioligand binding predicted by the law of mass action. Mol. Pharmacol., 25, 1–9.

2 Second Messenger Assays for G Protein-coupled Receptors: cAMP, Ca2+, Inositol Phosphates, ERK1/2 Karen J. Gregory, Patrick M. Sexton, Arthur Christopoulos and Caroline A. Hick Drug Discovery Biology Laboratory, Department of Pharmacology, Monash University, Clayton, Australia

2.1 Introduction Cells respond to their environment through a complex and interdependent series of signal transduction pathways. The largest family of cell surface proteins involved in the transduction of signals across biological membranes comprises G protein-coupled receptors (GPCRs). GPCRs have evolved to recognize a huge variety of different endogenous stimuli, ranging from lipids, peptides, proteins and nucleotides to ions and photons [1]. The predominant mechanism utilized by GPCRs for transducing extracellular stimuli to the intracellular environment involves the coupling of the receptor to intracellular heterotrimeric G proteins [2, 3]. The involvement of an intracellular coupling partner that can vary in subcellular localization, as well as between different cell types, generates enormous diversity in both the strength and nature of the resultant signal. It is these features that explain the prominent role of GPCRs as extracellular chemical sensors and their importance as targets for the development of drugs with wide clinical applications [4]. There is still much to be learnt about how GPCRs function and how they can be selectively modulated. Fortunately, the wide interest in GPCR research has facilitated the development of a huge range of screening G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

32

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

techniques. With the need in cellular screening to ﬁnd robust markers to monitor cells overexpressing the GPCR of interest, second messenger assays have become an essential tool in both facilitating GPCR drug development and understanding basic biology [5]. Second messengers are molecules that relay signals received by receptors on the cell surface to target molecules in the cytosol and the nucleus. Second messengers also serve to amplify the strength of a signal greatly. Activation of a GPCR induces the exchange of guanosine diphosphate for guanosine triphosphate on the G protein α subunit and dissociation of the α subunit from the βγ heterodimer [6]. Consequently, the Gα- and

b ai

as

aq/11

PI3K

g

Grb AC

PLC DAG

ATP

cAMP

PKA

Sos

PIP2 IP3

PKC

Ras

Ca2+ MEK

ERK1/2

Figure 2.1 Major intracellular signalling pathways arising from G protein coupling. The free Gαs subunit activates ACs, resulting in increased levels of intracellular cAMP, leading to the activation of downstream effectors, including protein kinase A (PKA). On the other hand, Gαi subunits inactivate AC, resulting in decreased levels of cAMP. The Gαq/11 family activate PLC, which hydrolyses phosphatidyl-inositol-4,5-bisphosphate (PIP2 ) to give diacylglycerol (DAG) and IP3 . IP3 acts at IP3 receptors present on intracellular stores to mobilize intracellular Ca2+ . Increased levels of intracellular Ca2+ and DAG can activate protein kinase C (PKC), which can activate the small G protein Ras, leading to the activation of the MAPK kinase (MEK), and subsequently activation of extracellular signal-regulated kinases 1 and 2 (ERK1/2). The βγ heterodimer can also activate ERK1/2. It does this through facilitating the localization, and activation of a number of kinases (e.g. PI-3-kinase) and adaptor molecules (e.g. Grb and Sos). ERK1/2 activation can also arise from Gαs coupling, the pathway being dependent on PKA and MEK.

2.2 METHODS AND APPROACHES

33

Gβγ-subunits stimulate effector molecules, which include adenylyl cyclase (AC), phosphodiesterases, phospholipase A2 (PLA2 ) and phospholipase C (PLC), thereby activating or inhibiting the production of a variety of second messengers [3]. There are four major classes of second messengers for which the majority of GPCR assays have been designed: cyclic nucleotides such as cyclic adenosine monophosphate (cAMP), inositol triphosphates (IP3 ), mitogen-activated protein kinases (MAPKs) and calcium ions. Second messengers are part of the signal transduction pathway that is initiated when GPCRs are activated. It is the coupling of the receptor to the G protein, particularly the Gα subunit, that determines how the stimulus is transduced in the intracellular environment. Although many Gα subunits have been identiﬁed, they can be divided into four main families, namely Gq/11 , Gi/o , Gs and G12/13 , each family modulating separate effectors in the signal transduction pathway (see Figure 2.1) [7, 8]. GPCR coupling to Gαq and the Gβγ subunits of Gi/o activate PLC-β, whose main function is to hydrolyse PIP2 to give DAG and IP3 . DAG is necessary for activation of PKC, whilst IP3 activates IP3 receptors mediating the release of calcium from intracellular stores. GPCR coupling to Gαs and Gαi/o is linked to activation and inhibition of AC respectively. AC triggers the production of cAMP from adenosine triphosphate (ATP), with increased levels of cellular cAMP leading to the activation of PKA. The activation of the small monomeric G protein Rho can be indicative of G12/13 coupling [9]; however, there is evidence that Gq/11 family members may also activate Rho [10, 11]. Thus, with the exception of the G12/13 family, second messenger assays for IP3 , calcium and cAMP are relatively speciﬁc measurements of functional GPCRs coupling to speciﬁc Gα subunits. It is widely accepted that many GPCRs are capable of coupling to multiple G protein families; as such, screening for the activation of a particular pathway may result in a bias for the modulation of a single aspect of receptor behaviour. Thus, from the perspective of ligand screening, it may be pertinent to assay for a marker of receptor activation that is a convergence of multiple pathways. Dissociated Gβγ subunits, as well as the activation of PKC- and PKA-dependent pathways, can result in the activation of MAPKs, particularly the activation by phosphorylation of ERK1/2) [12–21]. Therefore, assaying for receptor-mediated phosphorylation of ERK1/2 presents a measure of receptor function that represents a pathway with multiple inputs [22].

2.2 Methods and approaches 2.2.1 cAMP cAMP is a tightly regulated signalling molecule involved in the transduction of extracellular signals. The signal is initiated by the binding of a ligand to a GPCR followed by dissociation of the Gα from the βγ and direct interaction of the G protein subunit with the enzyme AC [23]. AC is an enzyme family with at least nine different isoforms having been identiﬁed [24]. The isoforms are expressed to different levels in different tissues, and although they are all activated by forskolin and interact with Gαs , each is under very distinct regulation [25]. AC is a membrane-bound enzyme that

34

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

catalyses the conversion of ATP to cAMP, which in turn modiﬁes cellular function by activating cAMP-dependent PKA enabling the phosphorylation of substrate proteins [26]. cAMP can also cause activation of the transcription factor CREB (cAMP response element binding protein), thereby turning on gene transcription [27]. GPCRs may be linked to the generation of cAMP by one of two routes. One group of receptors (such as β-adrenoceptors, A2 adenosine receptors and D1 dopamine receptors) is associated with an increase in AC activity and elevated cAMP levels mediated through Gαs . Activation of the second group of receptors (including A1 adenosine receptors, glutamate receptors and α2 -adrenoceptors) results in an inhibition of AC and a reduction in cAMP levels mediated through Gi/o proteins. There are many methods available to determine cAMP concentration. All of them rely on the construction of cAMP standard curves against which cAMP levels in the samples can be calculated. Stimulation of the cells with forskolin, which directly activates the AC enzyme, is used to assess the maximum cellular cAMP levels independent of receptor activation. Several methods also include an acetylation step, which increases the sensitivity of the assay. The four main methods are discussed below. 1 Radioimmunoassay (RIA) involves the competition between radioactively labelled [125 I]-cAMP and cellular cAMP in the sample for the binding site of a polyclonal cAMP antibody. Traditional RIAs require multiple steps and can involve handling large quantities of [125 I]-cAMP [28, 29]. More recently, a FlashPlate assay has been developed where the cAMP antibody is afﬁxed to scintillant-coated microplate wells and the counting of the bound fraction is dependent upon the distance of the material to the walls of the wells (see www.perkinelmer.com). This eliminates the need to separate the bound from the free antigen. Results can be read on a MicroBeta or TopCount microplate scintillation counter. 2 Competitive immunoassay kits utilize polyclonal antibodies to cAMP that bind the cAMP in the standards or samples, or to cAMP conjugated to a molecule that will produce either a ﬂuorescent or colorimetric change; for example, alkaline phosphatase, β-galactosidase or horseradish peroxidase [30]. The binding is competitive and the strength of signal generated is inversely proportional to the concentration of cAMP in the sample. Results can be read on any standard plate reader. 3 Fluorescence polarization assays utilize an antibody to cAMP and a ﬂuorescein-labelled cAMP tracer [31]. When the tracer is free in solution it rotates rapidly and when excited with polarized light it emits light of low polarization. However, when the tracer is bound to the cAMP antibody the rotation rate is slowed, causing an increase in polarization. Cellular cAMP in the sample competes with the tracer for the antibody, causing polarization values to decrease with increasing cAMP concentrations. Results can be read on a ﬂuorescence polarization plate reader; for example, Fusion or EnVision (Perkin Elmer). 4 AlphaScreen (Ampliﬁed Luminescent Proximity Homogeneous Assay) (PerkinElmer) contains donor and acceptor beads which generate a signal when they are in close proximity [32]. Upon laser excitation at 680 nm, ambient oxygen is converted to the singlet state by a photosensitizer contained within the

2.2 METHODS AND APPROACHES

35

donor bead. One donor bead can emit up to 50 000 singlet oxygen molecules per second. This results in very high signal ampliﬁcation, allowing assay miniaturization. The acceptor beads contain a thioxene derivative that reacts with the singlet oxygen to generate chemiluminescence at 370 nm. Energy transfer to ﬂuorescent acceptors in the same beads shifts the emission wavelength to 520–620 nm. The singlet oxygen can diffuse approximately 200 nm before it decays. In the case of the cAMP assay, biotinylated cAMP brings the two beads together to generate a signal. Exogenous cAMP from cell lysates competes with the biotinylated cAMP for binding to the acceptor bead, resulting in a decrease in signal with increasing cAMP concentration. Results can be read on a Fusion-α, AlphaQuest microplate analyser or EnVision (Perkin Elmer). The traditional RIA has long been the gold standard for cAMP assays; however, these assays are nonhomogeneous, requiring precipitation and multiple steps, making them unsuitable for handling large numbers of data points, and they have additional implications for health, safety and the environment. The other assay formats available give comparable results to the RIA, and the factors to consider when deciding on an assay format will involve the sensitivity of the assay, the plate reader available, the need for miniaturization and the cost. The AlphaScreen assay is described in Protocol 2.1.

PROTOCOL 2.1

cAMP AlphaScreen Assay

Equipment and Reagents • GPCR-expressing cell line • Versene (Invitrogen) • Phenol-red-free media (Invitrogen) • Haemocytometer • Bench-top centrifuge • Microplate analyser; for example, Fusion-α (PerkinElmer) • Multichannel pipette (Eppendorf) • 384-well, white Optiplates (PerkinElmer) • AlphaScreen cAMP assay kit (PerkinElmer) • Stimulation buffer (50 ml phenol-red-free media (Invitrogen), 50 mg bovine serum albumin (BSA) (Sigma), 100 µl of 500 mM 3-isobutyl-1-methylxanthine (IBMX)a (Sigma)) pH to 7.4 with NaOH • Lysis buffer (0.3% Tween 20 (Sigma), 5 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (Sigma), 0.1% BSA (Sigma)) pH to 7.4 with NaOH • Forskolin (Sigma)

36

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

• 2 × ligand stock solutions in stimulation buffer • cAMP (Sigma) dilutions in lysis bufferb • Topseal adhesive sealing ﬁlm (PerkinElmer).

Method 1 Maintain GPCR-expressing cells in T75 or T175 tissue culture ﬂasks; when 90% conﬂuent, harvest cells. 2 Detach cells with 3 ml versene for 5 min at 37 ◦ C. 3 Spin the cells at 2500 g for 4 min in a bench-top centrifuge, discard supernatant and resuspend the cells in 1.5 ml of phenol-red-free media. 4 Count the number of cells using a haemocytometer. Spin the cells at 2500 g for 4 min and adjust to required cell densityc with freshly made stimulation buffer. 5 Incubate the cells in stimulation buffer for 30 min at 37 ◦ C and 5% CO2 . 6 Following the recommendations in the AlphaScreen protocol, add 5 µl of agonist/antagonist/forskolind in stimulation buffer to the wells of a white 384-well plate to generate dose–response curves. 7 To generate a cAMP standard curve, add 10 µl of cAMP in lysis buffer to the wells of a white 384-well plate. 8 Add 5 µl of cells to all wells except those containing the cAMP standard curve. 9 Cover the plate with topseal to prevent evaporation and spin the plate at 1500 g for 30 s in a bench-top centrifuge. 10 Incubate plate for 30 min at 37 ◦ C.e 11 Add 10 µl lysis buffer to all wells and spin the plate at 1500 g for 30 s in a bench-top centrifuge. 12 Prepare acceptor beads and donor bead/biotinylated cAMP mix as recommended in the AlphaScreen protocol.f Donor bead/biotinylated cAMP mix must be incubated at room temperature for at least 30 min prior to addition to the plate. 13 Add 5 µl of acceptor beads to all wells and cover the plate with topseal to prevent evaporation. 14 Gently tap the bottom of the plate to mix and incubate the plate in the dark for exactly 30 min at room temperature. 15 Add 5 µl of donor bead/biotinylated cAMP mix to all wells and cover the plate with topseal to prevent evaporation. 16 Gently tap the bottom of the plate to mix and incubate the plate in the dark overnight at room temperature.g 17 Read the plate on a microplate analyser.

37

2.2 METHODS AND APPROACHES

18 Plot the AlphaScreen signal as a function of concentration of cAMPh (see Figure 2.2). 100

cAMP [nM]

80 60 40 20 0 −10

−9

−8

−7

−6

−5

−4

−3

Log [Drug] M

Figure 2.2 Typical dose–response curve for AlphaScreen cAMP assay. Chinese hamster ovary (CHO) cells stably transfected with the human glucagon-like-peptide (GLP1) receptor, which preferentially couples to the Gs family of G proteins, were stimulated with forskolin (•) and a receptor agonist (). This graph shows log drug concentration against cAMP concentration. The raw ﬂuorescence units measured in the Fusion-α microplate reader (PerkinElmer) have been converted to cAMP concentration by using a cAMP standard curve. Notes a IBMX

is a potent phosphodiesterase inhibitor. It reduces cAMP degradation during the assay, but it also competes with cAMP to bind the anti-cAMP antibodies, causing a decrease in maximal signal of 30%. IBMX concentration should be limited to 0.2–0.25 mM. b For

the cAMP standard curve, make up a 5 mM stock cAMP solution in phosphate-buffered saline (PBS) and store at −20 ◦ C. Make up a 100 µM stock in lysis buffer and keep on ice. Make up fresh serial dilutions to provide a concentration range from 3 × 10−6 to 3 × 10−11 M in 1/2 log units. There is a dilution of 1 : 3 in the assay plate, giving a ﬁnal concentration range of 1 × 10−6 to 1 × 10−11 M. c Cell

density is very important in this assay and must be optimized for each cell line. Once optimized, the same cell density should be used for all further experiments. As a general guide, 5000–10 000 cells per well (384-well plate) produce a good signal. High cell numbers, >25 000 cells per well, decrease the signal due to the high sensitivity of the AlphaScreen kit. d When

assaying receptors that couple through Gi/o causing inhibition of AC and a reduction in cAMP levels, make up forskolin plus agonist in stimulation buffer. A forskolin concentration producing 50% (EC50 ) of the maximum AC activation will be more sensitive to weak antagonists, while an assay performed at 80% (EC80 ) will give a larger signal window and a more robust assay. Agonists of Gi/o will reverse the effects of forskolin. e Stimulation time is critical for achieving optimal detection of cAMP; once the optimal cell number has been determined, a time course experiment for stimulation should be carried out.

38

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

f Donor

and acceptor beads are light sensitive and should be handled in a subdued light environment. It is possible to dilute the beads from the standard assay protocol. This should be determined empirically.

g Incubation of the plate after addition of donor bead/biotinylated cAMP mix can be from 4 h to overnight. Longer incubation times increase the signal and the signal-to-noise ratio in the assay. h AlphaScreen signal should be plotted as a function of concentration of cAMP due to the nonlinearity of the response in parts of the curve.

2.2.2 Inositol phosphates Coupling of GPCRs to the Gq/11 family of G proteins results in the activation of PLC, which subsequently hydrolyses PIP2 , a phospholipid that is located in the plasma membrane, to give IP3 (also commonly known as triphosphoinositol; abbreviated InsP3 or IP3 ), and DAG. DAG in turn activates PKC, whilst IP3 acts at IP3 receptors to mobilize Ca2+ from intracellular stores. The accumulation of IP3 is commonly used as a measure of GPCR function. An AlphaScreen-based assay that works on the same principle as the cAMP assay has been developed (see Protocol 2.1); however, in this case, cellular IP3 competes with biotinylated IP3 , with increasing amounts of IP3 corresponding to a decreased signal, assessed by a microplate reader. By far the most commonly used and reliable approach to measure the accumulation of IP3 relies on the incorporation or 3 H-inositol into the membrane phospholipids. Thus, if the agonist–receptor system of interest is coupled to the activation of PLCs, agonist stimulation will result in the hydrolysis of 3 H-inositol phospholipids in the plasma membrane, forming 3 H-inositol phosphates (3 H-IP). 3 H-IP is separated from 3 H-inositol and liquid scintillation counting can then be used to quantify the amount of 3 H-IP in a sample. Anion-exchange chromatography is the most commonly used method to isolate the 3 H-IP formed from 3 H-inositol (see Protocol 2.2, based on [33]). The separation of 3 H-IP from 3 H-inositol is time consuming, laborious and generates a signiﬁcant amount of radioactive waste. Semi-automated ﬁltration-based anion-exchange chromatography approaches have been described [34, 35] that are suitable for 96- to 384-well plate formats. More recently, the scintillation proximity assay (SPA ) technology has been used in conjunction with immobilized metal ion afﬁnity chromatography [36] or yttrium silicate [37] to separate 3 H-IP in a 96- to 384-well plate format. The positive charge of the metal ions (or yttrium silicate) binds the negatively charged 3 H-IP, facilitating the separation of 3 H-IP from the neutral 3 H-inositol. The radioactivity can be quantiﬁed using a β-counter capable of handling 96- to 384-well plates; for example, Topcount. The majority of techniques are based on the incorporation of radiolabelled inositol; as such, the handling of radioactive waste is an important consideration in selecting an assay format. Other important considerations are equipment access, the level of throughput required and the cost.

2.2 METHODS AND APPROACHES

PROTOCOL 2.2 Cells

Inositol Phosphate Accumulation Assay: Whole

Equipment and Reagents • GPCR-expressing cell line • Sterile 24-well plates (Nunc) • PBS (Invitrogen) • inositol-free Dulbecco’s modiﬁed Eagle medium (DMEM) (Invitrogen) • 3 H-myo-inositol (Amersham Biosciences) • 1 M NaOH • 1 M formic acid •

14

C-IP (synthesized to order by American Radiolabeled Chemicals)

• HEPES buffer (110 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2 ·2H2 O, 1.0 mM MgCl2 ·7H2 O, 25 mM glucose, 58.4 mM sucrose, pH 7.4) • Li+ buffer (same as HEPES buffer with 10 mM LiCl) • 1 M perchloric acid • 10 × ligand stocks • Resin AG 1-X8 (formate form) (anion-exchange columns, Bio-Rad) • Wash solution (60 mM ammonium formate/5 mM sodium tetraborate) • Elution buffer (1 M ammonium formate/0.1 M formic acid) • Scintillation vials • Scintillation cocktail appropriate for aqueous solutions (e.g. Perkin Elmer HiSafe 3) • Vortex • Liquid scintillation counter.

Method 1 Maintain GPCR-expressing cells in T75 or T175 tissue culture ﬂasks; when 90% conﬂuent, harvest cells and seed into 24-well ﬂat-bottom plates. 2 Grow cells in 24-well ﬂat-bottom plates until 70–80% conﬂuent. 24 h prior to assay, remove serum-containing DMEM by aspiration, wash each well twice with 1 ml PBS and replace with serum- and inositol-free DMEM containing 3 H-myo-inositol (1 µCi ml−1 ). 3 Regeneratea and calibrate columns by the addition of (in the following order): distilled water (dH2 O), 1 M NaOH, dH2 O, 1 M formic acid, dH2 O, dH2 O, 40 µl of diluted 14 C-IP standard.b

39

40

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

4 Add 40 µl of 14 C-IP, and 2 ml elution buffer to three scintillation vials. 5 Aspirate serum-free DMEM from wells,c washing twice with 1 ml HEPES buffer. 6 Add 450 µl of Li+ buffer/welld and then incubate for 15 min at 37 ◦ C. 7 Add 50 µl of 10 × ligand solution to wells and incubate for 30 min at 37 ◦ C. 8 Stop reaction by aspiration of ligand containing assay buffer, and replace with 500 µl of ice-cold 1 M perchloric acid/well. 9 Incubate for 15 min on ice. 10 Transfer supernatant to regenerated anion-exchange columns and allow solution to drip through. 11 Wash columns once with 10 ml of dH2 O, then twice with 10 ml of wash solution. 12 Elute columns with 2 ml elution buffer, collected into scintillation vials.e 13 Add 3 ml of scintillation ﬂuid to each scintillation vial. Cap vials, vortex and count using a dual 14 C/3 H counting protocol. 14 The normalized 3 H-IPb (disintegrations per minute) should be plotted against the log[agonist] M. Concentration–response curves should be ﬁtted to the data. Notes a

When washing and regenerating columns, ensure that each solution has dripped through completely before adding the next solution.

b Addition of 14 C-IP to each column prior to assay provides a control for variations in the efﬁciency of individual columns; as such, the 3 H-IP detected should be normalized against the corresponding 14 C-IP value for each individual sample. However, 14 C-IP is no longer stocked by most suppliers. c Serum-free

DMEM in step 4 is radioactive and should be disposed of in an appropriate manner.

d LiCl

is added to the HEPES-based buffer during the stimulation period as it inhibits inositol monophosphatases, thus allowing the accumulation of 3 H-inositol phosphates [33]. e

After completion of assay, columns should be washed with 1 M NaOH and dH2 O, so that they may be reused. To maintain efﬁciency, columns should be kept moist and regenerated prior to every experiment.

2.2.3 Intracellular calcium Calcium ions are probably the most widely assayed intracellular messengers. In response to many different signals, a rise in the concentration of Ca2+ in the cytosol triggers a variety of events, such as muscle contraction, release of neurotransmitters at synapses, secretion of hormones, activation of T cells and B cells, apoptosis and a variety of biochemical changes mediated by PKC [39]. GPCRs can trigger the release of Ca2+ from the endoplasmic reticulum (ER) via the binding of IP3 to IP3 receptors on the surface of the ER. Activation of GPCRs coupled to Gq cause activation of the

2.2 METHODS AND APPROACHES

41

enzyme PLCβ. PLC-β is a key enzyme in the PIP2 metabolism pathway; it hydrolyses PIP2 into two second messengers: IP3 and DAG [39]. There are many commercially available calcium indicators, and several factors should be taken into consideration when planning an experiment. The range of calcium concentrations that are to be evaluated, the instrumentation available and the loading requirements of the dye must be considered [40]. 1 Fura-2 is a UV-light-excitable ratiometric Ca2+ indicator. Upon binding Ca2+ , fura-2 exhibits an absorption shift that can be observed by scanning the excitation spectrum between 300 and 400 nm, while monitoring the emission at ∼510 nm. Ratiometric measurements of Ca2+ provide an extremely robust assay system, as they are not affected by dye loading, bleaching or illumination intensity. However, the assay does require a microplate instrument containing a UV light source; for example, FLIPR (Molecular Devices). 2 Visible-light-excitable Ca2+ indicators offer several advantages over UV-lightexcitable indicators. The major one is that efﬁcient excitation can be achieved with most laser-based bench-top instrumentation. There is also a reduced interference from sample autoﬂuorescence and less cellular photodamage and light scatter. However, the drawback is that ﬂuorescence intensity depends on several factors not related to calcium concentration, such as acquisition conditions and probe concentration. The two major indicators in this class are Fluo-3 and Fluo-4. Fluo-4 is an analogue of Fluo-3 with the two chlorine substituents replaced by ﬂuorine atoms. Fluo-4 has an increased ﬂuorescence excitation at 488 nm and, consequently, more than double the signal levels of Fluo-3 in microplate screening applications. Fura-2, Fluo-3 and Fluo-4 are available as cell impermeant salts which are useful for microinjection protocols or as acetoxymethyl (AM) esters which can passively diffuse across cell membranes; once inside the cell these esters are cleaved by intracellular esterases to yield cell impermeant ﬂuorescent indicators. AM esters are used in microplate assay formats and are suitable for high-throughput assays requiring robotics and miniaturization. An assay using Fluo-4 is described in Protocol 2.3.

PROTOCOL 2.3

Fluo-4 Calcium Assay

Equipment and Reagents • GPCR-expressing cell line and corresponding complete growth medium without antibiotic • Haemocytometer • Bench-top centrifuge • Fluorescent plate reader; for example, Flex Station (Molecular Devices) • Multichannel pipette (Eppendorf)

42

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

• 96-well, black clear-bottom plates (Nunc) • 96-well, v-bottomed plates (Nunc) • Buffer A (150 mM NaCl, 2.5 mM KCl, 1.2 mM MgCl2 , 1.5 mM CaCl2 , 10 mM HEPES, 10 mM glucose) • Probenecida (Sigma), prepare a 250 mM solution in 1 M NaOH • Assay buffer: add probenecid to buffer A to give a ﬁnal concentration of 2.5 mM; adjust assay buffer to pH 7.4 with NaOH • Fluo-4, AM, 1 mM solution in dimethyl sulfoxide (DMSO) (Invitrogen) • Pluronic F-127b , 20% solution in DMSO (Invitrogen) • Load buffer: mix Fluo-4 AM and Pluronic F-127 in equal volumes and add to assay buffer to give a ﬁnal concentration of 2–4 µM Fluo-4 • Ionomycinc (Sigma) • Uridine 5 -triphosphate (UTP)d (Sigma).

Method 1 Maintain GPCR-expressing cells in T75 or T175 tissue culture ﬂasks; when 90% conﬂuent, harvest cells. 2 Spin the cells at 2500 g for 4 min in a bench-top centrifuge and resuspend the cells in 10 ml of complete growth media without antibiotics. 3 Count the number of cells using a haemocytometer and seed the cells at 30 000–50 000 cells per well in 96-well black clear-bottom plates using 200 µl of cell suspension per well. 4 Incubate the cells in the 96-well plates for 24 h at 37 ◦ C and 5% CO2 . At the time of the assay the cells should be 90–100% conﬂuent. 5 Make up assay buffer on the day of the experiment; warm buffer to 37 ◦ C. 6 Remove media from the cellse by careful aspiration. 7 Wash cells by adding 100 µl prewarmed assay buffer to each well and aspirate. Repeat the wash step. 8 Make up load buffer and load cells by adding 100 µl load buffer to each well. 9 Protect the assay plate from light by wrapping the plate in foil. Return the plate to an incubator at 37 ◦ C and 5% CO2 for 1 h to load. 10 Whilst the plates are loading, prepare the drug plates by making up the appropriate 10× drug dilutions in assay buffer and aliquoting them into a 96-well v-bottomed plate. Include controls in the drug plate, such as assay buffer alone, ionomycin (100 µM) and UTP (1 mM). Keep the drug plate at 37 ◦ C until it is required in the assay. 11 Set up the ﬂuorescent plate reader (excitation at 485 nm and emission at 520 nm, 90 s read time, 20 µl compound transfer after 20 s). Prewarm the machine to 37 ◦ C.

43

2.2 METHODS AND APPROACHES

12 After the cells have loaded, wash to remove Fluo-4 by adding 100 µl prewarmed assay buffer to each well and aspirating. Repeat the wash step. 13 Add 180 µl assay buffer to each well. 14 Place the assay plate, drug plate and tips into the plate reader and run the assay.

Ca++ mobilisation (raw fluorescence units)

15 With this assay set-up you should see a baseline reading in the ﬁrst 20 s. Ionomycin should give a steep increase on addition then plateau at this high level. UTP should give a steep increase on addition then gradually return to baseline (see Figure 2.3). log10[agonist] M = −5

50 000 45 000

log10[agonist] M = −6

40 000

log10[agonist] M = −7

35 000

log10[agonist] M = −8

30 000 25 000

Not stimulated Basal

0

60

120

180

time(sec)

Figure 2.3 Representative Ca2+ mobilization trace using Fluo-4 AM indicator dye. This trace is taken from CHO cells stably transfected with the human M5 muscarinic acetylcholine receptor, which preferentially couples to the Gq/11 family of G proteins. Basal levels of ﬂuorescence are seen in the ﬁrst 10 s. In response to agonist stimulation, Ca2+ mobilization is characterized by an initial peak response, followed by a plateau phase. 16 To calculate a response from your test compounds, ﬁrst calculate the baseline over the ﬁrst 10 s, then calculate your maximum response between 20 and 90 s and use the following formula: [(Max response−Baseline)/(Baseline)] × 100 = percentage response. 17 To assay antagonists, preincubate the cells with the antagonist for 90 s in the plate reader before the addition of agonist.f Notes a Probenecid is a cation exchange inhibitor which stops the bound Ca–Fluo-4 complex from being pumped out of the cell. b Pluronic

F-127 is a nonionic detergent which can assist in dispersion of the nonpolar Fluo-4 AM ester in aqueous media, thereby increasing cell loading. c Ionomycin

is a calcium ionophore and can be used to record a maximum calcium response.

d

UTP acts on P2Y receptors endogenously expressed in most cell lines and can be used as an internal control. e When

washing the cells, handle the plates carefully to avoid dislodging the cells from the bottom of the wells.

44

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

f It may be necessary to perform an agonist dose–response curve in the presence of antagonist to determine the most suitable agonist concentration to use. An agonist concentration producing 50% (EC50 ) of the maximum calcium response will be more sensitive to weak antagonists, while an assay performed at 80% (EC80 ) will give a larger signal window and a more robust assay.

2.2.4 Mitogen-activated protein kinases Activation of cell-surface receptors can result in activation of multiple signalling pathways, many of which are associated MAPK cascades, ultimately resulting in the activation by phosphorylation of ERK1/2 [21, 22]. Gq coupled-receptors generally activate ERK1/2 via PKC and/or PLCβ in both Ca2+ -dependent and -independent manners [41–46]. Gi/o coupled-receptors generally mediate phosphorylated ERK1/2 (pERK1/2) via their βγ subunits [12–14]. Gs activation of ERK1/2 can be via a cAMP/PKA/B-Raf-dependent mechanism [15, 16, 47]. Measurement of a response that is a convergence of multiple pathways and not linked to a speciﬁc G protein subtype increases the likelihood of detecting a range of agonists. A read-out of receptor function that is ampliﬁed and not biased towards a speciﬁc pathway has the propensity to detect weak partial agonism that otherwise may be missed. With respect to ERK1/2, a very speciﬁc antibody is commercially available that can be used to determine relative levels of the doubly pERK1/2 using Western blotting (Protocol 2.4) as well as enzyme-linked immunosorbent assay (ELISA)-based assays. ERK1/2 are each phosphorylated at two sites: Thr202/Tyr204 for ERK1 and Tyr185/Thr187 for ERK2 [48]. In a Western blot, pERK1 and pERK2 can separated based on electrophoretic mobility, as they differ in size; thus, changes in pERK1 or pERK2 can be assessed independently of one another. However, both of these techniques are labour intensive and time consuming, requiring multiple washing steps.

PROTOCOL 2.4 pERK1/2 Detection Using Western Blotting-based Approach Equipment and Reagents • GPCR-expressing cell line • Sterile 6- or 12-well ﬂat-bottom plates (Nunc) • 10 × ligand stocks • Liquid nitrogen • PBS (Invitrogen); 137 mM NaCl, 3 mM KCl, 1.5 mM KH2 PO4 , 8 mM Na2 HPO4 • Serum-free DMEM (Invitrogen)

2.2 METHODS AND APPROACHES

• Lysis buffer (50 mM tris(hydroxymethyl)aminomethane hydrochloride (tris-HCl), 120 mM NaCl, 1 mM ethylenediaminetetraacetate, 50 mM NaF, 10 mM sodium pyrophosphate, 1 mM benzamidine (Sigma), 0.1 mM NaVO4 , 1% (v/v) Igepal (Sigma) and 0.2% (v/v) protease inhibitor cocktail III (Calbiochem)) • 4 × sample buffer (250 mM tris-HCl, 10% (w/v) sodium dodecyl sulfate (SDS), 10% (v/v) β-mercaptoethanol, 30% (v/v) glycerol and 0.05% (w/v) bromophenol blue (Sigma)) • Running buffer (25 mM tris(hydroxymethyl)aminomethane (tris-base), 250 mM glycine, 0.1% (w/v) SDS) • Transfer buffer (25 mM tris-base, 192 mM glycine, 20% (v/v) methanol, pH 8.5) • Protein assay kit (Bio-Rad) • BSA (Sigma) • Mini-Gel apparatus (Bio-Rad) • Gel transfer apparatus (Bio-Rad) • Polyvinylidene diﬂuoride PVDF membrane (Bio-Rad) • Tris-buffered saline containing 0.1% (v/v) Tween-20 (Sigma), 20 mM tris-base, 140 mM NaCl (TBST) • 2 M NaOH • Powdered low-fat skimmed milk • Phospho-ERK antibody (rabbit anti-phospho-p42/44 MAPK antibody, Cell Signalling Technologies) • Total-ERK antibody (rabbit anti-p42/44 MAPK antibody, Cell Signalling Technologies) • Secondary antibody (anti-rabbit horse radish peroxidase conjugated sheep raised immunoglobulin, Chemicon) • ECLTM Western Blotting Analysis System (Amersham) • Hyperﬁlm chemiluminescence ﬁlm (Amersham).

Method 1 Maintain GPCR-expressing cells in T75 or T175 tissue culture ﬂasks; when 90% conﬂuent, harvest cells and seed into 6- or 12-well ﬂat-bottom plates. 2 Allow cells to grow in plates until ∼80% conﬂuent. 3 16–24 h prior to stimulation, remove serum-containing DMEM by aspiration, wash twice with 1 ml/well PBS and replace with 1800 µl/well serum-free DMEM. 4 Prepare 10× ligand stocks to construct concentration response curves, or time courses, as ligands will be diluted (1 : 10) upon addition into plate, 200 µl/well. 5 Stimulate cells by addition of 200 µl/well of ligands for desired time at 37 ◦ C.a

45

46

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

6 After stimulation of cells with ligands, place cells onto ice, aspirate ligand-containing media and replace with 1 ml of ice-cold serum-free media. 7 Scrape cells from wells and transfer the contents of each well to pre-chilled 1.5 ml tubes. 8 Centrifuge (16 800 g) at 4 ◦ C for 1 min, aspirate supernatant, then snap-freeze cell pellet with liquid nitrogen.b 9 Resuspend the pellet in 50–100 µl of lysis bufferc and incubate for 45 min on ice. 10 Centrifuge for 5 min (16 800 g) and transfer the supernatant to a fresh set of 1.5 ml tubes. 11 Determine the protein concentration of each sample using BSA as the standard. 12 Dilute the samples in 15 µl 4× sample buffer, then denature for 5 min at 85 ◦ C. 13 Separate equal amounts of protein (10–20 µg) on a 10% SDS–polyacrylamide mini gel for 80 min at 200 V. 14 Transfer separated proteins to PVDF membrane (pretreated with methanol and pre-equilibrated in transfer buffer), for 35 min at 100 V. 15 Probe the membranes for pERK1/2 with phospho-ERK antibody (0.1% (v/v) in TBST with 5% milk powder) for 45 mind at room temperature, with continual agitation. 16 Wash membranes with TBST three times for 10 min at room temperature, with continual agitation.e 17 Probe with secondary antibody (0.2% (v/v) in TBST containing 5% milk powder) for 1 h at room temperature, with continual agitation. 18 Visualize bands using ECL Western Blotting Analysis System according to manufacturer’s instructions and Hyperﬁlm chemiluminescence ﬁlm.f 19 Membranes can then be stripped with 0.2 M NaOH for 45 min, before probing with total-ERK antibody (0.2% (v/v) in TBST with 5% milk powder) for 1 h. Membranes should be washed, exposed to secondary antibody and bands visualized as before. Notes a Receptor-mediated

pERK1/2 is a transient response and the time course for the peak pERK1/2 response should be established. In addition, the utmost care should be taken when handling cells prior and during stimulation, as shear forces and stress can result in activation ERK1/2 pathways.

b

Snap-frozen cell lysates can be stored at −20 ◦ C for at least 1 month.

c Protease

inhibitors should be added as required to the lysis buffer on day of assay for optimal

activity. d Concentrations

and exposure times for antibodies should be optimized, as should ﬁlm exposure times, as overexposure of ﬁlm may prevent accurate determination of relative band intensities. e To

limit nonspeciﬁc ﬂuorescence, length and number of washes should be optimized between primary and secondary antibody exposure.

2.2 METHODS AND APPROACHES

47

f Changes in the relative intensity of pERK1 and pERK2 can be quantiﬁed using a densitometric approach and may be interpreted singularly or as a whole. Best practice is to ensure loading of equivalent amounts of protein and also to control for the amount of nonphosphorylated ERK1/2 (total-ERK).

More recently, a number of relatively high-throughput assays have been developed for the determination of pERK1/2 based on electrochemiluminescence (MSD, Gaithersburg, MD), infrared ﬂuorescence (LI-COR, Lincoln, NE) and the proximity bead-based AlphaScreen technology (PerkinElmer, Boston, MA) detection systems [49, 50]. The SureFire cellular ERK1/2 AlphaScreen -based assay (TGR Biosciences, Adelaide, Australia), utilizes a two-antibody system, where one antibody is directed against pERK1/2 and the second to an invariant epitope. Only the pERK1/2 will be bound by both antibodies; thus, the AlphaScreen proximity-beads that bind the individual antibodies will only interact when pERK1/2 is present. The SureFire assay is described in Protocol 2.5 and is suitable for the detection of ERK1/2 phosphorylation mediated by both endogenously expressed and transfected (stable and transient) receptors in both adherent and suspended cells [50, 51]. In comparison with the other high-throughput kits available, the SureFire ERK1/2 kit does not require multiple washing steps or cell ﬁxation.

PROTOCOL 2.5 pERK1/2 Detection Using Sure Fire ERK AlphaScreen-based Assay for Adherent Cell Lines Equipment and Reagents • Sterile 96-well plates (Nunc) • PBS (Invitrogen) • Serum-free DMEM (Invitrogen) • 10× ligand stocks • Sure Fire ERK1/2 kit (including lysis buffer, activation buffer, reaction buffer and AlphaScreen beads) (TGR Biosciences) • 384-well, white Optiplates (Perkin Elmer) • Alphascreen plate reader; for example, Fusion-α (Packard BioScience).

Method 1 Maintain GPCR-expressing cells in T75 or T175 tissue culture ﬂasks; when 90% conﬂuent, harvest cells and seed into 96-well ﬂat-bottom plates at a suitable density.a 2 Allow cells to adhere for a minimum of 4 h, before aspiration of serum-containing DMEM, washing twice with PBS (100 µl/well) and then addition of 180 µl/well

48

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

serum-free DMEM (for interaction studies, reduce assay volume to 160 µl/well),16–24 h prior to assay.b 3 Prepare 10 × ligand stocks to construct concentration response curves, or time courses, as ligands will be diluted (1 : 10) upon addition into plate, 20 µl/well. 4 Stimulate cells with ligands (20 µl/well) for desired time; for antagonist interaction studies, a 30 min pre-incubation is recommended, at 37 ◦ C.c 5 Stop the assay by aspiration of ligand-containing serum-free DMEM and addition of lysis buffer, 100 µl/well, agitate for 1–2 min.d 6 Proceed according to manufacturer’s instructions.e

pERK1/2 (raw fluorescence units)

7 Read plate on an AlphaScreen plate reader, using standard AlphaScreen settings. Raw ﬂuorescence units can be corrected for basal ﬂuorescence and/or normalized to a positive control; for example 10% foetal bovine serum (see Figure 2.4). 20000 15000 10000 5000 0 0

5

10

15 20 time(min)

25

30

Figure 2.4 Typical time course for receptor-mediated ERK1/2 phosphorylation. CHO cells stably transfected with the human M2 muscarinic acetylcholine receptor, which preferentially couples to the Gi/o family of G proteins, were stimulated over the course of 30 min with 10% foetal bovine serum (•) and 1 µm acetylcholine (). Receptor-mediated ERK1/2 phosphorylation was measured in duplicate using the AlphaScreen-based SureFire assay, and was characterized as a transient response, peaking at 5 min and returning to baseline levels by 10 min. Notes cell density should ﬁrst be determined for different cellular backgrounds, ∼50 000 cell/well in a 96-well plate is recommended for transfected CHO cell lines. At the time of the assay, cells should be 90–100% conﬂuent.

a Optimal

b

Serum starvation prior to assay is recommended, as this ensures all cells are in a nongrowth phase of the cell cycle, it reduces background and increases the signal to noise. However, it is not a mandatory requirement.

c As

pERK1/2 is generally a transient response, a time course should ﬁrst be established for all ligands, such that consequent dose–response curves can be conducted at the peak time point. lysates and activated lysates may be stored at −20 ◦ C for at least 1 month prior to completing assay. d Cell

2.3 TROUBLESHOOTING

49

e Incorporation of a positive control into the experimental design is recommended that mediates pERK1/2 via a distinct mechanism (e.g. 10% foetal bovine serum). Basal levels of pERK1/2 should also be ascertained by the addition of serum-free media to wells, at the appropriate time point.

2.3 Troubleshooting • The most important aspect of a successful assay is the cells expressing your GPCR of interest. In order to achieve a robust cellular assay your GPCR needs to be expressed at a reasonable level in a cell line with minimal background activity in your chosen assay. GPCRs can be expressed either in a stable cell line or by transient transfection of cells. Transient transfection can give high expression levels and, therefore, a higher signal to background, whereas stable cell lines are important for large-scale screening campaigns where the same cell line is being repeatedly tested. GPCR expression levels in cell lines can often be determined by radioligand binding assays. Transfected receptor DNA constructs can also be modiﬁed to incorporate an epitope that is recognized by an antibody to determine expression. • The coupling proﬁle of GPCR to G protein can be cell line dependent; therefore, where possible, it is necessary to look at more than one output of receptor function or to look at function in a variety of cell backgrounds. • The overexpression of GPCRs in cell lines can lead to coupling through nonphysiological pathways. This is a particular problem when investigating the basic biology of GPCRs and their signalling pathways. However, this phenomenon, together with the use of promiscuous G proteins that will couple to any receptor, has led to the development of ultrahigh-throughput screens for GPCRs where the sole interest is the detection of receptor activation or inhibition. • Cell density is a critical factor in all of the assays and must be optimized for each condition. Cell density will not only affect the strength of the signal, but also the level of background noise detected within the assay. • As with all assays, the inclusion of informative controls, both positive and negative, is crucial. Important controls for all assays are measurement of basal levels, effect of vehicle addition and activation of the second messenger by a method independent of the GPCR of interest, whether this be by an endogenously expressed GPCR or nonreceptor mediated; for example, direct activation of ACs by forskolin. • GPCR overexpression can result in high levels of constitutive activity, thus reducing the signal window, or may result in the downregulation of a signalling pathway. Another concern is the inﬂuence of signal ampliﬁcation, whereby all agonists tested may appear as full agonists due to the inﬂuence of receptor reserve. Treatment with an antagonist/inverse agonist can decrease basal levels of GPCR activity, therefore increasing the signal-to-noise ratio. Receptor alkylation may be utilized to decrease

50

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

the number of binding sites for agonists, and may reveal the partial agonists, as receptor reserve is diminished.

References 1. Fredriksson, R., Lagerstrom, M.C., Lundin, L.G. and Schioth, H.B. (2003) The G-protein-coupled receptors in the human genome form ﬁve main families. Phylogenetic analysis, paralogon groups, and ﬁngerprints. Mol. Pharmacol., 63, 1256–1272. 2. Bourne, H.R. (1997) How receptors talk to trimeric G proteins. Curr. Opin. Cell Biol., 9, 134–142. 3. Cabrera-Vera, T.M., Vanhauwe, J., Thomas, T.O. et al. (2003) Insights into G protein structure, function, and regulation. Endocr. Rev., 24, 765–781. Review covering G-protein structure, function and regulation. 4. Drews, J. (2000) Drug discovery: a historical perspective. Science, 287, 1960–1964. 5. Thomsen, W., Frazer, J. and Unett, D. (2005) Functional assays for screening GPCR targets. Curr. Opin. Biotechnol., 16, 655–665. 6. Hamm, H.E. (1998) The many faces of G protein signaling. J. Biol. Chem., 273, 669–672. 7. Neer, E.J. (1995) Heterotrimeric G proteins: organizers of transmembrane signals. Cell , 80, 249–157. 8. Neves, S.R., Ram, P.T. and Iyengar, R. (2002) G protein pathways. Science, 31, 1636–1639. Review of G-protein signalling pathways. 9. Riobo, N.A. and Manning, D.R. (2005) Receptors coupled to heterotrimeric G proteins of the G12 family. Trends Pharm. Sci., 26, 146–154. 10. Chikumi, H., Vazquez-Prado, J., Servitja, J. et al. (2002) Potent activation of RhoA by Gαq and Gq -coupled receptors. J. Biol. Chem., 277, 27130–27134. 11. Vogt, S., Grosse, R., Schultz, G. and Offermanns, S. (2003) Receptor-dependent RhoA activation in G12 /G13 -deﬁcient cells: genetic evidence for an involvement of Gq /G11 . J. Biol. Chem., 278, 28743–28749. 12. Faure, M., Voyno-Yasenetskaya, T.A. and Bourne, H.R. (1994) cAMP and beta gamma subunits of heterotrimeric G proteins stimulate the mitogen-activated protein kinase pathway in COS-7 cells. J. Biol. Chem., 269, 7851–7854. 13. Crespo, P., Xu, N., Simonds, W.F. and Gutkind, J.S. (1994) Ras-dependent activation of MAP kinase pathway mediated by G-protein beta gamma subunits. Nature, 369, 418–420. 14. Lopez-Ilasca, M., Crespo, P., Pellici, P.G. et al. (1997) Linkage of G protein-coupled receptors to the MAPK signaling pathway through PI 3-kinase gamma. Science, 275, 394–397. 15. Schmitt, J.M. and Stork, P.J.S. (2002) PKA phosphorylation of Src mediates cAMP’s inhibition of cell growth via Rap1. Mol. Cell , 9, 85–94. 16. Schmitt, J.M. and Stork, P.J.S. (2002) Cyclic AMP-mediated inhibition of cell growth requires the small G protein Rap1. J. Biol. Chem., 277, 43024–43032. 17. Brock, C., Schaefer, M., Reusch, H.P. et al. (2003) Roles of Gβγ in membrane recruitment and activation of p110γ/p101 phosphoinositide 3-kinase γ. J. Cell Biol., 160, 89–99.

REFERENCES

51

18. Azzi, M., Charest, P.G., Angers, S. et al. (2003) β-Arrestin-mediated activation of MAPK by inverse agonists reveals distinct active conformations for G protein-coupled receptors. Proc. Natl. Acad. Sci. U. S. A., 100, 11406–11411. 19. Gutkind, J.S. (1998) The pathways connecting G protein-coupled receptors to the nucleus through divergent mitogen-activated protein kinase cascades. J. Biol. Chem., 273, 1839–1842. 20. Davis, R.J. (1993) The mitogen-activated protein kinase signal transduction pathway. J. Biol. Chem., 268, 14553–14556. 21. Werry, T.D., Sexton, P.M. and Christopoulos, A. (2005) ‘Ins and outs’ of seven-transmembrane receptor signalling to ERK. Trends Endocrinol. Metab., 16, 26–33. 22. Goldsmith, Z.G. and Dhanasekaran, D.N. (2007) G protein regulation of MAPK networks. Oncogene, 26, 3122–3142. 23. Hurley, J.H. (1999) Structure, mechanism, and regulation of mammalian adenylyl cyclase. J. Biol. Chem., 274, 7599–7602. 24. Hanoune, J. and Defer, N. (2001) Regulation and role of adenylyl cyclase isoforms. Annu. Rev. Pharmacol. Toxicol., 41, 145–174. 25. Lai, H.L., Lin, T.H., Kao, Y.Y. et al. (1999) The N terminus domain of type VI adenylyl cyclase mediates its inhibition by protein kinase C. Mol. Pharmacol., 56, 644–650. 26. Skalhegg, B.S. and Tasken, K. (1971) Speciﬁcity in the cAMP/PKA signaling pathway; differential expression, regulation, and subcellular localization of subunits of PKA. Front. Biosci., 1, 331–342. 27. Montminy, M. (1997) Transcriptional regulation by cyclic AMP. Annu. Rev. Biochem., 66, 807–822. 28. Steiner, A.L., Kipnis, D.M., Utiger, A. and Parker, C.W. (1969) Radioimmunoassay for the measurement of adenosine 3 , 5 -cyclic phosphate. Proc. Natl. Acad. Sci. U. S. A., 64, 367–373. Original RIA cAMP assay. 29. Steiner, A.L., Parker, C.W. and Kipnis, D.M. (1972) Radioimmunoassay for cyclic nucleotides. I. Preparation of antibodies and iodinated cyclic nucleotides. J. Biol. Chem., 247, 1106–1113. 30. Gabriel, D., Vernier, M., Pfeifer, M.J. et al. (2003) High throughput screening technologies for direct cyclic AMP measurement. Assay Drug Dev. Technol., 1, 291–303. 31. Pope, A.J., Haupts, U.M. and Moore, K.J. (1999) Homogeneous ﬂuorescence readouts for miniaturized high-throughput screening: theory and practice. Drug Discov. Today, 4, 350–362. 32. Ullman, E.F., Kirakossian, H., Singh, S. et al. (1994) Luminescent oxygen channeling immunoassay: measurement of particle binding kinetics by chemiluminescence. Proc. Natl. Acad. Sci. U. S. A., 91, 5426–5430. Description of AlphaScreen methodology. 33. Berridge, M.J., Downes, P.C. and Hanley, M.R. (1982) Lithium ampliﬁes agonist-dependent phosphatidylinositol responses in brain and salivary glands. Biochem. J., 206, 587–595. Anion-exchange chromatography method for IP3 . 34. Chengavala, M., Kostek, B. and Frail, D.E. (1999) A multi-well ﬁltration assay for quantitation of inositol phosphates in biological samples. J. Biochem. Biophys. Methods, 38, 163–170. 35. Tian, Y., Wu, L. and Chung, F. (1997) High throughput 96-well plate assay for receptor-mediated phosphatidylinositol turnover. J. Biomol. Screen., 2, 91–97.

52

CH 2 SECOND MESSENGER ASSAYS FOR G PROTEIN-COUPLED RECEPTORS

36. Liu, J.J., Hartman, D.S. and Bostwick, J.R. (2003) An immobilized metal ion afﬁnity adsorption and scintillation proximity assay for receptor-stimulated phosphoinositide hydrolysis. Anal. Biochem., 318, 91–99. 37. Brandish, P.E., Hill, L.A., Zheng, W. and Scolnick, E.M. (2003) Scintillation proximity assay of inositol phosphates in cell extracts: high-throughput measurement of G-protein-coupled receptor activation. Anal. Biochem., 313, 311–318. 38. Bootman, M.D., Collins, T.J., Peppiatt, C.M. et al. (2001) Calcium signalling - an overview. Semin. Cell Dev. Biol., 12, 3–10. 39. Berridge, M.J. and Irvine, R.F. (1989) Inositol phosphates and cell signalling. Nature, 341, 197–205. 40. Haugland, R.P. (2005) Indicators for Ca2+ , Mg2+ , Zn2+ and other metal ions, The Handbook: A Guide to Fluorescent Probes and Labelling Technologies, 10th edn, Invitrogen Corporation, http://www.invitrogen.com/site/us/en/home/References/Molecular-Probes-The-Handbook.html. Detailed description of calcium-sensitive dyes. 41. Ueda, Y., Hirai, S., Osada, S. et al. (1996) Protein kinase C δ activates the MEK-ERK pathway in a manner independent of Ras and dependent on Raf. J. Biol. Chem., 271, 23512–23519. 42. Schonwasser, D.C., Marais, R.M., Marshall, C.J. and Parker, P.J. (1998) Activation of the mitogen-activated protein kinase/extracellular signal-regulated kinase pathway by conventional, novel, and atypical protein kinase C isotypes. Mol. Cell. Biol., 18, 790–798. 43. Lev, S., Moreno, H., Martinez, R. et al. (1995) Protein tyrosine kinase PYK2 involved in Ca2+ -induced regulation of ion channel and MAP kinase functions. Nature, 376, 737–745. 44. Dikic, I., Tokiwa, G., Lev, S. et al. (1996) A role for Pyk2 and Src in linking G-protein-coupled receptors with MAP kinase activation. Nature, 383, 547–550. 45. Della Rocca, G.J., van Biesen, T., Daaka, Y. et al. (1997) Ras-dependent mitogen-activated protein kinase activation by G protein-coupled receptors. Convergence of Gi - and Gq -mediated pathways on calcium/calmodulin, Pyk2, and Src kinase. J. Biol. Chem., 272, 19125–19132. 46. Kolch, W., Heidecker, G., Kochs, G. et al. (1993) Protein kinase Cα activates RAF-1 by direct phosphorylation. Nature, 364, 249–252. 47. Vossler, M.R., Yao, H., York, R.D. et al. (1989) cAMP activates MAP kinase and Elk-1 through a B-Raf- and Rap1-dependent pathway. Cell , 89, 73–82. 48. Chen, Z., Gibson, T.B., Robinson, F. et al. (2001) MAP kinases. Chem. Rev., 101, 2449–2476. 49. Wong, S.K. (2004) A 384-well cell-based phospho-ERK assay for dopamine D2 and D3 receptors. Anal. Biochem., 333, 265–272. 50. Osmond, R.I.W., Sheehan, A., Borowicz, R. et al. (2005) GPCR screening via ERK 1/2: a novel platform for screening G protein-coupled receptors. J. Biomol. Screen., 10, 730–737. SureFire ERK assay methodology. 51. Lee, H.J., Mun, H., Lewis, N.C. et al. (2007) Allosteric activation of the extracellular Ca2+ -sensing receptor by l-amino acids enhances ERK1/2 phosphorylation. Biochem. J., 404, 141–149.

3 Use of the [35S]GTPγ S Binding Assay to Determine Ligand Efﬁcacy at G Protein-coupled Receptors Elodie Kara and Philip G. Strange School of Pharmacy, University of Reading, Reading, UK

3.1 Introduction The actions of drugs at receptors depend on two events: the binding of the drug to the receptor and the response triggered by the drug at the receptor and in the associated tissue. The binding of the drug is reﬂected in the afﬁnity with which the drug binds to the receptor. The ability of the drug to alter the activity of signalling systems linked to the receptor is often referred to as ‘efﬁcacy’ and is reﬂected in differences in the extent and potency of the response. We can envisage a scale of efﬁcacy from positive for agonists to negative for inverse agonists with neutral antagonists having zero efﬁcacy. In this chapter we shall consider methods for assessing efﬁcacy for agonists for the G protein-coupled receptors (GPCRs).

3.1.1 Efﬁcacy and its measurement It is very important to understand and quantify the efﬁcacy of drugs, as this has an important bearing on drug action. For example, several drugs with very low but positive

G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

54

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

efﬁcacy (partial agonists) have been introduced recently and found to have useful therapeutic properties. Examples here are buprenorphine (used for opiate addiction), buspirone (used for anxiety) and aripiprazole (used for schizophrenia) [1]. It seems that the very low agonist efﬁcacy expressed by these drugs is important for their actions, and so systems that can deﬁne this efﬁcacy are required. As efﬁcacy refers to the functional effects of a drug it must be deﬁned using functional tests. The effects of a set of drugs are, therefore, tested in a concentration/response experiment using a suitable functional response linked to the receptor. Responses used in the past have typically been intact tissue responses, such as smooth muscle contraction. More recently, however, second messenger assays (e.g. cyclic adenosine monophosphate) have been widely used (see Chapter 2). Newer biophysical assays for changes in receptor conformation (e.g. see [2]) are showing promise as well. For agonists, several measures of efﬁcacy are then accessible from this kind of experiment. First, the maximal effects Emax of the drugs can be determined. Comparisons of the Emax values for a set of drugs provide a useful measure of relative efﬁcacy, but this measure fails where the drugs produce a full (100%) response in the test system. For drugs that produce a full response, termed full agonists, it will be impossible to differentiate differences in relative efﬁcacy, and so other measures are required. One such measure is the KA /EC50 ratio [3–5]. This ratio provides an estimate of the extent to which the agonist response curve is shifted away from the agonist binding curve. The EC50 represents the concentration of agonist that produces a half-maximal response and KA is the dissociation constant for agonist binding to the receptors (also the concentration of agonist that occupies half the receptors). The KA /EC50 ratio provides a good estimate of relative efﬁcacy for agonists with moderate to full efﬁcacy but approaches unity for the weaker agonists and so provides little discrimination in this efﬁcacy range. The two measures may be combined to produce the parameter Emax KA /EC50 . This provides a continuous measure of efﬁcacy over the full range of agonist efﬁcacy (P.G. Strange unpublished results; see also [6]) and so is a very useful parameter. When it is used to analyse real data in the literature it also provides an excellent measure of efﬁcacy [7].

3.2 Methods and approaches 3.2.1 In vitro systems to measure agonist efﬁcacy: the [35 S]GTPγ S binding assay Whereas drug actions at many receptors can be assayed using functional tissue responses, it is useful to use in vitro tests based on cell-free preparations of receptors, expressed in recombinant cells for example. One very popular assay is the guanosine 5 -O-(3-thiotriphosphate) [35 S]GTPγ S binding assay (e.g. see [8]). This assay takes advantage of the normal G protein cycle associated with GPCRs [9]. In the G protein cycle, the inactive G protein (G) with guanosine diphosphate (GDP) bound

3.2 METHODS AND APPROACHES

55

is converted to the active G protein by combination with the agonist-occupied receptor (AR) to form the ternary complex ARG·GDP. In this complex, GDP release is facilitated and guanosine triphosphate (GTP) binds to activate the G protein. Dissociation of the G protein into α·GTP and βγ subunits occurs and these alter effector activity; for example, adenylyl cyclase. In the [35 S]GTPγ S binding assay, the GTP binding event is assayed using the nonhydrolysable analogue of GTP, [35 S]GTPγ S, and in some cases Eu-labelled GTPγ S (e.g. see [10]). The [35 S]GTPγ S binding assay is described in detail in Protocols 3.1–3.5. The preparation of membranes is initially described (Protocol 3.1). To obtain an optimal level of agonist-stimulated [35 S]GTPγ S binding, it is usually necessary to add GDP to assays. GDP reduces the basal [35 S]GTPγ S binding level by competing for [35 S]GTPγ S binding sites on G proteins and other nucleotide binding proteins. To determine the concentration of GDP to add in your experiment, the stimulation of [35 S]GTPγ S binding over the basal level by saturating agonist concentration should be calculated for different GDP concentrations. This is described in Protocol 3.2; the analysis forms Protocol 3.3.

PROTOCOL 3.1 Preparation of Membranes from Recombinant Cells Expressing G Protein-coupled Receptors Equipment and Reagents • GPCR-expressing cell line • Glass beads (2 mm diameter) (Sigma) • 50 ml plastic tubes (Greiner) • 50 ml Nalgene ultracentrifugation tubes (Fisher) • Ultra-Turrax homogenizer • Centrifuge (for 365 g) • Ultracentrifuge (for 31 500 g) • Buffer: 20 mM N-(2-hydroxyethyl)piperazine-N -ethanesulfonic acid (HEPES), 1 mM ethylene glycol bis(2-aminoethylether)-N,N,N ,N -tetraacetic acid (EGTA), 1 mM ethylenediaminetetraacetic acid (EDTA), pH 7.4. The buffer must be kept at 4 ◦ C until the experiment starts • 10% trichloroacetic acid • Lowry protein assay reagents: buffer A (2% Na2 CO3 , 0.1 M NaOH), buffer B (0.5% CuSO4 · 5H2 O, 1% NaK tartrate) and buffer C (Folin–Ciocalteu’s phenol reagent diluted to 1 : 1 in water).

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

56

Method The protocol provided here is for adherent cells cultured in 175 cm2 culture ﬂasks; for example, Chinese hamster ovary (CHO) cells expressing a recombinant GPCR. If the cells used are nonadherent, then put them directly in a 50 ml plastic tube and proceed to step 6. In order to obtain a good yield of protein at the end of the experiment, the adherent cells should be ∼95% conﬂuent when the membrane preparation is done. The ﬂasks are handled conveniently in batches of ﬁve. 1 Remove the medium from the tissue-culture ﬂasks and discard. 2 Add 7 ml buffer to each ﬂask to wash the cells and discard. 3 Add 5 ml buffer to each ﬂask. Add the glass beads to the ﬁrst ﬂask and shake. Then, add the beads and cells to the second ﬂask. Repeat this for all ﬁve ﬂasks. Cells may also be detached using phosphate-buffered saline (PBS)/4 mM EDTA or using a cell scraper. 4 Take the liquid from ﬂasks and add it to one 50 ml plastic tube (at this step, the total volume is around 25 ml). The samples should be kept in ice throughout. 5 Wash the ﬁve ﬂasks with 10 ml buffer, transferring the 10 ml from ﬂask to ﬂask and adding it to the 50 ml tube (the total volume of the tube at this step would be 25 ml from step 4 and 10 ml from step 5, giving a total volume of 35 ml). 6 Homogenize the cells and the buffer in the 50 ml plastic tube with the Ultra-Turrax homogenizer at 24 000 rpm, four times for 5 s each. 7 Centrifuge the tubes at 365 g for 10 min at 4 ◦ C. 8 Remove the supernatant and put it into 50 ml ultracentrifuge tubes. Balance the tubes. 9 Centrifuge at 31 500 g for 1 h at 4 ◦ C. 10 Discard the supernatant. Resuspend the pellet in 2 ml of buffer using the Ultra-Turrax homogenizer at 16 000 rpm for 3 s. 11 Wash the centrifuge tube with 500 µl of buffer and add this to membrane suspension. The total volume of membrane preparation from ﬁve 175 cm2 ﬂasks is 2.5 ml. If your cells were ∼95% conﬂuent, then the protein concentration should be around 3 mg ml−1 . 12 Divide into aliquots of 300 µl and store at −80 ◦ C. Special precautions are not required to freeze the samples. 13 Determine the protein concentration in the sample, protein (∼30 µg) should be precipitated using 10% trichloroacetic acid before determining protein concentration as described by Lowry et al. using bovine serum albumin (BSA) (10–100 µg) as a standard [11]. The precipitation step is employed to remove substances that interfere in the protein assay. Brieﬂy, add 1 ml of 10% trichloroacetic to samples, vortex mix and centrifuge 15 min at 2500 g, at room temperature. Discard the supernatant and allow tube to drain well. Ten minutes later, dilute Lowry B solution 1/50 (v/v) in Lowry A solution, to 1 : 50, and add 1 ml of this mixture into the tubes. Vortex mix and incubate 10 min at room temperature. Then add 100 µl of Lowry C and incubate 10 min more.

3.2 METHODS AND APPROACHES

Finally, add 2 ml of H2 O, vortex mix and incubate 30 min at room temperature. Read the absorbance at 760 nm. From the standard concentration–absorbance curve obtained, determine the protein concentration of your samples.

PROTOCOL 3.2 Determination of the Concentration of GDP Required in the [35 S]GTPγ S Binding Assay To obtain an optimal level of agonist stimulated [35 S]GTPγ S binding, it is usually necessary to add GDP to assays. GDP reduces the basal [35 S]GTPγ S binding level by competing for [35 S]GTPγ S binding sites on G proteins and other nucleotide binding proteins. To determine the concentration of GDP to add in your experiment, the stimulation of [35 S]GTPγ S binding over the basal level by saturating agonist concentration should be calculated for different GDP concentrations.

Equipment and Reagents • 5 ml round-bottom polystyrene tubes (Fisher) • Water bath • Brandel cell harvester (Semat International) (directions below are for a 24-place harvester) • GF/C glass-ﬁbre ﬁlters (Whatman – Schleicher and Schuell) • Scintillation tubes and caps • Scintillation ﬂuid (Ultima Gold, Perkin–Elmer) • [35 S]GTPγ S (GE Healthcare) (37 TBq mmol−1 ) • 10 × concentrations of GDP stock solutions • 10 × stock solutions of drugs required to stimulate the receptor • [35 S]GTPγ S binding assay buffer: 20 mM HEPES, 10 mM MgCl2 , 100 mM NaCl, pH 7.4. Keep at 4 ◦ C until the experiment starts • PBS: 140 mM NaCl, 10 mM KCl, 1.5 mM KH2 PO4 , 8 mM Na2 HPO4 . Keep at 4 ◦ C until the experiment starts.

Method 1 Prepare two sets of tubes: one for the assessment of the basal binding level and one for the measurement of the agonist stimulated binding. It is recommended to determine each data point in triplicate. That represents six tubes per concentration of GDP tested; Figure 3.1 shows a suggested rack layout.

57

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

58

GDP concentrations 1

2

3

4 basal stimulated

Figure 3.1 Rack organization for experiment to determine GDP dependency of [35 S]GTPγ S binding assay. 2 Prepare GDP at different concentrations (e.g. from 10−4 to 10−7 M) in assay buffer. The solutions should be prepared at 10 × the ﬁnal concentration in the assay, as 100 µl will be added in 1 ml ﬁnal assay volume. 3 Prepare the agonist solution(s). A saturating concentration should be used. Again, this solution must be prepared at 10 × the ﬁnal concentration as 100 µl will be added in a ﬁnal volume of 1 ml. 4 Prepare the protein suspension obtained as in Protocol 3.1. In recombinant systems expressing between 0.5–10 pmol mg−1 of receptors, it is recommended to add 20 µg of protein per tube. As 100 µl of protein solution will be added to the ﬁnal volume of 1 ml, the suspension prepared must be at 0.2 µg of protein per microliter.a 5 Prepare the [35 S]GTPγ S solution at a concentration of 1 nM. The ﬁnal concentration in the tubes will be ∼100 pM. The GraphPad radioactivity calculator is accessible online and may help to calculate the accurate amount of [35 S]GTPγ S to use based on the date of manufacture and half-life of 35 S (http://graphpad.com/quickcalcs/radcalcform. cfm). 6 In 5 ml round-bottom polystyrene tubes, add: (a) 700 µl of buffer in the tubes for the assessment of the basal and 600 µl in the tubes for the stimulated points (see Figure 3.1 for suggested layout of rack). (b) 100 µl of GDP in both ‘basal’ and ‘stimulated’ tubes for each concentration. (c) 100 µl of agonist in ‘stimulated’ tubes. (d) 100 µl membrane preparation in all tubes. Vortex mix and incubate in a water bath at 30 ◦ C for 30 min. At this time, the tubes contain:

Buffer GDP Agonist Membrane preparation

Basal

Stimulated

700 µl 100 µl – 100 µl

600 µl 100 µl 100 µl 100 µl

3.2 METHODS AND APPROACHES

7 Add 100 µl [35 S]GTPγ S solution to all the tubes. Vortex mix and incubate 30 min at 30 ◦ C. 8 Stop the reaction by ﬁltration through Whatman GF/C glass-ﬁbre ﬁlters using a Brandel cell harvester. Wash four times with 3 ml ice-cold PBS. 9 Place the ﬁlter disk in a scintillation tube and add 2 ml of scintillation liquid. Cover the tube with a cap, vortex mix, and incubate for at least 6 h. 10 Vortex mix the tubes and determine the radioactivity by liquid scintillation spectrometry. Notes For cells expressing at higher levels, less protein could be added (<10 µg). In that case, there may be loss of protein by adsorption to tubes, but this can be avoided by the addition of 0.1% BSA to all solutions. a

PROTOCOL 3.3

Data Analysis of GDP Requirements

Equipment • Computer program to analyse the data, such as GraphPad Prism 4.0.

Methods Two sets of data are obtained for each GDP concentration tested: the basal [35 S]GTPγ S binding level and the binding induced by the agonist stimulation, all expressed in disintegrations per minute (dpm). 1 Check that the values in your triplicate are similar. Then, take the average of the triplicates for all the data points. 2 In a GraphPad prism data sheet, enter the GDP concentrations in log[M] (x values), and the basal and stimulated values in two different columns (y values). Analyse it as a dose–response curve. Two curves will be obtained. The best GDP concentration to use is the concentration for which the difference between the basal and the stimulated values is greatest. That will represent the best stimulated over basal ratio. For example, on yeast (Pichia pastoris) membranes expressing a fusion protein involving the D2S receptor and the Gαo subunit of the G protein, the best GDP concentration to use in order to obtain a good stimulated over basal signal is 10−7 M (Figure 3.2). For the D2S receptor expressed in CHO cells, GDP should be at 10−6 –10−5 M depending on the cell preparation [12, 13].

59

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

60

[35S]GTPγ S binding (dpm)

5000

basal stimulated 2500

0

−8

−7

−6

−5

−4

−3

log[M] GDP

Figure 3.2 GDP dependency curve for [35 S]GTPγ S binding assay obtained on P. pastoris membranes expressing a fusion protein involving the D2S dopamine receptor and the Gαo subunit of the G protein. The data shown are for the [35 S]GTPγ S binding obtained under basal conditions and with stimulation by dopamine (100 µm). 3 To quantify the difference between the basal binding and the binding induced by agonist stimulation, calculate the ratio of stimulated over the basal for each GDP concentration, express it as a percentage and represent it on a graph (e.g. see Figure 3.3).

% of stimulation over basal

400

300

200

100

0

−8

−7

−6

−5

−4

log[M] GDP

Figure 3.3 GDP dependency curve for [35 S]GTPγ S binding assay obtained on P. pastoris membranes expressing the D2S dopamine receptor: Gαo fusion protein. The data are expressed as percentage stimulation over basal.

Assays are typically run as single time-point assays where membranes of cells expressing the receptor are mixed with agonist and [35 S]GTPγ S (∼100 pm) and the binding of [35 S]GTPS is assessed after 30 or 60 min (Protocol 3.4; the analysis is described in Protocol 3.5). The very low concentration of [35 S]GTPγ S used ensures

3.2 METHODS AND APPROACHES

61

that the rate of the agonist-stimulated [35 S]GTPγ S binding is measurable. This is a very convenient design and it can be run in the same format as used for ligand binding assays using membranes of recombinant cells expressing the receptor in question with ﬁltration to stop the assay and collect bound [35 S]GTPγ S. The assay works best for Gi/o proteins, but attempts have been made to use the assay for systems dependent on other G proteins [14]. The assay has also been used for puriﬁed receptors and G protein reconstituted into vesicles [15], and for receptors in living cells if the membranes are permeabilized to allow penetration of the [35 S]GTPγ S [16].

PROTOCOL 3.4 Agonist Stimulation of [35 S]GTPγ S Binding This assay, which examines the stimulation of [35 S]GTPγ S binding by agonists, allows the efﬁcacy and the potency of a range of ligands to be determined for a GPCR.

Equipment and Reagents • 5 ml round-bottom polystyrene tubes (Fisher) • Water bath • Brandel cell harvester (Semat International) (directions below are for a 24-place harvester) • GF/C glass-ﬁbre ﬁlters (Whatman – Schleicher and Schuell) • Scintillation tubes and caps • Scintillation ﬂuid (Ultima Gold, Perkin–Elmer) • [35 S]GTPγ S (GE Healthcare) (37 TBq mmol−1 ) • 10 × solution of GDP • 10 × stock solutions of drugs required to stimulate the receptor • [35 S]GTPγ S binding assay buffer: 20 mM HEPES, 10 mM MgCl2 , 100 mM NaCl, pH 7.4. Keep at 4 ◦ C until the experiment starts. • PBS: 140 mM NaCl, 10 mM KCl, 1.5 mM KH2 PO4 , 8 mM Na2 HPO4 . Keep at 4 ◦ C until the experiment starts.

Methods 1 Prepare the 10 × agonist dilutions in the assay buffer.a 2 Prepare the 10 × GDP solution at the concentration determined by the GDP dependency curve. 3 Dilute the membrane preparation in buffer, in order to have a concentration at 0.2 µg of protein per microlitre.

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

62

4 The following steps of this method are the same as the GDP dependency experiment, Protocol 3.2. Brieﬂy, in all tubes, add: (a) 700 µl of buffer in the three tubes (which will represent the basal binding level; Figure 3.4 shows a suggested rack layout) and 600 µl in all others. (b) 100 µl GDP (c) 100 µl drug (d) 100 µl membrane preparation. 5 Vortex mix and incubate 30 min in a water bath at 30 ◦ C. 6 During the 30 min incubation in step 5, prepare the [35 S]GTPγS solution at a concentration of 1 nM. The ﬁnal assay concentration will be ∼100 pM. 7 Add 100 µl of [35 S]GTPγ S. Vortex mix and incubate 30 min in a water bath at 30 ◦ C. 8 Stop the reaction by ﬁltration through GF/C glass-ﬁbre ﬁlters using a Brandel cell harvester. Wash four times with 3 ml ice-cold PBS. 9 Place the ﬁlter disk in a scintillation tube and add 2 ml of scintillation liquid. Close the tube with a cap, vortex mix and incubate for at least 6 h. 10 Vortex mix the tubes and determine the radioactivity by liquid scintillation spectrometry. basal

1

2

3 1-7 = drug concentrations

4

5

6

7

Figure 3.4 Rack organization for [35 S]GTPγ S binding assay. Notes a The

drug concentrations used should cover a range between the concentration of drug without any effect on the receptor and a saturating concentration which will have the full effect of the drug. This range may need to be determined for a new drug by testing a wide range of concentrations in an initial experiment and in subsequent experiments using a narrower range. A Brandel cell harvester for 5 ml tubes usually contains 24 aspiration units, so that 24 tubes can be harvested at the same time. In order to harvest all the tubes together, it is recommended to use seven different concentrations of drug, each concentration done in triplicate (that makes 7 × 3 = 21 tubes). The ﬁrst three tubes of your rack will not contain any agonist and will represent the basal binding level, which is essential to calculate the stimulation by the ligand (see Figure 3.4 for a suggested layout of racks).

63

3.2 METHODS AND APPROACHES

PROTOCOL 3.5 Data Analysis for Agonist Stimulation of [35 S]GTPγ S Binding Materials • Computer program to analyse the data, such as GraphPad Prism 4.0.

Methods 1 Express the radioactivity (in dpm) obtained for the stimulated points as a percentage of the basal binding level. 2 Open a data sheet in GraphPad Prism and enter the agonist concentrations used as x values (in log[M]) and the percentages of stimulation obtained for each drug concentration as y values. 3 Analyse the data using nonlinear regression. Use the equations for a sigmoidal dose–response curve with Hill coefﬁcient of one and with variable slope and compare the ﬁts using an F test (e.g. see Figure 3.5, where the data were ﬁtted best by a model with a Hill slope of one).a drug 1 = full agonist

700

drug 2 = partial agonist 600

drug 3 = partial agonist, more potent than the 2 other ligands

% of basal

500 400

Emax

300 200 100 0 −10.0

−7.5

−5.0

−2.5

0.0

log [M] ligand

EC50 drug3

EC50 drug1

Drug potency

Figure 3.5 Examples of concentration–response curves obtained by using [35 S]GTPγ S binding assays. The graph shows concentration–response curves for three agonists that have different maximal effects E max and potencies EC50 .

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

64

4 The maximal effects for each agonist in the assay may be compared to calculate their relative efﬁcacies. The relative efﬁcacy for different agonists is calculated as the maximal stimulation by the agonist expressed as a percentage of the maximal stimulation induced by the reference full agonist for a given receptor. To calculate the relative efﬁcacies, subtract the basal [35 S]GTPγ S binding from the [35 S]GTPγ S binding induced by the highest concentration of agonist. The value obtained for the full agonist represents 100% efﬁcacy. Express the maximal stimulation induced by the other agonists tested as a percentage of this value. 5 The values obtained in [35 S]GTPγ S binding experiments can also be converted to concentration of [35 S]GTPγ S bound per milligram of membrane protein. Convert the [35 S]GTPγ S bound from dpm to femtomoles, taking account of the speciﬁc activity of the radioisotope. Divide this value by the amount of protein in milligrams added into the assay to obtain the amount of [35 S]GTPγ S bound in femtomoles per milligram. Notes a

From the ﬁtting, the maximal effect of the agonist Emax is obtained from the upper asymptote of the curve. The ﬁtting also generates the EC50 for each agonist tested. The EC50 is the agonist concentration that gives half of the maximal effect. This value represents also the potency of a drug: the smaller the EC50 , the more potent is the agonist. For example, in Figure 3.5, drug 3 is more potent than the drug 1.

The assay has been used with a number of GPCRs, including opiate [17], cannabinoid [18], muscarinic acetylcholine [19] and CCR5 chemokine [20], and we have been using it to proﬁle ligands at the D2 dopamine receptor [3, 4, 12, 13, 21]. Based on concentration–response curves for a range of agonists and using Emax values, we have deﬁned a scale of efﬁcacy (Figure 3.6). The assay can be rather insensitive for deﬁning responses of low-efﬁcacy partial agonists or inverse agonists, and these may appear as antagonists in the normal assay conﬁguration. We have shown, however, that by substituting the standard 100 mm sodium ions in the assay for N -methyl-d-glucamine that the detection of partial agonism is considerably improved [22] and additional removal of GDP improves detection of inverse agonists for the D2 dopamine receptor [23]. For other GPCRs, inverse agonism may be detected under normal assay conditions (e.g. the 5-HT1A serotonin receptor and chemokine receptor CCR5 [24, 25]), and this may depend on the level of constitutive activation of the receptor concerned. It is important, however, to ask whether the relative efﬁcacy values determined using this assay reﬂect relative efﬁcacy values in other assays, including in vivo assays. This requires that we compare efﬁcacy values in different assays for one receptor. There are actually very few sets of data that enable this comparison to be performed. For the D2 dopamine receptor we have compared data for the Emax values for a range of ligands in [35 S]GTPγ S binding assays and in electrophysiology assays [26]. There is some agreement between the two sets of data (Figure 3.7), but the range of efﬁcacies is not really broad enough to allow a clear conclusion to be drawn. Nevertheless, the efﬁcacy values derived from using the [35 S]GTPγ S binding assay seem to be in agreement with relative efﬁcacies seen in other assays. For example, aripiprazole is a very low efﬁcacy partial agonist compared with dopamine in [35 S]GTPγ S binding

3.2 METHODS AND APPROACHES

65

response (% of dopamine)

100 80 60 40 20 0 d N m opa PA br -ty mi om ra ne m b- dih ocr ine ph yd ip en re tin yl xid e et in h p- yla e ty m ra in e (-) min -3 e ap -PP ar lind P ip o ip re ra z U ole H -2 3 AJ 2 -7 6

−20

Figure 3.6 A scale of efﬁcacy for agonist effects at the D2 dopamine receptor. The data show the maximal effects in [35 S]GTPγ S binding assays for a range of agonists expressed relative to dopamine. (NPA: N-propyl-norapomorphine hydrochloride; S-(−)-3-(3-hydroxyphenyl)N-propylpiperidine hydrochloride; UH-232: cis-(+)-5-methoxy-1-methyl-2-(di-N-propylamino) tetralin; AJ-76: (1S,2R)-cis-5-methoxy-1-methyl-2-(N-propylamino)tetralin.) The data are taken from [21, 22].

assays, whereas haloperidol exhibits inverse agonist effects in this assay [22, 23]. Aripiprazole and haloperidol have also been tested for their effects on prolactin secretion in experimental animals. Whereas haloperidol elicits a large increase in prolactin secretion, aripiprazole has only partial effects [27], in agreement with its efﬁcacy lying more towards the neutral point of the efﬁcacy scale.

3.2.2 Other assays for efﬁcacy at GPCRs Efﬁcacy of drugs at GPCRs may be assessed using a variety of different assays downstream of G protein activation. For example, for the D2 dopamine receptor, inhibition of adenylyl cyclase is a convenient assay for drugs at the receptor [13]. Alternatively, activation of extracellular signal-regulated kinase/mitogen-activated protein (MAP) kinase 120 100 Emax

80 60 GTPgammaS electrophysiology

40 20

le

in

ap

do

pi

ro

PP -P

qu

)-3 (+

pa om min or e ph (-) ine -3 -P PP

0

Figure 3.7 Comparison of efﬁcacy in [35 S]GTPγ S binding assays and in electrophysiology assays. The data are the maximal effects in the two assays and are taken from [3, 21, 26, 28].

66

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

provides another readout of the receptor [29] (see also Chapter 2, Protocols 3.4 and 3.5). As these assays occur after G protein activation, they tend to be well ampliﬁed. In consequence, many drugs will appear as full agonists on these responses. This means that Emax is not a good discriminator of agonist efﬁcacy, and other measures (e.g. KA /EC50 and Emax KA /EC50 ) may provide better measures of efﬁcacy. A further complication to these measurements is the apparent pathway dependence of some efﬁcacy measures. For example, for the D2 dopamine receptor, it has been reported that the pharmacological proﬁle of agonist responses for inhibition of adenylyl cyclase is different from that for stimulation of MAP kinase [29]. Many other examples of this apparent pathway dependence of efﬁcacy have been reported [30], and this needs to be borne in mind in making such measurements.

3.3 Troubleshooting • The [35 S]GTPγ S binding assay works well under the conditions described here for GPCRs coupled to Gi/o proteins. For GPCRs coupled to other G proteins, it may be necessary to employ an immunoprecipitation step to observe good signals over basal. • Membranes expressing GPCRs should be thawed for the assay and discarded after use. Repeated freeze–thaw cycles should never be employed. • The [35 S]GTPγ S will decay according to its half-life. This will lead to a gradual reduction in the chemical concentration (which can be compensated for; see Protocol 3.2) as well as radiolytic damage. After some time, some samples of [35 S]GTPγ S do not work well in the assay. • We ﬁnd that consecutive preparations of membranes do not behave identically. Basal and stimulated levels of [35 S]GTPγ S binding differ from preparation to preparation, so it is important to characterize each preparation carefully. Such variation in basal levels of [35 S]GTPγ S binding makes it difﬁcult to compare different receptors for the level of constitutive activation based on comparison of basal levels. • The assay may be performed in a ﬁnal volume of 200 µl in a 96-well plate if reagents are in short supply and a 96-well cell harvester is available. In this case, all of the volumes should be scaled down in proportion. If low protein concentrations are used (<10 µg ml−1 ) then 0.1% BSA should be included in assays to prevent protein loss by adsorption to tubes. For receptors for peptides and small proteins (e.g. chemokine receptors) it may be advantageous to include BSA in assays to prevent loss of the ligands. • The 30 min preincubation of agonist and membrane may be omitted without changing the results obtained. In this case, agonist, membrane and [35 S]GTPγ S are mixed and incubated for 30 min before terminating the assay. A preincubation will be required when antagonists are tested to allow equilibration of agonist and antagonist. • Where large assays spread over several racks are performed, addition of [35 S]GTPγ S should be staggered to allow for delays in ﬁltering racks of tubes.

REFERENCES

67

• Whereas many agonist stimulation curves in [35 S]GTPγ S binding assays give data that ﬁt well to a model of a sigmoidal dose–response curve with a Hill coefﬁcient of one, this is not always the case, and the more complex model with Hill coefﬁcient different from one may ﬁt better. This may be related to the mechanistic basis of the assay.

Acknowledgements We thank BBSRC for supporting this work and Cedric Fiez-Vandal for supplying membranes from P . pastoris expressing D2 -Go fusion proteins.

References 1. Ohlsen, R.I. and Pilowsky, L.S. (2005) The place of partial agonism in psychiatry: recent developments. J. Psychopharmacol., 19, 408–413. 2. Hoffmann, C., Gaietta, G., Bunemann, M. et al. (2005) A FlAsH-based FRET approach to determine G protein-coupled receptor activation in living cells. Nat. Methods, 2, 171–176. 3. Gardner, B. and Strange, P.G. (1998) Agonist action at D2(long) dopamine receptors: ligand binding and functional assays. Br. J. Pharmacol., 124, 978–984. 4. Gardner, B.R., Hall, D.A. and Strange, P.G. (1997) Agonist action at D2(short) dopamine receptors determined in ligand binding and functional assays. J. Neurochem., 69, 2589–2598. 5. Black, J.W. and Leff, P. (1983) Operational models of pharmacological agonism. Proc. R. Soc. Lond. B Biol. Sci., 220, 141–162. 6. Trzeciakowski, J.P. (1999) Stimulus ampliﬁcation, efﬁcacy, and the operational model. Part I – binary complex occupancy mechanisms. J. Theor. Biol., 198, 329–346. 7. Strange, P.G. (2007) Mechanisms underlying agonist efﬁcacy. Biochem. Soc. Trans., 35, 733–736. A review on the importance of considering different parameters in efﬁcacy assessment. 8. Harrison, C. and Traynor, J.R. (2003) The [35 S]GTPγ S binding assay: approaches and applications in pharmacology. Life Sci., 74, 489–508. An overview on the GTPγ S binding technique. 9. Birnbaumer, L., Abramowitz, J. and Brown, A.M. (1990) Receptor–effector coupling by G proteins. Biochim. Biophys. Acta, 1031, 163–224. 10. Labrecque, J., Anastassov, V., Lau, G. et al. (2005) The development of an europium–GTP assay to quantitate chemokine antagonist interactions for CXCR4 and CCR5. Assay Drug Dev. Technol., 3, 637–648. 11. Lowry, O., Rosebrough, N., Farr, A. and Randall, R. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem., 193, 265–275. 12. Gardner, B., Hall, D.A. and Strange, P.G. (1996) Pharmacological analysis of dopamine stimulation of [35 S]-GTPγ S binding via human D2short and D2long dopamine receptors expressed in recombinant cells. Br. J. Pharmacol., 118, 1544–1550. 13. Payne, S.L., Johansson, A.M. and Strange, P.G. (2002) Mechanisms of ligand binding and efﬁcacy at the human D2(short) dopamine receptor. J. Neurochem., 82, 1106–1117.

68

CH 3 USE OF THE [35 S]GTPγ S BINDING ASSAY

14. Milligan, G. (2003) Principles: extending the utility of [35 S]GTPγ S binding assays. Trends Pharmacol. Sci., 24, 87–90. 15. Asano, T., Pedersen, S.E., Scott, C.W. and Ross, E.M. (1984) Reconstitution of catecholamine-stimulated binding of guanosine 5 -O-(3-thiotriphosphate) to the stimulatory GTP-binding protein of adenylate cyclase. Biochemistry, 23, 5460–5467. 16. Bidlack, J.M. and Parkhill, A.L. (2004) Assay of G protein-coupled receptor activation of G proteins in native cell membranes using [35 S]GTPγ S binding. Methods Mol. Biol., 237, 135–143. 17. Selley, D.E., Sim, L.J., Xiao, R. et al. (1997) µ-Opioid receptor-stimulated guanosine-5 -O-(γ thio)-triphosphate binding in rat thalamus and cultured cell lines: signal transduction mechanisms underlying agonist efﬁcacy. Mol. Pharmacol., 51, 87–96. 18. Breivogel, C.S., Selley, D.E. and Childers, S.R. (1998) Cannabinoid receptor agonist efﬁcacy for stimulating [35 S]GTPγ S binding to rat cerebellar membranes correlates with agonist-induced decreases in GDP afﬁnity. J. Biol. Chem., 273, 16865–16873. 19. Lazareno, S., Farries, T. and Birdsall, N.J. (1993) Pharmacological characterization of guanine nucleotide exchange reactions in membranes from CHO cells stably transfected with human muscarinic receptors m1–m4. Life Sci., 52, 449–456. 20. Mueller, A., Mahmoud, N.G., Goedecke, M.C. et al. (2002) Pharmacological characterization of the chemokine receptor, CCR5. Br. J. Pharmacol., 135, 1033–1043. 21. Roberts, D.J., Lin, H. and Strange, P.G. (2004) Investigation of the mechanism of agonist and inverse agonist action at D2 dopamine receptors. Biochem. Pharmacol., 67, 1657–1665. 22. Lin, H., Saisch, S.G. and Strange, P.G. (2006) Assays for enhanced activity of low efﬁcacy partial agonists at the D2 dopamine receptor. Br. J. Pharmacol., 149, 291–299. This article shows the importance of assay conditions in assessing ligand efﬁcacy. 23. Roberts, D.J. and Strange, P.G. (2005) Mechanisms of inverse agonist action at D2 dopamine receptors. Br. J. Pharmacol., 145, 34–42. 24. McLoughlin, D.J. and Strange, P.G. (2000) Mechanisms of agonism and inverse agonism at serotonin 5-HT1A receptors. J. Neurochem., 74, 347–357. 25. Haworth, B., Lin, H., Fidock, M. et al. (2007) Allosteric effects of antagonists on signalling by the chemokine receptor CCR5. Biochem. Pharmacol., 74, 891–897. 26. Lahti, R.A., Figur, L.M., Piercey, M.F. et al. (1992) Intrinsic activity determinations at the dopamine D2 guanine nucleotide-binding protein-coupled receptor: utilization of receptor state binding afﬁnities. Mol. Pharmacol., 42, 432–438. 27. Cosi, C., Carilla-Durand, E., Assie, M.B. et al. (2006) Partial agonist properties of the antipsychotics SSR181507, aripiprazole and bifeprunox at dopamine D2 receptors: G protein activation and prolactin release. Eur. J. Pharmacol., 535, 135–144. 28. Quirk, K., Roberts, D.J. and Strange, P.G. (2007) Mechanisms of G protein activation via the D2 dopamine receptor: evidence for persistent receptor/G protein interaction after agonist stimulation. Br. J. Pharmacol., 151, 144–152. 29. Gay, E.A., Urban, J.D., Nichols, D.E. et al. (2004) Functional selectivity of D2 receptor ligands in a Chinese hamster ovary hD2L cell line: evidence for induction of ligand-speciﬁc receptor states. Mol. Pharmacol., 66, 97–105. 30. Kenakin, T. (2005) New concepts in drug discovery: collateral efﬁcacy and permissive antagonism. Nat. Rev. Drug Discov., 4, 919–927.

4 Quantitative Imaging of Receptor Trafﬁcking Andy R. James1 , Takeo Awaji2 , F. Anne Stephenson3 and Nicholas A. Hartell4 1 Department

of Pharmacology, The School of Pharmacy, University of London, London, UK of Physiology, Tokyo Women’s Medical University School of Medicine, Tokyo, Japan 3 Department of Pharmaceutical and Biological Chemistry, The School of Pharmacy, University of London, London, UK 4 Department of Cell Physiology and Pharmacology, University of Leicester, Leicester, UK 2 Department

4.1 Introduction Receptors provide cells with the means to respond to external signals, and the ability to regulate the density of receptors on the cell surface allows cells to modulate their responses over time. In the central nervous system (CNS), changes in the strength of signalling among neurons are thought to underpin memory formation. Conditions that lead to increases or decreases in synaptic signalling strength also produce concurrent increases or decreases in the numbers of receptors expressed at synapses, suggesting that dynamic receptor trafﬁcking may at least contribute to the biochemical basis for memory storage. The ability to visualize receptor trafﬁcking and to quantify changes in cell surface expression is of fundamental importance to many aspects of cell signalling. Methods which allow separation of intracellular and extracellular populations of receptors and that allow their distributions to be monitored over time would have signiﬁcant use for many biological systems, including synaptic transmission in the CNS. The introduction of ﬂuorescent proteins, in particular variants known as pHluorins that are sensitive to environmental pH, represents a signiﬁcant step towards this goal. Initially developed to visualize the pH changes associated with transmitter release [1], they are now being used to examine receptor trafﬁcking to and from the plasma membrane. The ﬂuorescence of pHluorins is quenched at acidic pH. By fusing pHluorin G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

70

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

to part of a receptor that is exposed to the extracellular space, receptors on the more alkaline cell surface appear brighter than those within acidic intracellular compartments. In this chapter, we describe the use of a ﬂuorescent-protein-based pH sensor that incorporates an additional, pH-insensitive ﬂuorophore. This allows the level of protein expression to be measured irrespective of the environmental pH and, as a consequence, the sensor can be calibrated. Surface-expressed receptors can be distinguished from those within intracellular compartments by the ratio of the two ﬂuorophores. Consequently, the two populations of receptors and their interactions can be visualized simultaneously. Although originally designed to examine glutamate receptor trafﬁcking in neurons, this sensor can be used to visualize receptor trafﬁcking in a range of biological systems; and because it can be calibrated, it allows receptor expression patterns to be compared directly between different cells.

4.2 Methods and approaches 4.2.1 Current methods of imaging receptor trafﬁcking The widespread introduction of laser scanning confocal microscopy over the last 20 years has revolutionized biological research; and, in combination with ﬂuorescent tagged antibodies, the spatial patterns of receptor expression have been explored. The spatial resolution of light microscopy, however, is generally limited to a few hundred nanometers, which does not allow complete discrimination of receptors on the cell surface from those inside cells. These limitations can be countered by labelling live cells to measure surface expression and then permeabilizing cells to reveal the total number of receptors. In this way, the proportions of each receptor population can be compared before and after experimental treatments. This technique represents a signiﬁcant technical advance over former, more laborious methods of radioligand binding and subcellular fractionation. However, both of these methods are limited because they give only an instantaneous snapshot of receptor distribution at a given point in time. An elegant adaptation of this method used thrombin cleavage in combination with antibody labelling to identify receptors newly inserted into the surface membrane and so provided a degree of temporal resolution [2–4]. Introduction of a histidine tag and a cleavage site for thrombin on the extracellular portion of the receptor allows, ﬁrst, surface receptor expression to be identiﬁed in live cells. Treatment with thrombin selectively cleaves the tag on surface receptors such that subsequent immunostaining for histidine at different time points reveals the time course of receptor insertion. Over the last decade, the introduction of recombinant, ﬂuorescent-protein-tagged receptors has facilitated more dynamic measurements of receptor movement. A number of techniques have been developed to monitor receptor movements. Fluorescence recovery after photobleaching can be used to measure the diffusion kinetics of receptors; but, as already stated, the spatial resolution limit of standard light microscopy techniques makes it difﬁcult to discern receptors on the cell membrane actively participating in cell-to-cell signalling from those involved in intracellular housekeeping.

4.2 METHODS AND APPROACHES

71

4.2.2 pH-sensitive ﬂuorescent proteins Initially, the pH sensitivity of wild-type green ﬂuorescent protein (GFP) and early variants (see [5] for a thorough review) hampered their use in measuring receptor distribution because their ﬂuorescence was quenched at the acidic pH of certain intracellular organelles involved in receptor trafﬁcking. This property was later harnessed: proteins engineered to have enhanced pH sensitivity within the physiological range of 5.5–7.4 were used to monitor the pH changes associated with transmitter release and endocytosis [1, 6, 7]. More recently, pH-sensitive ﬂuorescent proteins have been tagged to a number of ligand-gated ion-channel receptors and G protein-coupled receptors (GPCRs). pHluorins fused to the N-termini of glutamate receptor subunits 1 and 2 (GluR1 and GluR2) have been used to monitor the movement of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors expressed on the surface of cultured hippocampal neurons in response to treatments that give rise to long-term depression (LTD) and long-term potentiation (LTP) of synaptic transmission. These experiments reveal that chemically induced forms of LTP are accompanied by morphological changes in spine volume [8] and that chemical LTD triggers a reversible endocytosis of extrasynaptic receptors before a more slow but irreversible downregulation of receptors on spines [9, 10]. pHluorins have also been fused to the β3 subunit of the GABAA receptors to examine surface receptor diffusion [11] as well as to GPCRs including mGluR7, [12], calcium-sensing receptor [13], the β2 adrenergic receptor [14] and the CB1 cannabinoid receptor [15]. Ecliptic and superecliptic pHluorins, named because their ﬂuorescence is ‘eclipsed’ or quenched in acidic environments, allow direct visualization of the pH changes associated with the movement or exposure of proteins to environments at different pH values. These single-wavelength sensors, however, are not ideally suited to quantiﬁcation, particularly between different cells, because the baseline ﬂuorescence levels are dependent upon the overall protein expression levels. This cannot be readily controlled and may be highly variable. Therefore, whereas a difference in ﬂuorescence between neighbouring cells could arise from a selective difference in surface receptor density, it could equally indicate a difference in overall protein expression (Figure 4.1a). Quantiﬁcation of single-wavelength pH sensors, therefore, is restricted to relative changes with respect to arbitrary baselines. To overcome this limitation, we have developed sensors that comprise pH-sensitive and -insensitive components. YFpH consists of green (GFPUV ) and yellow (mEYFP) ﬂuorescent proteins joined together with a two amino acid linker; GFPUV is insensitive to pH over the physiological range [16] and mEYFP is highly pH sensitive between pH 5.0 and 8.0, with a pKa of 6.8. The addition of a pH-insensitive ﬂuorophore allows the total level of recombinant protein expression to be identiﬁed irrespective of the cellular compartment in which it resides. The ratio of mEYFP to GFPUV emission is proportional to pH; but crucially, because the two proteins are joined, and so co-expressed, the mEYFP : GFPUV ratio is independent of the level of protein expression, allowing absolute quantiﬁcation of pH. When attached to the extracellular terminus of a membrane receptor, cell-surface-expressed receptors can be distinguishable from those situated in more acidic endosomal compartments on the basis of their mEYFP : GFPUV

72

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

(a)

(b)

Figure 4.1 The uses of single and ratiometric ﬂuorescent protein pH sensors to monitor receptor trafﬁcking. (a) Single-wavelength pH sensors, such as pHluorin or superecliptic pHluorin, are quenched or eclipsed in acidic environments, such as those found in secretary vesicles (pH 5.5) or Golgi apparatus (pH 6.2). Receptors exposed to the relatively alkaline extracellular environment appear brighter than those in intracellular compartments. Since intracellular receptors are essentially invisible, it is not possible to distinguish between a cell that has low expression levels (left) or one that has high expression but which may have undergone LTD (centre). It is possible to monitor a change in receptor surface expression; for example, following the induction of LTP (right). (b) The ratiometric pH sensor YFpH has two ﬂuorophores. A pH-insensitive GFP is fused to a yellow ﬂuorescent protein pH sensor. The GFP signal indicates the total receptor number and the yellow ﬂuorescent protein signal indicates the environment. Receptors on the surface can be distinguished from those in intracellular compartments by the ratio of the two ﬂuorescent protein signals. A synapse that has low surface expression (centre) can be distinguished from one with low overall expression (left) because it will have a lower ratio. An increase in surface expression will be accompanied by an increase in the ratio (right).

ratio (Figure 4.1b). This information allows distinction between cells with low overall expression levels from those that have high expression levels but a low density of surface-expressed receptors. A further advantage is that surface-expressed receptors will have a higher mEYFP : GFPUV ratio than those in acidic compartments, and so the two populations can be separately identiﬁed, allowing the dynamic movement of the two populations to be monitored simultaneously in real time.

4.2.3 Characterization of YFpH fusion proteins The YFpH sensor has many potential uses, but it is well suited to the observation of membrane protein trafﬁcking, including the up- and down-regulation of ionotropic receptors and GPCRs. It is important to note that the spectral characteristics of YFpH

4.2 METHODS AND APPROACHES

73

may be inﬂuenced by the protein to which it is attached. The ﬂuorescence intensity of GFP is, for example, highly dependent on the folding properties of the protein sequence fused to either its N- or C-terminus [17]. Therefore, when fused to a membrane protein subunit, characterization of the spectral characteristics is essential to (i) determine the precise emission peaks of GFPUV and mEYFP, (ii) ensure limited ﬂuorescence resonance energy transfer (FRET) occurs between GFPUV and mEYFP and (iii) conﬁrm relative pH sensitivities for both GFPUV and mEYFP. Moreover, it is also essential to demonstrate that the protein tagged to YFpH shows similar functionality to the wild-type protein. Functional properties, therefore, should be shown for each fusion partner. In the case of ionotropic receptors, such as AMPA receptor GluR2 subunits, pharmacological and electrophysiological properties and expression patterns of YFpH-tagged GluR2 are similar to those of wild-type GluR2. The cholecystokinin receptor type A (CCKAR) is an example of a GPCR that has been tagged with GFP at its C-terminus and displayed similar ligand binding afﬁnity, similar signal transduction and similar patterns of surface expression to wild-type CCKARs [18]. The spectral characteristics of YFpH can be assessed using a confocal microscope equipped to carry out emission (lambda) scanning. We used an upright Leica TCS SP2 confocal microscope ﬁtted with a custom-built perfusion chamber and temperature controller. Standard molecular biological methods should be used to fuse YFpH to the extracellular N-terminal sequence of the protein of interest. If the receptor requires a signal peptide to aid surface expression, then YFpH cDNA should be inserted after this sequence. Plasmid cDNA of the YFpH-tagged receptor should be transfected into an adequate heterologous expression system for characterization experiments. Human embryonic kidney (HEK) 293 cells are a suitable model cell type for experiments with glutamate receptor subunits because they do not express endogenous glutamate receptors. HEK 293 cells were transfected with YFpH-tagged GluR2 using Effectene (Qiagen), following the manufacturer’s instructions. Transfected cells should be used no earlier than 48 h post-transfection to allow adequate receptor surface expression. Lambda scans of YFpH or YFpH fused constructs can be obtained in permeabilized cells by following Protocol 4.1.

PROTOCOL 4.1 Spectral Characterization of YFpH Equipment and Reagents • HEK 293 cells expressing YFpH or YFpH fused to protein/receptor of interest • Upright Leica TCS SP2 confocal microscope ﬁtted with perfusion chamber and temperature controller and standard blue (excitation 450–490 nm; 510 nm dichroic mirror; emission 515 nm long pass) and UV (excitation 340–380 nm; 400 nm dichroic mirror; emission 425 nm long pass) ﬁlter sets • Standard extracellular solution: NaCl, 135 mM; KCl, 5.4 mM; MgCl2 .2H2 O, 1 mM; CaCl2 , 1.8 mM; 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES), 10 mM , pH 7.4 [19]

74

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

• High K+ solution: KCl, 120 mM; NaCl, 20 mM; CaCl2 , 0.5 mM; MgSO4 , 0.5 mM; HEPES, 20 mM; containing nigericin and monensin (Sigma) both at 10 µg ml−1 [20] • The pH adjusted using NaOH and HCl for solutions ranging from pH 5.5 to 8.0.

Methods 1 Transfer YFpH-transfected HEK 293 cells to confocal microscope and perfuse with standard extracellular solution at pH 7.4. 2 Use ﬂuorescent light with either a 4 ,6-diamidino-2-phenylindole or ﬂuorescein isothiocyanate ﬁlter set to locate suitable cells with bright ﬂuorescence, indicating high YFpH expression. 3 Adjust the gain and offset for GFPUV (405 nm excitation) and mEYFP images (488 nm excitation).a 4 Perfuse cells with high K+ solution of lowest pHb (5.5) for 5 min. 5 Use 405 nm excitation and collect emission spectra between 450 and 650 nm for GFPUV .c 6 Use 488 nm excitation and collect emission spectra between 505 and 650 nm for mEYFP. 7 Perfuse with high K+ solution of pH 6.0 for 5 min. 8 Repeat steps 5–7 to obtain emission spectra for GFPUV and mEYFP up to pH 8.0. Notes a

Using the Leica TCS SP2 tunable emission ﬁlters, the collected wavelength bandwidth should be set to 5–10 nm and 10–20 separate steps deﬁned. It is important that the gain and offset of the photomultiplier tubes used to collect the data are kept constant throughout the experiment. Therefore, it is necessary to ensure that the gain and offset are adjusted at the outset so that signals are not saturated at the highest pH values or too low at the lowest pH values. Use an under-/over-ﬂow lookup table or an intensity histogram to ensure that the maximum grey level is approximately 50% of the maximum.

b Use

lowest pH ﬁrst to minimize photobleaching of chromophore.

c

Low laser powers, fast scan speeds and low resolution are preferred to minimize photobleaching of the chromophore. High-resolution images are not important for emission spectra measurements.

Most confocal microscopes equipped to do emission scanning have software packages that allow automatic measurement of the change in ﬂuorescence with wavelength, but not all adjust for background ﬂuorescence. We use custom procedures written in Igor Pro software (Wavemetrics Inc. Oregon, USA). Stacks of images for each excitation wavelength and for each pH buffer are loaded and absolute changes in ﬂuorescence within regions of interest (ROIs) are measured, including an area of background. Because the cells have to be scanned in several different pH buffers, precautions should be taken to minimize photobleaching. Laser power should be kept to an absolute minimum.

4.2 METHODS AND APPROACHES

75

ROIs should only encompass areas expressing YFpH, with the exception of the background ROI, which should ideally include a nontransfected cell. The mean ﬂuorescence for each ROI is then calculated at each wavelength and the corresponding background values subtracted. ROIs may then be pooled to form an average. Placing the background ROI on a nontransfected cell ensures compensation for wavelength-dependent changes in autoﬂuorescence. The same ROIs are used for each stack of images at each different pH. Emission scans at a given pH may then be compared directly or normalized with the maximum ﬂuorescence value for the entire data set. The relative intensity values at the peak emission value are then plotted and ﬁtted with a sigmoidal curve and the pKa estimated. Figure 4.2 provides an example of a lambda scan experiment carried out in permeabilized HEK 293 cells expressing YFpH. Lambda scans provide two essential pieces of information. First, they provide the optimum emission spectra for each ﬂuorophore in situ. This can be used to choose the most appropriate dichroic and emission ﬁlter combinations to best separate the two signals. Second, the data help to reveal whether there is any FRET between GFPUV (a)

(b)

(c)

(d)

Figure 4.2 Structure and characteristics of YFpH. (a) Schematic drawing of YFpH illustrating the predicted excitation and emission maxima of the composite ﬂuorophores GFPUV and mEYFP. (b) Confocal images of YFpH expressing HEK 293 cells, permeabilized with the H+ ionophores monensin and nigericin. The top set of images illustrates the ﬂuorescence of GFPUV (excitation 405 nm; emission 500–515 nm). The lower set illustrates the ﬂuorescence of mEYFP (excitation 488 nm; emission 518–540 nm). Lambda emission scans were obtained in the presence of a range of different pH buffers on HEK 293 cells expressing YFpH and permeabilized with H+ ionophores at excitation wavelengths of 405 nm (c) and 488 nm (d). Insets: the relationship between pH and the peak ﬂuorescence intensities measured at 509 nm and 527 nm are shown. Although mEYFP is optimally excited at 516 nm, we used 488 nm because this allowed us to use a dichroic mirror that allowed measurement of emission wavelengths above 500 nm.

76

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

and mEYFP. We have found that FRET is generally low and can be ignored, but it can vary between fusion constructs. If the percentage of FRET is too high, then ratiometric analysis becomes difﬁcult because the emission of GFPUV becomes dependent on the ﬂuorescence of mEYFP. Whilst this enhances the sensitivity of the pH sensor, the GFPUV is no longer independent of pH. For these reasons, it is important to determine the degree of FRET empirically.

4.2.4 Identifying surface-expressed receptors The ratio of the pH-sensitive to pH-insensitive components of YFpH should allow us to distinguish surface-expressed receptors from those in acidic intracellular compartments. To verify this, we fused YFpH to the N-terminus of GluR2 and observed its expression patterns in living and permeabilized HEK 293 cells in the presence of buffers ranging from pH 5.5 to 9.0. Images of GFPUV (excitation 405 nm; emission 490–515 nm) and mEYFP (excitation 488 nm; emission 517–560 nm) were taken alternately at suitable intervals, again taking care to reduce excitation laser power as much as possible to minimize photobleaching. Ratio images of mEYFP : GFPUV were then constructed. Ratiometric analysis must be carried out with great care to minimize the introduction of noise. The principles, described in Protocol 4.2, are essentially identical to those used for ratiometric calcium indicators or FRET analysis. A threshold value is ﬁrst applied to each image in a ratio pair to separate background ﬂuorescence from the desired signal. Pixels identiﬁed as background are then excluded from further analysis. A median ﬁlter can be applied to images before thresholding and ratioing to improve the signal-to-noise ratio if necessary. These parameters, which must be determined for each image set, are then applied to all images in the series and the ratio images of mEYFP/GFPUV generated.

PROTOCOL 4.2 Construction of Ratiometric Images Equipment • Suitable image analysis software.a

Methods 1 Load individual pairs of GFPUV and mEYFP images or stacks of image pairs taken over time. 2 Apply a median ﬁlter to each image in a set.b 3 Find the pixel intensity that best deﬁnes the background ﬂuorescence for each image set and subtract the background value from each pixel in the image. 4 Identify pixels that must be treated as ‘not a number’ (NAN).c 5 Divide the mEYFP image by the GFPUV image.d

4.2 METHODS AND APPROACHES

77

6 Repeat steps 2–5 for each image pair in a stack of images using the same parameters at each stage. Notes a This

analysis can be done using commercially available software that is capable of ratiometric calcium imaging FRET, such as Metamorph (Molecular Devices, Sunnyvale, CA, USA) or Image Pro Plus (Media Cybernetics, Bethesda, MD, USA). We have written procedures in Igor Pro that automate the processes and which make handling of large stacks of data relatively straightforward. ImageJ is a freeware program that can be used, but some programming may be required if dealing with large volumes of data. b

Images can be ﬁltered at this stage. We ﬁnd that a standard median ﬁlter helps to reduce the introduction of noise at later stages of ratio image construction. c

After background subtraction, some pixels will have values of zero or values approaching zero. Zero values in either the numerator image (mEYFP) or denominator image (GFPUV ) cause signiﬁcant problems and introduce large artefacts in subsequent analysis and so they have to be eliminated from further analysis. In Igor Pro, zero values or values below an arbitrary threshold can be identiﬁed as NANs and, therefore, they play no further role in image processing. d Images are generally in integer formats of 8-, 12- or 16-bit levels. Integer division loses signiﬁcant amounts of accuracy, and so the images must be converted into a format that allows values of less than one. In our procedures, images are converted into double-precision format, mEYFP is divided by GFPUV and then the image is rescaled and then reconverted into a suitable integer format. It should be noted that these steps are not necessary in commercial software because they are done automatically during image maths manipulations.

By choosing a cell line that does not normally express the receptor of interest, it is possible to test for the presence of the functional receptors by simply challenging with a suitable agonist and measuring the expected response. In this case, we used whole-cell patch clamp methods to measure inward currents in response to brief applications of glutamate. Having established that the cells express functional receptors on the cell surface, the next step involves correlating the pattern of surface expression with the mEYFP : GFPUV ratio. Protocol 4.3 describes speciﬁcally how to label cell surface YFpH-GluR2 using an anti-GluR2 antibody; this protocol can be adapted for use with GPCRs. To maintain pH gradients, live cells must be used.

PROTOCOL 4.3

Live Cell Antibody Labelling of Surface YFpH-GluR2

Equipment and Reagents • HEK 293 cells expressing YFpH-GluR2a • Tris-buffered saline (TBS): NaCl, 145 mM; trishydroxymethylaminomethane (tris), 19 mM; in double-distilled H2 O with a pH adjusted to 7.4 by using HCl and NaOH • Mouse anti-GluR2-N antibody (Chemicon)

78

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

• Goat anti-mouse CY3-conjugated secondary antibody (Molecular Probes) • Goat serum (e.g. Invitrogen) • Bovine serum albumin •

DL-Lysine

(e.g. Sigma)

• All steps to be conducted at 4 ◦ C.b

Method 1 Dilute mouse anti-GluR2-N antibody in TBS containing goat serum, 5% (v/v), bovine serum albumin, 2% (w/v) and DL-lysine, 0.1% (w/v) at a ratio of 1 : 1000 (antibody : dilutant). 2 Store the diluted mouse anti-GluR2-N antibody at 4 ◦ C until needed. 3 Wash coverslips containing YFpH-GluR2-transfected HEK 293 cells expressing with 250 µl of chilled TBS. 4 Wash the cells a further twice with 250 µl of chilled TBS. 5 Remove TBS and incubate cells with diluted mouse anti-GluR2-N antibody for 45 min at 4 ◦ C. 6 Meanwhile, dilute CY3-conjugated secondary antibody in TBS containing goat serum, 5% (v/v), bovine serum albumin, 2% (w/v), DL-lysine, 0.1% (w/v) at a ratio of 1 : 200 (antibody : dilutant) and store at 4 ◦ C in the darkc until needed. 7 After incubation, wash cells ﬁve times with chilled TBS. 8 Remove TBS and incubate with diluted CY3-conjugated secondary antibody for 75 minutes at 4 ◦ C 9 Wash cells ﬁve times with chilled TBS. 10 Proceed immediately with Protocol 4.3. Notes a Solution

volumes are based on cells growing on 13 millimeter diameter coverslips. If larger coverslips are used to culture cells, then increase solution volumes accordingly.

b To c

inhibit receptor trafﬁcking.

To reduce photobleaching of the chromophore.

Subsequent measurements can then be made using ROIs placed in identical positions on the original mEYFP and GFPUV images and on the ratio image. Static ROI measurements allow observation of receptors to and from the membrane or cytoplasm over time. An alternative approach is to identify pixels on the basis of their ratio value and place them into populations accordingly. Because the number of pixels in each population is known, this is useful for measuring the overall proportions of surface or intracellular receptors over time.

4.2 METHODS AND APPROACHES

79

This analysis depends upon accurately determining the ratio values that deﬁne surface receptors. Protocol 4.4 describes how this can be done. Ratio images are collected from nonpermeabilized live HEK 293 cells expressing YFpH-GluR2, ﬁrst at pH 7.4 and then after switching to a solution at pH 5.5. Brieﬂy lowering the extracellular pH has no effect on intracellular pH, but it selectively quenches mEYFP on the surface of cells (Figure 4.3a). GFPUV ﬂuorescence is unchanged. Frequency distribution plots of ratio data sampled at each pH reveal small but reproducible shifts in the overall distributions of ratio values: the numbers of pixels with low ratio values increase, whereas those with high ratio values decrease. This difference is very small and, therefore, is difﬁcult to see from the raw histogram because the number of low-ratio pixels far exceeds the number of high-ratio pixels (Figure 4.3b). Therefore, cumulative probability curves can be constructed. These reveal a signiﬁcant difference in distribution (Figure 4.3b inset; Kolmogorov–Smirnov test, P < 0.01). For the example shown, the difference becomes apparent at pixel ratio values greater than 1.3. This calibration technique gives a good estimate of the ratio that best separates the two populations of receptors. This procedure can be repeated using a weak base, such as ammonium chloride, which increases intracellular pH. Under these circumstances, free ammonia in solution crosses membranes and reacts with free protons, raising the pH of intracellular compartments. On washout, there is a rebound intracellular acidiﬁcation before a slow recovery. This method does not provide a precise intracellular pH change, but it produces a relatively selective change in pH of receptors within intracellular compartments.

PROTOCOL 4.4 Receptors

Calibration of Live Cells to Identify Extracellular

Equipment and Reagents • HEK 293 cells expressing YFpH-GluR2a • Extracellular solution: NaCl, 135 mM; KCl, 5.4 mM; MgCl2 ·2H2 O, 1 mM; CaCl2 , 1.8 mM; HEPES, 10 mM adjusted to pH 7.4 or 5.5–6.0 using NaOH.

Method 1 Take GFPUV and mEYFP image pairs at suitable intervals and challenge cells with 1 mM glutamate. 2 Take images of GFPUV and mEYFP at pH 7.4 and then in a solution at pH 5.5–6.0.a 3 Create ratio images according to Protocol 4.3. 4 Create frequency distribution histograms of all the ratio pixels in images at pH 7.4 and pH 5.5–6.0.b 5 Create cumulative probability graphs for each histogram.c 6 Establish the ratio which separates receptors that respond to extracellular acidiﬁcation. 7 Apply these ratio values to images collected during the application of glutamate.d

80

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

Notes a Brief

exposure of cells to acidic buffers has little effect.

b

Bin widths should be sufﬁciently large to produce a smooth histogram, but not so large as to miss small differences between histograms. The bin width will depend upon the bit level of the image.

c Subtracting the cumulative probability graphs can help identify precisely the ratio at which differences start. d

Ensure that the pH of the glutamate solution is 7.4.

One application of this technique is to monitor real-time internalization of receptors from the cell surface. Treatments such as application of insulin or glutamate receptor agonists have previously been shown to cause internalization of GluR2-containing AMPA receptors in HEK 293 cells [21, 22]. Figure 4.3c illustrates the effects of glutamate on surface and intracellular receptor populations. The advantages of ratiometric imaging can be highlighted by drawing parallels with calcium imaging techniques. Ratiometric dyes such as Fura-2 [23] allow absolute measurement of calcium concentrations because they emit a ﬂuorescent signal that is independent of calcium concentration as well as a calcium-dependent signal. Single-wavelength calcium dyes, however, only detect changes in calcium with respect to an unknown baseline level. Calibration is difﬁcult because the concentration of dye is not known. Ratiometric pH sensors can be readily calibrated because the expression level of the sensor is known. In addition to their use in monitoring receptor trafﬁcking, ratiometric pH sensors may be usefully tagged to vesicular proteins such as synaptobrevin or synaptotagmin. As well as allowing direct visualization of transmitter release from the readily releasable pool, a ratiometric pH sensor would additionally provide information about vesicles in the reserve pool that are not actively engaged in the release process. Therefore, it may be possible to estimate the release characteristics of a presynaptic terminal synapse by the mEYFP : GFPUV ratio emission signal.

4.3 Troubleshooting • Many receptors require a signalling peptide that helps to move the receptor across the endoplasmic reticulum to the plasma membrane. The peptide is cleaved during trafﬁcking, leaving the mature peptide in the membrane. In such cases, the ﬂuorescent protein sensor must be inserted between the signal peptide and the mature receptor sequence. For other receptors, such as the CB1 receptor, increasing the length of the N-terminus with a protein sensor may inﬂuence expression [24]. Efﬁcient surface expression requires the introduction of an artiﬁcial sequence derived from human growth hormone [15]. • Ratioing images can introduce a signiﬁcant level of noise, particularly if the ﬂuorescence intensity of the GFPUV signal is very weak. Illumination intensities should

4.3 TROUBLESHOOTING

81

(a)

(b)

(c)

Figure 4.3 Ratiometric identiﬁcation of extracellular and intracellular receptors. (a) The effects of changing the pH of the extracellular bathing solution of intact HEK 293 cells expressing YFpH-GluR2 are shown. Images of mEYFP and GFPUV signals were collected at 10 s intervals and, after a short period of baseline recording, the pH of the extracellular solution was lowered from 7.4 to 5.5. (b) Frequency distributions of the ratio values of individual image pixels at pH 7.4 (black traces) and pH 5.5 (grey traces) are shown. Inset: cumulative probability curves constructed at pH 7.4 (black) and 5.5 (grey). On moving from pH 7.4 to 5.5, pixel values below 1.3 increased in intensity, whereas those above 1.3 decreased in intensity. (c) The effects of a 10 min application of glutamate on HEK 293 cells expressing YFpH-GluR2 are illustrated. Ratio images of mEYFP : GFPUV signals were constructed and two sets of ROIs generated on the basis of threshold. Ratio values were normalized to their respective baseline levels to aid comparison. To the right, representative ratio images collected before and during glutamate application are shown. The upper image shows all pixels. The lower two images illustrate the effects of glutamate on pixels identiﬁed as being extracellular on the basis of their ratio.

82

CH 4 QUANTITATIVE IMAGING OF RECEPTOR TRAFFICKING

be adjusted to produce good contrast, bright images, but at the same time bearing in mind that higher intensities promote photobleaching. If protein expression levels are low, then it is better to lower the spatial resolution of the original images by increasing the size of the pinhole, reducing the confocal image format to 256 × 256 or, for camera-based systems, by pixel binning.

References 1. Miesenbock, G., De Angelis, D.A. and Rothman, J.E. (1998) Visualizing secretion and synaptic transmission with pH-sensitive green ﬂuorescent proteins. Nature, 394, 192–195. The original publication describing the use of synaptopHluorin to observe presynaptic release. 2. Daunt, D.A., Hurt, C., Hein, L. et al. (1997) Subtype-speciﬁc intracellular trafﬁcking of alpha2adrenergic receptors. Mol. Pharmacol., 51, 711–720. 3. Hein, L., Ishii, K., Coughlin, S.R. and Kobilka, B.K. (1994) Intracellular targeting and trafﬁcking of thrombin receptors. A novel mechanism for resensitization of a G protein-coupled receptor. J. Biol. Chem., 269, 27719–27726. 4. Passafaro, M., Piech, V. and Sheng, M. (2001) Subunit-speciﬁc temporal and spatial patterns of AMPA receptor exocytosis in hippocampal neurons. Nat. Neurosci., 4, 917–926. 5. Tsien, R.Y. (1998) The green ﬂuorescent protein. Annu. Rev. Biochem., 67, 509–544. 6. Sankaranarayanan, S., De Angelis, D., Rothman, J.E. and Ryan, T.A. (2000) The use of pHluorins for optical measurements of presynaptic activity. Biophys. J, 79, 2199–2208. 7. Sankaranarayanan, S. and Ryan, T.A. (2001) Calcium accelerates endocytosis of vSNAREs at hippocampal synapses. Nat. Neurosci., 4, 129–136. 8. Kopec, C.D., Li, B., Wei, W. et al. (2006) Glutamate receptor exocytosis and spine enlargement during chemically induced long-term potentiation. J. Neurosci., 26, 2000–2009. 9. Ashby, M.C., De La Rue, S.A., Ralph, G.S. et al. (2004) Removal of AMPA receptors (AMPARs) from synapses is preceded by transient endocytosis of extrasynaptic AMPARs. J. Neurosci., 24, 5172–5176. This paper demonstrates the use of ecliptic pHluorin to monitor GluR2 receptor trafﬁcking. 10. Ashby, M.C., Maier, S.R., Nishimune, A. and Henley, J.M. (2006) Lateral diffusion drives constitutive exchange of AMPA receptors at dendritic spines and is regulated by spine morphology. J. Neurosci., 26, 7046–7055. 11. Jacob, T.C., Bogdanov, Y.D., Magnus, C. et al. (2005) Gephyrin regulates the cell surface dynamics of synaptic GABAA receptors. J. Neurosci., 25, 10469–10478. 12. Pelkey, K.A., Yuan, X., Lavezzari, G. et al. (2007) mGluR7 undergoes rapid internalization in response to activation by the allosteric agonist AMN082. Neuropharmacology, 52, 108–117. 13. Bouschet, T., Martin, S. and Henley, J.M. (2005) Receptor-activity-modifying proteins are required for forward trafﬁcking of the calcium-sensing receptor to the plasma membrane. J. Cell Sci., 118, 4709–4720. 14. Yudowski, G.A., Puthenveedu, M.A. and von Zastrow, M. (2006) Distinct modes of regulated receptor insertion to the somatodendritic plasma membrane. Nat. Neurosci., 9, 622–627.

REFERENCES

83

15. McDonald, N.A., Henstridge, C.M., Connolly, C.N. and Irving, A.J. (2007) Generation and functional characterization of ﬂuorescent, N-terminally tagged CB1 receptor chimeras for live-cell imaging. Mol. Cell. Neurosci., 35, 237–248. 16. Crameri, A., Whitehorn, E.A., Tate, E. and Stemmer, W.P. (1996) Improved green ﬂuorescent protein by molecular evolution using DNA shufﬂing. Nat. Biotechnol., 14, 315–319. 17. Sacchetti, A. and Alberti, S. (1999) Protein tags enhance GFP folding in eukaryotic cells. Nat. Biotechnol., 17, 1046. 18. Tarasova, N.I., Stauber, R.H., Choi, J.K. et al. (1997) Visualization of G protein-coupled receptor trafﬁcking with the aid of the green ﬂuorescent protein. Endocytosis and recycling of cholecystokinin receptor type A. J. Biol. Chem., 272, 14817–14824. 19. Burnashev, N., Monyer, H., Seeburg, P.H. and Sakmann, B. (1992) Divalent ion permeability of AMPA receptor channels is dominated by the edited form of a single subunit. Neuron, 8, 189–198. 20. Awaji, T., Hirasawa, A., Shirakawa, H. et al. (2001) Novel green ﬂuorescent protein-based ratiometric indicators for monitoring pH in deﬁned intracellular microdomains. Biochem. Biophys. Res. Commun., 289, 457–462. 21. Man, H.Y., Lin, J.W., Ju, W.H. et al. (2000) Regulation of AMPA receptor-mediated synaptic transmission by clathrin-dependent receptor internalization. Neuron, 25, 649–662. 22. Wang, Y.T. and Linden, D.J. (2000) Expression of cerebellar long-term depression requires postsynaptic clathrin-mediated endocytosis. Neuron, 25, 635–647. 23. Grynkiewicz, G., Poenie, M. and Tsien, R.Y. (1985) A new generation of Ca2+ indicators with greatly improved ﬂuorescence properties. J. Biol. Chem., 260, 3440–3450. 24. Andersson, H., D’Antona, A.M., Kendall, D.A. et al. (2003) Membrane assembly of the cannabinoid receptor 1: impact of a long N-terminal tail. Mol. Pharmacol., 64, 570–577.

5 Production of Recombinant G Protein-coupled Receptor in Yeast for Structural and Functional Analysis Richard A.J. Darby, Mohammed Jamshad, Ljuban Grgic, William J. Holmes and Roslyn M. Bill Pharmaceutical & Biological Sciences, School of Life and Health Sciences, Aston University, Birmingham, UK

5.1 Introduction G protein-coupled receptors (GPCRs) are naturally present in their native membranes in low abundance. Consequently, recombinant protein production is an obligatory ﬁrst step en route to their structural and functional analysis. The notable exception to this is the case of rhodopsin, which can be isolated from bovine retina in very large (>100 mg) quantities, a feature that has undoubtedly been important in obtaining its structure to ˚ resolution. Interestingly, rhodopsin can also be produced recombinantly in the 2.2 A mammalian cell-line COS, which led to the recent report of a crystal structure of a thermostable mutant [1]. Such successes with rhodopsin are probably due in part to its enhanced ground-state stability on binding 11-cis-retinal. It is no surprise then, that similar successes with other, often less stable, GPCRs have not been forthcoming. In principle, recombinant GPCRs could be produced in a range of host systems, including bacteria, yeast, baculovirus-infected insect cells, virus-infected mammalian cells and cell-free systems. In practice, however, they are typically produced in only G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

86

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

three of these: yeast, insect cells and mammalian cells. Unfortunately, low expression levels, inefﬁcient puriﬁcation protocols and nonfunctional or unstable recombinant protein plague most attempts to emulate the case of rhodopsin. In fact, very few examples of GPCRs produced in functional form in sufﬁcient quantities for crystallization studies have been reported to date [2–4]. In these cases, the preferred hosts have been either the yeast Pichia pastoris or mammalian cells. It is very clear that the best expression system and conditions for optimal production are unpredictable and have to be determined empirically. In this chapter, we focus on yeast, which has the potential to produce stable, functional GPCRs in large quantities. Since it is a eukaryotic microbe, it combines the cost and speed beneﬁts of a eukaryote with the ease of handling of Escherichia coli [5].

5.2 Methods and approaches Features such as the requirement for chaperones [6, 7], glycosylation [8] or speciﬁc lipids (such as sterols [9]) must be taken into account when considering methods and approaches for the production of a recombinant GPCR. In particular, these considerations inﬂuence the molecular biology required. Thereafter, once a suitable production level has been attained, it is necessary to ensure that the protein is functional, stable and homogeneous. For example, µ-opioid receptor production in P . pastoris has been optimized so that the ratio of functional to nonfunctional receptor improved from 1 : 20 to 1 : 4.5 [10]. Intuitively, the GPCRs most likely to be produced in quantities amenable to crystallization trials are the ones which are best understood biochemically. This is certainly true of rhodopsin, for which there is a rich and detailed biochemical literature [1]. In addition, the careful analyses of the human serotonin and the adenosine 2a receptors [2, 4] have no doubt contributed to an understanding of how to handle them in detergent solution to maximize function and stability. One of the advantages of working with GPCRs is that they can be functionally characterized through the production and puriﬁcation process via ligand binding assays. This is an important part of any production pipeline, as it is essential that the puriﬁed recombinant protein retains its native characteristics [11]. The retention of native structural properties can also be assessed by measuring secondary-structure content by spectroscopy. In particular, a library of circular dichroism (CD) spectra of membrane proteins is being established which will become an increasingly important resource [12]. This chapter covers the selection of a suitable yeast species, cloning the gene encoding the target GPCR into an expression plasmid, clone selection, screening, scale-up, solubilization, puriﬁcation and, ﬁnally, analysis of the recombinant protein.

5.2.1 Choice of yeast host There are over 500 known species of yeast, yet only a few of these have been employed to produce heterologous proteins. The two most widely used species are P . pastoris and Saccharomyces cerevisiae.

5.2 METHODS AND APPROACHES

87

5.2.1.1 P. pastoris The P . pastoris expression system is available as a commercial kit from Invitrogen Corporation. In particular, the ability of this methylotrophic yeast to be cultivated to extremely high cell densities (e.g. measurement of optical density at 600 nm (OD600 ) in excess of 400) in deﬁned media makes it an ideal candidate for the production of GPCRs. The system is based around the promoter regulating the production of alcohol oxidase (AOX1 ), which is manipulated to drive integrated heterologous protein production. The initial step in the metabolism of methanol by P . pastoris is the oxidation of methanol by alcohol oxidase to yield formaldehyde and hydrogen peroxide using molecular oxygen. This takes place within the peroxisome in order to avoid hydrogen peroxide toxicity. Since alcohol oxidase has a poor afﬁnity for oxygen, P . pastoris compensates by generating large amounts of alcohol oxidase. In methanol-grown cells, approximately 5% of the polyA + RNA is from the AOX1 gene [13]. The regulation of the AOX1 gene involves a repression/derepression mechanism followed by an induction akin to the GAL1 gene in S. cerevisiae. Essentially, cultivating P . pastoris on glucose severely represses AOX1 transcription, even in the presence of methanol. Therefore, initial cultivation is conducted in a glycerol-containing medium to de-repress the gene, with induction being initiated via the addition of methanol (see Section 5.2.4 for a more detailed explanation of cultivation and [14]). The main advantages of P . pastoris, then, are its ability to be cultivated to obtain very high biomass yields and its potential to produce large amounts of GPCR in its membranes by virtue of the strong AOX1 promoter. The recent commercial availability of a genome sequence (www.integratedgenomics.com) should also allow genetic manipulation of this yeast to be carried out more readily.

5.2.1.2 S. cerevisiae More commonly known as Brewers’ or Bakers’ yeast, S. cerevisiae is an important industrial protein production host. Its value as the most important model eukaryote cannot be overstated: its genome has been fully sequenced and annotated and the sequence is freely available online (www.yeastgenome.org). In contrast to P . pastoris, there is an abundance of vectors available that generally fall into one of three categories, examples of which can be seen in Table 5.1. These include integrative plasmids, autonomous single copy number plasmids and autonomous multicopy plasmids. This means that a range of promoters of varying strengths can be tested in order to optimize the most suitable regime for recombinant protein production.

5.2.2 Preparation of expression plasmids Generally, the DNA sequence encoding the GPCR of interest is ampliﬁed via polymerase chain reaction (PCR) either from genomic DNA or cDNA. It is then cloned into an appropriate shuttle or expression vector with or without signal sequences and/or fusion partners, and subsequently sequence veriﬁed. The steps described below can be applied equally to P . pastoris or S. cerevisiae unless stated otherwise.

88

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

Table 5.1 Some of the available expression vectors for both P. pastoris and S. cerevisiae. Vector

Promoter Selection marker Promoter properties Yeast

pPICZ A,B,C

AOX1

Integrative

Tightly regulated

P . pastoris

pPIC3 ± K and pPIC9 ± K AOX1

HIS4

Tightly regulated

P . pastoris

pYES2

GAL1

URA3

Tightly regulated

S. cerevisiae

pYX212

TPI

URA3

Constitutive

S. cerevisiae

pYX222

TPI

HIS3

Constitutive

S. cerevisiae

pHybLex/Zeo

ADH2

Zeocin

Tightly regulated

S. cerevisiae

YEpCTHS

CUP1

Ampicilin

Weakly regulated

S. cerevisiae

5.2.2.1 Choice of vector The expression vector must contain an optimized promoter, terminator and an appropriate ribosome binding site. The fact that most of these elements have long been established for soluble protein production aids the choice and/or modiﬁcation of suitable GPCR expression vectors. Vectors are available containing either inducible promoters such as P . pastoris AOX1 and S. cerevisiae GAL1 , or constitutive promoters such as the S. cerevisiae TPI1 . Interestingly, there are no widely used, constitutive P . pastoris promoters (Table 5.1). Common to all vectors is a multiple cloning site (MCS) downstream of the promoter and upstream of the terminator sequence comprised of several unique restriction enzyme sites. For example, the MCS of the P . pastoris vector pPICZA contains the following sites: 5 -SfuI, EcoRI, Pml I, Sﬁ I, BsmBI, Asp718 I, KpnI, XhoI, SacII, NotI and ApaI-3 . The subsequent cloning of the gene into one or more of these sites places its transcription under the direct control of the upstream AOX1 promoter. The choice of restriction site is primarily governed by the DNA sequence of the GPCR gene, in that the chosen restriction sites must not be present within the gene. Of relevance here is that the vector will need to be linearized with SacI, PmeI or BstXI, prior to transformation (see Section 5.2.3), so it is important to check which, if any, of these sites are also present in the gene sequence. Once appropriate restriction sites have been selected they are normally incorporated at the 5 and 3 ends of the PCR primers used to amplify the GPCR sequence, as described below (Section 5.2.2.3). Particular attention should be paid to ensuring that the encoded mRNA is efﬁciently translated. Most critical to this is the 5 untranslated region (UTR) of the gene preceding the initiation codon. In short, the 5 UTR should not contain any of the following: • AUG codons preceding the initiation codon, as this will lead to mispriming by the ribosome and an incorrect translation of the transcript; • sequences that have the potential to form stable intramolecular secondary structures, as these stem-like structures can lead to signiﬁcant translational inhibition; • sequences containing runs of guanine nucleotides, as these lead to decreased translational efﬁciency.

5.2 METHODS AND APPROACHES

89

The sequence immediately before the initiation codon is the single most important feature to enable efﬁcient translation. Whilst many eukaryotic Kozak sequences are known, it is generally regarded that a run of at least ﬁve adenine nucleotides preceding the AUG start codon is the minimal requirement. It has been suggested, however, that translational efﬁciency can be improved by adding the following sequence, where the initiation codon is highlighted in bold: 5 -(A/U)A(A/C)AA(A/C)AUG [15].

5.2.2.2 Addition of signal sequences and tags To improve the chances of a GPCR being inserted into the yeast membrane, it is possible to direct the protein through the yeast secretory pathway. In order to achieve this, an amino-terminal hydrophobic signal peptide must be incorporated. The most common form of signal peptide used for this is the prepro region of the S. cerevisiae mating factor α1 (MFα1), which has been used to improve the production of GPCRs two- to three-fold [16]. The peptide contains a 22-residue ‘pre’ protein signal linked to a 61-residue ‘pro’ region. The pre-protein signal is cleaved by signal peptidases associated with the endoplasmic reticulum membrane, whilst the pro-region is subsequently cleaved by Kex proteases within the Golgi apparatus. However, the Kex protease recognizes and cleaves only at Lys–Arg junctions, so the amino terminus of the GPCR will not be the same as the native sequence and will at best contain an additional Arg residue. Several vectors are available which contain the MFα1 signal sequence, including pGS4 and pPIC9K (Table 5.1). It is also noteworthy that, in some cases, the signal sequence has not been cleaved, leading to a GPCR with compromised ligand-binding properties [17]. It may also be beneﬁcial to include one or more fusion tags to facilitate subsequent GPCR detection and/or puriﬁcation. To date, the most commonly used tags are polyhistidine (e.g. His6 and His10 ), Flag (DYKDDDDK), HA (YPYDVPDYA) and c-myc (EQKLISEEDL), all of which can be amino- and/or carboxy-terminal. The positioning of tags will ultimately be governed by the prediction of whether the termini are intra- or extra-cellular and whether the tags disrupt GPCR functionality. Without the aid of modelling or previous experience, ﬁnal decisions must be made following experimentation. Other, much larger, fusion partners, such as green ﬂuorescent protein (GFP), maltose binding protein (MBP), thioredoxin (TRX) and glutathione-S-transferase (GST), may also be used to potentially increase production levels, but they generally need to be removed to restore protein functionality.

5.2.2.3 PCR ampliﬁcation and cloning Once decisions have been made as to the vector to be used and the tags to be added, the steps leading up to cloning will typically be performed by PCR. PCR ampliﬁcation is well established and, therefore, will not be covered in detail here. Instead, we present a general overview, highlighting some of the major issues to consider and giving an experimental protocol (Protocol 5.1). Primer design is inﬂuenced by a number of features, including the restriction sites to be used, whether a 5 Kozak sequence will be included (Section 5.2.2.1) and whether small amino- and/or carboxy-terminal fusion tags will be encoded (Section 5.2.2.2). It is possible to amplify by PCR a GPCR

90

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

sequence with an appropriate Kozak and carboxy-terminal His6 tag using one set of primers. Of course, the ampliﬁed GPCR sequence can also be ligated into a vector that already contains appropriate sequences and tags. These decisions will be governed by the material to hand and the user’s capabilities.

PROTOCOL 5.1 Preparation of the Expression Plasmid Equipment and Reagents • DNA Engine (PTC-200) Peltier Thermo Cycler (Bio-Rad Laboratories) • PCR clean-up kit (Qiagen) • Gel extraction kit (Qiagen) • Pfu DNA polymerase (Promega) and 10× buffer provided with the Pfu polymerase • Oligonucleotide primers (Invitrogen) • 10 mM dNTP mixture (Sigma–Aldrich): a mixture of all four dNTPs (dATP, dCTP, dGTP, dTTP) • DNA template, typically a plasmid at 1 µg µl−1 containing the gene of interest • DNA ligase (Invitrogen) • 0.2 ml sterile PCR tubes (Sarstedt) • 0.5 ml microcentrifuge tubes (Sarstedt) • 0.8–1.0% agarose gel • Agarose gel tank (Bio-Rad Laboratories) • 240 V power pack (Bio-Rad Laboratories) • Restriction enzymes (New England Biolabs).

Method 1 Prepare a ‘master mix’ on ice as follows: (a) 1 µl forward primer (50 pmol) (b) 1 µl reverse primer (50 pmol) (c) 1 µl template DNA (0.1–1 ng) (d) 12 µl 10× buffer provided with the polymerase (e) 2 µl 10 mM dNTPs (f) 1 µl polymerase enzyme (e.g. Pfu from Promega) (g) 102 µl sterile water. 2 Aliquot 10 µl volumes of the master mix into thin-walled 0.2 ml sterile PCR tubes.

5.2 METHODS AND APPROACHES

91

3 Place the 0.2 ml PCR tubes in a thermocycler and run the following program:

Denature Anneal Extension Cycles

94 ◦ C; 1 min 45–65 ◦ C; 1 min (gradient PCR) 72 ◦ C; (500 base pairs/min) 35

4 Run 2 µl from each tube on a 0.8–1% agarose gel at 80 V until adequately separated to conﬁrm product size. Purify the remaining PCR product using a commercially available kit.a 5 Digest both the PCR product and vector with the appropriate restriction enzymes. Purify the digested vector and PCR product with a commercially available kit.b 6 Ligate the PCR product into the vector of choice. Conditions will be enzyme speciﬁc. The following example is based on the T4 DNA ligase enzyme and 5× buffer supplied by Invitrogen. • 30–60 fmol of vector • 90–180 fmol of insert • 4 µl 5× buffer • 1 µl T4 DNA ligase • up to 20 µl water. Perform the ligation in 0.5 ml tubes at 12 ◦ C for 20 h followed by 65 ◦ C for 20 min to inactivate the ligase.

Notes a Puriﬁcation of the PCR product can be achieved using one of two kits. If there is a single PCR product (as visualized on the agarose gel), then a PCR clean-up kit (Qiagen) is advised. If, however, there are multiple PCR products (as visualized on the agarose gel), then the remaining sample must ﬁrst be separated on another agarose gel, the relevant band excised and puriﬁed with a gel extraction clean-up kit (Qiagen). b Puriﬁcation

of the digested vector and PCR product can be achieved using one of two kits. If the fragments excised from the vector and PCR product are smaller than the lower retention limit of the PCR clean-up kit, then this product may be used. If the fragments are large enough to be retained, then the samples must ﬁrst be separated on an agarose gel and the relevant bands excised from the gel and puriﬁed with a gel extraction clean-up kit.

Generally, it is recommended to optimize PCR conditions for each construct made; by far the easiest method is that of temperature-gradient PCR, whereby it is possible to test 12 different annealing temperatures simultaneously. If this is not possible, then approximate annealing temperatures can be calculated from T m values for the

92

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

complementary regions of the primers. A good web site for calculating T m values is www.premierbiosoft.com/netprimer. PCR product puriﬁcation is an important step in improving cloning efﬁciency, as this effectively removes all traces of the polymerase and permits a buffer exchange with minimal loss of the ampliﬁed DNA. There are several commercially available kits on the market, from Qiagen and Promega for example. Once puriﬁed, the PCR product is ready to be digested and ligated into the appropriate vector in a single step using standard techniques. The resulting expression plasmid must be sequenced to conﬁrm the ﬁdelity of the inserted gene before proceeding to yeast transformation.

5.2.3 Transformation and strain selection Sufﬁcient high-quality plasmid DNA for transforming yeast cells can be obtained by ampliﬁcation in a bacterial host such as E. coli XL1-Blue. E. coli clones containing the yeast expression vector are cultured to provide microgram to milligram quantities of plasmid DNA. Bacterial cells are subsequently lysed and the plasmid DNA puriﬁed from the lysate using commercially-available kits such as QIAprep spin miniprep kit.

5.2.3.1 S. cerevisiae transformation Circular plasmid DNA vectors can be used for transforming S. cerevisiae cells by a number of methods, including lithium acetate (LiOAc) transformation and electroporation. Selection pressure must be maintained by growing on a nutrient-limited medium. The LiOAc method uses carrier DNA (e.g. sonicated salmon sperm DNA) to aid uptake of the vector. The S. cerevisiae cells are grown in a complex medium (e.g. yeast extract–peptone dextrose (YPD)) to mid/late log phase and then incubated in a plate solution (Protocol 5.2) with both the vector and carrier DNA. Following incubation the cells are removed from the transformation mix and plated on selective plates and incubated until colonies are visible. The exact composition of the selection plates is entirely dependent upon the selectable marker present in the vector.

PROTOCOL 5.2 Yeast Transformations Equipment and Reagents • Plasmid DNA from Protocol 5.1 • Carrier DNA: 7 mg ml−1 sonicated salmon testes DNA (Sigma–Aldrich) • Plate solution: 40% (w/v) PEG-3350, 1 mM ethylenediaminetetraacetate (EDTA), 10 mM tris(hydroxymethyl)aminomethane hydrochloride (pH 7.5), 0.1 M LiOAc • YPD medium: 1% yeast extract, 2% bacto peptone, 2% dextrose • BEDS solution: 10 mM bicine-NaOH, pH 8.3, 3% (v/v) ethylene glycol, 5% (v/v) dimethyl sulfoxide (DMSO), 1 M sorbitol

5.2 METHODS AND APPROACHES

• YPDS agar: 1% yeast extract, 2% peptone, 2% dextrose, 1 M sorbitol, 2% agar • Dithiothreitol (DTT) • Eppendorf multiporator (Hamburg) • S. cerevisiae: freshly plated • Yeast nitrogen base (YNB) medium: prepare 10× stock by dissolving 134 g of YNB (with ammonium sulfate or amino acids) in 1 l of water, ﬁlter sterilize and store at 4 ◦ C • 1.0 M sorbitol • YNB medium containing 2% dextrose, 1.0 M sorbitol • Zeocin (Invitrogen): when using zeocin plasmids.

Methods S. cerevisiae Transformation 1 Inoculate 10 ml of YPD medium with a loop full of freshly plated S. cerevisiae and culture overnight at 30 ◦ C at 230 rpm. 2 Centrifuge 0.5 ml of the culture in a 1.5 ml microcentrifuge tube at 5000g for 2 min at room temperature and discard the supernatant. 3 Add 2–5 µg of plasmid DNA and 10 µl carrier DNA to the S. cerevisiae cells. Add 0.5 ml plate solution and vortex to resuspend. 4 Incubate the tube stationary at room temperature overnight (approximately 14 h). 5 Withdraw the bottom 100–200 µl from the tube and plate on selective agar. Incubate the plate at 30 ◦ C for 2–3 days or until individual colonies are distinguishable.

Making Competent P. pastoris cellsa 1 Prepare a 5 ml overnight culture of P. pastoris cells in YPD medium at 30 ◦ C and agitate at 250 rpm. 2 Dilute the overnight culture to an OD600 of 0.15–0.20 in a volume of 50 ml YPD medium in a ﬂask large enough to provide good aeration. 3 Grow the yeast culture to an OD600 of 0.8–1.0 at 30 ◦ C with 250 rpm agitation. 4 Centrifuge the culture at 500g for 5 min at room temperature and pour off the supernatant. 5 Resuspend the pellet in 9 ml of ice-cold BEDS solution supplemented with 1 ml 1.0 M DTT. 6 Incubate the cell suspension for 5 min at 100 rpm in a 30 ◦ C shaking incubator. 7 Centrifuge the culture at 500g for 5 min at room temperature and resuspend the cells in 1 ml (0.02 volumes) of BEDS solution without DTT. Aliquot into 40 µl volumes. 8 The competent cells are now ready for transformation. Alternatively, the aliquots may be stored at −80 ◦ C for up to 6 months (do not snap-freeze in liquid nitrogen).

93

94

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

P. pastoris Transformation 1 Add approximately 4 µg of linearized plasmid DNA (usually linearized with SacI, PmeI or BstXI) to a 40 µl aliquot of competent cells in an electroporation cuvette. Incubate for 2 min on ice. 2 Electroporate samples using the following parameters: Eppendorf multiporator 1700 V, 15 ms pulse length. 3 Immediately after electroporation, resuspend the samples in 1 ml cold 1.0 M sorbitol and then plate on selective media (YNB containing 2% dextrose and 1.0 M sorbitol) for auxotrophic strains. Alternatively, if using zeocin-based plasmids, resuspend the samples in 0.5 ml 1.0 M sorbitol and 0.5 ml YPD, incubate in a 30 ◦ C shaker for 1 hb and then plate on media containing increasing concentrations of zeocin (100, 250, 500 or 1000 µg ml−1 ) for the selection of multicopy integrants. 4 Incubate the plates at 30 ◦ C for 3–4 days or until colonies are distinguishable. 5 Pick individual colonies and re-plate on YPDS agar with the same zeocin concentration as the initial selection plate and repeat the incubation in step 5.

Notes a Protocol

adapted from [18].

b

Note that increased numbers of transformants can be achieved for both types of selectable marker by incubating the resuspended cells in a 30 ◦ C shaker for longer periods of time (1–3 h). However, this is partly due to replication of transformants.

5.2.3.2 P. pastoris transformation The linearized plasmid can be integrated into the AOX1 locus by homologous recombination. This forms an integrated expression cassette which gives rise to an expressing strain. The restriction enzymes frequently used for linearization within AOX1 are SacI, PmeI and BstXI (see Section 5.2.2.1). The recommended technique for transformation when using Invitrogen’s expression vectors (using the zeocin resistance marker) is electroporation. This requires the preparation of electro-competent cells which are produced by washing mid-log-phase cells with a buffering solution (Protocol 5.2). This removes media components, such as salts, which could result in arcing during electroporation. Many protocols for the preparation of competent cells are time consuming, requiring multiple washes, centrifugations and incubations. There is, however, a condensed protocol requiring fewer manipulations [18] which yields comparable efﬁciencies when using auxotrophic markers and approximately 20-fold lower efﬁciencies with zeocin selection. Despite this, the method still provides sufﬁcient transformants for selecting multicopy integrants. Following electroporation, the cells are fragile and require careful handling. A ‘recovery’ step is required prior to exposure to selection pressure for positive transformants. Recovery in a rich medium gives the cells the opportunity to regain

5.2 METHODS AND APPROACHES

95

the integrity of the membrane; yeast cells require a minimum recovery time of 45 min (Protocol 5.2). If the method of selection relies on an antibiotic such as zeocin, then clones possessing multiple insertions can be isolated. This is achieved by plating out on agar with increasing zeocin concentrations: clones with multiple copies of the linearized plasmid will be more tolerant to the antibiotic. Such clones can give a two- to three-fold increase in expression [16].

5.2.4 Screening for GPCR production in P. pastoris Screening for heterologous GPCR production in P . pastoris is carried out in 24 deep-well plates using a method based on the high-throughput approach of Boettner et al. [19]: P . pastoris clones are grown in 5 ml buffered glycerol-complex medium (BMGY) overnight to generate biomass prior to induction in 3 ml buffered methanol-complex medium (BMMY). A siliconized rubber cap containing an Erlenmeyer ﬁlter vent permits efﬁcient gaseous exchange allowing the cells to reach OD600 ≈ 25 in the plates over the course of the 53 h induction (Protocol 5.3). Methanol levels must be maintained within the culture within a narrow range in order to attain maximal induction of the AOX1 promoter whilst preventing accumulation of methanol and, hence, cytotoxic shock [20]. GPCR production is subsequently detected by immunoblotting using a small whole-cell culture pellet. Positive clones are selected for scale-up as described in Section 5.2.5. This method has the potential to be extended to S. cerevisiae, although cells do not always grow as efﬁciently as P . pastoris in this set-up.

PROTOCOL 5.3 Screening for GPCR Production in P. pastoris Clones Equipment and Reagents • Uniplate 24 deep-well plates (Whatman) • Bugstopper siliconized rubber cap with Erlenmeyer vent (Whatman) • 20 ml sterile universal (Sarstedt) • BMGY (1% yeast extract, 2% peptone, 100 mM potassium phosphate, pH 6.0, 1.34% YNB, 4 × 10−5 % biotin, 1% glycerol). Prepare 1 l of medium by dissolving 10 g yeast extract and 20 g peptone in 700 ml of water and autoclaving for 20 min. Cool the solution to room temperature and add 100 ml 10× YNB, 100 ml 1 M potassium phosphate buffer, pH 6.0, 2 ml 500× biotin and 100 ml 10× glycerol. The shelf life of the medium is approximately 2 months when stored at 4 ◦ C. • BMMY (1% yeast extract, 2% peptone, 100 mM potassium phosphate, pH 6.0, 1.34% YNB, 4 × 10−5 % biotin, 0.5% methanol). Prepare this medium as described above for BMGY but using 10× methanol instead of 10× glycerol.

96

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

• Stock solutions for BMGY and BMMY: for 10× glycerol (10%), combine 900 ml of water with 100 ml of glycerol, sterilize by autoclaving and store at room temperature. To make 10× methanol (5%), add 95 ml of water to 5 ml of methanol, ﬁlter sterilize and maintain at 4 ◦ C. To prepare 500× biotin (0.02%) dissolve 20 mg biotin in 100 ml of water and ﬁlter sterilize (store at 4 ◦ C). For 1 M potassium phosphate buffer, pH 6.0, mix 868 ml of 1 M KH2 PO4 and 132 ml of 1 M K2 HPO4 (adjust pH with KOH or phosphoric acid to 6.0). Autoclave and store at room temperature. • Protran nitrocellulose transfer membrane (Whatman) • Tween-20 (Sigma–Aldrich) • Marvel (Premier Foods Ltd) • 6× His monoclonal antibody; albumin free (Clontech) • Goat anti-mouse IgG (Fab Speciﬁc) peroxidase conjugate (Sigma–Aldrich) • Phosphate-buffered saline (PBS) • Western transfer buffer (3.3 g Trizma-base, 14.4 g glycine, 20% methanol in 1 l water) • EZ-ECL chemiluminescence solution (Geneﬂow) • Sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis (PAGE) sample buffer (see [21]), SDS– PAGE gels and electrophoresis buffers • Blocking buffer (PBS, 5% Marvel).

Method 1 Select 10 P. pastoris clones for expression screeninga (see Protocol 5.2). 2 Inoculate 5 ml of BMGY media in a 20 ml sterile universal with a single colony of P. pastoris and incubate overnight at 30 ◦ C with shaking at 220 rpm. 3 Determine the OD600 of the overnight 3 ml culture in the morning. 4 Inoculate 3 ml BMMY with appropriate volume of the BMGY culture to achieve OD600 = 1 in the 24 deep-well plates.b 5 Cap the plates with the siliconized rubber caps and incubate for approximately 53 h at 30 ◦ C with shaking at 220 rpm (add methanol to 1% ﬁnal concentration at 24 and 48 h post induction). 6 Harvest 100 µl samples at 24 and 53 h and centrifuge at 10 000g for 5 min, snap-freeze the pellets in liquid nitrogen and store at −80 ◦ C until required. 7 Resuspend the 100 µl cell culture pellets in 60 µl of SDS–PAGE sample buffer. Heat the samples for 10 min at 98 ◦ C and centrifuge for 1 min at 13 000 rpm. 8 Load 15 µl of the supernatant for SDS–PAGE and immunoblotting. Immunoblotting can be carried out using a standard procedurec with the following changes. Carry out blocking in PBS + 5% Marvel for 1 h at room temperature. Add primary antibody to blocking buffer at a dilution of 1 : 5000 and incubate at room temperature for 1 h with

5.2 METHODS AND APPROACHES

97

moderate rocking. Wash the membrane in PBS + 0.2% Tween-20 twice for 5 min after primary and secondary antibody incubations. Add secondary antibody in blocking buffer at a dilution of 1 : 5000 and incubate at room temperature for 1 h with moderate rocking. Visualize bound proteins using EZ-ECL chemiluminescence solution according to manufacturer’s instructions. Notes a If

the GPCR is not detected after screening 40 colonies, then go back, reassess and modify other factors important for producing recombinant clones during earlier steps (see Section 5.2.2). b Method

adapted from Boettner et al. [19].

c See

http://tools.invitrogen.com/content/sfs/manuals/PI96-9045 Western%20Blot% 20Kit%20Rev%201008.pdf.

5.2.5 Scale-up and preparation of membranes After identifying positive P . pastoris clones as described in Section 5.2.4, the process of scaling up can begin. This procedure is also used directly for screening S. cerevisiae transformants, albeit on a 10–50 ml scale. The goal of the scale up process is to increase the yield of recombinant GPCR whilst maintaining its biological activity.

5.2.5.1 Scale-up Shake-ﬂask-scale cultures are typically performed to ﬁnd a strain or transformant that produces biologically active recombinant GPCR at an acceptable production level. In doing so, each growth curve should be compared with a nonproducing wild type. It is worth noting that the fastest growing clone might not necessarily be the best for production [22]. Supplementation of growth media with vitamins and minerals may also increase the growth rate and/or biomass yield. Further supplementation with other components, such as speciﬁc ligands, histidine and DMSO, may also affect the protein produced. For example, Andre et al. demonstrated that while GPCR yield in P . pastoris was not improved by supplementation, speciﬁc activity was increased [23]. The reason for this is not necessarily clear in the case of DMSO, which can alter the physical properties of the membrane bilayer [24] and also has an impact on intracellular regulatory biochemical pathways [25]. Adding speciﬁc ligands, however, probably acts to stabilize GPCRs. GPCR yield and activity should be checked throughout the growth curve. For S. cerevisiae, convenient sampling points are at the high-, mid- and low-glucose phases and during the ethanol utilizing phase [22]. For P . pastoris, sampling shortly after methanol induction then at 12, 24, 48 and 54 h post induction is suitable. Identiﬁcation by immunoblotting and assaying for activity should be performed using standard techniques. Scale-up from shake-ﬂasks to a fermenter/bioreactor can also be carried out to improve yields further. The experimental design needs to focus on deﬁning the variables that affect the quality of the recombinant GPCR. Fed-batch fermentation is most commonly used due to its simplicity and robustness. Particularly important is the dissolved

98

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

oxygen content, especially in the case of P . pastoris, since this can often be a rate limiting factor at high cell density. Consequently, it is useful to use air enriched with oxygen in order to maintain the dissolved oxygen level above 20% at all times.

5.2.5.2 Preparation of membranes Collection of yeast cells is achieved by ﬁltration or centrifugation. This is followed by cellular disruption and isolation of membrane fractions. Cellular disruption can either be carried out immediately or the wet cell mass can be snap-frozen in liquid nitrogen (in thin layers for fast cooling) and stored at −80 ◦ C until required. Different amounts of disruption may be required, depending both on the growth conditions and the growth phase in which the cells are harvested. Consequently, cells should be viewed under a microscope to ensure that disruption has been effective. A wide range of techniques can be applied to disrupt yeast cells and these can be grouped into either mechanical or nonmechanical disruption. Mechanical disruption systems include bead mills, homogenizers and jet streams. Nonmechanical disruption can be physical (sonication, thermolysis, decompression and osmotic shock), chemical (antibiotics, chaotropes, chelating agents and solvents) and enzymatic. Our preferred method is mechanical disruption. For small samples (less than 2 ml), the disruption with glass beads method using a FastPrep (FP120; Thermo Electron Corporation) is very effective (Protocol 5.4). The drawback of using this method is the build up of latent heat; thus, the samples must be cooled on ice between runs. If a FastPrep is not available, then it is possible to use a vortexer and to process the samples by hand. For the disruption of larger samples following fermentation, constant cell disruption systems are very convenient. The EmulsiFlex-C3 marketed by Avestin has a constant ﬂow rate of 3 l h−1 , adjustable homogenizing pressure between 500 and 30 000 psi and a heat exchanger enabling up to 100 ml of suspended cells to be continuously processed. It may be necessary to process the sample more than once (determined by observing the cells under a microscope and assessing the percentage disrupted). Too many passages, however, may lead to the production of ﬁne cell debris which may be difﬁcult to separate from membrane particles. The separation of cellular debris, soluble proteins and membrane particles is achieved in two steps: low-speed centrifugation to spin down cell debris (10 000g, 30 min) followed by high-speed centrifugation of the supernatant (100 000g, 60 min). The membrane fraction is then suspended, snap-frozen in liquid nitrogen and can be stored at –80 ◦ C until required (Protocol 5.4). Further isolation of the plasma membrane fraction can be achieved if required using standard sucrose gradient methods [26].

PROTOCOL 5.4 Preparation of Yeast Membranes Equipment and Reagents • Acid-washed glass beads (0.3–0.5 µm diameter)

5.2 METHODS AND APPROACHES

• Phenylmethylsulfonyl ﬂuoride (PMSF) • 1.5–2 ml screw-cap tube • FastPrep FP120 (Thermo Electron Corporation): for small-scale preparations • EmulsiFlex-C3 (Avestin): for large-scale preparations • Breaking buffer (50 mM sodium phosphate, pH 7.4, 5% glycerol, 2 mM EDTA, 100 mM NaCl), store at 4 ◦ C • Buffer A (20 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES) pH 5.5, 50 mM NaCl, 10% glycerol), store at 4 ◦ C • Homogenizer PTFE pestle/stainless steel rod 15 ml (Scientiﬁc Laboratory Supplies) • Yeast cells, from Protocol 5.3 for example.

Methods Unless stated otherwise, carry out all the procedures below on ice.

Small Scale (<0.5 g Cells) 1 Mix approximately 100–300 mg of wet yeast cells with 300–500 µl of acid-washed glass beads and 300–500 µl ice-cold breaking buffer in a 1.5–2 ml screw-cap tube. Add PMSF to a ﬁnal concentration of 2 mM (other protease inhibitors may be added according to the manufacturer’s instructions). 2 Agitate the tubes in a FastPrep FP120 set at level 6 for a period of 40 s. Place the tubes on ice for 1–2 min to dissipate any residual heat. 3 Repeat step 2 at least ﬁve times. 4 Following disruption, observe the cells under a light microscope to ascertain the extent of cell breakage (>70% should be disrupted). Repeat steps 2 and 3 if necessary. 5 Remove the supernatant from the glass beads and place in a fresh tube. Wash the beads with an equal volume of ice-cold breaking buffer and add to the recovered supernatant.a 6 Clarify the supernatant at 10 000g for 30 min at 4 ◦ C. 7 Isolate membranes from the clariﬁed supernatant at 100 000g for 60 min at 4 ◦ C and resuspend the pellet in 50 µl ice-cold buffer A.

Medium Scale (5–10 g Cells) 1 To 5 g of wet cells add 7.5 ml of acid-washed glass beads, 7.5 ml of ice-cold breaking buffer and PMSF to ﬁnal concentration of 2 mm. Vortex the suspension for 1–2 min and subsequently cool on ice for a period of 1–2 min. 2 Repeat the vortex steps for 10–20 cycles. 3 Determine the extent of cell breakage and repeat disruption if necessary, clarify the samples and recover the membranes as stated above.

99

100

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

Large Scale (>10 g) 1 Resuspend the cell pellet in ice-cold breaking buffer at a ratio of 2 : 1 buffer to pellet (v/w). 2 Pass the cell suspension through an Emulsiﬂex-C3 cell disrupter ﬁtted with a chilled heat exchanger (Avestin) four times according to the manufacturer’s instructions. Observe the cells under a light microscope to check the extent of cell breakage. The breaking efﬁciency should be >90% at a homogenizing pressure of 30 000 psi. 3 Remove the unbroken cells and cellular debris by centrifugation at 10 000g for 30 min at 4 ◦ C. 4 Centrifuge the clariﬁed supernatant at 100 000g for 90 min at 4 ◦ C to collect the membrane fraction. 5 Resuspend the membrane pellet in ice-cold buffer A using a glass homogenizer at a ratio of 10 ml buffer per gram of pellet.

Notes a The

amount of breaking buffer used to wash the glass beads can be adjusted so as not to exceed the combined ﬁlling volume of the tubes.

5.2.6 Solubilization and puriﬁcation An important step in the production and ultimate structural and functional characterization of GPCRs is the ability to solubilize the protein in detergent effectively prior to puriﬁcation. The correct choice of detergent is critical to ensure maximal return of biologically active, correctly folded protein. Detergent purity, its ability to maintain biological activity of the protein over varying solubilizing concentrations, its solubility at the desired working temperature, methods for its removal and potential conﬂicts with the needs of downstream puriﬁcation or crystallization all need to be considered. A key parameter is the critical micelle concentration (CMC), which is deﬁned as the lowest concentration above which detergent monomers cluster to form micelles. Proteins then incorporate into these micelles via hydrophobic interactions. Typically, it is necessary to work at or just above the CMC, since micelles form over a narrow concentration range. In order to identify the best detergent for GPCR solubilization, a panel of detergents each of varying concentrations needs to be evaluated. Solubilization screening with a panel of 10 different detergents, ranging from nonionic to zwitterionic molecules, is a good start point (Protocol 5.5). Nonionic detergents, such as n-octyl-β-d-glucopyranoside (β-OG), n-dodecylphosphocholine (DPC) and lauryldimethylamine oxide (LDAO), are often more successful in yielding soluble GPCRs than the zwitterionic cyclofos-4 (CYFOS-4) and foscholineiso9 (FC109). Recent studies have identiﬁed several detergents and optimized conditions for general membrane protein solubilization, including SoPIP2;1 using β-OG [27], PagP barrel with CYFOS [28], SERCA with DPC [29] and Arabidopsis thaliana leaf membrane proteins with

5.2 METHODS AND APPROACHES

101

Brij-n [30]. Furthermore, β-OG, LDAO, C8E4 and n-dodecyl-β-d-maltopyranoside (DDM) are routinely used amongst protein crystallographers and have facilitated numerous crystal structure studies (see http://www.mpibp-frankfurt.mpg.de/michel/ public/memprotstruct.html for an analysis of crystallization conditions for structures published up to 2006). It must be noted that there are no hard and fast rules regarding which detergents to choose, but research is revealing that certain proteins do solubilize better with a particular detergent. For example, White et al. [31] have suggested that membrane proteins possessing greater than 70% hydrophobic residues in their predicted transmembrane domains would solubilize better with β-OG and C8E4 than with other detergents compared with proteins with less than 70% hydrophobic residues. Note that it is also possible to solubilize efﬁciently in one detergent and then detergent exchange into a second more favourable detergent for downstream protein processing.

PROTOCOL 5.5 GPCR Solubilization and Puriﬁcation Equipment and Reagents • Fixed-speed roller mixer (Sturat SRT 9) • Reagents for immunoblotting (see Protocol 5.3) • Solubilization buffer (20 mM HEPES pH 7.0, 100 mM NaCl, 10% glycerol) • Homogenizer (homogenizer PTFE pestle/stainless steel rod 15 ml (SLS)) • DC protein assay (Bio-Rad Laboratories) • Fixed-speed ﬂatbed roller (Stuart) • Glass homogenizer (SLS) • BioRad DC protein assay kit • SDS–PAGE sample buffer [21], SDS– PAGE gels and electrophoresis buffers • Breaking buffer (50 mM sodium phosphate, pH 7.4, 5% glycerol, 2 mM EDTA, 100 mM NaCl), store at 4 ◦ C • 1 ml HisTrap HP column (GE Healthcare).

Method 1 Prepare the membranes as outlined in the large-scale section of Protocol 5.4 but omitting the resuspension step 5. 2 Resuspend the membrane pellet in solubilization buffer using a glass homogenizer and add 10 ml of buffer per gram of pellet. Determine the protein concentration using a BioRad DC protein assay kit according to the manufacturer’s instructions and resuspend the membrane pellet to a ﬁnal concentration of 10 mg protein per millilitre. 3 Proceed to step 4 or freeze in liquid nitrogen and store at −80 ◦ C until required.

102

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

4 For small-scale solubilisation, add 50 µl of the membrane fraction to 950 µl of solubilization buffer containing the required detergent concentration and incubate at room temperature for 1 h on a ﬁxed-speed roller mixer. 5 Pellet the nonsolubilized material at 100 000g, 1 h at 4 ◦ C and resuspend any visible pellet in 200 µl breaking buffer. 6 Load the supernatant and the resuspended pellet (mixed 1 : 1 in 2× SDS–PAGE sample buffer) for SDS–PAGE and immunoblot analysis.a 7 For a large-scale solubilization (50–100 ml), add the preferred detergent to ﬁnal working concentration (1–5%), as determined from the small-scale screen,b incubate and process as above. 8 Use the supernatant for puriﬁcation of the tagged receptor with IMAC using a 1 ml His Trap column according to manufacturer’s instructions (see Section 5.2.6.1). Notes a Refer

to Protocol 5.3, step 8, for instructions on immunoblotting.

bA

panel of detergents and working concentrations should be screened for effective solubilization of the recombinant protein using the following panel of detergents at 2% w/v on membrane suspensions: DDM, DPC, CYFOS-4, β-OG, foscholineiso9 (FC109), LDAO, pentaethyleneglycol-n-octylether (C8E5), DPC–cholesterolhemisuccinate and docosaethyleneglycol monohexadecylether (Brij 58; all purchased from Anatrace Inc). Analyse the solubilized (supernatant) and nonsolubilized (membrane pellet) material by immunoblot analysis for solubilization efﬁciency using a primary his monoclonal antibody and an anti-mouse IgG HRP-conjugated secondary antibody (Sigma).

5.2.6.1 Protein puriﬁcation Having successfully solubilized a GPCR from yeast membranes, the protein is now ready for puriﬁcation using afﬁnity chromatography dependent upon the tag added to the protein, as discussed in Section 5.2.2.2. One of the most popular methods is puriﬁcation using immobilized metal afﬁnity chromatography (IMAC). HisTrap HP prepacked Ni-Sepharose columns (GE Healthcare) allow quick and efﬁcient puriﬁcation and have binding capacities of at least 40 mg His6 -tagged protein per millilitre medium. These columns can be cleaned, stripped, recharged and reused, making them cost effective. Neutral or slightly alkaline buffers, such as sodium phosphate (pH 7–8), containing 0.5–1.0 m NaCl should be considered. Typically, 300 mm NaCl works well to prevent nonspeciﬁc, ionically bound host proteins in combination with the selected detergent at the lowest concentration to maintain protein solubilization and minimize aggregation. Addition of imidazole at low concentrations (∼20–30 mm) in the sample, binding and wash buffers is important to further minimize host cell protein binding to the puriﬁcation matrix. Elution of the protein is best achieved with a linear elution gradient of increasing imidazole to achieve high yield and high purity. Further detailed information can be found at www.gehealthcare.com/hitrap.

103

5.2 METHODS AND APPROACHES

5.2.7 Analysis of recombinant proteins To date there are X-ray crystal data for only one GPCR, that of bovine rhodopsin and its mutants [1, 32]. Electron microscopy has also been used to study rhodopsin in its native membranes [33, 34]. In addition, at least 13 other GPCRs have been characterized by noncrystallographic methods [35]. Various methods of protein analysis are discussed below, highlighting the basic principles and the advantages and disadvantages of each method. More detailed explanations for all the techniques can be found in the many specialized books available and within other chapters of this book. In particular, biological assays are covered elsewhere, but the importance of determining the amount of active GPCR in a sample, and the efforts to purify it ultimately in a homogeneous, fully functional form, cannot be overstated.

5.2.7.1 Immunoblot analysis Immunoblot analysis has already been mentioned in Protocol 5.3 and is a standard technique. Basically, the method involves the transfer of proteins from an SDS–PAGE gel (separated on the basis of size) to an inert membrane such as nitrocellulose or polyvinylidene ﬂuoride and the subsequent immunodetection of the recombinant GPCR visualized on either X-ray ﬁlm or digitally captured via a charge-coupled device camera. The most sensitive method of detection uses a peroxidise-tagged secondary antibody developed by chemiluminescence (allowing detection of <1 pg protein). The approximate sensitivity of other methods is shown below. Radiolabel Peroxidase chemiluminescence Peroxidase colorimetric Phosphatase chemiluminescence Phosphatase colorimetric

∼10 ng <1 pg ∼500 pg ∼10 pg ∼100 pg

The most common problem associated with immunoblotting is false positives that may arise from the nonspeciﬁc interaction of the primary and/or secondary antibodies with proteins other than the GPCR and/or its tag. This is usually caused by the use of poor quality polyclonal antibodies and can often be overcome by using higher purity monoclonal antibodies where possible and optimizing their dilutions. Whilst this is a useful method for initial detection of GPCR production and approximate size estimation, it does not give any information on biological activity and cannot generally be used to quantify yields.

5.2.7.2 Circular dichroism CD spectroscopy relies on the differential absorption of left and right circular polarized light by chromophores. The most signiﬁcant chromophore in a GPCR is that of the amide group of the polypeptide backbone, and the secondary structure of the GPCR (α-helix and β-sheet) imposes positional constraints on the CD transitions, resulting in

104

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

individual far-ultraviolet (UV) spectra (180–240 nm). The CD spectrum in the near-UV region (260–320 nm) arises from the amino acid side chains and can be interpreted to indicate the tertiary structure of the protein. Most conventional laboratory CD spectrophotometers use a xenon arc light source capable of producing wavelengths of 180–1000 nm; however, to generate spectra below 180 nm, vacuum UV or synchrotron radiation sources must be used [36]. The main advantages of this method of analysis are its speed and, owing to the nondestructive nature of the technique, it is often possible to recover the material for use in repeated experiments. Furthermore, it is possible to obtain a spectrum within 30 min, with far-UV studies requiring very small amounts of sample (100–500 µg). For measurements in the near-UV and visible spectrum, much larger quantities of protein are required (of the order of several milligrams), reﬂecting the lower amounts of chromophores compared with the peptide bonds. Instrumentation developments have also led to stopped-ﬂow CD spectroscopy, whereby it is now possible to record structural changes occurring on a tens of millisecond timescale, thus enabling the study of events occurring during the early stages of reactions. The main drawback of CD spectroscopy is that the structural information gained is of low resolution. Whilst it can give an estimate of the level of secondary and tertiary structure present, it gives little quaternary information. Furthermore, it does not give any indication where these regions occur within the protein and, as such, must be employed alongside other structural techniques. A recommended overview of CD spectroscopy is by Kelly and Price [37] and examples of GPCR peptide analysis can be found in the literature [38, 39].

5.2.7.3 Linear dichroism The principles of linear dichroism (LD) spectroscopy are similar to those of CD spectroscopy, except that the signal produced is generated from the differential absorption of linearly polarized radiation perpendicular and parallel to the direction of orientation. LD signals can be either positive (transitions where the polarization is parallel to the direction) or negative (where the transition is perpendicular). The LD spectrum of far-UV regions can generate information about the peptide backbone of the protein and, as such, can often be used to verify data obtained by CD. Furthermore, LD spectroscopy can indicate whether the GPCR is inserted within the membrane or bound to the surface. Prolonged measurements can be made to follow the dynamics of GPCR insertion. The use of LD spectroscopy for structure determination is discussed by Dafforn et al. [40].

5.2.7.4 Analytical ultracentrifugation Analytical ultracentrifugation (AUC) is a versatile technique that can be used to characterize the solution state of macromolecules, including detergent-solubilized GPCRs. It can be used to determine sample purity, stoichiometry and even conformational changes. Since the sample is visualized in real time during sedimentation, it allows for accurate determination of both hydrodynamic and thermodynamic parameters; and as the experiments are performed in solution, there are no complications arising from

5.2 METHODS AND APPROACHES

105

matrix interference, as there is with gel ﬁltration. In order to avoid protein charge effects, a minimum ionic strength of 50 mm is often recommended, with the samples being pre-equilibrated with an appropriate buffer by dialysis. Two types of experiment can be performed. • First, the sedimentation velocity technique is responsive to both the shape and mass of the protein. A moving boundary is generated by the centrifugal ﬁeld and the sample scanned in parallel to a reference cell at regular intervals to determine the rate of movement and broadening of the boundary as a function of time. • Second, sedimentation equilibration is a thermodynamic technique sensitive to the mass of the protein, but not its shape. Such experiments are often performed at lower centrifugal forces and measure the equilibrium concentration distribution of the protein that ultimately is generated when sedimentation is balanced by diffusion and can indicate the oligomeric nature of the sample. Both measurements can be used to provide complementary data. Useful books on AUC are those by Cole and Hansen [41], Scott et al. [42] and Schuster and Laue [43].

5.2.7.5 Fourier transform infrared spectroscopy The use of Fourier transform infrared spectroscopy (FTIR) to determine the structure of biological macromolecules has increased in recent years. Whilst the complete three-dimensional structure of a protein at high resolution can be determined by X-ray crystallography, this initially requires the molecule to form a well-ordered crystal, which is not always possible, especially for GPCRs. FTIR spectroscopy often requires small amounts of proteins (1 mm); therefore, high-quality spectra can be obtained with low background ﬂuorescence, light scattering and problems related to the size of the proteins. A Fourier transform is a mathematical operation used to translate a complex curve into its component curves. In a Fourier transform infrared instrument, the complex curve is an interferogram, or the sum of the constructive and destructive interferences generated by overlapping light waves, and the component curves are the infrared spectrum. The standard infrared spectrum is calculated from the Fourier-transformed interferogram, giving a spectrum in percent transmittance (%T) versus light frequency (cm−1 ). FTIR can often provide more sensitive details of beta-sheet structures than that of CD and it is possible to detect aggregation arising through formation of intermolecular beta sheets. FTIR has been used to investigate transmembrane helix 8 in bovine rhodopsin [44, 45].

5.2.7.6 Nuclear magnetic resonance spectroscopy An alternative to X-ray crystallography is multidimensional nuclear magnetic resonance (NMR) spectroscopy. Using NMR spectroscopy, structures of proteins can be determined in solution. However, the interpretation of the NMR spectra of large proteins

106

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

is very complex, so its application is often limited to small proteins of approximately 15–25 kDa. Protein NMR is carried out on highly puriﬁed aqueous protein (300–600 µl) at a concentration of between 0.1 and 3 mm that has been isotopically labelled. NMR spectroscopy exploits the fact that the most abundant isotopes of carbon (12 C) and oxygen (16 O) have no net nuclear spin. Whilst 14 N has a net nuclear spin, it also has a large quadrupolar moment which prevents high-resolution information being obtained. Therefore, protein NMR is restricted to utilizing the resonance based solely on protons and by labelling with the less common isotopes, 13 C and 15 N. Isotopic labelling is achieved by culturing the production host in a growth medium enriched with the desired isotopes. However, these isotope-enriched compounds can be prohibitively expensive. Essentially, yeasts are cultured on a deﬁned minimal medium, containing a single 13 C source (glucose, glycerol or methanol) and a 15 N source (ammonium sulfate or ammonium chloride). The GPCR is subsequently puriﬁed as previously discussed and diluted in a buffer adjusted to the desired solvent conditions. As yet, no high-resolution structure of a GPCR has been achieved. However, NMR spectroscopy has been used to investigate some GPCR fragments. These include S. cerevisiae Ste2p peptides [38–46], rat angiotensin II AT1A receptor fragments [47], human cholecystokinin-2 receptor fragments [48] and, more recently, complete chemokine receptors [49]. Useful books on NMR are those by Cavanagh et al. [50], Downing [51] and Rule and Hitchens [52].

5.3 Troubleshooting Whilst sometimes laborious and time consuming, PCR ampliﬁcation and cloning can usually be achieved using standard methods, and problems can be resolved using well-documented troubleshooting steps. By far the most challenging step, however, is obtaining an expressing clone. Once an expressing clone has been identiﬁed, in our experience it is then likely that it can be scaled up and that recombinant protein can be solubilized. The subsequent challenge is to obtain a stable, homogeneous sample of functional protein. Detection of the recombinant GPCR using antibodies raised against speciﬁc tags (e.g. His6 , HA) does not necessarily guarantee that the protein is correctly folded and inserted into the plasma membrane. It may be possible in a limited number of cases to use antibodies raised against speciﬁc topological regions of the GPCR, but this is only possible if the antibodies themselves or the time and resources to make them are available. The detection of false positive clones can be limited by not screening whole-cell extracts, but rather the membrane fractions. Another problem that may be encountered is that membrane proteins often have an electrophoretic mobility in SDS–PAGE that differs from that expected. In some cases, whether it is due to insufﬁcient detergent or the nature of the protein, it is possible to detect dimeric, trimeric, or multimeric structures and/or combinations of these along with the monomer. To overcome this problem, try increasing the denaturing time and/or adjusting the denaturation temperature in conjunction with different percentages of SDS–PAGE gels.

REFERENCES

107

Producing a stable, functional recombinant GPCR involves choosing the right puriﬁcation protocol and optimizing it. These days, afﬁnity chromatography, especially IMAC, is the method of choice. During the solubilization and puriﬁcation steps, delipidation can take place and the GPCR then becomes inactive. Post puriﬁcation it may be possible to reactivate the GPCR by reconstituting it with speciﬁc lipids, but in many cases the activity cannot be restored. One solution for this problem is adding lipids to the buffers used throughout the solubilization and puriﬁcation process. Following afﬁnity puriﬁcation, size-exclusion chromatography can be employed to determine whether the GPCR is homogeneous, which is often a prerequisite for crystallization trials. The addition of glycerol, ethylene glycol or other substrates to mimic osmotic pressure could also be considered in an effort to stabilize the protein. All these considerations must be viewed in the context of our opinion that the GPCRs most likely to be produced in quantities amenable to crystallization trials are the ones which are best understood biochemically; an investment in understanding the biochemistry of the target GPCR will usually pay dividends in the long run.

References 1. Standfuss, J. Xie, G., Edwards, P. et al. (2007) Crystal structure of a thermally stable rhodopsin mutant. J. Mol. Biol., 372, 1179–1188. The ﬁrst structure of a recombinant GPCR from a mammalian cell-line. 2. Fraser, N.J. (2006) Expression and functional puriﬁcation of a glycosylation deﬁcient version of the human adenosine 2a receptor for structural studies. Protein Expr. Purif., 49, 129–137. A good example of how a thorough understanding of the target GPCR is used to maximise functional yield. 3. Sarramegna, V. Muller, I., Mousseau, G. et al. (2005) Solubilization, puriﬁcation, and mass spectrometry analysis of the human mu-opioid receptor expressed in Pichia pastoris. Protein Expr. Purif., 43, 85–93. 4. Tate, C.G. (2001) Overexpression of mammalian integral membrane proteins for structural studies. FEBS Lett., 504, 94–98. An important overview of key considerations in this area. 5. Bill, R.M. (2001) Yeast – a panacea for the structure–function analysis of membrane proteins? Curr. Genet., 40, 157–171. 6. Tate, C.G., Whiteley, E. and Betenbaugh, M.J. (1997) Molecular chaperones improve functional expression of the serotonin (5-hydroxytryptamine) transporter in insect cells. Biochem. Soc. Trans., 27, 932–936. 7. Tate, C.G., Whiteley, E. and Betenbaugh, M.J. (1999) Molecular chaperones stimulate the functional expression of the cocaine-sensitive serotonin transporter. J. Biol. Chem., 274, 17551–17558. 8. Tate, C.G., Haase, J., Baker, C. et al. (2003) Comparison of seven different heterologous protein expression systems for the production of the serotonin transporter. Biochim. Biophys. Acta, 1610, 141–153. 9. Lagane, B. Gaibelet, G., Meilhoc, E. et al. (2000) Role of sterols in modulating the human mu-opioid receptor function in Saccharomyces cerevisiae. J. Biol. Chem., 275, 33197–33200.

108

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

10. Sarramegna, V., Demange, P., Milon, A. and Talmont, F. (2002) Optimizing functional versus total expression of the human mu-opioid receptor in Pichia pastoris. Protein Expr. Purif., 24, 212–220. 11. Lacapere, J.J., Pebay-Peyroula, E., Neumann, J.M. and Etchebest, C. (2007) Determining membrane protein structures: still a challenge! Trends Biochem. Sci., 32, 259–270. 12. Wallace, B.A., Lees, J.G., Orry, A.J. et al. (2003) Analyses of circular dichroism spectra of membrane proteins. Protein Sci., 12, 875–884. 13. Invitrogen Corporation (1997) A Manual of Methods for Expression of Recombinant Proteins in Pichia pastoris, Invitrogen Corporation, San Diego, CA. The key manual for P . pastoris. 14. Faber, K.N., Harder, W., Ab, G. and Veenhuis, M. (1995) Review: methylotrophic yeasts as factories for the production of foreign proteins. Yeast , 11, 1331–1344. 15. Miyasaka, H. (1999) The positive relationship between codon usage bias and translation initiation AUG context in Saccharomyces cerevisiae. Yeast , 15, 633–637. 16. Weiss, H.M., Haase, W., Michel, H. and Reilander, H. (1998) Comparative biochemical and pharmacological characterization of the mouse 5HT5A 5-hydroxytryptamine receptor and the human beta2-adrenergic receptor produced in the methylotrophic yeast Pichia pastoris. Biochem. J., 330, 1137–1147. 17. Zhang, R., Kim, T.K., Qiao, Z.H. et al. (2007) Biochemical and mass spectrometric characterization of the human CB2 cannabinoid receptor expressed in Pichia pastoris – importance of correct processing of the N-terminus. Protein Expr. Purif., 55, 225–235. 18. Lin-Cereghino, J., Wong, W.W., Xiong, S. et al. (2005) Condensed protocol for competent cell preparation and transformation of the methylotrophic yeast Pichia pastoris. Biotechniques, 38, 44–48. 19. Boettner, M., Prinz, B., Holz, C. et al. (2002) High-throughput screening for expression of heterologous proteins in the yeast Pichia pastoris. J. Biotechnol., 99, 51–62. 20. Gaurna, M.M., Lesnicki, G.J., Tam, B.M. et al. (1997) On-line monitoring and control of methanol concentration in shake-ﬂask cultures of Pichia pastoris. Biotechnol. Bioeng., 56, 279–286. 21. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.), 227, 680–685. 22. Bonander, N., Hedfalk, K., Larsson, C. et al. (2005) Design of improved membrane protein production experiments: quantitation of the host response. Protein Sci., 14, 1729–1740. 23. Andre, N., Cherouati, N., Prual, C. et al. (2006) Enhancing functional production of G protein-coupled receptors in Pichia pastoris to levels required for structural studies via a single expression screen. Protein Sci., 15, 1115–1126. 24. Murata, Y., Watanabe, T., Sato, M. et al. (2003) Dimethyl sulfoxide exposure facilitates phospholipid biosynthesis and cellular membrane proliferation in yeast cells. J. Biol. Chem., 278, 33185–33193. 25. Yu, Z.W. and Quinn, P.J. (1994) Dimethyl sulphoxide: a review of its applications in cell biology. Biosci. Rep., 14, 259–281. 26. Walworth, N.C., Goud, B., Ruohola, H. and Novick, P.J. (1989) Fractionation of yeast organelles. Methods Cell. Biol., 31, 335–356.

REFERENCES

109

27. Karlsson, M., Fotiadis, D., Sjovall, S. et al. (2003) Reconstitution of water channel function of an aquaporin overexpressed and puriﬁed from Pichia pastori s. FEBS Lett., 537, 68–72. 28. Hwang, P.M., Bishop, R.E. and Kay, L.E. (2004) The integral membrane enzyme PagP alternates between two dynamically distinct states. Proc. Natl. Acad. Sci. U. S. A., 101, 9618–9623. 29. Zamoon, J., Nitu, F., Karim, C. et al. (2005) Mapping the interaction surface of a membrane protein: unveiling the conformational switch of phospholamban in calcium pump regulation. Proc. Natl. Acad. Sci. U. S. A., 102, 4747–4752. 30. Luche, S., Santoni, V. and Rabilloud, T. (2003) Evaluation of nonionic and zwitterionic detergents as membrane protein solubilizers in two-dimensional electrophoresis. Proteomics, 3, 249–253. 31. White, M.A., Clark, K.M., Grayhack, E.J. and Dumont, M.E. (2007) Characteristics affecting expression and solubilization of yeast membrane proteins. J. Mol. Biol., 365, 621–636. 32. Palczewski, K., Kumasaka, T., Hori, T. et al. (2000) crystal structure of rhodopsin: a G protein-coupled receptor. Science, 289, 739–745. 33. Fotiadis, D., Jastrzebska, B., Philippsen, A. et al. (2006) Structure of the rhodopsin dimer: a working model for G-protein-coupled receptors. Curr. Opin. Struct. Biol., 16, 252–259. 34. Fotiadis, D., Liang, Y., Filipek, S. et al. (2004) The G protein-coupled receptor rhodopsin in the native membrane. FEBS Lett., 564, 281–288. 35. Yeagle, P.L. and Albert, A.D. (2007) G-protein coupled receptor structure. Biochim. Biophys. Acta Biomembr., 1768, 808–824. 36. Wallace, B.A. and Janes, R.W. (2001) Synchrotron radiation circular dichroism spectroscopy of proteins: secondary structure, fold recognition and structural genomics. Curr. Opin. Chem. Biol., 5, 567–571. 37. Kelly, S.M. and Price, N.C. (2000) The use of circular dichroism in the investigation of protein structure and function. Curr. Protein Pept. Sci., 1, 349–384. 38. Arevalo, E., Estephan, R., Madeo, J. et al. (2003) Biosynthesis and biophysical analysis of domains of a yeast G protein-coupled receptor. Biopolymers, 71, 516–531. 39. Cano-Sanchez, P., Severino, B., Sureshbabu, V.V. et al. (2006) Effects of N- and C-terminal addition of oligolysines or native loop residues on the biophysical properties of transmembrane domain peptides from a G-protein coupled receptor. J. Pept. Sci., 12, 808–822. 40. Dafforn, T.R., Rajendra, J., Halsall, D.J. et al. (2004) Protein ﬁber linear dichroism for structure determination and kinetics in a low-volume, low-wavelength couette ﬂow cell. Biophys. J., 86, 404–410. 41. Cole, J.L. and Hansen, J.C. (1999) Analytical ultracentrifugation as a contemporary biomolecular research tool, J. Biomol. Tech. 10, 163–176. 42. Scott D., Harding S.E. and Rowe A.J. (eds) (2005) Analytical Ultracentrifugation: Techniques and Methods, Royal Society of Chemistry. 43. Schuster T.M. and Laue T.M. (eds) (1994) Modern Analytical Ultracentrifugation: Acquisition and Interpretation of Data for Biological and Synthetic Polymer Systems, Emerging Biochemical & Biophysical Techniques, Birkhauser Verlag AG. 44. Fahmy, K. (2002) FTIR- and ﬂuorescence-spectroscopic analyses of receptor G-protein coupling in photoreception. Curr. Org. Chem., 6, 1259–1284.

110

CH 5 PRODUCTION OF RECOMBINANT G PROTEIN-COUPLED RECEPTOR IN YEAST

45. Lehmann, N., Alexiev, U. and Fahmy, K. (2007) Linkage between the intramembrane H-bond network around aspartic acid 83 and the cytosolic environment of helix 8 in photoactivated rhodopsin. J. Mol. Biol., 366, 1129–1141. 46. Naider, F., Arshava, B., Ding, F.X. et al. (2001) Peptide fragments as models to study the structure of a G-protein coupled receptor: the α-factor receptor of Saccharomyces cerevisiae. Biopolymers, 60, 334–350. 47. Franzoni, L., Nicastro, G., Pertinhez, T.A. et al. (1997) Structure of the C-terminal fragment 300–320 of the rat angiotensin II AT1A receptor and its relevance with respect to G-protein coupling. J. Biol. Chem., 272, 9734–9741. 48. Giragossian, C. and Mierke, D.F. (2003) Determination of ligand-receptor interactions of cholecystokinin by nuclear magnetic resonance. Life Sci., 73, 705–713. 49. Park, S.H., Prytulla, S., De Angelis, A.A. et al. (2006) High-resolution NMR spectroscopy of a GPCR in aligned bicelles. J. Am. Chem. Soc., 128, 7402–7403. 50. Cavanagh, J., Fairbrother, W.J., Palmer, A.G., Skelton, N.J. and Rance, M. (2006) Protein NMR Spectroscopy: Principles and Practice, 2nd edition, Academic Press. 51. Downing A.K. (ed.) (2004) Protein NMR Techniques, Methods in Molecular Biology, Humana Press. 52. Rule, G.S. and Hitchens, T.K. (2005) Fundamentals of Protein NMR Spectroscopy, Focus on Structural Biology, Springer Verlag.

6 Monitoring GPCR–Protein Complexes Using Bioluminescence Resonance Energy Transfer Werner C. Jaeger, Kevin D.G. Pﬂeger and Karin A. Eidne Laboratory for Molecular Endocrinology – GPCRs, Western Australian Institute for Medical Research and Centre for Medical Research, University of Western Australia, Perth, Australia

6.1 Introduction The formation of multiple protein complexes throughout the activation life cycle of G protein-coupled receptors (GPCRs) is essential for the wide-ranging array of functions mediated by this receptor superfamily. GPCRs form complexes with various proteins from initial origins at the endoplasmic reticulum (ER) through to their degradation. The process of maturation and piloting of some GPCRs from the ER to the cell surface has been shown to require single transmembrane domain chaperone proteins. In addition, these proteins may inﬂuence GPCR effector function. Differential functional activity is displayed by the calcitonin receptor-like-receptor, as it shows different peptide-responsive phenotypes depending on the interaction with three isoforms of the receptor activity-modifying proteins (RAMP1, RAMP2/3) at the ER [1–3]. Another recently discovered single-transmembrane protein, the melanocortin receptor accessory protein, has been shown to be important for proper cell membrane integration and signalling capabilities of the melanocortin receptor 2 (MC2R) [4, 5]. Moreover, mutations in this protein have exhibited effects on the function of MC2R [4, 6]. The binding of ligands, such as agonists and inverse agonists, is necessary to increase G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

112

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

or reduce the level of basal receptor activity respectively [7]. Upon activation of a GPCR, the onset of the deactivation/desensitization phase occurs rapidly and commonly through the interaction with a family of kinases, the G protein-coupled receptor kinases (GRKs) [8]. These proteins phosphorylate serine and/or threonine residues on the carboxy-terminal tail or third intracellular loop and dramatically increase the binding afﬁnity for a family of multifunctional adaptor proteins, the β-arrestins [9, 10]. Interactions with β-arrestins are being studied intensively due to their wide-ranging list of associated functions [11, 12]. Interactions between β-arrestin 1 or 2 and GPCRs are primarily implicated in desensitization and internalization of activated GPCRs. These processes are responsible for terminating the initial GPCR–G protein interaction and thus attenuating the G protein-mediated signalling pathway, trafﬁcking of activated GPCR complexes to various compartments within the cell and recycling the receptor back to the cell surface [13]. The activity of GPCRs during these phases is characterized by two different GPCR–β-arrestin-binding kinetic proﬁles [14] and further evidence exists for a third class [15, 16]. These proﬁles characterize the localization, degradation and signalling pathways that internalized GPCR complexes exhibit. The extended monitoring of dynamic GPCR–β-arrestin interactions has provided a wealth of information [13, 17], allowing insight into the intracellular activities of GPCRs and how the physiology of a cell changes in response to external stimuli. GPCR-mediated cellular signalling has traditionally been observed to occur through the activation of G proteins [18–21]. The multifunctional roles of β-arrestins have been extended through characterization of G protein-independent, β-arrestin-mediated signalling pathways [22]. GPCR-mediated signalling events have recently been discovered through the ability of GPCR–β-arrestin complexes to scaffold kinases involved in downstream signalling cascades [23–27]. These signalling activities are secondary to, but not necessarily isolated from, the initial G protein signalling event and include activation of the family of mitogen-activated protein kinases (MAPKs) involved in transcriptional regulation and cell cycle progression [28]. Covalent modiﬁcations of GPCRs, including phosphorylation, palmitoylation and ubiquitination, are dynamic interactions that play key roles in cellular sorting, membrane association and general regulation of GPCR function [29]. The dynamic relationship of GPCRs and β-arrestins with ubiquitin has recently provided a greater depth of knowledge concerning the regulation of GPCR internalization, degradation and signalling processes [30–34]. Indeed, through the study of β-arrestin with wild-type and mutant chimeric forms of two well-studied GPCRs, β2-adrenergic receptor and V2 vasopressin receptor [27, 32, 35], it has become apparent that ubiquitination is indispensable for differential compartmentation of GPCR complexes for lysosomal or proteasomal degradation. In addition, the dynamics of β-arrestin ubiquitination directly regulate the scaffolding of signalling kinases such as the MAPKs (e.g. extracellular signal-regulated kinases 1 and 2), thus providing an important determining factor in the level of G protein-independent signalling between the two β-arrestin-binding classes of GPCRs [27]. This overview of vital interactions between GPCRs and associated proteins illustrates the necessity to characterize spatiotemporal interactions that inevitably result in a change in cell physiology. The characterization of dynamic GPCR–protein interactions

6.1 INTRODUCTION

113

has recently been approached with renewed vigour due in part to the number of putative GPCRs and interacting protein sequences gathered from the success of the human genome project [36, 37], as well as the evolution of more efﬁcient and functionally relevant techniques to characterize these interactions.

6.1.1 Measurement of GPCR–protein interactions and the potential of bioluminescence resonance energy transfer (BRET) In order to deﬁne novel effector functions of a particular GPCR and characterize the activity of key molecular associations between GPCRs and their interacting proteins, biochemical and biophysical techniques constitute the mainstream set of methods employed. The most recent and exciting development of these methodologies involves using the biophysical principle of resonance energy transfer (RET). Two core forms of RET have been characterized and developed: ﬂuorescence resonance energy transfer (FRET) and the sister technique bioluminescence resonance energy transfer (BRET). These methods have been increasingly employed by researchers investigating GPCR–protein interactions due to the distinct advantages that they offer [38, 39]. In particular, BRET techniques provide investigators with several further beneﬁts that will be described in detail throughout this chapter. BRET technology has evolved through the generation of improved luminophore, ﬂuorophore and substrate constituents since its initial use to investigate the interaction between circadian clock proteins in 1999 [40]. This has resulted in a series of sub-methods being designed for certain applications. Extended temporal monitoring of GPCR–protein interactions [13, 17] and the use of mutant luciferases to increase detection sensitivity [41] are examples of two recent progressions in BRET methodology that enable researchers to monitor long-term or elusive GPCR–protein interactions. In addition, the ability of BRET technology to scale from small single-plate sample sizes through to multi-plate high-throughput analysis of GPCR–protein interactions exempliﬁes the highly versatile nature of this technology [42, 43]. In this way, BRET technology can be implemented by small laboratories and commercial entities alike to aid in the discovery and development of therapeutic GPCR-targeted substances [43]. The application of BRET technology will provide many researchers with the tools required to investigate the formation of GPCR–protein complexes within and between any of the three major families of GPCRs.

6.1.2 Advantages and constraints of biochemical and resonance energy transfer techniques Conventional biochemical methods aimed at detecting protein–protein interactions such as glutathione-S-transferase pull down, co-immunoprecipitation and yeast-twohybrid assays have provided much inferred information on the formation of certain GPCR–protein complexes in vitro [44]; however, the physiological relevance of this information is often questionable due to the high degree of preparation and processing [45]. The isolation of cell extracts and puriﬁcation of membrane proteins involves chemical processes using various detergents [45]. These processes could potentially

114

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

denature GPCR–protein complexes due to the highly hydrophobic nature of GPCRs, resulting in either artefactual aggregation or separation of physiological complexes. Nevertheless, the use of a stringent control base, such as the use of a range of different processing detergents, can further validate biochemical means of detecting protein interactions. Under optimal conditions, these biochemical techniques do have the ability to infer an interaction directly between a GPCR and another protein. RET techniques, on the other hand, can only infer this interaction indirectly based on their respective changes in proximity throughout an assay. However, the condition that BRET is reliant upon (namely the distance between the two molecules) is highly favourable for suggesting a legitimate interaction, since this distance is almost synonymous with an intermolecular interaction between two proteins when suitably labelled with luminescent or ﬂuorescent BRET species. Nonetheless, controversy exists over the interpretation of BRET data relating to constitutive protein interactions [46]. These interpretations involve altering the relative or absolute protein concentration in the cell to provide further validation of a bona ﬁde GPCR–protein interaction and will be discussed in further detail in Section 6.2.2.7. The use of BRET in a study may not eliminate the need for biochemical assays as a primary means of determining a GPCR–protein interaction, since these techniques can be used in association to fortify the veriﬁcation of the interaction. Biochemical studies can provide a means of identifying and providing evidence as to the presence and size of a certain GPCR–protein complex, depending on the speciﬁcity of the antibody used. Biophysical analysis by means of BRET can then determine a greater near-physiological speciﬁcity of the interaction by secondary veriﬁcation using a range of kinetic models and suitable controls [47]. This provides a more conclusive method of characterizing the nature of a GPCR–protein interaction, and can be extended to monitor the temporal dynamics of the interaction. Deﬁning the relative, temporal nature of GPCR–protein interactions using biochemical techniques is possible, but measurements made at multiple time points require signiﬁcantly more serial processing time and resources. BRET techniques overcome these problems by eliminating the level of post-assay cellular processing, thereby increasing the physiological relevance of the assay data. In addition, BRET offers the ability to obtain data of GPCR–protein complex formation dynamics as they occur physiologically at 37 ◦ C in live cells and in real time.

6.2 Methods and approaches 6.2.1 BRET principles Bioluminescence has evolved as a primordial form of communication to signal a behavioural response from one organism to another [48]. Cnidarian sea creatures, such as the jellyﬁsh Aequorea victoria and the sea pansy Renilla reniformis, commonly exhibit this biophysical process that is the basis for the contemporary technique of BRET currently used by many laboratories to study the cellular biology of protein–protein interactions. BRET allows both qualitative and quantitative monitoring to be carried out on putative GPCR–protein complexes. The process occurs via oxidation of a coelenterazine substrate by a donor enzyme (e.g. Renilla luciferase, Rluc). If sufﬁciently close to an acceptor ﬂuorophore (e.g. enhanced green ﬂuorescent protein,

6.2 METHODS AND APPROACHES

115

EGFP), energy emitted from this reaction is transmitted in a nonradiative dipole–dipole manner at wavelengths corresponding to the energy excitation spectrum of the ﬂuorophore. The ﬂuorophore will consequently emit light energy at longer wavelengths, according to its distinct emission spectrum [47]. The light energy emitted by both of these processes can be monitored and compared using a dual-ﬁlter luminometer to determine if energy transfer has occurred between the donor and acceptor. In this way, an increased level of energy emitted from the acceptor ﬂuorophore compared with that emitted at the same time point from the oxidation of coelenterazine is indicative of the proximity favouring RET between the species. The physical distance over which RET occurs (<10 nm) is signiﬁcant, since distances within this range are indicative of physiological protein–protein interactions [49]. Figure 6.1 illustrates how BRET occurs and how it can be used to assess the interaction between a GPCR and an interacting protein of interest. (a)

(b)

Figure 6.1 Measurement of BRET that occurs from the interaction of a GPCR with a putative intracellular protein tagged with a donor or an acceptor. A GPCR is C-terminally tagged with a donor enzyme, Rluc, and the intracellular protein is tagged with an acceptor ﬂuorophore, yellow ﬂuorescent protein (YFP). Upon addition of a coelenterazine-based substrate and in the presence of molecular oxygen, the substrate undergoes oxidation through enzymatic activity with Rluc, emitting light and CO2 . If the intracellular protein does not come into close proximity with the GPCR, then conditions prevent RET occurring (a). In this example, addition of ligand promotes interaction between the GPCR and intracellular protein. Provided that the relative tag distance and orientation are favourable for RET to occur, light energy from both the donor and acceptor are measured at their characteristic wavelengths (b).

116

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

The principles involved for detecting protein–protein interactions using BRET were initially described by Ward and Cormier [50]. However, the ﬁrst application and protein–protein interaction discovery using BRET was only relatively recently revealed [40] and laid out the foundation for studies involving interactions between protein species in bacterial [51, 52], plant [53, 54] and mammalian cells [39]. The dipole–dipole-induced RET rate and efﬁciency between a donor and an acceptor have been set out according to the theory proposed in the seminal paper on the relationship between ﬂuorescence and intermediate molecular energy states [55]. The spectroscopic properties of the donor and acceptor, the orientation of the two dipoles, the quantum yield and the lifetime of the donor are critical determining factors of RET. The F¨orster distance, or the critical distance at which energy transfer between the donor and acceptor is equal to the decay rate of the donor, is speciﬁc to the donor–acceptor pair used and is an important determinant of RET efﬁciency [49]. The overlap integral, or spectral overlap between the donor emission and acceptor absorption spectra, is important in resolving the signal-to-noise ratio and is again dependent on the spectral properties of the RET pair. Critically, the efﬁciency of RET is dependent on the distance to the sixth power between donor and acceptor [49]. Investigation of the conformation of β-arrestin by monitoring the intramolecular distance between the carboxyl and amino termini during inactivated and activated states is an example of the exquisite distance dependency of BRET [56].

6.2.2 Measuring GPCR–protein interactions using BRET 6.2.2.1 Assessing the suitability of BRET versus FRET to detect GPCR–protein interactions BRET is a sensitive technique that enables GPCR–protein interactions to be detected that are either constitutively formed or which form upon the addition of a substance such as an agonist. FRET involves the same set of physical principles as BRET; however, many differences are responsible for their application towards a particular use. FRET does not require the genetic fusion of ﬂuorescent proteins with the protein of interest, allowing the interaction between endogenous proteins to be detected, provided a suitable antibody is available or epitope is present [57]. Although one of the proteins in a BRET assay could be conceivably immunolabelled with a ﬂuorescent probe, the other ‘partner’ protein would require genetic fusion to the luciferase enzyme. Autoﬂuorescence, or the ﬂuorescence of endogenous cellular substances upon stimulation with an external light source, leads to higher background noise and lower signal resolution. This is a problematic issue when utilizing FRET. Derivatives of the FRET technique, such as time-resolved FRET (TR-FRET) and FRET ﬂuorescence lifetime imaging microscopy (FRET-FLIM), attempt to eliminate background autoﬂuorescence and increase the signal-to-noise ratio by allowing background ﬂuorescence to decay. Donor and acceptor species with greater ﬂuorescent half-lives are used to measure the desired protein interaction in the case of TR-FRET [58]. Alternatively, FRET-FLIM can monitor energy transfer through changes in ﬂuorescent lifetime; however, complex interpretation of this data may be involved. BRET is superior to FRET in that no external light source is required, resulting in the elimination of

6.2 METHODS AND APPROACHES

117

fluorophore photobleaching. This was found to be particularly beneﬁcial for detecting the interaction between proteins that are light sensitive, such as circadian clock proteins [40]. Indeed, the necessity of this requirement seems to have given rise to the foundation of the BRET technique. In addition, the ability to monitor dynamic interactions between GPCRs and putative interacting proteins over extended periods of time (many hours) with no interference of the cell system, and little decrease in the signal-to-noise ratio, has been particularly beneﬁcial in the characterization of internalization and trafﬁcking pathways of GPCRs [13, 17]. Measurement of BRET assays in a more automated fashion by the use of electronically controlled injectors to add reagents such as substrates and ligands accurately can be implemented in addition to maintaining and controlling assay temperature throughout the measurement period. Control over both of these variables may lead to lower intra- and inter-assay variability and less attributed error. This is especially useful for high-throughput BRET assays. As the substrate used to initiate BRET diffuses throughout the cellular environment, compartmentalization of GPCR–protein interactions cannot be determined at this stage. However, taking advantage of recent advances in cell imaging through the use of high-resolution supercooled charge-coupled device cameras, mutated Rluc enzymes to adjust the wavelength and quantum yield of light emission [59, 60] in conjunction with a range of modiﬁed ﬂuorescent acceptor proteins [61], it has become possible to visually monitor GPCR–protein interactions in single cells. However, FRET remains the dominant RET method for assessing the cellular compartmentation of GPCR–protein interactions [39]. In vivo animal imaging typically employs the FRET technique due to the near-infrared wavelengths that can be applied. Tissue attenuation of short-wavelength (<600 nm) emissions using the substrate Rluc obstructs the application of BRET for these assays [60]. However, a recent study has presented novel Rluc mutations that result in bathochromic shifts in the wavelength of light emitted upon enzymatic oxidation of coelenterazine, resulting in an increase in the amount of tissue-penetrable light available for measurement [60]. This presents a possible alternative in monitoring protein–protein interactions by in vivo animal imaging.

6.2.2.2 Enabling proteins to be tagged for BRET detection A key difference between FRET and BRET is the requirement for the cDNA sequence of the GPCR and the putative interacting protein under investigation to be genetically fused with cDNA encoding either the luminophore enzyme or a ﬂuorophore when using BRET. Immunolabelling of ﬂuorescent proteins for BRET detection is a possibility; however, the availability and speciﬁcity of antibodies to the proteins of interest often excludes these procedures. The carboxyl (C) terminus of GPCRs is typically tagged for investigations of intracellular protein interactions. However, provided that the tag does not interfere with the normal structural folding or physiological function of the protein, the tag may be integrated into any suitable region of the protein. This may be particularly beneﬁcial if both the N and C termini of the receptor or protein are known to be structurally or functionally involved in a known process. The types of BRET tag to be conjugated onto each protein will depend on the particular BRET method adopted, which will be discussed in more detail in Section 6.2.2.9 (see Table 6.1).

118

Many hours

Up to 1 h

DeepBlueC Seconds to (dideoxycoelminutes enterazine h)

EnduRen

Up to 2 h

Coelenterazine h Up to 1 h

Rluc2/Rluc8

Rluc/hRluc

Rluc/hRluc/ Rluc2/Rluc8

Rluc2/Rluc8

Rluc/hRluc

∼400

∼480

∼480

370–450

440–500

400–475

440–500

400–475

GFP2 (class 1)/ GFP10 (class 3)

YFP/Venus (class 4)

EGFP (class 2)

YFP/Venus (class 4)

EGFP (class 2)

Suitable Donor enzymec Donor Suggested donor Acceptor measurement emission measurement ﬂuorophore time frameb peak (nm)d ﬁlter range (nm)e (ﬂuorescent classf )

510–590 500–530

∼510

500–550

510–590

500–550

∼530

∼510

∼530

∼510

c Within each sub-method, any suggested donor enzyme can be combined with any suggested acceptor ﬂuorophore. d The donor emission peak values are relatively conserved between wild-type Rluc, codon-humanized Rluc (hRluc) and the two mutants Rluc2 and Rluc8 [59]. e The donor measurement ﬁlter range varies due to the spectral properties of the ﬂuorophore used. Minimal bleed-through of measured signal is required for optimal BRET signals. f Fluorophores are classiﬁed based on distinct chemical groups that govern the excitation and emission properties of these molecules [62].

∼110

∼50

∼30

∼50

∼30

Acceptor Suggested acceptor Donor/Acceptor emission measurement emission peak peak (nm) ﬁlter range separation (nm) (nm)

a Extended BRET involves the use of EnduRen, a protected form of coelenterazine h that exhibits the same spectral properties as the unprotected, activated form [17]. b The suitable measurement time frame for each sub-method is dependent upon the donor enzyme used.

BRET2

eBRETa

BRET1

BRET Substrate sub-method

Table 6.1 Characteristics of the BRET sub-methods.

6.2 METHODS AND APPROACHES

119

Figure 6.2 A typical scenario for the construction of BRET fusion cDNA constructs. Suitable expression vectors containing the cDNA sequence of the protein of interest are genetically fused in-frame, at either the N- or C-terminal, with either the donor or acceptor tag. The stop codon between the cDNA sequences is also removed to ensure generation of a fusion protein. The plasmid is transformed into competent bacterial cells (e.g. TOP10; Invitrogen) and cultured. The DNA is then extracted, puriﬁed and is ready to be transfected for a BRET assay.

Molecular biological techniques are required to integrate the BRET tags in-frame with the cDNA sequence encoding the GPCR or putative interacting protein under investigation (see Figure 6.2). Suitable cloning vectors are used for this process in concert with oligonucleotide primers, restriction endonucleases, ligases and the ability of bacterial cells to amplify the vectors coding for each fusion protein. Accordingly, the cDNA sequence of both proteins under investigation must be known. The size of the putative interacting proteins relative to the size of the luminophore or ﬂuorophore BRET tags should be taken into consideration. The luminophore Rluc is 36 kDa, and a ﬂuorescent protein such as EGFP is 27 kDa. Attachment of these constructs to proteins of similar or much smaller size may affect normal physiological function. Recently, it was revealed that fusing green ﬂuorescent protein (GFP) on the C-terminus of MC2R resulted in a loss of the ability to be inserted properly into the plasma membrane for normal function [5]. On the other hand, if the proteins are much larger than the BRET tag, then interactions may not be detected, or rendered weak, due to shielding or extended distances, resulting in the impairment of RET. The addition of linker regions between the BRET tag and the protein can be utilized in an attempt to counteract proximity problems by increasing freedom of movement and separating the BRET tag from functional domains in the protein of interest. A range of possible

120

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

assays for GPCRs can be used to determine whether a loss or gain of function is attributable to the fusion of a BRET tag. Mechanisms for investigating these aspects will be elaborated further in Section 6.2.2.3.

6.2.2.3 Validation of BRET fusion proteins Determining changes in the functional attributes of the protein of interest due to the addition of a BRET tag is critical for generating meaningful conclusions. Any data generated from a BRET experiment must be interpreted in light of possible differences in the activity of the protein by comparing function with the untagged or native form. To test the function of BRET-tagged GPCRs, a common method is to measure the potency/efﬁcacy of agonists to generate second messengers. Dose–response curves generated from the activation of BRET-tagged GPCRs can be compared against the wild-type GPCR to achieve this. Assays are available for the measurement of commonly generated second messengers, such as inositol phosphates and cyclic adenosine monophosphate. Since GPCRs should be principally present at the plasma membrane, confocal microscopy is often an excellent mechanism to conﬁrm visually that the receptor is capable of being adequately trafﬁcked to the cell surface. Fluorophore-tagged GPCRs (i.e. acceptors) may be visualized directly; however, luminophore-tagged GPCRs (i.e. donors) may require attachment of immunochemical tags such as myc, FLAG or haemagglutinin and labelling with suitable antibodies for detection. Confocal microscopy may also be used to indicate whether the activity of the putative interacting protein has been compromised, particularly if the protein should be localized to a particular cellular compartment. Fusion of immunochemical tags may also be of use in assays, such as enzyme-linked immunosorbent assays (ELISAs). Cell-based ELISAs [63] can be particularly beneﬁcial in determining membrane expression of proteins. Any other testing would be at the investigator’s discretion, depending on the function of the particular protein.

6.2.2.4 BRET expression systems and transfection of fusion proteins After the BRET fusion cDNA constructs have been assembled, a suitable expression system is required for both proteins to be synthesized. It should be noted that certain cell lines might contain endogenous levels of the proteins under investigation, which may affect the BRET signal by providing competition for the tagged proteins. Levels of endogenous protein can typically be assessed by performing western blots, and any such ﬁndings should be considered when interpreting the results. Heterologous cell lines such as HEK 293 or COS-7 cells are commonly used; however, a range of endogenous GPCRs have been discovered in these cells [64] and may need to be accounted for in the interpretation of BRET data. Expression of fusion proteins in various cell types requires the use of a suitable transfection method. For mammalian cells, an example would be the use of cationic lipids to aid in the packaging and transportation of DNA across the bi-lipid plasma membrane into the cytosol. Importantly, transfection reagents should be tested for efﬁciency and cytotoxicity with the speciﬁc mammalian cell line to be used. Basic selection of successfully transfected cells may be achieved through the use of antibiotics

6.2 METHODS AND APPROACHES

121

targeted towards selectable markers present in the cDNA construct; however, in the case of double transfections required for BRET assays, the use of two antibiotics targeted towards each BRET fusion construct would be preferable. Mammalian cells that are transiently transfected with fusion protein-coding cDNA can result in a broad range of protein expression levels. Transient cDNA transfections are suitable for most BRET experiments; however, if the protein of interest does not express well, then the generation of stably transfected cells may be a preferred alternative. Stable cell lines are produced through continual selection of cells that have been successfully transfected with one or both BRET-tagged cDNA constructs. This method should decrease the range of variability in protein expression over that observed for a transiently expressing cell population. The generation of monoclonal rather than polyclonal stable cell lines is preferred, as this will further reduce the variability of protein expression. It should be noted that some cDNA constructs will express at a higher level than others and that, due to the level of overexpression of the constructs, high expression of one construct might equate to lower expression of the other construct due to the constraints placed on the cellular protein machinery. This should be taken into account when deciding appropriate amounts and ratios of cDNA to transfect. A drawback of the BRET detection system is the enhanced expression, or overexpression, of proteins in quantities that may be greater than that observed physiologically. This may result in the attainment of artiﬁcial BRET signals, simply due to the close proximity of donor and acceptor fusion proteins. An alternative approach may be carried out by homologous recombination of BRET fusion protein sequences targeted towards speciﬁcally designed recipient cell lines, to generate isogenic stable cell lines [65]. This technique has already been used in a study to investigate the effects of a particular GPCR–protein interaction [5], but it has yet to be utilized for a BRET assay. Its use may result in better control and more physiologically relevant levels of expressed protein. Methods are available to measure these expression levels in a cell population for BRET assays and will be discussed in the Section 6.2.2.5.

6.2.2.5 Methods for measuring protein expression The relative measurement of donor and acceptor protein concentrations in cell cultures for BRET assays is crucial for the analysis and interpretation of results. The ability to replicate BRET assays with low inter-assay variability is statistically beneﬁcial for the generation of meaningful conclusions. If the GPCR or protein under investigation has been fused with a suitable antigenic determinant that is accessible to an introduced antibody, cell-based ELISA techniques can be used to compare the GPCR cell surface expression between two populations and also intracellular levels if a suitable permeabilization agent is used. This will allow protein expression to be compared by ratiometric means. Comparative levels of acceptor ﬂuorescence may also be determined by measuring the total ﬂuorescence emitted by a predetermined number of cells using a band-pass ﬁlter appropriate for the ﬂuorescent probe used. Alternatively, absolute levels of either donor or acceptor protein may be determined using either membrane preparations or puriﬁed protein extracts of the transfected cells. The equation, methodology and variables required for this analysis have been previously

122

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

determined [66], though not for use in a BRET assay. The ﬂuorescent or luminescent intensity of transfected cells can be measured using a ﬂuorometer or luminometer with a suitable band-pass ﬁlter set or using a scanning spectrometer (see Section 6.2.2.9; Table 6.1). Alternatively, ﬂuorescence-activated cell sorting (FACS) is capable of determining both the percentage of expressing cells and level of acceptor concentration per cell [47]. By extension, this process can be used to purify cells that contain a certain desired level of protein expression. Unless a suitable ﬂuorescently conjugated antibody can be targeted towards the donor protein, measurement of donor protein expression cannot be determined using this technique. Conjugation of immunochemical tags to the donor protein may be required to measure concentration of Rluc-tagged proteins and aid in determining donor protein expression.

6.2.2.6 Determining optimal expression of BRET-tagged proteins Establishing the optimal ratio of donor and acceptor BRET-tagged proteins is critically important for obtaining appreciable BRET signals, as well as in generating overall conclusions. Two variables exist: the overall level of BRET protein expression and the relative expression levels of donor and acceptor BRET-tagged proteins. As mentioned earlier, the overall level of protein overexpression should be kept to a minimum to reduce the level of nonspeciﬁc interactions that may occur. Adjustment of the ratio between acceptor and donor expression will not only lead to more resolute BRET signals, but can also be used to indicate speciﬁcity, elaborated on in Section 6.2.2.7. Generally, the ratio of acceptor to donor protein expression should be higher than unity and typically ranges from 2 : 1 to 4 : 1. However, these values depend on a range of characteristics speciﬁc to the GPCR–protein pair under investigation, including the relative afﬁnity between the GPCR–protein pair and the known or likely stoichiometry of the interacting pair [47]. For example, the stoichiometry of ubiquitinated GPCRs is dependent on either the formation of monoubiquitinated or polyubiquitinated complexes. The formation of polyubiquitinated protein complexes has recently been optimized for detection using BRET [31]. Importantly, the relative expression levels of acceptor and donor BRET constructs are highly dependent upon the efﬁciency of transfection (the amount of protein expressed for a given amount of transfected acceptor or donor construct cDNA). Assessment of protein expression can be determined using the techniques described in Section 6.2.2.5; however, titration assays are often useful to determine the optimal acceptor–donor BRET construct ratio.

6.2.2.7 Methods for assessing BRET speciﬁcity Alteration of the total concentration and ratio between the donor and acceptor BRET-tagged proteins is often used to validate the speciﬁcity of a GPCR–protein interaction. Two different methods have been formulated to assess the speciﬁcity of a putative GPCR–protein interaction [67]. Saturation or titration assays can be used to measure constitutive interactions and dynamic interactions at a single time-point per individual assay. This technique involves the expression of a constant amount of donor protein, with an increasing level of acceptor protein. A speciﬁc

6.2 METHODS AND APPROACHES

123

GPCR–protein interaction should result in the generation of a hyperbolic curve with an asymptote tending to the maximal BRET value. BRETmax is the BRET ratio value that corresponds to the horizontal asymptotic value at high acceptor-to-donor ratios. This value indicates the tendency towards a saturation point, where all available donor proteins are interacting with acceptor proteins, such that addition of more acceptor protein will not result in a greater number of interactions. BRET50 values are typically used to compare the speciﬁcity of a GPCR–protein interaction by indicating the ratio of acceptor- over donor-conjugated proteins required for a half-maximal BRET value to be attained. In contrast, nonspeciﬁc interactions result in a near-linear relationship between BRET ratio and increasing acceptor-to-donor ratios [67]. An alternative procedure involves the displacement or competition of acceptorconjugated proteins by employing an increasing amount of either the untagged ‘acceptor’ protein or another noninteracting protein while keeping the concentration and ratio of both the donor and acceptor BRET-tagged proteins the same. If the interaction is speciﬁc, then successively increasing the amount of untagged ‘acceptor’ protein will result in a sigmoidal decrease in BRET ratio. If an increasing concentration of a known noninteracting protein is included in successive assays, then the BRET ratio should remain the same for speciﬁc interactions between the donor and acceptor BRET-tagged proteins [67]. Although these methods have been used to provide evidence for a number of GPCR–protein interactions, controversy exists over their use [46]. The overexpression of donor and acceptor proteins is a primary concern for interpreting the physiological relevance of such interactions, and the use of stringent controls is critical.

6.2.2.8 BRET controls The use of suitable controls for BRET assays is vital to generate signiﬁcant conclusions. Typically, cells expressing the donor protein only are assayed alongside donorand acceptor-expressing cells. Upon the addition of substrate, the donor-only control will indicate the background BRET signal. Alternatively, if a substance such as a ligand alters the GPCR–protein interaction, matched donor- and acceptor-expressing cell populations can be used. These are measured and treated in parallel with either the ligand or the vehicle in which the ligand was diluted [68]. Positive and negative experimental controls should be used in all BRET assays to ascertain the relevance of data obtained. Positive controls may take the form of a known GPCR–protein interaction, particularly an interaction containing a similar GPCR to the one being tested. For example, the interaction between the angiotensin type 1a receptor and β-arrestin 2 produces strong and stable BRET ratios over extended periods of time, typical of a class B receptor β-arrestin binding proﬁle [17]. Preferably, positive control BRET pairs should also be expressed in the same respective cellular compartment as the proteins being studied. A suitable negative control should involve assaying a protein of similar size, shape and cellular localization to the interacting protein under investigation. For example, the known inability of gonadotrophin-releasing hormone receptor to bind to β-arrestin due to the lack of an intracellular C-terminal tail [69] makes this pair a suitable negative control candidate when investigating GPCR–β-arrestin interactions. In addition, the innate inability of angiotensin 2 receptor to bind to β-arrestin 2

124

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

upon activation [70] is a similar example of a suitable negative control for these investigations. Alternative negative control candidates may be developed by utilizing mutant forms that exhibit a loss-of-function mutation and result in the inability to bind with the complementary protein under investigation. These are potentially more powerful as negative controls, since a greater number of variables, such as size, shape and cellular localization, are held constant.

6.2.2.9 Choice of BRET method and procedure Currently, three BRET sub-methods have been developed, categorized principally by the coelenterazine-based substrate used. A summary of these methods, including the substrate donor tags and acceptor tags used, is presented in Table 6.1. Protocol 6.1 may be used for any of these. Importantly, the choice of BRET sub-method will directly inﬂuence the construction of the donor and acceptor BRET-conjugated proteins. Multiple ﬂuorophores [62], as well as multiple mutated forms of Rluc [59, 60], are available to optimize the signal acquired between the GPCR and putative interacting protein of interest. Two variables can be altered in a BRET assay, depending on the dynamic properties of the GPCR–protein interaction pair under investigation. For constitutive interactions, measuring BRET at a single time point may be beneﬁcial, for example, in association with the methods employed in Section 6.2.2.7. In this case, a range of concentrations of donor and acceptor protein may be used, but the assay time must be kept constant. On the other hand, long-term GPCR–protein dynamics can be monitored over a range of time points, but sequential measurements must be carried out under the same protein concentration conditions. Typically, once a suitable concentration and ratio of donor and acceptor proteins is attained, time-course studies may subsequently be performed.

PROTOCOL 6.1 BRET Using a Mammalian Cell System This protocol can be used in conjunction with all three BRET sub-methods, the use of cells in suspension or adhered to the bottom of a tissue culture plate. Finally, single and multiple time-point BRET measurements can also be made.

Equipment and Reagents • Cell line suitable for the desired study (e.g. HEK 293, COS-7) • Growth medium for cells (e.g. Dulbecco’s modiﬁed Eagle’s medium (DMEM) containing 0.3 mg ml−1 glutamine (Gibco), 100 IU/ml penicillin, 100 µg ml−1 streptomycin (Gibco) and 5–10% foetal calf serum (FCS; Gibco) for HEK 293 or COS-7 cells (termed ‘complete medium’)) • Phenol-red-free complete medium (as above without phenol red) for HEK 293 or COS-7 cells • eBRET only – 25 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES) buffer (Gibco) is required

6.2 METHODS AND APPROACHES

• Dulbecco’s phosphate-buffered saline (D-PBS) containing 0.1 g l−1 CaCl2 , 0.1 g l−1 MgCl2 ·6H2 O and 1 g l−1 D-glucose (Gibco). • Six-well tissue culture plates (BD Falcon) • 96-well white tissue culture plates (Nunc) • Poly-L-lysine (≥70 000 MW) (used for coating plates – may not be required separately) • Transfection system or reagent (e.g. GeneJuice (Novagen)) • 0.05% trypsin, 0.53 mM ethylenediamine tetraacetic acid (EDTA) (Gibco) • Coelenterazine-based substrate (e.g. coelenterazine h (Molecular Probes), DeepBlueC (PerkinElmer), EnduRen (Promega)) • Microplate luminometer capable of either sequential or simultaneous light measurement through ﬁlters that detect donor and acceptor emission spectra (e.g. VICTORlight (PerkinElmer); Mithras LB 940 (Berthold Technologies))a • Cell culture hood and incubator • Optional: FACS equipment • Optional: scanning spectrometer to visualize spectral output (e.g. Cary Eclipse (Varian); Flexstation II (Molecular Devices)).

Method 1 If required, coat six-well clear tissue culture plates and 96-well white tissue culture plates with poly-L-lysine according to the manufacturer’s instructions. 2 Culture cells according to the necessary criteria as follows: (a) Transient transfection (adherent/suspension): Culture cells using complete medium in a six-well plate at a density depending on the growth rate of the cell line such that general conﬂuency after 18–24 h will be suitable for transfection. Ensure an even distribution of cells in each well. Proceed to step 3 (b) Stable transfection (adherent): Plate 60 000–80 000 cells into each well of a white 96-well plate using phenol-red-free complete medium with 25 mM HEPES buffer if media is not to be changed before an assay. Optional: aliquots of cells can be prepared to determine the relative and/or absolute concentrations of BRET-tagged proteins by FACS (see Section 6.2.2.5). Proceed to step 3. (c) Stable transfection (suspension): Plate 80 000–100 000 cells into each well of a white 96-well plate. The type of diluent is dependent upon the BRET method employed as follows: (i) For cells to be analysed using BRET1 or BRET2 methods, use supplemented D-PBS. (ii) For cells to be analysed using the eBRET method, use phenol-red-free complete medium with 25 mM HEPES. Optional: aliquots of cells can be prepared to determine the relative and/or absolute concentrations of BRET-tagged proteins by FACS (see Section 6.2.2.5). Proceed immediately to step 7.

125

126

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

3 Incubate cells at 37 ◦ C/5% CO2 overnight. For transiently transfected cells, proceed to step 4; for stably transfected cells proceed to step 7. 4 Transfect the cells with donor and acceptor constructs using a transfection reagent suitable for the cell line, according to the manufacturer’s instructions (e.g. Genejuice (Novagen); Metafectene (Biontex)). Suitable controls such as donor-only transfected cells should be included, if required. 5 Incubate cells at 37 ◦ C/5% CO2 . 6 Split cells depending on the chosen assay format: (a) Adherent cell assay: Approximately 18–24 h after transient transfection, split 60 000–80 000 cells from the six-well plate into each well of a white 96-well plate using phenol-red-free complete medium with 25 mM HEPES buffer if media is not to be changed before an assay. Incubate cells at 37 ◦ C/5% CO2 overnight. (b) Suspended cell assay: Approximately 36–48 h after transient transfection, or when the desired level of cell conﬂuency has been reached, split approximately 80 000–100 000 cells from the six-well plate into each well of a 96-well plate. The type of diluent is dependent upon the BRET method employed as follows: (i) For cells to be analysed using BRET1 or BRET2 methods, use supplemented D-PBS. (ii) For cells to be analysed using the eBRET method, use phenol-red-free complete medium with 25 mM HEPES. Optional: include additional aliquots of cells if subsequent FACS analysis is to be carried out (see Section 6.2.2.5). 7 For assays of cells in suspension, substrate must be added directly to the well. For adherent cells, substrate can be either directly added to each well, or further diluted in 5% FCS phenol-red-free complete medium with 25 mM HEPES/supplemented D-PBS and added subsequent to aspirating the previous media. Take care not to disturb the cell monolayer when aspirating media. Add the substrate to each of the wells according to the following guidelines: (a) BRET1 – add coelenterazine h substrate (5 µM optimal ﬁnal concentration) and measure immediately; (b) BRET2 – add DeepBlueC b substrate and measure immediately (5 µM optimal ﬁnal concentration); (c) eBRET – add EnduRen substratec (60 µM ﬁnal optimal concentration) and allow to incubate for at least 1.5 h at 37 ◦ C/5% CO2 protected from light before assaying. 8 Place the 96-well plate into an appropriate dual-ﬁlter luminometer and take measurements simultaneously or sequentially for between 0.5–5 s per ﬁlter before proceeding to the next well.d Alternatively, use a scanning spectrometer to generate spectral data. 9 If a non-ligand-driven interaction is being measured (i.e. a constitutive interaction), then no intervention is required during the assay. If the interaction is ligand modulated, then allow the luminescence and ﬂuorescence values to be read over a short

6.2 METHODS AND APPROACHES

127

period before adding the ligand to establish a baseline BRET level according to the BRET method used as follows: (a) ∼5–15 min for BRET1 ; (b) ∼1–2 min for BRET2 using Rluc/hRluc luminophores, or 5–10 min using BRET2 with Rluc2/Rluc8 luminophores; (c) ∼10–30 min for eBRET. 10 If a substance (such as ligand) is to be added to the cells, add this after the baseline BRET signals have been read. 11 Allow the luminescence and ﬂuorescence values to be read over the required time period. If using a scanning spectrometer, obtaining a spectrum before and after substance addition may be appropriate. 12 Store the acquired luminescence and ﬂuorescence values for analysis.

Analysis of Results Depending on the type of BRET controls used in the assay, two different methods have been formulated to determine the BRET ratio. First, for assays with donor-only controls, the following equation is used: BRETratio =

long-wavelength emission (donor & acceptor pair) short-wavelength emission (donor & acceptor pair) −

long-wavelength emission (donor-only control) short-wavelength emission (donor-only control)

For assays involving an added substance (i.e. ligand) along with vehicle controls (i.e. phosphate-buffered saline), the following equation should be used: BRETratio =

long-wavelength emission (ligand treated) short-wavelength emission (ligand treated) −

long-wavelength emission (vehicle treated) short-wavelength emission (vehicle treated)

Statistical analysis of BRET assays may involve either using analysis of variance with suitable post tests or Student’s t-tests.

Notes a Injectors may also be required, particularly for the BRET2 method, due to the short half-life of the substrate DeepBlueC [71]. b DeepBlueC

has a very short luminescent half-life (seconds) with Rluc; therefore, immediate measurement is required. The use of injectors is preferred. c Stock

EnduRen substrate must be warmed to 37 ◦ C to dissolve before use.

d Note that longer measurement times through each ﬁlter to generate a single data acquisition point may result in higher acquired values, but decreased relevance due to changing conditions over time.

128

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

6.3 Troubleshooting 6.3.1 Low or no protein expression • BRET cDNA construct is not expressed in a vector suitable for expression in the cell line used. • Check sequencing of the construct for unwanted mutations and/or ensure the cDNA sequences are in-frame. • Optimize use of transfection reagent/system. • Optimize amount and ratio of transfected BRET fusion cDNA.

6.3.2 Low or no BRET signal • Improper trafﬁcking of the protein to the correct cellular compartment (e.g. lack of GPCR expressed at the cell surface). • Low expression levels of one or both BRET fusion proteins. • Suboptimal ﬁlter set combination. • Presence of a reducing agent, resulting in signal quenching. • Instrument not properly calibrated.

6.3.3 Variable BRET signal • Suboptimal positioning of the BRET fusion tag (see Section 6.2.2.2). • eBRET – EnduRen substrate not sufﬁciently dissolved (must be dissolved at 37 ◦ C). • Unfavourable environmental conditions (i.e. temperature and/or CO2 concentration).

6.3.4 High background BRET level • Suboptimal ﬁlter set combination. • Excessive expression of BRET constructs resulting in bystander BRET.

References 1. H´eroux, M. Breton, B. Hogue, M. and Bouvier, M. (2007) Assembly and signaling of CRLR and RAMP1 complexes assessed by BRET. Biochemistry, 46, 7022–7033. 2. McLatchie, L.M. Fraser, N.J. Main, M.J. et al. (1998) RAMPs regulate the transport and ligand speciﬁcity of the calcitonin-receptor-like receptor. Nature, 393, 333–339.

REFERENCES

129

3. Udawela, M. Hay, D.L. and Sexton, P.M. (2004) The receptor activity modifying protein family of G protein coupled receptor accessory proteins. Semin. Cell Dev. Biol., 15, 299–308. 4. Metherell, L. Chapple, J. Cooray, S. et al. (2005) Mutations in MRAP, encoding a new interacting partner of the ACTH receptor, cause familial glucocorticoid deﬁciency type 2. Nat. Genet., 37, 166–170. 5. Roy, S. Rached, M. and Gallo-Payet, N. (2007) Differential regulation of the human adrenocorticotropin receptor [melanocortin-2 receptor (MC2R)] by human MC2R accessory protein isoforms α and β in isogenic human embryonic kidney 293 cells. J. Mol. Endocrinol., 21, 1656–1669. Use of isogenic stable cell lines in a protein interaction study. 6. Modan-Moses, D. Ben-Zeev, B. Hoffmann, C. et al. (2006) Unusual presentation of familial glucocorticoid deﬁciency with a novel MRAP mutation. J. Clin. Endocrinol. Metab., 91, 3713–3717. 7. Violin, J. and Lefkowitz, R. (2007) β-Arrestin-biased ligands at seven-transmembrane receptors. Trends Pharmacol. Sci., 28, 416–422. 8. Reiter, E. and Lefkowitz, R. (2006) GRKs and β-arrestins: roles in receptor silencing, trafﬁcking and signaling. Trends Endocrinol. Metab., 17, 159–165. 9. DeWire, S. Ahn, S. Lefkowitz, R. and Shenoy, S. (2007) β-Arrestins and cell signaling. Annu. Rev. Physiol., 69, 483–510. 10. Gurevich, E.V. and Gurevich, V.V. (2006) Arrestins: ubiquitous regulators of cellular signaling pathways. Genome Biol., 7, 236.1–236.10. 11. Buchanan, F.G. and DuBois, R.N. (2006) Emerging roles of β-arrestins. Cell Cycle, 5, 2060–2063. 12. Moore, C. Milano, S. and Benovic, J. (2006) Regulation of receptor trafﬁcking by GRKs and arrestins. Annu. Rev. Physiol., 69, 451–482. 13. Pﬂeger, K. Dalrymple, M. Dromey, J. and Eidne, K. (2007) Monitoring interactions between G-protein-coupled receptors and β-arrestins. Biochem. Soc. Trans., 35, 764–766. Study of GPCR–β-arrestin interactions using BRET. 14. Oakley, R. Laporte, S. Holt, J. et al. (2000) Differential afﬁnities of visual arrestin, βarrestin1, and βarrestin2 for G protein-coupled receptors delineate two major classes of receptors. J. Biol. Chem., 275, 17201–17210. 15. Hamdan, F. Rochdi, M. Breton, B. et al. (2007) Unraveling G protein-coupled receptor endocytosis pathways using real-time monitoring of agonist-promoted interaction between β-arrestins and AP-2. J. Biol. Chem., 282, 29089–29100. 16. Simaan, M. B´edard-Goulet, S. Fessart, D. et al. (2005) Dissociation of β-arrestin from internalized bradykinin B2 receptor is necessary for receptor recycling and resensitization. Cell Signal., 17, 1074–1083. 17. Pﬂeger, K. Dromey, J. Dalrymple, M. et al. (2006) Extended bioluminescence resonance energy transfer (eBRET) for monitoring prolonged protein-protein interactions in live cells. Cell Signal. 18, 1664–1670. Seminal article on the use of the eBRET method. 18. Gilman, A.G. (1987) G proteins: transducers of receptor-generated signals. Annu. Rev. Biochem., 56, 615–649. 19. Neves, S. Ram, P. and Iyengar, R. (2002) G protein pathways. Science, 296, 1636–1639. 20. Rashid, A. O’Dowd, B. and George, S. (2004) Minireview: Diversity and complexity of signaling through peptidergic G protein-coupled receptors. Endocrinology, 145, 2645–2652.

130

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

21. Wettschureck, N. and Offermanns, S. (2005) Mammalian G proteins and their cell type speciﬁc functions. Physiol. Rev., 85, 1159–1204. 22. Brzostowski, J. and Kimmel, A. (2001) Signaling at zero G: G-protein-independent functions for 7-TM receptors. Trends Biochem. Sci., 26, 291–297. 23. DeFea, K.A. Vaughn, Z.D. O’Bryan, E.M. et al. (2000) The proliferative and antiapoptotic effects of substance P are facilitated by formation of a β-arrestin-dependent scaffolding complex. Proc. Natl. Acad. Sci. U. S. A., 97, 11086–11091. 24. DeFea, K.A. Zalevsky, J. Thoma, M.S. et al. (2000) β-Arrestin-dependent endocytosis of proteinase-activated receptor 2 is required for intracellular targeting of activated ERK1/2. J. Cell. Biol., 148, 1267–1281. 25. Luttrell, L. Ferguson, S. Daaka, Y. et al. (1999) β-Arrestin-dependent formation of β2 adrenergic receptor-Src protein kinase complexes. Science, 283, 655–661. 26. Luttrell, L.M. Roudabush, F.L. Choy, E.W. et al. (2001) Activation and targeting of extracellular signal-regulated kinases by β-arrestin scaffolds. Proc. Natl. Acad. Sci. U. S. A., 98, 2449–2454. 27. Tohgo, A. Choy, E. Gesty-Palmer, D. et al. (2003) The stability of the G protein-coupled receptor–β-arrestin interaction determines the mechanism and functional consequence of ERK activation. J. Biol. Chem., 278, 6258–6267. 28. Caunt, C.J. Finch, A.R. Sedgley, K.R. and McArdle, C.A. (2006) Seven-transmembrane receptor signalling and ERK compartmentalization. Trends Endocrinol. Metab., 17, 276–283. 29. Torrecilla, I. and Tobin, A. (2006) Co-ordinated covalent modiﬁcation of G-protein coupled receptors. Curr. Pharm. Des., 12, 1797–1808. 30. Haglund, K. and Dikic, I. (2005) Ubiquitylation and cell signaling. EMBO J., 24, 3353–3359. 31. Perroy, J. Pontier, S. Charest, P. et al. (2005) Real-time monitoring of ubiquitination in living cells by BRET. Nat. Methods, 1, 203–208. Study measuring ubiquitin-β-arrestin interactions using BRET. 32. Shenoy, S. and Lefkowitz, R. (2003) Trafﬁcking patterns of β-arrestin and G protein-coupled receptors determined by the kinetics of β-arrestin deubiquitination. J. Biol. Chem., 278, 14498–14506. 33. Shenoy, S. McDonald, P. Kohout, T. and Lefkowitz, R. (2001) Regulation of receptor fate by ubiquitination of activated β2 -adrenergic receptor and β-arrestin. Science, 294, 1307–1313. 34. Wojcikiewicz, R. (2004) Regulated ubiquitination of proteins in GPCR-initiated signaling pathways. Trends Pharmacol. Sci., 25, 35–41. 35. Shenoy, S. Barak, L. Xiao, K. et al. (2007) Ubiquitination of β-arrestin links 7-transmembrane receptor endocytosis and ERK activation. J. Biol. Chem., 282, 29549–29562. 36. Hopkins, A. and Groom, C. (2002) The druggable genome. Nat. Rev. Drug Discov., 1, 727–730. 37. Takeda, S. Kadowaki, S. Haga, T. et al. (2002) Identiﬁcation of G protein-coupled receptor genes from the human genome sequence. FEBS Lett., 520, 97–101. 38. Pﬂeger, K. and Eidne, K. (2003) New technologies: bioluminescence resonance energy transfer (BRET) for the detection of real time interactions involving G-protein coupled receptors. Pituitary, 6, 141–151. Review of BRET.

REFERENCES

131

39. Pﬂeger, K. and Eidne, K. (2005) Monitoring the formation of dynamic G-protein-coupled receptor-protein complexes in living cells. Biochem. J., 385, 625–637. Review of BRET and comparison with FRET. 40. Xu, Y. Piston, D.W. and Johnson, C.H. (1999) A bioluminescence resonance energy transfer (BRET) system: application to interacting circadian clock proteins. Proc. Natl. Acad. Sci. U. S. A., 96, 151–156. 41. Kocan, M. See, H.B. Seeber, R.M. et al. (2008) Demonstration of improvements to the bioluminescence resonance energy transfer (BRET) technology for the monitoring of G protein-coupled receptors in live cells. J. Biomol. Screen., 13, 888–898. Use of recently enhanced BRET technology. 42. Boute, N. Jockers, R. and Issad, T. (2002) The use of resonance energy transfer in high-throughput screening: BRET versus FRET. Trends Pharmacol. Sci., 23, 351–354. 43. Roda, A. Guardigli, M. Pasini, P. and Mirasoli, M. (2003) Bioluminescence and chemiluminescence in drug screening. Anal. Bioanal. Chem., 377, 826–833. 44. Rios, C. Jordan, B. Gomes, I. and Devi, L. (2001) G-protein-coupled receptor dimerization: modulation of receptor function. Pharmacol. Ther., 92, 71–87. 45. Park, P. Filipek, S. Wells, J. and Palczewski, K. (2004) Oligomerization of G protein-coupled receptors: past, present, and future. Biochemistry, 43, 15643–15656. 46. James, J. Oliveira, M. Carmo, A. et al. (2006) A rigorous experimental framework for detecting protein oligomerization using bioluminescence resonance energy transfer. Nat. Methods, 3, 1001–1006. Critical review of the usage of BRET to infer protein–protein interactions. 47. Pﬂeger, K. and Eidne, K. (2006) Illuminating insights into protein–protein interactions using bioluminescence resonance energy transfer (BRET). Nat. Methods, 3, 165–174. Review and implementation of BRET. 48. Greer, L.F. and Szalay, A.A. III (2002) Imaging of light emission from the expression of luciferases in living cells and organisms: a review. Luminescence, 17, 43–74. 49. Wu, P. and Brand, L. (1994) Resonance energy transfer: methods and applications. Anal. Biochem., 218, 1–13. 50. Ward, W.W. and Cormier, M.J. (1976) In vitro energy transfer in Renilla bioluminescence. J. Phys. Chem., 80, 2289–2291. 51. Nieto, C. Pellicer, T. Balsa, D. et al. (2006) The chromosomal relBE2 toxin–antitoxin locus of Streptococcus pneumoniae: characterization and use of a bioluminescence resonance energy transfer assay to detect toxin–antitoxin interaction. Mol. Microbiol., 59, 1280–1296. 52. Shimizu, T.S. Delalez, N. Pichler, K. and Berg, H.C. (2006) Monitoring bacterial chemotaxis by using bioluminescence resonance energy transfer: absence of feedback from the ﬂagellar motors. Proc. Natl. Acad. Sci. U. S. A., 103, 2093–2097. 53. Subramanian, C. Kim, B.H. Lyssenko, N.N. et al. (2004) The Arabidopsis repressor of light signaling, COP1, is regulated by nuclear exclusion: mutational analysis by bioluminescence resonance energy transfer. Proc. Natl. Acad. Sci. U. S. A., 101, 6798–6802. 54. Xu, X. Soutto, M. Xie, Q. et al. (2007) Imaging protein interactions with bioluminescence resonance energy transfer (BRET) in plant and mammalian cells and tissues. Proc. Natl. Acad. Sci. U. S. A., 104, 10264–10269.

132

CH 6 MONITORING GPCR–PROTEIN COMPLEXES

55. F¨orster, T. (1948) Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann. Phys., 437, 55–75. 56. Charest, P. Terrillon, S. and Bouvier, M. (2005) Monitoring agonist-promoted conformational changes of beta-arrestin in living cells by intramolecular BRET. EMBO Rep., 6, 334–340. Describes the use of BRET to measure intramolecular conformational changes of β-arrestin. 57. Giepmans, B. Adams, S. Ellisman, M. and Tsien, R. (2006) The ﬂuorescent toolbox for assessing protein location and function. Science, 312, 217–224. 58. Selvin, P.R. (2000) The renaissance of ﬂuorescence resonance energy transfer. Nat. Struct. Mol. Biol., 7, 730–734. 59. De, A. Loening, A.M. and Gambhir, S.S. (2007) An improved bioluminescence resonance energy transfer strategy for imaging intracellular events in single cells and living subjects. Cancer Res., 67, 7175–7183. Describes current advances in optimizing BRET in general. 60. Loening, A. Wu, A. and Gambhir, S. (2007) Red-shifted Renilla reniformis luciferase variants for imaging in living subjects. Nat. Methods, 4, 641–643. 61. Shaner, N.C. Steinbach, P.A. and Tsien, R.Y. (2005) A guide to choosing ﬂuorescent proteins. Nat. Methods, 2, 905–909. 62. Tsien, R.Y. (1998) The green ﬂuorescent protein. Annu. Rev. Biochem., 67, 509–544. Overview of ﬂuorescent proteins and their classes. 63. Sedgwick, J.D. and Czerkinsky, C. (1992) Detection of cell-surface molecules, secreted products of single cells and cellular proliferation by enzyme immunoassay. J. Immunol. Methods, 150, 159–175. 64. Thomas, P. and Smart, T. (2005) HEK 293 cell line: a vehicle for the expression of recombinant proteins. J. Pharmacol. Toxicol. Methods, 51, 187–200. 65. Liu, W. Xiong, Y. and Gossen, M. (2006) Stability and homogeneity of transgene expression in isogenic cells. J. Mol. Med., 84, 57–64. 66. Remy, I. Wilson, I.A. and Michnick, S.W. (1999) Erythropoietin receptor activation by a ligand-induced conformation change. Science, 283, 990–993. 67. Marullo, S. and Bouvier, M. (2007) Resonance energy transfer approaches in molecular pharmacology and beyond. Trends Pharmacol. Sci., 28, 362–365. Recent update on BRET technology and uses. 68. Pﬂeger, K.D. Seeber, R.M. and Eidne, K.A. (2006) Bioluminescence resonance energy transfer (BRET) for the real-time detection of protein–protein interactions. Nat. Protoc., 1, 337–345. 69. Pﬂeger, K. Kroeger, K. and Eidne, K. (2004) Receptors for hypothalamic releasing hormones TRH and GnRH: oligomerization and interactions with intracellular proteins. Semin. Cell Dev. Biol., 15, 269–280. 70. Turu, G. Szidonya, L. Gaborik, Z. et al. (2006) Differential β-arrestin binding of AT1 and AT2 angiotensin receptors. FEBS Lett., 580, 41–45. 71. Harrison, C. and van der Graaf, P.H. (2006) Current methods used to investigate G protein coupled receptor oligomerisation. J. Pharmacol. Toxicol. Methods, 54, 26–35.

7 Using Intramolecular Fluorescence Resonance Energy Transfer to Study Receptor Conformation Cornelius Krasel1 and Carsten Hoffmann2 1 School

of Pharmacy, University of Reading, Reading, UK of Pharmacology, University of W¨urzburg, W¨urzburg, Germany

2 Department

7.1 Introduction G protein-coupled receptors may undergo conformational changes upon ligand binding that are transmitted to their intracellular surface where they affect the interaction of the receptor with effector proteins such as heterotrimeric G proteins, G protein-coupled receptor kinases or arrestins. A large number of reports suggest that these conformational changes involve rearrangements of the transmembrane helices of the receptor. To measure the kinetics of this conformational change in real time, spectroscopic methods are particularly suitable as they allow a nondestructive readout of some aspect of receptor conformation. Rhodopsin, the visual pigment, was the ﬁrst (and is still the most popular) G protein-coupled receptor that was subjected to spectroscopic studies. This is due to two reasons. First, rhodopsin is comparatively easy to purify and prepare in large quantities. Second, rhodopsin carries a covalently bound ligand (11-cis-retinal) that is amenable to spectroscopic detection. From a large number of studies, several intermediate conformations have been identiﬁed (reviewed in [1]). Brieﬂy, light induces the isomerization of 11-cis-retinal to its all-trans isomer. This initial process occurs on the G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

134

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

femtosecond timescale. The energized chromophore–opsin complex proceeds through at least two other states: photorhodopsin and bathorhodopsin. After light activation, the transition to bathorhodopsin occurs on the nanosecond timescale. Bathorhodopsin then thermally relaxes to form lumirhodopsin, a process that takes microseconds. Finally, conformations of rhodopsins are reached that are capable of acting with G proteins, namely metarhodopsin I and II. This process takes place on the millisecond timescale. Observations of conformational changes in other G protein-coupled receptors have not yet reached the same level of detail. This is mostly because ﬂuorescent groups have to be introduced artiﬁcially into these receptors. This feat was ﬁrst achieved by the group of Brian Kobilka in human β2 -adrenergic receptors puriﬁed from Sf9 cells. While they initially labelled native receptors [2], much greater versatility was achieved later by constructing receptors containing only the minimal amount of cysteine and no lysine residues [3, 4]. Cysteines and lysines could be introduced into these receptors at arbitrary sites and then chemically labelled with suitable probes. These experiments have been reviewed recently by Kobilka [5]. From a number of different experiments, it was suggested that the β2 -adrenergic receptor goes through at least three different conformations when binding a full agonist, for example, noradrenaline or isoproterenol. These include: (i) conformational changes in the aromatic amino acids in helix 6, triggered by the binding of the catechol hydroxyl groups to the serine residues in helix 5 (called the ‘rotamer toggle’); (b) binding of the β-hydroxyl group to the aspartate counterion in helix 3; (iii) break of the ionic lock within the DRY motif in the second intracellular loop below helix 3. However, the mutation and puriﬁcation of G protein-coupled receptors may affect their properties. It is desirable, therefore, to measure the kinetics of conformational changes in living cells. Fluorescent labelling of proteins in vivo was ﬁrst made possible with the cloning of green ﬂuorescent protein (GFP) and its spectral variants and later with the development of cell-permeable dyes that could be used to label short peptides. The ﬁrst of these was the ‘ﬂuorescein-based arsenical hairpin binder’ (FlAsH) [6], now commercially available as TC-FlAsH (previously Lumio Green) from Invitrogen, later followed by a red variant, ‘resoruﬁn-based arsenical hairpin binder’ (ReAsH) [7], commercially available as TC-ReAsH (previously Lumio Red) (reviewed in [8]). These two dyes bind selectively to the short peptide sequence CCPGCC. Very recently, a Cy3-based biarsenical dye has been reported that binds to a distinct peptide, CCKAEAACC [9]. The ﬂuorescent properties of some of the frequently used ﬂuorophores in G protein-coupled receptor research are summarized in Table 7.1; more information about other ﬂuorescent proteins can be found in [10]. The purpose of this chapter is to elaborate further on the successful creation of these receptor sensors. Carefully selected probes could be used for ﬂuorescence resonance energy transfer (FRET) measurements which make it possible to read out changes in distance or orientation between two ﬂuorophores [14]. Initially, we used this approach to construct G protein-coupled receptors carrying two ﬂuorescent proteins: one in the third intracellular loop and one in the C-terminal tail [15]. Later, the ﬂuorescent protein in the third intracellular loop was replaced with an acceptor peptide sequence for FlAsH [16]. At the time of writing, this technique has been applied successfully to the parathyroid hormone receptor [15], the α2A -adrenergic receptor [15, 17, 18], the A2A adenosine

135

7.1 INTRODUCTION

Table 7.1 Excitation and emission maxima of selected ﬂuorophores. Excitation (nm)

Emission (nm)

Cerulean [11]

433

475

Emerald [12]

487

509

FlAsH/Lumio Green [13]

508

528

Enhanced YFP [12]

514

527

ReAsH/Lumio Red [13]

593

608

receptor [16] and the β1 -adrenergic receptor [19]. There have also been reports of a β2 -adrenergic receptor construct [20]; in that paper, FRET traces were shown but no individual ﬂuorescence traces were shown, and the kinetics of agonist-induced FRET change seems unusually slow. So far, all the receptors that were investigated showed a reduction in FRET for agonist binding (with the exception of the report on the β2 -adrenergic receptor [20]) and an increase in FRET for binding of inverse agonists. This might be coincidental. The kinetics of conformational change are different for each receptor–agonist combination and obviously depend on ligand concentration. Noradrenaline activated the α2A -adrenergic receptor with a half-life of less than 30 ms [15], and another full agonist, UK-14304, showed similar kinetics [17, 18]. Interestingly, all the partial agonists tested at the α2A -adrenergic receptor showed much slower half-lives: moxonidine and dopamine have half-lives of about 70 ms; octopamine, norphenylephrine, clonidine and oxymetazoline show half-lives of between 250 and 350 ms [18]; and the inverse agonist yohimbine shows a half-life of about 1.1–1.2 s [17]. This was taken as evidence that full agonists, partial agonists and inverse agonists switched the receptor into different conformations. Adenosine activated the A2A adenosine receptor with a half-life of about 40 ms [16], similar to the kinetics of norepinephrine at the β1 -adrenergic receptor [19]. Time constants for inverse agonists at the β1 -adrenergic receptor were not reported. Parathyroid hormone at its receptor showed a half-life of 0.7 s [15]. Both the in vitro and the in situ techniques measure conformational changes in G protein-coupled receptors. The advantages of labelling puriﬁed proteins are that many different ﬂuorophores can be used (cell permeability is not an issue) and that predictions can actually be made of how a receptor labelled at a particular position will react to the application of a ligand. Results of these measurements have been successfully combined with molecular modelling [21]. In contrast, the use of variants of GFP gives little spatial information on where the ﬂuorophore is located and yields insufﬁcient information about the exact nature of the conformational change. This is slightly alleviated with the use of FlAsH/ReAsH, and progress in the ﬁeld of small cell-permeable ﬂuorophores is quite rapid, although even here the acceptor inserted into the receptor is bigger than just the mutation of a single side chain. It is known that β2 -adrenergic receptors activate Gs within a few hundred milliseconds [22] and bind arrestins within a few seconds [23]. This suggests that the kinetics of conformational change measured in situ are related to subsequent signalling processes.

136

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

The kinetics of conformational changes measured in vitro are often much slower (in the range of tens of seconds). Future research will have to address how these rates relate to each other.

7.2 Methods and approaches 7.2.1 Selection of good insertion sites for the ﬂuorophores Results using this technique have so far only been reported for FRET pairs positioned in the third intracellular loop and the C-terminal tail. However, the length of the third intracellular loop varies widely between G protein-coupled receptors. Therefore, to allow optimal reporting of changes in FRET that occur upon agonist stimulation, each receptor construct has to be individually optimized with respect to the relative positioning of both ﬂuorophores. In this chapter, we will give some general advice and report examples from our own experience to aid with this task. When using GFP variants for FRET measurements one needs to be aware that the ˚ away actual ﬂuorophore is buried within the β-barrel structure and approximately 15 A from the site where the ﬂuorescent protein is actually fused to the target protein. This is different when using small ﬂuorophores like FlAsH, where the ﬂuorophore binds directly to its target sequence which is encoded in the receptor sequence. For G protein-coupled receptors with a long C-terminus, a truncation of the C-terminus may be necessary to obtain optimal FRET between the two ﬂuorophores. However, if the C-terminus plays an important role for cell surface expression, then truncation may prevent proper targeting of the receptor, making it much more difﬁcult to generate working receptor constructs. For receptors with a short C-terminus, it may be useful to insert a short linker to provide convenient cloning sites and to optimize the ﬂuorophore distance to obtain good FRET signals. We use the same approach for the C-terminus regardless of whether we work with FlAsH/cyan ﬂuorescent protein (CFP) or CFP/yellow ﬂuorescent protein (YFP). The treatment of the third intracellular loop depends on which combination of ﬂuorophores is being used. If inserting a GFP variant, then we usually prefer to use YFP, as it seems to cause less mistargeting of the mutated receptor than CFP. The N˚ apart and C-termini of GFP in the resolved crystal structure are approximately 26 A (as measured for the carbonyl oxygen for Glu-6 and Ile-229). Although the ﬁrst ﬁve N-terminal amino acids of GFP were not resolved in the crystal structure, this distance has to be accommodated by the third intracellular loop when inserting the ﬂuorophore. This can be achieved by appropriate truncations. In contrast, if using FlAsH as second ﬂuorophore, then a sequence of six amino acids (CCPGCC) is sufﬁcient for speciﬁc binding of the ﬂuorophore to the target protein. The insertion site of the ﬂuorophore should not be too close to the ends of transmembrane helix 6. In the crystal structure of rhodopsin this helix extends further into the cytoplasm than estimated from hydropathy plots. Therefore, we assume that this will be the case for other G protein-coupled receptors as well. This helical extension is thought to be important for G protein coupling [24]; therefore, it is advisable not to disturb it by introducing the bulky YFP or the turn-inducing CCPGCC motif

137

7.2 METHODS AND APPROACHES NH2

TM1

TM2

TM3

TM4

TM5

TM6

TM7

COOH S

S

S

As HO

S As

O

O

COOH

Figure 7.1 Schematic representation of a GPCR-FRET sensor. CFP is fused to the C-terminus of the receptor, while YFP or the FlAsH-binding sequence is introduced into the third intracellular loop. CFP and YFP can also be swapped. The small box represents FlAsH in its approximate size relation to CFP. The enlarged box shows the chemical structure of FlAsH.

(Figure 7.1). YFP was inserted into the α2A -adrenergic receptor by replacing a large part of the third intracellular loop from Cys-239 to Ala-357 with two restriction sites which were used to insert the YFP protein [25]. To achieve this, two restriction sites were inserted when removing the third intracellular loop, and subsequently YFP was cloned between them. In contrast, for the β1 -adrenergic receptor, YFP was inserted between two adjacent amino acids [19], residues 273 and 274. In the FlAsH-labelled adenosine A2A receptor, the CCPGCC motif was inserted around a naturally occurring PG motif (residues 217 and 218) in the third intracellular loop. Thus, amino acids 215, 216, 219 and 220 were mutated to Cys. The C-terminus of the A2A receptor was truncated from 122 to 50 amino acids behind residue 340 to obtain good FRET signals upon receptor stimulation. This was possible since it had been described that this large part was not necessary for surface expression of the receptor [26]. For these examples, the distances were: • α 2A -adrenergic receptor – TM5–YFP, 22 amino acids; YFP–TM6, 17 amino acids; • β1 -adrenergic receptor – TM5–YFP, 28 amino acids; YFP–TM6, 52 amino acids; • adenosine A2A receptor – TM5–CCPGCC, 16 amino acids; CCPGCC–TM6, 14 amino acids.

138

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

7.2.2 Biarsenical ﬂuorophores Biarsenical ﬂuorophores, such as FlAsH, bind to a recognition motif that usually has to be inserted into the target protein by mutagenesis. The core sequence of this binding motif consists of the CCPGCC sequence [13], which is sufﬁcient for speciﬁc binding. Motifs with higher afﬁnity for biarsenicals have been identiﬁed that include speciﬁc ﬂanking sequences [27], and the motif FLNCCPGCCMEP was shown to be of particular high afﬁnity for FlAsH. Here, we will brieﬂy describe how to insert the sequence into the protein sequence by mutagenesis. As outlined above, it is necessary to stay away from the border of the transmembrane domain and the intracellular loop to avoid the generation of constitutive activity or interference with G protein coupling. We use the information about transmembrane domains and loop regions provided by the GPCRDB at http://www.GPCR.org for most receptors. To introduce the CCPGCC sequence into the third intracellular loop, we use standard polymerase chain reaction (PCR) techniques. If the CCPGCC sequence is encoded by the nucleotide sequence TGT TGC CCG GGC TGC TGT, this sequence contains a CCCGGG which can be cut blunt end by SmaI or with overlapping fragments by XmaI (see Figure 7.2). By selecting primer combinations of your speciﬁc receptor and the sequence to be added, one ends up with two PCR fragments and the vector; thus, a three-piece ligation will generate the desired construct and the inserted SmaI/XmaI site can be used for colony screening. This strategy can, of course, only be used if no endogenous SmaI/XmaI site is present in the receptor sequence. Otherwise, standard techniques for mutagenesis can be used (e.g. see descriptions in [28]). For the adenosine A2a receptor the CCPGCC sequence was created by ﬂanking an endogenous PG in the third intracellular loop with two cysteines on each site. If no such sequence exists then one should be aware that the CCPGCC most likely forms a hairpin structure [13] and, thereby, can cause tension within the loop structure if it is replacing endogenous amino acids. As shown for rhodopsin [29], a small, third intracellular loop does not provide much space to accommodate this motif, since a part of the sequence is an extension of the transmembrane helices. If a replacement of endogenous amino acids leads to an intracellular retention of the receptor, then one could try to insert the whole sequence rather than substituting it for endogenous amino acids. It is also possible to insert the enhanced 12 amino acid motif and, thereby, further reduce the possible tensions introduced by a hairpin sequence. This would be a larger change to the receptor compared with the six amino acid motif, but still signiﬁcantly less of a change than a ﬂuorescent protein in size.

7.2.3 Cellular labelling with FlAsH FlAsH is commercially available through Invitrogen and is sold as TC-FlAsH (previously Lumio Green). Cells are labelled with FlAsH by following the manufacturer’s instructions. Since there are a number of different applications which vary in terms of labelling, we brieﬂy summarize the individual steps for labelling GPCRs and comment on important steps (Protocol 7.1). For all our experiments using the superfusion system it is necessary to ensure that the cells are appropriately attached. Therefore, cells are seeded onto 25 mm

139

7.2 METHODS AND APPROACHES

Task:

Site I

CCPGCC

Site II

ATG

Site I

PCR 1:

Site II

ATG ACAACGGGCCCGACGACA SmaI / XmaI

Site I

SmaI / XmaI

Site II

TGTTGCCCGGGCTGCTGT

PCR 2:

ATG

Site I

Digest:

SmaI / XmaI

PCR 1 fragment:

Site II PCR 2 fragment:

SmaI / XmaI

Site I

Site II

Vector:

Site I

Ligation:

ATG

SmaI / XmaI

Site II

CCPGCC

Figure 7.2 Cloning strategy for insertion of the CCPGCC sequence into a GPCR. The ﬁrst task is to identify two unique restriction sites that are conveniently located within the receptor of interest. These enzymes should yield incompatible ends; otherwise, the three-fragment ligation further down is not possible. In a second step, two individual PCR reactions are performed that add the CCPGCC motif using the coding sequence given in the text. In a third step, the ampliﬁed PCR fragments and the vector are digested with the appropriate restriction enzymes and subsequently puriﬁed. Finally, the three puriﬁed fragments are combined in a single ligation reaction (three-fragment ligation) which is transformed into competent bacteria by the usual methods. To screen colonies for positives, the SmaI/XmaI site can be used.

140

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

glass cover-slips that have been treated with poly-d-lysine for 15–30 min prior to use. Mostly, we use six-well plates for labelling. This ensures that 1 ml volume is sufﬁcient to cover all cells and also allows one to close the plate with a good sealing lid to prevent the unpleasant odour of ethanedithiol (EDT) from spreading. Using standard protocols cells are transfected 24–48 hours prior to labelling. We have used diethylaminoethyl-dextran, calcium phosphate, Effectene and Lipofectamine for transfection and each of these transfection methods has worked in combination with FlAsH labelling. We also used a variety of standard cell lines, like HEK-293, HEK-TSA, Cos-7, HeLa and PC12, and found them all useful in combination with this labelling technique. Time of transfection prior to labelling has to be individually tested and is only a general recommendation. However, cell density should be between 60 and 90% conﬂuency to ensure the best labelling results. This is also recommended to allow perfused ligands to reach the cell-surface-located receptor without diffusion problems; thus, changes in FRET would report accurate kinetics. Labelling with FlAsH is always accompanied by background staining. This is unavoidable, but can be reduced to a great extent. However, it has to be kept in mind that proteins with high local concentrations, like membrane-localized GPCRs, have a great advantage for visualization of labelling above background. Since EDT has an unpleasant odour, all labelling steps should be done under a fume hood.

PROTOCOL 7.1 Labelling Cells with FlAsH Equipment and Reagents • Cell culture hood and incubator • phosphate-buffered saline (PBS): 8 mM Na2 HPO4 , 1.5 mM KH2 PO4 , 137 mM NaCl, 2.7 mM KCl, 0.90 mM CaCl2 , 0.49 mM MgCl2 ; pH 7.4 • HEPES-buffered saline (HBS, sometimes also called HBSS): 150 mM NaCl, 10 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES); 2.5 mM KCl, 4 mM CaCl2 , 2 mM MgCl2 ; adjust to pH 7.2 with NaOH. This solution can optionally be supplemented with 10 mM glucose. • TC-FlAsH in-cell tetracysteine tag detection kit (Invitrogen) • 1× British anti-Lewisite (BAL, 2,3-dimercaptopropanol) washing solution: 250 µM BAL in HBS or PBS, diluted according to manufacturer’s instructions from 100× stock provided with the TC-FlAsH kit. Previous versions of the kit (sold under the Lumio label) did not have the BAL solution included. • Normal culture medium for the cells.

Method 1 Aspirate off the medium.  2 Wash the cells once with PBS or HBS (Invitrogen recommends using Opti-Mem for all steps, including the labelling; we have not noticed any difference to using HBS supplemented with 10 mM glucose).

7.2 METHODS AND APPROACHES

141

3 Aspirate off the PBS or HBS. 4 Add 1 ml labelling solution to each well of a six-well plate. The manufacturer recommends adding FlAsH-EDT2 solution at a concentration of 2.5 µM FlAsH. The FlAsH stock solution provided is 2 mM and called 800×. Thus, adding 1.25 µl to 1 ml buffer solution results in 1 ml FlAsH labelling solution of the appropriate concentration. We have used 1 µM concentrations of FlAsH for the adenosine A2A receptor and have not observed labelling problems by using lower concentrations. 5 Incubate the cells for 30 min at room temperature. This time is recommended by Invitrogen. Further increase in incubation time may result in an increase in unspeciﬁc binding. Our experience is that the time can be increased to 60 min without problems, and in fact an incubation time of 30 min would not have been sufﬁcient to label the control construct fully [16] and, therefore, would have led to false data when determining FRET efﬁciency. We also prefer to label the cells at 37 ◦ C in the cell incubator. This automatically protects FlAsH from light exposure, and the cells are kept in better conditions for later experiments. 6 Carefully aspirate off the labelling solution. 7 Wash the cells once with 2 ml 1× BAL washing solution. This step removes the nonspeciﬁcally bound FlAsH. The concentration of 1× BAL is 250 µM and sufﬁciently high to remove background staining. The manufacturer recommends a simple wash with BAL, but we prefer to incubate the cells for 5–10 min with the wash solution at 37 ◦ C in the cell incubator for the same reasons as mentioned in step 5. The 5–10 min incubation time does, in our hands, result in a better reduction of background staining than a simple wash by adding and aspirating. 8 Carefully aspirate the wash solution. 9 Wash the cells twice (or more) with HBS or PBS to remove BAL. This step ensures the removal of the remaining BAL in your samples. Although the odour of BAL is not as unpleasant as that of EDT, it is still a substance that can easily be noticed. 10 Add normal culture medium and place the cells in the incubator until further needed. We have used cells that have been labelled this way after 8 h without problems. However, labelled protein still undergoes turnover; therefore, the time delay for measuring FRET efﬁciencies should be kept very short to ensure proper results. Measurements of GPCR activation can be done at later time without problems, since individual corrections are made for each experiment (see below).

7.2.4 Measuring ligand-mediated changes in FRET This is described in Protocol 7.2. The microscope should be equipped with a 63× or 100× lens to make it possible to visualize single cells. The cells are illuminated with the appropriate wavelength. A ﬂash lamp (e.g. Till Polychrome) is used to minimize photobleaching, compared with continuous illumination. We use a photometry system from Till Photonics as a detector, which simultaneously allows us to acquire ﬂuorescence intensity of an area at two wavelengths. The detector is equipped with a beamsplitter that guides the incident light to two avalanche photodiodes. For CFP/YFP, the

142

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

beamsplitter is a DCLP505 (Chroma Technology). The CFP light is passed through a 480 ± 20 nm bandpass ﬁlter. For YFP, either a 535 ± 15 nm bandpass ﬁlter or an LP515 longpass ﬁlter can be used. The photometry system is equipped with a viewﬁnder which, in its current version, does ﬁlter out light >600 nm; therefore, it cannot be used for measuring FRET at longer wavelengths. The signals produced by the photodiodes are digitized with a Digidata 1322 (Axon Instruments, Union City, CA, USA) and acquired on a PC with Clampex8.1 or Axioscope software (Axon Instruments). Clampex/Axioscope also controls the perfusion system (see below). Alternatively, it is possible to use a camera to detect the ﬂuorescence. We have used the Coolsnap HQ in combination with a MultiSpec-Imager image splitter (Optical Insights). The disadvantage, compared with the photometry system, is that the sensitivity of the camera and, therefore, the temporal resolution are lower. This should be even more dramatic for a ﬁlter wheel system. For the application of the ligand of choice, we use a perfusion system (ALA-VM8, ALA Scientiﬁc Instruments) that is driven by compressed nitrogen/air. This system has the advantage of very short switching times (<10 ms). The tip of the perfusion system is positioned close to the cell using a micromanipulator (e.g. M¨arzh¨auser MM33).

PROTOCOL 7.2 Measuring Ligand-mediated Changes in FRET Equipment and Reagents • Inverted microscope (e.g. Axiovert135, Axiovert200 or Nikon TE2000) • Fluorescence ﬂash lamp (e.g. Polychrome IV or Polychrome V from Till Photonics) • Detector (e.g. photometry system from Till Photonics) • Perfusion system (e.g. ALA BPS-8SP from ALA Scientiﬁc Instruments) • Appropriate ligand • Attoﬂuor holder (Invitrogen) for mounting 25 mM diameter cover-slips • HBS: 150 mM NaCl, 10 mM HEPES; 2.5 mM KCl, 4 mM CaCl2 , 2 mM MgCl2 ; adjust to pH 7.2 with NaOH. This solution can optionally be supplemented with 10 mM glucose.

Method 1 Fill the syringes of the perfusion system. One syringe needs to be ﬁlled with HBS; the other syringes can be ﬁlled with the ligand at a desired concentration dissolved in HBS. 2 Mount a cover-slip in an Attoﬂuor holder on the microscope in HBS. It may be advantageous to include glucose in the HBS. 3 Using excitation at either 436 nM (CFP) or 505 nM (YFP), look for a cell that shows proper ﬂuorescence localization. Preferentially, look for a cell which shows good membrane staining with little to no internal ﬂuorescence. As soon as such a cell is found, switch

7.3 TROUBLESHOOTING

143

off the illumination, to avoid excessive photobleaching of the ﬂuorophores before the measurement. 4 Use the viewﬁnder to make sure that the cell is actually seen by the photometry system. 5 Start Clampex/Axioscope. Switch on the perfusion system and perfuse the cell with HBS. FRET will decrease over time, as the YFP or FlAsH acceptor usually bleaches faster than the CFP donor. 6 After about 30–60 s, start perfusing the cells with agonist. Hopefully, a FRET change will occur when the ligand is applied. YFP/FlAsH and CFP traces move in opposite directions.

7.3 Troubleshooting • The most frequently encountered problem is that there is no FRET change when the cells are perfused with agonist. This is often caused either by a general problem of the construct or by some failure to perfuse the cell properly. Constructing a receptor in the way described above does in no way guarantee a construct which will show agonist-induced FRET changes. For example, to get the parathyroid hormone receptor sensor, more than half a dozen different constructs were made, only one of which showed agonist-induced changes in FRET. On the other hand, sometimes we were very lucky, in that the ﬁrst construct made did show agonist-induced changes in FRET. • The thin tubing of the perfusion system sometimes clogs. This can either be caused by air bubbles or by particles. To get rid of these, refer to the manufacturer’s instructions. Clogging can often be avoided by storing the perfusion system in 20% ethanol. • The tip of the perfusion system must be quite close to the cell. With a 100× lens we get good results if the tip is just outside the ﬁeld of vision. If the tip is positioned too close to the cell, then the cell may be washed away. • Sometimes, an unacceptable amount of noise is encountered. This can be reduced by installing the microscope plus the perfusion system on an antivibration table. Noise is also caused by dirt in the solutions, so ﬁltration of the solutions through a 0.22 µm ﬁlter might be considered.

Note Recently we have become aware of the creation of conformational sensors for the human B2 bradykinin receptor [30] and the murine M1 muscarinic acetylcholine receptor [31]. For both of these sensors, eYFP was inserted into the third intracellular loop and Cerulean [11] attached to the C-terminus; in the case of the B2 bradykinin receptor, the C-terminus was truncated before attaching the Cerulean to increase intramolecular

144

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

FRET [30]. The M1 sensor is the ﬁrst sensor to display a FRET increase upon agonist application [31] rather than the decrease observed for all other sensors so far. The following table provides an overview of all published conformational sensors for G protein-coupled receptors. Receptor

A2A adenosine

Position of insertion in third intracellular loop (numbering based on Swissprot)

CCPGCC replacing Pro-215 to Arg-220 α 2A -adrenergic YFP replacing Cys-239 to Ala-357 α 2A -adrenergic CCPGCC inserted between Ala-238 and Ser-360 α 2A -adrenergic FLNCCPGCC MEP inserted between Ala-246 and Arg-257 α 2A -adrenergic FLNCCPGCC MEP inserted between Ser-297 and Arg-308 α 2A -adrenergic FLNCCPGCC MEP inserted between Gly-350 and Arg-361 YFP inserted between Pro-273 β 1 -adrenergic and Ser-274 YFP inserted between Phe-232 B2 bradykinin and Lys-233 M1 acetylcholine YFP inserted between Ala-223 and Val-358

Distance Distance Reference from TM5 from TM6 16

14

[16]

22 24

17 15

[25] [16]

28

128

[32]

80

67

[32]

133

13

[32]

28

52

[19]

11

8

[30]

14

9

[31]

References 1. Okada, T., Ernst, O.P., Palczewski, K. and Hofmann, K.P. (2001) Activation of rhodopsin: new insights from structural and biochemical studies. Trends Biochem. Sci., 26, 318–324. 2. Gether, U., Lin, S. and Kobilka, B.K. (1995) Fluorescent labeling of puriﬁed β2 -adrenergic receptor. Evidence for ligand-speciﬁc conformational changes. J. Biol. Chem., 270, 28268–28275. 3. Gether, U., Lin, S., Ghanouni, P. et al. (1997) Agonists induce conformational changes in transmembrane domains III and VI of the β2 adrenoceptor. EMBO J., 16, 6737–6747. 4. Parola, A.L., Lin, S. and Kobilka, B.K. (1997) Site-speciﬁc ﬂuorescence labeling of the β2 adrenergic receptor amino terminus. Anal. Biochem., 254, 88–95. 5. Kobilka, B.K. (2007) G protein coupled receptor structure and activation. Biochim. Biophys. Acta, 1768, 794–807. Reviews ﬂuorescence labelling of puriﬁed β2 -adrenergic receptors. 6. Grifﬁn, B.A., Adams, S.R. and Tsien, R.Y. (1998) Speciﬁc covalent labeling of recombinant protein molecules inside live cells. Science, 281, 269–272. First description of FlAsH reagent.

REFERENCES

145

7. Gaietta, G., Deerinck, T.J., Adams, S.R. et al. (2002) Multicolor and electron microscopic imaging of connexin trafﬁcking. Science, 296, 503–507. 8. Giepmans, B.N., Adams, S.R., Ellisman, M.H. and Tsien, R.Y. (2006) The ﬂuorescent toolbox for assessing protein location and function. Science, 312, 217–224. Useful review of available ﬂuorescence techniques. 9. Cao, H., Xiong, Y., Wang, T. et al. (2007) A red Cy3-based biarsenical ﬂuorescent probe targeted to a complementary binding peptide. J. Am. Chem. Soc., 129, 8672–8673. 10. Shaner, N.C., Steinbach, P.A. and Tsien, R.Y. (2005) A guide to choosing ﬂuorescent proteins. Nat. Methods, 2, 905–909. 11. Rizzo, M.A., Springer, G.H., Granada, B. and Piston, D.W. (2004) An improved cyan ﬂuorescent protein variant useful for FRET. Nat. Biotechnol., 22, 445–449. 12. Tsien, R.Y. (1998) The green ﬂuorescent protein. Annu. Rev. Biochem., 67, 509–544. 13. Adams, S.R., Campbell, R.E., Gross, L.A. et al. (2002) New biarsenical ligands and tetracysteine motifs for protein labeling in vitro and in vivo: synthesis and biological applications. J. Am. Chem. Soc., 124, 6063–6076. Optimization of the original FlAsH labelling protocol. 14. Stryer, L. (1978) Fluorescence energy transfer as a spectroscopic ruler. Annu. Rev. Biochem., 47, 819–846. 15. Vilardaga, J-P., B¨unemann, M., Krasel, C. et al. (2003) Measurement of the millisecond activation switch of G protein-coupled receptors in living cells. Nat. Biotechnol., 21, 807–812. First cellular expression of a conformational GPCR sensor. 16. Hoffmann, C., Gaietta, G., B¨unemann, M. et al. (2005) A FlAsH-based FRET approach to determine G protein-coupled receptor activation in living cells. Nat. Methods, 2, 171–176. First conformational GPCR sensor using FlAsH. 17. Vilardaga, J.-P., Steinmeyer, R., Harms, G.S. and Lohse, M.J. (2005) Molecular basis of inverse agonism in a G protein-coupled receptor. Nat. Chem. Biol., 1, 25–28. 18. Nikolaev, V.O., Hoffmann, C., B¨unemann, M. et al. (2006) Molecular basis of partial agonism at the neurotransmitter α2A -adrenergic receptor and Gi -protein heterotrimer. J. Biol. Chem., 281, 24506–24511. 19. Rochais, F., Vilardaga, J.P., Nikolaev, V.O. et al. (2007) Real-time optical recording of β1 -adrenergic receptor activation reveals supersensitivity of the Arg389 variant to carvedilol. J. Clin. Invest., 117, 229–235. 20. Nakanishi, J., Takarada, T., Yunoki, S. et al. (2006) FRET-based monitoring of conformational change of the β2 adrenergic receptor in living cells. Biochem. Biophys. Res. Commun., 343, 1191–1196. 21. Yao, X., Parnot, C., Deupi, X. et al. (2006) Coupling ligand structure to speciﬁc conformational switches in the β2 -adrenoceptor. Nat. Chem. Biol., 2, 417–422. 22. Hein, P., Rochais, F., Hoffmann, C. et al. (2006) Gs activation is time-limiting in initiating receptor-mediated signaling. J. Biol. Chem., 281, 33345–33351. 23. Krasel, C., B¨unemann, M., Lorenz, K. and Lohse, M.J. (2005) β-Arrestin binding to the β2 -adrenergic receptor requires both receptor phosphorylation and receptor activation. J. Biol. Chem., 280, 9528–9535. 24. Wess, J. (1998) Molecular basis of receptor/G-protein-coupling selectivity. Pharmacol. Ther., 80, 231–264.

146

CH 7 INTRAMOLECULAR FRET TO SENSE GPCR CONFORMATIONAL CHANGES

25. Vilardaga, J.P., B¨unemann, M., Krasel, C. et al. (2003) Measurement of the millisecond activation switch of G protein-coupled receptors in living cells. Nat. Biotechnol., 21, 807–812. 26. Palmer, T.M. and Stiles, G.L. (1997) Identiﬁcation of an A2a adenosine receptor domain speciﬁcally responsible for mediating short-term desensitization. Biochemistry, 36, 832–838. 27. Martin, B.R., Giepmans, B.N., Adams, S.R. and Tsien, R.Y. (2005) Mammalian cell-based optimization of the biarsenical-binding tetracysteine motif for improved ﬂuorescence and afﬁnity. Nat. Biotechnol. 23, 1308–1314. Further optimization of FlAsH labelling. 28. Sambrook, J. and Russell, D.W. (2001) Molecular Cloning: A Laboratory Manual . Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 29. Schertler, G.F. (2005) Structure of rhodopsin and the metarhodopsin I photointermediate. Curr. Opin. Struct. Biol. 15, 408–415. 30. Chachisvillis, M., Zhang, Y.L. and Frangos, J.A. (2006) G protein-coupled receptors sense ﬂuid shear stress in endothelial cells. Proc. Natl. Acad. Sci. USA, 103, 15463–15468. 31. Jensen, J.B., Lyssand, J.S., Hague, C. and Hille, B. (2009) Fluorescence changes reveal kinetic steps of muscarinic receptor-mediated modulation of phosphoinositides and Kv7.2/7.3 K+ channels. J. Gen. Physiol. 133, 347–359. 32. Z¨urn, A., Zabel, U., Vilardaga, J.-P., Schindelin, H., Lohse, M.J. and Hoffmann, C. (2009) Fluorescence resonance energy transfer analysis of α 2a -adrenergic receptor activation reveals distinct agonist-speciﬁc conformational changes. Mol. Pharmacol. 75, 534–541.

8 A Disulﬁde Cross-linking Strategy Useful for Studying Ligand-induced Structural Changes in GPCRs Jian Hua Li, Stuart D.C. Ward, Sung-Jun Han, Fadi F. Hamdan and Jurgen Wess ¨ Molecular Signaling Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, USA

8.1 Introduction The superfamily of G protein-coupled receptors (GPCRs) represents the largest group of cell-surface receptors found in nature [1–5]. A better understanding of how GPCRs function at the molecular level is critical for the design of novel classes of drugs that can modulate signalling through speciﬁc GPCR subtypes. The structural features of the ligand binding and G protein-coupling domains have been studied extensively for many different GPCRs [1–3, 6]. Moreover, biophysical and biochemical studies have provided considerable insight into the conformational changes that accompany the activation of the photoreceptor rhodopsin [7–11]. Most of these studies were carried out with mutant versions of rhodopsin following a series of receptor solubilization and puriﬁcation steps. However, as discussed recently [11], the structural and dynamic properties of the solution state of rhodopsin may not be identical with that found in native disk membranes.

G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

148

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

In contrast to the wealth of structural information that is available for bovine rhodopsin [7–11], much less is known about the sequence of molecular events that trigger the activation of GPCRs that bind diffusible ligands, such as neurotransmitters and hormones. Fluorescence-based biophysical studies carried out with puriﬁed, mutationally modiﬁed versions of the β2 -adrenergic receptor have detected activity-dependent movements of various intracellular receptor domains, although with limited molecular resolution (see [12] for a review). Thus, there is a clear need to examine GPCR structure and ligand-dependent changes in receptor conformation in greater structural detail. Ideally, the method to be employed for such studies should allow the analysis of receptors present in their native membrane environment (in situ). To address this issue, we have used the rat M3 muscarinic acetylcholine (ACh) receptor (M3 mAChR), a prototypic class I GPCR [13, 14], as a model system. Agonist occupation of the M3 mAChR leads to the preferential activation of G proteins of the Gq family, triggering the activation of different subtypes of phospholipase Cβ, which in turn leads to the generation of the second messengers diacylglycerol and inositol 1,4,5-trisphosphate (IP3 ) [13, 14]. To monitor ligand-induced changes in M3 mAChR structure, we developed a disulﬁde cross-linking strategy that allows the detection of disulﬁde bond formation between two Cys residues that are adjacent to each other in the three-dimensional (3D) structure of the receptor [15–20]. One major advantage of this strategy is that ligand-dependent conformational changes can be studied in receptors present in their native membrane environment. The differences in disulﬁde cross-linking patterns observed in the absence or the presence of muscarinic ligands can then guide predictions regarding the molecular nature of the ligand-induced conformational changes. In previous studies, these predictions were aided by a 3D model of the inactive state of the M3 mAChR which was established via homology modelling using the high-resolution X-ray structure of the dark (inactive) state of bovine rhodopsin as a template [17, 21]. Recently, a high-resolution X-ray structure of the human β2 -adrenergic receptor bound to a partial inverse agonist (carazolol) has been reported [22, 23]. Since the M3 mAChR shares a higher degree of sequence homology with the β2 -adrenergic receptor than with rhodopsin, we are currently reﬁning our M3 mAChR model based on the atomic coordinates of the newly solved β2 -adrenergic receptor structure. Cys scanning mutagenesis approaches followed by oxidative disulﬁde cross-linking were ﬁrst successfully employed to study the structure of bacterial chemotactic receptors [24–27]. In many subsequent studies, this strategy has been used to explore the structure of the native (dark) structure of bovine rhodopsin, as well as the structural changes that accompany light-induced rhodopsin activation (reviewed in [8, 11]). In the vast majority of cases, the distance between the α-carbon atoms of Cys ˚ [28, 29]. The general residues engaged in a disulﬁde bridge ranges from 4.4 to 6.8 A assumption, therefore, is that two Cys residues have the potential to form a disulﬁde bond (under the appropriate experimental conditions) when the distance between their ˚ two α-carbon atoms is less than ∼7 A.

8.2 METHODS AND APPROACHES

149

8.2 Methods and approaches 8.2.1 Generation of a modiﬁed version of the M3 mAChR suitable for disulﬁde cross-linking studies To render the M3 mAChR suitable for disulﬁde cross-linking studies, the receptor protein was modiﬁed as outlined below (also see Figure 8.1). After each individual step, we conﬁrmed that the introduced structural changes did not interfere with ligand binding and receptor/G protein coupling (see below).

Figure 8.1 Structure of the M3 (3C)-Xa receptor which served as a template for disulﬁde cross-linking studies. The M3 (3C)-Xa receptor represents a modiﬁed version of the rat M3 mAChR. The M3 (3C)-Xa construct contains only three remaining native Cys residues: C140, C220 and C532 (ﬁlled squares). All other native Cys residues were replaced with either Ser or Ala residues (open squares). In addition, the central portion of the i3 loop (A274–K469) was replaced with two adjacent factor Xa cleavage sites (underlined). The ﬁve potential N-glycosylation sites present in the N-terminal portion of the receptor protein (N6, N15, N41, N48 and N52) were replaced with Gln residues (open circles). In addition, a haemagglutinin (HA) epitope tag was inserted after the initiating Met residue. A rabbit polyclonal antibody (referred to as anti-C-M3) was raised against the indicated C-terminal receptor sequence [30]. Numbers refer to amino acid positions in the rat M3 mAChR sequence [31].

150

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

(a)

(b)

(c)

(d)

(e)

8.2 METHODS AND APPROACHES

151

1 An HA epitope tag was added to the N-terminus of the receptor protein. 2 The ﬁve Asn residues present within the extracellular N-terminal domain of the M3 mAChR, which are predicted to serve as sites for N-linked glycosylation [13, 14], were replaced with Gln residues. These point mutations were made to prevent the appearance of multiple immunoreactive M3 mAChR species on immunoblots caused by heterogeneous glycosylation. 3 The central portion of the third intracellular loop (i3 loop; A274–K469) was replaced with two adjacent factor Xa cleavage sites. For reasons that are not entirely clear, the A274–K469 deletion facilitates the detection of the M3 mAChR via western blotting. We speculated that the presence of two factor Xa cleavage sites, rather than a single site, would ensure a more efﬁcient proteolytic cleavage by factor Xa of the modiﬁed M3 mAChR. 4 In the next step, all remaining Cys residues, except for C140, C220 and C532, were replaced with either Ala or Ser, depending on which of these two amino acids was better tolerated in terms of receptor expression and function. C140 and C220 are predicted to be involved in an intramolecular disulﬁde bond which is critical for proper M3 mAChR folding and function [30]. In the context of all other Cys substitutions and structural modiﬁcations, substitution of C532 with Ala, Ser or several other amino acids led to a dramatic loss in ligand binding activity [32]. In the following, we refer to this mutationally modiﬁed M3 mAChR as M3 (3C)-Xa receptor. Importantly, the M3 (3C)-Xa receptor shows ligand binding afﬁnities and G protein-coupling properties similar to those of the wild-type M3 mAChR [32].

8.2.2 General strategy used to study disulﬁde cross-link formation in double Cys mutant M3 mAChRs In the next step, we reintroduced pairs of Cys residues into the M3 (3C)-Xa background receptor, one Cys N-terminal and the other one C-terminal of the factor Xa cleavage site (Figure 8.2). When two Cys residues face each other in the 3D structure of the receptor, Figure 8.2 Scheme outlining the strategy used to detect the formation of disulﬁde bonds between vicinal Cys residues in mutant M3 mAChRs. Pairs of Cys residues were introduced into the M3 (3C)-Xa background receptor (see Figure 8.1), one Cys N-terminal and the other one C-terminal of the factor Xa cleavage site. When two Cys residues lie adjacent to each other in the 3D structure of the receptor, they have the potential to form a disulﬁde bridge, either spontaneously or in the presence of oxidizing agents such as Cu-Phen or molecular iodine (a, c). Upon cleavage with factor Xa, the disulﬁde bridge will keep the two cleavage products covalently linked (a, c). As a result, a full-length receptor band (∼38 kDa in size) will appear on western blots run under nonreducing conditions (e, left lane). In contrast, when the two introduced Cys residues are not close to each other in the 3D structure of the receptor, they are unable to form a disulﬁde bridge (b, d). As a consequence, factor Xa digestion will yield two separate cleavage products and a full-length receptor band will not appear on western blots run under nonreducing conditions (e, right lane).

152

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

they have the potential to form a disulﬁde bridge (Figure 8.2), either spontaneously or in the presence of oxidizing agents such as Cu(II)-(1,10-phenanthroline)3 (Cu-Phen) or molecular iodine. Upon cleavage with factor Xa, the disulﬁde bridge will keep the two cleavage products covalently linked. As a result, a full-length receptor band, ∼38 kDa in size, will appear on western blots run under nonreducing conditions (Figure 8.2).

8.2.3 Summary of activity-dependent conformational changes deduced from cross-linking studies with double Cys mutant M3 mAChRs During the past few years, we used the approach summarized in Figure 8.2 to analyse more than 100 different double Cys mutant M3 mAChRs [15–20]. The positions that we targeted by Cys substitution mutagenesis are highlighted in Figure 8.3. To ensure

Figure 8.3 Agonist-promoted disulﬁde cross-link formation in double Cys mutant M3 mAChRs. All Cys substitutions were introduced into the M3 (3C)-Xa receptor (see Figure 8.1). Muscarinic agonists promoted the formation of disulﬁde cross-links between Cys residues present at positions 151 (3.36) and 532 (7.42) (red circles [18]), 88 (1.53) and 543 (7.53) (brown circles [17]), 91 (1.56) and 545 (7.55)/546 (7.56) (orange circles [17]), 169 (3.54) and 484 (6.29)/488 (6.33) (purple circles [19]), and 254 (5.62) and 489 (6.34)/490 (6.35)/491 (6.36)/492 (6.37) (green circles [15]). In contrast, muscarinic agonists inhibited the formation of disulﬁde cross-links between Cys residues present at positions 91 (1.56)/92 (1.57) and 549 (7.59)/550 (7.60) (blue circles [20]). Numbers refer to amino acid positions in the rat M3 mAChR sequence [31]. A colour reproduction of this ﬁgure can be viewed in the colour plate 1 towards the centre of the book.

153

8.2 METHODS AND APPROACHES

that the double Cys mutant receptors were properly folded, all receptors were transiently expressed in COS-7 cells and studied in radioligand binding assays for their ability to bind agonist and antagonist ligands. Moreover, to conﬁrm that the various Cys substitutions did not interfere with receptor-mediated G protein activation, all double Cys mutant receptors were examined for their ability to stimulate agonist-mediated phosphatidyl inositol hydrolysis. One key observation that we made was that muscarinic agonists promoted the formation of a disulﬁde bond between C532 (position 7.42 according to the Ballesteros–Weinstein nomenclature of GPCRs [33]) and a Cys residue introduced at position 151 (position 3.36) (Figure 8.3). In conjunction with a 3D model of the inactive state of the M3 mAChR, this observation supported the concept that agonist activation of the M3 mAChR pulls the extracellular parts of transmembrane domain (TM) VII and III closer to each other (Figure 8.4a). It is likely that this movement represents (a)

(b)

(c)

Figure 8.4 Models of the M3 mAChR depicting agonist-induced changes in receptor structure. A 3D model of the inactive state of the rat M3 mAChR was built via homology modelling using the high-resolution X-ray structure of bovine rhodopsin as a template [17, 21]. (a) Extracellular view of the M3 mAChR parallel to the path of the exofacial segment of TM III. For the sake of clarity, only selected TM helices are shown. We found that muscarinic agonists promote the formation of a disulﬁde bond between C532 (7.42) and a Cys residue introduced at position 151 (3.36) [18]. This observation supports the concept that agonist activation of the M3 mAChR pulls the extracellular segments of TM VII and III closer to each other [18]. (b) View of the cytoplasmic ends of TM I–VII of the M3 mAChR. Published disulﬁde cross-linking data suggest that M3 mAChR activation involves a conformational change that moves the cytoplasmic end of TM VII closer to that of TM I and that this movement is accompanied by a clockwise rotation of the cytoplasmic segment of TM VII [17]. Moreover, agonist activation of the M3 mAChR allows the cytoplasmic portion of TM VI to move closer to that of TM V [15]. This movement is predicted to be accompanied by a clockwise rotational movement of the cytoplasmic end of TM VI [15, 19], consistent with ﬁndings obtained with bovine rhodopsin [7, 9–11] and the β2 -adrenergic receptor [12]. (c) Cytoplasmic view of a selected region of the intracellular surface of the M3 mAChR. We recently showed [20] that muscarinic agonists increase the distance between Cys residues introduced at (i) positions 91 (1.56) and 549 (7.59) and (ii) positions 92 (1.57) and 550 (7.60) (these four residues are highlighted in yellow). In contrast, inverse muscarinic agonists are predicted to reduce the distance between these residues [20]. The agonist-induced structural changes are thought to expose previously inaccessible receptor residues or surfaces (e.g. on the cytoplasmic ends of TM III, VI and VII), ultimately resulting in productive receptor/G protein coupling. A colour reproduction of this ﬁgure can be viewed in the colour plate 1 towards the centre of the book.

154

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

one of the early conformational events that trigger subsequent structural changes on the intracellular receptor surface where G protein coupling is known to occur. The analysis of a large number of mutant M3 mAChRs containing Cys substitutions on the cytoplasmic side of the receptor led to the identiﬁcation of nine double Cys mutant receptors which displayed increased disulﬁde cross-linking in the presence of agonists (Figure 8.3). On the other hand, in two of the double Cys mutant receptors investigated, agonist administration impaired disulﬁde bond formation (A91C/T549C and F92C/F550C [20]; Figure 8.4c). On the basis of these cross-linking data and a 3D model of the M3 mAChR, agonist binding is predicted to cause the following structural changes on the intracellular surface of the M3 receptor (Figure 8.4b,c): 1 The cytoplasmic end of TM VI undergoes a rotational movement and moves closer to the corresponding segment of TM V [15, 19], a movement similar to that proposed for other class I GPCRs, including rhodopsin and the β2 -adrenergic receptor (reviewed in [11, 12]). 2 The cytoplasmic end of TM VII moves closer to the corresponding region of TM I, accompanied by a rotational movement of TM VII [17]. 3 The N-terminal portion of helix 8 moves away from the cytoplasmic end of TM I [20]. Interestingly, in the same study [20], we demonstrated that inverse muscarinic agonists increase the proximity between helix 8 and the cytoplasmic end of TM I. These observations provide a structural basis for the opposing biological effects of muscarinic agonists and inverse agonists. Most likely, the agonist-induced conformational changes outlined above uncover previously inaccessible receptor surfaces or residues that are able to recognize G proteins with high afﬁnity and selectivity, ultimately triggering G protein activation.

8.2.4 Cell culture, transfection and membrane preparation All disulﬁde cross-linking studies were carried out with membranes prepared from COS-7 cells transiently expressing the different receptor constructs. Below are the protocols (Protocols 8.1 and 8.2) that we routinely use to culture and transfect COS-7 cells and to prepare membranes from transfected cells.

PROTOCOL 8.1 Cell Culture and Transfection Equipment and Reagents • Tissue culture hood and incubator • Inverted microscope • 175 cm2 ﬂask (Corning) • Atropine sulfate (Sigma)

8.2 METHODS AND APPROACHES

• 100 mm tissue culture dish (Falcon) • Lipofectamine and Plus reagent (Invitrogen) • 1 × Dulbecco’s phosphate-buffered saline (DPBS) with or without Ca2+ and Mg2+ (GIBCO) • 1 × 0.05% trypsin–ethylenediaminetetraacetate (EDTA) (GIBCO) • Dulbecco’s modiﬁed Eagle’s medium (DMEM; GIBCO) • Foetal bovine serum (FBS; GIBCO) • 100 × penicillin–streptomycin–glutamine (GIBCO) • 1 × complete medium (50 ml FBS, 5 ml 100 × penicillin–streptomycin–glutamine, and 445 ml DMEM) • 2 × complete medium (100 ml FBS, 10 ml 100 × penicillin–streptomycin–glutamine, and 390 ml DMEM) • 10 mM atropine stock solution in sterile distilled water (sdH2 O).

Method 1 Culture and maintain COS-7 cells in a 175 cm2 ﬂask containing ∼30 ml 1 × complete medium at 37 ◦ C in a humidiﬁed 5% CO2 incubator. 2 About 24 h prior to transfection, seed COS-7 cells into 100 mm dishes at a density of ∼(1–1.5) × 106 cells/dish.a 3 Immediately prior to transfections, examine the cells under an inverted microscope to conﬁrm that the cells look healthy and are ∼60–80% conﬂuent. 4 Mix 4 µg of receptor plasmid DNA with 750 µl DMEM and 30 µl of lipofectamine. Keep the mixture at room temperature (RT) for 15 min. 5 Premix 20 µl plus reagent with 750 µl DMEM. Add 770 µl of this solution to the DNA–lipofectamine mixture (step 4). Mix well by vortexing and keep the solution at RT for 15 min. 6 Immediately prior to transfections, wash the cells once with DPBS (10 ml per dish)b and then add 3.5 ml DMEM per dish. 7 Add 1.5 ml of the DNA–lipofectamine mixture from step 5 to each dish. Incubate at 37 ◦ C for 3 h in a humidiﬁed CO2 incubator. 8 Add 5 ml of 2 × complete medium and return the dishes to the 37 ◦ C CO2 incubator. Let the cells grow for ∼48 h and then prepare membranes as described in Protocol 8.2. 9 To increase the expression levels of Cys-substituted mutant M3 mAChRs, we routinely add atropine (ﬁnal concentration 1 µM) to the culture medium 24 h prior to harvesting of the cells.c Dilute a 10 mm atropine stock solution 1 : 100 with 1 × complete medium. Add 100 µl of this solution to each dish (this will result in a ﬁnal atropine concentration of 1 µM in the medium). The diluted atropine solution is stable for about 2 weeks when stored at 4 ◦ C.

155

156

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

Notes conﬂuent 175 cm2 tissue culture ﬂask usually yields enough COS-7 cells to prepare ∼12 dishes (100 mm) for transfections. To detach the cells from the bottom of the ﬂask, add 2 ml of 0.05% trypsin–EDTA and incubate for ∼3–5 min at 37 ◦ C. aA

b Use

DPBS containing Ca2+ and Mg2+ because these ions will help keep the cells attached to the dishes. Use DPBS without Ca2+ and Mg2+ when washing cells prior to adding trypsin.

c Several other studies have shown that treatment of transfected cells with antagonist ligands can lead to signiﬁcant increases in the expression levels of certain mutant GPCRs [34, 35].

PROTOCOL 8.2 Membrane Preparation Equipment and Reagents • 14 ml polypropylene (PP) Falcon tube (Falcon) • 25 cm cell scraper (Sarstedt) • Cell homogenizer (Polytron) • Sorvall centrifuge (RC 50 Plus) • Ice-cold phosphate-buffered saline (PBS; pH 7.4) • Ice-cold buffer A (25 mM sodium phosphate, 5 mM MgCl2 , pH 7.4).

Method 1 Remove the 100 mm dishes containing transfected COS-7 cells from the incubator ∼48 h after transfection and aspirate the medium. 2 Wash the cells twice with 10 ml of ice-cold PBS (10 min each wash) to ensure the complete removal of atropine (see Protocol 8.1, step 9). 3 Add 1 ml of ice-cold buffer A to each dish and incubate at 4 ◦ C for 15 min. 4 Scrape off the cells with a cell scraper and transfer them to a chilled 14 ml PP Falcon tubea . Wash the dish with 1 ml of ice-cold buffer A to maximize the yield of transferred cell material. 5 Homogenize the cells using a Polytron homogenizer (setting 5) for 20 s. Place the tubes on ice. 6 Centrifuge the samples at 18 000g (12 000 rpm) for 10 min at 4 ◦ C in a Sorvall centrifuge using an SM-24 rotor.b Discard the supernatants. 7 Resuspend the pellets in 2 ml of ice-cold buffer A and rehomogenize (see step 6). 8 Carry out cross-linking experiments, or snap-freeze the membrane homogenates on dry ice for 5–10 min and store at −70 ◦ C until needed.

8.2 METHODS AND APPROACHES

157

Notes a Cells b

scraped off from different plates transfected with the same plasmid can be pooled.

Note that speeds above this setting may result in damage to the tubes.

8.2.5 Urea treatment of membrane preparations prior to disulﬁde cross-linking To exclude the possibility that the ability of a speciﬁc double Cys mutant receptor to form an intramolecular disulﬁde bond is caused indirectly by ligand-induced dissociation of precoupled receptor/G protein complexes, it is recommended to treat receptor-containing membranes with a high concentration (5 m) of the chaotropic agent urea prior to disulﬁde cross-linking (Protocol 8.3; also see [15, 19, 20]). Previous studies have shown that this strategy leads to the almost complete inactivation or removal of heterotrimeric G proteins while leaving uncoupled receptors fully functional [36, 37].

PROTOCOL 8.3

Urea Treatment of Receptor-containing Membranes

Equipment and Reagents • Urea (Sigma) • Buffer A (25 mM sodium phosphate, 5 mM MgCl2 , pH 7.4) • Refrigerated centrifuge (Eppendorf 5417 R).

Method 1 Prepare a 5 M urea solution by dissolving the proper amount of urea in buffer A. 2 Thaw frozen membranes prepared from transfected COS-7 cells at RT. Resuspend membranes harvested from one 100 mm dish in 1 ml of buffer A containing 5 M urea (include samples that do no contain urea as controls). Incubate samples on ice for 30 min. 3 Centrifuge at 8000 g for 30 min at 4 ◦ C. Discard the supernatant. 4 Resuspend the pellet in 1 ml of buffer A, followed by a 30 min incubation on ice. 5 Centrifuge at 8000 g for 10 min at 4 ◦ C. Discard the supernatant. 6 Resuspend the pellet in 1 ml of buffer A. Centrifuge at 8000 g for 10 min at 4 ◦ C. Discard the supernatant. Samples can then be subjected to disulﬁde cross-linking (see Section 8.2.6).

158

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

8.2.6 Disulﬁde cross-linking and detection of disulﬁde bonds via factor Xa digestion of solubilized receptors and western blotting To promote the formation of disulﬁde bonds, receptors contained in membranes prepared from receptor-expressing COS-7 cells are subjected to oxidizing conditions (Protocol 8.4). Subsequently, receptor proteins are solubilized and digested to completion with factor Xa (Protocol 8.5), followed by western blotting studies in order to detect the formation of disulﬁde bonds (Protocol 8.6; also see Figure 8.2).

PROTOCOL 8.4 Disulﬁde Cross-linking Equipment and Reagents • Rack rotator (RKVS; Appropriate Technical Resources, Inc.) • CuSO4 (Sigma) • 1,10-Phenanthroline (Sigma) • N-Ethylmaleimide (NEM; Sigma) • 5 mM Cu-Phen stock solution in sdH2 O (mix 4 ml of 5 mM CuSO4 with 4 ml of 15 mM 1,10-phenanthroline)a,b • Stock solutions of muscarinic agonists (e.g. carbachol) and muscarinic antagonists/inverse agonists (e.g. atropine) in sdH2 O • 250 mM NEM stock solution in sdH2 O • 500 mM EDTA stock solution in sdH2 O (pH 8.0; Quality Biological Inc.).

Method 1 Thaw frozen membranes harvested from three dishes at RT, homogenize as described in Protocol 8.2 (step 5) and adjust the volume to 3 ml with buffer A. Aliquot 880 µl into three 1.5 ml microcentrifuge tubes. 2 Add 20 µl of the proper stock solutions of muscarinic ligands or sdH2 O (as a control) to the 880 µl membrane aliquots.c Mix by inverting the tubes several times and keep the samples at RT for 10 min. 3 Add 100 µl of sdH2 O (negative control)d or 100 µl of a Cu-Phen stock solution to achieve the desired ﬁnal Cu-Phen concentration (2.5–100 µM).e Incubate for 10 min at RT with end-over-end rotation (use an Appropriate Technical Resources rack rotator). 4 Stop the reaction by adding 20 µl of 500 mM EDTA (pH 8.0) (ﬁnal concentration ∼10 mM) and 40 µl of 250 mM NEM stock solution (ﬁnal concentration ∼10 mM). Mix by inverting the tubes and incubate the samples on ice for 10 min. Notes a Throughout

this chapter, the concentrations indicated for Cu-Phen refer to molar copper concentrations.

8.2 METHODS AND APPROACHES

159

b To

accelerate the dissolution of 1,10-phenanthroline in sdH2 O, brief heat pulses (5 s at a time) can be applied in a microwave oven.

c The types of ligand and the ﬁnal concentrations to be tested depend on the speciﬁc aims of a particular study. d To

reduce the number of samples to be processed, the sdH2 O control may be omitted, as long as the M3 (3C)-Xa background receptor (template for Cys substitution mutagenesis) is included in each experiment. Cu-Phen concentrations (25–100 µM) appear to be required in order to cross-link vicinal Cys residues that are not easily accessible on the receptor surface. As discussed in more detail in Section 8.3, it is advisable to use oxidizing agents that are smaller than Cu-Phen, such as molecular iodine, when the Cys residues to be cross-linked are located within the ligand binding crevice or are buried more deeply in the TM receptor core (e.g. see [15, 18]). e Higher

PROTOCOL 8.5

Receptor Solubilization and Digestion

Equipment and Reagents • Refrigerated centrifuge (Eppendorf 5417 R) • Rack rotator (RKVS; Appropriate Technical Resources, Inc.) • Digitonin (Sigma) • Micro BCA protein assay kit (Pierce) • Factor Xa protease (Roche) • Factor Xa buffer (50 mM tris(hydroxymethyl)aminomethane (tris)-HCl, 100 mM NaCl, 1 mM CaCl2 , pH 8.0) • Protease inhibitor cocktail (Sigma)a • Dithiothreitol (DTT; Sigma) • 24% digitonin stock solution in dimethyl sulfoxide (DMSO) • 2% digitonin in PBS (dilute the proper amount of 24% digitonin stock solution in PBS, pH 7.4; store on ice) • 2% digitonin in factor Xa buffer (dilute the proper amount of 24% digitonin stock solution in factor Xa buffer; store on ice) • 5 × loading buffer (62.5 mM tris base (pH 6.8), 2% sodium dodecyl sulfate (SDS), 0.02% bromphenol blue, 10% glycerol; Quality Biological Inc.).

Method 1 Centrifuge the cross-linked membrane samples (see Protocol 8.4, step 5) at 8000 g for 10 min at 4 ◦ C. Discard the supernatant.

160

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

2 Resuspend the pellet in 250 µl of ice-cold 0.2% digitonin solution (in PBS), and incubate on ice for 5 min.b 3 Centrifuge at 8000 g for 10 min at 4 ◦ C, discard the supernatant, and resuspend the pellet in 100 µl of 1.2% digitonin solution (in factor Xa buffer). Rotate the samples gently at 4 ◦ C (20 rpm) with end-over-end rotation for at least 2 h.c 4 Centrifuge at 14 000 rpm (20 000g) for 30 min at 4 ◦ C. 5 Transfer supernatants into new tubes and store samples at −70 ◦ C until needed. 6 Measure protein concentrations in the supernatants using the Micro BCA protein assay kit according to the manufacturer’s instructions. 7 Reconstitute 100 µg of factor Xa protease in 110 µl of factor Xa buffer and store this solution at 4 ◦ C. The solution is stable for at least 2 weeks. 8 Mix 23 µl of membrane extract (∼15 µg protein) with 2 µl of factor Xa solution (total reaction volume 25 µl). If needed, use factor Xa buffer to dilute membrane extracts. Mix and incubate at RT for 16–20 h. 9 Stop the protease reaction by adding 1 µl of protease inhibitor cocktail and 6 µl of 5 × loading buffer, followed by a 30 min incubation at 37 ◦ C.d The samples are then used directly for SDS–polyacrylamide gel electrophoresis and western blotting or frozen at −70 ◦ C until needed. Notes a

Protease inhibitor cocktail (supplied in DMSO) should be aliquoted in a small volume and stored at −20 ◦ C. Stocks should be thawed and frozen only a few times, since frequent thawing may result in reduced inhibitor activity.

b During

this step, peripheral membrane proteins, but not TM proteins (like the M3 mAChR), will be solubilized. c Samples can also be solubilized overnight. Most importantly, all samples need to be treated in an identical fashion. d When necessary, add DTT to the samples (ﬁnal concentration 50 mM) at the beginning of the 30 min incubation (37 ◦ C) in order to generate reducing conditions needed to break up disulﬁde bonds. Do not boil samples, since this will lead to nonspeciﬁc aggregation of receptor proteins.

PROTOCOL 8.6 Western Blotting Analysis Equipment and Reagents • XCell SureLock Novex mini-cell electrophoresis system (Invitrogen) • Bio-Rad Mini blotting apparatus (Bio-Rad) • Orbital shaker (model 3520; Lab-line Instruments Inc.) • Precast 10–20% tris–glycine polyacrylamide gel (Invitrogen)

8.2 METHODS AND APPROACHES

• 10 × tris–glycine SDS running buffer (Invitrogen) • 25 × tris–glycine transfer buffer (Invitrogen) • Methanol (Mallinckrodt) • Hybond-ECL nitrocellulose membranes (Amersham Biosciences) • PBS containing 0.05% Tween 20 (PBS-T) • 1 × running buffer (mix 100 ml of 10 × tris–glycine SDS running buffer with 900 ml dsH2 O) • 1 × transfer buffer (mix 40 ml of 25 × tris–glycine transfer buffer, 200 ml methanol, and 760 ml dsH2 O) • Blocking buffer (PBS-T containing 5% non-skim milk) • Donkey anti-rabbit immunoglobulin G conjugated with horseradish peroxidase (Amersham Biosciences) • SuperSignal West Pico chemiluminescent substrate (Pierce) • Hyperﬁlm ECL (Kodak).

Method 1 Assemble the precast 10–20% tris–glycine polyacrylamide gel in an XCell SureLock electrophoresis cell and add 1 × running buffer to the gel chamber. Load ∼10 µl of the different receptor-containing membrane lysates onto the gel (∼5 µg protein/well). 2 Run the gel at 125 V for ∼2.5 h in 1 × tris–glycine SDS running buffer. 3 Transfer proteins to a nitrocellulose membrane. Carry out the transfer using the Bio-Rad Mini blotting apparatus (100 V for 1 h at 4 ◦ C), according to the manufacturer’s instructions. 4 After the transfer, wash the blot brieﬂy with PBS-T. Subsequently, incubate the blot on an orbital shaker (800 rpm) in 10 ml of blocking buffer for 30 min. 5 Add the anti-C-M3 rabbit polyclonal antibody (see Figure 8.1; [30]) to the blocking buffer (1 : 5000 dilution) and incubate for 1 h at RT on an orbital shaker (800 rpm). 6 Wash with 20 ml of 1 × PBS-T for three times (10 min each wash) on an orbital shaker (800 rpm). 7 Add 10 ml fresh blocking buffer containing an anti-rabbit secondary antibody conjugated with horseradish peroxidase (1 : 2000 dilution) and incubate at RT for 1 h on an orbital shaker (800 rpm). 8 Wash with 20 ml of 1 × PBS-T for three times (10 min each time) on an orbital shaker (800 rpm). 9 Develop the blot with SuperSignal West Pico chemiluminescent substrate at RT for 5 min under gentle shaking (use equal volumes of peroxide and enhancer solutions; 8–10 ml total volume per blot) and expose to Hyperﬁlm ECL.a,b

161

162

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

Notes a The

blot can be stripped (as long as it is still moist) for subsequent probing with a different antibody. For example, the use of a β-actin antibody is recommended to ensure that similar amounts of proteins were loaded in the different lanes.

b The

intensities of immunoreactive bands can be quantitated by scanning densitometry using the program NIH ImageJ (or an equivalent image analysis program). This software, including the necessary instructions for users, can be downloaded from the National Institutes of Health (NIH) web site (http://rsb.info.nih.gov/ij/).

8.3 General considerations, caveats and troubleshooting • Vector-transfected COS-7 cells should be included in each experiment to rule out any nonspeciﬁc signals. Moreover, it is essential to include the M3 (3C)-Xa background receptor as a control in all cross-linking experiments. Most importantly, the absence of a full-length receptor band in the M3 (3C)-Xa-containing samples following factor Xa treatment conﬁrms that factor Xa digestion was complete. In order to interpret negative data properly (lack of disulﬁde cross-linking), it is also advisable to include receptor-containing samples that were subjected to all steps of the cross-linking protocol outlined above, except for the factor Xa digestion step. Usually, western blotting will show that these samples contain detectable amounts of full-length receptor protein, conﬁrming that the lack of a disulﬁde cross-linking signal following factor Xa digestion is not caused by very low amounts of receptor protein. • In most of our previous studies, we used Cu-Phen as the oxidizing agent to induce the formation of disulﬁde bonds in receptor-containing membrane preparations. Although the mechanism by which this agent promotes the formation of disulﬁde bonds is not well understood, it is thought to involve the transitory reduction of Cu2+ to Cu+ [38]. When used at high concentrations, Cu-Phen may induce the formation of disulﬁde cross-links of proteins in nonnative conformations [8] (FY Zeng and J Wess, unpublished observations). If possible, therefore, we recommend the use of relatively low concentrations of Cu-Phen (2.5–100 µm). • One caveat inherent in all disulﬁde cross-linking approaches is that negative results are difﬁcult to interpret. It is known that the rate of disulﬁde bond formation is not only determined by the distance between the two Cys residues under investigation, but also by their relative orientation and the environment surrounding the Cys residues, which can strongly affect the pK a values of the SH groups [28, 39]. It is possible, therefore, that two Cys residues, despite being close to each other in ˚ do not the 3D structure of the receptor (distance between α-carbon atoms <7 A), readily form a disulﬁde bond, even under strong oxidizing conditions. • It has also been observed that two Cys residues whose α-carbon atoms are predicted ˚ can be to be relatively far apart in the 3D structure of the receptor (e.g. >7 A)

8.3 GENERAL CONSIDERATIONS, CAVEATS AND TROUBLESHOOTING

163

linked by a disulﬁde bridge [8, 9]. Such cross-links can form when the Cys residues are contained in receptor regions endowed with a high degree of conformational ﬂexibility. Consistent with this notion, disulﬁde cross-linking studies carried out with Cys-substituted versions of a bacterial chemoreceptor of known structure indicate that disulﬁde bonds can readily form when the two Cα carbon atoms are less than ˚ apart [40]. Structural and biophysical evidence indicates that the intracellular ∼12 A surface of GPCRs, including the C-terminal tail and the i3 loop, are characterized by a high degree of structural ﬂexibility [11, 12, 21–23, 41]. Given the caveats outlined in the last two bullet points, extreme caution should be exercised in interpreting cross-linking data obtained with a small number of double Cys mutant receptors. It is strongly advisable to perform Cys scans in which consecutive residues in two receptor regions of interest are replaced, one by one, with Cys. Oxidative cross-linking analysis of the resulting collection of double Cys mutant receptors is likely to provide a more comprehensive picture of the dynamic changes associated with receptor activation. • Under certain experimental conditions, it may be advantageous to replace Cu-Phen with smaller oxidizing agents, such as molecular iodine. In one of our previous studies [16], we observed that muscarinic ligands prevented the formation of a Cu-Phen-catalysed disulﬁde bond between a Cys residue introduced into TM I at position L77 (1.42) and C532 in TM VII (7.42; Figure 8.3). We speculated that this phenomenon was caused by competition of muscarinic ligands with the relatively bulky Cu-Phen moiety for access to the hydrophilic binding crevice. Although the mechanism by which the Cu-Phen complex catalyses disulﬁde bond formation is not well understood, previous work [38] suggests that Cu-Phen-catalysed disulﬁde bond formation does not require the generation of hydrogen peroxide as an oxidizing intermediate. This observation is in agreement with the concept (but does not prove it) that the Cu-Phen complex must be present in the direct vicinity of the SH groups to be cross-linked. Strikingly, when we used molecular iodine (which has a molecular ˚ instead of Cu-Phen as the oxidizing agent, the efﬁciency of disulﬁde size of <2.7 A) cross-linking observed between L77C and C532 was not signiﬁcantly affected by the presence of muscarinic ligands [16]. Therefore, we recommend the use of molecular iodine or other small oxidizing agents in cross-linking studies involving Cys residues present in the vicinity of the ligand binding crevice. Mercury(II) salts (HgCl2 ) which can bridge vicinal pairs of Cys residues have also been used successfully to achieve cross-linking between Cys residues buried in the TM receptor core [42, 43]. • Disulﬁde cross-linking can occur by trapping rather transient receptor conformations (for example, conformations that are intermediate between the ground state and the fully active state of the receptor) in which the two Cys residues are temporarily close to each other in the 3D structure of the receptor. In contrast, most biophysical approaches, including site-directed spin labelling techniques and studies involving the use of ﬂuorescent reporter molecules, provide information only about average receptor conformations. The two different techniques, therefore, yield results that are complementary in nature.

164

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

• Like other GPCRs [44], the M3 mAChR is known to form dimers or oligomers [45, 46]. It is important, therefore, to conﬁrm that the disulﬁde cross-link observed with a speciﬁc double Cys mutant receptor is due to the formation of an intramolecular rather than an intermolecular disulﬁde bond. One experimental approach that we have repeatedly used to address this issue is to co-express two different mutant receptors that contain the two Cys substitutions individually [15, 17, 19, 20]. If the disulﬁde cross-link observed with the double Cys mutant receptor is due to the formation of an intramolecular disulﬁde bridge, then no cross-linking signal will be observed following co-transfection of the two single Cys mutant receptors [15, 17, 19, 20].

Acknowledgements The studies performed in the laboratory of J. W. were supported by the Intramural Research Program of the NIH, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). We thank Dr Soo-Kyung Kim and Dr Kenneth A. Jacobson (NIH, NIDDK) for carrying out the M3 mAChR modelling studies.

References 1. Bockaert, J. and Pin, J.P. (1999) Molecular tinkering of G protein-coupled receptors: an evolutionary success. EMBO J., 18, 1723–1729. 2. Gether, U. (2000) Uncovering molecular mechanisms involved in activation of G protein-coupled receptors. Endocr. Rev., 21, 90–113. 3. Kristiansen, K. (2004) Molecular mechanisms of ligand binding, signaling, and regulation within the superfamily of G-protein-coupled receptors: molecular modeling and mutagenesis approaches to receptor structure and function. Pharmacol. Ther., 103, 21–80. 4. Foord, S.M., Bonner, T.I., Neubig, R.R. et al. (2005) International Union of Pharmacology. XLVI. G protein-coupled receptor list. Pharmacol. Rev., 57, 279–288. 5. Fredriksson, R. and Schioth, H.B. (2005) The repertoire of G-protein-coupled receptors in fully sequenced genomes. Mol. Pharmacol., 67, 1414–1425. 6. Wess, J. (1998) Molecular basis of receptor/G-protein-coupling selectivity. Pharmacol. Ther., 80, 231–264. 7. Farrens, D.L., Altenbach, C., Yang, K. et al. (1996) Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science, 274, 768–770. 8. Struthers, M. and Oprian, D.D. (2000) Mapping tertiary contacts between amino acid residues within rhodopsin. Methods Enzymol., 315, 130–143. A review of a disulﬁde cross-linking approach useful to study the structural and dynamic properties of rhodopsin. 9. Meng, E.C. and Bourne, H.R. (2001) Receptor activation: what does the rhodopsin structure tell us? Trends Pharmacol. Sci., 22, 587–593.

REFERENCES

165

10. Sakmar, T.P., Menon, S.T., Marin, E.P. and Awad, E.S. (2002) Rhodopsin: insights from recent structural studies. Annu. Rev. Biophys. Biomol. Struct., 31, 443–484. 11. Hubbell, W.L., Altenbach, C., Hubbell, C.M. and Khorana, H.G. (2003) Rhodopsin structure, dynamics, and activation: a perspective from crystallography, site-directed spin labeling, sulfhydryl reactivity, and disulﬁde cross-linking. Adv. Protein Chem., 63, 243–290. A comprehensive review of biophysical and biochemical techniques (including disulﬁde cross-linking approaches) to study rhodopsin structure, dynamics and activation. 12. Kobilka, B.K. (2006) G protein coupled receptor structure and activation. Biochim. Biophys. Acta, 768, 794–807. 13. Hulme, E.C., Birdsall, N.J. and Buckley, N.J. (1990) Muscarinic receptor subtypes. Annu. Rev. Pharmacol. Toxicol., 30, 633–673. 14. Wess, J. (1996) Molecular biology of muscarinic acetylcholine receptors. Crit. Rev. Neurobiol., 10, 69–99. 15. Ward, S.D., Hamdan, F.F., Bloodworth, L.M. et al. (2002) Conformational changes that occur during M3 muscarinic acetylcholine receptor activation probed by the use of an in situ disulﬁde cross-linking strategy. J. Biol. Chem., 277, 2247–2257. The original publication describing the in situ disulﬁde cross-linking strategy discussed in this chapter. 16. Hamdan, F.F., Ward, S.D., Siddiqui, N.A. et al. (2002) Use of an in situ disulﬁde cross-linking strategy to map proximities between amino acid residues in transmembrane domains I and VII of the M3 muscarinic acetylcholine receptor. Biochemistry, 41, 7647–7658. Reports that muscarinic ligands can prevent Cu-Phen-mediated cross-link formation involving Cys residues located close to the ligand binding pocket of the M3 mAChR. 17. Han, S.J., Hamdan, F.F., Kim, S.K. et al. (2005) Pronounced conformational changes following agonist activation of the M3 muscarinic acetylcholine receptor. J. Biol. Chem., 280, 24870–24879. 18. Han, S.J., Hamdan, F.F., Kim, S.K. et al. (2005) Identiﬁcation of an agonist-induced conformational change occurring adjacent to the ligand-binding pocket of the M3 muscarinic acetylcholine receptor. J. Biol. Chem., 280, 34849–34858. 19. Ward, S.D., Hamdan, F.F., Bloodworth, L.M. and Wess, J. (2006) Use of an in situ disulﬁde cross-linking strategy to study the dynamic properties of the cytoplasmic end of transmembrane domain VI of the M3 muscarinic acetylcholine receptor. Biochemistry, 45, 676–685. 20. Li, J.H., Han, S.J., Hamdan, F.F. et al. (2007) Distinct structural changes in a G-protein-coupled receptor caused by different classes of agonist ligands. J. Biol. Chem., 282, 26284–26293. Reports that full and inverse muscarinic agonists can have opposite effects on disulﬁde cross-link formation in selected double Cys mutant M3 mAChRs. 21. Palczewski, K., Kumasaka, T., Hori, T. et al. (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science, 289, 739–745. 22. Cherezov, V., Rosenbaum, D.M., Hanson, M.A. et al. (2007) High-resolution crystal structure of an engineered human β2 -adrenergic G protein-coupled receptor. Science, 318, 1258–1265. 23. Rosenbaum, D.M., Cherezov, V., Hanson, M.A. et al. (2007) GPCR engineering yields high-resolution structural insights into β2 -adrenergic receptor function. Science, 318, 1266–1273. 24. Falke, J.J. and Koshland, D.E. Jr (1987) Global ﬂexibility in a sensory receptor: a site-directed cross-linking approach. Science, 237, 1596–1600.

166

CH 8 A DISULFIDE CROSS-LINKING STRATEGY

25. Lynch, B.A. and Koshland, D.E. Jr (1991) Disulﬁde cross-linking studies of the transmembrane regions of the aspartate sensory receptor of Escherichia coli . Proc. Natl. Acad. Sci. U. S. A., 88, 10402–10406. 26. Pakula, A.A. and Simon, M.I. (1992) Determination of transmembrane protein structure by disulﬁde cross-linking: the Escherichia coli Tar receptor. Proc. Natl. Acad. Sci. U. S. A., 89, 4144–4148. 27. Lee, G.F., Burrows, G.G., Lebert, M.R. et al. (1994) Deducing the organization of a transmembrane domain by disulﬁde cross-linking. The bacterial chemoreceptor Trg. J. Biol. Chem., 269, 29920–29927. 28. Kosen P.A. (1992) Disulﬁde bonds in proteins, in Stability of Protein Pharmaceuticals, Part A: Chemical and Physical Pathways of Protein Degradation (eds T.J. Ahern and M.C. Manning), Plenum Press, New York, pp. 31–67. 29. Klein-Seetharaman, J., Hwa, J., Cai, K. et al. (2001) Probing the dark state tertiary structure in the cytoplasmic domain of rhodopsin: proximities between amino acids deduced from spontaneous disulﬁde bond formation between Cys316 and engineered cysteines in cytoplasmic loop 1. Biochemistry, 40, 12472–12478. 30. Zeng, F.Y., Soldner, A., Sch¨oneberg, T. and Wess, J. (1999) Conserved extracellular cysteine pair in the M3 muscarinic acetylcholine receptor is essential for proper receptor cell surface localization but not for G protein coupling. J. Neurochem., 72, 2404–2414. 31. Bonner, T.I., Buckley, N.J., Young, A.C. and Brann, M.R. (1987) Identiﬁcation of a family of muscarinic acetylcholine receptor genes. Science, 237, 527–532. 32. Zeng, F.Y., Hopp, A., S¨oldner, A. and Wess, J. (1999) Use of a disulﬁde cross-linking strategy to study muscarinic receptor structure and mechanisms of activation. J. Biol. Chem., 274, 16629–16640. 33. Ballesteros, J.A. and Weinstein, H. (1995) Integrated methods for construction three-dimensional models and computational probing of structure–function relations in G protein-coupled receptors. Methods Neurosci., 25, 366–428. 34. Lu, Z.-L. and Hulme, E.C. (1999) The functional topography of transmembrane domain 3 of the M1 muscarinic acetylcholine receptor, revealed by scanning mutagenesis. J. Biol. Chem., 274, 7309–7315. 35. Morello, J.-P., Salahpour, A., Laperriere, A. et al (2000) Pharmacological chaperones rescue cell-surface expression and function of misfolded V2 vasopressin receptor mutants. J. Clin. Invest., 105, 887–895. 36. Hartman, J.L. and Northup, J.K. IV (1996) Functional reconstitution in situ of 5-hydroxytryptamine2c (5HT2c ) receptors with αq and inverse agonism of 5HT2c receptor antagonists. J. Biol. Chem., 271, 22591–22597. 37. Lim, W.K. and Neubig, R.R. (2001) Selective inactivation of guanine-nucleotide-binding regulatory protein (G-protein) α and βγ subunits by urea. Biochem. J., 354, 337–344. 38. Kobashi, K. (1968) Catalytic oxidation of sulfhydryl groups by o-phenanthroline copper complex. Biochim. Biophys. Acta, 158, 239–245. 39. Shaked, Z., Szajewski, R.P. and Whitesides, G.M. (1980) Rates of thiol–disulﬁde interchange reactions involving proteins and kinetic measurements of thiol pKa values. Biochemistry, 19, 4156–4166.

REFERENCES

167

40. Careaga, C.L. and Falke, J.J. (1992) Thermal motions of surface alpha-helices in the d-galactose chemosensory receptor. Detection by disulﬁde trapping. J. Mol. Biol., 226, 1219–1235. 41. Kobilka, B.K. and Deupi, X. (2007) Conformational complexity of G-protein-coupled receptors. Trends Pharmacol. Sci., 28, 397–406. 42. Soskine, M., Steiner-Mordoch, S. and Schuldiner, S. (2002) Crosslinking of membrane-embedded cysteines reveals contact points in the EmrE oligomer. Proc. Natl. Acad. Sci. U. S. A., 99, 12043–12048. Reports the use of oxidative disulﬁde cross-linking techniques to map the structural and dynamic properties of the D2 dopamine receptor dimer interface. 43. Guo, W., Shi, L., Filizola, M. et al. (2005) Crosstalk in G protein-coupled receptors: changes at the transmembrane homodimer interface determine activation. Proc. Natl. Acad. Sci. U. S. A., 102, 17495–17500. 44. Angers, S., Salahpour, A. and Bouvier, M. (2002) Dimerization: an emerging concept for G protein-coupled receptor ontogeny and function. Annu. Rev. Pharmacol. Toxicol., 42, 409–435. 45. Zeng, F.Y. and Wess, J. (1999) Identiﬁcation and molecular characterization of m3 muscarinic receptor dimers. J. Biol. Chem., 274, 19487–19497. 46. Goin, J.C. and Nathanson, N.M. (2006) Quantitative analysis of homo- and heterodimerization of muscarinic acetylcholine receptors in live cells: regulation of receptor down-regulation by heterodimerization. J. Biol. Chem., 281, 5416–5425.

9 Use of Fluorescence Correlation Spectroscopy to Study the Diffusion of G Protein-coupled Receptors Stephen J. Briddon, Jonathan A. Hern and Stephen J. Hill Institute of Cell Signalling, School of Biomedical Sciences, University of Nottingham, Nottingham, UK

9.1 Introduction There is an increasing need for information about the way in which G protein-coupled receptors (GPCRs) and their associated signalling proteins are spatially organized within the cell [1, 2]. This has led to the application of new high-resolution imaging techniques to this area. In particular, the use of confocal imaging and associated methods, such as the ﬂuorescence and the bioluminescence resonance energy transfer techniques [3–6] and ﬂuorescence recovery after photobleaching (FRAP) [7–9], have provided useful information on the molecular organization of GPCRs. More recently, ﬂuorescence correlation spectroscopy (FCS) has been used to investigate the membrane mobility of both native and ligand-occupied GPCRs (reviewed in [10]). FCS is based on statistical analysis of ﬂuorescence ﬂuctuations obtained from a small deﬁned detection volume, which can be positioned precisely within the cell [11, 12]. Analysis of these ﬂuctuations using autocorrelation statistics yields quantitative information about both the diffusion coefﬁcient and concentration of the ﬂuorescent species of interest. As with other imaging techniques, FCS can be applied to individual living cells to study GPCRs in their native membrane environment [10]. The single-photon counting nature of the detection also means that FCS is a highly sensitive technique. Thus, even G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

170

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

though FCS is not a true ‘single-molecule’ technique (unlike single-particle tracking, e.g. [13, 14]), the average number of receptors present within the detection volume during a measurement is usually in the range 0.5–150. One particular advantage of FCS is its quantitative nature, directly providing dwell times (diffusion coefﬁcients) and particle numbers (concentration) of the measured species [12]. In addition, ﬂuctuations can be correlated over a wide temporal range (0.1 µs to 300 ms), allowing simultaneous detection of very fast moving species (e.g. ligands), as well as receptors themselves. For membrane receptors, FCS has been used in conjunction with directly labelled GPCRs (e.g. using ﬂuorescent proteins such as green ﬂuorescent protein (GFP)) to obtain diffusion coefﬁcients of receptors within small areas (∼0.1 µm2 ) of cell membrane [15–17]. FCS is also particularly well suited to the use of ﬂuorescent ligands to monitor speciﬁcally the diffusion of agonist- and antagonist-occupied receptors [18–29]. In such cases, a single measurement allows the determination of both free and bound ligand concentrations in similar small areas of membrane [30]. A range of FCS-based approaches have been developed (e.g. [31–38]), but for reasons of simplicity we will conﬁne ourselves to describing the use of single-point FCS using continuous-wave laser excitation. We describe the basic techniques required for the alignment and calibration of a simple FCS system. We provide protocols for obtaining diffusion coefﬁcients for labelled GPCRs using FCS and, in particular, how to use FCS in conjunction with ﬂuorescently labelled GPCR ligands to perform quantitative pharmacology in small areas of living cell membranes. Detailed methodological reviews by Kim et al. [39] and Bacia and Schwille [40] provide good additional sources of general technical advice for FCS measurements, and also cross-correlation FCS, which is not covered in this article.

9.2 Methods and approaches 9.2.1 Principles of ﬂuorescence correlation spectroscopy The principles underlying FCS were ﬁrst described in the 1970s [41]. However, it was not until nearly 20 years later that advances in optical and photon-counting technology enabled the application of this technique to living cells (e.g. [36, 42, 43]). Most current FCS equipment is based around a standard confocal microscope set-up [44]. A laser is focused through a high numerical aperture objective lens to a diffraction-limited spot. In the case of excitation with a continuous-wave laser, a pinhole placed in the confocal image plane creates a small detection volume (∼0.25 ﬂ) which is Gaussian–Lorentzian in shape. The FCS detection volume is held ﬁxed in the chosen measurement position. Fluorescent species diffusing through the volume become excited and emit photons. These are detected in a time-correlated way by single-photon counting detectors placed behind the pinholes. Over a period of time, this results in a time-dependent ﬂuctuation in the detected ﬂuorescent intensity (Figure 9.1a). Such ﬂuctuations can result from translational diffusion of the species within the volume, changes in ﬂuorescence intensity (e.g. through environmental changes) or from photophysical events within the ﬂuorophore. The statistical analysis of these ﬂuctuations by autocorrelation analysis allows quantiﬁcation of diffusion times and concentrations. In autocorrelation analysis,

171

9.2 METHODS AND APPROACHES (a) G(τ)=1+

t+δt

t

<δI(t ).δI(t + τ)> < I >2

Intensity

δt δI Time (b) G(τ)

Diffusion co-efficient

G(τ)

G(0)

τD

τ

1/N

Particle number

Dwell time

τ

Figure 9.1 Principles of autocorrelation analysis. (a) Example ﬂuorescence ﬂuctuations are shown, such as those collected during a typical FCS experiment. The general form of the autocorrelation function G(τ ) is also shown. The deviation in ﬂuorescence δI from the average ﬂuorescence intensity I at a time t is compared with the deviation δI(t + τ ) at a small time τ later. The product of these intensities is normalized to I 2 , for each value of t and for a range of τ , from 0.1 µs to 1 s, to give the full autocorrelation function G(τ ). (b) The autocorrelation function represented graphically with τ plotted on the abscissa and G(τ ) as the ordinate, indicating that the dwell time τD of a ﬂuorescent species in the volume can be determined from the midpoint of the decay of G(τ ). Increased dwell time (or decreased diffusion coefﬁcient) results in a right shift in G(τ ) and a corresponding increase in τD . (c) Similarly, the intercept on the ordinate axis G(0) is equal to the inverse of the particle number, such that an increase in the amplitude of the curve represents a decrease in the particle number.

the deviation in ﬂuorescence intensity δI from the average intensity I at a given time point t is compared with the deviation at a small time (t + τ ) later. The product of these deviations is normalized to the square of the average intensity. When this is performed for all t values in the read, and for a range of τ values, the autocorrelation function G(τ ) is described. For a single species diffusing with three-dimensional (3D) Brownian motion, this function describes a sigmoidal decay function. The midpoint of the decay (on the abscissa) represents the average dwell time τD of the species within the detection volume during the measurement. Thus, as the speed of diffusion slows, so the dwell time increases (i.e. the smaller the diffusion coefﬁcient) and the more right-shifted the curve becomes (Figure 9.1b). Likewise, the intercept on the ordinate is proportional to the reciprocal of the average number of particles 1/N in the volume during the measurement. Thus, the more concentrated the species, the higher the particle number and the lower the amplitude of the curve. The autocorrelation curve, therefore, is of greater amplitude and better deﬁned at lower concentrations, giving FCS a dynamic range of approximately 0.1–400 nm. Values for τD and N are

172

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

determined by nonlinear curve ﬁtting of the autocorrelation curve to an appropriate diffusion model. Importantly, autocorrelation functions for species with different dwell times are simply additive. Thus, if fast-moving (e.g. ligand) and slow-moving (e.g. ligand-occupied receptor) species coexist within the detection volume, then a biphasic autocorrelation function will be observed [10, 30]. The contribution of each component to the amplitude of the curve allows the number of particles of each component to be calculated. Differences in diffusion time of the order of 1.6-fold can be detected (this equates to an approximate sixfold difference in molecular weight) [45].

9.2.2 Equipment required for ﬂuorescence correlation spectroscopy One of the main disadvantages of live-cell FCS is the requirement for specialist equipment. Commercial equipment capable of performing FCS is available from a number of sources. Some microscope manufacturers offer optional FCS detection as an integral part of their confocal microscopes. Alternatively, a number of specialist manufacturers offer photon-counting technology and analysis software which can be added to existing confocal microscopes. It is perfectly possible to build a bespoke FCS system, and some considerations on design and components can be found in the review by Sengupta et al. [44]. Whilst this is cheaper and more ﬂexible, it also requires more technical expertise. In this article, we describe the use of a commercially available system from Carl Zeiss, the Confocor 2, which we have modiﬁed to perform live-cell work [17, 46]. The basic components of an FCS system are illustrated in Figure 9.2. The system itself should be placed in a room with good temperature stability (±2 ◦ C). For live-cell work, the system should be based around a research-grade inverted microscope, positioned on an antivibration table. Upright systems are possible, but require the use of dipping lenses. FCS components can be mounted on a microscope camera port, or can be ﬁbre coupled from the confocal detection unit. Excitation light is provided by a laser, which is collimated and directed, via an appropriate dichroic mirror, to the back aperture of a high numerical aperture objective lens. Choosing the appropriate collimation can lead to under- or over-ﬁlling of the back of the objective lens, which can affect the size of the detection volume used [47]. Generally, slight underﬁlling is used for single-photon excitation, as this causes less volume distortion [48]. The choice of objective is crucial to the creation of a small and correctly shaped detection volume. High numerical aperture apo-chromatic lenses provide a smaller diffraction-limited spot, whilst water-immersion lenses equipped with a correction collar create less distortion when using samples in aqueous media. Fluorescence emission light is usually collected in the epi-ﬂuorescent mode; that is, through the same objective lens. Emission light passes through the excitation dichroic and then through a further emission ﬁlter, chosen to limit detection to the emission of the chosen ﬂuorophore. Correct choice of emission ﬁlter is also important, as this can affect the amount of background/autoﬂuorescence signal which is detected, and also the amount of interference from photophysical effects of the ﬂuorophores. For continuous-wave laser excitation, a pinhole is located in between the emission ﬁlter and the photon detector. As with confocal optics, only photons which originate inside the desired detection volume pass through the pinhole. Varying the size of this pinhole will change the size of the

173

9.2 METHODS AND APPROACHES

Diffraction limited illumination spot Coverglass

Excitation light path Emission light path

Objective Lens

Collimating Lens or Beam expander

Dichroic Mirror

Laser

Emission Filter

Computer

Lens

Pinhole Correlator

Optical Fibre

Detector

Figure 9.2 The basic components of a single-channel FCS microscope using continuous-wave laser excitation are indicated. The excitation light path is indicated by the dotted line and the emission light path by the solid line. The individual components are described in the text.

174

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

detection volume, although not the volume illuminated, which is purely a function of the wavelength of light and the numerical aperture of the objective lens. Photons passing through the pinhole are detected by a single-photon counting detector (ﬁbre feed or separate lens). These are usually avalanche photodiodes, although photomultiplier tubes can also be used, though they tend not to be as sensitive at longer wavelengths. Detectors should have a high time resolution, short dead time and low dark count. The detectors feed, via a transistor–transistor logic pulse, to a hardware correlator, which records the time intervals between photons and uses multiple τ values to create a correlation curve. Appropriate capture software can be used to view the curves in real time on a normal mid- to high-speciﬁcation computer.

9.2.3 Routine calibration of a ﬂuorescence correlation spectroscopy microscope The proper alignment and calibration of an FCS system is crucial to achieving useful FCS data from cells, where the signal-to-noise ratio S : N may well be low. The purpose of calibration is twofold. First, to ensure that the beam path and pinholes are in the optimum position to create a correctly shaped detection volume. This is very important for accurate quantiﬁcation [47, 48]. Second, it allows us to calculate the dimensions of the detection volume to allow calculation of absolute ﬂuorophore concentrations (Protocol 9.1). Note that the size of the diffraction-limited detection volume changes with excitation wavelength. In addition, each combination of dichroic and emission ﬁlters may have a slightly different optimal pinhole position in x, y and z planes. Separate calibration should, therefore, be performed for each combination of beam path and wavelength used. Calibration is achieved using aqueous solutions of a bright ﬂuorophore with known diffusion coefﬁcient. Generally, rhodamine 6G (R6G) is useful, as it can be used for argon (458, 477, 488 and 514 nm) and green HeNe (543 nm) lasers. Other dyes, such as AlexaFluor488 or AlexaFluor546, can be used (e.g. for 561 nm illumination). For the red lasers (e.g. 633 nm HeNe and 647 nm KrAr), Cy5 or AlexaFluor633 are commonly used. A note should be made of the requirement for appropriate carriers for FCS measurements. Sample holders should contain a high-quality #1.0 or #1.5 coverglass which is scratch and distortion free [47]. It is also important for calibration purposes that ﬂuorescence-free high-performance liquid chromatography (HPLC)-grade water is used for dilutions, as this reduces background ﬂuorescence. Calibration is performed in two stages. Initially, a concentrated solution of ﬂuorophore is used to optimize the system alignment (Protocol 9.1, steps 1–10). The ﬁrst step is to position the measurement volume inside the calibration solution. This is achieved using the reﬂected laser light from the top surface of the coverslip to determine the focal position in the z plane. Once this reﬂection is found, the detection volume should be moved 200 µm upwards into the solution. The objective correction collar (if present), should then be optimized for the thickness of coverslip by adjusting the laser power to obtain a count rate of 250–500 kHz. This count rate is then maximized by adjusting the correction collar, ensuring the maximum molecular brightness value η is obtained. This η value normalizes the count rate to the number of molecules present (i.e. has units of counts per molecule and second, or kilohertz) and gives the

9.2 METHODS AND APPROACHES

175

best indication of a good S : N for the system. The pinhole position can now be optimized in the x and y planes, and ﬁnally in the z plane to obtain a maximum count rate or η value. Once the z position has been optimized, the x and y alignments are repeated. It may be necessary to adjust the collimator/beam expander position to ensure that the optimal z position is within the travel of the pinhole.

PROTOCOL 9.1

Alignment and Calibration of the FCS Microscope

Equipment and Reagents • FCS-capable microscope  • Water (high-purity, low-ﬂuorescence water for HPLC, e.g. Choromasolv ) • Lab-Tek eight-well chambered coverglass (No. 1 thickness, Nalge Nunc International) • Cy5 (10 mM aqueous solution)a or R6G (10 mM ethanolic solution).a

Method 1 Switch on the microscope system, lasers and air conditioning (if present) and allow 1–2 h for the system to equilibrate. 2 Add 10 µl of the required ﬂuorophore stock solution (10 mM) to 990 µl of water, mix and repeat this serial dilution to make a 1 µM solution. Add 20 µl of this to 980 µl of water to give a 20 nM solution. 3 Add 200 µl of each of the 1 µm and 20 nM solutions to separate chambers of a chambered coverglass. 4 Add a drop of water to the objective lens (as immersion ﬂuid), then mount the chambered coverglass on the microscope stage. 5 Position the chamber containing the 1 µM ﬂuorophore solution over the objective. 6 Adjust the focus of the microscope in order to position the focal position of the laser on the upper surface of the coverglass.b Move the focal position approximately 200 µm upwards into the solution. 7 Select the appropriate beampath (laser, main dichroic and emission ﬁlters) for the GPCR tag being used.c Expose the sample to laser light and adjust the laser power to obtain a count rate of 250–500 kHz. 8 Adjust the correction collar on the objective lens to maximize the count rate. 9 Check the pinhole diameter is set to 1 Airy unit. Adjust the pinhole in the x direction by moving its position to obtain the maximum count rate. Repeat this in the y and ﬁnally the z planes. Note the pinhole positions for future reference. 10 If a peak count rate is not obtained within the travel distance of the pinhole in the z plane, then the collimator should be adjusted and the alignment repeated. Once the z alignment is satisfactory, repeat the x and y alignments to conﬁrm their positions.

176

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

11 Move the chamber containing the 20 nM ﬂuorophore solution over the objective. Adjust the laser intensity to obtain a count rate of 20–100 kHz (η = 30 − 90 kHz).d 12 Start the data collection, collecting for 10 × 10 s. 13 Fit the data using a model for one-component, free 3D Brownian diffusion with a pre-exponential to allow for triplet state using the following equation: G(τ ) = 1 + AN−1 (1 + τ/τD )−1 [1 + τ/(S2 τD )]−0.5 where A N τD S Tτ τt

= = = = = =

1 + (Tτ e−τ/τt )(1 − Tτ )−1 particle number diffusion time structure parameter triplet fraction triplet relaxation time.

14 If this ﬁt yields 3 < S < 8, then the system is satisfactorily aligned. 15 From the diffusion time value τ D , calculate the radius of the detection volume ω0 , using ω0 = (4Dτ D )0.5 , where D is the literature value for the diffusion coefﬁcient of the calibration ﬂuorophore (D = 2.8 × 10−10 m2 s−1 for R6G and 3.1 × 10−10 m2 s−1 for Cy5 at 22 ◦ C). 16 Subsequently, the half-height of the volume is given by ω1 = Sω0 , and the detection volume V can be calculated using V = π 1.5 ω02 ω1 . This volume can be used to calculate concentrations and diffusion coefﬁcients from N and τD respectively in subsequent experiments. Notes a

Fluorophore is only required for the beampath being used during the experiment. R6G sufﬁces for 458, 477, 488, 514, 543 and 561 nm excitations. Cy5 is required for 633 and 647 nm excitation. Other ﬂuorophores may be used, at concentrations giving the appropriate count rate.

b The

exact procedure for this will vary between microscopes. Use the reﬂection of the laser light from the lower and subsequently upper surface of the coverslip as the focal position is moved upwards.

c Separate

calibration procedures are required for each combination of excitation wavelength and dichroic and emission ﬁlters. Remember to calibrate the beampath to be used in the experiment, not the one optimum for the calibration ﬂuorophore. d

If the software allows, then it is best to adjust the laser power using a real-time readout of molecular brightness η, rather than count rate. This is independent of the ﬂuorophore concentration, and should be set to 30–90 kHz counts per molecule per second.

Following alignment, a lower concentration of ﬂuorophore (10–50 nm) is used to obtain a calibration autocorrelation curve (Protocol 9.1, steps 11–16). The laser power should be adjusted to obtain an η value of 30–80 kHz, and data collected for 10 × 10 s for example. This curve should be ﬁtted to a simple one-component 3D diffusion model

9.2 METHODS AND APPROACHES

177

and the parameters noted. Parameters from this ﬁt can be used to conﬁrm alignment of the system and calculate the detection volume. One of these parameters is the structure parameter S. This represents the ratio of the height to diameter of the detection volume, and should be 3–7 for a well-aligned single-photon system. A value outside this range for a calibration read may indicate poor pinhole alignment or an incorrect correction collar adjustment. In this case, the alignment procedure should be repeated until a satisfactory value is obtained. This value for S can then be ﬁxed in any subsequent curve ﬁtting. The diffusion time τD obtained from this reading can also be used to calculate the volume radius ω0 and, subsequently, the half-height z0 and volume V (see Protocol 9.1, steps 15–16).

9.2.4 Measuring GPCR diffusion in the cell membrane 9.2.4.1 Choice of ﬂuorescent label The choice of ﬂuorophore for FCS can be complex, as photophysical events (such as triplet-state and dark-state formation) and photobleaching can signiﬁcantly inﬂuence autocorrelation curves [49, 50]. For direct measurement of GPCR diffusion using FCS, ﬂuorescent protein fusions have been predominantly used [15–17]. There is now a wide range of ﬂuorescent proteins available, with excitation/emission maxima ranging from the ultraviolet to the red wavelengths [51, 52]. Since cellular autoﬂuorescence is generally more prominent at shorter wavelengths, a red-shifted protein may be effective. Other considerations include quantum yield, photobleaching, environmental sensitivity (e.g. pH dependence), size, photophysical properties (e.g. blinking) and propensity for dimerization [53–57]. Other recently developed labelling strategies may prove more ﬂexible. For instance, SNAP-tag [4, 58–60] and the tetra-cysteine-based ‘ﬂuorescein-based arsenical hairpin binder’ labelling systems [61, 62] provide ways in which a single fusion construct can be exogenously labelled with probes of varying wavelengths. The position of the ﬂuorescent label also needs to be considered. With GPCRs, labelling on the C-terminus has proved widely successful. However, there is always a danger, particularly with larger tags, that such labelling will sterically interfere with the interaction of the C-terminus with interacting proteins such as scaffolding proteins and cytoskeleton [63]. This could signiﬁcantly effect translational membrane diffusion. On the other hand, successful expression of GPCRs which are tagged on their extracellular N-terminus often results in an inability of the cell to deliver the protein to the cell surface. Some success has been obtained by including a short signal sequence upstream of the N-terminal label [59].

9.2.4.2 Cell culture and manipulation Maintaining cell health is crucial to minimizing autoﬂuorescence in FCS experiments, and this issue is cell-type dependent [12, 39, 42]. For instance, Chinese hamster ovary (CHO) and human embryonic kidney cells have relatively low autoﬂuorescence, whilst primary cells often have high levels. Some cell types are also less tolerant of glass as a substrate, which can lead to cell stress and higher autoﬂuorescence. Cells should be grown in culture medium which has no phenol red, as this can contribute signiﬁcantly

178

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

to the background ﬂuorescence. Similarly, buffers should be made from HPLC-grade water with low ﬂuorescence. Cells should be washed in buffer two to three times prior to use to remove as much particulate matter as possible and the cells left to equilibrate to the desired temperature for 5–10 min. Generally, FCS measurements on cell membranes are easier to perform at room temperature (22–24 ◦ C), since there is less cell membrane movement at this temperature, and this can introduce signiﬁcant artefacts into autocorrelation analysis [64, 65].

9.2.4.3 Autoﬂuorescence All cellular systems exhibit some degree of autoﬂuorescence. This can be minimized, for instance, by using phenol-red-free culture media, keeping cells healthy and using labels with excitation maxima at longer wavelengths [42]. The level of autoﬂuorescence should be determined directly by taking FCS measurements in parent cells which contain no ﬂuorescently labelled receptor. Under these circumstances, positioning of the volume on the membrane is difﬁcult and cytosolic measurements should be used. This can be achieved by positioning the measurement volume in x − y over the cytosol and then performing an intensity scan in z to position it within the cytosol. Ideally, these measurements should be taken using the same read time and laser power as those used for tagged receptor and averaged over a number (10–20) of cells (see Section 9.2.4.4). Autoﬂuorescence count rates can be 0.5–3 kHz. If this autoﬂuorescence gradually decreases during a read, then exposing the cells to laser light prior to an FCS read (known as a pre-bleach, see Section 9.2.4.4) may remove the interference. If the autoﬂuorescence signal is stable, and does not correlate, then this can be dealt with as background correction (see Section 9.2.4.4) [42, 43]. If the autoﬂuorescence signal autocorrelates, but is <10% of that obtained in the presence of receptor, then it can usually be discounted.

9.2.4.4 Taking a membrane ﬂuorescence correlation spectroscopy measurement To make an FCS measurement of a tagged GPCR in a cell membrane, an appropriate cell is ﬁrst chosen (Protocol 9.2). Cells can be visualized using either wideﬁeld illumination with a mercury/xenon arc-lamp or via scanning confocal illumination. It is important that this initial visualization period is short to minimize bleaching. Cells should be used which show normal morphology and distinctive membrane ﬂuorescence; this should help precise positioning of the detection volume on the membrane. Any excessive intracellular ﬂuorescence may indicate poor protein folding or poor cell health (although some Golgi/endoplasmic reticulum ﬂuorescence may be expected). The optimum ‘brightness’ of cells can be determined empirically. Very bright cells may yield an autocorrelation curve of low amplitude, which is subsequently very difﬁcult to ﬁt. In contrast, dim cells will yield low S : N and the signal may not be sufﬁciently above background levels. In time, it may be possible to set the camera exposure or ampliﬁer gain such that cells with an appropriate level of ﬂuorescence are easy to identify. Using cell populations of mixed brightness (e.g. transient transfections) or stably transfected mixed-cell populations gives maximum ﬂexibility in this respect.

9.2 METHODS AND APPROACHES

PROTOCOL 9.2 Taking an FCS Measurement of a GFP-tagged GPCR on the upper Cell Membrane Equipment and Reagents • FCS-capable microscope • R6G (10 mM ethanolic solution)  • Water (high-purity, low-ﬂuorescence water for HPLC, e.g. Choromasolv ) • Lab-Tek eight-well chambered coverglass (No. 1 thickness, Nalge Nunc International) • HEPES-buffered saline solution (HBSS; 147 mM NaCl, 24 mM KCl, 1.3 mM CaCl2 , 1 mM MgSO4 , 1 mM sodium pyruvate, 1 mM NaHCO3 , 10 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES), pH 7.4)a • Cell culture medium, without phenol red • Cell line expressing your GPCR–GFP construct of interest at the cell surface.

Method 1 Seed cells onto a chambered coverglass in phenol-red-free culture medium 24–48 h prior to experimentation. Cells should be attached ﬁrmly for measurements. Seed at a density which will achieve 60–70% conﬂuency on the day of experimentation. 2 Align the microscope and perform calibration readings and calculations as described in Protocol 9.1. 3 In addition, check the HBSS for correlating ﬂuorescence by reading 10 × 10 s at the laser power used for calibration. This should be noncorrelating and <0.5–1 kHz. 4 Remove the cell culture medium from cells and wash them three times with ∼500 µl per well of warmed HBSS (37 ◦ C). Add 360 µl HBSS to each chamber and allow the cells to equilibrate to room temperature for 15 min. 5 Apply immersion water to the objective lens and mount the chambered coverglass on the microscope with the appropriate chamber in the measurement position. 6 Locate a suitable healthy cell with discrete membrane ﬂuorescence. Using either white light or ﬂuorescent illumination, position the cell so that the measurement volume in x –y is over the optical axis of the microscope,b,c Positioning over the cell nucleus is recommended for the most discrete membrane peak in step 7. 7 Using minimum laser power, take an intensity scan through the cell in the z-axis at 0.5 µm intervals. This should result in two peaks, representing the lower and upper cell membranes. Adjust the z positioning of the confocal volume so that it lies directly on the peak intensity of the upper membrane. 8 Take an FCS measurement for 5 × 10 s, at a laser power which gives a count rate at least 10-fold above autoﬂuorescence levels.

179

180

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

9 Repeat the z-axis intensity scan (step 7) to check that the cell membrane has not moved signiﬁcantly from the centre of the confocal volume. 10 Repeat steps 8 and 9 on 5–10 cells and determine the amount of bleaching seen in the ﬁrst few reads. Use this to decide how long a pre-bleach period is necessary, if any. 11 Now take FCS measurements as above, using the optimum pre-bleach time/laser power combination, followed by 2 × 15 s of data collection. 12 Repeat this for a range of laser powers to determine the optimum power which provides a good S : N, but no left shift in the autocorrelation curve. Use this protocol in subsequent experiments for consistency. 13 Fit the data according to an equation representing a single two-dimensional (2D) diffusion component, with a pre-exponential term to account for photophysics if necessary, as follows: G(τ ) = 1 + AN−1 (1 + τ/τD )−1 where A N τD S Tb τb

= = = = = =

1 + (Tb e−τ/τb )(1 − Tb )−1 particle number diffusion time structure parameter blinking fraction blinking relaxation time.

14 If this produces a poor ﬁt, then a second component can be added as follows: G(τ ) = 1 + AN−1 [F1 (1 + τ/τD1 )−1 + F2 (1 + τ/τD2 )−1 ] where F1 , τD1 and F2 , τD2 are the relative fractions and diffusion times of components 1 and 2 respectively. 15 Similarly, an anomalous diffusion model, incorporating an anomaly factor, 0 < α < 1, can be used as follows: G(τ ) = 1 + AN−1 (F1 (1 + (τ/τD1 )α )−1 + F2 (1 + (τ/τD2 )α )−1 ) 16 Diffusion coefﬁcients can be calculated from dwell times according to the equation D = ω02 /4τD

Notes a This

solution will be suitable for many cell lines. However, a suitable buffered live-cell solution should be chosen, optimized for the cells used.

b The

method for determining the exact location of the laser beam in x –y is described in the

text. c Depending

on the design of the experiment, illumination by either ﬂuorescent light or normal white light may be preferable (i.e. whether you want to select cells of known ﬂuorescence or select cells blindly to avoid bias).

181

9.2 METHODS AND APPROACHES (a)

(c) Rate (kHz)

100 75 50 25

0

5

10

15

20

25

30

Time (s)

(b) Coverslip

Lower Upper Membrane Membrane

1.20

τD1=523µs (46%)

Chosen Z-position

10

Component 2 (diffusion)

1.10 1.05

Blinking

20

τD2=59ms (54%)

Component 1 (photophysics)

1.15 G (t)

Count Rate (kHz)

30

τb

1.00 10−5

0 1430

1435 z-position (µ)

1440

10−4

10−3

10−2

10−1

100

Time (s)

Figure 9.3 FCS measurement of the diffusion of a GFP-tagged GPCR (β3 -adrenergic receptor). (a) CHO cells were transiently transfected to express the β3 -adrenergic receptor tagged on its C-terminus with enhanced GFP (eGFP). The location of the detection volume in x –y is indicated by the white cross. (b) An intensity scan in z is shown, with peaks corresponding to the lower and upper membranes clearly identiﬁed. The chosen measurement position is indicated by the broken line. (c) Fluorescence ﬂuctuations (top) are recorded for 30 s, and the subsequent autocorrelation analysis (bottom) is indicated. Nonlinear curve ﬁtting indicates two diffusion components: τD1 (523 µs, 46%), representing an eGFP photophysical process; τD2 (59 ms, 54%), representing translational diffusion of the receptor. A faster photophysical phenomenon is represented by τb , which is accounted for by a pre-exponential term in the curve ﬁtting.

Having chosen a cell, the volume should be positioned in the x and y planes using the chosen method of visualization (Figure 9.3a). The position of the volume in x –y can be predetermined by bleaching a spot in a glass coverslip coated with R6G (e.g. applying an ethanolic solution of 100 nm R6G to the surface and allowing the liquid to evaporate). The best results are obtained if the detection volume is located close to the optical axis of the lens. To obtain the best deﬁnition of peaks for the upper and lower membranes, positioning the volume in x –y over the nucleus is recommended. A subsequent intensity scan in z, in 0.5 µm intervals, should reveal distinct peaks, representative of the lower and upper membranes (Figure 9.3b). To prevent bleaching, this z-scan should be carried out at the lowest possible laser power. The detection volume should then be positioned directly on the upper membrane peak. Measurement on the lower membrane peak is also possible, although measurement this close to the coverslip can be affected by scattering from the coverslip reﬂection.

182

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

A measurement can now be triggered. The optimum laser power and measurement time to use should be determined empirically. It is usual that, in the initial 5–15 s of exposure to laser light, the count rate will gradually decrease (bleaching). This can be caused by immobile receptor, and adds a noncorrelating background signal to the count rate. This has two consequences. First, the added background decreases the S : N ; second, the drop in count rate causes a large deviation of the autocorrelation curve at high values of τ , leading to an uninterpretable curve [50]. This background can be removed using a ‘pre-bleach’ period, where the sample is exposed to laser light prior to the start of data collection. The optimum period for this pre-bleach can be determined by monitoring this initial bleach over a series of measurements. Data collection times themselves will vary depending on the dwell time of the receptor, but will generally be in the region of 30–100 s. These data can be collected in two to ﬁve sets (e.g. 3 × 20 s). This has the advantage that any datasets which contain large aggregates, excessive bleaching or membrane movement can subsequently be removed from the analysis (see Section 9.2.4.4). However, it should be noted that, for autocorrelation analysis, 2 × 20 s of data is less statistically powerful than a single 40 s data collection [66]. For a statistically valid sample, ideally 1000 transitions of ﬂuorophore through the volume should be recorded (e.g. 10 s data collection for a τD of 10 ms). An optimum laser power should also be chosen empirically. Laser power needs to be high enough to provide a good S : N above that from background and autoﬂuorescence. However, too high a laser power can result in both global photobleaching and spot bleaching. Spot bleaching occurs when a ﬂuorescent particle is bleached directly as it passes through the excitation beam. This results in an artiﬁcially short dwell time, as the particle is effectively prematurely ‘removed’ from the volume. Initially, therefore, it is recommended that a series of measurements is performed with varying laser powers. The optimum power is the one which provides the best brightness value η, with no left shift in the autocorrelation curve. Ideally, such a series should be performed on a number of individual cells. Once a satisfactory autocorrelation curve has been obtained, appropriate parameters (diffusion times, etc.) can be obtained using nonlinear curve ﬁtting (see Section 9.2.4.4). In the case of ﬂuorescent-protein-tagged receptors, it is likely that these curves will consist of at least two components, with one representing some kind of intramolecular photophysical event, such as blinking (τD ≈ 50 − 400 µs) [53, 54, 57], and the other (τD ≈ 1 − 80 ms) representing translational diffusion of the receptor [10, 15–17]. This second component may actually consist of more than one species, which can be delineated by adding additional diffusion components during curve ﬁtting [16].

9.2.5 Measuring the diffusion of a ligand-occupied GPCR Using a ﬂuorescent receptor ligand in conjunction with FCS provides a powerful way to investigate the diffusion characteristics of different functional receptor forms. In addition, this method can be used with endogenously expressed receptors; for example, in primary cells and mixed cell populations. The lack of protein labelling required also means that there is no risk of altering the binding of receptor to signalling enzymes or scaffolding proteins.

9.2 METHODS AND APPROACHES

183

However, this approach does have limitations. First, it requires the design, synthesis and characterization of a suitable ﬂuorescent ligand (see Section 9.2.5.1). Second, FCS generally works best at low concentrations, and experiments are generally carried out under conditions of low receptor occupancy, where ligands which have a strong pharmacological preference for either active (agonists) or inactive receptors (inverse agonists) will predominantly label these conformations. In addition, as FCS is a live-cell technique, intracellular concentrations of guanosine triphosphate are high and receptor conformations which couple to G proteins will, therefore, have short lifetimes under these conditions [10]. Therefore, data needs to be interpreted in the light of these pharmacological constraints.

9.2.5.1 Design of ﬂuorescent ligands Most criteria for choosing an appropriate ﬂuorophore for labelling of ligand are similar to those for choosing a ﬂuorescent protein label. The label should be bright (have a high quantum yield), photostable, environmentally insensitive and have suitable photophysical properties. A wide variety of small-molecule ﬂuorophores are available for this purpose, with series such as the BODIPY, AlexFluor and cyanine dyes offering a range of excitation/emission wavelengths. From a cell phototoxicity and autoﬂuorescence point of view, red-shifted dyes provide a more suitable option. In general, the more blue shifted the dye, the more likely problems will occur with autoﬂuorescence, complications with photophysics and, particularly at ultraviolet wavelengths, phototoxicity. Much of the initial work using FCS to study ligand–receptor interactions was carried out using ﬂuorescently labelled peptide ligands labelled with amine-reactive ﬂuorophores [18–21, 24]. In such cases, the relative size of ligand to ﬂuorophore means that there is no signiﬁcant impact on the pharmacophore. However, due consideration still needs to be given to the position of the label, and the pharmacological activity of the labelled peptide should be conﬁrmed. Fluorescent labelling of small-molecule ligands for class A GPCRs is a more complex process, since the ﬂuorophore is often of a similar size to the pharmacophore [67, 68]. Structure–activity relationships need to be studied to determine a suitable labelling position [25–27]. This may require resynthesizing or derivitizing the ligand to incorporate, for example, an amine or thiol group to facilitate ﬂuorophore attachment [26, 27]. A linker may also be required to separate the ﬂuorophore and pharmacophore [17, 26, 27]. The length and composition of this linker (e.g. in terms of ﬂexibility and water solubility) can signiﬁcantly alter the subsequent pharmacological activity of the labelled ligand ([67] and references cited therein). These parameters will need to be optimized for each ligand family at each receptor. Finally, and perhaps most importantly, the pharmacological activity of the labelled ligand should be comprehensively studied. Afﬁnity, efﬁcacy and subtype selectivity can be substantially affected by addition of a ﬂuorophore, and the effects will vary with the parameters deﬁned above (ﬂuorophore structure, labelling position, linker length and composition) [26, 27]. For all ligands, high levels of purity (>99%) using HPLC, for example, are required, as contaminants can affect autocorrelation curves. Dissociation constants of 50 nm or

184

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

less are desirable to obtain high enough receptor occupancies at the concentrations which are amenable to FCS measurements.

9.2.5.2 Taking a ﬂuorescence correlation spectroscopy measurement of ligand-occupied receptors First, the diffusion time of the free ligand in the appropriate buffer should be determined. This can then be ﬁxed (along with the structure parameter S) in the subsequent curve ﬁt. In essence, performing ligand binding measurements with FCS (Protocol 9.3) is similar to the procedure described in Section 9.2.4.4. Cells should be cultured and prepared in the same way (Section 9.2.4.2), and with the same consideration of autoﬂuorescence (Section 9.2.4.3). Choice of ligand concentration and incubation time will be ligand and receptor dependent, but this is generally a compromise between obtaining a high degree of speciﬁc membrane labelling (i.e. minimizing any intracellular ligand) and highest molecular brightness value η, coupled with a large amplitude for the autocorrelation curve. A 10–20 min incubation with 5–50 nm ligand is a suitable starting point.

PROTOCOL 9.3 Taking an FCS Measurement of a GPCR Labelled Using Fluorescent Ligand on the upper Cell Membrane Equipment and Reagents • FCS-capable microscope • R6G (10 mM ethanolic solution) or Cy5 (10 mM aqueous solution)  • Water (high-purity, low-ﬂuorescence water for HPLC, e.g. Choromasolv ) • Lab-Tek eight-well chambered coverglass (No. 1 thickness, Nalge Nunc International) • HBSS (147 mM NaCl, 24 mM KCl, 1.3 mM CaCl2 , 1 mM MgSO4 , 1 mM Na pyruvate, 1 mM NaHCO3 , 10 mM HEPES, pH 7.4) • Cell culture medium, without phenol red • Cell line expressing your GPCR of interest at the cell surface • Fluorescent ligand (10 mM or 1 mM stock solution).

Method 1 Align and calibrate the FCS microscope for the appropriate beampath using either R6G or Cy5 as per Protocol 9.1. 2 Prepare a solution of ﬂuorescent ligand in HBSS at a ﬁnal concentration of 20 nM, and add 200 µl to an empty chamber of the coverglass, and position over the objective.

9.2 METHODS AND APPROACHES

185

3 Move the focal position to 200 µm above the coverslip. Adjust the laser power to obtain a brightness value η of 20–60 kHz. Take an FCS read for 10 × 10 s. Fit to a single-component 3D diffusion model and determine the diffusion time of free ligand. 4 Remove the cell medium and wash three times in warmed HBSS, replacing ﬁnally with 360 µl of HBSS. Allow cells to equilibrate to room temperature for 15 min. 5 Place the chambered coverglass onto the microscope in the appropriate well. 6 Add 40 µl of ﬂuorescent ligand at a concentration 10-fold above the ﬁnal concentration required.a 7 Locate a suitable cell and position the measurement volume over the nucleus in x –y. 8 After an appropriate incubation period,b scan the cell for intensity in the z-axis at 0.5 µm intervals with minimal laser power. Two peaks should be obtained for lower and upper membranes. 9 Position the confocal volume on the point at which the peak intensity of the upper membrane decays to 50%. 10 Perform an FCS measurement with a 15 s pre-bleach, followed by 4 × 15 s data collection.c 11 Repeat the z-axis intensity scan (step 7) to check the cell membrane has not moved signiﬁcantly from the centre of the confocal volume. 12 Move the measurement position 2–3 µm upwards in the z-axis. Repeat the FCS read, to conﬁrm that only free ligand is present. 13 Analyse the data with the ligand diffusion time ﬁxed to that obtained in step 1 and the structure parameter ﬁxed from the calibration read (see Protocol 9.1). 14 For an autocorrelation curve containing free ligand and two binding components, a model assuming one 3D component, two 2D components and a pre-exponential A for triplet should be used as follows: G(τ ) = 1 + AN−1 {F1 (1 + τ/τD )−1 [1 + τ/(S2 τD1 )]−0.5 + F2 (1 + τ/τD2 )−1 + F3 (1 + τ/τD3 )−1 } where F1 and τD1 respectively represent the fractional contribution and the dwell time of the free ligand, and F2 , F3 and τD2 , τD3 respectively represent the fractional contributions and the dwell times for the ligand-bound receptor.

Notes a As

described in the text, the concentration of ligand used will depend on its afﬁnity and molecular brightness value. A suitable starting concentration will be in the 5–25 nM range. b As

above, the optimum incubation time will be ligand and receptor dependent. Bear in mind that the incubation needs to be long enough to give a good signal, but not high enough to give excessive intracellular ligand concentration. c Optimum

pre-bleach times and laser powers should be determined as in Protocol 9.1, steps 10–12.

186

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

Again, cells can be visualized by wideﬁeld or confocal illumination. Alternatively, a wideﬁeld white light image can be used to choose cells ‘blind’. This is a useful method to use in experiments using a nonﬂuorescent competing ligand to avoid bias towards particularly bright or dim cells. The measurement volume should be positioned in x –y, as described previously, with positioning over the nucleus providing the most deﬁned peak in the z-scan (Figure 9.4a), which should be performed at the lowest possible laser intensity. Positioning the volume in z requires more care than with tagged receptor measurements. It is usually advisable to position the volume at the point were the upper membrane peak has decayed to 50% of its peak value. This results in the membrane resting approximately two-thirds of the way towards the bottom of the volume. Care should be taken with this approach, however, since membrane measurements are prone to artefacts from membrane movement, and these are exaggerated for membrane which is further form the centre of the illumination volume [64, 65]. In such instances, where free ligand concentrations are not required, the volume can be placed directly on the upper membrane peak. Measurements can also be made on the lower membrane, although ligand access to this area can often be limited, depending on the nature of the cell attachment. Problems can also occur with scattered or reﬂected light and the ‘unstirred layer’ effect of ligand with limited movement. This can be particularly troublesome if the ligand interacts with or sticks to the glass carrier. Similar considerations apply when determining the optimal laser power, length of data collection and application of a pre-bleach period, as described in Section 9.2.4.4. However, in most cases ﬂuorescent ligands are less prone to global bleaching problems than ﬂuorescent protein is. This is partly because of their photophysical properties, but also because, if readings are taken in the presence of free ligand, there is a large pool of replacement ﬂuorophore available. Spot bleaching can still be a problem, however, and the power dependence of the diffusion time should be investigated. A suggested starting point would be a 15 s pre-bleach followed by 4 × 15 s measurements. Following the FCS measurement, the cell can be rescanned in z and the confocal volume repositioned 2–3 µm above the cell (1–1.5 times the confocal volume height). Performing an FCS measurement here should detect a single fast-diffusing component. This conﬁrms that the previous measurement position was on the membrane, and it can also be used to calculate the free ligand concentration. The resulting autocorrelation curve will contain at least two components. A fastdiffusing component with a similar diffusion time to that seen for free ligand in buffer should be seen in the left part of the curve (τD ≈ 50 − 100 µs). If the volume has been positioned correctly in z, then this should constitute approximately one-third of the amplitude of the curve. The remaining two-thirds of the amplitude will be made up of slower diffusing components. These represent speciﬁc receptor-bound ligand, nonspeciﬁcally bound ligand and intracellular ligand. Interpretation of which component is represented by which species is complicated, but may be clariﬁed by the following experiments. First, cytosolic measurements will determine the concentration of intracellular ligand and its diffusion time. Second, at least in the case of transfected receptor systems, the equivalent experiment should also be performed in cells not expressing the receptor of interest. Care should be taken in these experiments that the peak seen in the z-scan really does represent the cell membrane and not the

187

9.2 METHODS AND APPROACHES (a) Lower Coverslip Membrane

Upper Membrane

Rate (kHz)

150

M

E

100

50

0 1210

1215

1220

1225

Position (µm) Membrane

Extracellular

160 140 120 100 80 60

Rate (kHz)

Rate (kHz)

(b)

0

5

10

15

20

25

100 80 60 40 20 0

30

5

10

τD1=63µs (50%) τD2=3.6ms (21%) τD3=93ms (29%) 1.10

Component 1 (free ligand)

Component 2 (bound)

25

30

10−1

100

τD1=60µs (100%)

Component 3 (bound)

Component 1 (free ligand)

1.6

1.06 1.04

1.4 1.2

1.02 1.00

20

1.8

G(t)

G(t)

1.08

15 Time (s)

Time (s)

10−5

10−4

10−3 Time (s)

10−2

10−1

1.0 100

−5

10

10−4

10−3

10−2

Time (s)

Figure 9.4 FCS measurement of the diffusion of an agonist-occupied GPCR (adenosine-A3 receptor with ABEA-X-BY630). (a) CHO cells expressing the human adenosine-A3 receptor were incubated with 5 nM ABEA-X-BY630 for 10 min at room temperature. The measurement volume was positioned in x –y as indicated by the white cross (left). An intensity scan in z (right) reveals peaks for the lower and upper membranes. The position of the measurement volume for data collection on the membrane (M) and extracellular solution (E) is shown. (b) Fluctuations and autocorrelation analysis from 30 s of data collection on the membrane (left) and in the extracellular solution (right) are shown. Also indicated are the diffusion times, their relative contributions and their assignment. For the membrane measurement, this includes free ligand (τD1 ) and two slower diffusing receptor–ligand species (τD2/3 ). Above the cell, only free ligand was detected.

188

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

cytosol. Third, displacement or competition experiments can be performed using a nonﬂuorescent ligand which is speciﬁc to the receptor of interest. Classically, a concentration of competitor 500–1000-fold above the equilibrium dissociation constant KD would be used to provide complete displacement. In FCS experiments, since a complete displacement of ligand would make it impossible to position the volume on the membrane, a concentration of 50–100 times KD should be used. This should give signiﬁcant displacement of speciﬁc binding, but leave sufﬁcient signal to position the volume accurately. However, it should be noted that, at the low receptor occupancies used, a competitor of widely differing efﬁcacy to the labelled ligand (e.g. an inverse agonist displacing a full agonist) may show a much reduced KD to radioligand binding experiments.

9.2.6 Data analysis Diffusion times and particle numbers are derived from autocorrelation data using nonlinear curve-ﬁtting techniques. A Marquardt–Levenberg nonlinear least-squares ﬁtting routine is recommended. This kind of routine is often available within manufacturer’s software (such as Zeiss’s AIM software), but can also be performed in more commonly used graphing/statistical packages (e.g. GraphPad Prism, Origin, Igor or MatLab). Fitting should be performed using the most appropriate biophysical model. For solution-based calibrations, this is a model assuming a single diffusing species showing simple 3D Brownian motion (Protocol 9.1, step 13). A pre-exponential term is used to incorporate any photophysical phenomena, such as triplet-state formation. Fitting of the calibration data to this equation should provide a value for the structure parameter S of 3–7. A value of S outside these values suggests poor system alignment. This value of S should be noted and ﬁxed in subsequent analysis. The calibration data should also provide a value for τD which can be used to calculate the volume parameters, as deﬁned by the point at which the excitation intensity drops to 1/e2 value of the peak intensity (see Protocol 9.1, steps 15 and 16). Triaging the experimental data is an important part of the analysis procedure. For data which consist of a number of repeated measurements, poor datasets need to be removed from the subsequent analysis. For instance, where the average count rate drops by more than 10% of the initial value through global bleaching, the ﬁtted value for diffusion time will be skewed to longer values. Likewise, if any strong deviation in count rate (more than two standard deviations from the mean value) is seen during the read (e.g. because of bright aggregates or substantial membrane movement), then the measurement should be discarded. Several models may need to be compared when ﬁtting data from membrane measurements. Data from tagged GPCRs can initially be ﬁtted to a 2D diffusion model (note that the 2D equation does not require a value for S). This will usually require a pre-exponential term to account for photophysics, in particular with ﬂuorescent proteins, to account for blinking. Initial ﬁtting should assume a single component. If the ﬁt is poor, then a further diffusion component may be added. With GFP and yellow ﬂuorescent protein, for instance, a second component is usually needed to account

9.2 METHODS AND APPROACHES

189

for complex photophysical events [53, 54, 57]. The ﬁt should then also return the fractional contribution of each component. Particle numbers for each component can then be calculated simply by multiplying the fractional component by the total particle number from the ﬁt. Where simple 2D diffusion still appears to provide a poor ﬁt, an anomalous diffusion model can be applied [64]. This introduces an exponential factor of between 0 and 1 to the diffusion time to account for restricted movement. Goodness of ﬁt can be assessed by using the χ 2 value (sum of squares). Different ﬁt models should be compared using an F -test to determine which ﬁt is the more statistically valid. In addition, visual inspection of the residuals from the ﬁt can give an indication whether a ﬁt model is good – large deviations in the residuals are acceptable providing they are nonsystematic. For ligand binding data, free ligand diffusion time should be initially determined using a single-component 3D Brownian diffusion ﬁt. This diffusion time τD1 , along with S, should be ﬁxed in the subsequent ﬁt of binding data to minimize the number of variable parameters. Generally, the model used for binding data contains, in addition to a pre-exponential, a 3D diffusion term (τD1 , for free ligand) and one or more 2D diffusion terms for membrane-bound ligand (τD2/3 ) [30]. Concentrations for each individual species can be calculated from their fractional contributions and total particle number, as above. Free ligand concentration can be calculated from this curve, using half the detection volume to calculate concentration. More commonly, a separate measurement 2 µm above the cell membrane can be taken to calculate free ligand concentration directly. Where a number of identical measurements have been taken (i.e. where the diffusion time would be expected to remain the same between cells), globally ﬁtting the diffusion time data signiﬁcantly improves the statistical power of the analysis. Global ﬁtting simultaneously compares all datasets and returns a value for the globally ﬁt parameter which best ﬁts all the data [69]. Such global ﬁtting routines are available in many of the commercially available software packages. In the presence of autoﬂuorescence or a high noncorrelated background, the autocorrelation curve may be offset. This can be accounted for in ﬁtting, by using a simple additional offset value in the ﬁt. Likewise, if the background ﬂuorescence is known (this can equally apply to autoﬂuorescence), this can be corrected for in the ﬁt, by using a correction value of (1 − IB /IT )2 , where IB is the background intensity and IT is the total intensity. For ligand binding, one particularly important consideration is any difference in quantum yield between free and bound ligand. This consideration also applies to any data which has large (bright) aggregates involved. Where multiple components are present, the amplitude of the autocorrelation curve is skewed towards the brighter species by the square of the relative brightness. This can lead to an overestimation of the particle number for very bright species, and a subsequent underestimation of species with low quantum yield. If the quantum yield or relative brightness of the two species is known, then this can also be accounted for in the ﬁt, by introducing a term of the square of the fractional quantum yield for free and bound ligand in front of the appropriate diffusion term in the ﬁt equation [30].

190

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

9.3 Troubleshooting • If good calibration data cannot be obtained from an aqueous ﬂuorophore solution (consistent diffusion time, high η value, 3 < S < 7), then this suggests problems with the system alignment. Check the correction collar on the objective, the pinhole adjustment and the collimator position. Optical stability is helped by allowing 2 h for stabilization of system components prior to use, and maintaining an air temperature of ±2–3 ◦ C. Laser stability can be tested using a concentrated solution of ﬂuorophore (10−5 m) and monitoring output; this should achieve a steady count rate. • In the case of low values for η from cellular measurements, system alignment should be checked initially. If this is satisfactory, then this may be a result of high levels of autoﬂuorescence or scattering from cellular structures. Make sure cells are healthy, use phenol-red-free cell medium and try measuring at least 4–5 µm above the coverslip to reduce reﬂected light. Increase the laser power to obtain a higher S : N , but not such that bleaching is introduced. If possible, move to a cell line with lower autoﬂuorescence, or consider bleaching the autoﬂuorescence out before taking a measurement. • A high uncorrelated background signal will also cause a substantial drop in S : N . This may be a result of a high immobile fraction of receptor, and can be checked using FRAP [13, 70], which may be a more suitable technique in these circumstances. Alternatively, related techniques such as image correlation spectroscopy [70] or scanning FCS [31, 32] could be considered, as these are better at detecting slow-moving populations. • Bleaching (both global and spot) can be a signiﬁcant problem, especially with ﬂuorescent proteins. Use the minimum laser power which gives a workable η value and shorten read times to the minimum necessary for the diffusion time of the species under study. Pre-bleaching may help by removing immobile receptor, which bleaches more quickly. Using multiphoton excitation, if available, will help reduce global bleaching (and will also help with autoﬂuorescence issues) [12, 36]. However, spot bleaching is not usually affected. • Where very high count rates are obtained (this can be a problem with higher concentrations of very bright ligands, for instance), afterpulsing at the detectors may occur. Whilst this is not usually a problem for slow diffusing species (the effects are seen at very short τ values), try reducing the laser power to a minimum. If the problem persists, try cross-correlating the signal across two identical detection channels – the afterpulsing will not be correlated in two different detectors. • If differences in diffusion times following exposure to drugs or treatments are being assessed, then it may be that these differences are too small to be picked up by FCS. The change in diffusion time is proportional to the cube root of the mass difference, meaning that a sixfold mass change is necessary to give a 1.6-fold difference in diffusion time [45]. This is the minimum difference which can be reliably detected using

REFERENCES

191

autocorrelation analysis. In these instances, the alternative analysis of ﬂuctuation data using photon counting histogram analysis [71, 72] can be applied. This analysis still provides particle number, but calculates molecular brightness directly, rather than diffusion time. This is sensitive enough to allow, for instance, discrimination of monomeric and dimeric receptor species.

References 1. Ostrom, R.S. and Insel, P.A. (2004) The evolving role of lipid rafts and caveolae in G proteincoupled receptor signaling: implications for molecular pharmacology. Br. J. Pharmacol., 143, 235–245. A review of the importance of compartmentalization in GPCR signalling. 2. Ostrom, R.S., Post, S.R. and Insel, P.A. (2000) Stoichiometry and compartmentation in G proteincoupled receptor signaling: implications for therapeutic interventions involving Gs . J. Pharmacol. Exp. Ther., 294, 407–412. 3. H´ebert, T.E., Gales, C. and Rebois, R.V. (2006) Detecting and imaging protein–protein interactions during G protein-mediated signal transduction in vivo and in situ by using ﬂuorescence-based techniques. Cell Biochem. Biophys., 45, 85–109. A useful comparison of FRAP, FRET and BRET techniques for studying GPCRs. 4. Maurel, D., Comps-Agrar, L., Brock, C. et al. (2008) Cell-surface protein–protein interaction analysis with time-resolved FRET and snap-tag technologies: application to GPCR oligomerization. Nat. Methods, 5, 561–567. 5. Hoffman, C., Gaietta, G., Bunemann, M. et al. (2005) A FlAsH-based FRET approach to determine G protein-coupled receptor activation in living cells. Nat. Methods, 2, 163–164. 6. Pfelger, K.D. and Eidne, K.A. (2005) Monitoring the formation of dynamic G-protein-coupled receptor–protein complexes in living cells. Biochem. J., 385, 625–637. 7. Charalambous, C., Gstander, I., Keuerleber, S. et al. (2008) Restricted collision coupling of the A2A receptor revisited: evidence for physical separation of two signaling cascades. J. Biol. Chem., 283, 9276–9288. 8. Pucadyil, T.J. and Chattopadhyay, A. (2007) Cholesterol depletion induces dynamic conﬁnement of the G-protein coupled serotonin1A receptor in the plasma membrane of living cells. Biochim. Biophys. Acta, 1768, 655–668. 9. Wheeler, D., Sneddon, W.B., Wang, B. et al. (2007) NHERF-1 and the cytoskeleton regulate the trafﬁc and membrane dynamics of G protein-coupled receptors. J. Biol. Chem., 282, 25076–25087. 10. Briddon, S.J. and Hill, S.J. (2008) Pharmacology under the microscope: the use of ﬂuorescence correlation spectroscopy to determine the properties of ligand–receptor complexes. Trends Pharmacol. Sci., 28, 637–645. An up-to-date review of the application of FCS to studying receptor diffusion. 11. Vukojevi´c, V., Parmanik, A., Yakovleva, T. et al. (2005) Study of molecular events in cells by ﬂuorescence correlation spectroscopy. Cell Mol. Life Sci., 62, 535–550.

192

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

12. Schwille, P. (2001) Fluorescence correlation spectroscopy and its potential for intracellular applications. Cell Biochem. Biophys., 34, 383–408. Detailed review of the practical and theoretical applications of FCS to cell biology. 13. Chen, Y., Lagerholm, B.C., Yang, B. and Jacobson, K. (2006) Methods to measure the lateral diffusion of membrane lipids and proteins. Methods, 39, 147–153. 14. Garcia-Saez, A.J. and Schwille, P. (2007) Single molecule techniques for the study of membrane proteins. Appl. Microbiol. Biotechnol., 76, 257–266. 15. Licht, S.S., Sonnleitner, A., Weiss, S. and Schultz, P.G. (2003) A rugged energy landscape mechanism for trapping of transmembrane receptors during endocytosis. Biochemistry, 42, 2916–2925. 16. Philip, F., Sengupta, P. and Scarlata, S. (2007) Signaling through a G protein-coupled receptor and its corresponding G protein follows a stoichiometrically limited model. J. Biol. Chem., 282, 19203–19216. 17. Briddon, S.J., Middleton, R.J., Cordeaux, Y. et al. (2004) Quantitative analysis of the formation and diffusion of A1 -adenosine receptor-antagonist complexes in single living cells. Proc. Natl. Acad. Sci. U. S. A., 101, 4673–4678. 18. Zhong, Z.H., Parmanik, A. and Ekberg, K. et al. (2001) Insulin binding monitored by ﬂuorescence correlation spectroscopy. Diabetologia, 44, 1184–1188. 19. Pramanik, A. and Rigler, R. (2001) Ligand–receptor interactions in the membrane of cultured cells monitored by ﬂuorescence correlation spectroscopy. Biol. Chem., 382, 371–378. 20. Rigler, R., Pramanik, A., Jonasson, P. et al. (1999) Speciﬁc binding of proinsulin C-peptide to human cell membranes. Proc. Natl. Acad. Sci. U. S. A., 96, 13318–13323. One of the ﬁrst examples of the use of FCS to characterize ligand-occupied receptors in living cell membranes. 21. Pramanik, A., Olsson, M., Langel, U. et al. (2001) Fluorescence correlation spectroscopy detects galanin receptor diversity on insulinoma cells. Biochemistry, 40, 10839–10845. 22. Hegener, O., Prenner, L., Runkel, F. et al. (2004) Dynamics of β2 -adrenergic receptor–ligand complexes on living cells. Biochemistry, 43, 6190–6199. 23. Meissner, O. and H¨aberlein, H. (2003) Lateral mobility and speciﬁc binding to GABAA receptors on hippocampal neurons monitored by ﬂuorescence correlation spectroscopy. Biochemistry, 42, 1667–1672. 24. Henriksson, M., Pramanik, A., Shafgat, J. et al. (2001) Speciﬁc binding of proinsulin C-peptide to intact and to detergent-solubilized human skin ﬁbroblasts. Biochem. Biophys. Res. Commun., 280, 423–427. 25. Hegener, O., Jordan, R. and Haberlein, H. (2004) Dye-labeled benzodiazepines: development of small ligands for receptor binding studies using ﬂuorescence correlation spectroscopy. J. Med. Chem., 47, 3600–3605. 26. Wohland, T., Friedrich, K., Hovius, R. and Vogel, H. (1999) Study of ligand–receptor interactions by ﬂuorescence correlation spectroscopy with different ﬂuorophores: evidence that the homopentameric 5-hydroxytryptamine type 3As receptor binds only one ligand. Biochemistry, 38, 8671–8681. 27. Middleton, R.J., Briddon, S.J., Cordeaux, Y. et al. (2007) New ﬂuorescent adenosine A1 -receptor agonists that allow quantiﬁcation of ligand–receptor interactions in microdomains of single living cells. J. Med. Chem., 50, 782–793.

REFERENCES

193

28. Briddon, S.J., Middleton, R.J., Yates, A.S. et al. (2003) Application of ﬂuorescence correlation spectroscopy to the measurement of agonist binding to a G-protein coupled receptor at the single cell level. Faraday Discuss., 126, 197–207. 29. Cordeaux, Y., Briddon, S.J., Alexander, S.P.H. et al. (2008) Agonist-occupied A3 adenosine receptors exist within heterogeneous complexes in membrane microdomains of individual living cells. FASEB J., 22, 850–860. 30. Pramanik A. and Rigler R. (2001) FCS-analysis of ligand–receptor interactions in living cells, in Fluorescence Correlation Spectroscopy: Theory and Applications (eds E. Elson and R. Rigler), Springer, Heidelberg, pp. 101–129. A detailed and comprehensive discussion of the use of FCS to measure ligand binding in living cells. 31. Digman, M.A., Brown, C.M., Horwitz, A.F. and Gratton, E. (2005) Measuring fast dynamics in solutions and cells with a laser scanning microscope. Biophys. J., 89, 1317–1327. 32. Ries, J. and Schwille, P. (2006) Studying slow membrane dynamics with continuous wave scanning ﬂuorescence correlation spectroscopy. Biophys. J., 91, 1915–1924. 33. Schwille, P., Meyer-Almes, F.J. and Rigler, R.J. (1997) Dual-color ﬂuorescence cross-correlation spectroscopy for multicomponent diffusional analysis in solution. Biophys. J., 72, 1878– 1886. 34. Heinze, K.G., Koltermann, A. and Schwille, P. (2000) Simultaneous two-photon excitation of distinct labels for dual-color ﬂuorescence crosscorrelation analysis. Proc. Natl. Acad. Sci. U. S. A., 97, 10377–10382. 35. Hwang, L.C., Gosch, M., Lasser, T. and Wohland, T. (2006) Simultaneous multicolor ﬂuorescence cross-correlation spectroscopy to detect higher order molecular interactions using single wavelength laser excitation. Biophys J., 91, 715–727. 36. Schwille, P., Haupts, U., Maiti, S., and Webb, W.W. (1999) Molecular dynamics in living cells observed by ﬂuorescence correlation spectroscopy with one- and two-photon excitation. Biophys J., 77, 2251–2265. 37. Haustein, E. and Schwille, P. (2007) Fluorescence correlation spectroscopy: novel variations of an established technique. Annu. Rev. Biophys. Biomol. Struct., 36, 151–169. 38. Thompson, N.L. and Steele, B.L. (2007) Total internal reﬂection with ﬂuorescence correlation spectroscopy. Nat. Protoc., 2, 878–890. 39. Kim, S., Heinze, K. and Schwille, P. (2007) Fluorescence correlation spectroscopy in living cells. Nat. Methods, 4, 963–973. Provides a thorough technical and troubleshooting guide to FCS measurements in cells. 40. Bacia, K. and Schwille, P. (2007) Practical guidelines for dual-color ﬂuorescence cross-correlation spectroscopy. Nat. Protoc., 2, 2842–2856. 41. Elson, E.L. (2004) Quick tour of ﬂuorescence correlation spectroscopy from its inception. J. Biomed. Opt., 9, 857–864. An essential overview of the development and history of FCS form the 1970s onwards. 42. Brock, R., Hink, M.A. and Jovin, T.M. (1998) Fluorescence correlation microscopy of cells in the presence of autoﬂuorescence. Biophys. J., 75, 2547–2557. 43. Brock, R. and Jovin, T.M. (1998) Fluorescence correlation microscopy (FCM) – ﬂuorescence correlation spectroscopy (FCS) taken into the cell. Cell Mol. Biol., 44, 847–856.

194

CH 9 USE OF FLUORESCENCE CORRELATION SPECTROSCOPY

44. Sengupta, P., Balaji, J. and Maiti, S. (2002) Measuring diffusion in cell membranes by ﬂuorescence correlation spectroscopy. Methods, 27, 374–387. Detailed information on the constituent components for bulding a bespoke FCS system. 45. Meseth, U., Wohland, T., Rigler, R. and Vogel, H. (1999) Resolution of ﬂuorescence correlation measurements. Biophys. J., 76, 1619–1631. 46. Weisshart, K., Jungel, V. and Briddon, S.J. (2004) The LSM 510 META–ConfoCor 2 system: an integrated imaging and spectroscopic platform for single-molecule detection. Curr. Pharm. Biotechnol., 5, 135–154. 47. Enderlein, J., Gregor, I., Patra, D. et al. (2005) Performance of ﬂuorescence correlation spectroscopy for measuring diffusion and concentration. ChemPhysChem, 6, 2324–2336. An important study on the effects of system set-up on the detection volume and quantiﬁcation. 48. Hess, S.T. and Webb, W.W. (2002) Focal volume optics and experimental artifacts in confocal ﬂuorescence correlation spectroscopy. Biophys J., 83, 2300–2317. 49. Davis, L.M. and Shen, G. (2006) Accounting for triplet and saturation effects in FCS measurements. Curr. Pharm. Biotechnol., 7, 287–301. 50. Widengren, J. and Thyberg, P. (2005) FCS cell surface measurements – photophysical limitations and consequences on molecular ensembles with heterogenic mobilities. Cytometry A, 68, 101–112. An important study of the effects of photobleaching on FCS measurements. 51. Shaner, N.C., Patterson, G.H. and Davidson, M.W. (2007) Advances in ﬂuorescent protein technology. J. Cell Sci., 120, 4247–4260. 52. Olenych, S.G., Claxton, N.S., Ottenberg, G.K., and Davidson, M.W. (2006) The ﬂuorescent protein color palette. Curr. Protoc. Cell Biol., 21.5, 1–33. 53. Schwille, P., Kummer, S., Heikal, A.A. et al. (2000) Fluorescence correlation spectroscopy reveals fast optical excitation-driven intramolecular dynamics of yellow ﬂuorescent proteins. Proc. Natl. Acad. Sci. U. S. A., 97, 151–156. 54. Haupts, U., Maiti, S., Schwille, P. and Webb, W. (1998) Dynamics of ﬂuorescence ﬂuctuations in green ﬂuorescent protein observed by ﬂuorescence correlation spectroscopy. Proc. Natl. Acad. Sci. U. S. A., 95, 13573–13578. 55. Remington, S.J. (2006) Fluorescent proteins: maturation, photochemistry and photophysics. Curr.Opin. Struct. Biol., 16, 714–721. 56. Sinnecker, D., Voigt, P., Hellwig, N. and Schaefer, M. (2005) Reversible photobleaching of enhanced green ﬂuorescent proteins. Biochemistry, 44, 7085–7094. 57. Jung, G. and Zumbusch, A. (2006) Improving autoﬂuorescent proteins: comparative studies of the effective brightness of green ﬂuorescent protein (GFP) mutants. Microsc. Res. Tech., 69, 175–185. 58. Keppler, A., Gendreizig, S., Gronemeyer, T. et al. (2003) A general method for the covalent labeling of fusion proteins with small molecules in vivo. Nat. Biotechnol., 21, 86–89. 59. Keppler, A., Pick, H., Arrivoli, C. et al. (2004) Labeling of fusion proteins with synthetic ﬂuorophores in live cells. Proc. Natl. Acad. Sci. U. S. A., 101, 9955–9959. 60. Bohme, I., Morl, K., Bamming, D. et al. (2007) Tracking of human Y receptors in living cells – a ﬂuorescence approach. Peptides, 28, 226–234. 61. Grifﬁn, B.A., Adams, S.R. and Tsien, R.Y. (1998) Speciﬁc covalent labeling of recombinant protein molecules inside live cells. Science, 281, 269–272.

REFERENCES

195

62. Grifﬁn, B.A., Adams, S.R., Jones, J. and Tsien, R.Y. (2000) Fluorescent labeling of recombinant proteins in living cells with FlAsH. Methods Enzymol., 327, 565–578. 63. McLean, A.J. and Milligan, G. (2000) Ligand regulation of green ﬂuorescent protein-tagged forms of the human β1 − and β2 -adrenoceptors; comparisons with the unmodiﬁed receptors. Br. J. Pharmacol., 130, 1825–1832. 64. Weiss, M., Hashimoto, H. and Nilsson, T. (2003) Anomalous protein diffusion in living cells as seen by ﬂuorescence correlation spectroscopy. Biophys. J., 84, 4043–4052. 65. Fradin, C., Abu-Arish, A., Granek, R. and Elbaum, M. (2003) Fluorescence correlation spectroscopy close to a ﬂuctuating membrane. Biophys. J., 84, 2030–2042. 66. Wohland, T., Rigler, R. and Vogel, H. (2001) The standard deviation in ﬂuorescence correlation spectroscopy. Biophys. J., 80, 2987–2999. 67. Middleton, R.J. and Kellam, B. (2005) Fluorophore-tagged GPCR ligands. Curr. Opin. Chem. Biol., 9, 517–525. An excellent summary of considerations in designing ﬂuorescent GPCR ligands. 68. Daly, C.J. and McGrath, J.C. (2003) Fluorescent ligands, antibodies, and proteins for the study of receptors. Pharmacol. Ther., 100, 101–118. A useful overview of ﬂuorescent GPCR ligands. 69. Rao, R., Langoju, R., Gosch, M. et al. (2006) Stochastic approach to data analysis in ﬂuorescence correlation spectroscopy. J. Phys. Chem. A, 110, 10674–10682. 70. Bates, I.R., Wiseman, P.W. and Hanrahan, J.W. (2006) Investigating membrane protein dynamics in living cells. Biochem. Cell Biol., 84, 825–831. 71. Chen, Y., Muller, J.D., So, P.T. and Gratton, E. (1999) The photon counting histogram in ﬂuorescence ﬂuctuation spectroscopy. Biophys. J., 77, 553–567. 72. Kask, P., Palo, K., Ullman, D. and Gall, K. (1999) Fluorescence-intensity distribution analysis and its application in biomolecular detection technology. Proc. Natl. Acad. Sci. U. S. A., 96, 13756–13761.

10 Identiﬁcation and Analysis of GPCR Phosphorylation Kok Choi Kong, Sharad C. Mistry and Andrew B. Tobin Department of Cell Physiology and Pharmacology, University of Leicester, Leicester, UK

10.1 Introduction G protein-coupled receptors (GPCRs) comprise ∼1% of the mammalian genome, encoded by ∼1000 genes [1]. They are involved in regulating a plethora of biological responses. Activated upon binding of an agonist, the response is switched off (desensitized) by receptor phosphorylation. Classically, GPCR phosphorylation is associated with receptor internalization and trafﬁcking, involving both second messenger-regulated protein kinases such as protein kinase C and protein kinase A and receptor-speciﬁc kinases of the G protein-coupled receptor kinase (GRK) family [2]. More recently, GPCR phosphorylation has also been shown to mediate coupling of the receptor to various signalling pathways, such as the mitogen-activated protein kinase pathway [3, 4]. Furthermore, recent studies from our laboratory have demonstrated that GPCRs can be phosphorylated in an agonist-dependent manner by a diverse range of kinase families, not just the GRK family, and that receptor phosphorylation by a deﬁned kinase determines a speciﬁc signalling outcome [5]. Conventionally, studies of phosphorylation of GPCRs generally require the puriﬁcation of the receptors, which can be problematic, as the receptors are usually present at very low levels and difﬁcult to solubilize [6]. The use of receptor-speciﬁc antibodies for immunoprecipitation has overcome this problem. This chapter describes the protocols successfully applied in our laboratory for the study of M3 -muscarinic receptor phosphorylation. The generation of epitope-tagged receptor and receptor-selective antibodies

G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

198

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

will be covered, as will radiolabelling cells with [32 P]orthophosphate, immunoprecipitation of the receptor and subsequent analysis of the phosphorylation signature of the receptor by two-dimensional (2D) phosphopeptide mapping. Phosphoamino acid (PAA) analysis and Edman degradation of isolated phosphopeptides, which can reveal information that can lead to the direct identiﬁcation of GPCR phosphorylation sites, will also be discussed. Although the protocols use the example of the M3 -receptor for illustrative purposes, they can be readily adapted to most other GPCRs.

10.2 Methods 10.2.1 Generation of anti-receptor antibodies for use in immunoprecipitation studies 10.2.1.1 Epitope tagging receptors The key to successful immunoprecipitation of any protein, but in particular of GPCRs, which are usually expressed at low levels and run as broad bands on sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis (PAGE) gels, is the quality of the antibody. Generating antisera to GPCRs has proven to be difﬁcult and certainly is a time-consuming and rather hit-and-miss process. There are now a number of commercially available receptor-speciﬁc antibodies; some of these can be excellent, but others can be totally inadequate. To circumvent the need for receptor-speciﬁc antibodies, many laboratories have used recombination cDNA techniques to add an epitope tag to the receptor either at the N- or C-terminus. Commercially available antibodies can then be used against the tag to immunoprecipitate or western blot the receptor, in addition to use in immunocytochemistry. Many different epitope tags can be engineered into recombinant receptors; the most commonly used are haemagglutanin (HA), FLAG , HIS, c-Myc, VSV-G, V5 and HSV, the sequences for which are given in Table 10.1.

Table 10.1 Examples of commonly used epitope tags and their sequences. Tag

Sequence

HA

YPYDVPDYA

FLAG

DYKDDDDK

HIS

HHHHHH

c-MYC

EQKLISEEDL

HSV

QPELAPEDPED

V5

GKPIPNPLLGLDST

VSV-G

YTDIEMNRLGK

199

10.2 METHODS (a)

YPYDVPDYA HA-tag NH2

IV V

III

II

VI

VII

Vasopressin −

+

I

COOH

(b) +

200− 100− 75− 50− 37−

Figure 10.1 Phosphorylation of the HA-tagged human vasopressin V1A -receptor expressed in Chinese hamster ovary (CHO) cells. (a) The human V1A -receptor was tagged with an HA epitope tag on the N-terminus. (b) This receptor was expressed stably in CHO cells which were labelled with [32 P]-orthophosphate and stimulated in the presence or absence of [Arg8 ]vasopressin (1 µm) for 5 min. The cells were then lysed in radioimmunoprecipitation assay (RIPA) buffer and the receptor immunoprecipitated using the 12CA5 monoclonal antibody. The phosphorylated form of the receptor can be seen as a broad band running at ∼90 kDa.

We have successfully used an HA-tagged V1a vasopressin receptor generated by Hawtin [7] in immunoprecipitation studies where the phosphorylation status of the receptor was determined (Figure 10.1). In these experiments, the commercially available mouse monoclonal anti-HA antibody (12CA5, Roche) was used to immunoprecipitate the receptor [7].

10.2.1.2 Generation and puriﬁcation of receptor-speciﬁc antibodies Whereas epitope tagging may seem a quick solution to the problems of a paucity of high-quality receptor-speciﬁc antibodies, there are a number of caveats. First, the addition of a tag either N- or C-terminus may affect the processing, trafﬁcking or signalling properties of the receptor. Second, the researcher is restricted to investigation of the recombinant receptor expressed from plasmid transfection or viral infection. It is not possible to study the native receptor in primary cell cultures or tissue preparations. These problems are circumvented by raising a receptor-speciﬁc antibody. We have experience with raising a number of antibodies to the M3 -muscarinic receptor

200

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

from glutathione-S-transferase (GST)-fusion proteins of a variant region of the third intracellular loop. Hence, Protocol 10.1 is based on the antisera we have raised against the region S344 –L462 of the mouse M3 -muscarinic receptor third intracellular loop. This region was cloned into a bacterial expression plasmid that produced an N-terminal tagged GST-receptor fusion protein which was used to inoculate New Zealand white rabbits (Harlan Sera-Lab). The resulting antisera were tested in immunoprecipitation studies and shown to react speciﬁcally with the mouse M3 -muscarinic receptor (Figure 10.2).

PROTOCOL 10.1 Preparation of Anti-receptor Antibodies Equipment and Reagents • Preparative ultracentrifuge (Beckman) • Sonicator • NZY media: (per litre) 5 g of NaCl, 2 g of MgSO4 ·7H2 O, 5 g of yeast extract, 10 g of N-Z Amine (casein hydrolysate), adjust pH to 7.5 with NaOH and add deionized H2 O to a ﬁnal volume of 1 l • Phosphate-buffered saline (PBS): 0.9% (w/v) NaCl in 10 mM K2 HPO4 /KH2 PO4 at pH 7.4. Adjust the pH of 10 mM K2 PO4 with 10 mM KH2 PO4 to pH 7.4 and dissolve 0.9 g per 100 ml of NaCl in the ﬁnal buffer • Bacteria lysis buffer: tris(hydroxymethyl)aminomethane hydrochloride (tris-HCl), 10 mM at pH 7.4; ethylenediaminetetraacetate (EDTA), 5 mM; glycerol 10%, dithiothreitol, 1 mM • Elution buffer: tris-HCl, 50 mM at pH 8.0; glutathione, 5 mM • 8 m urea (Fisher Scientiﬁc) • Isopropyl-β-D-thiogalactopyranosid (IPTG) (Sigma) • 6 ml glutathione-–sepharose afﬁnity column (GE Healthcare) • Complete adjuvant (Harlan Sera-Lab) • Incomplete adjuvant (Harlan Sera-Lab) • New Zealand white rabbits (Harlan Sera-Lab).

Method 1 Using standard polymerase chain reaction/cloning techniques, clone the coding region of interest into a bacterial expression plasmid that allows for the inducible production of a fusion protein that contains the region of the receptor to which the antibody will be raised and a tag that allows for puriﬁcation. There are many commercially available plasmids for this purpose. We have extensively used pGEX-2T from GE Healthcare (formerly Amersham Pharmacia), but other vectors that produced a maltose binding protein domain fusion, a HIS-tag fusion or a Nus-A fusion are equally as applicable. The

10.2 METHODS

201

example we give here is using pGEX-2T, which is one of the pGEX series of vectors. Mouse M3 -muscarinic receptor third intracellular loop (S344 –L462 ) was cloned into the BamHI/EcoRI sites in pGEX-2T (GE Healthcare) so that GST was fused at the C-terminal with the muscarinic receptor fragment. This construct drives the expression of the GST-receptor fusion protein from a Lac promoter that is inducible with the lactose derivative IPTG. 2 Once a bacterial transformant clone is generated, inoculate the NZY medium (400 ml) with a 5 ml saturated culture of the bacterial clone. 3 On reaching log phase (A600 = 0.6), add IPTG (0.1 mM ﬁnal) and continue the culture for 4 h at 30 ◦ C. 4 Harvest the bacterial pellet by centrifugation (12 000g for 10 min at 4 ◦ C). 5 Resuspend the pellet in 24 ml of bacterial lysis buffer and lyse by sonicating at 4 ◦ C with three 10 s pulses at the maximum setting of the microprobe. 6 Collect the bacterial occlusion bodies by centrifugation at 40 000g for 10 min (4 ◦ C) and dissolve the pellet in 8 m urea (20 ml) at room temperature for 15 min. 7 Clear the sample by centrifugation at 10 000g for 10 min at 4 ◦ C. 8 Dialysis of the supernatant against (i) 1 m urea 2–4 h, (ii) 0.1 m urea overnight, (iii) two changes of PBS over 4 h. 9 Pass the dialysed sample over a 6 ml glutathione–sepharose afﬁnity column equilibrated with PBS. 10 Elute the GST-fusion protein with 10 ml of elution buffer. 11 Check the purity of the GST-fusion protein by SDS–PAGE (see Protocol 10.3) and adjust protein concentration to 1 mg ml−1 . 12 Primary immunization: inject New Zealand white rabbits subcutaneously (SC) with 0.25 mg antigen in a volume of 0.5 ml with 0.5 ml of complete adjuvant. 13 Secondary immunizations are carried out every 14 days for ﬁve immunizations in total using 0.25 mg antigen in a volume of 0.5 ml with 0.5 ml of incomplete adjuvant injected SC. 14 Test bleeds are taken between each boost and the reactivity in western blots and immunoprecipitation tested against a pre-immune bleed control (see Protocol 10.3).

10.2.1.3 Validation of the receptor antibody by immunoprecipitation of the biotinylated receptor The receptor antibody can be used in western blots to determine speciﬁcity and can also be used in immunoprecipitation. To validate the antibody speciﬁcity in immunoprecipitation, we recommend biotinylation of all cell-surface proteins followed by preparation of a cell lysate from which the receptor is immunoprecipitated using the receptor-speciﬁc antisera. The receptor can then be detected using streptavidin

202

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION Cloning of M3-receptor 3iloop region into bacterial expression plasmid NH2 Purified GST - 3iloop fusion protein

III IV V

II VI

VII

I

75− 3iloop

50− 37−

COOH

lac

1

tac GST-coding sequence

3iloop/GST GST

25− 20−

BamHI (931) Aval (936)

laclq

Xmal (936) Smal (938) Eco RI (941)

GEX-2T 4948 bp

Apa U (3628)

2

Apa U (1472) Pst I (1902)

Rep Origin 1

Biotinylation immunoprecipitation M1

M2

M3

M4

M5 3

100 − 75 −

Antisera Production

Figure 10.2 Generation of an M3 -muscarinic receptor speciﬁc rabbit polyclonal antibody. (1) A region of the third intracellular loop of the mouse M3 -muscarinic receptor S344 –L462 was cloned into the bacterial expression plasmid pGEX-2T so that a fusion protein consisting of GST–3iloop was produced. A Coomassie-stained gel of puriﬁed GST and GST–3iloop proteins is shown. (2) The fusion protein was used to immunize New Zealand white rabbits. (3) To validate the quality of the antisera for immunoprecipitation, CHO cells expressing the mouse M1 –M5 muscarinic receptor subtypes were biotinylated, solubilized and the muscarinic receptors immunoprecipitated using the M3 -speciﬁc antisera. Equal amounts of muscarinic receptor were used in the immunoprecipitation as determined by radioligand binding. The antibody immunoprecipitated the M3 -muscarinic receptor but not any of the other muscarinic receptor subtypes.

10.2 METHODS

203

conjugated to horseradish peroxidase as described in Protocol 10.2. We prefer this method, since it mimics closely the protocol used for immunoprecipitation of receptors from radiolabelled cells.

PROTOCOL 10.2 Validation of Antibodies by Biotinylation Equipment and Reagents • Cell culture facilities • Scintillation counter (Wallac) • Biotin (Pierce Chemical Co.) dissolved in Krebs/ 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES) buffer • Streptavidin conjugated horseradish peroxidase (Pierce Chemical Co.) • Krebs/HEPES: HEPES, 10 mM at pH 7.4; NaCl, 118 mM; KCl, 4.3 mM; MgSO4 , 1.17 mM; CaCl2 , 1.3 mM; NaHCO3 , 25 mM; KH2 PO4 , 1.18 mM; glucose, 11.7 mM • RIPA buffer: Tris-HCl, 10 mM at pH 7.4; EDTA, 2 mM; β-glycerophosphate, 20 mM, NaCl, 160 mM; 1% Nonidet P-40 (NP-40) 0.5% deoxycholate • PBS: 0.9% (w/v) NaCl in 10 mM K2 HPO4 /KH2 PO4 at pH 7.4. Adjust the pH of 10 mM K2 PO4 with 10 mM KH2 PO4 to pH 7.4 and dissolve 0.9 g per 100 ml of NaCl in the ﬁnal buffer • Tris-buffered saline (TBS): tris(hydroxymethyl)aminomethane (tris-base) 10 mM at pH 7.4; NaCl 100 mM • [3 H]N-methyl scopolamine ([3 H]-NMS) (PerkinElmer) • Atropine (Sigma); prepare a 100 times stock solution with water • PROTRAN nitrocellulose membranes (Whatman) • ECL Plus kit (GE Healthcare).

Method 1 Grow the cells in a six-well plate to sub-conﬂuence (60–80%) and incubate in Krebs/HEPES buffer containing 1 mM biotin for 30 min at 37 ◦ C. 2 Wash the biotin from cells with 1 ml Krebs/HEPES buffer containing 100 mM glycine to quench further biotinylation. 3 Solubilize and immunoprecipitate the receptor as described in Protocol 10.3, Section B. 4 In parallel, measure M3 -muscarinic receptor levels by incubating cells contained on a duplicate six-well dish with Krebs/HEPES buffer containing a saturating concentration of [3 H]-NMS (0.5 nM) in the presence or absence of 10 µM atropine for 1 h at 37 ◦ C. Wash the cells three times with cold PBS and solubilize receptors with RIPA buffer before counting using a scintillation counter. The counts in the absence of atropine represent total [3 H]-NMS binding and the counts in the presence of atropine represent nonspeciﬁc

204

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

binding. Subtracting the total binding from the nonspeciﬁc binding gives the speciﬁc binding, from which the molar quantity of receptor/milligram of protein can be easily determined, since the speciﬁc activity of [3 H]-NMS is known. The amount of lysate used in the immunoprecipitation can then be adjusted so that an equal number of receptors are used. 5 Resolve the immunoprecipitate by SDS–PAGE (Protocol 10.3, Section C), transfer to nitrocellulose membranes and block with 5% non-fat dry milk in 0.1% Tween 20 in TBS (TBST) for 1 h at room temperature with gentle rocking. 6 Incubate membranes with 50 ng ml−1 streptavidin conjugated to horseradish peroxidase (Pierce Chemical Co.) for 30 min. 7 Detect biotinylated proteins with chemiluminescence reagent (ECL Plus, GE Healthcare).

10.2.2 Detecting phosphorylated receptor from [32 P]-Orthophosphate-labelled Cells We have carried out many experiments on transfected cell lines and native tissue (particularly cerebella granule neurones) where the M3 -muscarinic receptor was immunoprecipitated from cells labelled with [32 P]-orthophosphate. Protocol 10.3 describes this process, based on three simple steps: (i) radiolabelling of cells, (ii) solubilization and immunoprecipitation of the receptor and (iii) resolving the receptor by SDS–PAGE.

PROTOCOL 10.3 Detecting Phosphorylated Receptor from [32 P]-Orthophosphate-labelled Cells Equipment and Reagents • Cell culture facilities • Benchtop microcentrifuge with fridge (Eppendorf) • Shaking tray • Gel electrophoresis apparatus (BioRad) • Gel drier • Cold room • Phosphate-free Krebs/HEPES: HEPES, 10 mM at pH 7.4; NaCl, 118 mM; KCl, 4.3 mM; MgSO4 , 1.17 mM; CaCl2 , 1.3 mM; NaHCO3 , 25 mM; glucose, 11.7 mM • [32 P]-Orthophosphate was from GE Healthcare (cat. no. PBS11), at 10 mCi ml−1 • RIPA buffer: tris-HCl, 10 mM at pH 7.4; EDTA, 2 mM; β-glycerophosphate, 20 mM, NaCl, 160 mM; 1% Nonidet P-40 (NP-40) 0.5% deoxycholate • TEG buffer: tris-HCl, 10 mM at pH 7.4; EDTA, 2 mM; β-glycerophosphate, 20 mM

10.2 METHODS

• Protein A–Sepharose from GE Healthcare (cat. no. 17-0780-01), 1.5 g resuspended in 50 ml of TEG buffer • Laemmli buffer: tris-HCl, 125 mM at pH 6.8; β-mercaptoethanol, 10 mM; 4% SDS, 20% glycerol; 0.05% bromophenol blue • 30% acrylamide and 0.8%N,N-bis-methylacrylamide solution mix • 1.5 m tris-base solution at pH 8.8 • 10% SDS solution • 10% (w/v) ammonium persulfate (APS) • Tetramethylethylenediamine (TEMED) (Sigma) • 0.5 m tris-base solution at pH 6.8 • Isopropanol (Sigma) • Coomassie stain: 40% methanol, 10% acetic acid, 0.2% Coomassie brilliant blue R250 • Destain: 40% methanol, 10% acetic acid • Agonist at 100× stock in phosphate-free Krebs/HEPES • 1.5 ml microcentrifuge tubes • X-ray ﬁlm (GE Healthcare) • X-Ograph Compact X2 ﬁlm developer (Xograph).

Method A. [32 P]-Orthophosphate-labelling of Cells in Culture 1 Grown cells to 80–95% conﬂuency in six-well dishes and wash twice with 1 ml of phosphate-free Krebs/HEPES. Leave with 1 ml of phosphate-free Krebs/HEPES for 10 min to equilibrate. 2 Remove the buffer and label cells with 1 ml phosphate-free Krebs/HEPES containing 50–200 µCi ml−1 of [32 P]-orthophosphate for 60–90 mina at 37 ◦ C. 3 Stimulate cells by directly adding 10 µl of agonist made up as a 100 × stock in phosphate-free Krebs/HEPES. 4 Terminate stimulation by removing medium and lysing cells with 1 ml of cold RIPA buffer. B. Solubilization and Immunoprecipitation of Receptor All steps in this section should be carried out on ice or with a centrifuge precooled to 4 ◦ C. 5 Leave the cells on the plate/dish to lyse in cold RIPA buffer on ice for 10 min. 6 Wash the cells down on the plate/dish by pipetting up and down. Transfer everything from one well of the six-well dish in to a 1.5 ml microfuge tube. Screw-cap microfuge tubes can be used to reduce contamination of the microfuge.

205

206

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

7 Centrifuge at 16 000g for 5 min at 4 ◦ C. 8 Dilute antibody stock with TEG buffer, 1–5 µg of in-house anti-M3 -muscarinic receptor antibody (see Protocol 10.1) per 100 µl TEG buffer per sample. 9 Transfer 900 µl of supernatant into a fresh tube and add 100 µl of the above TEG buffer with antibody. 10 Leave on ice for 60–90 min (the supernatant can be left at 4 ◦ C overnight, but this is not our standard practice). 11 Add 180 µl of protein A–Sepharose slurry to each sample. Rock in 4 ◦ C room for 15 min. 12 Pellet down the protein A–Sepharose beads by centrifuging at 500g for 30 s at 4 ◦ C. 13 Carefully aspirate off the supernatant without disturbing the beads. Leave ∼100 µl of the supernatant in the tube as a precaution. 14 Wash beads three times with 1 ml of cold TEG buffer. Centrifuge at 500g for 30 s at 4 ◦ C each time to pellet down beads and aspirate off supernatant carefully without losing beads. It is better to have more wash steps than to lose sample by sucking up the protein A pellet. 15 During the ﬁnal wash, remove all the supernatant by aspiration with a ﬁne-tipped gel loading pipette. 16 Add 20 µl of Laemmli sample buffer to the protein A pellet (the sample can be stored at −20 ◦ C, if desired, overnight). 17 Mix samples by ﬂicking the tubes and then heating at 50–60 ◦ C for 2–3 min. Do not boil the samples, as the receptor may aggregate.b 18 Centrifuge the samples brieﬂy at high speed to pellet down the beads and load onto an SDS–PAGE (see Section C) gel. C. Resolving the Immunoprecipitated Receptor Using SDS–PAGE 19 Prepare an 8% polyacrylamide separating gel from a stock solution of 30% acrylamide and 0.8%N,N-bis-methylacrylamide mix diluted to the required concentration using 6.9 ml of distilled water, 3.8 ml of 1.5 m tris-base at pH 8.8, 150 µl of 10% (w/v) SDS, 150 µl of 10% (w/v) APS and 9 µl of TEMED, making up to a total volume of 15 ml. The latter two components are used to polymerize the gels and, therefore, are added last. 20 Layer the top of the gel with some isopropanol so that the gel sets evenly. Leave at room temperature for gel to set. 21 Rinse off isopropanol with distilled water. Prepare a 10-well, 4% stacking gel by mixing 750 µl of 30% acrylamide and 0.8%N,N-bis-methylacrylamide mix with 3.15 ml of distilled water, 1.25 ml of 0.5 m tris-base at pH 6.8, 50 µl of 10% (w/v) SDS, 50 µl of 10% (w/v) APS and 5 µl of TEMED. 22 After polymerization of the separating and stacking gels in the electrophoresis kit, ﬁll the equipment with electrophoresis buffer containing 25 mM tris-base, 190 mM glycine and 0.1% (w/v) SDS.

10.2 METHODS

207

23 Load the samples onto the gel and run at 200 V until the bromophenol blue front (diaphragm) reaches about 1 cm from the bottom of the gel (this usually take about 40–45 min). Prestained molecular weight markers should be run in parallel to facilitate the identiﬁcation of the receptor. 24 Cut off the stacking gel and stain the separating gel with Coomassie stain by rocking at room temperature for 15–30 min. 25 Destain the gel with destain buffer, rocking at room temperature for 60–90 min. 26 Dry the gel with a gel dryer for 40 min at 80 ◦ C. 27 Wrap gel with cling ﬁlm and expose to X-ray ﬁlm overnight at −80 ◦ C, depending on the amount of radioactivity on gel. 28 Develop the ﬁlm by feeding it into the Xograph Compact X2 developer in the darkroom. Notes a Although isotopic equilibrium point has not been reach with short-term labelling, longer labelling time was found to have no beneﬁcial effects for the studies of the phosphorylation of the receptor; rather, it produced detrimental effects on the cells. Thus, we ﬁnd 1–2 h labelling time is sufﬁcient for the purpose of the studies of receptor phosphorylation. b Owing to hydrophobic nature of GPCRs and most membrane proteins, boiling of samples for denaturation is not recommended, as this can cause receptor aggregation.

10.2.3 Biochemical approaches to analysis of [32 P]-labelled GPCRs Although mass spectrometry techniques are now being applied successfully to the identiﬁcation of phosphorylation sites on GPCRs [8–10], these are restricted by the amount of material that is required for such analysis. In Protocol 10.4, we use biochemical techniques that, rather than being limited by the amount of material, are limited by the amount of radiolabel that is incorporated into the receptor. Thus, provided that the receptor is well labeled, then the phosphopeptide mapping, Edman degradation and phospho-amino acid analysis described here can be carried out even though the actual molar quantities of the receptor protein are vanishingly low. The scheme of work shown can reveal an abundance of information that can lead directly to the determination of phosphorylation sites [5, 11] (Figure 10.3). First, the phosphopeptide map reveals the putative number of phosphorylation sites and gives what we have called a phosphorylation signature. By comparing the phosphorylation signature of the same receptor expressed in different tissues, we have revealed the diversity of receptor phosphorylation [5]. Edman degradation can be performed on a hot spot from the phosphopeptide map, where each cycle from the degradation is collected, dried and spotted onto ﬁlter paper and exposed in a phosphorimager to reveal whether that amino acid is phosphorylated.

208

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

In this way one can determine which position in the peptide is phosphorylated. PAA analysis reveals whether the spot is phosphorylated on a serine, threonine or tyrosine. By combining this data with the predicted proteolytic digest products, it is possible to determine which residue is phosphorylated (Figure 10.3).

A

I

II III IV V VI VII

In vivo Phosphorylation

3iloop

(−)

(+)

Protease

Chromatography

Immunoprecipitation

Phospho-peptide Map Electrophoresis

Protease Digest

15

Edman degradation

NNNDAAASLENSASSDEEDIGSETRAIY 1

2

3 4 5

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Figure 10.3 Generation of a phosphopeptide map and the Edman degradation of a speciﬁc spot derived from the in vivo phosphorylation of the human M3 -muscarinic receptor. Human M3 -muscarinic receptors expressed in CHO cells labelled with [32 P]-orthophosphate were immunoprecipitated and the immunoprecipitate resolved by SDS–PAGE. The gel was then transferred to nitrocellulose and the labelled receptor band cut out and digested with chymotrypsin. The peptides from the digest were resolved on a TLC plate, ﬁrst by electrophoresis and second by TLC. The gel was exposed in a phosphorimager to obtain a phosphopeptide map. A single phosphopeptide from the map was scrapped off and Edman degradation revealed that residue 15 on the peptide was phosphorylated. (Previous studies had determined that the receptor was only phosphorylated on serine residues.) The predicted chymotryptic peptide fragments showed that there is only one peptide were a serine residue exists in position 15. These data strongly suggest that this serine in position 15 is an in vivo phospho-acceptor site in the human M3 -muscarinic receptor.

10.2 METHODS

PROTOCOL 10.4 Biochemical Approaches to Analysis of [32 P]-labelled GPCRs Equipment and Reagents • Hunter thin-layer peptide mapping electrophoresis unit from CBS Scientiﬁc INC., Del Mar, CA, USA • STORM phosphorimager (GE Healthcare) • Sequelon-arylamine (AA) disk (Applied Biosystems, part number: POROSCUSTOM) • 1-(3-Dimethylaminopropyl)-3-ethyl-carbodiimide hydrochloride (EDC) (Fluka) • 2-(N-Morpholino)ethanesulfonic acid (MES) (Sigma–Aldrich) • Bio-Rad Mini-Blot transfer kit • SpeedVac (Eppendorf) • Fan • Drying oven • Transfer buffer: 25 mM tris-base and 190 mM glycine in 20% methanol • 0.5% polyvinylpyrrolidone K30 in 0.6% acetic acid • Trypsin solution: 1 µg of sequencing-grade modiﬁed trypsin from Promega (cat. no. V5111) dissolved in 50–150 µl freshly prepared 50 mM NH4 HCO3 solution • pH 1.9 buffer: 88% formic acid : acetic acid : water 25 : 78 : 897 (v/v/v) • pH 3.5 buffer: acetic acid : pyridine : water 100 : 10 : 1794 (v/v/v) • Isobutyric acid chromatography buffer: isobutyric acid : n− butanol : pyridine : acetic acid : water 1250 : 38 : 96 : 58 : 558 (v/v/v/v/v) • 30% formic acid (Sigma) • Triﬂuoroacetic acid (Sigma) • 6 m HCl • Asp-N or Glu-C protease (Sigma) • PAA standards: phosphoserine, phosphothreonine and phosphotyrosine (stock of 1.5 mg ml−1 of each; the stocks are stable for years at −20 ◦ C) • Ninhydrin (0.25% w/v in acetone) • Radioactive ink • PROTRAN nitrocellulose membrane (Whatman) • Thin-layer chromatography (TLC) plates from VWR Int. (20 cm × 20 cm, cat. no. 155232T) • 20 cm × 20 cm Whatman ﬁlter paper

209

210

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

• X-ray ﬁlm (GE Healthcare) and developing solutions (can be replaced by a phosphorimager) • Cling ﬁlm • 1.5 ml microcentrifuge tube.

Method A. Phosphopeptide Mapping of Immunoprecipitated [32 P]-labelled Receptor 1 Label, immunoprecipitate and separate the receptor as in Protocol 10.3. 2 After SDS–PAGE (Protocol 10.3, Section C), instead of staining and drying the gel, transfer the proteins on the gel to nitrocellulose membrane by using a Bio-Rad Mini-Blot transfer kit. Fill the tank with cold transfer buffer and transfer at 100 V for 1 h. Use the ice block provided and stir the buffer to prevent overheating. 3 Wrap the membrane in cling ﬁlm and expose to X-ray ﬁlm or a phosphorimager. The membrane must not be dried, as this will make the digestion impossible. An overnight ﬁlm exposure or 2–4 h on a phosphorimager should be sufﬁcient to visualize the bands. A ﬂuorescent ruler marker can be used as a guide to the position and orientation of the membrane. 4 Superimpose the autoradiogram or a 1 : 1 printout of the phosphorimager analysis on the nitrocellulose membrane and cut out the bands of interest. Contours of bands may be marked using a needle or a pen. Precision of the cuts should be veriﬁed by re-exposing the membrane. 5 Block the membrane pieces with 200 µl of 0.5% polyvinylpyrrolidone K30 in 0.6% acetic acid for 30 min at 37 ◦ C. 6 Aspirate off blocking solution and wash membrane pieces three times with distilled water. 7 Digest the proteins on membrane pieces by incubating overnight at 37 ◦ C in the smallest volume possible (50–150 µl) of trypsin solution. 8 Transfer the supernatant to a fresh tube and wash the membrane once or twice with water for 15–30 min with shaking at 1500 rpm at room temperature. Pool all the supernatants and dry down completely with a SpeedVac at room temperature (2–6 h). 9 If the membrane pieces still contain a high amount of radioactivity, then they can be digested again and the samples pooled. 10 Redissolve the dried pellet with 25–50 µl of pH 1.9 buffer and dry again with a SpeedVac. 11 Dissolve the pellet in 5–10 µl of pH 1.9 buffer, vortex intensely and centrifuge for 1 min. 12 Apply the supernatant in very small portions onto a cellulose TLC plate. Use a fan with no heating to dry the portions. Hot air would ‘bake’ the peptides to the plate and it is essential to get the smallest spot possible, as this will ensure generation of high-quality

10.2 METHODS

maps. The spot should be 2 cm from one end of an edge of the plate and 4 cm from the adjacent other. 13 Wet the plate with pH 1.9 buffer using a same size (20 cm × 20 cm) Whatman paper with a circular hole of 1 cm in diameter at the position of the phosphopeptide spot. Wet the areas around the sample ﬁrst. If done properly, all the buffer converges on the centre of the circle and acts to concentrate the spotted sample. Pat the paper over the rest of the area to make sure the entire surface of the plate is wet. The plate should look dull grey and not shiny because of too much buffer on the plate. Areas where the buffer has puddled should be blotted carefully with a paper towel. 14 Electrophorese for 30–40 min at 2000 V. 15 Air-dry the plate extensively in a fume hood. 16 Scrape off a lane of cellulose 3 cm from top of the plate using a scraper. 17 Run ascending chromatography overnight in isobutyric acid chromatography buffer. 18 Dry the plate extensively in a fume hood, label the corners with radioactive ink and wrap in cling ﬁlm and expose to ﬁlm or a phosphorimager. B. Edman Degradation of Hot Phosphopeptides Isolated from Phosphopeptide Maps 19 Scrape the spot(s) of interest on the 2D map identiﬁed by autoradiography from the TLC plate and transfer to a microfuge tube. 20 Extract two to three times with 200 µl of pH 1.9 buffer.a Pool all the supernatant. (There are no general rules that predict which buffer to use. pH 1.9 buffer is often used as a starting point because most peptides are soluble in it.) 21 10–20% of the samples can be reserved for PAA analysis, and if possible, 10–20% for secondary cleavage with Asp-N or Glu-C protease. 22 Before subjecting the remaining samples to Edman degradation (see steps 31–35), they need to be covalently attached to a Sequelon-AA disk. 23 Dry down the phosphopeptide samples in microfuge tubes using a SpeedVac. 24 Redissolve the phosphopeptide in 20 µl of 50% aqueous acetonitrile solution containing 0.1% triﬂuoroacetic acid (caution: very toxic and corrosive!). Heating the samples at 55 ◦ C with occasional mixing (vortex or sonication bath) may facilitate sample to dissolve. 25 Rest a sheet of Mylar (or glass slide) on heat block at 55 ◦ C and place a Sequelon-AA disk on top. 26 With a micropipette, apply the samples to the disk, a small drop at a time and allowing sample to dry in between. Then allow the disk to dry thoroughly (10–15 min). 27 Remove the Mylar (or glass slide) with the disk on top from the heat block, being careful not to touch the disk. 28 Just before use, weigh out approximately 1 mg of EDC (water-soluble carbodiimide) in a microfuge tube and add 100 µl of 0.1 M MES pH 5.0 containing 15% acetonitrile.

211

212

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

29 Carefully apply 5 µl of the freshly prepared EDC solution to the disk. Discard the remaining EDC solution. 30 Allow the reaction to proceed at room temperature for 20 min. Let the disk dry and then wash in 1 ml of 50% aqueous acetonitrile–0.1% triﬂuoroacetic acid. The disk is now ready for Edman degradation. 31 Edman degradation is performed in an Applied Biosystems Procise 49x protein sequencer. Place the Sequelon-AA dick in a vertical Blott cartridge of an Applied Biosystems 49x sequencer. An anilinothiazolinone (ATZ) amino acid collection chemistry cycle is used, which is a modiﬁcation of the standard pulsed liquid cycle as described by Campbell and Morrice [12]. 32 ATZ amino acid is extracted and transferred at each sequencing cycle using 90% methanol–10% water solution, which is placed in the X3 bottle position on the Procise. A transfer line (Applied Biosystems part number 602930) is attached to the ATZ/FC port (or port 39) of the Procise and connected to the arm of an external fraction collector. The fraction collector can be interfaced electronically to the Procise and advanced by introducing a Relay 1 pulse function (function 253) into the sequencing program or operated by time-based collection (setting a time which is equal to the total time of a sequencing cycle). The ATZ amino acid is transferred to the fraction collector with two extractions, of 0.2 ml with a 20 s wait between extractions. 33 Collect each cycle of degradation and dry with a SpeedVac. 34 Dissolve in 5–10 µl of 50% aqueous acetonitrile–0.1% triﬂuoroacetic acid and apply onto a sheet of 3 mM Whatman ﬁlter. A square grid (10 mM × 10 mM) is marked out in pencil on the ﬁlter paper; at the centre of each square, a 2–3 mM diameter circle is marked where each ATZ amino acid fraction is applied in a tight spot. 35 Expose for 2–3 days on a phosphorimager. Erase the screen twice to obtain a really low background on the screen. Alternatively, autoradiographing on Hyperﬁlm (GE Healthcare) with an intensifying screen can be performed, but a much longer exposure is needed (1–2 weeks), as the sensitivity is decreased. C. PAA Analysis 36 PAAs can be analysed by scraping off spot(s) of interest on the 2D map identiﬁed by autoradiography from the TLC plate and extracted the same way as described in Section B of this protocol. Alternatively, tryptically digested peptides collected from nitrocellulose membranes as described in Section A of this protocol can be dried and used as a start material in the PAA analysis. 37 Dry the samples and redissolve in 100 µl of 6 M HCl. 38 Incubate at 110 ◦ C for 1 h to hydrolyse the proteins/peptides. 39 Cool on a bench to room temperature. 40 Dilute samples with 400 µl of distilled water, mix and then dry with a SpeedVac (4–6 h). 41 Add 400–500 µl of distilled water and centrifuge for 5 min. 42 Transfer 90% of the supernatant to new tubes, carefully avoiding any insoluble material and dry again to get rid of the salt and residual HCl.

REFERENCES

213

43 Dissolve samples in 5–10 µl of pH 1.9 buffer and add 1.5 µg (i.e. 1 µl) of each of the nonradioactive PAA standards. 44 Apply the samples in a similar way as described in Section A of this protocol onto a TLC plate and dry every application with a cold-air fan. Four samples can be analysed simultaneously on a 20 cm × 20 cm cellulose TLC plate. Refer to Boyle et al. [13] for detailed instructions and outlines of TLC plates. 45 Wet the entire TLC plate with pH 1.9 buffer as described in Section A of this protocol using Whatman paper. 46 Connect anode to the left and cathode to the right and run ﬁrst dimension in pH 1.9 buffer at 2000 V for 20 min. 47 Air-dry the TLC plates in a fume cupboard. 48 Wet the plate again with Whatman paper, but this time with pH 3.5 buffer, leaving out the areas where the phosphopeptides have run in the ﬁrst dimension. 49 Turn plate at 90◦ to the left and run the second dimension in pH 3.5 buffer at 1800 V for 20 min. 50 Dry plate with a dryer (hot air fan or oven). 51 Spray with ninhydrin and heat until amino acid markers are stained. Distinct purple spots will appear corresponding to the position of the standards: phosphoserine (top left), phosphothreonine (middle) and phosphotyrosine (bottom). 52 Expose to phosphorimager for 1–3 days. 53 Identify the phosphorylated amino acid(s) by overlaying a 1 : 1 phosphorimager printout with the TLC plate and comparison with the stained standards. Notes a If extraction of peptide is of poor efﬁciency, an alternative method is to extract with 30% formic acid or isobutyric acid chromatography buffer.

Acknowledgements The authors would like to thank the Wellcome Trust for their support (grant 047600).

References 1. Torrecilla, I. and Tobin, A.B. (2006) Co-ordinated covalent modiﬁcation of G-protein coupled receptors. Curr. Pharm. Des., 12 (14), 1797–1808. 2. Pitcher, J.A., Freedman, N.J. and Lefkowitz, R.J. (1998) G protein-coupled receptor kinases. Annu. Rev. Biochem., 67, 653–692. 3. Budd, D.C., Willars, G.B., McDonald, J.E. and Tobin, A.B. (2001) Phosphorylation of the Gq/11-coupled M3-muscarinic receptor is involved in receptor activation of the ERK-1/2 mitogen-activated protein kinase pathway. J. Biol. Chem., 276 (7), 4581–4587.

214

CH 10 IDENTIFICATION AND ANALYSIS OF GPCR PHOSPHORYLATION

4. Rakhit, S., Conway, A.M., Tate, R. et al. (1999) Sphingosine 1-phosphate stimulation of the p42/p44 mitogen-activated protein kinase pathway in airway smooth muscle. Role of endothelial differentiation gene 1, c-Src tyrosine kinase and phosphoinositide 3-kinase. Biochem. J., 338 (Pt 3), 643–649. 5. Torrecilla, I., Spragg, E.J., Poulin, B. et al. (2007) Phosphorylation and regulation of a G protein-coupled receptor by protein kinase CK2. J. Cell Biol., 177 (1), 127–137. Demonstrates the identiﬁcation of phosphorylation sites in M3 -muscarinic receptors by using the methods described here. 6. Tobin, A.B. (1997) Protocols employed in the investigation of G protein-coupled receptor phosphorylation. Methods Mol. Biol., 83, 227–234. 7. Hawtin, S.R. (2005) Charged residues of the conserved DRY triplet of the vasopressin V1a receptor provide molecular determinants for cell surface delivery and internalization. Mol. Pharmacol., 68 (4), 1172–1182. Demonstrates the use of an HA-tagged GPCR. 8. Karoor, V. and Malbon, C.C. (1996) Insulin-like growth factor receptor-1 stimulates phosphorylation of the β2 -adrenergic receptor in vivo on sites distinct from those phosphorylated in response to insulin. J. Biol. Chem., 271 (46), 29347–29352. 9. Papac, D.I., Oatis, J.E. Jr, Crouch, R.K. and Knapp, D.R. (1993) Mass spectrometric identiﬁcation of phosphorylation sites in bleached bovine rhodopsin. Biochemistry, 32 (23), 5930–5934. 10. Trester-Zedlitz, M., Burlingame, A., Kobilka, B. and von Zastrow, M. (2005) Mass spectrometric analysis of agonist effects on posttranslational modiﬁcations of the β − 2 adrenoceptor in mammalian cells. Biochemistry, 44 (16), 6133–6143. The ﬁrst demonstration of phospho-aceptor site on the β2 -adrenergic receptor using mass spectrometry. 11. Blaukat, A., Pizard, A., Breit, A. et al. (2001) Determination of bradykinin B2 receptor in vivo phosphorylation sites and their role in receptor function. J. Biol. Chem., 276 (44), 40431–40440. Demonstrates the identiﬁcation of phosphorylation sites in bradykinin B2 receptors by using the methods described here. 12. Campbell, D.G. and Morrice, N.A. (2002) Identiﬁcation of protein phosphorylation sites by a combination of mass spectrometry and solid phase Edman sequencing. J. Biomol. Tech., 13 (3), 119–130. 13. Boyle, W.J., van der Geer, P. and Hunter, T. (1991) Phosphopeptide mapping and phosphoamino acid analysis by two-dimensional separation on thin-layer cellulose plates. Methods Enzymol., 201, 110–149. Provides detailed instruction and outlines for TLC plates, as well as useful information for the protocols.

11 Measurement and Visualization of G Protein-coupled Receptor Trafﬁcking by Enzyme-linked Immunosorbent Assay and Immunoﬂuorescence Stuart J. Mundell, Shaista P. Nisar and Eamonn Kelly Department of Physiology & Pharmacology, University of Bristol, Bristol, UK

11.1 Introduction Internalization (also termed endocytosis) of G protein-coupled receptors (GPCRs) is a major mechanism for regulating the signalling levels achieved by this receptor protein superfamily [1–5]. The temporal and spatial aspects of internalization have been investigated for many GPCRs [1]. GPCRs are targeted to discrete sites on the plasma membrane from where they are internalized [6]. Internalization can serve multiple functions [3], including sequestration, which is deﬁned as rapid and reversible removal of the receptor from the cell surface. Internalization is also associated with the processes of desensitization [7] and resensitization [1] for many GPCRs. In addition, internalization is often a ﬁrst step in GPCR downregulation, where the internalized GPCR is targeted to lysosomes or proteosomes where subsequent degradation can occur. Finally, internalization can also lead to signal generation through vesicular signalling scaffolds [3]. In the past, internalization of cell-surface receptors was detected primarily by the use of radioligands [8, 9]. This method does allow internalization of endogenously expressed GPCRs to be determined, provided that the receptors are expressed at a G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

216

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

high enough concentration. Although still widely applicable today [10], this approach can also be limited by the availability of radioligands of high enough receptor afﬁnity or speciﬁcity. In the early 1990s, the use of ﬂuorescently labelled ligands allowed for a more direct observation of internalization [11]. More recently, the use of GPCR–green ﬂuorescent protein (GFP) fusions has been particularly useful because, with the aid of confocal microscopy, the ﬂuorescent signal can be followed into endosomal compartments [12, 13]. The use of GPCR–GFP fusions provided an advantage over the ﬂuorescently labelled ligands, primarily because the ligand can become detached from the receptor upon encountering the acidic environment of endosomal compartments, thus potentially preventing an accurate assessment of receptor localization [11–13]. It is generally assumed that the GPCR–GFP fusion protein trafﬁcs like the original GPCR; whilst this seems to be the case generally, the assumption does not always hold true [14]. Colourless substrate

Yellow product

Secondary antibody Anti HA antibody HA

HA

Figure 11.1 How the ELISA works. In intact cells the epitope-tagged GPCRs are expressed at the cell surface and the primary antibody (in grey; anti-HA in this example) can bind to the HA epitope in the N-terminus of the GPCR. The secondary antibody, which binds to the primary, is coupled to alkaline phosphatase such that when the substrate is added a yellow colour will develop in the medium. The degree of yellow colour development will depend upon the amount of secondary antibody bound and, hence, relates to the number of epitope-tagged GPCRs at the cell surface. The primary antibody cannot access GPCRs that have been internalized, so the technique provides a reliable measure of GPCR internalization in response to agonist.

11.2 METHODS AND APPROACHES

217

In this chapter we will discuss another method for examining GPCR trafﬁcking: enzyme-linked immunosorbent assay (ELISA) of cell-surface epitope-tagged GPCRs. Although this approach does require the initial generation of an epitope-tagged GPCR, data generation using this strategy is both rapid and robust. Clearly, this approach is limited, in that the tagged GPCR needs to be subsequently expressed in a cell line or tissue of interest, and it cannot be used for endogenously expressed GPCRs unless a high-afﬁnity antibody for the extracellular N-terminus of the GPCR exists, which is usually not the case. In order for the outlined protocols to work the receptor must be epitope tagged on its extracellular N-terminus, and the precise positioning of the epitope can be extremely important. For many GPCRs, simply placing the tag at the very start of the N-terminus sequence will yield a workable construct [15, 16]. For some GPCRs, however, Golgi export and/or cell-surface localization sequences are present in the N-terminus. In this case, the epitope must be carefully positioned within the N-terminus in order to allow expression of receptor at the cell surface [17]. There are a number of different choices of epitope, including haemagglutinin (HA) [17], FLAG [18] and myc [19]. Although these epitopes appear to give broadly comparable results for internalization assays, the FLAG epitope is also useful for measuring the reappearance of internalized GPCRs at the cell surface, as FLAG antibody bound to cell-surface GPCRs can be removed by washing cells in a Ca2+ -free buffer [20]. Once an epitope-tagged receptor construct has been generated, speciﬁc ELISA methods [15, 17, 21] allowing the quantiﬁcation of receptor internalization and receptor recycling can be used. In addition the use of immunoﬂuorescence to determine the endocytic trafﬁcking route taken by a GPCR is possible [16, 17]. Protocols outlining each of these methods are described in this chapter. A combination of these methodologies should allow the reader to quantify the kinetics of receptor internalization and recycling and successfully determine the endocytic trafﬁcking route taken by a particular GPCR.

11.2 Methods and approaches 11.2.1 Loss of surface GPCR by ELISA ELISA requires the expression of an epitope-tagged GPCR in the cell line of choice. In Protocol 11.1 we assume the use of transiently transfected HEK 293 cells with HA-tagged GPCR constructs. The user also can generate cell lines stably expressing receptor construct to undertake such studies. GPCR-expressing cells are exposed to agonist and surface expression is subsequently measured in ﬁxed cells using a primary antibody to the epitope tag on the GPCR followed by a secondary antibody conjugated to alkaline phosphatase. Levels of alkaline phosphatase activity can then be measured by absorbance using a colorimetric assay. Internalized GPCRs are not accessible to the antibodies, and so agonist-induced GPCR internalization will reduce the amount of primary and, hence, secondary antibody that binds at the cell surface, thus reducing the alkaline phosphatase activity, seen as reduced colour development in the assay (Figure 11.1). The use of a plate reader to measure colour development means that many factors, such as time-dependency and agonist concentration dependency of internalization, can be investigated in the same assay. The user should note

218

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

that this technique, although it examines total surface receptor loss, does not measure internalization in isolation, since receptor recycling as well as insertion of newly synthesized receptors into the membrane may also contribute to changes in surface receptor levels. However, pharmacological tools which inhibit protein synthesis and recycling will allow the user to gauge the relative contribution of these processes to changes in cell-surface receptor expression and, hence, provide a more accurate measure of receptor internalization.

PROTOCOL 11.1 Measurement of Surface Receptor Loss Equipment and Reagents • Rocking platform • Plate reader (e.g. Dynex MRX Revelation) • Cell culture hood and incubator • Transfection reagent (e.g. Lipofectamine (Invitrogen)) • Dulbecco’s modiﬁed Eagle’s medium (DMEM) supplemented with and without 10% foetal bovine serum (FBS). • Poly-L-lysine (0.1 mg ml−1 ) (Sigma) • 24-well and 96-well plates • 3.7% formaldehyde in tris-buffered saline (TBS) • TBS: 20 mM tris(hydroxymethyl)aminomethane (tris), pH 7.5, 150 mM NaCl • Bovine serum albumin (BSA) (Sigma) • Primary antibody (see note d below) • Secondary antibody conjugated to alkaline phosphatase (Sigma) • Alkaline phosphatase kit (Biorad) • 0.4 M NaOH • Appropriate agonist.

Method 1 Unless stably transfected cell lines are used, transiently transfect the receptor construct by the method of choice into the cell line of choice and leave for at least 18 h.a 2 Treat a 24-well plate with poly-L-lysine (0.1 mg ml−1 ) for 2 min and allow to dry for at least 30 min in a tissue culture hood. 3 Split the cells from the Petri dish evenly among the wells and incubate overnight at 37 ◦ C in DMEM + 10% FBS.b 4 Replace the media on cells with 0.5 ml of prewarmed (to 37 ◦ C)c DMEM alone (i.e. nothing added). Incubate at 37 ◦ C for 15 min.

11.2 METHODS AND APPROACHES

219

5 Add the agonist at intervals to give incubation times of, for example, 2 h, 1 h, 30 min and 10 min, as well as no agonist as a control. 6 Remove the media and ﬁx the cells by addition of 250 µl 3.7% formaldehyde for 5 min. Do not exceed time. 7 Wash the cells three times by addition and aspiration of 500 µl of TBS. 8 After aspirating the ﬁnal TBS wash, block the cells for 45 min at room temperature with 250 µl of 1% BSA in TBS. 9 Make a dilution of the primary antibodyd in TBS and replace the blocking solution in each well with 250 µl of this solution. Incubate on a rocking platform at room temperature for 1 h. 10 Wash the cells three times with 500 µl TBS. 11 Remove the ﬁnal TBS wash and block the cells for 15 min at room temperature with 250 µl 1% BSA in TBS. 12 Make a dilution of the secondary antibodye in TBS and replace the blocking solution in each well with 250 µl of this solution. 13 Wash the cells three times with 500 µl TBS. 14 Prepare a ‘developing solution’ from the alkaline phosphatase kit by combining p-nitrophenyl phosphate tablets, 5× diethanolamine buffer and double-distilled H2 O in a ratio of one tablet : 1 ml : 4 ml (e.g. for 2 × 24-well plates, use three tablets : 3 ml : 12 ml). Vortex well to dissolve tablets. Remove the ﬁnal TBS wash from the cells and add 250 µl of developing solution to each well. 15 Incubate the plates at 37 ◦ C until a colour change in the developing solution is evidentf (i.e. a deeper yellow colour develops). 16 When the developing solution has turned a yellow colour, transfer 100 µl of the solution from each well to the wells of a 96-well plate (for reading in the plate reader) containing 100 µl of 0.4 M NaOH to terminate the colour reaction. 17 Measure the absorbance of the samples at 405 nM using a plate reader. The user should be aiming for readings between 0.8 and 1.5 in untreated cells. The user should also run control cells that are vector expressing alone in order to determine background readings.g Notes we transfect 5 µg of epitope-tagged GPCR DNA into a 100 mM plastic Petri dish of subconﬂuent (70–90% conﬂuent) cells. In addition, cells transfected with plasmid vector alone are required for background readings in the assay. In our hands, although it is perfectly acceptable to use cells transiently expressing the epitope-tagged GPCR construct of interest, stable expression generally gives more consistent results in the ELISA. In addition, stable expression can reduce expense, since less transfection reagent is required, although a number of weeks are required to generate stably transfected cells. On the other hand, transient expression will be easier if, for example, the internalization of a number of GPCR mutants needs to be compared.

a Typically,

220

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

b We ﬁnd that a 100 mM dish of HEK 293 cells can be split evenly amongst a 24-well plate to provide an adequate ELISA signal the following day; however, this will vary depending on cell type. Too many cells per well may make it difﬁcult to detect internalization, whereas too few cells will give a poor signal. It may be advisable to test a range of cell densities to begin with (e.g. 1 × 105 − 4 × 104 cells per well for a 24-well plate). c Unless

stated otherwise, all solutions used are at room temperature.

d

The dilution of the primary antibody is epitope speciﬁc; for HA we use monoclonal anti-HA 11 from Covance at 1 : 1000; for FLAG we use monoclonal M2 anti-FLAG from Sigma at 1 : 500. e Since we use a monoclonal primary antibody, we use an anti-mouse immunoglobulin G alkaline phosphatase, for example, from Sigma at 1 : 1000. Incubate on a rocking platform at room temperature for 1 h. f

The speed of colour change is dependent on receptor expression, as well as on the quality of the antibodies. In most cases, a developing time of between 5 and 45 min is required. It is important for reproducibility that the reaction is stopped in the linear phase of colour development. If necessary, an initial experiment can be undertaken to determine this by stopping the reaction at various times of colour development and measuring absorbance.

g

In a typical experiment using one 24-well plate, 21 of the wells would contain epitope-tagged GPCR-transfected cells whilst the other three wells would contain plasmid vector-transfected cells for background readings (Figure 11.2). Of the wells containing epitope-tagged GPCR-transfected cells, three wells would be for control receptor levels in the absence of drug, allowing up to six time-points or drug concentrations if triplicates are used.

Control

20 min

5 min

30 min

10 min

60 min

15 min

Background

Figure 11.2 An example of an experimental setup for ELISA in a 24-well plate. The white wells all contain cells expressing the epitope-tagged GPCR, with the control not being treated with agonist. The background wells in grey contain cells transfected with the plasmid vector used for the GPCR construct. The background readings must be subtracted from all the other readings. For examples of experimental results for this assay, see [14, 16] for example.

11.2.2 Measurement of receptor recycling In Protocol 11.2 the user is able to measure receptor recycling following agonistinduced internalization. In addition, this method allows the detection of constitutive internalization of GPCR if present. Brieﬂy, cells expressing epitope-tagged GPCR are prelabelled at 4 ◦ C with antibody against the epitope on the GPCR. At this temperature

11.2 METHODS AND APPROACHES

221

all endocytic events are blocked. Following removal of unbound antibody, cells are rapidly warmed to 37 ◦ C and agonist added. Following agonist addition, cell-surface antibody is stripped, leaving only internalized, receptor-bound antibody. In Protocol 11.2 using FLAG-tagged GPCR, we make use of the Ca2+ dependence of FLAG-M1 antibody binding to the FLAG epitope [20, 22] and remove surface antibody by a wash in a Ca2+ -free buffer. In addition, it may also be possible to use a low-pH wash to remove surface-bound antibody (not just FLAG-M1), in the same way that a low-pH wash is used to remove surface-bound peptide ligands for GPCRs [23]. However, we have not used this approach and will not discuss it further here. Following removal of surface antibody, cells are then left in the absence of agonist and receptor recycling back to the cell surface is monitored by ELISA. This strategy also allows the user to estimate the potential of a receptor to downregulate, since any receptor not recycled is likely to be targeted for degradation. Using this technique it is also possible to measure constitutive receptor internalization by examining changes in surface receptor levels in cells left at 4 ◦ C versus those heated up to 37 ◦ C but not exposed to agonist.

PROTOCOL 11.2 Receptor Recycling Equipment and Reagents • Transfection reagent (e.g. Lipofectamine (Invitrogen)) • DMEM supplemented with 10% FBS • DMEM supplemented with 1% BSA • Poly-L-lysine (0.1 mg ml−1 ) • 24-well and 96-well plates • 3.7% formaldehyde in TBS • TBS (20 mM tris, pH 7.5, 150 mM NaCl) supplemented with either 1 mM CaCl2 or 1 mM ethylenediaminetetraacetate (EDTA) • BSA • Rocking platform • Primary antibody (anti-FLAG M1 monoclonal; Sigma) • Secondary antibody conjugated to alkaline phosphatase (Sigma) • Alkaline phosphatase kit (Biorad) • Plate reader (e.g. Dynex MRX Revelation) • 0.4 M NaOH • Appropriate agonist.

222

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

Method 1 Follow steps 1–3 of Protocol 11.1. 2 Make a dilution of the primary antibody in ice-cold DMEM containing 1% BSA. If using anti-FLAG M1 monoclonal antibody,a dilute 1 : 500. Note that this antibody requires the presence of 1 mM CaCl2 (present in DMEM) in order to bind to the FLAG epitope. Wash the cells twice with ice-cold DMEM–BSA and then incubate with 250 µl DMEM containing primary antibody (1 : 200) at 4 ◦ C for 1 h. 3 Wash off unbound antibody using three washes with 500 µl TBS + 1 mM CaCl2 . 4 Warmb the cells to 37 ◦ C and add agonist to promote receptor internalization for, for example, 30–60 min. 5 Remove surface-bound anti-FLAG antibody using two washes with TBS + 1 mM EDTA to ensure Ca2+ -free conditions. Important: if using anti-FLAG M1 antibody, make sure Ca2+ -containing solutions are used from now on. 6 Remove the Ca2+ -free wash, add 500 µl TBS (+1 mM CaCl2 ) and warm to 37 ◦ C. 7 Remove the TBS at different time points (e.g. 10, 30, 60 min) and ﬁx the cells in 250 µl 3.7% formaldehyde for 5 min. Do not exceed this time. 8 Follow steps 7–8 on ELISA Protocol 11.1. 9 Follow steps 12–17 on ELISA Protocol 11.1.c Notes binding of the anti-FLAG M1 monoclonal antibody to its epitope is Ca2+ dependent; do not confuse this antibody with the anti-FLAG M2 antibody mentioned in Protocol 11.1, whose binding to the epitope is Ca2+ independent.

a The

b We either do this by adding the agonist and immediately placing the tray in the CO incubator 2 at 37 ◦ C, or by adding the agonist and immediately placing the wells carefully in a water bath at 37 ◦ C. Placing the cells in the incubator means that it will take them a few minutes to warm up fully; but, if, for example, comparing drug-treated with untreated on the same plate, then the rate of warming is probably not crucial. On the other hand, placing the tray in a water bath on the laboratory bench means that they will warm to 37 ◦ C faster than in an incubator, but they will not be subjected to the levels of CO2 they normally experience. In our experience, either way appears to give satisfactory and reproducible results. c For

some GPCRs (e.g. the P2Y1 and P2Y12 purinergic receptors) we obtain around 50–75% recycling of receptor within 1 h. If recycling is not observed, then further experiments to determine the fate of the internalized receptors can be undertaken (e.g. the visualization protocol below or by radioligand binding or western blotting).

11.2.3 Preparation of cells for visualization of receptor trafﬁcking Immunoﬂuorescence is a technique that allows the visualization of a particular protein, such as a GPCR, in cells by binding a speciﬁc antibody chemically conjugated with a

11.2 METHODS AND APPROACHES

223

Table 11.1 Markers of endosomal compartments. Endocytic compartment

Marker

Early endosomes

Transferrin, EAA1, Rab5

Sorting endosomes

Transferrin, Rab5

Recycling endosomes

Rab4 and Rab11

Late endosomes

Rab7, LAMP1

Lysosomes/proteosomes

Lysotracker red, LAMP1

Apart from lysotracker red and transferrin, antibodies are commercially available to selectively label these endocytic marker proteins and are added to ﬁxed, permeabilized cells. On the other hand, lysotracker red (a ﬂuorescent dye that labels acidic compartments) and rhodamine-conjugated transferrin (rhodamine being a ﬂuorescent dye) are both added to live cells.

ﬂuorescent dye such as ﬂuorescein isothiocyanate (FITC). There are two major types of immunoﬂuorescence staining method: (i) direct immunoﬂuorescence staining, in which the primary antibody is labelled with ﬂuorescent dye [16]; (ii) indirect immunoﬂuorescence staining, in which a secondary antibody labelled with ﬂuorochrome is used to recognize a primary antibody [16]. Clearly, the choice between direct and indirect immunoﬂuorescence will be dependent upon the availability of the epitope-recognizing antibody conjugated to ﬂuorescent dye. Using immunoﬂuorescence, the user should readily be able to visualize the endocytic route taken by a GPCR by examining the co-localization of the GPCRs and known endocytic markers. There are a large number of endocytic markers that are commercially available (see Table 11.1). Although this list below is by no means exhaustive, it should allow the user to choose suitable endocytic marker proteins to examine the route of GPCR internalization [24]. Protocol 11.3 describes how cells are prepared for indirect immunoﬂuorescence so that the pathway of GPCR internalization can be examined. Brieﬂy, epitope-tagged GPCR-expressing cells grown on coverslips are prelabelled at 4 ◦ C with antibody against the epitope on the GPCR. Following removal of unbound antibody, cells are warmed to 37 ◦ C and agonist added. Cells are then ﬁxed and permeabilized in order to allow access to either secondary antibody conjugated to a ﬂuorescent moiety or to allow a primary antibody to label an endocytic compartment. Note that there must be a species difference (e.g. mouse monoclonal versus rat polyclonal) between antibody labelling of the GPCR and the endocytic compartment in order to allow secondary antibody discrimination. Equally, the secondary antibody of each species must be conjugated to a different ﬂuorescent moiety; for example, monoclonal conjugated to ﬂuorescein (green) versus polyclonal conjugated to rhodamine (red). Following antibody labelling and washes, cells are again ﬁxed and coverslips mounted for immunﬂuorescent microscopy. The actual microscopy is beyond the scope of this chapter and, anyway, to some extent depends upon the microscope system used. However, if you have access to an imaging facility, there should be expertise available to help you successfully visualize the GPCRs and the subcellular compartments through which they trafﬁc.

224

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

PROTOCOL 11.3 Preparation of Cells for Visualization of Internalized GPCRs by Microscopy Equipment and Reagents • Transfection reagent (e.g. Lipofectamine (Invitrogen)) • DMEM supplemented with 10% FBS • DMEM supplemented with 1% and 0.5% BSA • Glass coverslips (e.g. 60 mm diameter) • Poly-L-lysine (0.1 mg ml−1 ) • Six- or 12-well plates • 30 ml centrifuge tube (e.g. Nalgene) • 3.7% formaldehyde in TBS • TBS (20 mM tris, pH 7.5, 150 mM NaCl) • BSA • 0.05% Triton-X-100 in phosphate-buffered saline (PBS) • 5% dried milk in 0.05% Trition-X-100–PBS • Agonist solution • Rocking platform • Primary antibodies as appropriate for epitope-tagged GPCRs and endocytic marker proteins as described in Table 11.1. Appropriate secondary antibodies. • Lysotracker red (Invitrogen) • Rhodamine-conjugated transferrin (Invitrogen) • 1,4-Diazabicyclo-[2,2,2]-octane (DABCO) (Sigma) • Mowiol mounting medium (Calbiochem)b • Microscope system; for example, we use an upright Leica TCS-NT confocal laser-scanning microscope attached to a Leica DM IRBE epiﬂuorescence microscope with phase contrast and a Plan-Apo 40 × 1.40 numerical aperture oil-immersion objective. • Bench-top microcentrifuge • Hotplate stirrer.

Method 1 Day 1. Unless stably transfected cell lines are used, transiently transfect the receptor construct by the method of choice into the cell line of choice and leave for at least 18 h. Typically, we transfect 3 µg receptor DNA into a 60 mM dish of subconﬂuent cells.

11.2 METHODS AND APPROACHES

2 Day 2. Place a coverslip in each well of a six- or 12-well plate and coat with 0.1 mg ml−1 poly-L-lysine. Remove poly-L-lysine and allow coverslips to dry for at least 30 min. 3 Split the cells evenly among the wells and incubate overnight at 37 ◦ C in DMEM + 10% FBS. We ﬁnd that the 60 mM dish of HEK 293 cells can be split amongst the wells of a six-well plate to provide adequate cells for the following day.a 4 Day 3. Remove the media from the wells and wash the cells twice with 1 ml PBS. 5 Add 1 ml of primary antibody (against the GPCR epitope) diluted 1 : 200 in ice-cold DMEM + 1% BSA and incubate the cells for 1 h at 4 ◦ C. 6 Remove the primary antibody solution and wash the cells twice with 1 ml ice-cold PBS. 7 Add 1 ml DMEM + 0.5% BSA to each well and incubate cells with or without agonist at 37 ◦ C for the appropriate amount of time. NB: the endocytic markers rhodamineconjugated transferrin (1 µM) and lysotracker red (0.1 µM) can be added at this stage, along with agonist. Under these conditions, in live cells the lysotracker red and rhodamine-conjugated transferrin will accumulate in their respective intracellular compartments. 8 Remove the agonist-containing media and wash cells twice with 1 ml PBS. 9 Fix the cells to stop further agonist-induced activation by adding 1 ml 3.7% formaldehyde in PBS and incubating at room temperature for 30 min. 10 Remove formaldehyde and wash the cells twice with 1 ml PBS. 11 Permeabilize the cells by incubation with 1 ml 0.05% Triton–PBS for 10 min at room temperature. 12 Block the cells by incubation with 5% dried milk in 0.05% Triton–PBS for 30 min at 37 ◦ C. 13 If using a primary antibody to label an endocytic compartment, add the antibody at this stage and incubate for 1 h. Then wash three times with 1 ml PBS. 14 To prepare the secondary antibody, ﬁrst dilute an appropriate volume in 1 ml 5% dried milk in 0.05% Triton–PBS in a microcentrifuge tube and spin in a microcentrifuge at 14 000 rpm for ∼30 s (to remove unconjugated ﬂuorophore and, hence, reduce subsequent background ﬂuorescence). Remove the supernatant to a 30 ml centrifuge tube and dilute further to give a ﬁnal dilution of 1 : 200, antibody : 5% dried milk in 0.05% Triton–PBS. If using two secondary antibodies, these can be added together and should not cross-react if from different species. 15 Add 1 ml secondary antibody solution to each well and incubate at 37 ◦ C for 1 h. NB: the procedure is now photosensitive, so try to minimize exposure to light, by putting samples in a box for example. 16 Remove antibody solution and wash the cells ﬁve times in 0.05% Triton–PBS. 17 Leave the last wash on for 30 min at 37 ◦ C. 18 Remove the wash and ﬁx the cells a second time by incubation with 1 ml 3.7% formaldehyde in PBS at room temperature for 10 min.

225

226

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

19 Remove formaldehyde and wash the cells twice with 1 ml PBS. 20 Prepare a DABCO/Mowiol solution by adding a small amount of DABCO (∼2.5% v/v) to a 1 ml aliquot of Mowiol and place a drop on a microscope slide.b 21 Remove coverslips containing cells from the PBS and place upside down over the drop of DABCO/Mowiol on the slide. 22 Store the slides in the dark at 4 ◦ C until ready to perform confocal microscopy. 23 Image within 7 days to limit fading.c Notes Make sure cells are not overconﬂuent on day of assay (ideally should be ∼30–50%), as this will make visualization of individual cells difﬁcult.

a

b

Mowiol is an aqueous mounting medium compatible with immunoﬂuorescence. To prepare, add 2.4 g Mowiol to 6 g glycerol and stir. Add 6 ml water and leave at room temperature for several hours. Add 12 ml 0.2 M tris (pH 8.5) and heat to 50 ◦ C for 10 min, with occasional mixing. Once dissolved, clarify by centrifugation at 5000 g for 15 min. Aliquot and store at −20 ◦ C until required [25]. DABCO is added to Mowiol in order to prevent quenching of ﬂuorescence. Add a small amount (∼2.5%) to Mowiol solution just prior to mounting of coverslips.

c FITC

and tetramethyl rhodamine isothiocyanate (TRITC) have overlapping excitation–emission spectra; therefore, when dual staining with these ﬂuorophores, it is advisable to use separate FITC and TRITC ﬁlters, as opposed to a joint FITC–TRITC ﬁlter, to avoid bleed through from adjacent channels.

11.3 Troubleshooting • Poor triplicates in the ELISA: First, ensure that the cells are properly dispersed in the medium when pipetting into wells, and pipette cell suspension randomly into wells to avoid systematic error; if necessary, a protein assay can be undertaken to check how evenly the cells are being seeded into different wells. Second, take care washing cells on the 24-well plates, as overvigorous washing can cause cells to detach. If necessary, direct the washing solutions onto the side of the well to ensure gentler washing of cells. Third, avoid overgrowth of the cells in the wells, as this causes them to be detached more easily during washes. • High background readings in the ELISA: Under normal conditions absorbance values of 0.7–1.2 should be obtained, being equal to the control cell reading minus the background reading. High backgrounds may be due to not enough washes, so increase washes from three times to four. Otherwise, reduce primary and secondary antibody dilutions. • Low control readings in the ELISA: This is usually due to poor plasma membrane expression of the epitope-tagged GPCR, and so transfection efﬁciency should be checked by immunoﬂuorescence microscopy. In addition, check that the pH of wash solutions is correct and also ensure that the ﬁxation time does not exceed 5 min.

REFERENCES

227

Low control readings may in some cases be remedied by increasing the secondary antibody concentration. • Lack of labelled cells in immunoﬂuorescence microscopy: Check transfection efﬁciency. Also check that the microscope has lasers/ﬁlter to recognize ﬂuorophore. • High background in immunoﬂuorescence microscopy: Possibly not enough washes, so increase the number of washes. Otherwise, reduce the primary and secondary antibody dilutions.

References 1. Ferguson, S.S. (2001) Evolving concepts in G protein-coupled receptor endocytosis: the role in receptor desensitization and signalling. Pharmacol. Rev., 53, 1–24. Comprehensive review of GPCR trafﬁcking. 2. von Zastrow, M. (2003) Mechanisms regulating membrane trafﬁcking of G protein-coupled receptors in the endocytic pathway. Life Sci., 74, 217–224. 3. Drake, M.T., Shenoy, S.K. and Lefkowitz, R.J. (2006) Trafﬁcking of G protein-coupled receptors. Circ. Res., 99, 570–582. 4. Moore, C.A., Milano, S.K. and Benovic, J.L. (2007) Regulation of receptor trafﬁcking by GRKs and arrestins. Annu. Rev. Physiol., 69, 451–482. 5. Hanyaloglu, A.C. and von Zastrow, M. (2008) Regulation of GPCRs by endocytic membrane trafﬁcking and its potential implications. Annu. Rev. Pharmacol. Toxicol., 48, 537–568. Good review of the molecular mechanisms that underlie GPCR internalization and recycling. 6. Yudowski, G.A., Puthenveedu, M.A. and von Zastrow, M. (2006) Distinct modes of regulated receptor insertion to the somatodendritic plasma membrane. Nat. Neurosci., 9, 622–627. 7. Beaumont, V., Hepworth, M.B., Luty, J.S. et al. (1998) Somatostatin receptor desensitization in NG108-15 cells: a consequence of receptor sequestration? J. Biol. Chem., 273, 33174–33183. 8. Hertel, C., Coulter, S.J. and Perkins, J.P. (1985) A comparison of catecholamine-induced internalization of beta-adrenergic receptors and receptor-mediated endocytosis of epidermal growth factor in human astrocytoma cells. Inhibition by phenylarsine oxide. J. Biol. Chem., 260, 12547–12553. 9. Harden, T.K., Petch, L.A., Traynelis, S.F. and Waldo, G.L. (1985) Agonist-induced alteration in the membrane form of muscarinic cholinergic receptors. J. Biol. Chem., 260, 13060–13066. 10. Lawrence, J., Mundell, S.J., Yun, H. et al. (2005) Centaurin-a1, an ADP-ribosylation factor 6 GTPase activating protein, inhibits b2-adrenoceptor internalization. Mol. Pharmacol., 67, 1822–1828. 11. Daly, C.J. and McGrath, J.C. (2003) Fluorescent ligands, antibodies, and proteins for the study of receptors. Pharmacol. Ther., 100, 101–118. 12. Barak, L.S., Zhang, J., Ferguson, S.S. et al. (1999) Signaling, desensitization, and trafﬁcking of G protein-coupled receptors revealed by green ﬂuorescent protein conjugates. Methods Enzymol., 302, 153–171. 13. Kallal, L. and Benovic, J.L. (2000) Using green ﬂuorescent proteins to study G protein-coupled receptor localization and trafﬁcking. Trends Pharmacol. Sci., 21, 175–180.

228

CH 11 MEASUREMENT AND VISUALIZATION OF G PROTEIN-COUPLED RECEPTOR TRAFFICKING

14. Madziva, M. and Edwardson, M.J. (2001) Trafﬁcking of green ﬂuorescent protein-tagged muscarinic M4 receptors in NG108-15 cells. Eur. J. Pharmacol., 428, 9–18. 15. Mundell, S.J., Matharu, A.L., Kelly, E. and Benovic, J.L. (2000) Arrestin isoforms dictate differential kinetics of A2B adenosine receptor trafﬁcking. Biochemistry, 39, 12828–12836. 16. Mundell, S.J., Luo, J., Benovic, J.L. et al. (2006) Distinct clathrin-coated pits sort different G protein-coupled receptor cargo. Trafﬁc, 7, 1420–1431. 17. Mundell, S.J., Matharu, A.L., Pula, G. et al. (2001) Agonist-induced internalization of the metabotropic glutamate receptor 1a is arrestin- and dynamin-dependent. J. Neurochem., 78, 546–551. 18. Walker, J.K.L., Premont, R.T., Barak, L.S. et al. (1999) Properties of secretin receptor internalization differ from those of the b2 -adrenergic receptor. J. Biol. Chem., 274, 31515–31523. 19. Self, T.J., Oakley, S.M. and Hill, S.J. (2005) Clathrin-independent internalization of the human histamine H1-receptor in CHO-K1 cells. Br. J. Pharmacol., 146, 612–624. 20. Parent, J.L., Labrecque, P., Driss Rochdi, M. and Benovic, J.L. (2001) Role of the differentially spliced carboxyl-terminus in TXA2 receptor trafﬁcking: identiﬁcation of a distinct motif for tonic internalization. J. Biol. Chem., 276, 7079–7085. 21. Daunt, D.A., Hurt, C., Hein, L. et al. (1997) Subtype-speciﬁc intracellular trafﬁcking of α2 -adrenergic receptors. Mol. Pharmacol., 51, 711–720. Original description of ELISA used to quantify GPCR trafﬁcking. 22. Einhauer, A. and Jungbauer, A. (2001) Afﬁnity of the monoclonal antibody M1 directed against the FLAG peptide. J. Chromatogr. A, 921, 25–30. 23. Briscoe, C.P., Plevin, R. and Wakelam, M.J.O. (1994) Rapid desensitization and resensitization of bombesin-stimulated phospholipase D activity in Swiss 3T3 cells. Biochem. J., 298, 61–67. 24. Seachrist, J.L. and Ferguson, S.S. (2003) Regulation of G protein-coupled receptor endocytosis and trafﬁcking by Rab GTPases. Life Sci., 74, 225–235. 25. Harlow, E. and Lane, D. (1999) Using Antibodies: A Laboratory Manual, Cold Spring Harbour Press, Cold Spring Harbor, NY, pp. 183–184.

12 Substituted Cysteine Accessibility Method (SCAM) George Liapakis1 and Jonathan A. Javitch2 1 Faculty 2

of Medicine, University of Crete, Crete, Greece Center for Molecular Recognition, Columbia University, New York, USA

12.1 Introduction The substituted cysteine accessibility method (SCAM) [1] provides an approach to map systematically the residues on the water-accessible surface of a protein. These residues are identiﬁed by substituting them with cysteine (Cys), and assessing for the reaction of charged, hydrophilic, sulfhydryl reagents with the substituted Cys (engineered Cys). By applying SCAM, channel-lining residues in a variety of ion channels and transporters have been mapped, including the nicotinic acetylcholine receptor [1–3], the GABAA receptor [4, 5], the cystic ﬁbrosis membrane-spanning conductance regulator [6], the UhpT transporter [7] and potassium channels [8]. SCAM has also been used to investigate structural alterations in different functional states of proteins, such as the β2 -adrenergic receptor (see Section 12.2.7) Using this approach we mapped systematically residues on the surface of the binding-site crevice in the dopamine D2 receptor, a member of the G protein-coupled receptor (GPCR) superfamily [9–15]. The binding-site crevice of D2 and the other GPCRs in the rhodopsin-like subfamily is a water-accessible crevice formed among their seven, mostly hydrophobic, membrane-spanning segments (TMx, x = 1 − 7) and extending from the extracellular surface of the receptor into the membrane-spanning domain [16]. The surface of this crevice is formed by residues that can contact speciﬁc agonists and/or antagonists (binding-site residues) and by other residues that may play a structural role and affect binding indirectly. G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

230

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

A different approach for the determination of binding-site residues is site-directed mutagenesis. 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 on protein synthesis, folding, processing and structure. In contrast to typical mutagenesis experiments, SCAM does not rely on the functional effects of a given mutation, but rather 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. Residues that are likely to be very near the binding site could also be identiﬁed in afﬁnity labelling experiments. However, this approach has many disadvantages, including 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. In contrast to afﬁnity labelling, SCAM can be applied systematically to any binding site, and because the engineered Cys is known, the labelled residue does not need to be laboriously identiﬁed with protein chemical techniques. Other advantages of SCAM include the ability to probe binding sites by assessing the ability of agonists or antagonists to retard the reaction of sulfhydryl reagents with a particular substituted Cys (see Section 12.2.4.5) and the ability to probe the steric constraints and electrostatic potential of sites by comparing the rates of reaction using reagents of varying size and charge (see Section 12.2.6).

12.2 Methods and approaches 12.2.1 The substituted Cys accessibility method and its assumptions SCAM provides an approach to map systematically residues of the membrane-spanning regions, lying on the water-accessible surface of the binding-site crevice of GPCRs. To perform SCAM, every residue in a membrane-spanning segment is mutated, one at a time, to Cys (engineered Cys). The mutant receptors are then expressed in heterologous cells and assessed as to whether the substituted Cys residues react with small, hydrophilic, charged sulfhydryl 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) [17]. These reagents differ somewhat in size, with MTSET > MTSES> MTSEA. The MTS reagents form mixed disulﬁdes with the Cys sulfhydryl, covalently + − linking –SCH2 CH2 X, where X is NH+ 3 , N(CH3 )3 , or SO3 (see Figure 12.1). The MTS reagents are speciﬁc for Cys sulfhydryls and do not react with disulﬁde-bonded Cys or with other residues. In order to interpret the results of SCAM, we make a number of assumptions. Given that, in membrane-spanning segments, the sulfhydryl of a Cys can face into the binding-site crevice, into the interior of the protein or into the lipid bilayer, we ﬁrst assume that the highly polar MTS reagents react much faster with sulfhydryls lying at the water-accessible surface of a protein than those facing into lipid bilayer or into protein interior. They react with the ionized thiolate (–S− ) more than a billion times faster

231

12.2 METHODS AND APPROACHES

O

X = NH3+ (MTSEA) N(CH3)3+ (MTSET) − SO3 (MTSES)

O S

X S

S−

S

X S

N H

N H O

O

Figure 12.1 Reaction of an MTS reagent with a Cys sulfhydryl at the water-accessible surface of a receptor.

than with the un-ionized thiol (–SH) [18]. Furthermore, the MTS reagents are very hydrophilic, with a relative solubility in water : octanol greater than 2500 : 1 [1, 18]. Experimental support for the validity of this assumption comes from a study in the aspartate receptor of the accessibility of engineered Cys to reaction with another hydrophilic, sulfhydryl-speciﬁc alkylating reagent. [19]. In the α2 helix of the periplasmic domain, a striking correlation was observed between the measured chemical reactivity of each engineered Cys and the calculated solvent accessibility of the β carbon at the corresponding position in the crystal structure. We further assume that, in membrane-spanning segments, the access of highly polar reagents to side chains is only through the binding-site crevice, that the addition of –SCH2 CH2 X to a Cys at the surface of the binding-site crevice is likely to alter binding irreversibly (see Sections 12.2.2 and 12.2.4.3), and, reciprocally, that agonists and antagonists should retard the reaction of the MTS reagents with substituted Cys that line the binding site (see Sections 12.2.2 and 12.2.4.5). As mentioned above, to perform SCAM, residues in the membrane-spanning segments are mutated to Cys (engineered Cys), one at a time, and the mutant receptors are assessed as to whether the substituted Cys residues react with MTS reagents. We assume that the engineered Cys is an accurate reporter for the water accessibility of the corresponding wild-type residue. If ligand binding to a Cys-substitution mutant is near normal and, thus, Cys-substitution mutant is functional, its overall three-dimensional structure is likely to be similar to the structure of the wild-type receptor and the substituted Cys will lie in a similar orientation to that of the wild-type residue. In general, Cys substitution is remarkably well tolerated (see Section 12.2.4.2). Nonetheless, local changes at the site of the engineered Cys could, in principle, alter the accessibility of the residue relative to the accessibility of the wild-type residue. A strength of SCAM results from studying entire membrane-spanning segments where regular patterns of accessibility can be identiﬁed. Given the general consistency of the results obtained in the various receptors and channels studied to date, it is likely that, in most cases, the position of the Cys residue is similar to that of the wild-type residue it replaced. In several cases, irregular patterns have been observed. We cannot be certain whether

232

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

the secondary structure in such regions is irregular in the native structure, whether the protein structure ﬂuctuates in such a region alternately exposing multiple residues or whether the Cys substitution has disrupted the local secondary structure, making the Cys accessible when the wild-type residue is not.

12.2.2 Detection of reaction of MTS reagents with engineered Cys Reaction can be detected either 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 upon the direct detection of reaction. Instead, we use the irreversible modiﬁcation of function to assay the reaction. In a receptor, the reaction of an MTS reagent with an engineered Cys in the binding-site crevice should alter binding irreversibly (Figure 12.2). Additionally, reaction with a Cys near the binding site should be retarded by the presence of bound ligand. The functional effect of the addition of –SCH2 CH2 X to the engineered Cys could be a result of steric block, electrostatic interaction or indirect structural changes. Regardless, although we do not know the detailed mechanism of the alterations in binding, an irreversible effect is evidence of reaction and, therefore, of the accessibility of the engineered Cys. Although reaction usually inhibits binding, it can also potentiate binding. This can be illustrated in the dopamine D2 receptor by the mutation of Asp-108. [10]. Mutation to Cys of this residue at the extracellular end of the third

–SH

–SSEtX –SH

–SH

ligand

−SH

–SSEtX –SH binding

–SH no binding

Figure 12.2 Schematic representation of the reaction of an MTS reagent with a Cys exposed in the binding-site crevice. The membrane is represented by the shaded rectangle, the binding-site crevice by the white area within the plane of the membrane, and ligand by the solid oval. SEtX of + an MTS reagent (MTSEtX) represents SCH2 CH2 X, where Et is CH2 CH2 and X is the NH+ 3 , N(CH3 )3 or − SO3 . SEtX is covalently linked to the water-accessible Cys sulfhydryl. In the bound state (lower left panel), ligand is reversibly bound at the binding site within the binding-site crevice. In the unbound state (upper left), the binding site is unoccupied. After irreversible reaction with MTSEtX (upper right), ligand can no longer bind (lower right). The Cys-sulfhydryl-facing lipid or the interior of the protein does not react signiﬁcantly with MTSEtX. MTSEtX only reacts with a sulfhydryl in the binding-site crevice of ligand-free receptor. Thus, ligand binding retards the rate of reaction of receptor with MTSEtX and protects subsequent ligand binding.

12.2 METHODS AND APPROACHES

233

membrane-spanning segment reduced the afﬁnity of the receptor for antagonist binding about threefold. Reaction of the positively charged MTSEA or MTSET at this position inhibited binding signiﬁcantly. In contrast, reaction of the negatively charged MTSES restored the negative charge at this position and shifted the afﬁnity towards that of the 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 that results from an increase in ligand afﬁnity could be missed by measuring binding at too high a ligand concentration relative to the Kd . If an MTS reagent has no effect on a mutant, then interpretation of the results must be made with caution. The temptation is to infer that the engineered Cys is inaccessible to the MTS reagents and, therefore, is not on the water-accessible surface of the protein. While this is the most likely explanation, there are other possibilities. First, electrostatic or steric factors may alter the reactivity of the MTS reagents with a water-accessible residue. Second, while it seems unlikely that a residue forming the surface of the binding-site crevice could be modiﬁed covalently by the addition of the charged –SCH2 CH2 X without interfering with binding, such a result is nonetheless possible. In the dopamine D2 receptor, we have observed that the reaction of MTSEA at certain positions has a much greater effect on the binding of particular ligands; for example, reaction of MTSEA with Cys118, the highly reactive endogenous Cys, causes a negligible decrease in the afﬁnity of the receptor for particular ligands but a large decrease in its afﬁnity for other ligands [20]. To decrease the likelihood of such a false-negative determination, we typically screen for effects with antagonists from two different structural classes. Alternatively, the inability of reaction of an MTS reagent with a substituted Cys on the water-accessible surface of a receptor to produce a change in the binding of an antagonist could be to due to the inability of the antagonist to detect conformational changes of receptor associated with MTS reaction. In such a case, direct radiolabelled agonist binding or agonist competition of radiolabelled antagonist binding may be a more sensitive indicator of a disruption of receptor structure and, thus, a more sensitive assay of reaction. Indeed, we have identiﬁed an endogenous Cys in the β2 -adrenergic receptor that reacts with MTSEA without altering antagonist binding (Liapakis and Javitch, in preparation). We were able to detect reaction In this case by an alteration in the afﬁnity of the agonist isoproterenol in competition with radiolabelled antagonist. All the above potential complications further demonstrate the importance of systematically mutating to Cys consecutive residues along an entire membrane-spanning segment; while mutation of any individual residue might be subject to potential pitfalls due to steric or electrostatic factors or silent reaction, this is unlikely to be a systematic problem affecting the overall pattern of accessibility of multiple residues in a membrane-spanning segment.

12.2.3 Endogenous Cys Before applying SCAM, it must be ascertained that binding to wild-type receptor is not altered by addition of the appropriate sulfhydryl reagents. For example, the binding of [3 H]CGP-12177 to wild-type β2 -adrenergic receptor is not altered after

234

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

MTSEA reaction. Nevertheless, when one uses a receptor with endogenous Cys as a background construct, one must be alert to the possibility that a new engineered Cys might produce its effect, not because the engineered Cys is accessible, but rather because of a mutation-induced alteration of the accessibility of an endogenous Cys, thereby resulting in a false-positive determination of accessibility. Such a scenario has been observed in the serotonin transporter [21]. In contrast to β2 -adrenergic receptor, the binding of N -[3 H]methylspiperone to D2 receptor is affected by MTS reagents. In such a case, the ideal starting point would be to create a Cys-less protein with normal expression and function. Such a construct has been possible with the lactose permease [22], the NhaA Na+ /H+ antiporter [23] and a glutamate transporter [24], but a dopamine D2 receptor with all ﬁve putative membrane-spanning Cys simultaneously substituted by other residues expressed too poorly for further study. Nonetheless, in the dopamine D2 receptor, replacement of a single endogenous Cys (Cys118) with serine resulted in a 100-fold decrease in the reactivity of the receptor with MTSEA and MTSET [9]. Thus, C118S mutant was used as the background for further Cys substitutions, given that this receptor expresses normally and has unaltered binding properties [10–15].

12.2.4 Application of the substituted Cys accessibility method To determine residues that form the surface of the binding-site crevice of a GPCR by applying the SCAM, we: 1 perform mutagenesis experiments in order to substitute Cys, one at a time, for residues in a putative membrane-spanning segment (thus generating Cys mutants) and express Cys mutant receptors in heterologous cells; 2 determine the effect of Cys mutations on the functional properties of receptor; 3 determine the effect of MTS reagents on ligand binding to Cys mutant receptors; 4 determine the rate of reaction, if MTS reagents react by altering ligand binding; 5 test for the ability of ligand to retard the rate of reaction of an MTS reagent at the reactive positions (protection of substituted Cys by bound ligand).

12.2.4.1 Generate Cys mutant receptors and express them in heterologous cells To generate a Cys-substitution mutant (e.g. the S200C mutant of the β2 -adrenergic receptor) by site-directed mutagenesis we use the polymerase chain reaction with Pfu polymerase and with oligonucleotides containing the desired mutation (Ser200Cys in the sixth membrane-spanning segment, TM6, of β2 -adrenergic receptor) as well as an altered restriction site, which facilitates screening and veriﬁcation of mutants after subcloning. Subsequently, we screen mutants by restriction mapping and conﬁrm mutations by DNA sequencing. Next, we express the receptor (S200C mutant) in heterologous cells (see Protocol 12.1) and harvest the cells (see Protocol 12.2) for ligand binding, MTS reaction and protection experiments (see Sections 12.2.4.2–12.2.4.5).

12.2 METHODS AND APPROACHES

PROTOCOL 12.1 Transfection of HEK 293 Cells Equipment and Reagents • Dulbecco’s modiﬁed Eagle’s medium (DMEM) and nutrient mix F12 (1 : 1), containing 3.15 g l−1 glucose (Invitrogen) • Bovine calf serum (BCS) (Hyclone) • Lipofectamine (Invitrogen) • OPTIMEM (Invitrogen).

Method • Grow human embryonic kidney cells (HEK 293) in DMEM–F12 and 10% BCS at 37 ◦ C and 5% CO2 . • Transfect HEK 293 cells at 60–80% conﬂuence in 35 mM dishes with 2–3 µg of plasmid DNA containing the sequence encoding the S200C mutant of β2 -adrenergic receptor, using 5 µl of lipofectamine and 1 ml of OPTIMEM. • After 5 h, replace the OPTIMEM with DMEM–F12 containing 10% BCS and keep cells in culture at 37 ◦ C and 5% CO2 24 h before harvesting them.

PROTOCOL 12.2 Harvesting Cells Equipment and Reagents • Ca2+ - and Mg2+ -free Dulbecco’s phosphate-buffered saline (CMF-DPBS) • CMF-DPBS containing 1 mM ethylenediaminetetraacetate (EDTA) (CMF-DPBS–EDTA) • Binding buffer (140 mM NaCl, 5.4 mM KCl, 1 mM EDTA, 0.006% bovine serum albumin (BSA),a 25 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES), pH 7.4).

Method 1 Wash cells, 24 h after transfection, with 5 ml CMF-DPBS. 2 Aspirate the solution and treat cells for ∼30 s with 1 ml CMF-DPBS–EDTA. 3 Aspirate the solution and tap the plate several times on the bench top to loosen cells. 4 Pipette cells up and down about ﬁve times in 5 ml CMF-DPBS to dissociate. 5 Pellet cells for 5 min at 1000g, 4 ◦ C.

235

236

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

6 Resuspend cell pellet in 400 µl binding bufferb for ligand binding studies (see Protocol 12.3), or treatment with MTS reagents (see Protocol 12.4), or protection experiments (see Protocol 12.5).

Notes a

Store binding buffer without BSA for up to 1 month at 4 ◦ C. Add BSA fresh before each use.

b Adjust,

if required, the amount of cells used (by suspending whole cells in more or less than 400 µl binding buffer) to provide an adequate signal and to ensure that the speciﬁc binding (see Protocols 12.3, 12.4 and 12.7) is always equal to or less than 10% of the total concentration of the added radioligand, thus avoiding ligand depletion. The adjustment of cell density depends upon the level of receptor expression.

12.2.4.2 Determination of the effect of mutation of a residue to Cys on the functional properties of the receptor To determine the effect of a Cys substitution on the functional properties of receptor, we determine the afﬁnity of the Cys mutant receptor for the radiolabelled ligand from competition studies in intact, dissociated cells (see Protocol 12.3). The afﬁnity of a Cys mutant must be similar (within a few fold) to that of the wild-type receptor. For example, the afﬁnity of S200C mutant is similar to that of wild-type β2 -adrenergic receptor. This allows the assumption that the structure of the mutant is similar to wild -type and that the substituted Cys is an accurate reporter of the accessibility of the wild-type side chain, as mentioned in Section 12.2.1.

PROTOCOL 12.3 Ligand Binding Equipment and Reagents • Binding buffer (140 mM NaCl, 5.4 mM KCl, 1 mM EDTA, 0.006% BSA, 25 mM HEPES, pH 7.4) • Washing buffer (120 mM NaCl, 10 mM tris(hydroxymethyl)aminomethane hydrochloride (Tris-HCl), pH 7.4 at 4 ◦ C) • Radioligand, [3 H]CGP-12177 (NEN Life Science Products) • Brandel cell harvester • Liquid scintillation counter • Curve-ﬁtting package (e.g. Graphpad Prism; GraphPad Software, San Diego, CA) • Polypropylene minitubes • Whatman 934AH glass-ﬁbre ﬁlters, particle retention 1.5 µm (Brandel) • Alprenolol (Sigma–Aldrich)

12.2 METHODS AND APPROACHES

237

• Isoproterenol (Sigma–Aldrich) • Cell pellet prepared from Protocol 12.3.

Method • Suspend whole cells from a 35 mM dish in 400 µl binding buffer (see Protocol 12.2). • Add into polypropylene minitubes (triplicates), in a ﬁnal volume of 0.5 ml, 300 µl of a 20-fold dilution (in binding buffer) of cell suspension and 200 µl binding buffer containing 0.7 nM [3 H]CGP-12177 with 0 nM (total binding), 0.1 nM, 1 nM, 10 nM, 100 nM, 1 µM, 10 µM isoproterenol (binding in the presence of displacer) and 1 µM alprenolol (nonspeciﬁc binding). • Incubate the mixtures at room temperature for 60 min and then ﬁlter through Whatman 934AH glass-ﬁbre ﬁlters, using a Brandel cell harvester • Wash the ﬁlters three times with 1 ml of washing buffer at 4 ◦ C. • Quantify the radioactivity (and hence binding) on the washed ﬁlters using liquid scintillation spectrometry. • Determine the speciﬁc binding of [3 H]CGP-12177 by subtracting the nonspeciﬁc binding from total binding or from the binding in the presence of displacer. • Analyse the data by nonlinear regression analysis using GraphPad Prism and obtain IC50 values by ﬁtting the data to a one-site competition model. Determine the afﬁnity of isoproterenol binding (Ki value) using the equation Ki = IC50 /(1 + L/Kd ), where L is the concentration of radioligand and Kd is the afﬁnity of radioligand.

The ability, therefore, to substitute Cys residues for other residues and still obtain functional receptor is central to this approach. In the seven membrane-spanning segments of the dopamine D2 receptor, 144 of 157 Cys-substitution mutants bound antagonist with near-normal afﬁnity properties [10–15]. Only seven mutants had greater than a ﬁvefold reduction in afﬁnity, and only four of these seven bound antagonist with greater than a 10-fold reduction in binding. Three of these four mutations were of presumed contact residues. The tolerated substitutions were for hydrophobic residues (alanine, leucine, isoleucine, methionine and valine), polar residues (asparagine, serine and threonine), neutral residues (proline), acidic residues (aspartate and glutamate), aromatic residues (phenylalanine, tryptophan and tyrosine) and glycine and histidine. Thus, Cys is a remarkably well-tolerated substitution. A Cys-substitution mutant that does not function cannot be studied by the SCAM (or by traditional site-directed mutagenesis). Residues that cannot be mutated to Cys without loss of receptor function either are accessible in the binding-site crevice 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 receptor. 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 it is likely to be accessible as well. If it is not accessible, then the

238

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

functional effect of its mutation is likely due to an indirect effect on structure. The use of an epitope-tagged receptor allows us to determine whether the rare mutant receptor that does not bind antagonist is expressed at the cell surface. Even the presence in the membrane of a nonbinding receptor mutant, however, does not prove that the mutated residue contacts ligand, as the residue could still interfere with binding indirectly.

12.2.4.3 Determination of the effects of MTS reagents on ligand binding to Cys mutant receptors To determine the effect of MTS reagents on ligand binding we follow Protocol 12.4. If an MTS reagent irreversibly alters ligand binding to a Cys mutant receptor, then we infer that the reagent reacted with the engineered Cys and that this Cys is at the water-accessible surface of the receptor (see Section 12.2.2). The MTS reagents we use are the positively charged MTSEA and MTSET and the negatively charged MTSES (see Section 12.2.1).

PROTOCOL 12.4 Reaction with MTS Reagents Equipment and Reagents • MTS reagents (Toronto Research Chemicals or Biotium): – 2-aminoethyl methanethiosulfonate hydrobromide (mol. wt 236.2) – 2-(trimethylammonium)ethyl methanethiosulfonate bromide (mol. wt 278.2) – sodium 2-sulfanatoethyl methanethiosulfonate (mol. wt 242.3) • Reagents and equipment to perform ligand binding (see Protocol 12.3).

Method 1 Suspend whole cells from a 35 mM dish in 400 µl binding buffer (see Protocol 12.2). 2 Add to 45 µl suspended cells 5 µl of water (water-treated cells) or freshly prepared 10 × stock solutiona,b,c of 2-aminoethyl methanethiosulfonate hydrobromide, 2-(trimethylammonium)ethyl methanethiosulfonate bromide or sodium 2-sulfanatoethyl methanethiosulfonated (MTS-treated cells). Mix and incubate for 2 mine at room temperature. 3 Dilutef cell suspension 15-fold in binding buffer. 4 Add 300 µl aliquots of water-treated cells into polypropylene minitubes (triplicates), in a ﬁnal volume of 0.5 ml containing 0.7 nMg [3 H]CGP-12177 either without (total binding without MTS reagent) or with 1 µM alprenolol (nonspeciﬁc binding). 5 Add 300 µl aliquots of MTS-treated cells into polypropylene minitubes (triplicates), in a ﬁnal volume of 0.5 ml containing 0.7 nM [3 H]CGP-12177 (total binding with MTS reagent).

239

12.2 METHODS AND APPROACHES

6 Incubate the mixtures at room temperature for 60 min and then ﬁlter, using a Brandel cell harvester, through Whatman 934AH glass-ﬁbre ﬁlters to separate bound radioligand from free radioligand. 7 Wash the ﬁlters three times with 1 ml of 120 mM NaCl, 10 mM Tris-HCl, pH 7.4 at 4 ◦ C. 8 Quantify the radioactivity (and hence binding) on the washed ﬁlters using liquid scintillation spectrometry. 9 Determine the speciﬁc binding of [3 H]CGP-12177 by subtracting the nonspeciﬁc binding from total binding without MTS reagent (speciﬁc binding without MTS reagent) and from total binding with MTS reagent (speciﬁc binding with MTS reagent). 10 Calculate the fraction of control as (speciﬁc binding with MTS reagent)/(speciﬁc binding without MTS reagent). For example, the fraction of control after reaction of 2.5 mM 2-aminoethyl methanethiosulfonate hydrobromide with the S200C mutant of β2 -adrenergic receptor is 477 cpm/3625 cpm = 0.13 (see Table 12.1).

Table 12.1 Data analysis. Data from a typical SCAM experiment to measure the effect of MTSEA on binding of [3 H]CGP-12177 to S200C mutant of the β2 -adrenergic receptor. Bound [3 H]CGP-12177 (cpm) NSBa Total 10 mM 2.5 mM

1 mM

MTSEA MTSEA MTSEA

0.25 mM 0.1 mM 0.025 mM MTSEA

MTSEA

MTSEA

Sample 1

88

3568

603

508

595

2505

3468

3718

Sample 2

81

3758

557

555

543

2313

3199

3308

Sample 3

91

3810

546

627

502

2471

3101

3817

Average

87

3712

569

563

547

2430

3256

3614

Speciﬁc binding

NAb

3625

482

477

460

2343

3169

3528

Fraction of control

NA

1

0.13

0.13

0.13

0.65

0.87

0.97

a NSB: b NA:

nonspeciﬁc binding. not applicable.

Notes Store stock MTS reagents desiccated at 4 ◦ C. Keep a frequently used stock desiccated at room temperature. The frequently used stock can be replenished from the 4 ◦ C stock after being warmed to room temperature. a

make a 10 × stock solution of an MTS reagent, weigh an appropriate amount of the reagent and dissolve it in water immediately before use.

b To

c At pH 7 and 22 ◦ C, MTS reagents rapidly hydrolyse with a half-life of 5–20 min [25]. If it is necessary to dissolve the reagents or perform an intermediate dilution in buffer at a physiological pH, then this should be done immediately before starting the reaction. At lower pH and lower temperature, hydrolysis is appreciably slower. A solution in distilled water appears to be stable for several hours at 4 ◦ C.

240

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

d

For screening, use ﬁnal concentrations of 2.5 mM 2-aminoethyl methanethiosulfonate hydrobromide, 1 mM 2-(trimethylammonium)ethyl methanethiosulfonate bromide and 10 mM sodium 2-sulfanatoethyl methanethiosulfonate to normalize for the intrinsic reactivities of the reagents with sulfhydryls in solution [17]. e

Two minutes is an arbitrary, but convenient, choice for the reaction time. Five-minute reactions can be used if preferred. The critical issue is that the various mutants and controls be treated identically. If incubation times are changed, then the rate calculations (see Section 12.2.4.4) must be adjusted accordingly.

f To

slow the reaction, the concentration of the MTS reagents must be decreased by dilution, centrifugation or ﬁltration. Do not stop the reaction with a reducing agent such as dithiothreitol or 2-mercaptoethanol, as this will reduce the newly formed disulﬁde bonds between the MTS reagents and the substituted Cys. Do not quench the reaction with solutions containing free sulfhydryls, as they can undergo disulﬁde exchange with the newly formed disulﬁde bond.

g The

concentration of radioligand used is close to its afﬁnity (Kd ) for the wild-type receptor, as well as for Cys-substitution mutants with near-normal function. As discussed in Section 12.2.2, the radioligand concentration must not be too high relative to its Kd . At a high ligand concentration one might miss changes in binding that result from an increase or decrease in ligand afﬁnity caused by modiﬁcation of a substituted Cys with an MTS reagent.

12.2.4.4 Determine the rate of reaction If an MTS reagent reacts with a Cys mutant receptor by altering ligand binding, then we next determine the rate of reaction by determining the effect of several different concentrations of reagent on ligand binding. Typically, 10 µm, 25 µm, 100 µm, 250 µm, 1 mm, 2.5 mm MTSEA and 10 mm MTSEA can be used, but the amounts must be appropriate to the reactivity of the substituted Cys. The extent of reaction is taken to be the extent of inhibition of binding after a ﬁxed time (typically 2 min) with ﬁve or six concentrations of reagent. The fraction of control (fraction of initial binding) Y is ﬁt to Y = (span × e−Kc ) + plateau where plateau is the fraction of residual binding at saturating concentrations of MTS reagent, span = 1 − plateau, c is the concentration of MTS reagent and K = kt, where k (m−1 s−1 ) is the second-order rate constant and t (s) is the time (120 seconds). See Table 12.1 and Figure 12.3 for results from a representative experiment of the S200C mutant of β2 -adrenergic receptor.

12.2.4.5 Protection of substituted Cys by bound ligand Reaction of an MTS reagent with a Cys near a binding site should be retarded in the presence of ligand (protection of substituted Cys by bound ligand), as mentioned in Sections 12.2.1 and 12.2.2. Thus, if an MTS reagent reacts with a Cys mutant receptor, we next determine the ability of antagonists or agonists to slow the reaction.

241

12.2 METHODS AND APPROACHES

r2=0.992

Specific binding

1.0

span

0.5

plateau 0.0 0.0000 0.0025

0.0050 0.0075 MTSEA (M)

0.0100

0.0125

Figure 12.3 A representative experiment showing the inhibition of ligand binding by the reaction of MTSEA with a substituted Cys (S200C, in TM6) in the β2 -adrenergic receptor. Data (from Table 12.1) were ﬁt by nonlinear regression using Prism (GraphPad) to a one-phase exponential decay function, Y = (span × e−Kc ) + plateau, where Y is the fraction of initial binding, c is the concentration of MTSEA and span + plateau = 1. The curve starts at (span + plateau) = 1 and decays to plateau at saturating MTSEA. In this sample ﬁt, span = 0.9, K = 2333 m−1 , plateau = 0.1 and r 2 = 0.992. Since K = kt, where k(m−1 s−1 ) is the second-order rate constant and t (s) is the time (120 s) of the reaction with the MTSEA, k = K/t = 2333/120 = 19.5 m−1 s−1 .

If antagonist or agonist slows reaction with the sulfhydryl reagents, then we infer that the residue is accessible in the binding-site crevice. Each and every residue that is protected, however, need not contact ligand; ligand could protect residues deeper in the crevice by binding above them and blocking the passage of the MTS reagent from the extracellular medium to the cytoplasmic end of the crevice. In addition, we cannot rule out indirect protection through ligand-mediated propagated structural rearrangement. To probe the reaction of an MTS reagent with a Cys mutant receptor in the presence of ligand, we: 1 perform the protecting ligand binding reaction (Protocol 12.5); 2 perform the MTS reaction (Protocol 12.6); 3 perform radioligand binding and analyse the data (Protocol 12.7).

PROTOCOL 12.5 The Protecting Ligand Binding Reaction Equipment and Reagents • Reagents and equipment to perform ligand binding (see Protocol 12.3) • 96-well multiscreen plate containing GF/B ﬁlters (Millipore).

242

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

Method 1 Suspend whole cells from a 35 mm dish in 400 µl binding buffer (see Protocol 12.2). 2 Pre-wet a 96-well multiscreen plate containing GF/B ﬁlters with 100 µl binding buffer. 3 Filter under vacuum to remove buffer. 4 Disconnect vacuum and dry the bottom of the multiscreen plate to prevent wicking. 5 Add 100 µl cells expressing the desired receptor and 50 µl binding buffer, or binding buffer containing protecting (unlabelled) ligand.a,b Include controls (with and without ligand) that will not be treated with MTS reagent.c Use triplicate determinations for each condition. See the sample layout in Table 12.2.

Table 12.2 Protection experiment. Layout of a typical protection experimenta in a 96-well multiscreen plate. 1–3

4–6

7–9

10–12

A

B/EA (NSB)b

B/EA (T)b

L/EA (NSB)

L/EA (T)

Mutant 1

B

B/B (NSB)

B/B (T)

L/B (NSB)

L/B (T)

Mutant 1

C

B/EA (NSB)

B/EA (T)

L/EA (NSB)

L/EA (T)

Mutant 2

D

B/B (NSB)

B/B (T)

L/B (NSB)

L/B (T)

Mutant 2

E

B/EA (NSB)

B/EA (T)

L/EA (NSB)

L/EA (T)

Mutant 3

F

B/B (NSB)

B/B (T)

L/B (NSB)

L/B (T)

Mutant 3

G

B/EA (NSB)

B/EA (T)

L/EA (NSB)

L/EA (T)

Mutant 4

H

B/B (NSB)

B/B (T)

L/B (NSB)

L/B (T)

Mutant 4

a First

incubation (20 min): B, binding buffer; L, protecting ligand. Second incubation (2 min): B, binding buffer; EA, MTSEA. b Binding assay after washes. NSB: nonspeciﬁc binding; T: total binding.

6 Incubate for 20 min at room temperature.

Notes ﬁnal concentration of protecting ligand is typically ∼1000 times its afﬁnity Ki . Thus, the concentration may need to be adjusted for different mutants with differing Ki values.

a The b

The protecting ligand should be relatively hydrophilic to facilitate complete removal prior to the determination of residual binding.

c

The MTS-free controls will be used to test for inhibition of radioligand binding by residual protecting ligand remaining after the ﬁlter washing step (see Protocol 12.7).

12.2 METHODS AND APPROACHES

243

PROTOCOL 12.6 The MTS Reaction Equipment and Reagents • Equipment and reagents from Protocols 12.4 and 12.5.

Method 1 Prepare the appropriate 10 × MTS stock solutiona (see Protocol 12.4), dilute in 2.5 × binding buffer and immediately add 50 µl to the experimental wells in the continued presence or absence of ligand (see Protocol 12.5, step 5). Add an equivalent volume of water diluted in buffer to the control wells. 2 Incubate for 2 mina at room temperature and stop the reaction by removing the reagents from all wells by ﬁltration. 3 Add 250 µl binding buffer to each well. Mix gently for 5 min on a plate shaker at room temperature and then vacuum ﬁlter. 4 Repeat the wash three times.

Notes a It

is important to remember that, although the protecting ligand binds reversibly, the reaction of the MTS reagents is irreversible. 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, ∼70% of the maximal effect.

PROTOCOL 12.7 Radioligand Binding and Data Analysis Equipment and Reagents • Equipment and reagents from Protocols 12.3 and 12.5.

Method 1 Add 100 µl binding buffer and 50 µl unlabelled ligand in binding buffer to the appropriate wells (nonspeciﬁc binding). 2 Add 150 µl binding buffer to the appropriate wells (total binding). 3 Shake the plate on a plate shaker for 5 min at room temperature to resuspend the cells. 4 Add 100 µl radioligand solution (0.7 nM, [3 H]CGP-12177) to all wells.

244

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

5 Incubate the mixtures at room temperature for 60 min. 6 Filter to separate bound radioligand from free. Wash twice with ice-cold binding buffer without BSA. 7 Punch ﬁltersa into scintillation vials and quantify the radioactivity using liquid scintillation spectrometry. 8 Analyse the data by determining speciﬁc binding (by subtracting nonspeciﬁc binding from total binding) and inhibition of speciﬁc binding by MTS reagents and by calculating protection as 1 − [(inhibition in the presence of ligand)/(inhibition in the absence of ligand)].

Notes a The

Hydropure membrane support in Millipore 96-well ﬁlter plates avidly binds many small radioligands. Thus, it is imperative to determine if the ligand used binds to the Hydropure backing. If so, then the glass-ﬁbre ﬁlters must be counted without the backing. This is facilitated by complete drying of the ﬁlters prior to punching.

12.2.5 Secondary structure To infer a secondary structure, we must assume, as mentioned in Section 12.2.2, 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-site crevice. In an α-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 α helix with a stripe of about 140◦ facing the binding-site crevice [10]. In contrast, in an antiparallel β 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 informative. For example, we observed an unusual pattern of accessibility in the seventh membrane-spanning segment (TM7) of the dopamine D2 receptor. The overall pattern of exposure was not consistent with a simple secondary structure of either α helix or β strand. TM7, however, contains the highly conserved residues Asn-Pro in the middle of the membrane-spanning segment. In soluble proteins, these residues have been observed to introduce kinks and twists in α helices. In molecular modelling work, we found that the pattern of exposure of the Cys mutant receptors to MTSEA can be explained if TM7 is a kinked and twisted α helix and is consistent as well with the 3–10 helix structure observed in TM7 of the rhodopsin structure [12, 26, 27]. In contrast, accessibility data in TM6 are consistent with a kinked but not twisted α helix, again consistent with the rhodopsin crystal structure [13, 26, 27]. Thus, these ‘irregular’ patterns of accessibility can lead to new insights or directions for further experimental pursuit. In addition to membrane-spanning segments, structural information for the extracellular loops of GPCRs can also be obtained by applying SCAM. Speciﬁcally, Shi and

12.2 METHODS AND APPROACHES

245

Javitch [28], using SCAM, have shown that the pattern of accessibility in the second extracellular loop of dopamine D2 receptor is consistent with a structure similar to that of bovine rhodopsin, in which the region C-terminal to the conserved disulﬁde bond is deeper in the binding-site crevice than is the N-terminal part of the loop. These results suggested that the second extracellular loop of D2 receptor is likely to contribute to the binding site in the receptor.

12.2.6 Electrostatic potential Because positively charged MTSET and negatively charged MTSES are similar in size, differences in their reactivities with engineered Cys are likely to be due to differences in the electrostatic potential of the binding-site crevice. For example, in TM3 of the D2 receptor, MTSES did not react with any engineered Cys more cytoplasmic than Val111, whereas MTSET reacted with several residues more cytoplasmic than this position [10]. This reﬂects the negative electrostatic potential deeper in the binding-site crevice. By subsequent applications of positively and negatively charged reagents, we can rule out the possibility that reaction has occurred without alteration of function in the case of addition of one but not the other charged moiety. For example, we determined that MTSES did not react with Cys118 in D2 receptor because subsequent application of MTSET still inhibited binding. If MTSES had reacted silently with Cys118, then it would have prevented the Cys from subsequent reaction with MTSET. In contrast, the reactivity of MTSET and MTSES with F110C and V111C, residues located near the extracellular end of TM3, is similar. This indicates that the electrostatic potential near these residues is not as negative as it is below Val111. The results in TM2 are also consistent with a negative electrostatic potential [14], but this has been less apparent in the other membrane-spanning segments studied [10–13, 15]. This suggests that Asp80 and Glu95 in TM2 and Asp108 and Asp114 in TM3 are signiﬁcant contributors to the negative potential in this region and that the potential is not uniform throughout the crevice at similar depths. This distribution of the ﬁeld likely helps to orient ligand within the binding-site crevice with the protonated amine towards Asp114.

12.2.7 Conformational changes associated with receptor activation Conformational changes in a protein may result in changes in the accessibility of substituted Cys 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 [1, 2]. 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 residues forming the surface of the binding-site crevice, we sought to determine the relative accessibilities of a series of engineered Cys 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 Cys. However, the

246

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

activated state of the receptor can be achieved by using a constitutively active mutant (CAM) receptor as a background for further Cys substitution. A CAM receptor is intrinsically active and has a higher afﬁnity for agonist than the wild-type receptor [29]. 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 isomerizes more easily to the active state). Thus, we can compare the resting and active forms of the receptor by determining the accessibility of substituted Cys in the binding-site crevice in these two states using wild-type receptor and a CAM as background constructs. We have chosen to pursue initial studies in the β2 -adrenergic receptor because of the availability of a well-characterized CAM [29] (kindly provided by R. Lefkowitz). MTSEA had no effect on antagonist binding to the wild-type β2 -adrenergic receptor expressed in HEK 293 cells. This suggests that no endogenous Cys are accessible in the binding-site crevice (or that reaction takes place but is without functional effect). In contrast, in the CAM β2 -adrenergic receptor, MTSEA inhibited antagonist binding signiﬁcantly and isoproterenol slowed the rate of reaction of MTSEA [30]. This implies that at least one endogenous Cys becomes accessible in the binding-site crevice of the CAM β2 -adrenergic receptor. We found that Cys285, in the TM6, is responsible for the inhibitory effect of MTSEA on ligand binding to the CAM β2 -adrenergic receptor [30]. 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 (Figure 12.4). Such a movement of TM6 upon receptor activation is consistent with the results of ﬂuorescence spectroscopy studies in β2 -adrenergic receptor [31, 32] and spin-labelling studies in rhodopsin [33] and suggests that SCAM in a CAM background is a powerful approach for probing conformational changes in these receptors.

C285

TM6 TM5

TM7 TM1 TM3 TM2

TM4

Figure 12.4 An illustration of the rotation and/or tilting of the sixth membrane-spanning segment, TM6, associated with the activation of the β2 -adrenergic receptor. The rearrangement indicated brings Cys285 to the margin of the binding-site crevice of β2 -adrenergic receptor and allows it to react with MTSEA to inhibit ligand binding. The arrangement of the membrane-spanning segments is based on the projection structure of rhodopsin [34]. The accessible surface of TM6 of the dopamine D2 receptor at the level of the aligned Cys is shaded [13]. Adapted from Javitch et al. [30].

12.3 TROUBLESHOOTING

247

More recently, Yan et al. [35] have also detected conformational changes in the κ-opioid receptor by applying SCAM and determining the accessibilities of substituted Cys in TM6 and TM7, in the presence and absence of overexpression of different Gα-subunits with this receptor. Thus, overexpressing G proteins with a GPCR is another approach to stabilizing an active state of receptor. Speciﬁcally, in TM7 of κ-opioid receptor, Cys substituted for Ser311 and Asn326 were shown to become more reactive in the presence of Gα16 overexpression, whereas Cys substituted for Tyr313, Asn322, Ser323 and Leu329 became more reactive with Gαi2 overexpression. In contrast to the TM7 residues, the V296C in TM6 became insensitive after Gαi2 overexpression.

12.2.8 Structural bases of pharmacological speciﬁcity Conserved features of the sequences of dopamine receptors and of homologous GPCRs point to regions, and amino acid residues within these regions that contribute to their ligand-binding sites. Differences in binding speciﬁcities among the catecholamine receptors, however, must stem from their non-conserved residues. Using SCAM, we have identiﬁed the residues that form the surface of the water accessible binding-site crevice in the dopamine D2 receptor. Of ∼90 membrane-spanning residues that differ between the D2 and D4 receptors, only 20 were found to be accessible, and six of these 20 are conservative aliphatic substitutions. We targeted these accessible residues not conserved in the homologous dopamine D4 receptor as candidates for the structural determinants of pharmacological speciﬁcity. We reasoned that mutation of these candidate residues in the D2 receptor to the aligned residues in the D4 receptor would generally be well tolerated, based on our previous experience in mutating them to Cys. In a D2 receptor background, we mutated to the aligned residues in the D4 receptor, individually or in combinations, the 14 accessible, nonconserved residues [36]. We also made the reciprocal mutations in a D4 receptor background. The combined substitution of four to six of these residues was sufﬁcient to switch the afﬁnity of the receptors for several chemically distinct D4 -selective antagonists by 3 orders of magnitude in both directions (D2 − D4 -like and D4 − D2 -like). The mutated residues are in TM2, TM3 and TM7 and form a cluster in the binding-site crevice. Mutation of a single residue in this cluster in TM2 was sufﬁcient to increase the afﬁnity for clozapine to D4 -like levels. We can rationalize data in terms of a set of chemical moieties in the ligands interacting with a divergent aromatic microdomain in TM2–TM3–TM7 of the D2 and D4 receptors.

12.3 Troubleshooting • While Cys substitution is well tolerated in the great majority of cases, it is not universally tolerated. The use of an N-terminally epitope-tagged receptor allows one to assess receptor insertion into the plasma membrane, but even presence in the membrane does not ensure that protein folding and other posttranslational modiﬁcations are correct. In rare cases, an apparently successful cohesive-ended ligation of a fragment bearing a sequenced mutation can result in a loss of base pairs at

248

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

the site of ligation, resulting in a frame shift and complete loss of expression for a Cys mutant. Therefore, before concluding that a Cys substitution results in a loss of expression or binding, it is important to check not only for the presence of the mutation, but also for the presence of the expected restriction sites at the sites of ligation, either by restriction mapping or sequencing. • MTS reagents are sensitive to hydrolysis, and reagents that have not been stored or handled properly may no longer be reactive. Methods are available to assay the purity and reactivity of the reagents if such questions should arise [25].

References 1. Akabas, M.H., Stauffer, D.A., Xu, M. and Karlin, A. (1992) Acetylcholine receptor channel structure probed in cysteine-substitution mutants. Science, 258, 307–310. 2. Akabas, M.H., Kaufmann, C., Archdeacon, P. and Karlin, A. (1994) Identiﬁcation of acetylcholine receptor channel-lining residues in the entire M2 segment of the α subunit. Neuron, 13, 919–927. 3. Akabas, M.H. and Karlin, A. (1993) Identiﬁcation of acetylcholine receptor channel-lining residues in the M1 segment of the α subunit. Biochemistry, 34, 12496–12500. 4. Xu, M., Covey, D.F. and Akabas, M.H. (1995) Interaction of picrotoxin with GABAA receptor channel-lining residues probed in cysteine mutants. Biophys. J., 69, 1858–1867. 5. Xu, M. and Akabas, M.H. (1993) Amino acids lining the channel of the γ -aminobutyric acid type A receptor identiﬁed by cysteine-substitution. J. Biol. Chem., 268, 21505–21508. 6. Akabas, M.H., Kaufmann, C., Cook, T.A. and Archdeacon, P. (1994) Amino acid residues lining the chloride channel of the cystic ﬁbrosis transmembrane conductance regulator. J. Biol. Chem., 269, 14865–14868. 7. Yan, R.T. and Maloney, P.C. (1995) Residues in the pathway through a membrane transporter. Proc. Natl. Acad. Sci. U. S. A., 92, 5973–5976. 8. Pascual, J.M., Shieh, C.C., Kirsch, G.E. and Brown, A.M. (1995) Multiple residues specify external tetraethylammonium blockade in voltage-gated potassium channels. Biophys. J., 69, 428–434. 9. Javitch, J.A., Li, X., Kaback, J. and Karlin, A. (1994) A cysteine residue in the third membrane-spanning segment of the human D2 dopamine receptor is exposed in the binding-site crevice. Proc. Natl. Acad. Sci. U. S. A., 91, 10355–10359. 10. Javitch, J.A., Fu, D., Chen, J. and Karlin, A. (1995) Mapping the binding-site crevice of the dopamine D2 receptor by the substituted-cysteine accessibility method. Neuron, 14, 825–831. The original publication ﬁrst describing the application of SCAM to GPCRs. 11. Javitch, J.A., Fu, D. and Chen, J. (1995) Residues in the ﬁfth membrane-spanning segment of the dopamine D2 receptor exposed in the binding-site crevice. Biochemistry, 34, 16433–16439. SCAM applied to GPCRs. 12. Fu, D., Ballesteros, J.A., Weinstein, H. et al. (1996) Residues in the seventh membrane-spanning segment of the dopamine D2 receptor accessible in the binding-site crevice. Biochemistry, 35, 11278–112785. SCAM applied to GPCRs.

REFERENCES

249

13. Javitch, J.A., Ballesteros, J.A., Weinstein, H. and Chen, J. (1998) A cluster of aromatic residues in the sixth membrane-spanning segment of the dopamine D2 receptor is accessible in the binding-site crevice. Biochemistry, 37, 998–1006. SCAM applied to GPCRs. 14. Javitch, J.A., Ballesteros, J.A., Chen, J. et al. (1999) Electrostatic and aromatic microdomains within the binding-site crevice of the D2 receptor: contributions of the second membrane-spanning segment. Biochemistry, 38, 7961–7968. SCAM applied to GPCRs. 15. Javitch, J.A., Shi, L., Simpson, M.M. et al. (2000) The fourth transmembrane segment of the dopamine D2 receptor: accessibility in the binding-site crevice and position in the transmembrane bundle. Biochemistry, 39, 12190–12199. SCAM applied to GPCRs. 16. Strader, C.D., Fong, T.M., Tota, M.R. et al. (1994) Structure and function of G protein-coupled receptors. Annu. Rev. Biochem., 63, 101–132. 17. Stauffer, D.A. and Karlin, A. (1994) Electrostatic potential of the acetylcholine binding sites in the nicotinic receptor probed by reactions of binding-site cysteines with charged methanethiosulfonates. Biochemistry, 33, 6840–6849. 18. Roberts, D.D., Lewis, S.D., Ballou, D.P. et al. (1986) Reactivity of small thiolate anions and cysteine-25 in papain toward methyl methanethiosulfonate. Biochemistry, 25, 5595–5601. 19. Danielson, M.A., Bass, R.B. and Falke, J.J. (1997) Cysteine and disulﬁde scanning reveals a regulatory α-helix in the cytoplasmic domain of the aspartate receptor. J. Biol. Chem., 272, 32878–32888. 20. Javitch, J.A., Fu, D. and Chen, J. (1996) Differentiating dopamine D2 ligands by their sensitivities to modiﬁcation of the cysteine exposed in the binding-site crevice. Mol. Pharmacol., 49, 692–698. 21. Stephan, M.M., Kamdar, G., Rudnick, G. and Penado, K.M.Y. (1999) A functional link between transmembrane span 7 and extracellular loop 1 of the serotonin transporter: effects of Na+ , Li+ , and methansulfonate reagents. Soc. Neurosci. Abstr., 25, 1699. 22. Jung, K., Jung, H., Wu, J. et al. (1993) Use of site-directed ﬂuorescence labeling to study proximity relationships in the lactose permease of Escherichia coli . Biochemistry, 32, 12273–12278. 23. Olami, Y., Rimon, A., Gerchman, Y. et al. (1997) Histidine 225, a residue of the NhaA-Na+ /H+ antiporter of Escherichia coli is exposed and faces the cell exterior. J. Biol. Chem., 272, 1761–1768. 24. Seal, R.P. and Amara, S.G. (1996) Residues involved in substrate interactions with a sodium-dependent glutamate transporter identiﬁed using cysteine scanning mutagenesis. Soc. Neurosci. Abstr., 22, 1575. 25. Karlin, A. and Akabas, M.H. (1998) Substituted-cysteine accessibility method. Methods Enzymol., 293, 123–145. 26. Ballesteros, J.A., Shi, L. and Javitch, J.A. (2001) Structural mimicry in G-protein-coupled receptors: implications of the high-resolution structure of rhodopsin for structure–function analysis of rhodopsin-like receptors. Mol. Pharmacol., 60, 1–19. 27. Palczewski, K., Kumasaka, T., Hori, T. et al. (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science, 289, 739–745. 28. Shi, L. and Javitch, J.A. (2004) The second extracellular loop of the dopamine D2 receptor lines the binding-site crevice. Proc. Natl. Acad. Sci. U. S. A., 101, 440–445.

250

CH 12 SUBSTITUTED CYSTEINE ACCESSIBILITY METHOD (SCAM)

29. Samama, P., Cotecchia, S., Costa, T. and Lefkowitz, R.J. (1993) A mutation-induced activated state of the ββ2 -adrenergic receptor. Extending the ternary complex model. J. Biol. Chem., 268, 4625–4636. 30. Javitch, J.A., Fu, D., Liapakis, G. and Chen, J. (1997) Constitutive activation of the β2 adrenergic receptor alters the orientation of its sixth membrane-spanning segment. J. Biol. Chem., 272, 18546–18549. Application of SCAM to detect conformational changes of GPCRs associated with their activation. 31. Gether, U., Lin, S., Ghanouni, P. et al. (1997) Agonists induce conformational changes in transmembrane domains III and VI of the β2 adrenoceptor. EMBO J., 16, 6737–6747. 32. Gether, U., Lin, S. and Kobilka, B.K. (1995) Fluorescent labeling of puriﬁed β2 adrenergic receptor. Evidence for ligand-speciﬁc conformational changes. J. Biol. Chem., 270, 28268–28275. 33. Farrens, D.L., Altenbach, C., Yang, K. et al. (1996) Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science, 274, 768–770. 34. Schertler, G.F., Villa, C. and Henderson, R. (1993) Projection structure of rhodopsin. Nature, 362, 770–772. 35. Yan, F., Mosier, P.D., Westkaemper, R.B. and Roth, B.L. (2008) Gα-subunits differentially alter the conformation and agonist afﬁnity of κ-opioid receptors. Biochemistry, 47, 1567–1578. 36. Simpson, M.M., Ballesteros, J.A., Chiappa, V. et al. (1999) Dopamine D4 /D2 receptor selectivity is determined by a divergent aromatic microdomain contained within the second, third, and seventh membrane-spanning segments. Mol. Pharmacol., 56, 1116–1126.

13 Homology Modelling of G Protein-coupled Receptors John Simms Department of Pharmacology, University of Monash, Clayton, Australia

13.1 Introduction The functional characterization of a protein sequence is one of the most frequent problems in biology. This task is usually facilitated by a three-dimensional (3D) structure of the protein studied. However, when a structure is not available, as is in the case for G protein-coupled receptors (GPCRs), homology modelling can be used as a tool for making quantitative and qualitative conclusions about the query sequence. Over 1000 sequences likely to encode GPCRs are currently available in publicly accessible and proprietary databases and this number is likely to grow with the reﬁnement of the human genome. Furthermore, the biological importance of these receptors is reﬂected in the enormous interest they have to the pharmaceutical industry as the targets of approximately 30% of existing medications [1]. Despite numerous efforts, however, only bovine rhodopsin (bRh [2]) and more recently the β2a adrenergic receptor (β2a R [3]) have been solved to high resolution. Homology modelling and prediction of GPCR structure is, therefore, expected to be a valuable tool for understanding the function of this class of membrane proteins. At present, most homology modelling protocols start from the assumption that, except for insertions and deletions, the backbone of the target is identical to the backbone of the template structure. In practice, however, domain motions and bending of parts of the molecule with respect to each other is often seen. Even in cases of signiﬁcant distortion, short-range interactions will not differ very much and homology modelling has been shown to be adequate for the interpretation of experimental data [4] and rational drug design [5]. In recent years, homology modelling has become G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

252

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

increasingly common; although the number of techniques has increased, any homology modelling procedure consists of four sequential elements: 1 2 3 4

Template recognition. Alignment of the target sequence and template structure(s). Model building. Model veriﬁcation, which may be used iteratively with steps 1–3.

13.2 Methods and approaches 13.2.1 Template selection for homology modelling The ﬁrst step of any homology modelling protocol begins with the selection of a suitable structural template from the Protein Data Bank (PDB, http://www.pdb.org[6]). If the target sequence shares a moderate sequence identity (>30%) to the template, then homology detection is reasonably straightforward and can be achieved by software such as BLAST (http://www.ncbi.nlm.nih.gov/blast/ [7]). However, at low sequence similarities, hits from BLAST may not be reliable; as such, a number of alternative strategies have been developed. Methods such as PSI-BLAST (http://www.ncbi.nlm.nih.gov/Education/BLASTinfo/psi1.html [8]) and hidden Markov models [9], as implemented in the SAM (http://www.soe.ucsc.edu/compbio/SAM T06/ T06-query.html [10]) and HMMER (http://hmmer.janelia.org/ [11]) packages, have vastly improved the accuracy of sequence alignments and have extended the boundaries of detectable sequence similarity. There are many tools publicly available in this area, which can be found at the Critical Assessment of Techniques for Protein Structure Prediction (CASP) 7 web site (http://predictioncenter.gc.ucdavis.edu/casp7/Casp7.html). In addition, GPCRs may be considered as multidomain structures for which a number of templates may be available for modelling discrete domains of the receptor and combined to generate a complete model.

13.2.1.1 Transmembrane domain including interconnecting loops To date, the only suitable templates for the transmembrane (TM) region of a GPCR are the X-ray crystal structures of bRh and β2a R. It is important not to assume blindly that the β2a R structure is a better template than bRh for modelling GPCRs as a whole. Sequence alignment may be used as a guide to which sections of each structure are appropriate templates for the query sequence. This is especially important in loop regions, which exhibit the greatest variability in the structure. In addition, despite the fact that a number of bRh structures are available, subtleties in the crystallization procedures from different groups have resulted in slight differences between structures. This is most apparent in the region between TMs V and VI of bRh which show differences in the proximal and distal ends of the loop (PDB code 1U19 [12] compared with 1GZM [13]).

13.2 METHODS AND APPROACHES

253

13.2.1.2 Large N-terminal domains A characteristic feature of Family B and C GPCRs, as well as some Family A GPCRs, is the large N-terminal domain that provides an epitope for high-afﬁnity ligand binding. However, despite this common feature, the structures of the N-terminal domains from the respective families are distinct, and different templates are required for homology modelling. A number of Family A GPCRs are characterized by a large N-terminal domain containing leucine-rich-repeats (LRRs) that are capped by N-terminal and C-terminal cysteine-rich regions. To date, a number of structures have been reported containing this motif but exhibit different properties, including number of LRRs and position of cysteine residues. In this case, a number of possible templates should be identiﬁed, which can then be combined when generating homology models. Examples of LRRs containing proteins which share high sequence similarity with the thyroid-stimulating hormone receptor (TSHR) are at PDB references 1OZN [14] and 2BNH [15]. In contrast to the N-terminal region of Family A GPCRs, only a limited number of templates (PDB references 1U34 [16], 2JOD [17], 2JND [18] and 2QKH [19]) have been reported for the equivalent domain of Family B GPCRs. Consistent with Family A GPCRs, however, these domains share little overall sequence homology, whilst maintaining a similar overall fold; as such, the best template(s) should be identiﬁed and used for modelling. In the case for Family C GPCRs, a number of structures for the large globular N-terminal domain of the metabotropic glutamate receptor have been reported (e.g. 1ISR [20] and 2E4U [21]) in a number of liganded states and may be used as templates for this region in other receptors.

13.2.2 Sequence and structure alignments for homology modelling Despite recent advances in methods used to generate homology models, a critical step is still the initial alignment between the target sequence and template structure(s) (Protocol 13.1). A common method to align the target and the template is by using a pairwise sequence alignment. This method enumerates a number of possible matches between the two protein sequences that are then scored using a substitution probability matrix. For water-soluble proteins, two common substitution matrices are Gonnet [22] and BLOSUM [23]. Since the amino acid composition of TM domains differs from water-soluble proteins, several groups have introduced membrane-speciﬁc-scoring matrices, including PHAT [24] and SLIM [25]. However, a recent study has shown that matrices derived for the alignment of water-soluble proteins are equally applicable to the alignment of membrane proteins [26]. A derivative of pairwise sequence alignments is the bipartite alignment method, which utilizes separate substitution matrices for the TM region and water-exposed regions of the receptor, such as loops [27]. In spite of the accuracy of the pairwise method, multiple sequence and proﬁle alignments are the most robust methods for generating alignments [26]. Methods such as T-Coffee [28] and ClustalW [29] are accurate at high sequence identities (>40%), and although they have been used for aligning water-soluble proteins, these methods have been shown to

254

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

be equally applicable to membrane proteins. At lower sequence identities, as is the case for aligning other families of GPCRs against bRh or β2a R, alignment techniques which include secondary structure information can improve the accuracy of the alignment [26]. In addition, the use of multiple sequence alignments has also given rise to a number of different nomenclatures for identifying residues across divergent GPCR families. One popular system [30] uses the most conserved residue within a TM as a reference for other residues within the same helix. This nomenclature substitutes the residue’s absolute position number by a numerical identiﬁer that is composed of (i) a number from 1 to 7 that identiﬁes the helix and (ii) a number associated with a position in that particular helix. The second number is relative to the most conserved residue in a particular helix that is given the number 50. For example, using the bRh sequence Asn1.50 refers to Asn55 in TMI, Asp2.50 refers to Asp83 in TMII, Arg3.50 refers to Arg135 TMIII, Trp4.50 refers to Trp161 in TMIV, Pro5.50 refers to Pro215 in TMV, Pro6.50 refers to Pro267 in TMVI and Pro7.50 refers to Pro303 in TMVII. Residues located C-terminally to the conserved residue are identiﬁed with numbers increasing from 50, whereas decreasing numbers are allocated to residues located N-terminally to the conserved residue.

PROTOCOL 13.1 Aligning the Sequences of the Human Vasopressin1a Receptor with bRh The following procedure uses MODELLER to align the human vasopressin 1a receptor sequence (huv1ar) and the rat vasopressin1a receptor (rv1ar) sequence in ﬁle ‘huv1ar.seq’ with the bRh structure, 1U19. The resulting ﬁle ‘huv1ar-1U19.pir’ is then used in Protocol 13.2 to generate a homology model.

Requirements and Conventions for Protocols [13.1–13.3] Hardware • A computer running Linux/Unix, Microsoft Windows98/NT/2000/XP or Apple Mac OS X, 512 MB RAM or higher and approximately 100 MB of free hard-disk space for the software and output ﬁles. Software to obtain and install • MODELLER, available from http://salilab.org/modeller/download installation.html • DeepView [31], available from http://au.expasy.org/spdbv

Computer Skills Although MODELLER and DeepView can run under UNIX-based, Windows or Apple Mac operating systems, UNIX-based operating systems offer a framework and scripting system for the

13.2 METHODS AND APPROACHES

255

manipulation of ﬁles. However, whilst knowledge of Linux/UNIX is not necessary for modelling, some knowledge of the scripting system would be useful. In addition, MODELLER does not have a graphical user interface and is run from the command line; therefore, a basic knowledge of command-line skills is necessary to follow the protocol described in this chapter.

Conventions Followed in the Text The sequence for the query structure, for which a homology model is being calculated, is referred to as the ‘target’. A ‘template’ is an experimentally (nuclear magnetic resonance (NMR) or X-ray crystallography) derived structure used for comparative modelling. Files with a ‘.seq’ extension refer to one or more unaligned sequences in the PIR format. Files with ‘.ali’ extensions contain the alignment of two or more sequences and/or structures. Files with ‘.pir’ extension contain the alignment of two or more sequences and/or structures in PIR format. Files with a ‘.py’ extension are MODELLER scripts. MODHOME is the location of the MODELLER home directory. The extensions presented here are not mandatory, and the user may adopt any systematic convention deemed necessary. The version of MODELLER is 9.1. In the methods section’ the symbol ‘>’ refers to a submenu located under a main menu title in DeepView. The symbols ‘$>’ refer to commands which should be applied directly to the command line and do not form part of a script. UNIX commands are represented in the main text as italics. Text in Courier New font refers to commands entered at the command line, in a script or the contents of a ﬁle and does not form part of the main body of text.

Requirements for Protocol 13.1 • Internet access • a text editor • DeepView installed on a local computer • MODELLER installed on either a local or networked computer.

Method 1 Download the amino acid sequences for the huv1ar (accession number P37288) and rv1ar (accession number P30560) from http://au.expasy.org/sprot/. 2 Download the ﬁle 1U19 from the PDB repository (http://www.rcsb.org/pdb/home/home.do). 3 Open the ﬁle 1U19.pdb using the File > Open PDB ﬁle function in DeepView and save the amino acid sequence using the File > Save > Sequence (FASTA) function. 4 Edit a text ﬁle such that it contains, in PIR formata (see Figure 13.1) (i) the sequence of the template (1U19.pdb), (ii) the sequence of the protein to be modelled (huv1ar) and (iii) other homologues obtained from a database search (rv1ar) and save as huv1ar.seq.

256

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

>P1;1U19.pdb structure:1U19.pdb:1:A:348:A::::: MNGTEGPNFYVPFSNKTGVVRSPFEAPQYYLAEPWQFSMLAAYMFLLIMLGFPINFLTLYVTVQHKKLR TPLNYILLNLAVADLFMVFGGFTTTLYTSLHGYFVFGPTGCNLEGFFATLGGEIALWSLVVLAIERYVV VCKPMSNFRFGENHAIMGVAFTWVMALACAAPPLVGWSRYIPEGMQCSCGIDYYTPHEETNNESFVIYM FVVHFIIPLIVIFFCYGQLVFTVKEAAAQQQESATTQKAEKEVTRMVIIMVIAFLICWLPYAGVAFYIF THQGSDFGPIFMTIPAFFAKTSAVYNPVIYIMMNKQFRNCMVTTLCCGKNPLGDDEASTTVSKTETSQV APA* >P1;huv1ar sequence:huv1ar::::::::: MSEAAHVLITGAAGQIGYILSHWIASGELYGDRQVYLHLLDIPPAMNRLTALTMELEDCAFPHLAGFVA TTDPKAAFKDIDCAFLVASMPLKPGQVRADLISSNSVIFKNTGEYLSKWAKPSVKVLVIGNPDNTNCEI AMLHAKNLKPENFSSLSMLDQNRAYYEVASKLGVDVKDVHDIIVWGNHGESMVADLTQATFTKEGKTQK VVDVLDHDYVFDTFFKKIGHRAWDILEHRGFTSAASPTKAAIQHMKAWLFGTAPGEVLSMGIPVPEGNP YGIKPGVVFSFPCNVDKEGKIHVVEGFKVNDWLREKLDFTEKDLFHEKEIALNHLAQGG* >P1;rv1ar sequence:rv1ar:::::::: MSFPRGSQDRSVGNSSPWWPLTTEGSNGSQEAARLGEGDSPLGDVRNEELAKLEIAVLAVIFVVAVLGN SSVLLALHRTPRKTSRMHLFIRHLSLADLAVAFFQVLPQLCWDITYRFRGPDWLCRVVKHLQVFAMFAS AYMLVVMTADRYIAVCHPLKTLQQPARRSRLMIATSWVLSFILSTPQYFIFSVIEIEVNNGTKTQDCWA TFIQPWGTRAYVTWMTSGVFVAPVVVLGTCYGFICYHIWRNIRGKTASSRHSKGDKGSGEAVGPFHKGL LVTPCVSSVKSISRAKIRTVKMTFVIVSAYILCWAPFFIVQMWSVWDENFIWTDSENPSITITALLASL NSCCNPWIYMFFSGHLLQDCVQSFPCCHSMAQKFAKDDSDSMSRRQTSYSNNRSPTNSTGMWKDSPKSS KSIRFIPVST*

Figure 13.1 An example of a PIR formatted ﬁle containing the sequences of the template (bRh, PDB identiﬁer 1U19.pdb), the target (huv1ar) and a homologous protein to the target (rv1ar). For clarity, this example contains only three sequences, whereas multiple sequence alignments for homology modelling should contain a number of homologous proteins of the target sequence.

5 Edit a ﬁle such that it contains the following MODELLER Python script and save the ﬁle as salign.py:b from modeller import* log.verbose() env = environ() env.io.atom_files_directory=’./’ aln = alignment(env, file=’huv1ar.seq’, align_codes=’all’) aln.salign(rr_file=’$(LIB)/as1.sim.mat’, output=’’, max_gap_length=20, gap_function=True, feature_weights=(1., 0., 0., 0., 0., 0.), gap_penalties_1d=(-200, 0), gap_penalties_2d=(3.5, 3.5, 3.5, 0.2, 4.0, 6.5, 2.0, 0.0, 0.0),

13.2 METHODS AND APPROACHES

257

output_weights_file=’salign.mtx’ similarity_flag=True) aln.write(file=’huv1ar-1U19.pir’, alignment_format=’PIR’)

6 Execute the MODELLER Python script salign.py with the command $> MODHOME/modeller9v1/bin/mod9v1 salign.py

7 Manually edit the alignment ﬁle ‘huv1ar-1U19.pir’ with a text editor such that large gaps or regions of uncertainty in the alignment and the corresponding regions in other sequences are removedc (see Figures 13.2 and 13.3). >P1;1U19.pdb structure:1U19.pdb:1 :A:348 :A:::-1.00:-1.00 MNGTEGPNFYVPFSNK------TGVVRSPFEA--------PQYYLAEPWQFSMLAAYMFLLIMLGFPIN FLTLYVTVQHKKLRTPLNYILLNLAVADLFMVFGGFTTTLYTSLHGYFVFGPTGCNLEGFFATLGGEIA LWSLVVLAIERYVVVCKPMSNFRFGENHAIMGVAFTWVMALACAAPPLVGWS--RYIPEGM—-QCSCGI DYYTPHEETNNESFVIYMFVVHFIIPLIVIFFCYGQLVFTV------KEAAA----Q----QQES---------------ATTQKAEKEVTRMVIIMVIAFLICWLPYAGVAFYIFTHQGS---DFGPIFMTIPAFF AKTSAVYNPVIYIM-MNKQFRNCMVTTLCC----------GKNPLG------DDEASTTVSK--------TETSQVAPA--* >P1;huv1ar sequence:huv1ar: : : : :::-1.00:-1.00 MRLSAGPDAGPSGNSSPWWPLATGAGNTSREAEALGEGNGPPRDVRNEELAKLEIAVLAVTFAVAVLGN SSVLLALHRTPRKTSRMHLFIRHLSLADLAVAFFQVLPQMCWDITYRFRGPDWLCRVVKHLQVFGMFAS AYMLVVMTADRYIAVCHPLKTLQQPARRSRLMIAAAWVLSFVLSTPQYFVFS—-MIEVNNVTKARDCWA TFIQPW---GSRAYVTWMTGGIFVAPVVILGTCYGFICYNIWCNVRGKTASR----QSKGAEQAGVAFQ KGFLLAPCVSSVKSISRAKIRTVKMTFVIVTAYIVCWAPFFIIQMWSVWDPMSVWTESENPTITITALL GSLNSCCNPWIYMFFSGHLLQDCVQSFPCCQNMKEKFNKEDTDSMSRRQTFYSNNRSPTNSTGMWKDSP KSSKSIKFIPVST* >P1;rv1ar sequence:rv1ar: : : : :::-1.00:-1.00 MSFPRGSQDRSVGNSSPWWPLTTEGSNGSQEAARLGEGDSPLGDVRNEELAKLEIAVLAVIFVVAVLGN SSVLLALHRTPRKTSRMHLFIRHLSLADLAVAFFQVLPQLCWDITYRFRGPDWLCRVVKHLQVFAMFAS AYMLVVMTADRYIAVCHPLKTLQQPARRSRLMIATSWVLSFILSTPQYFIFSVIEIEVNNGTKTQDCWA TFIQPW---GTRAYVTWMTSGVFVAPVVVLGTCYGFICYHIWRNIRGKTASSRHSKGDKGSGEAVGPFH KGLLVTPCVSSVKSISRAKIRTVKMTFVIVSAYILCWAPFFIVQMWSVWDENFIWTDSENPSITITALL ASLNSCCNPWIYMFFSGHLLQDCVQSFPCCHSMAQKFAKDDSDSMSRRQTSYSNNRSPTNSTGMWKDSP KSSKSIRFIPVST*

Figure 13.2 The salign.py-generated, PIR-formatted multiple sequence alignment ﬁle (huv1ar-1U19.pir) containing the sequences for the template (bRh, 1U19.pdb), the target sequence (human V1aR, huv1ar) and a homologous protein to the target sequence (rat V1aR, rv1ar). Highlighted in grey are large gaps or regions of uncertainty in the alignment. Modelling of large inserts requires the use of specialist techniques and should be addressed separately; as such, these regions are deleted from the alignment (see Figure 13.3).

258

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

>P1;1U19.pdb structure:1U19.pdb:1 :A:287 :A:::-1.00:-1.00 PQYYLAEPWQFSMLAAYMFLLIMLGFPINFLTLYVTVQHKKLRTPLNYILLNLAVADLFMVFGGFTTTL YTSLHGYFVFGPTGCNLEGFFATLGGEIALWSLVVLAIERYVVVCKPMSNFRFGENHAIMGVAFTWVMA LACAAPPLVGWS--RYIPEGM--QCSCGIDYYTPHEETNNESFVIYMFVVHFIIPLIVIFFCYGQLVFT V-ATTQKAEKEVTRMVIIMVIAFLICWLPYAGVAFYIFTHQGS---DFGPIFMTIPAFFAKTSAVYNPV IYIM-MNKQFRNCMVTTLCC* >P1;huv1ar sequence:huv1ar: : : : :::-1.00:-1.00 PPRDVRNEELAKLEIAVLAVTFAVAVLGNSSVLLALHRTPRKTSRMHLFIRHLSLADLAVAFFQVLPQM CWDITYRFRGPDWLCRVVKHLQVFGMFASAYMLVVMTADRYIAVCHPLKTLQQPARRSRLMIAAAWVLS FVLSTPQYFVFS--MIEVNNVTKARDCWATFIQPW---GSRAYVTWMTGGIFVAPVVILGTCYGFICYN I/KSISRAKIRTVKMTFVIVTAYIVCWAPFFIIQMWSVWDPMSVWTESENPTITITALLGSLNSCCNPW IYMFFSGHLLQDCVQSFPCC* >P1;rv1ar sequence:rv1ar: : : : :::-1.00:-1.00 PLGDVRNEELAKLEIAVLAVIFVVAVLGNSSVLLALHRTPRKTSRMHLFIRHLSLADLAVAFFQVLPQL CWDITYRFRGPDWLCRVVKHLQVFAMFASAYMLVVMTADRYIAVCHPLKTLQQPARRSRLMIATSWVLS FILSTPQYFIFSVIEIEVNNGTKTQDCWATFIQPW---GTRAYVTWMTSGVFVAPVVVLGTCYGFICYH I/KSISRAKIRTVKMTFVIVSAYILCWAPFFIVQMWSVWDENFIWTDSENPSITITALLASLNSCCNPW IYMFFSGHLLQDCVQSFPCC*

Figure 13.3 The edited multiple sequence alignment ﬁle (huv1ar-1U19.pir) containing the sequences for the template (bRh, 1U19.pdb), the target sequence (huV1aR) and a homologous protein to the target sequence (rV1aR). Large insertions or regions of uncertainty have been deleted and replaced by the symbol ‘/’ (shown in grey) which denotes a break in the protein backbone where a large inserted region has been removed. In addition, the number of residues in ﬁeld 5 of the structure entry (1U19.pdb) has been changed to reﬂect the changes made in the alignment. See the MODELLER manual for more information. 8 Use the program DeepView to open and edit the template ﬁle so that it corresponds to changes made in the edited sequence alignment (see Figures 13.2 and 13.3). This can be achieved by highlighting the relevant residues in the control panel (Window > Control Panel) and deleting them by using the Build > Remove Selected Residues function. 9 Use DeepView to renumber the edited template ﬁle starting from 1 using the Edit > Rename Current Layer function. 10 Save the DeepView edited ﬁle using the Save > Layer function. Notes a The

PIR format (see Figure 13.1) for use in MODELLER is as follows. The ﬁrst line contains the sequence identiﬁer, in the format ‘>P1;seqname’, where seqname is the name of the sequence. The identiﬁer must be unique for all proteins and sequences in the ﬁle. The second line contains 10 ﬁelds separated by colons. However, only ﬁelds 1–6 are required for MODELLER use. Field 1 speciﬁes whether or not the data is from a 3D structure (structure) or is sequence (sequence) data. Field 2 (1U19.pdb, huv1ar and rv1ar respectively) does not have to be unique but must correspond to the variable ‘sequence’ in the MODELLER Python script. Fields 3–6 are the residue

13.2 METHODS AND APPROACHES

259

and chain identiﬁers for the ﬁrst (ﬁelds 3 and 4) and last residue (ﬁelds 5 and 6) of the structure ﬁle. Fields 7–10 are optional; see MODELLER manual for more information. The remainder of the data contains the sequence of protein, with an asterisk marking the end of the entry. The standard upper case, one-letter amino acid codes are used. b

This script will align the sequences in the ﬁle huv1ar.seq and outputs the ﬁle huv1ar-1U19.pir which contains the sequence alignment in PIR format. The salign module accounts for structural information from the template when constructing a sequence alignment using a gap penalty function that inserts gaps in loop regions that are outside regions of secondary structure; see MODELLER manual for more information regarding format. c A further consideration is the length of inserted regions, denoted by gaps (-) between the target and template sequences. Whereas short insertions in the template alignment may be tolerated (one to four residues), larger insertions may result in ‘knots’ in the ﬁnal structure. Therefore, it is a prudent step to delete any region of large insertions in the sequence alignment (shaded in grey; see Figures 13.2 and 13.3). It is important to maintain the overall alignment, as changes made to one sequence must also be reﬂected in the other sequences. Special attention should be paid to ensure that conserved residues are aligned between the sequences. Any errors introduced at this point will seriously affect the ﬁnal model.

13.2.3 Homology modelling of a GPCR Once a suitable template has been found and an alignment between the target and the template structures has been generated, a model of the target protein can be created (Protocol 13.2). The most common method to generate homology models is through the satisfaction of spatial constraints, much in the same way as NMR structures are generated (Protocol 13.3). An example of a program that uses this style of homology modelling is MODELLER [32]. Spatial restraints can be obtained from a number of sources that include homology-derived restraints and molecular-mechanics-based or statistically derived preferences for bonded and nonbonded interactions. In addition, restraints may also be obtained from a number of external source, such as spin labelling [33], ﬂuorescence labelling [34], cysteine cross-linking [33], engineered Zn-binding sites [35] and double mutant constructs [36]. A number of publicly accessible programs are available to verify homology models, and these generally belong to one of two categories. The ﬁrst category (e.g. PROCHECK [37] and WHATIF [38]) checks for proper protein stereochemistry, such as symmetry checks, geometry checks and structural packing quality. The second category (e.g.VERIFY3D [39] and PROSAII [40]) checks the ﬁtness of sequence to structure and assigns a score for each residue ﬁtting its current environment. Finally, of course, the best discriminator between good and bad models is through human intervention, and this should be employed at each stage of the modelling process. Generating a homology model using MODELLER consists of the following steps: 1 preparing an input MODELLER Python script; 2 ensuring that all required ﬁles, including sequences, structures and/or alignments, are located the same directory;

260

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

3 executing a MODELLER input to generate the homology models; 4 analysing the output model and scoring.

PROTOCOL 13.2 Generating a Homology Model for the Human Vasopressin1a Receptor Based on the Structure 1U19 The following procedure uses MODELLER to generate ﬁve homology models of the huv1ar using the edited alignment generated in Protocol 13.1 and the bRh structure 1U19.

Requirements for Protocol 13.2 • The edited alignment ﬁle (huv1ar-1U19.pir) from Protocol 13.1. • The edited structural template 1U19.pdb from Protocol 13.1. • A text editor. • MODELLER installed on either a local or networked computer.

Method 1 Edit a text ﬁle such that it contains the following MODELLER Python script and save the ﬁle as huv1ar-model.py: from modeller import* from modeller.automodel import* env = environ() a = automodel(env, alnfile=’huv1ar-1U19.pir’, knowns=’1U19.pdb’, sequence=’huv1ar’) a.starting_model = 1 a.ending_model = 5 a.make()

2 Execute the MODELLER Python script with the commanda $> MODHOME/modeller9v1/bin/mod9v1 huv1ar-model.py &

3 Once the models have been generated, order them on the basis of the MODELLER objective function using the following command at the UNIX prompt:b $> grep ‘MODELLER OBJECTIVE FUNCTION:’ huv1ar.B9999*.pdb| sort +1

4 Visually inspect the top, lowest scoring, models using DeepView and observe differences between the side-chain packing of the models. Notes a Once

an alignment between the target and template has been generated, MODELLER can calculate a basic 3D model of the target sequence automatically using the automodel routine.

13.2 METHODS AND APPROACHES

261

In this script, the line ‘alnﬁle’ deﬁnes the ﬁlename for the alignment between target and template in PIR format (huv1ar-1U19A.pir), ‘knowns’ deﬁnes the ﬁlename of the template structure (1U19.pdb) and sequence deﬁnes the name of the target sequence (huv1ar) in the alignment ﬁle. The two variables starting_model and ending_model limit the number of models that are calculated (ﬁve in the above example). However, more models should be generated (∼200) for homology modelling and only ﬁve models were generated with the above script for the purposes of illustration and speed. In addition to a number of housekeeping ﬁles (see MODELLER manual for more information), the ﬁles containing the coordinates of the models are deﬁned with the sufﬁx B9999000∗ .pdb, where ∗ is the model number. Visual inspection of the models can then be performed using a package such as DeepView. The symbol ‘&’ allows the program to run in the background, which allows the user to log out of the system. b One

method of obtaining a good structure is to use the MODELLER objective function found in the model PDB ﬁle. This value is not an absolute measure of model quality and can only be used to rank models calculated from the same alignment. As with most objective functions, the lower the value, the better the model is. Although reliable, the MODELLER objective function should be used in combination with other energy/scoring functions, thus forming a consensus approach to picking the best model. In this way, the relative strengths and weaknesses of the various scoring functions will not bias the results of the modelling exercise. However, it is important to note that, before any external evaluation of the model, one should check the log ﬁle from the MODELLER run for runtime errors and restraint violations (see the MODELLER manual). grep is a command-line utility that is found in most UNIX-style operating systems, and given a list of ﬁles (huv1ar.B9999∗ .pdb) grep searches for lines of text that match a regular expressions (MODELLLER OBJECTIVE FUNCTION) and outputs only the matching lines. The output of grep is sent, or piped, using the symbol ‘|’ to a second UNIX command called sort. This command is able to order the output of grep numerically, such that it is easy to identify the best (lowest) scoring model.

PROTOCOL 13.3 Generating a Homology Model for the Human Vasopressin1a Receptor Based on the Structure 1U19 Incorporating External Restraints The following procedure uses MODELLER to generate ﬁve homology models of the huv1ar using the edited alignment generated in Protocol 13.1 and the bRh structure 1U19. In addition, a number of external distance constraints will also be included which will be incorporated into the ﬁnal model.

Requirements for Protocol 13.3 • The edited alignment ﬁle (huv1ar-1U19.pir) from Protocol 13.1. • The edited structural template 1U19.pdb from Protocol 13.1. • A text editor. • MODELLER installed on either a local or networked computer.

262

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

Method 1 Edit a text ﬁle such that it contains the following MODELLER Python script and save the ﬁle as huv1ar-model-user-restraints.py: # Addition of restraints to the default ones from modeller import * from modeller.automodel import* # Load the automodel class log.verbose() env = environ() # directories for input atom files env.io.atom_files_directory = ‘./:../atom_files’ class mymodel(automodel): def special_restraints(self, aln): rsr = self.restraints at = self.atoms # Residues 1 through 10 should be an alpha helix: rsr.add(secondary_structure.alpha(self.residue_range(’1:’ ‘10:’))) # Use a harmonic potential between residues 29 and 57. rsr.add(forms.gaussian(group=physical.xy_distance, feature=features.distance(at[’CA:29’], at[’CA:57’]), mean=6.7, stdev=0.1)) a = mymodel(env, alnfile = ‘huv1ar-1U19.pir ‘, knowns = ‘1U19.pdb’, sequence = ‘huv1ar’) a.starting_model= 1 a.ending_model =1 a.make()

2 Execute the MODELLER Python script with the commanda $> MODHOME/modeller9v1/bin/mod9v1 huv1ar-model-user-restraints.py. &

3 Once the models have been generated, order them on the basis of the MODELLER objective function using the following UNIX command at the UNIX prompt:b $> grep ‘MODELLER OBJECTIVE FUNCTION:’ huv1ar.B9999*.pdb| sort +1

4 Visually inspect the top models using DeepView and observe the effects of the additional distance restraints on the ﬁnal structure.

13.2 METHODS AND APPROACHES

263

13.2.4 Loop prediction Unlike the transmembrane domains, loop regions connecting the helices are quite diverse in size and amino acid composition, which makes homology modelling of these regions unreliable. Furthermore, the importance of the loops is such that any model of a GPCR without adequate consideration of these regions would be incomplete. The goal of loop prediction is to identify the correct conformation of a protein fragment from a series of decoys in the context of the remaining protein. This is not a trivial task, as loop regions, which can included both insertions and deletions, often share little sequence homology with the template. Traditionally, two approaches have been used for loop prediction: database methods [41, 42] and ab initio methods [43, 44]. Database methods attempt to ﬁnd protein fragments from existing structures that are approximately the same size as the desired loop section. This is followed by the evaluation of suitable candidates and optimization by means of an energy function. However, for database methods, loops longer than ﬁve residues often have problems identifying near-native conformations in the template library, which limits their usefulness in a number of cases [42]. Ab initio methods, however, do not require a homologous protein fragment, but involve the generation of a large ensemble of candidate loop structures from which a singleconformation or cluster of conformations is picked using an energy-based scoring function. Current methods have shown that ab initio prediction of regions of up to 12 residues give good agreement with experimental data [45]. However, one limit to the effectiveness of ab initio loop prediction is the scoring function [43]. Traditionally, detailed atomic force ﬁelds, either with or without a solvation term, have been used successfully to identify native loop fragments. However, although effective, these methods are often computationally intensive and require an initial ﬁlter step to remove nonnative loop fragments [44]. One such advance in loop prediction is through the development of knowledge-based or statistical potentials, which can be as effective at predicting the correct loop conformation as more rigorous force-ﬁeld-based approaches, but at a fraction of the computational cost. A recently developed statistical potential is the distance-scaled, ﬁnite ideal-gas reference (DFIRE) potential [46], which was shown to be comparable to that of an Assisted Model Building and Energy Requirement (AMBER) generalized Born surface area (GB/SA) force ﬁeld for short loops of two to eight residues [47]. Once programs such as DFIRE ﬁlter out nonnative loop conformations, traditional physical-based scoring methods can be used to identify the native loop conformation. One program that has been shown to be effective for ab initio loop prediction is RAPPER [44], which is a conformational search algorithm for restraint-based protein modelling. Although RAPPER can be accessed through the Internet (http://mordred.bioc.cam.ac.uk/∼rapper/loop2.php), a Linux binary is also available that allows for greater control over the loop generation process and can be accessed from the UNIX command line. A typical ab initio loop prediction protocol using RAPPER consists of the generation of a large number of loop conformations with idealized stereochemistry for the heavy atoms (N, Cα , C, O). Atomic van der Waals radii are scaled down by 25% to ensure good packing and loop fragments generated that contain clashing atoms are discarded. A distance geometry check is also performed on the ﬁnal modelled loop fragment such that the N-terminal Cα

264

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

anchor atom of the framework and the C-terminal Cα atom of the loop fragment are ˚ To ensure conformational diversity of the loop ensemble generated, within 3.7 A. no two fragments are allowed which have a global root-mean-squared deviation ˚ The loop fragments generated are then joined to the backbone of the of <0.2 A. remaining protein and ﬁltered using the DFIRE statistical potential, from which the top 50 are retained and scored using a physical-based scoring method such as the AMBER GB/SA force ﬁeld as implemented in the TINKER program package (http://dasher.wustl.edu/tinker/). Minimization is performed for 100 steps or a 0.1 kcal mol−1 cut-off point is reached.

PROTOCOL 13.4 Modelling Loops in GPCRs To demonstrate the effectiveness of ab initio loop prediction, the following example is to predict the conformation of extracellular loop 1 from the bRh crystal structure 1U19.

Requirements for Protocol 13.4 Internet access • A text editor. • DeepView installed on a local computer. Software to obtain and install • RAPPER http://mordred.bioc.cam.ac.uk/∼rapper/ • DFIRE (available to academic groups on request) http://sparks.informatics.iupui.edu/index.php?pageLoc=Downloads • TINKER http://dasher.wustl.edu/tinker

Computer Skills Unlike MODELLER, the programs RAPPER, DFIRE and TINKER are only easily available under a UNIX-based operating system. Furthermore, the above programs do not contain a graphical user interface and, as such, a basic knowledge of the command line is necessary to follow the protocol in this section.

Conventions Followed in the Text The variables RAPPERHOME and TINKERHOME refer to the path of the RAPPER and TINKER home directories respectively and should be typed in full. Text in Courier New font refers to commands that are entered at the command line, as a script or are the contents of a ﬁle. Comments between # and # are for information only and are not required to be typed in as part of the command. The symbol ‘∗’ refers to a wildcard and may represent any character. Separate commands for either scripts or commands should be typed as a single line and are only shown as separate lines for clarity. UNIX commands are represented in the main text as italics. The symbols ‘$>’ refers to commands which should be applied directly to the command line and do

13.2 METHODS AND APPROACHES

265

not form part of a script. Files with a ‘.key’ are keyﬁles used with the TINKER suite of programs and contain arguments for use with the programs contained in this protocol. See the TINKER manual for more information on the keyﬁle arguments.

Method 1 Obtain and install, under a UNIX-based operating system, the programs RAPPER, DFIRE and TINKER. 2 Download the PDB ﬁle 1U19 from the PDB repository (http://www.rcsb.org/pdb/home/home.do). 3 Copy the ﬁle 1U19.pdb to the RAPPERHOME/rappermc/distrib directory.a 4 Generate 1000 loop models for the region between residues 98 and 107 of extracellular loop 1 for the PDB ﬁle 1U19 using RAPPER with the following command line:a $> RAPPERHOME/rappermc/distrib/rapper_Linux_i386 RAPPERHOME/rappermc/distrib/params_Linux_i386.xml model-loops-benchmark --rapper-dir RAPPERHOME/rappermc --verify-setup false --models 1000 #number of models to be generated # --pdb 1U19.pdb #PDB code of template # --start 98 #start residue # --stop 107 #end residue # --population-max-passes 2000 --divide-and-conquer false --ignore-hetatms true --sidechain-mode smart --sidechain-library RAPPERHOME/rappermc/data/richardson.lib --sidechain-radius-reduction 0.75 --verify-structures true --enforce-strict-anchor-geometry true --verify-structures-for-energy-calculations true --use-contact-filters true --write-individual-models true >&rhod-ecl1& # enables RAPPER to run in the background #

5 Once RAPPER has generated 1000 loop conformations for the region 98–107 of the PDB ﬁle 1U19, join the fragment PDB ﬁles to the framework PDB ﬁle (both are generated by RAPPER) by editing a text ﬁle such that it contains the following lines of script and save the ﬁle as ‘join’:b cd RAPPERHOME/rappermc/distrib/TESTRUNS foreach pdb (looptest*pdb) RAPPERHOME/rappermc/distrib/rapper_Linux_i386 RAPPERHOME/rappermc/distrib/params_Linux_i386.xml joinpdb

266

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

--pdb1 framework.pdb --pdb2 $pdb --pdb-out $pdb.pdb --rapper-dir RAPPERHOME/rappermc end

6 Execute the script ‘join’ by using the command $> csh join

7 Whilst in the directory RAPPERHOME/rappermc/distrib/TESTRUNS, execute the following UNIX command to generate a ﬁle containing all the complete receptor PDB ﬁles:c $> ls * > loops

8 Edit the ﬁle ‘loops’ using the UNIX command sed such that it contains the number 10 after each of the entries:d $> sed ‘s/looptest.*/& 10/g’ loop > loop.in

9 Copy the program DFIRE and associated ﬁles to the RAPPERHOME/rappermc/distrib/TESTRUNS directory. 10 Execute the program DFIRE using the ﬁle loop.in as input and redirect the resulting output to the ﬁle loop.out using the following UNIX command line: $> ./dfire < loop.in > loop.out

11 Numerically sort the ﬁle loop.out using the UNIX command sort on the basis of the DFIRE score and retain the ﬁlenames of the 50 best-scoring loop predictions in a ﬁle called best-loops:e $> sort +1 – g loop.out | head – n 50 >best-loops

12 Copy the top 50 best-scoring loop predictions ﬁles to a separate directory. 13 Covert these PDB ﬁles to TINKER XYZ format by editing a text ﬁle such that it contains the following lines of script and save the ﬁle as pdb2xyz:f foreach pdb (*pdb) TINKERHOME/tinker/bin/pdbxyz.x – k pdbxyz.key $pdb end

14 Edit a ﬁle such that it contains the following statements and save it as ‘pdbxyz.key’: parameters TINKERHOME/tinker/params/amber99.prm

13.2 METHODS AND APPROACHES

15 Execute the script ‘pdb2xyz’ by using the command $> csh pdb2xyz

16 Once the ﬁle format conversion has been completed, minimize the predicted loop using the AMBER99 GB/SA force ﬁeld as implemented in TINKER by editing a text ﬁle such that it contains the following lines of script and save the ﬁle as ‘minimize-loop’: foreach xyz (*xyz) TINKERHOME/tinker/bin/minimize.x – k minimize.key $xyz 0.1 end

17 Edit a ﬁle such that it contains the following statements and save it as ‘minimize.key’: parameters TINKERHOME/tinker/params/amber99.prm solvate still inactive – 1 1576 # atom numbers to be restrained active – 1577 1729 # atom numbers to minimize inactive – 1730 5468 # atom numbers to be restrained maxiter 100

18 Execute the script ‘minimize-loop’ by using the command $> csh minimize-loop >& loop-minimize &

19 Analyse the resulting minimized loop structures using the AMBER99 GB/SA forceﬁeld using the following UNIX script (analyze) and keyﬁle (analyze.key): foreach xyz_2 (*.xyz_2) TINKERHOME/tinker/bin/analyze.x – k analyze.key $xyz_2 e > $xyz_2.analysed end

the keyﬁle ‘analyze.key’ contains the following statements: parameters TINKERHOME/tinker/params/amber99.prm solvate still

20 Order the loop models on the basis of the energy function obtained in the previous step: $> grep ‘Energy’ *xyz_2.analysed | sort +1

21 Retain and visually inspect the top 10 loop structures using DeepView and compare them with the region 98–107 of the structure 1U19.

267

268

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

Notes a Further

information regarding the functions of the input ﬂags can be obtained using the help function under RAPPER. Whilst a link may be made to the RAPPER programs and associated ﬁles, this example is run directly from the RAPPERHOME/rappermc/distrib directory. Depending on the size of the loop to be modelled, the run time of RAPPER may be between hours or days. The last line of the command (>&rhod-ecl1&) allows RAPPER to run as a background job. b The

script ‘join’ uses the UNIX command foreach to merge the framework.pdb ﬁle generated by RAPPER with each of the loop decoys (looptest*.pdb). foreach is usually used in place of a standard for statement and essentially performs a task on everything in a set, rather than performing a speciﬁc task x times to one ﬁle. The output of the script is a set of ﬁles with the sufﬁx ∗ pdb.pdb. These ﬁle contain the complete receptor protein and should be archived for future use. inspection of the ﬁle ‘loops’ should reveal one entry for each of the ∗ pdb.pdb ﬁles located in the RAPPERHOME/rappermc/distrib/TESTRUNS directory as shown below.

c Visual

The ﬁle ‘loop’ looptest-0.pdb.pdb looptest-100.pdb.pdb looptest-101.pdb.pdb looptest-102.pdb.pdb ... ... looptest-9.pdb.pdb d The

product of the sed command is the ﬁle ‘loop.in’. Visual inspection of this ﬁle should reveal the same list of entries as found in the ‘loop’ ﬁle with the number 10 after each looptest∗ pdb.pdb. The command sed (stream editor) refers to a UNIX utility for parsing text ﬁles and applying the operation which has been speciﬁed via the command line. The outputs are then directed to either the screen or a ﬁle. The number 10 which was applied to the ﬁle loop as part of the sed command is arbitrary and is a format requirement for the correct operation of the program DFIRE. The ﬁle ‘loop.in’ looptest-0.pdb.pdb 10 looptest-100.pdb.pdb 10 looptest-101.pdb.pdb 10 looptest-102.pdb.pdb 10 ... ... looptest-9.pdb.pdb 10 e The

UNIX command sort is used here to order the output from DFIRE numerically, the results of which are piped to the UNIX command head that retains the top 50 entries and saves them under the ﬁlename ‘best-loops’. At this point the directory RAPPERHOME/rappermc/distrib/TESTRUNS should be archived to a permanent storage medium.

f The

PDB ﬁles, keyﬁle script and executable script should all be present in the same directory.

13.4 AUTOMATED METHODS FOR GENERATING MODELS OF GPCRs

269

13.3 Troubleshooting • Knots in ﬁnal homology structure: Initially, knots, or regions of entangled protein chains, may be a direct result of a poor initial alignment. Visual inspection and editing of the alignment should be performed and adjusted or deleted accordingly. If the alignment is correct and regions of uncertainty have been deleted, then one solution is to calculate independently many models and to ensure that a population exists which is free of tangled protein. In addition, extending the optimization time for the model-building process may reduce the occurrence of the knots. • Python formatting matters: Failure of the MODELLER Python script to initiate may be as a result of a poorly formatted ‘.py’ ﬁle. Familiarity with Python expressions may save a lot of time and frustration. There are many resources for learning Python itself, such as a series of comprehensive tutorials at http://www.python.org/doc/.

13.4 Automated methods for generating models of GPCRs Whilst GPCR models can be generated using the techniques described above, a number of automated servers have been developed that provide a fast and reliable route for the creation of homology models and require just the target sequence or alignment with a suitable template. The Expert Protein Analysis System (ExPASy) web site (http://ca.expasy.org/ [48]) can be used for the retrieval of a protein sequence or more complex tasks, including the analysis and identiﬁcation of related sequences. Several dedicated databases for studying GPCRs, such as the GPCRDB (http://www.gpcr.org/7tm/ [49]), supplement ExPASy-like web sites and include specialized tools for examining this class of proteins. As with manual modelling methods, automated modelling servers often require an alignment between the template and target sequences. To facilitate this, a number of freely accessible servers are available, including T-COFFEE (http://www.igs.cnrs-mrs.fr/Tcoffee/tcoffee cgi/index.cgi) and CLUSTALW (http://www.ebi.ac.uk/Tools/clustalw/). The subsequent alignment can then be submitted to an automated modelling server to generate a homology model of the protein of interest. Two examples of automated modelling servers are MODBASE (http://salilab.org/modbase/ [50]) and SWISS MODEL (http://swissmodel.expasy.org/SWISS-MODEL.html [51]), both of which also act as repositories for previously generated models. Whilst both servers are capable of generating high-quality models of GPCRs, SWISS MODEL does have the advantage of supporting a dedicated GPCR modelling section. In addition, SWISS MODEL also enables the user to choose a speciﬁc template from a list of available and suitable structures. The use of different templates can then lead to a series of models that can be ﬁltered further using methods such as experimental data. Whilst the above methods can be used for automatically generating homology models of related sequences, distantly related targets require a subtly different

270

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

approach. One service is the Structure Prediction Meta Server (http://meta. bioinfo.pl/submit wizard.pl [52]), which provides access to various fold recognition, function prediction and local structure prediction methods. The output from the server includes the PDB codes of the hits, alignments, reliability scores for every server and provides a jury prediction based on the results collected from other services. Allied methods to this include the servers from the Fiser lab (http://www.ﬁserlab.org/servers table.htm [53]), which provides access to various fold recognition, function prediction and local structure prediction methods. In cases where there is no obvious template for the target, ab initio modelling servers may be used to generate of a series models which can be ﬁltered using knowledge-based approaches. An example of this is the ROBETTA server (http://robetta.bakerlab.org/ [54]), which is an automated front end to the successful ROSETTA method [55]. This service requires only the protein sequence to generate molecular models, and with the success of ROSETTA in recent CASP experiments, this method is a good starting point for generating models which may be tested experimentally. Whilst these methods may appear to be a black-box approach to generate protein structure, a series of checks needs to be performed before the models are used to analyse or suggest experiments. These checks include stereochemistry checks using modelling servers such as WHATIF and PROCHECK and a careful examination of the alignments used to generate models, as well as a visual examination of the model using programs such as DeepView. External to these checks are services such as EVA (http://eva.compbio.ucsf.edu/∼eva/ [56]), which provides a continuous, fully automated and statistically signiﬁcant analysis of structure prediction servers. This could be checked prior to using a particular server to assess the validity of the modelling routines for the target of choice.

References 1. Gudermann, T., Nurnberg, B. and Schultz, G. (1995) Receptors and G proteins as primary components of transmembrane signal transduction. Part 1. G-protein-coupled receptors: structure and function J. Mol. Med., 73, 51–63. 2. Palczewski, K., Kumasaka, T., Hori, T. et al. (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science, 289, 739–745. The ﬁrst three-dimensional structure of a GPCR, which has enabled the generation of homology models for this family of membrane proteins. 3. Cherezov, V., Rosenbaum, D.M., Hanson, M.A. et al. (2007) High-resolution crystal structure of an engineered human β2 -adrenergic G protein-coupled receptor. Science, 318, 1258–1265. 4. Forrest, L.R., Tavoulari, S., Zhang, Y.W. et al. (2007) Identiﬁcation of a chloride ion binding site in Na+ /Cl− -dependent transporters. Proc. Natl. Acad. Sci. U. S. A., 104, 12761–12766. 5. Wieman, H., Tøndel, K., Anderssen, E. and Drabløs, F. (2004) Homology-based modelling of targets for rational drug design. Mini Rev. Med. Chem., 4, 793–804. 6. Berman, H.M., Westbrook, J., Feng, Z. et al. (2000) The Protein Data Bank. Nucleic Acids Res. 28, 235–242.

REFERENCES

271

7. Altschul, S.F., Gish, W., Miller, W. et al. (1990) Basic local alignment search tool. J. Mol. Biol., 215, 403–410. 8. Altschul, S.F., Madden, T.L., Sch¨affer, A.A. et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res., 25, 3389–3402. 9. Krogh, A., Brown, M., Mian, I.S. et al. (1994) Hidden Markov models in computational biology: applications to protein modeling. J. Mol. Biol., 235, 1501–1531. 10. Karplus, K., Barrett, C. and Hughey, R. (1998) Hidden Markov models for detecting remote protein homologies. Bioinformatics, 14, 846–856. 11. Eddy, S.R. (1998) Proﬁle hidden Markov models. Bioinformatics, 14, 755–763. 12. Okada, T., Sugihara, M., Bondar, A.N. et al. (2004) The retinal conformation and its environment ˚ crystal structure. J. Mol. Biol., 342, 571–583. in rhodopsin in light of a new 2.2 A 13. Li, J., Edwards, P.C., Burghammer, M. et al. (2004) Structure of bovine rhodopsin in a trigonal crystal form. J. Mol. Biol., 343, 1409–1438. 14. He, X.L., Bazan, J.F., McDermott, G. et al. (2003) Structure of the Nogo receptor ectodomain: a recognition module implicated in myelin inhibition. Neuron, 38, 177–185. 15. Kobe, B. and Deisenhofer, J. (1996) Mechanism of ribonuclease inhibition by ribonuclease inhibitor protein based on the crystal structure of its complex with ribonuclease A. J. Mol. Biol., 264, 1028–1043. 16. Grace, C.R., Perrin, M.H., DiGruccio, M.R. et al. (2004) NMR structure and peptide hormone binding site of the ﬁrst extracellular domain of a type B1 G protein-coupled receptor. Proc. Natl. Acad. Sci. U. S. A., 101, 12836–12841. The ﬁrst reported structure of the N-terminal domain of a Family B GPCR. 17. Sun, C., Song, D., Davis-Taber, R.A. et al. (2007) Solution structure and mutational analysis of pituitary adenylate cyclase-activating polypeptide binding to the extracellular domain of PAC1-RS. Proc. Natl. Acad. Sci. U. S. A., 104, 7875–7880. 18. Grace, C.R., Perrin, M.H., Gulyas, J. et al. (2007) Structure of the N-terminal domain of a type B1 G protein-coupled receptor in complex with a peptide ligand. Proc. Natl. Acad. Sci. U. S. A., 104, 4858–4863. 19. Parthier, C., Kleinschmidt, M., Neumann, P. et al. (2007) Crystal structure of the incretin-bound extracellular domain of a G protein-coupled receptor. Proc. Natl. Acad. Sci. U. S. A., 104, 13942–13947. 20. Tsuchiya, D., Kunishima, N., Kamiya, N. et al. (2002) Structural views of the ligand-binding cores of a metabotropic glutamate receptor complexed with an antagonist and both glutamate and Gd3+ . Proc. Natl. Acad. Sci. U. S. A., 99, 2660–2665. 21. Muto, T., Tsuchiya, D., Morikawa, K. and Jingami, H. (2007) Structures of the extracellular regions of the group ii/iii metabotropic glutamate receptors. Proc. Natl. Acad. Sci. U. S. A., 104, 3759–3764. 22. Benner, S.A., Cohen, M.A. and Gonnet, G.H. (1994) Amino acid substitution during functionally constrained divergent evolution of protein sequences. Protein Eng., 7, 1323–1332. 23. Henikoff S. and Henikoff J.G. (1992) Amino acid substitution matrices from protein blocks. Proc. Natl. Acad. Sci. U. S. A., 89, 10915–10919. 24. Ng, P.C., Henikoff, J.G. and Henikoff, S. (2000) PHAT: a transmembrane-speciﬁc substitution matrix. Predicted hydrophobic and transmembrane. Bioinformatics, 16, 760–796.

272

CH 13 HOMOLOGY MODELLING OF G PROTEIN-COUPLED RECEPTORS

25. Muller, T., Rahmann, S. and Rehmsmeier, M. (2001) Non-symmetric score matrices and the detection of homologous transmembrane proteins. Bioinformatics, 17 ( Suppl 1), S182–S189. 26. Forrest, L.R., Tang, C.L. and Honig, B. (2006) On the accuracy of homology modeling and sequence alignment methods applied to membrane proteins. Biophys. J , 91, 508–517. 27. Shafrir, Y. and Guy, H.R. (2004) STAM: simple transmembrane alignment method. Bioinformatics, 20, 758–769. 28. Notredame, C., Higgins, D.G. and Heringa, J. (2000) T-Coffee: a novel method for fast and accurate multiple sequence alignment. J. Mol. Biol., 302, 205–217. 29. Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-speciﬁc gap penalties and weight matrix choice. Nucleic Acids Res., 22, 4673–4680. 30. Ballesteros, J.A., Shi, L. and Javitch, J.A. (2001) Structural mimicry in G-protein-coupled receptors: implications of the high-resolution structure of rhodopsin for structure-function analysis of rhodopsin-like receptors. Mol. Pharmacol., 60, 1–19. An interesting article which draws ideas from the structure of bovine rhodopsin to rationalize the function of GPCRs as a whole. 31. Guex, N. and Peitsch, M.C. (1997) SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis, 18, 2714–2723. 32. Sali, A. and Blundell, T.L. (1993) Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol., 234, 779–815. 33. Farrens, D.L., Altenbach, C., Yang, K. et al. (1996) Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science, 274, 768–770. 34. Gether, U., Lin, S., Ghanouni, P. et al. (1997) Agonists induce conformational changes in transmembrane domains III and VI of the β2 adrenoceptor. EMBO J., 16, 6737–6747. 35. Elling, C.E., Thirstrup, K., Nielsen, S.M. et al. (1997) Metal-ion sites as structural and functional probes of helix–helix interactions in 7TM receptors. Ann. N. Y. Acad. Sci., 814, 142–151. 36. Zhou, W., Flanagan, C., Ballesteros, J.A. et al. (1994) A reciprocal mutation supports helix 2 and helix 7 proximity in the gonadotropin-releasing hormone receptor. Mol. Pharmacol., 45, 165–170. 37. Laskowski, R.A., Rullmannn, J.A., MacArthur, M.W. et al. (1996) AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR. J. Biomol. NMR, 8, 477–486. 38. Hooft, R.W., Vriend, G., Sander, C. and Abola, E.E. (1996) Errors in protein structures. Nature, 381, 272. 39. Eisenberg, D., L¨uthy, R. and Bowie, J.U. (1997) VERIFY3D: assessment of protein models with three-dimensional proﬁles. Methods Enzymol., 277, 396–404. 40. Sippl, M.J. (1993) Recognition of errors in three-dimensional structures of proteins. Proteins, 17, 355–362. 41. Fiser, A., Do, R.K. and Sali, A. (2000) Modeling of loops in protein structures. Protein Sci., 9, 1753–1773. 42. Deane, C.M. and Blundell, T.L. (2001) CODA: a combined algorithm for predicting the structurally variable regions of protein models. Protein Sci., 10, 599–612.

REFERENCES

273

43. Soto, C.S., Fasnacht, M., Zhu, J. et al. (2008) Loop modeling: sampling, ﬁltering, and scoring. Proteins, 70, 834–843. 44. de Bakker, P.I., DePristo, M.A., Burke, D.F. and Blundell, T.L. (2003) Ab initio construction of polypeptide fragments: accuracy of loop decoy discrimination by an all-atom statistical potential and the AMBER force ﬁeld with the generalized Born solvation model. Proteins, 51, 21–40. 45. Zhu, K., Pincus, D.L., Zhao, S. and Friesner, R.A. (2006) Long loop prediction using the protein local optimization program. Proteins, 65, 438–452. 46. Zhou, H. and Zhou, Y. (2002) Distance-scaled, ﬁnite ideal-gas reference state improves structure-derived potentials of mean force for structure selection and stability prediction. Protein Sci., 11, 2714–2726. The ﬁrst report of a statistical potential with comparable accuracy to full atomic force ﬁelds. 47. Zhang, C., Liu, S. and Zhou, Y. (2004) Accurate and efﬁcient loop selections using DFIRE-based all-atom statistical potential. Protein Sci., 13, 391–399. 48. Gasteiger, E., Gattiker, A., Hoogland, C. et al. (2003) ExPASy: the proteomics server for in-depth protein knowledge and analysis. Nucleic Acids Res., 31, 3784–3788. 49. Horn, F., Bettler, E., Oliveira, L. et al. (2003) GPCRDB information system for G protein-coupled receptors. Nucleic Acids Res., 31, 294–297. 50. Pieper, U., Eswar, N., Davis, F.P. et al. (2006) MODBASE: a database of annotated comparative protein structure models and associated resources. Nucleic Acids Res., 34, D291–D295. 51. Schwede, T., Kopp, J., Guex, N. and Peitsch, M.C. (2003) SWISS-MODEL: an automated protein homology-modeling server Nucleic Acids Res., 31, 3381–3385. 52. Ginalski, K., Elofsson, A., Fischer, D. and Rychlewski, L. (2003) 3D-Jury: a simple approach to improve protein structure predictions. Bioinformatics, 19, 1015–1018. 53. Rai, B.K., Madrid-Aliste, C.J., Fajardo, J.E. and Fiser, A. (2006) MMM: a sequence-to-structure alignment protocol. Bioinformatics, 22, 2691–2692. 54. Kim, D.E., Chivian, D. and Baker, D. (2004) Protein structure prediction and analysis using the Robetta server. Nucleic Acids Res., 32, W526–W531. 55. Rohl, C.A., Strauss, C.E., Misura, K.M. and Baker, D. (2004) Protein structure prediction using Rosetta. Methods Enzymol., 383, 66–93. 56. Eyrich, V.A., Mart´ı-Renom, M.A., Przybylski, D. et al. (2001) EVA: continuous automatic evaluation of protein structure prediction servers. Bioinformatics, 17, 1242–1243.

Appendix Site-directed Mutagenesis and Chimeras Alex Conner1 , Mark Wheatley2 and David R. Poyner3 1 School

of Medicine, Warwick University, Coventry CV4 7AL, UK of Biosciences, Birmingham University, Birmingham B15 2TT, UK 3 School of Life and Health Sciences, Aston University, Birmingham B4 7ET, UK 2 School

A.1 Introduction The ability to alter the structure of G protein-coupled receptors (GPCRs) at will by site-directed mutagenesis or by generation of receptor chimeras is an essential technique in molecular pharmacology. This appendix will concentrate on the design of mutants for point changes or chimeras, as well as brieﬂy reviewing the main experimental techniques.

A.2 Why mutagenesis? The search for reliable structural data, whether through crystal studies, nuclear magnetic resonance (NMR) or biophysics, is a primary goal towards understanding GPCRs. However, this will not provide information about the movements and contacts made and broken during the transition stages of receptor activation. This also holds true for the contacts made between the receptor and the countless interacting molecules that GPCRs encounter during their surface localization, activation and desensitization. GPCRs are increasingly being recognized not as a simple lock in a ‘lock and key’ mechanism, but as the crux of a ﬂuid receptor-complex heavily dependent on the cellular environment and the available molecular partners. Examples of GPCRs are now widely recognized as functional homodimers and heterodimers [1], and several G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

276

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

are known to require accessory proteins [2]. Their pharmacology can be altered by the membrane constituents, including cholesterol-lipid rafts and the invaginated caveolae [3], and the emphasis on non-G protein signalling of receptors is now established [4]. Clearly, exploring the structural and biochemical requirements of integral domains and individual receptors of GPCRs is crucial towards modelling the shape of activation, a key factor both for targeted therapeutics and fundamental receptor biology. By far the most convenient and powerful technique for altering and assessing GPCR biochemistry involves the alteration of individual amino acids or groups of residues in a manner thought to negate their inﬂuence or to change the inﬂuence in such a way that the role becomes clear. Such mutagenesis has been used successfully for 30 years and can be broadly split into two main approaches. These include a comprehensive random (saturation) mutagenesis, essentially ﬁshing for disrupted or enhanced constraints. The second is a targeted mutagenesis approach, substituting or deleting residues or whole regions within a GPCR based on a hypothesis-led appraisal of key functional parts of the protein.

A.2.1 Random (saturation) mutagenesis Random mutagenesis is a nonprejudiced approach for the isolation of speciﬁc but untargeted residues of structural or functional relevance. This technique results in a signiﬁcant volume of data for analysis and requires a subsequent selection from that pool of data. Originally used in whole organisms more than 50 years ago using radiation or chemical mutagens to alter overt phenotypes, random mutagenesis has recently undergone an upsurge in experimental popularity, moving towards in vitro methods more in line with the directed analysis used elsewhere. This involves the polymerase chain reaction (PCR) misincorporation or ‘gene-shufﬂing’ through recombination strategies for localized random mutagenesis of speciﬁc gene targets or their regulatory elements; see [5] for a review. A recent development in the power of random mutagenesis for the analysis of GPCRs came from the group of Jurgen Wess, studying the M3 muscarinic acetylcholine receptor (M3R). They utilized a combination of low-frequency random mutagenesis of the entire M3R coding sequence, followed by the application of a new yeast genetic screen that allows the recovery of inactivating M3R single-point mutations [6]. This gives a comprehensive understanding of multiple residues in the second extracellular loop, critical for receptor activation [7]. It remains to be seen whether this is more efﬁcient for the systematic analysis of GPCR structure function.

A.2.2 Targeted mutagenesis Targeting residues for mutation takes many forms and is a widely used and powerful form of functional protein analysis leading to the discovery of innumerable speciﬁc functional and structural domains required for a host of protein functions. It is clear that targeted mutagenesis does not necessarily need to start with a precise biochemical understanding of the functional or structural regions to be assessed. On a macroscale,

A.2 WHY MUTAGENESIS?

277

a mining approach to narrow down regions of interest has been consistently and successfully used in the GPCR ﬁeld. Taking two examples, this has led to the discovery of agonist-speciﬁc binding domains in the V1a vasopressin receptor [8] and speciﬁc signalling loci in the metabotropic glutamate receptors [9]. Mutations altering large areas of a protein can take many forms. The relatively simple deletion mutants are, perhaps, easier to design via single PCRs (often through intermediary cloning vectors) and subcloning. C-terminal deletions have the added possibility of point mutations to introduce stop codons which makes them easier to make, although the possibility of missing this codon might require more than one consecutive stop codon; western blots are useful to show the size differential, conﬁrming the deletion status. Peptide mimetics are, in essence, an extreme form of protein deletion using peptide synthesis technology to create small fragments of a protein, either for competition with the natural protein or to analyse interaction with speciﬁed partners in vivo and in vitro. Chimeras are more difﬁcult, but often more revealing when combined with sophisticated data retrieval. Straight hybrids, an N-terminal piece of one receptor fused to the C-terminus of another, are technically challenging but can be made with cloning vectors in a similar way to deletions but with an extra ligation step (see Protocol A.1). The particularly complicated aspect arises when introducing a new piece in the centre of a gene of interest (this is also true for deletions of a central region of a protein). Several methods are used in the literature, including the attractive but costly ‘outsourcing’ to one of a number of companies specializing in speciﬁc gene manipulation [10]. A simple PCR/ligation approach can be used when the domains to be fused are derived from very closely related GPCRs. This technique, which incorporates into the oligonucleotides the same sites for restriction endonucleases at each end of the PCR product as needed for the exchange of corresponding domains, was used successfully in an early study of the functional domains of the Family A vasopressin V2 and oxytocin receptors [11] as well as the similar Family B parathyroid hormone and secretin receptors [12]. This is a relatively easy method for chimera production, but relies heavily on the presence or introduction of common restriction endonuclease sites, which is often limited to the formation of constructs from closely related receptors. Where the precise endonuclease sites are not available, these can sometimes be introduced as silent mutations and incorporated into the ends of the oligonucleotides used to generate a fragment for insertion from the donor receptor. This has been used successfully to create chimeras between members of the bradykinin receptor subfamily [13] and less similar rhodopsin and β2 adrenergic receptors [14]. In Protocol A.1, a generic method is described for production of chimeric mutants. A second example has been described [15] to include a more comprehensive three-step PCR method which can used even where the insertion of suitable sites is not feasible. The gene sequence to be inserted (I) can be ampliﬁed using oligonucleotides designed to have ﬂanking sequences from the target gene (T) corresponding to the region of insertion. This fragment, when ampliﬁed, can be mixed with the gene to be targeted (in its plasmid vector) and the mixture used as a template for a second round of PCR. This requires the separate ampliﬁcation of two products, including the entire upstream sequence of the target gene and the entire downstream sequence of the target gene both

278

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

fused to the insert sequence. This is shown in Figure A.1. The ﬁnal step requires the mixing of both sequences and the subsequent ampliﬁcation of the entire chimeric gene using the extreme oligonucleotide primers only. This method was described previously [15], although not necessarily for GPCRs, and is one of several similar methods of PCR-based chimeric production used, including blunt cloning (Protocol A.1), the introduction of restriction sites for domain swapping [8] and 5 -overhang plasmid extension [16]. For all of these methods, the precise molar ratios have to be adjusted and titrated for optimum efﬁciency. It is a very important consideration, particularly when designing chimeras between closely related receptors, to note that several individual point mutations may lead to a signiﬁcant chimeric effect if the amino acid conservation is high. The technique is much simpler and is described below (Protocol A.2). Furthermore, the problems associated with the construction and analysis of chimeric proteins can outweigh the insight they provide. Predominantly, this includes topology-folding problems and that, due to the spread of important residues which can come together on tertiary/quaternary folding, ‘key areas’ of importance identiﬁed through chimeric receptor analysis often require a subsequent alanine sweep of the individual residues.

PROTOCOL A.1 Production of Chimeric Receptors Equipment and Reagents • Standard PCR thermo cycler (many available) • Pfu DNA polymerase (Promega) and 10 × buffer provided with the Pfu polymerase • Oligonucleotide primers (Invitrogen) • 10 mm deoxyribonucleotide triphosphate (dNTP) mixture (Sigma–Aldrich): a mixture of all four dNTPs (2 -deoxyadenosine 5 -triphosphate (dATP), 2 -deoxycytidine 5 -triphosphate (dCTP), 2 -deoxyguanosine 5 -triphosphate (dGTP), 2 -deoxythymidine 5 -triphosphate (dTTP)) • DNA template plasmid at 100 ng µl−1 containing the gene of interest • 0.5 ml sterile PCR tubes (Sarstedt) • 0.8–1.2% agarose gel • Seaplaque agarose (Lonza, UK) • Agarose gel tank (Bio-Rad Laboratories) • 240 V power pack (Bio-Rad Laboratories) • Desktop microcentrifuge (many available) • Gel extraction kit (Qiagen) • T4 DNA ligase (Invitrogen) • Srf I restriction enzyme (Promega, UK)

A.2 WHY MUTAGENESIS?

• Blunt-cloning plasmid pCR-Script (Stratagene, UK) • Competent cells; for example, XL10-GOLD cells (Stratagene, UK) • Blue–white selective agents isopropyl-β-d-thiogalactopyranoside (IPTG) and X-gal (Stratagene, UK) • Microbiological incubator (many available) • Luria–Bertani (LB) medium (Oxoid, UK) • l-Agar (Oxoid, UK) • Ampicillin (sodium salt; Sigma, UK) • Plasmid isolation kit (Wiz-prep; Promega, UK).

Method 1 Identify the fragments required from each receptor and design PCR primers for their ampliﬁcation, ensuring the upstream sense primer and the downstream antisense primer have suitable restriction sites for the ﬁnal cloning step into the multiple cloning site of the desired expression vector. Oligonucleotide primer design considerations should follow those described for routine PCR. Each fragment should be ampliﬁed separately using a blunt polymerase (e.g. Pfu; others are available). 2 Prepare the ampliﬁcation mix. This should contain the following: (a) 5–10 ng plasmid (as for general PCR); mini-prep DNA is adequately pure and concentrated for the reaction to be successful. (b) 1–10 pm each primer (in excess). (c) 1–10 U Pfu polymerase (according to manufacturers instructions) with the associated buffer. (d) Excess dNTPs (typically 1 µl at 2.5 mm starting concentration). A ﬁnal volume of 50 µl is advised, overlaid with an equivalent volume of mineral oil to prevent evaporation/condensation. The melting temperature has to be optimized; however, a general estimate of 60 ◦ C is suitable for most primers. Following an initial denaturation step of 95 ◦ C for 1 min, cycle as follows, for 30 cycles:

95 ◦ C 55–60 ◦ C 75 ◦ C

1 min 1 min 1 min

Typically, two reactions for each desired fragment is sensible. 3 Run all on a 1.2% agarose gel (seaplaque agarose is more pure and has a lower melting temperature; this is desirable for gel elution). Remove the gel fragment using a clean, sterile scalpel, retaining as little gel as possible. 4 Purify the sample using a commercially available resin-retention kit (Promega, UK, Qiagen, UK, and others). This routinely results in clean, eluted samples of between 30

279

280

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

and 50 µl of DNA at a concentration of 20–200 ng µl−1 ; however, it is necessary to run an aliquot (1 µl is usually sufﬁcient depending on PCR ampliﬁcation) on a 1.2% agarose gel to conﬁrm band presence, as gel-clean kits can occasionally fail, leading to a loss of DNA. 5 Ligate the two clean bands. Mix approximately 100 ng each fragment, 1–10 U T4 DNA ligase (according to manufacturer’s instructions) and add the appropriate ligase buffer as directed. Use a typical ﬁnal volume of 10–20 µl. Reports differ as to the optimum length and temperature of ligation. Experience suggests 16 h at 14 ◦ C is particularly efﬁcient, although 1 h at 37 ◦ C is known to be sufﬁcient. 6 Carry out the second PCR step to create the chimeric fragment. Using approximately 1–2 µl of the ligation product as a template, repeat the PCR as above using the N-terminal sense primer and the C-terminal antisense primer of the desired chimera. Include individual primer controls. Run, excise and clean the PCR product as described above. The product can then be ligated into a blunt-ended cloning vector such as the pCR-Script blunt-cloning vector (Stratagene) with blue–white selection. 7 Perform the blunt cloning of the fragment into a plasmid vector. Pre-digest the vector with a restriction site producing blunt ends (any single EcoRVI-containing plasmid can be used, but in this case there is an Srf I site located in the middle of a β-lactamase gene) and then ligate into the plasmid. Digestion and ligation are according to manufacturer’s instructions, but brieﬂy: approximately 100 ng plasmid is digested in appropriate buffer with 5–10 U appropriate restriction enzymes for 1 h at 37 ◦ C followed by a 20 min 75 ◦ C enzyme denaturation step. Ligation can be titrated for efﬁciency with a starting point of 1 : 10 plasmid : fragment ratio, at concentrations of approximately 10 : 100 ng respectively. The ligation conditions are as described above (step 5). 8 Transform any normal Escherichia coli competent cells (unlike the site-directed point mutagenesis described below, no special considerations are necessary). This is followed by overnight 37 ◦ C incubation on l-agar plates with the required antibiotic selection agent (ampicillin in the case of pCR-Script). To identify successfully ligated transformants,a spread 100 µl of 100 mm IPTG (for β-lactamase promoter) and 100 µl 2% X-gal (β-lactamase substrate) onto the agar, allowing the components to dry individually to prevent aggregation. White colonies are selected (include one blue colony as an unligated plasmid control). 9 Plasmids obtained from overnight cultures are sequenced using appropriate sequencing primers (refer to http://www.stratagene.com/manuals/211190.pdf for a speciﬁc pCR-Script protocol (Stratagene, UK)). Subsequent subcloning into an expression vector is routine. Notes of the fragment results in the disruption of the β-lactamase gene. Plating transformants onto X-gal containing agar will result in blue and white colonies for those without inserts and with a disruptive insert respectively.

a Ligation

281

A.2 WHY MUTAGENESIS?

STEP 1

Donor gene Target gene

PCR

STEP 2

PCR

PCR

STEP 3

PCR

Sub-clone into expression vector.

Figure A.1 A three-step PCR process for chimeric-gene formation [15].

A.2.3 Point mutations From the early experiments of the British Canadian chemist Michael Smith in the 1970s (for which he was awarded a Nobel Prize), site-directed point mutations have arguably become the single biggest tool in the identiﬁcation of biologically important loci of the GPCR superfamily. Their use, combined with bioinformatics techniques and the large number of protein sequences in the GPCR family have, to some extent, circumvented the problems encountered by the lack of further structural data following on from the publication of the rhodopsin crystal structure in 2000 [17, 18]. The choice for residue substitution has been predominantly alanine replacement, either individually or as part of a less-targeted ‘alanine scan’. This is due to alanine residues having a small inert side chain (a single methyl group) without introducing

282

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

the ﬂexible propensity observed with the smaller glycine residue. Alanine scans have been used in abundance for biological GPCR analysis by ourselves and numerous other groups studying each of the main subfamilies [19–21]. Earlier studies have been comprehensively reviewed [17]. Cysteine-substitution is a second highly used technique with the advantage of using the natural sulfhydryl reactivity and disulﬁde cross-linking for inter- and intra-molecular analysis. This has been used to observe ligand-binding interactions and activation mechanisms in rhodopsin [22] and other receptors [23]. A similar approach creating intramolecular constraints through zinc binding to histidine substitutions throughout GPCRs from Family A and Family B showed a very clear suggestion of the constraints applied to these receptors upon ligand-induced activation, and was hugely inﬂuential in our understanding of the ‘ﬂowering’ effect seen during this conformational change in the GPCR superfamily [24, 25]. There are, of course, no right answers, and the choice of substitution can always be criticized. Alanine scans are usually preferred to glycine as an approximate ‘null

Figure A.2 The Stratagene QuikChange mutagenesis method (based on http://www.stratagene. com/manuals/200518.pdf).

A.2 WHY MUTAGENESIS?

283

substitution’, although it has been suggested (not without justiﬁcation) that many proteins can tolerate a single alanine substitution, even in functionally important domains. Conversely, the introduction of cysteine or histidine residues clearly create large reactive side chains which could form whole new intramolecular interactions with no clear model to describe them, often resulting in problems with expression, rather than receptor physiology. Unlike chimeric cloning, over the last 10–15 years, a version of the Stratagene QuikChange mutagenesis method (Figure A.2) has become, almost exclusively, the site-directed mutagenesis method of choice. For GPCR research, an analysis of publications including mutagenic technology (targeted) reveals that almost all of the papers used this protocol compared with previous methods. The reason for this is a good one. The method is cheap, quick and reliable and requires no speciﬁc plasmid vector, multiple cloning site or multiple ligation/transformation reactions. Plasmid size is limiting, but there is a relatively broad range (predicted to be approximately 2–10 kb). Primer design is a key consideration, but standard, commercially available oligonucleotides provide ample material of sufﬁcient quality for most mutagenesis reactions. An example method is given in Protocol A.2.

PROTOCOL A.2 Site-directed Mutagenesis Method Equipment and Reagents • Standard PCR thermo cycler (many available) • Pfu DNA polymerase (Promega) and 10 × buffer provided with the Pfu polymerase • Oligonucleotide primers (Invitrogen) • 10 mm dNTP mixture(Sigma–Aldrich): a mixture of all four dNTPs (dATP, dCTP, dGTP, dTTP) • DNA template plasmid at 100 ng µl−1 containing the gene of interest • 5 ml sterile PCR tubes (Sarstedt) • 8–1.2% agarose gel • Agarose gel tank (Bio-Rad Laboratories) • 240 V power pack (Bio-Rad Laboratories) • Desktop microcentrifuge (many available) • DpnI restriction enzyme (New England Biolabs) • Competent cells; for example, XL10-GOLD cells (Stratagene, UK) • Microbiological incubator (many available) • LB medium (Oxoid, UK) • l-Agar (Oxoid, UK)

284

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

• Ampicillin (sodium salt; Sigma, UK) • Plasmid isolation kit (Wiz-prep; Promega, UK).

Method 1 Design appropriate oligonucleotide primers; this is arguably the most important design aspect to site-directed mutagenesis. The key rules include: (a) Incorporating 9–15 nucleotides either side of the mutation. (b) Using codon degeneracy to reduce the number of changed bases; codon usage is not thought to be a major issue. (c) Checking for obvious secondary structure and potential mispriming with a BLAST search. Practically, when many primers are being designed, it is common to perform these checks only if a problem with individual mutagenesis using certain primers is encountered. (d) Trying to avoid multiple runs of individual bases; this has been known to cause mispriming but is often an unavoidable issue. (e) Trying to terminate the primer at the 3 end with a so-called GC clamp;a that is, end the primer at the 3 end with two Gs, Cs or a combination. 2 The ampliﬁcation mix contains the following: (a) 100 ng plasmid (this is not PCR therefore a signiﬁcant starting mass is required); mini-prep DNA is adequately pure and concentrated for the reaction to be successful. (b) 10 pm each primer. (c) 5–10 U Pfu polymerase with the associated buffer. (d) Excess dNTPs (typically 1 µl at 2.5 mm ﬁnal concentration). A ﬁnal volume of 50 µl is advised. Always include a control with one or both oligonucleotides omitted. 3 The melting temperature has to be optimized, although a general estimate of 60 ◦ C is suitable for most primers. Following an initial denaturation step of 95 ◦ C for 1 min, cycle as follows, for 12 cycles:

95 ◦ C 60 ◦ C 75 ◦ C

1 min 1 min 2 min per kb (14 min for a typical GPCR in a pcDNA3 expression plasmid)b

To digest the template fragment from the ampliﬁcation mix, add 2–10 U DpnI, pipette up and down and spin brieﬂy in a microcentrifuge for 1 min. Incubate at 37 ◦ C for 1 h.

A.4 CONCLUSION

285

4 Transform any commercially available competent cells,c using typically 1–5 µl ampliﬁcation mix; include a DpnI-treated plasmid transformation as a control for DpnI digestion). 5 Following transformation, it is advisable to incubate in 0.5–1 ml LB medium (with no selective agent) at 37 ◦ C for 1–1.5 h to promote initial antibiotic resistance in free media. This is followed by overnight 37 ◦ C incubation on l-agar plates with the required antibiotic selection agent (ampicillin in the case of pCR-Script). 6 To identify successful transformants, plasmids obtained from overnight cultures should be sequenced using appropriate sequencing primers as for the chimera production (Protocol A.1, step 9). If restriction sites have been disrupted by the mutation change, then a digestion analysis of transformants can be a useful initial screen to avoid unnecessary sequencing costs. Notes a

The A–T interaction has two hydrogen bonds, whereas the G–C interaction has three. The movement of polymerase from the 3 end of the annealed primer requires a ﬁrm double-stranded grip. No such interaction is required at the 5 end. As such, ending with a GC clamp can enhance the ampliﬁcation step signiﬁcantly. check that the ampliﬁcation has been successful it is useful to remove 5 µl and compare with the pre-DpnI control on a 1.2% agarose gel. This is not essential and occasionally the product is at too low a concentration for electrophoretic analysis. b To

c

Competence can be a limiting factor with transformations, and the higher the cell competence, the more chance of success there is.

A.3 Troubleshooting The conditions for transformation of the DpnI-digested ampliﬁcation reaction (in concert with the oligonucleotide-free control) is a key consideration for successful routine mutagenesis. It has been noted by several researchers (including the ﬁrst author of this appendix) that sedimentation and precipitation of several components of the ampliﬁcation reaction is a strong negative inﬂuence. Following thawing, concerted vortexing of all components is advisable. Almost all E. coli strains used to produce the initial template are (dam) methylase positive, perfect for DpnI-mediated mutagenesis. On rare occasions, strains do not have this capability and the plasmids produced are not suitable templates.

A.4 Conclusion Mutagenesis is a sophisticated, crucial tool for the structure–function analysis of GPCRs which has provided more information on the pharmacology and signalling of this superfamily than any other technique. It is important to be aware that the information gleaned from these types of study can only be as good as the experimental

286

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

design. The choice of mutation, from chimera and deletions to individual alanine substitutions, needs a thorough and preferably model-driven process. No technique can be considered best, and the subsequent assaying technique is a major problem due to the promiscuous nature of this receptor family. Furthermore, it is important to consider that, as the primary amino acid structure folds so many times, the mining technique or even multiple residue-substitution can result in mutations far apart in the fully folded protein structure. Following the huge interest in GPCR mutagenesis, it is critical that the data retrieved is not wasted. A recent review discussing roles for information collection, management and integration in structure–function studies of GPCRs insightfully covers the requirement for data-mining and meta-analyses in the ﬁeld [26]. Also, whilst individual studies have revealed many interesting and useful facets of receptor biology, the future of this research lies in combining these techniques with structural analysis (through crystals and NMR), iterative modelling and detailed spectroscopy (including ﬂuorescence resonance energy transfer and bioluminescence resonance energy transfer). It is important that GPCR mutagenesis is not a stopgap whilst waiting for the breakthrough in three-dimensional GPCR analysis; rather, it is taken forward to assist the understanding of receptor activation, antagonist binding and the speciﬁc impact of the local environment on GPCR biology. Genomics suites have increased the turnaround time and, in our experience, the analysis has by far overtaken the molecular biology as the rate-limiting factor. This puts pressure on the high-throughput screening techniques to catch up.

References 1. Milligan, G. (2004) G protein-coupled receptor dimerization: function and ligand pharmacology. Mol. Pharmacol., 66, 1–7. 2. Sexton, P.M., Morﬁs, M., Tilakaratne, N. et al. (2006) Complexing receptor pharmacology: modulation of family B G protein-coupled receptor function by RAMPs. Ann. N. Y. Acad. Sci., 1070, 90–104. 3. Insel, P.A., Head, B.P., Ostrom, R.S. et al. (2005) Caveolae and lipid rafts: G protein-coupled receptor signaling microdomains in cardiac myocytes. Ann. N. Y. Acad. Sci., 1047, 166–172. 4. Barak, L.S., Wilbanks, A.M. and Caron, M.G. (2003) Constitutive desensitization: a new paradigm for g protein-coupled receptor regulation. Assay Drug Dev. Technol., 1, 339–346. 5. Beukers, M.W. and Ijzerman, A.P. (2005) Techniques: how to boost GPCR mutagenesis studies using yeast. Trends Pharmacol. Sci., 26, 533–539. 6. Li, B., Scarselli, M., Knudsen, C.D. et al. (2007) Rapid identiﬁcation of functionally critical amino acids in a G protein-coupled receptor Nat. Methods, 4, 169–174. 7. Scarselli, M., Li, B., Kim, S.K. and Wess, J. (2007) Multiple residues in the second extracellular loop are critical for M3 muscarinic acetylcholine receptor activation J. Biol. Chem., 282, 7385–7396.

REFERENCES

287

8. Hawtin, S.R., Wesley, V.J., Parslow, R.A. et al. (2000) Critical role of a subdomain of the N-terminus of the V1a vasopressin receptor for binding agonists but not antagonists; functional rescue by the oxytocin receptor N-terminus. Biochemistry, 39, 13524–13533. 9. Havlickova, M., Blahos, J., Brabet, I. et al. (2003) The second intracellular loop of metabotropic glutamate receptors recognizes C termini of G-protein α-subunits. J. Biol. Chem., 278, 35063– 35070. 10. Gupte, J., Cutler, G., Chen, J.L. and Tian, H. (2004) Elucidation of signaling properties of vasopressin receptor-related receptor 1 by using the chimeric receptor approach. Proc. Natl. Acad. Sci. U. S. A., 101, 1508–1513. 11. Postina, R., Kojro, E. and Fahrenholz, F. (1996) Separate agonist and peptide antagonist binding sites of the oxytocin receptor deﬁned by their transfer into the V2 vasopressin receptor. J. Biol. Chem., 271, 31593–31601. 12. Turner, P.R., Bambino, T. and Nissenson, R.A. (1996) A putative selectivity ﬁlter in the G-proteincoupled receptors for parathyroid hormone and secretion. J. Biol. Chem., 271, 9205–9208. 13. Yu, J., Polgar, P., Lubinsky, D. et al. (2005) Coulombic and hydrophobic interactions in the ﬁrst intracellular loop are vital for bradykinin B2 receptor ligand binding and consequent signal transduction. Biochemistry, 44, 5295–5306. GPCR chimera construction and analysis between close members of a GPCR subfamily. 14. Kim, J.M., Hwa, J., Garriga, P. et al. (2005) Light-driven activation of β2 -adrenergic receptor signaling by a chimeric rhodopsin containing the β2 -adrenergic receptor cytoplasmic loops. Biochemistry, 44, 2284–2292. GPCR chimera construction and analysis between more distant members of a GPCR subfamily. 15. Grandori, R., Struck, K., Giovanielli, K. and Carey, J. (1997) A three-step PCR protocol for construction of chimeric proteins. Protein Eng., 10, 1099–1100. A clear multistep process for unrestricted chimeric construct formation. 16. Hoffmann, C., Soltysiak, K., West, P.L. and Jacobson, K.A. (2004) Shift in purine/pyrimidine base recognition upon exchanging extracellular domains in P2Y 1/6 chimeric receptors. Biochem. Pharmacol., 68, 2075–2086. 17. Kristiansen, K. (2004) Molecular mechanisms of ligand binding, signaling, and regulation within the superfamily of G-protein-coupled receptors: molecular modeling and mutagenesis approaches to receptor structure and function. Pharmacol. Ther., 103, 21–80. A comprehensive review of GPCR mutagenesis before 2004. 18. Lu, Z.L., Saldanha, J.W. and Hulme, E.C. (2002) Seven-transmembrane receptors: crystals clarify. Trends Pharmacol. Sci., 23, 140–146. 19. Conner, A.C., Simms, J., Conner, M.T. et al. (2006) Diverse functional motifs within the three intracellular loops of the CGRP1 receptor. Biochemistry, 45, 12976–12985. 20. Conner, A.C., Simms, J., Howitt, S.G. et al. (2006) The second intracellular loop of the calcitonin gene-related peptide receptor provides molecular determinants for signal transduction and cell surface expression. J. Biol. Chem., 281, 1644–1651. Use of QuikChange mutagenesis for an alanine scan analysis of a GPCR. 21. Conner, M., Hawtin, S.R., Simms, J. et al. (2007) Systematic analysis of the entire second extracellular loop of the V1a vasopressin receptor: key residues, conserved throughout a G-protein-coupled receptor family, identiﬁed. J. Biol. Chem., 282, 17405–17412.

288

APPENDIX: SITE-DIRECTED MUTAGENESIS AND CHIMERAS

22. Hubbell, W.L., Altenbach, C., Hubbell, C.M. and Khorana, H.G. (2003) Rhodopsin structure, dynamics, and activation: a perspective from crystallography, site-directed spin labeling, sulfhydryl reactivity, and disulﬁde cross-linking. Adv. Protein Chem., 63, 243–290. 23. Han, S.J., Hamdan, F.F., Kim, S.K. et al. (2005) Pronounced conformational changes following agonist activation of the M3 muscarinic acetylcholine receptor. J. Biol. Chem., 280, 24870–24879. 24. Sheikh, S.P., Vilardarga, J.P., Baranski, T.J. et al. (1999) Similar structures and shared switch mechanisms of the β2 -adrenoceptor and the parathyroid hormone receptor. Zn(II) bridges between helices III and VI block activation. J. Biol. Chem., 274, 17033–17041. 25. Sheikh, S.P., Zvyaga, T.A., Lichtarge, O. et al. (1996) Rhodopsin activation blocked by metal-ion-binding sites linking transmembrane helices C and F. Nature, 383, 347–350. 26. Shi, L. and Javitch, J.A. (2006) A role for information collection, management, and integration in structure–function studies of G-protein coupled receptors. Curr. Pharm. Des., 12, 1771–1783.

Index

Numbers in italics refer to Protocols A2A adenosine receptors 86, 134–5, 137–8, 144 ab initio modelling 263, 270 acetoxymethyl (AM) esters 41 adenine nucleotides 89 adenosine 14, 86, 134–5, 137–8, 144, 187 adenosine triphosphate (ATP) 33–4 adenylyl cyclase (A) 32–4, 49, 65–6 adherent mammalian cells 8–9 Aequorea victoria 114 afﬁnity chromatography 102, 107 afﬁnity labelling 230 agonists 3–4, 54–5, 61–4, 197, 277 binding assays 5–6, 13–14, 17, 19 binding curve 54 BRET 111, 116, 120 disulﬁde cross-linking 148, 152–4 FCS 170, 183, 187 FRET 135–6, 143–4 full 49, 54, 66, 134–5, 188 ligand efﬁcacy 53–5, 60–1, 64–7 partial 50, 54, 64, 135 partial inverse 148 response curve 54 SCAM 229–31, 233, 240–1, 245–6 second messenger assays 49–50 trafﬁcking 77, 80, 216–17, 220–1 see also inverse agonists alanine 237, 281–3, 286 scans 281–2 alcohol oxidase (AOX 1) 87, 94–5

alignment of target sequence 252–4, 254–9, 259, 269–70 alkylation 49 allosteric ligands 4–5 allosteric modulator 17–19, 19–21, 25–6, 26–7 alpha2A (α2A ) adrenergic receptor 134–5, 137, 144 AlphaScreen 34–5, 35–8, 47, 47–9 amine 183 amino acids 71, 104, 207, 247, 276, 286 disulﬁde cross-linking 149, 151–2 FRET 134, 136–8 homology modelling 253, 263 ammonium chloride 79 ampliﬁcation 89–92, 106 analytical ultracentrifugation (AUC) 104–5 angiotensin 123 anion-exchange chromatography 38 antagonists 2–4, 17, 66, 153, 170, 286 SCAM 229–31, 233, 237–8, 240–1, 246 antibiotics 120–1 anti-receptor antibodies 197–200, 200–1, 202, 203–4 Arabidopsis thaliana 100 aripiprazole 54, 64–5 arrestins 133, 135 ascorbic acid 6 asparagines 237 aspartate 239 association rate 6–7, 22, 23–4 constant 1–2, 7, 22 atropine 10, 12

G Protein Coupled Receptors Edited by David R. Poyner and Mark Wheatley  2010 John Wiley & Sons, Ltd.

290 autocorrelation analysis 171 FCS 169–72, 178, 181–2, 187–8, 191 autocorrelation curves 176–8, 182–4, 189 autoﬂuorescence 116, 178 FCS 177–8, 182–4, 189–90 bacteria 85, 92, 139, 148, 163, 200 BRET 116, 119 bathorhodopsin 134 beta1 (β1 ) adrenergic receptor 135, 137, 144 beta2 (β2 ) adrenergic receptor 71, 112, 277 disulﬁde cross-linking 148, 153–4 FRET 134–5 homology modelling 251–2, 254 SCAM 229, 233–4, 236, 240–1, 246 beta3 (β3 ) adrenergic receptor 181 beta (β) arrestins 112, 116, 123 beta2 (β2 ) bradykinin receptor 143–4 biarsenical ﬂuorophores 138 biochemistry 114, 147, 207–9, 209–13, 276 bioinformatics 281 bioluminescence resonance energy transfer (BRET) 111–24, 124-7, 127–8 controls 123–4 FCS 169 mutagenesis 286 principles 114–16 speciﬁcity 122–3 sub-methods 118, 124 biophysics 54, 147–8, 275 biotinylated receptor 201–3, 203–4 bipartite alignment method 253 bradykinin 277 buffers 5–7, 9, 74 FCS 178, 184, 186 recombinant GPCR 92, 94, 102, 105, 107 buprenorphine 54 buspirone 54 calcitonin 111 calcium 6, 32–3, 40–1 Fluo-4 assay 41–4 trafﬁcking 71, 80 cAMP (cyclic adenosine monophosphate) 32–5, 35–8, 44, 54, 120 cAMP response element binding protein (CREB) 34 cannabinoid 64, 71 carazolol 148 catecholamine 6, 247

INDEX

cDNA 117, 119, 120–1, 128, 198 cell culture and transfection 154–6 cell density 49 cellular labelling 138–40 central nervous system (CNS) 69 centrifugation 7, 97 cerebella granule neurones 204 Cerulean 135, 143 chaperons 86, 111 chemiluminescence 103 Cheng–Prusoff equation 3, 13 chimeras 275, 277–8, 278–80, 281, 283, 286 Chinese hamster ovary (CHO) cells 177, 181, 187 phosphorylation 199, 202, 208 cholecystokinin receptor type A (CCKAR) 73 chromophore 103–4 chromophore–opsin complex 134 chymotrypsin 208 circadian clock proteins 113, 117 circular dichroism (CD) 86, 103–5 clonidine 135 cloning 86–92, 95, 97, 106, 139 clozapine 247 coelenterazine 114–15, 117–18, 124 co-immunoprecipitation 113 colorimetric assay 217 competition binding 3–4, 13–15, 15–16 competitive immunoassay 34 concentration–occupancy relationship 2–3 concentration–response curves 64 confocal microscopy 120, 173 FCS 169–70, 172–3, 178, 186 trafﬁcking 70, 73–5, 82, 216 conformational changes in SCAM 245–7 constitutively active mutant (CAM) 246 continuous-wave laser excitation 172–3 critical micelle concentration (CMC) 100 C-termini 73, 198–9, 245, 277 BRET 115, 117, 119 disulﬁde cross-linking 149–51, 163 FCS 177, 181 FRET 134, 136–7, 143 homology modelling 253–4, 264 Cu–Phen 150–2, 162–3 cyan ﬂuorescent protein (CFP) 136–7, 141 cyclic adenosine monophosphate 54, 120 cyclic adenosine monophosphate (cAMP) 32–5, 35–8, 44, 54, 120 Cys residues 148–9, 151–4, 157, 162–4, 233–4

INDEX

291

database modelling 263 deactivation 112 DeepBlueC 118 DeepView 254–9, 270 degradation 198, 207–8, 221 density (concentration) of receptors 1–3, 7 desensitization 112, 215, 275 detergents 100–1, 113–14 DFIRE 263–4, 264–5, 268 diacylglycerol (DAG) 32, 38, 41, 148 diffusion 169–91 coefﬁecients 170–1 digestion of factor Xa 158, 159–60, 162 dimers 164, 191 dissociation constant 1–2, 7, 22, 26, 54 binding kinetics 22, 24–7 distance-scaled, ﬁnite ideal-gas reference (DFIRE) 263–4, 264–6, 268 disulﬁde cross-linking 147–58, 158-9, 159–64, 282 DNA sequencing 234 dopamine 6, 64–6, 135 SCAM 229, 232–4, 237, 244–7 dose-response curves 67, 120 drugs and medication 31–2, 53–4, 65–6, 147, 190, 251

enhanced green ﬂuorescent protein (EGFP) 114–15, 119 enhanced yellow ﬂuorescent protein (EYFP) 135, 143 trafﬁcking 7–3, 75–81 environment 128, 275, 286 trafﬁcking 69–70, 72 enzyme-linked immunosorbent assay (ELISA) 44, 120–1 trafﬁcking 215–27 epinephrine 6 epitope tagging 49, 116, 197–9, 253 disulﬁde cross-linking 149–51 SCAM 238, 247 trafﬁcking 216–17, 220–1, 223, 226 equilibrium binding experiments 17–19 equilibrium dissociation constant 2, 5, 9, 11, 13, 17–18, 22 Escherichia coli 86, 92, 285 ethanedithiol (EDT) 140 ethanol 181 ethylenediaminetetra acetate (EDTA) 6, 9, 13 ethyleneglycoltetra acetate (EGTA) 6, 9, 13 excitation maxima 135, 177–8, 183 expression vectors 87–9 extended BRET (eBRET) 118, 128 external restraints on homology modelling 261–2 extracellular receptors 79–80, 81 extracellular signal-related kinases 1 and 2 (ERK 1/2) 32–3, 44 phosphorylated 44, 44–7, 47, 47–9

Edman degradation 198, 207–8 efﬁcacy 53–67 measurement 53–4 electrochemiluminescence 47 electron microscopy 103 electrophysiology assays 64–5 electroporation 92, 94 electrostatic potential 232–3, 245 Emerald 135 emission maxima 135, 177, 183 endocytic markers 223 endocytosis 71, 215 endogenous Cys 233–4 endosomes 223 EnduRen 118, 128 engineered cysteine 229–34, 238, 245

Family A 5, 253, 277, 282 Family B 253, 277, 282 Family C 6, 253 Fed-batch fermentation 97 FLAG 120, 198, 217, 221 Fluo-3 41 Fluo-4 41, 41–4 ﬂuorescein 223 ﬂuorescein-based arsenical hairpin binder (FlAsH) 134–40, 140–1, 177 cellular labelling 138–40, 140–1 ﬂuorescein isothiocyanate (FITC) 223 ﬂuorescence 47, 105, 177, 246, 259 BRET 116–17, 121–2 disulﬁde cross-linking 148, 163 trafﬁcking 69–76, 79–81, 216, 223 ﬂuorescence-activated cell sorting (FACS) 122

cysteine 6, 134, 138, 177, 253, 259, 282–3 engineered 229–34, 238, 245 SCAM 229–48 substitution 282 cytosol 32, 40, 120 FCS 178, 186, 188

292 ﬂuorescence correlation spectroscopy (FCS) 169–84, 184–5, 185–91 calibration of microscope 174–5, 175-6, 176–7 equipment 172–4 principles 170–2 upper cell membrane 179–80 ﬂuorescence polarization assays 34 ﬂuorescence recovery after photobleaching (FRAP) 169, 190 ﬂuorescence resonance energy transfer (FRET) 113–14, 169, 286 compared with BRET 116–17 ﬂuorescence lifetime imaging microscopy (FLIM) 116 intramolecular 133–44 ligand mediated changes 141–2, 142–3, 143 trafﬁcking 73, 75–6 ﬂuorochromes 223 ﬂuorophores biarsenical 138 BRET 113, 115, 117–20, 124 FCS 170, 172, 174, 176–7, 182–3, 186, 190 FRET 134–8 good insertion sites 136–7 trafﬁcking 70, 75, 227 formaldehyde 87 forskolin 33–4, 49 F¨orster distance 116 Fourier transform infrared spectroscopy (FTIR) 105 FRET ﬂuorescence lifetime imaging microscopy (FRET-FLIM) 116 full agonists 49, 54, 66, 134–5, 188 functional analysis 85–107 Fura-2 41 fusion proteins 120–1 Gα subunit 32–3 Gβγ subunits 32–3 G protein-coupled receptor kinase (GPK) 197 glutamate 234, 237, 253, 277 trafﬁcking 70, 71, 77, 80–1 glutathione-S -transferase (GST) 89, 113, 200, 202 glycine 237, 282 glycosylation 86, 151 Golgi export 217 gondotrophin 123

INDEX

green ﬂuorescent protein (GFP) 89, 119 FCS 170, 179–80, 181, 188 FRET 134–6 trafﬁcking 71–3, 75–81, 216 guanine nucleotides 3–4, 14, 88 guanosine 5 -O-(3-thiotriphosphate) ([35 S] GTPγ S) 3, 6, 54–5, 55–64, 64–7 guanosine diphosphate (GDP) 3, 32, 54–5 guanosine triphosphate (GTP) 3, 32, 55, 183 haemagglutinin (HA) 120, 149, 198–9, 216–17 haloperidol 65 harvesting cells 234, 235–6 HEPES-based buffers 5, 9 heterodimers 275 heterologous cells 234 high-performance liquid chromatography (HPLC) 174, 178, 183 Hill coefﬁcient 2–4, 14–15, 67 Hill–Langmuir binding isotherm 2 Hill–Langmuir occupancy equation 10 histidine 70, 97, 237, 282–3 homodimers 275 homology modelling 251–4, 254–9, 259–60, 260–2, 263–4, 264–8, 269–70 automated methods 269–70 horseradish peroxidise 34, 203 human embryonic kidney (HEK) cells 177, 217, 235, 246 trafﬁcking 73, 75–6, 79–81 human genome 113, 251 hydrogen peroxide 87, 163 image correlation spectroscopy 190 imidazole 102 immobilized metal afﬁnity chromatography (IMAC) 38, 102, 107 immunoblotting 97, 103 immunocytochemistry 198 immunoﬂuorescence 217, 222–3, 226 immunoprecipitation 197–204, 208 incubation time 6–7 infrared ﬂuorescence 47 initiation codon 88–9 inositol phosphates (IP) 32–3, 38, 120 accumulation assay 39–40 inositol triphosphates (IP3 ) 33, 38, 39–40, 41, 148 insect cells 85–6 insulin 80

293

INDEX

interferogram 105 internalization 112, 117, 197 trafﬁcking 215–18, 220–1, 223, 224–6 intracellular signalling 32 inverse agonists 5, 49, 111, 135, 153–4 FCS 183, 188 ligand efﬁcacy 53, 64–5 ionized thiolate 230–1 ionotropic receptors 73 isoleucine 237 isomerization 133, 246 isoproterenol 134, 233, 246 isotopic dilution 25–6 isotopic labelling 106 jellyﬁsh 114 kinases 112, 133 kinetic radioligand binding assays 22, 22–5, 25–6, 26–7 knots in homology structure 269 Kolmogorov–Smirnov test 79 lactose permease 234 lambda scanning 73, 75 laser scanning confocal microscopy 70 law of mass action 1–3 leucine 237 leucine-rich repeats (LRRs) 253 ligands and ligand binding 86, 97, 253, 282 allosteric 4–5 BRET 111, 117, 123 depletion 7, 10–11 design of ﬂuorescent 183–4 efﬁcacy 53–67 FCS 169–70, 172, 182–4, 184–5, 185–8 FRET 133, 135, 141–2, 142–3, 143 inducing structural changes 147–64 orthosteric 4–5, 17–18, 19–21, 25 protection 240–1, 241–4 SCAM 231–4, 236–7, 237–8, 238–40, 240–1, 241–4, 244–7 trafﬁcking 73, 216, 221 ligand–receptor complex (AR) 1–3, 5, 10 ligand–receptor interactions 1–27, 183 light microscopy 70 linear dichroism (LD) 104 lipids 31, 86 live cell antibody labelling 77–8 long-term depression (LTD) 71–2

long-term potentiation (LTP) 71–2 loop predictions 263–4 loop regions 263–4, 264–8 luciferases 113, 116 luminescence 114, 122 luminophores 113, 117, 119–120 Lumio Green 134–5, 138 Lumio Red 134–5 lumirhodopsin 134 lysines 134 lysosomes 215, 223 lysotracker red 223 M1 acetylcholine receptor 143–4 magnesium 6 maltose binding protein (MBP) 89 mammalian cells 85–6, 197 adherent 8–9 BRET 116, 120–1, 124–7 mass spectrometry 207 melanocortin receptors (MC2R) 111, 119 membrane preparation 8, 8–9, 55, 154, 156–7 urea treatment 157 membrane-spanning segments 229–34, 237, 244–7 memory 69 metarhodopsin 134 methane thiosulfate (MTS) 232–3 SCAM 230–4, 238, 238–40, 240–1, 243, 243–4, 248 methane thiosulfate ethylammonium (MTSEA) 230, 233–4, 238, 241, 244–6 methane thiosulfate ethylsulfonate (MTSES) 230, 233, 238, 245 methane thiosulfate ethyltrimethylammonium (MTSET) 230, 233–4, 238, 245 methanol 87, 95 methionine 237 microscopy 70, 103, 116, 223, 224–6 see also confocal microscopy mitogen-activated protein (MAP) kinase 32–3, 44, 65–6, 112 phosphorylation 197 MODELLER 254–9, 259–60, 260–2, 262, 269 molecular iodine 150–2, 163 monensin 75 monoclonal antibody 199 monomeric receptors 191 moxonidine 135 muscarinic acetylcholine 64

294 muscarinic receptor (mAChR) 151–4, 276 disulﬁde cross-linking 148–54, 164 phosphorylation 197–200, 202, 204, 208 mutagenesis 138, 275–8, 278–80, 280–3, 283–5, 285–6 disulﬁde cross-linking 148, 152–4 point mutations 281–3 random (saturation) 276 SCAM 230, 232, 234, 236–8, 238–40, 240–1, 247–8 site directed 275, 283–5 targeted 276–8 myc 120, 217 nigericin 75 nitrocellulose 103 nonnative conformations 162 nonnative loops 263 nonspeciﬁc binding 9–11 noradrenaline 134–5 norepinephrine 6, 135 normalized binding 17–18 norphenylephrine 135 N-termini 136, 177, 277 BRET 117, 119 disulﬁde cross-linking 149–51, 154 homology modelling 253–4, 263 large domains 253 phosphorylation 198–200 SCAM 245, 247 trafﬁcking 71, 73, 76, 216–17 nuclear magnetic resonance (NMR) spectroscopy 105–6, 259 mutagenesis 275, 286 octopamine 135 opiates 54, 64 [32 P] orthophosphate 198–9, 204, 204–7, 207–9, 209–13 orthosteric ligands 4–5, 17–18, 19–21, 25 overexpression 49, 121–3, 247 oxymetazoline 135 palmitoylation 112 parathyroid hormone 134–5, 143, 277 partial agonists 50, 54, 64, 135 partial inverse agonists 148 peptide mimetics 277 pH 5, 71–2 trafﬁcking 69–81, 221, 226 pharmacophore 183

INDEX

phenol red 177–8, 190 phenylalanine 237 pHluorins 69–72 phosphatise 216–17 phosphoamino acid (PAA) 198, 207–8 phosphopeptides 198, 207–8 phosphorimager 208 phosphorylated ERK 1/2 44, 44–7, 47, 47–9 phosphorylation 197–200, 200–1, 201–4, 204–7, 208, 209–13, 213 BRET 112 photobleaching 117, 141 FCS 177, 181–2, 188, 190 ﬂuorescence recovery after (FRAP) 169, 190 trafﬁcking 70, 74, 76, 82 photometry 141–2 photon counting 169–70, 172, 174, 177 histogram analysis 191 photophysics 188 photorhodopsin 134 phototoxicity 183, 186 Pichia pastoris 86–8, 93–4, 94–5, 95–7, 98 PIP2 32, 38, 41 plasmids 86–7, 90–1, 92, 95, 199 mutagenesis 278, 282–3 PLC 32, 38 point mutations 281–3 polyacrylamide gel electrophoresis (PAGE) 198 SDS 103, 106, 198, 204, 208 polymerade chain reaction (PCR) 87, 138–9 mutagenesis 276–8, 281 polyvinylidene ﬂuoride 103 potassium 6 pre-bleach 178, 182, 186, 190 preparation of membranes 97–8, 98–100 proline 237 protease digest 208 protein expression measurement 121–2, 128 protein kinase A (PKA) 32, 34, 44, 197 protein kinase C (PKC) 32, 38, 40, 44, 197 proteosomes 215, 223 protonated amine 245 puriﬁcation 86, 100–1, 101–2, 102–3, 107 disulﬁde cross-linking 147–8 phosphorylation 197 Python formatting 259, 269 quantitative imaging 69–82 query sequence 251–2

295

INDEX

radioimmunoassay (RIA) 34–5 radioimmunoprecipitation assay (RIPA) 199 radiolabelling 2–3, 198, 204, 207 radioligands 1–3, 5–8, 49, 153 competition binding 13–14, 15–16 equilibrium binding experiments 17 FCS 188 measurement of kinetics 22–5, 26–7 normalized 17–18 orthosteric 4–5, 17–18, 19–21, 25 phosphorylation 202 saturation 9–11, 11–13 SCAM 241, 243–4 separating bound from free 7 trafﬁcking 70, 215–16 RAPPER 263, 264–6, 268 ratiometric analysis 76, 76–7, 80–1, 121 receptor speciﬁc antibodies 197–201 recombinant protein analysis 86–7 recycling 217–18, 220–1, 221–2 regions of interest (ROIs) 74–5, 78, 81 Renilla luciferase (Rluc) 114–15, 117–19, 122, 124 Renilla reniformis 114 resensitization 215 resonance energy transfer (RET) 113–16, 119 resoruﬁn-based arsenical hairpin binder (ReAsH) 134–5 restraint-based modelling 263 restriction mapping 234, 248 rhodamine 223 rhodopsin 85–6, 103, 105, 147–8, 153–4 FRET 133, 136, 138 homology modelling 251–2, 254–9, 260 mutagenesis 277, 281–2 SCAM 229, 244–6 Saccharomyces cerevisiae 86–9, 97, 106 transformation 92, 93 saturation binding assays 9–11, 11–13, 13, 122 scaffolding proteins 177, 182 scale-up 86, 97–8, 106 scanning FCS 190 Schild plot 17 scintillation proximity assay (SPA) 38 screening 86, 95–7 sea pansy 114 second extracellular loop 245 second messengers 31–50, 54, 120, 148, 197 secondary structure 244–5

secretin 277 sedimentation equilibration 105 sedimentation velocity techniques 105 sequestration 215 serine 6, 112, 208, 234, 237 serotonin 86, 234 sigmoidal dose–response curve 67 signal peptides 89 signal-regulated kinase 65–6 signal sequence addition 89, 90 signal-to-noise ratio (S:N ) 49, 116–17 FCS 174–5, 178, 182, 190 signal transduction pathways 31, 33, 49 signalling pathways 112, 197 single photon counting 169–70, 172, 174, 177 single transmembrane proteins 111 sodium 5–6, 13 sodium dodecyl sulphate (SDS) 198 PAGE 103, 106, 198, 204, 208 sodium phosphate 102 solubilization 86, 100–1, 101–2, 106–7, 147, 159–60 phosphorylation 202, 204 speciﬁc binding 9–10 spectroscopy 86, 133, 286 spot bleaching 181–2, 186, 190 stereochemistry 259, 270 steric block 232–3 sterols 86 stoichiometry 104, 122 stop codon 119, 277 Stratagene QuikChange 282–3 streptavidin 201 subcellular fractionation 70 substituted cysteine accessibility method (SCAM) 229–48 substitution matrices 253 substrates 114–15, 117–18, 124 sulfhydryl reagents 229–32, 233, 241, 245, 282 SureFire 47, 47–9 surface-expressed receptors 71–2, 76–7, 77–8, 78–80, 217–18 surface receptor loss 218–20 synaptic transmission 69, 71, 80 synaptobrevin 80 synaptotagmin 80 tag addition 89, 90 tagging proteins 117, 119–24

296 target sequence 252–4, 259, 269–70 aligning 254–9, 269–70 TC-FlAsH 134, 138 TC-ReAsH 134 temperature 5–7, 91, 128 FCS 178, 187 trafﬁcking 220–1, 223 templates for homology modelling 252, 263, 269–70 alignment 252–3, 254–9, 269–70 selection 252–3, 259 ternary complex model (TCM) 4–5 thiol 183 thioredoxin (TRX) 89 third intracellular loop 200, 202 disulﬁde cross-linking 149, 151, 163 FRET 134, 136–8, 143–4 threonine 112, 208, 237 thrombin 70 thyroid stimulating hormone receptor (TSHR) 253 time-resolved FRET (TR-FRET) 116 TINKER 264, 264–7 titration assays 122 total binding (TB) 9–11 trace metals 5–6 trafﬁcking 215–18, 218–20, 220–1, 221–2, 222–4, 224–6, 226–7 BRET 117, 128 phosphorylation 197, 199 quantitative imaging 69–82 transfection 49, 154–6, 199, 204, 235 BRET 120–2, 128 disulﬁde cross-linking 154–6, 162, 164 FCS 178, 181, 186 FRET 140 trafﬁcking 217, 220, 226–7 transferring 223 transformation 92, 92–4, 94–5, 97 transient transfection 49 transmembrane (TM) domains 252–4, 263 transmitter release 71

INDEX

tryptophan 237 tyrosine 208, 237 ubiquitin 112, 122 ultra-violet 41, 183 urea treatment 157 vacuum ﬁltration 7 valine 237 vasopressin 112, 199, 254–9, 260–2, 277 veriﬁcation 252, 259 visible-light excitation 41 western blotting 44, 44–7, 120, 277 disulﬁde cross-linking 150–2, 158, 160–2 phosphorylation 198, 201 whole-cell patch clamp methods 77 wild-type proteins 4, 120 SCAM 231–3, 236, 246 trafﬁcking 72–3 xenon 104 X-ray crystallography 105 yeast 85–107 choice 86–7 preparation of membranes 98–100 transformation 92–4 yeast-two-hybrid assays 113 yellow ﬂuorescent protein (YFP) 71–3, 115, 188 FRET 136–7, 141–2, 144 YFpH-GluR2 73, 77, 77–8, 78–9, 81 YFpH fusion proteins 72–3, 73–4, 74–7, 77–8 spectral characterization 73–4 yohimbine 135 zeocin 94–5 zinc 282 zwitterionic molecules 100

G protein-coupled receptors in drug discovery

Read more

G Protein-Coupled Receptors in Drug Discovery

Read more

Ligand Design for G Protein-coupled Receptors

Read more

G Protein-Coupled Receptors in Drug Discovery

Read more

Methods in Enzymology: Nuclear Receptors

Read more

G Proteins (Methods in Neurosciences)

Read more

Essential Demographic Methods

Read more

Essential numerical computer methods

Read more

The Chemokine Receptors (The Receptors)

Read more

Toll-Like Receptors Methods and Protocols

Read more

Immune Receptors (Methods in Molecular Biology, v748)

Read more

The Cannabinoid Receptors (The Receptors)

Read more

The Dopamine Receptors (The Receptors)

Read more

Toll-Like Receptors: Methods and Protocols

Read more

Toll-Like Receptors Methods and Protocols

Read more

Thyroid Hormone Receptors. Methods and Protocols

Read more

Nuclear Receptors, Volume 364 (Methods in Enzymology)

Read more

G Protein-Coupled Receptors: Structure, Signaling, and Physiology

Read more

The Cannabinoid Receptors (The Receptors)

Read more

The Chemokine Receptors (The Receptors)

Read more

The Melanocortin Receptors (The Receptors)

Read more

G Protein-Coupled Receptors: Structure, Function, and Ligand Screening (Methods in Signal Transduction)

Read more

Rhodopsin and G-Protein Linked Receptors, Part A,

Read more

Methods for the Discovery and Characterization of G Protein-Coupled Receptors (Neuromethods 60)

Read more

Ligand Design for G Protein-coupled Receptors (Methods and Principles in Medicinal Chemistry)

Read more

The G Protein-Coupled Receptors Handbook (Contemporary Clinical Neuroscience)

Read more

G

Read more

PAF Receptors

Read more

Essential mathematical methods for physicists

Read more

Animal Cell Culture: Essential Methods

Read more

Recommend Documents

G protein-coupled receptors in drug discovery

METHODS IN M O L E C U L A R B I O L O G Y TM Series Editor John M. Walker School of Life Sciences University of Her...

G Protein-Coupled Receptors in Drug Discovery

G ProteinCoupled Receptors in Drug Discovery Drug Discovery Series Series Editor Andrew Carmen Johnson & Johnson PRD...

Ligand Design for G Protein-coupled Receptors

Nuclear Receptors as Drug Targets Edited by Eckhard Ottow and Hilmar Weinmann Methods and Principles in Medicinal Ch...

G Protein-Coupled Receptors in Drug Discovery

METHODS IN M O L E C U L A R B I O L O G Y TM Series Editor John M. Walker School of Life Sciences University of Her...

Methods in Enzymology: Nuclear Receptors

Preface In 1975 Bert O’Malley and Joel Hardman edited ﬁve volumes in the Methods in Enzymology series on the topic of h...

G Proteins (Methods in Neurosciences)

G Proteins G Proteins This Page Intentionally Left Blank G Proteins Edited by Patrick C. Roche Department of Exp...

Essential Demographic Methods

Essential Demographic Methods Kenneth W. Wachter Department of Demography University of California 2232 Piedmont Avenue...

Essential numerical computer methods

Essential Numerical Computer Methods Reliable Lab Solutions This page intentionally left blank Essential Numerical ...

The Chemokine Receptors (The Receptors)

The Chemokine Receptors TH E RE C E P T O R S Kim A. Neve, Series Editor The Chemokine Receptors, edited by Jeffr...

Toll-Like Receptors Methods and Protocols

METHODS IN MOLECULAR BIOLOGY™ Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfi...