MHC Volume 1: A Practical Approach (Practical Approach Series) (Vol 1)

MHC Volume 1

The Practical Approach Series SERIES EDITORS

D. RICKWOOD Department of Biology, University of Essex, Wivenhoe Park, Colchester, Essex CO43SQ, UK B. D. HAMES Department of Biochemistry and Molecular Biology University of Leeds, Leeds LS2 9JT, UK

* indicates new and forthcoming titles

*

* if

if

Affinity Chromatography Affinity Separations Anaerobic Microbiology Animal Cell Culture (2nd edition) Animal Virus Pathogenesis Antibodies I and II Antibody Engineering Applied Microbial Physiology Basic Cell Culture Behavioural Neuroscience Biochemical Toxicology Bioenergetics Biological Data Analysis Biological Membranes Biomechanics—Materials Biomechanics—Structures and Systems Biosensors Calcium-Pi signalling Carbohydrate Analysis (2nd edition) Cell-Cell Interactions

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

* DNA Cloning 4: Mammalian Systems (2nd edition) Electron Microscopy in Biology Electron Microscopy in Molecular Biology Electrophysiology Enzyme Assays * Epithelial Cell Culture Essential Developmental Biology Essential Molecular Biology I and II Experimental Neuroanatomy if Extracellular Matrix Flow Cytometry (2nd edition) * Free Radicals Gas Chromatography Gel Electrophoresis of Nucleic Acids (2nd edition) Gel Electrophoresis of Proteins (2nd edition) Gene Probes 1 and 2 Gene Targeting Gene Transcription Glycobiology Growth Factors Haemopoiesis Histocompatibility Testing HIV Volumes 1 and 2 Human Cytogenetics I and II (2nd edition) Human Genetic Disease Analysis * Immunochemistry 1 * Immunochemistry 2 Immunocytochemistry In Situ Hybridization

*

* *

if if

lodinated Density Gradient Media Ion Channels Lipid Analysis Lipid Modification of Proteins Lipoprotein Analysis Liposomes Mammalian Cell Biotechnology Medical Bacteriology Medical Mycology Medical Parasitology Medical Virology MHC Volume 1 MHC Volume 2 Microcomputers in Biology Molecular Genetic Analysis of Populations Molecular Genetics of Yeast Molecular Imaging in Neuroscience Molecular Neurobiology Molecular Plant Pathology I and II Molecular Virology Monitoring Neuronal Activity Mutagenicity Testing Neural Cell Culture Neural Transplantation Neurochemistry (2nd edition) Neuronal Cell Lines NMR of Biological Macromolecules Non-isotopic Methods in Molecular Biology Nucleic Acid Hybridization Nucleic Acid and Protein Sequence Analysis

Oligonucleotides and Analogues Oligonucleotide Synthesis PCR 1 PCR2 Peptide Antigens Photosynthesis: Energy Transduction Plant Cell Biology Plant Cell Culture (2nd edition Plant Molecular Biology Plasmids (2nd edition) * Platelets Pollination Ecology Postimplantation Mammalian Embryos Preparative Centrifugation Prostaglandins and Related Substances Protein Blotting Protein Engineering if Protein Function (2nd edition) Protein Phosphorylation

Protein Purification Applications Protein Purification Methods Protein Sequencing if Protein Structure (2nd edition) if Protein Structure Prediction Protein Targeting Proteolytic Enzymes Pulsed Field Gel Electrophoresis Radioisotopes in Biology Receptor Biochemistry Receptor-Ligand Interactions RNA Processing I and II if Subcellular Fractionation Signal Transduction Solid Phase Peptide Synthesis Transcription Factors Transcription and Translation Tumour Immunobiology Virology Yeast

MHC Volume 1 A Practical Approach Edited by N. FERNANDEZ Department of Biological and Chemical Sciences, University of Essex

and G. BUTCHER Babraham Institute, Babraham, Cambridge

IRL PRESS at

OXFORD UNIVERSITY PRESS Oxford New York Tokyo

Oxford University Press, Great Clarendon Street, Oxford OX2 6DP Oxford New York Athens Auckland Bangkok Bogota Bombay Buenos Aires Calcutta Cape Town Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madras Madrid Melbourne Mexico City Nairobi Paris Singapore Taipei Tokyo Toronto Warsaw and associated companies in Berlin Ibadan Oxford is a trade mark of Oxford University Press Published in the United States by Oxford University Press Inc., New York © Oxford University Press, 1997 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press. Within the UK, exceptions are allowed in respect of any fair dealing for the purpose of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms and in other countries should be sent to the Rights Department, Oxford University Press, at the address above. This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out, or otherwise circulated without the publisher's prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser. Users of books in the Practical Approach Series are advised that prudent laboratory safety procedures should be followed at all times. Oxford University Press makes no representation, express or implied, in respect of the accuracy of the material set forth in books in this series and cannot accept any legal responsibility or liability for any errors or omissions that may be made. A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data MHC1: a practical approach/edited by N. Fernandez and G. Butcher. (Practical approach series; 180) Includes bibliographical references. 1. Major histocompatibility complex—Laboratory manuals. I. Fernandez, Nelson, Dr. II. Butcher, G. (Geoffrey W.) III. Title: MHC Volume 1 IV. Series. [DNLM: 1. Major Histocompatibility Complex. 2. Antigen Presentation. 3. Antigen-Presenting Cells. 4. Proteins—chemistry. QW 568 M617 1997] QR184.315.M462 1997 616.07'9—dc21 97-2434 ISBN 0 19 963554 4 (Hbk) ISBN 0 19 963553 6 (Pbk) Two volume set ISBN 0 19 963558 7 (Hbk) ISBN 0 19 963557 9 (Pbk) Typeset by Footnote Graphics, Warminster, Wilts Printed in Great Britain by Information Press Ltd, Eynsham, Oxon.

Preface The study of the MHC began its career in the fields of tumour biology and transplantation; it achieved glory in the elucidation of some central mysteries in immunology, namely T lymphocyte recognition of antigen and the mechanisms of antigen presentation. Along the way it has played vital or exciting roles in both pragmatic and fundamental biology—from clinical organ transplantation to such diverse research areas as evolutionary genetics and studies of social behaviour. The value of MHC studies is yet to wane: currently new excitement surrounds topics such as the role of MHC class I molecules in regulating natural killer cell function, and the coordinated functioning of proteins involved in the class I and II presentation pathways, so many of which are found to be encoded within the MHC. Many of the central issues in MHC research are continuously 'revisited' by the application of more sensitive and novel techniques; at the same time, new issues require the best contemporary approaches. This is, in fact, the reason why we set out to edit this book. We present, in a single volume, the most important techniques for MHC research in biochemistry and cellular biology. In doing so, we hope to help researchers who intend to apply new techniques to their particular field, being the MHC or a related field. Unfortunately constraints of space have made it necessary for us to leave out some sub-topics; however we have tried to select the most important methods for the key investigative procedures. We thank all the contributors to this volume for their kind submissions and for writing their protocols in a clear and intelligible way. This has made our job as editors so much easier and enjoyable. Finally, we would like to thank the staff at OUP for their support and advice in the course of the editing procedures, and Dr David Rickwood for inspiring help in the initial stages of this project. Colchester Cambridge June 1997

N.F. G. W. B.

This page intentionally left blank

Contents Contributors

xv

Abbreviations

ix

1. Isolation and functional properties of antigen presenting cells

1

Brigitta Stockinger 1. Introduction

1

2. Isolation and functional properties of antigen presenting cells Dendritic cells Macrophages B cells

2 2 7 9

3. Functional assays of antigen presentation Presentation requirements for activation of CS-specific T cell hybrids and clones Readouts for T cell activation

11

References

15

2. Labelling methods for analysis of class II MHC endocytosis and peptide turnover

11 12

17

Pamela A. Reid and Colin Watts 1. Introduction

17

2. Surface labelling of MHC molecules Lactoperoxidase-catalysed labelling on tyrosine residues Labelling of amino groups by succinimidyl ester based reagents Methods utilized for study of the endocytosis of class I and II glycoproteins in B-lymphoblastoid cells Study of the endocytosis of class I and II MHC glycoproteins in B-lymphoblastoid cells using cleavable cell surface labelling reagents Endocytosis of the MHC glycoproteins in the presence of primaquine Measurement of the recycling kinetics of endocytosed class I and II MHC glycoproteins Study of the antigenic peptides associated with class II MHC glycoproteins

18 18 19 22 24 26 27 27

Contents Analysis of the turnover time of class II MHC and its associated peptides

31

Acknowledgements

31

References

31

3. Methods for detecting signalling via MHC class II molecules which can lead to either activation or programmed cell death

33

N. Mooney, J. P. Truman, and D. Charron 1. Introduction

33

2. Changes in intracellular calcium levels in signal transduction

34

3. Inositol phospholipid turnover during lymphocyte activation

36

4. Signal transduction by kinases and phosphatases

38

5. Detection of tyrosine kinase activity

40

6. Immunoprecipitation of tyrosine-phosphorylated substrates

42

7. Detection of programmed cell death via MHC class II molecules

44

8. A flow cytometric method for the quantification of cell death

44

9. Visualization of programmed cell death by fluorescence microscopy

45

Summary

46

References

46

4. Peptide translocation into the ER

49

Jacques Neefjes, Frank Momburg, Gunter Hammerling, and Joost Roelse 1. Introduction

49

2. TAP-dependent peptide translocation Peptide translocation in streptolysin O-permeabilized cells Peptide translocation in microsomes

50 50 54

3. Peptide substrates Peptide size The sequence of peptides Peptide modifications Competition for peptide translocation

56 56 56 56 56

References

57

x

Contents

5. The 20S proteasome and antigen presentation

59

Angela Seelig and Peter M.-Kloetzel 1. Introduction

59

2. Biochemical characterization of the 20S proteasome Purification Analysis of proteolytic activities of the 20S proteasome

60 60 63

3. Isolation and analysis of proteasomes from cell lines before and after IFN-7 induction

65

References

69

6. Analysis of MHC class II-specific T cell clones

71

Graham Pawelec, Fumiya Obata, David Sansom, Hilke Friccius, Thomas Daikeler, Medi Adibzadeh, Kurt Schaudt, and Heike Pohla

1. Introduction

71

2. Source of cells Isolation and activation of T cells

72 72

3. T cell cloning

75

4. Culture media

78

5. Cryopreservation

79

6. Restimulation of MHC class II-specific TCC and data analysis Proliferation assay Assay of cytokine mRNA by polymerase chain reaction (PCR)

80 80 82

7. Sequencing T cell receptors for antigen

85

Acknowledgements

88

References

89

7. Methods for studying the role of the major histocompatibility complex in NK cell cytotoxicity

91

Jose Pena, Rafael Solana, Francisco Borrego, and Bent Rolstad 1. Introduction

91

xi

Contents 2. Definition of NK cells Cell surface markers Tissue distribution and turnover Cytokine production and response to cytokines Functions Receptors Generation of IL-2 activated NK cells, lines and clones in vitro

92 92 92 94 94 95 95

3. In vitro assay for NK cell cytotoxicity against normal allogeneic cells and tumor cell lines

99

4. Description of the human, rat and mouse MHC regions relevant for NK alorecognition Rat and mouse MHC Human MHC

101 101 102

5. Concluding remarks

103

References

104

8. The interaction of superantigens with MHC class II molecules

107

Carol Morgan and John D. Fraser 1. Introduction

107

2. Production of recombinant bacterial superantigens

109

3. Binding of staphylococcal enterotoxins to MHC class II

113

4. Zinc binding by type I staphylococcal enterotoxins, SEA, SEE, and SED

115

5. T cell stimulation by superantigens

117

Acknowledgements

119

References

119

9. Methods for the generation of T cell clones and epithelial cell lines from excised human biopsies or needle aspirates 123 G. De Libero 1. Introduction

123

2. Establishment of T cell clones

123

3. Preparation of cells from tissue biopsies

127

xii

Contents 4. Isolation of epithelial cells from human biopsies Medium for expansion of epithelial cells Medium for expansion of 3T3 cells 5. Use of T cell clones to study tissue compartmentalization of human TCR -yS cells

132 134 136

Acknowledgements References

139 139

10. The application of mass spectrometry to the analysis of peptides bound to MHC molecules

136

141

Andrea L. Cox, Eric L. Huczko, Victor H. Engelhard, Jeffrey Shabanowitz, and Donald F. Hunt 1. Introduction

141

2. Preparing samples for analysis by electrospray ionization mass spectrometry Immunoaffinity purification of MHC-associated peptides

142 142

3. Reverse-phase HPLC (RP-HPLC) of eluted peptides First dimension chromatography Second dimension chromatography Column effluent splitting device 4. Mass spectrometry for peptide analysis Electrospray ionization mass spectrometry for peptide mass analysis Sequencing peptides with tandem mass spectrometry

146 146 147 149 149

5. Chemical modification of peptides Acknowledgements References

157 159 159

11. The molecular basis of alloreactivity Liz Lightstone, Sarah Deacock, Tricia Heaton, and Robert Lechler 1. Introduction 2. Allo-reactive T cells 3. Cell lines for use as allo-stimulator cells 4. Transfection of class II molecules into adherent and non-adherent cells Preparation of DNA xiii

149 155

161 161 163 164 165 165

Contents Which cells to use for transfection? Choice and preparation of selection agents

166 166

5. Measurement of frequencies of alloreactive cells Theoretical basis of LDA Outline of LDA Measurement of alloreactive Th cells Technical notes Statistical analysis of LDA

175 176 177 178 182 183

6. Applications Generation of allo-specific T cell clones using mouse L-cells expressing human class II molecules Detecting precursor frequencies of alloreactive T cells—a tool in transplantation The nature of the allo-ligand The indirect pathway of allo-recognition Determining the nature of co-stimulatory molecules required for allo-reactivity Investigating the allo-specific T cell repertoire

184

7. Summary

195

References

195

12. Elution and analysis of peptide pools from MHC class I molecules Simon J. Powis and Geoffrey

184 184 185 193 194 194

199

W. Butcher

1. Introduction

199

2. Affinity purification of class I molecules

201

3. HPLC separation and sequencing of peptide pools

204

4. Use of peptide motifs

207

Acknowledgements

208

References

208

Al List of suppliers

211

Index

215

xiv

Contributors MEDI ADIBZADEH

Section for Transplantation Immunology and Immunohematology, University of Tubingen Medical School, D-72076 Tubingen, Germany. F. BORREGO

Department of Physiology and Immunology, Faculty of Medicine, Hospital "Reina Sofia", University of Cordoba, Spain. GEOFFREY BUTCHER

Department of Immunology, Babraham Institute, Babraham, Cambridge, CB2 4AT. D. CHARRON

Inserm U396, Institut Biomedical des Cordeliers, 15 rue de 1'Ecole de Medecine, 75006 Paris, France. ANDREA L. COX

University of Virginia, Department of Chemistry, Charlottesville, VA, 22901, USA. THOMAS DAIKELER

Section for Transplantation Immunology and Immunohematology, University of Tubingen Medical School, D-72076 Tubingen, Germany. SARAH DEACOCK

Royal Postgraduate Medical School, Hammersmith Hospital, Ducane Road, London W12 OHS. GENNARO DE LIBERO

Department of Research, University Hospital, Hebelstrasse 20, CH4031 Basel, Switzerland. VICTOR H. ENGELHARD

Biological Mass Spectrometry Lab., University of Virginia, Dept. of Chemistry, McGormick Road, Charlottesville, Virginia 22901, USA. NELSON FERNANDEZ

University of Essex, Department of Biological and Chemical Sciences, Wivenhoe Park, Colchester, CO4 3SQ. JOHN D. FRASER

Department of Molecular Medicine, University of Auckland School of Medicine, Auckland, New Zealand.

Contributors HILKE FRICCIUS

Section for Transplantation Immunology and Immunohematology, University of Tubingen Medical School, D-72076 Tubingen, Germany. GUNTER HAMMERLING

Netherlands Cancer Institute, Plesmanlaan 21, 1066 Amsterdam, The Netherlands. TRICIA HEATON

Royal Postgraduate Medical School, Hammersmith Hospital, Ducane Road, London W12 OHS. CAROLHORGAN

Department of Molecular Medicine, University of Auckland School of Medicine, Auckland, New Zealand. ERIC L.HUCZKO

Biological Mass Spectrometry Lab., University of Virginia, Dept. of Chemistry, McGormic Road, Charlottesville, Virginia 22901, USA. D. F. HUNT Biological Mass Spectometry Laboratory, University of Virginia, McGormick Road, Virginia 22901, USA. PETER M.-KLOETZEL

Humboldt-Universitat zu Berline-Charite, Hessiche Str.3/4,10115 Berlin. ROBERT LECHLER

Postgraduate Medical Hospital, Hammersmith Hospital, Ducane Road, London W12 OHS. LIZ LIGHTSTONE

Royal Postgraduate Medical School Hammersmith Hospital Ducane Road London, W12 OHS. N. MOONEY

Inserm U396, Institut des Cordeliers, Escalier A - ler etage, 15, rue de 1'Ecole de Medicine 75006, Paris, France. FRANK MOMBURG

Netherlands Cancer Institute, Plesmanlaan 21, 1066 Amsterdam, The Netherlands. JAQUES NEEFJES

Netherlands Cancer Institute Plesmanlaan 121, 1066 Amsterdam, The Netherlands. FUMIYA OBATA

Laboratory of Immunology, Kitstato University School of Medicine, Sagamihgara, Japan. xvi

Contributors GRAHAM PAWELEC

Section for Transplantation Immunology and Immunohematology, University of Tubingen Medical School, D-72076 Tubingen, Germany. J. PENA

Department of Physiology and Immunology, Faculty of Medicine, Hospital "Reina Sofia", University of Cordoba, Spain. HEIKE POHLA

Present address: Urology Clinic, University of Munich, Germany. Previous address: Section for Transplantation Immunology and Immunohematology, University of Tubingen Medical School, D-72076 Tubingen, Germany. SIMON J. POWIS

Department of Biochemistry, University of Dundee, Dundee, DD1 4HN. PAMELA A. REID

Department of Biochemistry, University of Dundee, Dundee, DD1 4HN. JOOST ROELSE

Netherlands Cancer Institute, Plesmanlaan 121, 1066 Amsterdam, The Netherlands. B. ROLSTAD

Immunobiological Laboratory, Department of Anatomy, Institute of Basic Medical Sciences, University of Oslo, Norway. DAVID SANSOM

Bath Institute of Rheumatic Diseases, Bath. KURT SCHAUDT

Present address: Cellular and Molecular Biology Laboratory, Department of Biology, University of Sussex, Brighton. Previous address: Section for Transplantation Immunology and Immunohematology, University of Tubingen Medical School, D-72076 Tubingen, Germany. ANGELA SEELIG

AFRC, Babraham Institute, Babraham, Cambridge, CB2 4AT. JEFFREY SHABANOWITZ

Biological Mass Spectrometry Lab., University of Virginia, Dept. of Chemistry, McGormick Road, Charlottesville, Virginia 22901, USA. RAFAEL SOLANA

Department of Immunology, Faculty of Medicine, Hospital Universitatio 'Reina Sofia', Cordoba, Spain. xvii

Contributors BRIGITTA STOCKINGER

Division of Molecular Immunology National Institute for Medical Research Mill Hill London, NW7 1AA. J.-P. TRUMAN

Inserm, U316, Institut des Cordeliers, 15, rue de 1'Ecole de Medicine, 75006, Paris, France. COLIN WATTS

Department of Biochemistry, University of Dundee, Dundee, DD1 4HN.

xviii

Abbreviations ADCC ALC APC DAG DTT EBV EDTA EGTA ELISA ER GM-CSF IL MHC MLC NK PAGE PBMC PBS PCR PHA PKC PLC PMA PMSF SDS SHPP TCC TCR TNF

antibody-dependant cellular cytotoxicity allogenic lymphocyte cytotoxicity antigen presenting cell cAMP cyclic adenosine monophosphate diacyl glycerol dithiothreitol Epstein-Barr virus ethylene diamine tetraacetic acid ethylene glycobis(B-aminoethyl) ether tetroacetic acid enzyme-linked immunosorbent assay endoplasmic recticulum granulocyte-macrophage colony stimulating factor interleukin mAB monoclonal antibody major histocompatibility complex mixed lymphocyte culture natural killer poly acrylamide gel electrophoresis peripheral blood mononuclear cells phosphate buffered saline polymerase chain reaction phytohaemagglutinin protein kinase C phospholipase C phorbol myristate acetate phenylmethyl sulfonyl fluoride sodium dodecyl sulfate succinimidyl-3-(4-hydroxyphenyl) propionate TAP Transporter associated with Antigen Presentation T cell clone T cell receptor tumour necrosis factor

This page intentionally left blank

1 Isolation and functional properties of antigen presenting cells BRIGITTA STOCKINGER

1. Introduction T cell recognition of antigen depends crucially on degradation of protein and binding of protein fragments to MHC molecules which 'present' these fragments to T cells. The term 'antigen presentation' is used to summarize the complex events of antigen internalization, processing of antigen into peptide fragments, peptide transport, and MHC/peptide interaction until the final appearance of peptide/MHC complexes on the surface of antigen presenting cells. Two separate pathways exist using largely non-overlapping sources of antigen and distinct intracellular compartments for binding to different MHC molecules. Most biosynthesized, intracellular antigens are degraded in the cytosol and peptides are translocated into the endoplasmic reticulum (ER), where they bind to MHC class I molecules and after transport to the cell surface are recognized by class I restricted cytotoxic T cells. The MHC class II presentation pathway does not make use of peptides in the ER since the peptide binding site of MHC class II is blocked by the invariant chain. Instead MHC class II molecules are re-routed to endosomal compartments where the invariant chain is degraded, so that peptides from exogenous proteins that have been internalized by antigen presenting cells (APCs) or from endogenous sources of protein that can enter these compartments bind to invariant chain-free MHC class II for recognition by class II restricted T cells. MHC class I molecules are ubiquituously expressed throughout the body, which at least theoretically makes any cell a potential APC for class I restricted T cells. In contrast, class II molecules are (in the mouse) expressed on mainly three cell types only which are classified as 'professional' APC: B cells, macrophages, and dendritic cells. Since most antigens which are to be presented with class II molecules have to be internalized by an APC there are a number of variables which influence the outcome and the efficiency of antigen presentation to MHC class II restricted cells. These variables are: the concentration and nature of the antigen the efficiency of antigen internalization by a particular APC

Brigitta Stockinger processing MHC class II biosynthesis surface expression of MHC class II

2. Isolation and functional properties of antigen presenting cells In the following the main characteristics of the three major antigen presenting cell classes and what is known with respect to the variables listed above are summarized. Detailed descriptions for their isolation and purification are given in the various protocols. A number of reviews are cited at the start of each section.

2.1 Dendritic cells Dendritic cells (1, 2, 3) were discovered relatively recently as a small subpopulation of cells in lymphoid organs. They are derived from bone marrow and constitute less than 1% of total cells in spleen, thymus and lymph nodes. Their characteristic morphology—with many protrusions or long motile processes—is variable in different tissues, a fact which obscured the common lineage and led them to acquire a number of names: Langerhans cells in the skin, interdigitating cells in the medulla of the thymus and in the T cell areas of lymph nodes or veiled cells in afferent lymph. The cells differ in various locations with respect to functional behaviour and their developmental relationships with each other are not entirely resolved. Langerhans cells in the skin resemble more immature precursors of lymphoid dendritic cells, based on the expression of markers that place them closer to the macrophage lineage and similar to their peers developing out of bone marrow cultures, i.e. Fc receptors and F4/80. Upon culture Langerhans cells develop functional and morphological characteristics similar to dendritic cells in spleen and lymph nodes. These cells do not proliferate further. The relative immaturity of skin Langerhans cells is often cited as the reason for their inability to induce strong mixed lymphocyte reactions (MLR). This readout indicates a paucity of adhesion molecules on Langerhans cells which prevents them forming clusters or rosettes with surrounding lymphocytes as an initial step in activation, a function in which mature dendritic cells excel. On the other hand the capacity to present exogenous protein to T cells is diminished or absent in mature spleen dendritic cells or cultured Langerhans cells. However, upon culture, MHC class II and invariant chain synthesis ceases in dendritic cells (4) and the absence of freshly synthesized MHC class II, which was shown to be of prime importance in presentation of exogenous antigen (5), may fully explain their failure to present while it may not necessarily reflect an in vivo characteristic of mature dendritic cells. Unfortunately, the standard isolation technique for spleen dendritic cells involves an overnight 2

1: Isolation and functional properties of antigen presenting cells incubation step (see Protocol 1) which might initiate this process and reduce their efficiency of presentation. This problem can be overcome by including the antigen in the overnight culture step. The technical difficulties in obtaining large numbers of pure dendritic cells and their heterogeneity as indicated by surface markers (6) has hindered the analysis of antigen uptake and processing. Dendritic cells were originally described as lacking efficient fluid phase endocytic activity (7) and acidified endosomal compartments, whereas later evidence indicated that their endocytic activity is as high as that of other APC (8). Phagocytic activity is absent in mature lymphoid dendritic cells whereas Langerhans in the skin possess this ability. The gaps concerning antigen uptake and processing will no doubt be filled in the near future following the discovery that large numbers of dendritic cells can be generated from bone marrow cultures under the influence of GM-CSF (granulocytemacrophage colony stimulatory factor) (see refs. 9 and 10, and Protocol 2). Perhaps the most striking characteristic of dendritic cells which sets them apart from the other APC is their constitutive expression of 'co-stimulatory' molecules like B7 (CD80) (11) and the high levels of MHC class II. The latter is remarkably stable even in conditions under which new synthesis of class II ceases. Thus dendritic cells in culture retain for at least a few days MHC class II/peptide complexes which were generated either in vivo (4) or in vitro (12) on the cell surface in marked contrast to macrophages (see Section 2.2.2.2.) which rapidly lose surface class II once placed into culture. This attribute of dendritic cells emphasizes their importance for antigen presentation to MHC class II restricted cells and lends support to the hypothesis that one of thenfunctions is to trap and process antigen in non-lymphoid organs like the skin and subsequently transport it to lymph nodes and spleen for presentation to T cells (13).

Protocol 1. Isolation of dendritic cells from spleen Equipment and reagents DNase (Sigma D-25): prepare 1%stock solution in serum-free medium and freeze at -20°C BSS No.1 UOx): dextrose (anhydrous),10 g (or (x 1 H2O)10.42 g); KH2PO4, 0.4 g; make up to 1 litre; Na2HPO4 (anhydrous), 2.6 g; phenol red (0.5%), 20 ml BSS No.2 (10X): CaCI2 (anhydrous), 1.41 g (or CaCI2 x2 H20, 1.86 g); KCB, 4.00 g; NaCI, 80 g; MgCI2 anhydrous, 1.04 g (make up to 1 litre) (or MgCI2 x 6 H2O, 2.25 g); MgS04 anhydrous, 1.74 g (or MgS04 x 7 H2O, 3.57 g) BSS No.3 (10x): dextrose (anhydrous),10 g (or x 1 H2O, 10.42 g); Kh2PO4 0.4 g (make up to 1 litre); H3PO4, 1.79 g; phenol red (0.5%), 20 ml

Culture medium: IMDM (Iscoves modified Dulbecco medium, 25 mM HEPES, 5% heat-inactivated fetal calf serum (FCS), 2X10"3 M L-glutamine, 5X10-5 M mercaptoethanol, 100 ngml-1 streptomycin, 100 UmT1 penicillin Air buffered IMDM for cell washing: IMDM with Hepes, penicillin and streptomycin as above. Per 1 liter add 2.1 g NaCI and 2.5 ml 5 M NaOH Collagenase (Worthington CLS4): prepare 8 mg ml-1 stock in serum-free medium and freeze at -20°C BSS No.4 (10x): same as BSS No.2, make up to 950 ml and add 50 ml 1 M HCI (vary this amount if the Percoll pH changes)

3

Brigitta Stockinger Protocol 1. Continued Stock Percoll: 20 ml Percoll (Pharmacia); 2.5 ml BSS No.3 (10x), pH 7-7.4; 0.5 ml BSS No.4 (10X); 2 ml BSS No.2 (10x)

Working BSS (W-BSS): 80 ml H20; 10 ml BSS No.1; 10 ml BSS No.2 (add in this order, otherwise BSS No.2 might precipitate)

A. Digestion of spleen with enzyme cocktail 1. Place the spleens in a 50 mm non-tissue culture treated petri dish and make one superficial cut with a needle along the spleen surface to facilitate access of the enzymes. 2. Incubate spleens in 5 ml enzyme cocktail for 30 min at 37°C in a CO2 incubator. Up to 5 spleens can be digested with 10 ml cocktail consisting of 2 ml collagenase stock, 1 ml DNase stock, 7 ml serum free IMDM medium. 3. Collect released cells in a tube, place on ice and incubate spleens for another 45 min in 5 ml enzyme cocktail. 4. After the final incubation dissociate spleens very gently preferably by repeated pipetting using a 25 ml pipette. Avoid extensive mechanical disruption which will destroy dendritic cells. 5. Fill up the tube with 5% PCS containing IMDM medium and spin cells at 300 x gfor 10 min. 6. Wash once in serum free air buffered medium. B. Percoll density centrifugation 1. Make up gradient solutions of the following densities: 1. = 2. = 3. = 4. =

1.09 1.085 1.075 1.060

2.5 ml stock Percoll + 2.5 ml stock Percoll + 2.5 ml stock Percoll + 2.5 ml stock Percoll +

0.25 ml W-BSS 0.5 ml W-BSS 1 ml W-BSS 1.75 ml W-BSS

2. After the last wash resuspend the cells in solution No.2, carefully layer onto solution No.1 with a 5 ml pipette and then layer solutions 3 and 4 on top. Finish with 2 ml of W-BSS. 3. Using 16 x 125 mm polystyrene tubes (Falcon 2025), spin for 20 min at 1300 X g without brake. 4. Pipette off the top layer of cells between W-BSS and solution No.4, then wash three times in air buffered medium. This fraction contains dendritic cells, macrophages, activated B cells and very few T cells. C. Separation of B cells from macrophages and dendritic cells by adherence 1. Place cells onto tissue culture treated petri dishes (maximally 5 x 107 cells per 90 mm dish) and incubate for 3 h at 37°C in a C02 incubator. 4

1: Isolation and functional properties of antigen presenting cells 2. Wash off non-adherent cells (mainly B cells) with pre-warmed culture medium. Macrophages and most dendritic cells remain adherent. 3. Place the dish back in the incubator for an overnight incubation step. If you want to use dendritic cells for presentation of an exogenous protein, include it in the overnight incubation step. D. Recovery of dendritic cells 1. Dendritic cells detach from the plastic surface after overnight culture and can be collected by thorough washing of the dish. Most macrophages remain firmly adherent and can only be detached by incubation with Trypsin/EDTA solution. 2. To further purify dendritic cells, remove contaminating B cells (about 20%) by panning on anti-lg coated plastic dishes. These can be prepared a day in advance by coupling goat-anti-mouse IgG+IgM (Jackson Immunochemicals) at 10 (xgmM in 0.05 mM Tris buffer pH 8.5 to the plastic surface for 40 min at room temperature followed by 5 washes in PBS containing 10% FCS to block remaining sites on the plastic. Add maximally 5 x 107 cells per 90 mm dish) and leave for 90 min at 4°C. The yield of dendritic cells is about 1% of total spleen cells and their puritiy is usually 95% as checked by FACS analysis.

It is now well established that dendritic cells play a crucial role in the initiation of immune responses in vivo (14) and in fact may be the only APC that can activate naive T cells in vivo. They can be used as physiological adjuvants for the deliberate induction of immune responses in vivo, abolishing the need for other often harmful substances to increase the immunogenicity of certain antigens. Their antigen presentation potency is so strong and they act in such small numbers that it is a major problem excluding small contaminations with dendritic cells when isolating and assessing the functions of other APC from mixtures of cell populations. Dendritic cells play a decisive role not only in initiating immune responses, but also in the induction of self tolerance in the thymus. Thymic dendritic cells do not enter the thymus as mature dendritic cells from the blood, but are generated in situ from precursors colonizing the thymus at the same time as T cell precursors (15). They are localized in the medulla and at the corticomedullary boundary, but are absent from the cortex. This may have important consequences for tolerance induction in MHC class II restricted T cells, which appear to be dependent on presentation of self antigen by dendritic cells (see ref. 12). Therefore encounter with self antigen may for MHC class II restricted T cells occur at a later, post-cortical stage in development, compared with MHC class I restricted T cells which seem to be sensitive to tolerogenic presentation by a variety of MHC class I expressing cells including thymocytes. 5

Brigitta Stockinger It can fairly be stated that dendritic cells, which for many years played a background role and were regarded by some as an exotic but dispensable APC population, have been promoted to a key role in the initiation of immune responses and self tolerance. Protocol 2. Generation of dendritic cells in bone marrow cultures Reagents See Protocol 1 A. Preparation of bone marrow suspensions 1. Excise femura and clean of flesh using forceps and a scalpel. 2. Flush out bone marrow with air buffered wash medium in a syringe with needle and wash three times in wash medium with centrifugation steps for 10 min at 300 x g. 3. Adjust the cell suspension to 10X106 cells ml-1 in air buffered medium. B. Removal of T and B cells from the bone marrow suspension 1. Incubate cells in a 37°C waterbath with antibodies to Thy1 and the B cell marker B220 for 20 min followed by addition of rabbit complement (Cedarlane) to a final concentration of 1:10 and further incubation for 60 min. 2. Wash cells three times and adjust to 5 x 105 cells mT-1 in culture medium containing 25 ng ml-1 GM-CSF. Culture cells in a total volume of 10 ml in 90 mm tissue culture treated petri dishes. C. Partial depletion of granulocytes 1. On day 3 of culture carefully remove non-adherent cells, which are mostly granulocytes, by pipetting off without extensive swirling of the dish. 2. Add back fresh GM-CSF containing medium. This procedure will lead to some loss of dendritic cells, but will greatly reduce the number of contaminating granulocytes. Omit this step if you require higher yields of dendritic cells and the purity is not important (granulocytes do not express MHC class II molecules). Maximal yield of dendritic cells can be obtained between day 6 and 8 of culture. 3. Since by day 7 the original petri dishes contain an almost confluent monolayer of macrophages some of them will contaminate the nonadherent population of dendritic cells. They can be removed by readhering the washed off population on fresh petri dishes 1-2 h prior to an assay. The yield and purity of bone marrow dendritic cells can be assessed by FACS analysis. 6

1: Isolation and functional properties of antigen presenting cells

2.2 Macrophages Macrophages (16, 17) differentiate from bone marrow derived monocytes which as they leave the blood and settle down in different tissues develop diverse phenotypic and functional characteristics. Early cells of the macrophage lineage proliferate in bone marrow and tissues, but as they mature they stop dividing. Antigen presentation is only one of many tasks of the macrophage. Their unique potential to internalize extracellular particles by phagocytosis makes them a first line defence against invading bacteria, and a critical component in the disposal of senescent red blood cells, thymocytes destined to undergo apoptosis, or antibody-coated antigens. Macrophages are highly secretory cells and products like proteases, complement components or cytokines are important as inflammatory mediators. Cytokine and complement receptors on their surface involve them in interactions with lymphocytes which can result in dramatic changes in metabolic state. Macrophages activated by e.g. "y-interferon released by T lymphocytes are highly microbicidal and tumoricidal due to increased synthesis of proteins like TNF and IL-1, up-regulation of catabolic enzyme synthesis and generation of reactive oxygen intermediates. The role of macrophages in antigen presentation has been extensively studied. Although they can take up antigen by fluid phase pinocytosis, internalization is enhanced 4000-fold for particulate antigen (18), which is rapidly phagocytosed. In areas of inflammation a2 macroglobulin activated by local proteolysis may form complexes with antigens, facilitating their internalization via high affinity receptors on the surface of macrophages (19). Once inside phagolysosomes antigen is degraded by a cascade of proteolytic enzymes in increasingly acidic conditions until resulting peptides are rescued from complete catabolism by interaction with MHC class II molecules. Interference with the pH gradient by lysosomotropic agents like chloroquin or ammonium chloride abolishes presentation. In a similar way bacteria such as Listeria monocytogenes can inhibit their own degradation with the help of the protein listeriolysin which forms pores in the membranes of phagolysosomes, destroying the pH gradient and allowing the escape of Listeria into the cytoplasm where they can replicate undisturbed by lysosomal enzymes (20). Macrophages are the only APC which can process external antigens not only for MHC class II presentation, but also by an as yet undefined mechanism can shunt them into the MHC class I presentation pathway which is normally inaccessible to exogenous antigen (21, 22). The expression of MHC class II on macrophages is not constitutive as it is on dendritic cells, but is under regulation. The ratio of MHC class II positive to negative macrophages varies considerably among different tissues. For instance, macrophages in the spleen red pulp are mostly MHC class II positive, whereas those in the white pulp are negative. Peritoneal macrophages and thymic macrophages are mostly MHC class II negative. Although macrophage class II is induced by interaction with T cells through 7

Brigitta Stockinger -ylFN, the basic distribution of class II on macrophages is T cell independent as shown in athymic nude mice or scid mice. Once placed in culture macrophages rapidly lose MHC class II and antigen presentation function unless expression is re-induced by -yIFN. Macrophage antigen presentation in vivo is crucial for the destruction of particulate pathogens like bacteria. Certain pathogens seem to preferentially induce Thl-type mediated immunity which correlates with the production of IL-12 by macrophages (23). It is likely, but not proven, that T cell responses to particulate antigens have to be initiated by dendritic cells as well, which presents the dilemma that these are (apart from immature precursors) unable to phagocytose. It is possible, however, that macrophages make bacterial antigens accessible for presentation by dendritic cells through secretion of proteases. The role of macrophages in the induction of tolerance in the thymus is less clear. Presentation of self antigens in the context of MHC class II is not possible because they express very little if any class II. They might present self antigens for negative selection in the context of MHC class I, but their most crucial role in the thymus seems to be the disposal of thymocytes that are destined to die by apoptosis, either because of failure to be positively selected or because of negative selection through self antigen recognition. Protocol 3. Isolation of macrophages from peritoneum Reagents See Protocol 1 A. Non-induced macrophages 1. Kill mice by C02 rather than cervical dislocation to avoid any bleeding. 2. Pull back fur and inject 5 ml ice cold Ca2+ Mg2+ free PBS into the abdomen with a small needle. 3. Massage the abdomen gently and then withdraw the fluid with a large needle bevel pointing down avoiding trapping of fat. About 90% of the injected fluid can normally be recovered. 4. Non-induced peritoneal lavage contains about 30% macrophages which can be purified by adherence and subsequent detachment with Trypsin/EDTA. B. Activated macrophages 1. Inject mice 5-10 days before sacrifice with 2 ml 3% thioglycollate broth (Difco). This will increase the yield of macrophages at least 10-fold. 8

1: Isolation and functional properties of antigen presenting cells C. Induction of MHC class II positive macrophages 1. Non-elicited as well as thioglycollate-elicited peritoneal macrophages contain only about 5-15% MHC class II positive cells. Injection with 100 ug KLH (keyhole limpet haemocyanin) i.p. followed 1 week later by i.p. injection of 1 ml 10% proteose-peptone (Difco) will lead to MHC class II induction on 60-90% of peritoneal macrophages which are harvested 3 days later.

2.3 B cells B cells (24,25) are the effector cells of humoral immunity and their function in antigen presentation is entirely self-centred. MHC-restricted cognate interaction with T cells has long been known to be a prerequisite for antibody production. The molecular nature of this interaction was shown to be the Ig receptor-mediated internalization of antigen and its subseqent presentation as peptides on MHC class II molecules. Ig receptor-mediated endocytosis is very efficient, allowing presentation at concentrations of antigen which occupy only about 0.05% of surface Ig (26). In contrast, non-specific presentation by B cells is very inefficient since these cells have a 100- to 1000-fold lower capacity to pinocytose than macrophages (27) and are non-phagocytic. Contact with antigen activates a resting B cell, resulting in up-regulation of co-stimulatory molecules like B7 (CD80). This allows induction and binding of CD40 ligand on T cells, the subseqent cognate interaction via T cell receptor recognition of MHC class II bound peptides, and eventually the delivery of cytokines to the B cell. Although it is known that B cells cannot activate a resting T cell (28), and in some cases might even tolerize them (29), the issue of how the sequential interaction of T cells with dendritic cells and B cells is orchestrated in vivo is at present unresolved. MHC class II molecules are expressed on all B cells with the exception of plasma cells and pre-B cells. The levels of expression are lower than on dendritic cells, but can be up-regulated by contact with IL-4. MHC class II molecules recycle on many B cell lines studied in vitro and it is a matter of some debate whether peptides from internalized and processed antigens associate with freshly synthesized MHC class II molecules or with class II molecules recycling from the surface. In the latter case it was assumed that class II might regenerate its binding site and exchange peptides during recycling (30). On the other hand a very detailed biochemical analysis of peptide-class II interaction detected binding of radiolabelled antigen fragments to newly synthesized rather than surface class II molecules (5). Furthermore it was shown that the half-life of a particular peptide/class II complex on the surface of an EBV-transformed B cell as detected by T cell recognition correlated with the half-life of class II itself (31) supporting a stable association of a given peptide with class II. This leaves the possibility that the relative usage of 9

Brigitta Stockinger recycled vs. newly synthesized class II for peptide binding might depend on the extent of processing required to generate a given T cell epitope. Apart from their proposed role in inducing peripheral tolerance, B cells are involved in central tolerance induction in the thymus. A small sub-population of CDS B cells, resident in the medulla cells and comprising about 0.5% of total thymus cells, was reported to be responsible for tolerance induction to Mis determinants, but its involvement in tolerance induction to other self antigens is unknown. Protocol 4. Isolation of B cells from spleen Reagents See Protocol 1 A. Preparation of cell suspensions 1. To minimize contamination with dendritic cells, mechanically tease spleens with needles instead of using enzyme digestion. 2. Wash the cell suspension twice and lyse red cells with Gey's solution (Gibco). B. Depletion of T cells, macrophages and dendritic cells 1. Deplete T cells by two rounds of anti-Thyl antibody and complement as described in Protocol 2. 2. Deplete macrophages by adherence for 2 h on tissue culture treated petri dishes. 3. If virtually complete removal of macrophages is desired, run the cell suspension through a Sephadex G-10 column. 4. Complete removal of dendritic cells is the most difficult task. As few as 100 contaminating dendritic cells can still present antigen which can lead to overestimation of B cell antigen presentation capacity.We routinely use a combination of adherence and depletion by incubation with biotinylated N418 (a dendritic cell-specific antibody ATCC HB 224) followed by removal on Streptavidin-coated Dynabeads. C. Isolation of low densitiy (activated) and high density (resting) B cells (32) 1. Centrifuge spleen cell suspensions on Percoll gradients as described in Protocol 1. An additional layer of density 0.08 gml-1 is introduced between densities 0.085 and 0.075 g ml-1. 2. Recover high density resting B cells at the 1.08-1.085 gml-1 interphase and activated low density B cells at the 1.06 to 1.075 gml-1 interphase.

10

1: Isolation and functional properties of antigen presenting cells

3. Functional assays of antigen presentation Activation of T cells can be measured in terms of antigen-specific proliferation, lymphokine secretion, cytotoxicity, or in the very early stage as elevation of intracellular calcium levels. The readout used to assess T cell activation depends on the type of T cell and the antigen in question. The clone size for any given T cell specificity in vivo is usually very small (in the order of 1 in 105) with the exception of alloreactive T cells which are composed of a large number of cells recognizing a multitude of peptides presented by foreign MHC molecules. This means that in order to detect a T cell response against a foreign protein in vitro, an in vivo immunization step which serves to increase the clone size of antigen-specific T cells has to precede culture in vitro. Injecting the antigen in complete Freund's adjuvant or as alum precipitate or infecting an animal with virus will usually guarantee successful priming such that a specific response can be recalled in vitro. However, in this case it is not possible to control which APC gets access to the antigen and how this might influence the subsequent response. As mentioned earlier (Section 2.1) injection with antigen-pulsed dendritic cells is a powerful means of T cell priming which obviates the need for adjuvant. Subsequent repeated re-stimulation in vitro can generate antigen-specific T cell clones and these can be fused to the thymoma BW to obtain T cell hybrids. Neither T cell clones nor hybrids can be considered naive T cells and antigen presentation requirements will differ from those for resting T cells in vivo, i.e. they may not be dependent on co-stimulatory interactions involving B7. The development of transgenic technology has made it possible to create mice with a single T cell receptor specificity which can be used to address directly questions like antigen presentation requirements for activation of resting T cells in vitro. In the following, examples are given of T cell activation assays using different APC and various readouts. These examples concern T cell responses against a serum protein, the fifth component of complement (C5). Obviously it is not possible to give a standard protocol that applies to all T cells and all antigen specificities. Cell numbers, antigen concentrations and assay times will have to be determined freshly for other T cells with different specificities, but the general points relating to antigen uptake, its dependence on the nature of the antigen and the relative numbers of APC required for this task are likely to be similar for other proteins which stimulate MHC class II restricted responses.

3.1 Presentation requirements for activation of C5-specific T cell hybrids and clones The differential capacities of APC to present an antigen are only fully evident in a concentration range that mimicks physiological conditions. In vast antigen excess even APC with a very poor non-specific endocytosis rate like B cells will get hold of enough antigen to activate a T cell clone. A soluble 11

Brigitta Stockinger protein like C5 is presented very efficiently by dendritic cells in antigen concentrations as low as 1 ngmT1. Presentation by B cells requires at least 5-10 ixgmr1 of C5. However, the C5 concentration needed for B cell presentation can be dramatically reduced (almost to that presented by dendritic cells) if C5 can be internalized by the Ig receptor. As an example this situation was mimicked using B cells isolated from a transgenic mouse with a TNP-specific B cell receptor and C5 coupled with the TNP hapten as antigen (12). In the same context it can be illustrated that macrophage antigen presentation depends largely on the physical shape of the antigen (33). The lowest dose of soluble C5 which initiated a detectable IL-2 response by a C5-specific T cell hybrid was 1 ug ml-1. In the presence of CS-specinc antibodies, forming immune complexes, the C5 dose required for a minimal response was lowered to 0.2 ug ml-1. However, when C5 was turned into a particulate antigen by binding to latex beads which had been coupled with C5-specific antibodies (34), a T cell response could now be induced with as little as 10 ng ml-1 C5.

3.2 Readouts for T cell activation Assays for T cell activation are conveniently set up in 96-well tissue culture plates. When working with T cell hybrids which proliferate constitutively it is advisable to use flat bottom plates to avoid overcrowding during the 24 h incubation time with APC and antigen. T cell clones or T cells from transgenic mice on the other hand are best cultured with APC in round bottom plates to guarantee optimal cell contact for activation. Responder cell numbers for hybrids are optimal around 5xl04 cells/well whereas the concentrations for T cell clones should be between 2 and 5X104 cells/well. Dendritic cells as APC will function in numbers as low as 500 cells/well, but optimal responses are achieved with concentrations around 0.5-1 X104 cells/well. Increasing the dose does not increase antigen presentation, but neither does it result in inhibition. In contrast the dose of macrophages should not exceed 1x105 cells/well; usually 5xl04cells are sufficient. Higher doses tend to inhibit responses, probably due to the release of prostaglandins, nitric oxide, and other inhibitory or toxic molecules. The most convenient readouts for T cell activation are lymphokine release (the only option for T cell hybrids), antigen-specific proliferation, or for MHC class I restricted clones, cytotoxic killing of labelled target cells. To measure release of IL-2 IL-3 or IL-4 (Protocol 5) T cell and APC are co-cultured for 24-48 h (the optimum has to be determined for each different T cell). Culture supernatant is removed and added to indicator cells which are dependent for their growth on the relevant lymphokine to be tested. IL-2 dependent cells are CTLL (ATCC TIB 214), IL-3 dependent cells are PB-3c (35) or 80FD/C.1 (36) and IL-4 dependent cells are CT4S (37). Release of -ylFN tends to peak later at about 72 h after the start of the culture and is assayed by an ELISA capture assay (Protocol 6). It is important to point out that T cell hybrids switch off 12

1: Isolation and functional properties of antigen presenting cells ylFN production and can therefore only be assayed for IL-2, 3,or 4 release. Antigen-specific proliferation is measured by 3H-thymidine incorporation and the optimal incubation time has to be determined for each T cell type. The killing of target cells by cytotoxic T cells is conventionally assayed by incubation of activated T cells (optimum development of killing activity between day 3 and 6 of culture) with chromium 51-labelled target cells which release the radioactive label when they are killed. Background due to nonspecific leakage of label is often a problem depending on the type of target cells used. The only reliable targets with low background and high uptake of label are tumour cells which are not available for all MHC types. An alternative to killing of chromium labelled target cells is the measurement of apoptosis in target cells labelled with 3H-thymidine (Protocol 7). This test measures surviving rather than dead cells. Delivery of the lethal hit by a cytotoxic effector cell will lead to fragmentation of DNA in the target cell. Fragmented DNA will be washed off during harvesting of the cultures while non fragmented DNA of surviving cells will be retained on the filter (38). Protocol 5. Measurement of IL-2, 3 and 4 release Reagents See Protocol 1 Method 1. Transfer 50 ul of culture supernatant 24-48 h after initiation of T cell stimulation with ARC and antigen into fresh flat bottom 96-well plates. 2. To eliminate any contribution of accidentally co-transferred cells from the original culture irradiate the plates with 2000 rad or freeze before continuing the assay. 3. Harvest IL-2 dependent CTLL cells (or IL-4 dependent CT4S or IL-3 dependent PB-3c) and wash at least 3x in air buffered medium with centrifugation steps at 300 x g in between. It is important that any trace of lymphokine in these cell lines is washed out for the assay. 4. Add between 5 x 103 and 1 x 104 lymphokine-dependent cells per well to supernatant to be assayed for activity (optimum number of cells should be tested) and incubate for 24-30 h. Control wells should be set up with titrations of IL-2 and with indicator cells in medium. 5. Add 3H-thymidine when controls wells with cells in medium are dead (usually 24-30 h after addition of indicator cells. 6. Leave thymidine for a minimum of 8 h before harvesting in an automated 96-well Betaplate harvester. 7. Measure counts in a liquid scintillation B counter.

13

Brigitta Stockinger Protocol 5.

Continued

8. Important: CTLL cells proliferate to some extent in IL-4 and CT4S cells proliferate in IL-2. To allow unequivocal identification of the lymphokine responsible for proliferation, antibodies specific for the alternative lymphokine have to be present i.e. anti-IL-2 (S4B6 ATCC HB 8794) or anti-IL-4(11B11 ATCC HB 188)

Protocol 6.

Capture ELISA for detection of -ylFN

Equipment and reagents 96-well flat bottom flexible plates (Dynatech) Antibodies: AN18 (rat) anti-mouse -ylFN (Pharmingen), R4-6A2 (rat -y1) anti mouse -y IFN (ATCC HB 170) (one of the two antibodies is biotinylated, it does not matter which one) ELISA wash solution: PBS, 0.05% Tween20 Coating buffer: 0.2 M Borate buffer ph 8.5

Blocking buffer: PBS with 5% PCS, 5% horse serum Developing reagents: Streptavidin-horseradish peroxidase (Southern Biotechnology) 1:500 ABTS substrate Sigma A-1888 Substrate buffer: 0.1 M citric acid in dH20, pH 4.35

Method 1. Add 150 mg substrate to 500 ml substrate buffer, check pH, aliquot into 11 ml vials and store at -20°C. Add 10 ul 30% H2O2/vial prior to use. 2. Coat wells of 96-well flexiplates with first antibody at 10 ug ml-1 (50 ul/well) in coating buffer. 3. Incubate overnight in a humidified box at 4°C or 2 h at 37°C. 4. Block remaining sites on the wells with blocking buffer and leave at room temperature for 1 h. 5. Add 50-100 ul of test supernatants, incubate at room temperature for 2 h or overnight at 4°C. 6. Wash at least three times with wash solution. Add biotinylated second antibody (about 5 ug ml-1 in PBS, 50 ul/well, but concentration needs to be tested for every new batch), leave for 1 h at room temperature, and wash as before. 7. Add developing Streptavidin-HRPO conjugate 1:500 in PBS for 1 h at room temperature, wash again, and add 100 ul/well substrate. The reaction develops fast (within 5-10 min) and should be read immediately in an ELISA reader at an O.D. of 414 nm. Use wells with just substrate solution as blanks.

14

1: Isolation and functional properties of antigen presenting cells Protocol 7. JAM test for measurement of cytotoxic activity A. Pulsing of target celts? 1. To tumour targets add 5 (uCiml-1 3H-thymidine for 3-4 h (use in early log phase; 0.3 x 106 cells/ml) 2. To ConA blasts add 10 Urnl-1 IL-2 and 5 uCiml-1 3H-thymidine for 4-6 h. (2 x 106 spleen cells/ml, 2.5 ug ConA, cultured upright in 10 ml volume in 25 ml tissue culture flasks for 24 h) 3. To LPS blasts add 5 (xCimr1 3H-thymidine for 4-6 h. (30 x 106 spleen cells in 30 ml volume with 300 ug LPS cultured for 48 h) B. Assay 1. Wash cultured cells twice and add to 96 round bottom microtiter wells. 2. Incubate with 104 pulsed target cells per well in different effector to target ratios, which have to be determined for each type of responder cell. (For CTL clones, start at E:T 10:1 and set up serial dilutions; for in vitro boosted cells from primed mice use E:T around 100:1 as starting point.) Incubate for 2-4 h depending on effector cell type. C. Harvesting and calculation of results 1. For each assay set up a spontaneous release (SR) control (target cells only, incubated and harvested as the experimental set up (Ex)) and a control for the total counts present (targets which are harvested immediately after the pulsing period at the time of addition to the experimental wells). Although the total counts do not enter the equation they serve as an indication of the extent of background apoptosis. 2. Harvest the plates in an automated 96-well Betaplate harvester and measure counts in a liquid scintillation p counter. Calculate the results as: SR — Ex x SR

100 = % specific lysis

aTarget cells have to be proliferating cells.

References 1. Metlay, J.P., Pure, E., and Steinman, R.M. (1989). Adv. in Immunology, 47,45. 2. Inaba, K., Hosono, M., and Inaba, M. (1990). Int. Rev. Immunol, 6,117. 3. Agger, R., Crowley, M.T., and Witmer-Pack, M.D. (1990). Int. Rev. Immunol., 6,89.

15

Brigitta Stockinger 4. Pure, E., Inaba, K., Crowley, M.T., Tardelli, L., Witmer, P.M., Ruberti, G., Fathman, G., and Steinman, R.M. (1990). J. Exp. Med., 172,1459. 5. Davidson, H.W., Reid, P.A., Lanzavecchia, A., and Watts, C. (1991). Cell, 67,105. 6. Aiba, S. and Katz, S.I. (1990). J. Immunol, 145,2791. 7. Schuler, G. and Steinman, R.M. (1985). J. Exp. Med., 161, 526. 8. Levine, T.P. and Chain, B.M. (1993). Proc. Natl Acad. Sci. USA, 89, 8342. 9. Schleicher, C., Mehlig, M., Zecher, R., and Reske, K. (1992). J. Immunol. Methods, 154,253. 10. Inaba, K., Inaba, M., Romani, N., Aya, H., Deguchi, M., Ikehara, S., Muramatsu, S., and Steinman, R.M. (1992).J. Exp. Med., 176,1693. 11. Linsley, P.S. and Ledbetter, J.A. (1993). Annu. Rev. Immunol, 11,191. 12. Stockinger, B. and Hausmann, B. (1994). Int. Immunol., 6,247. 13. Macatonia, S.E., Knight, S.C., Edwards, A.J., Griffiths, S., and Fryer, P. (1987). /. Exp. Med., 166,1654. 14. Inaba, K., Metlay, J.P., Crowley, M.T., Witmer-Pack, M., and Steinman, R.M. (1990). Int. Rev. Immunol, 6,197. 15. Ardavin, C., Wu, L., Li, Ch. -L., and Shortman, K. (1993). Nature, 362,761. 16. Unanue, E.R. and Allen, P.M. (1987). Science, 236, 551. 17. Gordon, S., Fraser, I., Nath, D., Hughes, D., and Clarke, S. (1992). Curr. Opinion in Immunol., 4,25. 18. Steinman, R.M. and Conn, Z.A. (1972). J. Cell. BioL, 55,616. 19. Chu, C.T. and Pizzo, S.V. (1993). J. Immunol., 150,48. 20. Brunt, L.M., Portnoy, D.A., and Unanue, E.R. (1990). J. Immunol., 145,3540. 21. Debrick, J.E., Campbell, P.A., and Staerz, U.D. (1991). J. Immunol., 147, 2846. 22. Kovacsovics-Bankowski, M., Clark, K., Benacerraf, B., and Rock, K.L. (1993). Proc. Natl Acad. Sci. USA, 90,4942. 23. Hsieh, Ch-S., Macatonia, S.E., Tripp, C.S., Wolf, S.F., O'Garra, A., and Murphy, K.M. (1993). Science, 260,547. 24. Lanzavecchia, A. (1987). Immunol. Rev., 99,39. 25. Clark, E.A. and Ledbetter, J.A. (1994). Nature, 367,425. 26. Lanzavecchia, A. (1985). Nature, 314,537. 27. Chesnut, R.W., Colon, S.M., and Grey, H.M. (1982). J. Immunol., 128,1764. 28. Ronchese, F. and Hausmann, B. (1993). J. Exp. Med., 177,679. 29. Fuchs, E.J. and Matzinger, P. (1992). Science, 258,1156. 30. Adorini, L., Appella, E., Doria, G., Cardinaux, F., and Nagy, Z.A. (1989). Nature, 343,800. 31. Lanzavecchia, A., Reid, P.A., and Watts, C. (1992). Nature, 357,249. 32. Ratcliffe, M.J.H., and Julius, M.H. (1982). Eur. J. Immunol., 12,634. 33. Stockinger, B. (1992). Eur. J. Immunol., 22,1271. 34. Wirbelauer, C., Meding, S., Stockinger, B., Gillard, S., and Langhorne, J. (1989). Immunol. Letters, 23, 257. 35. Erb, P., Kennedy, M., Huegli, G., and Fluri, M. (1988). Prog. Leucocyte BioL, 7,11. 36. LeGros, G.S., Gillis, S., and Watson, J.D. (1985). J. Immunol., 139,4009. 37. Hu-Li, J., Ohara, J., Watson, C, Tsang, W., and Paul, W.E. (1989). J. Immunol., 142,800. 38. Matzinger, P. (1991). J. Immunol. Methods, 145,185.

16

2 Labelling methods for analysis of class II MHC endocytosis and peptide turnover PAMELA A.REID and COLIN WATTS

1. Introduction MHC glycoproteins are peptide binding and transport proteins which give a continual update, in peptide form, of the cellular and environmental protein composition on class I or class II MHC molecules respectively (see ref. 1 for a review). Cells expressing these molecules can therefore be screened by T lymphocytes for the presence of foreign peptides. A response will only be elicited if the level of a particular peptide/MHC complex is above the threshold needed for activation of the responding T cells. Whether this threshold is reached will depend on the relative rates of production and loss of peptide/MHC complexes from the cell surface. Production rates will be governed by variables such as the rate of MHC biosynthesis, level of antigen expression or capture and efficiency of antigen processing etc. (2). Rates of loss will depend on (i) the absolute lifetime on the cell surface of MHC molecules and (ii) the lifetime of particular peptide species on MHC molecules. Clearly, if the lifetime of peptides on MHC molecules is much longer than the absolute lifetime, loss rates will simply reflect how long MHC molecules reside on the cell surface. On the other hand, if peptide lifetime is much shorter than absolute MHC lifetime on the cell surface then the rate of loss of complexes will be largely independent of absolute lifetime. This latter scenario raises the possibility that the 'empty' MHC molecules generated might be re-used to present additional peptides during their tenure on the cell surface. Since peptide production for MHC binding is not thought to occur on the cell surface, an endocytic cycle of some kind would seem to be necessary to permit re-utilization. These considerations raise basic questions about the dynamic behaviour of MHC molecules on cell surfaces and of peptides bound to such molecules. In this chapter we describe some methods which allow several parameters to be assessed. These include: (i) the absolute lifetime of surface MHC molecules,

Pamela A. Reid and Colin Watts (ii) the lifetime of peptides bound to class II MHC molecules, and (iii) endocytosis and recycling rates. We focus mostly on class II MHC molecules but most methods would apply equally well, and have in fact been used, to study class I MHC molecules. Only external labelling methods and not biosynthetic methods are presented here since there is evidence that a proportion of newly synthesized MHC never reaches the cell surface, and so half-lives of such material are likely to be a composite of internal turnover as well as loss from the cell surface.

2. Surface labelling of MHC molecules Radioiodination is the method of choice for labelling cell surfaces since it is inexpensive, produces labelled proteins of high specific activity and the isotope (125I, t1/2= 60 days) is easily detected without the use of scintillants. Labelling can be achieved on both tyrosine or amino groups by different methods.

2.1 Lactoperoxidase-catalysed labelling on tyrosine residues In the presence of 125I and H2O2 (usually generated continuously by the action of glucose oxidase on glucose) lactoperoxidase (EC 1.11.1.7) can catalyse the incorporation of 125I into tyrosine residues to produce mostly mono-iodotyrosine (3). The method we have utilized, based on the that of Hubbard and Cohn (4), is given in Protocol 1. Protocol 1. Labelling of cell surface tyrosine residues with lactoperoxidase Method 1. Harvest the cells (2 x 107, 4 min, 1500 r.p.m.), wash twice with PBS/ 5 mg ml-1 BSA, then wash twice with PBS. 2. On ice in a volume of 250 uJ prepare a mix containing 40 p-gmT-1 lactoperoxidase (Sigma, L8257), 200 ngmT-1 glucose oxidase (type V-S, Sigma, G6891), 20 mM B-D( + ) glucose (Sigma, G5250) and 2 mCimr-1 125l-iodine (Amersham, IMS 30).a 3. Add the labelling mix to the cells in PBS and allow reaction to proceed for 30 min. 4. Wash cells three times with PBS/1 mM Kl and then once with PBS/ 5 mgmr-1 BSA before aliquoting as appropriate. a125 l is a powerful gamma-emitting radionuclide and hence appropriate care must be exercised in its use. Wear disposable gloves at all times, perform iodination in a ventilated fume cupboard using a lead screen (e.g. Scotlab, SL-1104), use a sensitive monitoring device, and wear a radiation film badge.

18

2: Labelling methods for analysis of class II MHC endocytosis Experiments performed using this protocol showed that while we were able to get labelling of the MHC glycoproteins with lactoperoxidase, higher specific activities were obtained if we used reagents which covalently reacted with free amino groups on the cell surface.

2.2 Labelling of amino groups by succinimidyl ester based reagents Two basic types of reagents can be utilized in studies of the MHC glycoproteins (Figure 1). Basically they consist of reactive JV-succinimidyl esters linked via spacers to a phenolic ring which can be labelled with 125I (see below). Reagent (a) is sulpho-SHPP, a commercial reagent which permits labelling of a high specific activity (5) (Protocol 2). Reagents of type (b) have been developed by M.S. Bretscher at the Laboratory for Molecular Biology, Cambridge, for endocytosis studies (6, 7) and differ from type (a) in that they have one or more disulphide bonds incorporated in their spacers. Selective removal of the cell surface bound reagent can be achieved with, for example, membrane impermeant reducing agents such as glutathione (6, 7). Cleavage of the disulphide bond

Figure 1. Non-cleavable and cleavable cell surface labelling reagents. Both types are N-hydroxysuccinimide ester-based reagents which are iodinated prior to cell surface or protein labelling, (a) sulpho-SHPP (5) (b) DPSgt/gtc (6).

19

Pamela A. Reid and Colin Watts causes selective removal of the 125I-tyrosine labelled portion of the reagent so that following immunoprecipitation of the molecule of interest and SDS polyacrylamide gel analysis these molecules will no longer be visible by autoradiography. Three gel reagents have been developed (6,7), the most simple of which, 3,3'dithiopropionyl, 1 sulphosuccinimidyl, 1'glycyl-tyrosine (DPSgt) can be prepared from commercially available components (see Protocol2) and used without further purification. Reagents DPSgt and DPSgtc (which differs from DPSgt in that it also contains a cholamine residue) contain one disulphide bond whilst DPSch contains two disulphides (7).

Protocol 2. Preparation of non-cleavable cell surface labelling reagent sulpho-SHPP3 and of the cleavable cell surface labelling reagents DPSgt and DPSgtcb A. Sulpho-SHPP 1. Prepare a 0.37 mgmr-1 solution of sulphosuccinimidyl-3-(4-hydroxyphenyl) propionate (sulpho-SHPP, Pierce and Warriner, 27712) in DMSO. B. DPSgt 1. Prepare solutions of 0.1 M glycyl-tyrosine, 1 M sodium phosphate pH 7 and 0.05 M acetic acid 2. At 4°C to 20 ul 0.1 M glycyl -tyrosine and 40 ul 1 M sodium phosphate pH 7 add 2.43 mg 3,3'-dithiobis(sulphosuccinimidyl) propionate (DTSSP, Pierce and Warriner, 21577) in 40 ul water. 3. Vortex the mixture and leave for 15 min on ice. 4. Add 2.3 ml ice-cold 0.05 M acetic acid and store at -70°C in small aliquots (approximately 25 ul). C. DPSgtc 1. Prepare 0.1 M glycyl-tyrosyl-cholamine chloride (gtc, ref. 6), 0.025 M disodium hydrogen phosphate and 0.05 M acetic acid. 2. At 4°C to 20 ul 0.1 M gtc and 160 ul 0.0125 M disodium hydrogen phosphate add 2.43 mg DTSSP in 60 ul water. 3. Vortex the mix and leave 15 min on ice. 4. Add 2.3 ml ice cold 0.05 M acetic acid and store at -70°C in small aliquots (approximately 25 |xl). ' Unlabelled reagent stable for at least one month at -70°C "Reagents prepared as described in ref. 6 are stable for at least 8 months at -70°C. It is recommended that the small stored aliquots of the reagents are used only once and not subjected to repeated freezing and thawing.

20

2: Labelling methods for analysis of class IIMHC endocytosis The labelling reaction with 125I proceeds by a two-step process to produce proteins of high specific activity. Initially, the N-succinimidyl 3-(4-hydroxyphenyl) propionate is radioiodinated using chloramine T to form the Bolton and Hunter reagent (10). The W-succinimidyl group of the reagent then reacts with free amino groups in the proteins, e.g. the epsilon amino group of lysine residues, to form a conjugate in which the radioiodinated phenol group of the reagent is covalently linked to the protein by an amide bond. The main disadvantage of this technique is the fact that the labelled reagents will eventually be hydrolysed in aqueous solution. Therefore the reaction must be carried out in a small volume and as quickly as possible under alkaline conditions to minimize hydrolysis (see Protocols 3 and 4).

Protocol 3. Preparation of 125l-labelled reagents Method 1. On ice prepare the labelling mixture containing 10 ul water, 10 ul iodine-125 (1 mCi), 1 ul 3 M NaCI, 2 ul 1 M sodium phosphate pH 7, 1 ul reagent see and 1 ul 5 ml-1 chloramine T (Sigma, C9887). Retain the mix on ice for 15 min. The chloramine T solution should be prepared in distilled water immediately before use. 2. Add 2 (ul 1 M sodium p-hydroxybenzoate (Sigma, H3766) containing 0.1 M sodium iodide (Sigma, S8379) and leave on ice for a further 4 min before use. (The purpose of this step is to destroy excess chloramine T and to chase all reactive species into inert derivatives of hydroxybenzoate and prevent further labelling reactions occurring). The solution of 1 M sodium p-hydroxybenzoate / 0.1 M sodium iodide can be stored frozen and used repeatedly.

Protocol 4. Radiolabelling cells with the 125l-reagentsa/b Method 1. Collect suspension cells by centrifugation for 4 min at 1500 r.p.m. Remove adherent cells from tissue culture plastic using a 5-10 min incubation at 37 °C in a pre-warmed solution of PBS containing 5 mM EDTA. 2. Wash all cells twice with PBS. 3. Resuspend the cells in 50 ul 0.1 M Na2HPO4 until 125l-labelled reagent has been prepared. Immediately prior to labelling (at the stage where

21

Pamela A. Reid and Colin Watts Protocol 4. Continued the 1 M sodium p-hydroxybenzoate / 0.1 M sodium iodide is add to the 125I labelled reagent) spin down the cells and resuspend in a fresh aliquot of 50 (ul 0.1 M Na2HP04. 4. Add cells to labelled reagent (density 108-109 cells ml-1) so that they are resuspended in a final volume of 80-100 ul. 5. Label the cells on ice for 20 min then wash three times with PBS containing 10% fetal calf serum or 1% BSA. "Prepare 125I labelled reagents as described in Protocols. 6 For some cell types the inclusion of Ca2+ and Mg2t in all PBS buffers is advantageous.

2.3 Methods utilized for study of the endocytosis of class I and II glycoproteins in B-lymphoblastoid cells Several methods have been applied to study of the endocytosis of the MHC glycoproteins. These include the use of 125I antibodies or Fab fragments (11-13) or fluorescent antibodies (14). Other researchers have used lactoperoxidase to tag the cell surface population and study its subsequent fate on warming the cells to 37 °C (15). A further modification of the lactoperoxidase labelling protocol involves treating the 125I labelled surface population with neuraminidase. In this approach the terminal sialic acid residues of cell surface MHC glycoproteins are removed so these proteins would show a characteristic shift on two-dimensional SDS-PAGE. In contrast, molecules which have been endocytosed would be protected from desialylation by the impermeant neuraminidase. Using this method two groups studied the endocytosis of class I and II MHC glycoproteins (16,17). In all the studies above it was concluded that there was little endocytosis of the MHC glycoproteins. However, as these methods might not detect small endocytosed pools we employed the cleavable reagents described in Section 2.2 to see whether there was any internalization of the MHC glycoproteins from the cell surface. Figure 2 outlines the protocol for measurement of endocytosis and recycling using these reagents. The three available reagents were characterized (a) for incorporation of 125I into the glycoproteins and (b) for the ability to remove the 125I component of the reagent with reduced glutathione at 4 °C. The reagents differ both in their net charges and the number of disulphide bonds they contain. For example, unlike DPSgt and DPSgtc, DPSch contains two disulphide bonds between the reactive ester group and the tyrosine labelled with 125I. This property of the reagent should in theory lead to a lower background remaining after treatment with reduced glutathione. Additionally, DPSgtc carries a fixed positive charge since it contains a cholamine residue and might be expected to be less permeant (6, 7). 22

Figure 2. Endocytosis and recycling detected using a cleavable label. Experimental scheme shows sequential cell surface labelling at 0°C, incubation at 37°C to allow endocytosis to occur, removal of remaining cell surface label with reducing agent (Glutathione 1°), re-culture to allow recycling of the intracellular pool, and quantitation of this by further reduction {Glutathione 2°).

Pamela A. Reid and Colin Watts All three reagents could effectively incorporate I2S1 into class I and II MIIC and in most cases the l25I labelled portion of the reagent could he cleaved with reduced glutathione (> 90% removal). However, in the case of the pj-micnjgloblin chain of class I MHC the label was inefficiently removed (66-77%) and therefore study of the intraceltular pools of this particular molecule would be difficult to assess by this method.

2.4 Study of the endocytosis of class I and II MHC glycoproteins in B-lymphoblastoid cells using cleavable cell surface labelling reagents The preliminary studies described in Section 2.3 suggested that although all three reagents could be used tor studies on the endocytosis of the MHC glycoproteins, DPSgtc was the most efficient. EBV-lransformed B-lymphoblastoid A46 cells (18) were labelled with 125I-DPSglc as described in Protocol 4 and then altquots were warmed to 37 °C for various times before the cells were chilled and treated with reduced glulathione (see Protocol 5 and Figure .?). The class I and II MHC glycoproteins were sequentially immunoprecipitated

Figure 3. Small but detectable pools of endocytosed class II MHC molecules exist on human lymphoblastoid cells. Efficient endocytosis of the transferrin receptor produces a large intracellular pool [track 3}. MHC molecules are endocytosed less efficiently such that at steady state approximately 7% of class II (track 6) and 3% (track 9) of class I MHC can be detected intracellularly.

24

2: Labelling methods for analysis of class II MHC endocytosis with specific antibodies, recovered using protein A-Sepharose and the labelled molecules resolved on either non-reducing 15% (class II) or 17.5% SDS polyacrylamide gels (class I) and the dried gels autoradiographed (Figure 3). It is important to use non-reducing conditions in these experiments, otherwise the 125I-labelled portion of the molecule will undergo reduction and will not be visible by autoradiography. Quantitation of the proportion of labelled molecules which had been internalized was performed following removal of the counts from an aliquot of cells which had been held on ice and then treated with reduced glutathione (see Figure 3). Protocol 5. Measurement of the endocytosis of the class I and II MHC glycoproteins Method 1. Label cells with 125l-DPSgtc as described in Protocols 2 to 4. 2. Aliquot washed cells in RPMI 1640/20 mM Hepes pH 7.5 containing 10% FCS into 10 ml conical tubes. (Use 5 x 106 B cells per experimental point.) 3. Retain two aliquots of cells on ice and warm other aliquots to 37°C in the presence or absence of primaquine (see Protocol 6) for various times. Stop the 37°C incubation rapidly by adding 10 volumes of chilled RPMI/Hepes/FCS. 4. Treat all cell aliquots except one of the unwarmed aliquots with 46 mM reduced glutathione (free acid form, Sigma, G4251) for 30 min on ice. Resuspend cells (1.6 x 106 cells/ml) in 3 ml of the glutathione buffer, containing 45.5 mg glutathione, 2.64 ml water, 44 ul 5 M NaCI, 22 ul 10 M NaOH and 0.293 ml FCS. 5. Wash the cells twice with PBS / 5 mg ml-1 BSA to remove unreacted glutathione. 6. Treat the cells for 10 min on ice with 0.1 M NaCI, 0.05 M Tris pH 7.4, 5 mM MgCI2, 1 mM CaCI2, 1 mM PMSF, 5 mgmr1 iodoacetamide and 40 ug ml-1 soya bean trypsin inhibitor (type 1-S, Sigma, T9003), then lyse for 1 h on ice by the addition of 1.67% NP-40. Centrifuge the lysates at 13000 x g for 15 min and centrifuge the supernatant further in a Beckman airfuge at 100000 x g for 30 min. 7. Pre-clear cell lysates for 2 h at 4°C using either protein A-Sepharose or Pansorbin and recover the class I or II MHC glycoproteins by sequential immunoprecipitation at 4°C with appropriate primary antibodies8 for 1 h, with affinity purified rabbit anti-mouse immunoglobulin (Serotec) overnight and then with protein A Sepharose for 2h.

25

Pamela A. Reid and Colin Watts Protocol 5. Continued 8. Wash protein A beads once with 0.15 M NaCI, 0.05 M Tris pH 7.4, 1 mM PMSF, 1 mgmr1 iodoacetamide and 1% NP-40 and three times with this buffer diluted 10-fold, then elute immune complexes using non-reducing SDS sample buffer. "For immunoprecipitation studies monoclonal antibodies DA6.231 or DA6.147 which are directed against class II MHC (23) and for class I MHC W6/32 (24) were used.

Our results (8, 9) suggested that following a 30 min incubation at 37 °C, 7±1.8 and 3±1.1% (s.d., n=7) respectively of the labelled cell surface population of class II and I MHC became resistant to glutathione reduction following endocytosis into the cells. These small intracellular pools could be detected using this sensitive assay but may explain why other authors concluded that there was no internalization of the MHC glycoproteins in B Cells (16,17). However, this small intracellular pool size made measurement of the initial rates of endocytosis from the cell surface difficult, as equilibration was reached within 3 to 4 min of warming to 37 °C. These problems were overcome if the cells were warmed in the presence of primaquine.

2.5 Endocytosis of the MHC glycoproteins in the presence of primaquine Primaquine, a lysosomotrophic amine has been shown to slow the recycling of the transferrin and asialoglycoprotein receptors back to the plasma membrane thus increasing their intracellular pool size (19). We investigated the effect of primaquine (prepared as described in Protocol 6) on the intracellular pool size of the MHC glycoproteins in A46 cells (8) and found that using 300 u-M primaquine 23.6 ± 4.2% of class II and 19 ± 3.6% of class I MHC was protected from glutathione reduction. These larger intracellular pools took between 10 and 15 min to reach equilibrium, making it possible to measure the initial rates of endocytosis from the cell surface. On warming the B cells to 37 °C in the presence of 300 uM primaquine (8), class II MHC was internalized at 2.1 ±1.0% and class I MHC at 1.7±0.4% per minute (8). Protocol 6. Preparation of stock solution of primaquine Method 1. Prepare primaquine (diphosphate salt, Sigma, P9504) as a 3 mM stock in RPMI 1640 / 20 mM Hepes. 2. Adjust the final pH to 7.6 before the addition of 10% PCS. Prepare fresh stock solutions for each experiment. 26

2: Labelling methods for analysis of class II MHC endocytosis

2.6 Measurement of the recycling kinetics of endocytosed class I and IIMHC glycoproteins As described in Section 2.5, class I and II MHC glycoproteins were endocytosed from the cell surface at 1-2% per minute and the intracellular pools were approximately 3% and 7% respectively. To maintain these small intracellular pools, the molecules must be efficiently recycled back to the cell surface following endocytosis. B cells were labelled with 125I-DPSgtc and recycling studied as described in Protocol 7 (see also Figure 2). The rapid recycling of both class I and II MHC was detected both in the presence and absence of primaquine (8, 9) and calculations suggested that the half-lives of the intracellular pools were 2-3 min (8). Therefore cleavable cell surface labelling reagents can be used to demonstrate that class I and II MHC glycoproteins are constitutively endocytosed and recycled in B-lymphoblastoid cells (8, 9). Protocol 7. Measurement of recycling of endocytosed class I and II MHC glycoproteins to the cell surface Method 1. Label B-lymphoblastoid cells with 125l-DPSgtc as described in Protocol 4. 2. Wash the labelled cells, warm to 37°C for 30-60 min to allow endocytosis, chill to 4°C, and then treat with reduced glutathionine as described in Protocol 5. 3. Hold aliquots of the cells on ice or return to 37°C for up to 10 min to permit recycling. 4. After rewarming, rapidly chill the cells to 4°C and then treat for a second time with reduced glutathione (Protocol 5). 5. Wash cells with PBS / 5 mgmr1 BSA, lyse in NP-40 lysis buffer containing iodoacetamide and treat further as described in Protocol 5.

2.7 Study of the antigenic peptides associated with class II MHC glycoproteins SDS polyacrylamide gel analysis of immunoprecipitated class II molecules revealed 125I label running at the dye front (Figure 3). We reasoned that small reagents specific for ammo groups should be able to label the amino terminus or any available lysine residue of peptides found in the binding cleft of MHC class II, particularly if these peptides extend out of the groove. A twodimensional gel analysis protocol was devised to study the association of 125Ilabelled peptides with class II MHC molecules (see Figure 4). The aim was to

27

Figure 4. Two-dimensional analysis of SDS stable and unstable cell surface class II MHC glycoproteins. The peptide pools bound to SDSstable and SDS-unstable molecules migrate in two different positions at the dye front. Those from SDS-stable complexes migrate at position P whilst the latter migrate at the end of the diagonal (P', possibly together with other material. Lifetimes on living cells of both a and p chains as welf as peptide pools can be measured.

2: Labelling methods for analysis of class IIMHC endocytosis Table 1. Composition of 15% SDS polyacrylamide gels used for class II MHC analysisa

Resolving gel AnalaR urea 40% acrylamide 1% bis-acrylamide 2 M Tris pH8.8 10% SDS 10% ammonium persulphate TEMED water

15% gel (30ml)

15%+8M urea (50ml)

none 11.25ml 2.6ml 5.625 ml 300 nl 200 |il 20 |il 10ml

24 g 18.75 ml 4.34 ml 9.375 ml 500 |il 166.9 |il 33.4 ul none

Stacking gel (15ml) 30% acrylamide/0.8% bis-acrylamide 1 M Tris-HCI pH6.8 10% SDS 10% ammonium persulphate TEMED water

2.5ml 1.875ml 150 ul 75 ul 15 ul 10.385 ml

a

The volumes shown are used to form slab gels of 14.5 x 19 x 0.1cm (0.2cm for second dimension urea gels). When preparing urea gels add solid urea to bottom of beaker, add acrylamide, bisacrylamide, SDS, and Tris-HCI. Heat to dissolve the urea, then immediately before pouring add the ammonium persulphate and TEMED. To prevent the instant setting of the stacking gel the urea was omitted from this recipe.

isolate stable class II afi complexes in the first dimension and then to subsequently dissociate these complexes and study the properties of any peptides bound to class II MHC. Since it had been observed that the a and p chains of class II MHC glycoproteins would stay associated in sodium dodecyl sulphate (SDS) unless the samples were boiled (20,21), stable ap complexes were isolated in the first dimension using a modified sample buffer containing 1% SDS and immunoprecipitated class II MHC eluted at room temperature (Protocol 8) prior to running on a 15% SDS polyacrylamide gel (see Figure 4, Table 1). The gel slice was excised, treated at 37 °C for 30 min in 0.1% SDS / 8 M urea, then further electrophoresed horizontally in the 2nd dimension on a 15% gel containing 8 M urea (see Table 1) . As shown in Figure 4, both class II molecules and some small molecular weight peptide material running at the gel front could be detected using this protocol. Protocol 8. Labelling and turnover of MHC class II and bound peptides Method 1. Label B-lymphoblastoid cells with either 125l-DPSgtc or 125l-sulphoSHPP (Protocol 4), wash with RPMI/bicarbonate/10% PCS and then either reculture in this medium at 37°C in a C02 gassed incubator or

29

Pamela A. Reid and Colin Watts Protocol 8. Continued lyse with NP-40 containing buffer. Process for immunoprecipitation as described in Protocols. a2. Remove recultured cells from the incubator at various times after

labelling and wash, lyse and immunoprecipitate as described in Protocol 5. 3. Wash the immune complexes once with 0.15 M NaCI, 0.05 M Tris pH 7.4,1 mM PMSF, 1 mg ml-1 iodoacetamide, and 1% NP-40, and then three times with this buffer diluted 10-fold. Then elute from the beads using a modified sample buffer consisting of 10% glycerol, 0.125 M Tris-HCI pH6.8, 1% SDS and 0.01% bromophenol blue. Do not boil the samples but leave for 30 min at room temperature before eluting through a small hole pierced in the bottom of the tube with a needle (25G x 5/8 inch) and by centrifugation in a bench-top centrifuge. 4. Load the samples on a 15% SDS Polyacrylamide gel prepared as described in Table 1 and run the gel so that the bromophenol tracking dye migrates only partially into the gel. It is important that the first dimension gel piece is 1-2 cm shorter than the horizontal width between the teeth of the second dimension comb, since some swelling occurs in the urea containing equilibration buffer (see below).6 5. Carefully excise the track/piece of gel containing the molecules of interest. This is most easily done by exposing the wet gel to autoradiographic film at -70°C overnight and using the film as a template for the excision. 6. Equilibrate the gel slice at 37°C for 30 min in 0.125 M Tris-HCI pH 6.8, 0.1% SDS, and 8 M urea (BDH, AnalaR, 10290). 7. Load the gel slice on to the top of a 15% SDS polyacrylamide gel containing 8 M urea (see Table 1) 8. Run the gel at 200 V for approximately 5-6 h, fix for 10 min with 45% methanol, 10% acetic acid, and then dry down without further attempts to remove the methanol. This protocol minimizes diffusion of the small peptide material from the gel. 9. Using the autoradiographic film as a template, excise the pieces of gel corresponding to the class II «p complexes, a, p chains and peptide spots and count in the gamma counter. 'To facilitate reculturing of 125l-labelled cells, filter sterilize all labelling mix components (Protocol 3), except 125I, prior to use. ''For ease of use the spacers used in the second dimension gel should be at least 0.5 mm thicker than those used for the first dimension gel.

30

2: Labelling methods for analysis of class IIMHC endocytosis

2.8 Analysis of the turnover time of class II MHC and its associated peptides As described above, the use of a two-dimensional gel protocol enabled the peptides associated with class II MHC to be isolated. To see whether these peptides were turned over faster or at the same rate as class II molecules themselves, we analysed SDS-stable a3 complexes at various times following labelling on ice with 125I-sulpho-SHPP. The cells were labelled, washed with sterile RPMI 1640/10% PCS and then recultured in the incubator for up to 36 h. At various times the cells were harvested, lysed and their class II MHC molecules immunoprecipitated under conditions designed to retain aB/peptide complexes together (see above and Protocol 8) and analysed by the twodimensional system (see Protocol 8 and Figure 4). In EBV cells, we found that the peptides were stably bound to the aB chains and persisted for the same length of time that the class II molecules themselves. The half-life of the class II molecules and bound peptides on these cells was 25-30 h (22). Thus in spite of the peripheral cycling of class II molecules, this does not seem to be accompanied by peptide turnover. The implications of this result are discussed elsewhere (22).

Acknowledgements This work was supported by the Medical Research Council and The Wellcome Trust. We thank Dr Mark S. Bretscher, MRC, Cambridge for generous provision of his reagents and for many helpful discussions.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Germain, R.N. and Margulies, D.H. (1992). Ann. Rev. ImmunoL, 11,403. Lanzavecchia, A. (1990). Ann. Rev. ImmunoL, 8,773. Morrison, M. and Bayse, G.S. (1970). Biochemistry, 9,2995. Hubbard, A.L. and Cohn, Z.A. (1975). J. Cell BioL, 64,438. Thompson, J.A., Lau, A.L. and Cunningham, D.D. (1987). Biochemistry, 26,743. Bretscher, M.S. and Lutter, R. (1988). EMBO J., 7,4087. Bretscher, M.S. (1989). EMBOJ.,», 1341. Reid, P.A. and Watts, C. (1990). Nature, 346,655. Reid, P.A. and Watts, C. (1992). Immunology, 77, 539. Bolton, A.E. and Hunter, W.M. (1973). Biochem. J., 133,529. Guagliardi, L.E., Koppelman, B., Blum, J.S., Marks, M.S.,Cresswell, P and Brodsky, P.M. (1990). Nature, 343,133. 12. Capps, G.G., Van Karnpen, M., Ward, C.L. and Zuniga, M.C. (1989). J. Cell BioL, 108,1317. 13. Harding, C.V. and Unanue, E.R. (1989). J. ImmunoL, 142,12. 14. Pernis, B. (1988). In Processing and presentation of antigens (ed. B.Pernis, S.C. Silverstein and H.J. Vogel), pp. 247-259. Academic Press, London. 31

Pamela A. Reid and Colin Watts 15. Salamero, J., Humbert, M., Cosson, P. and Davoust, J. (1990). EMBO J., 9,3489. 16. Neefjes, J.J., Stollorz, V., Peters, P.J., Geuze, H.J., and Ploegh, H.L.(1990). Cell, 61,171. 17. Davis, J.E. and Cresswell, P. (1990). J. ImmunoL, 144,990. 18. Lanzavecchia, A. (1985). Nature, 314, 537. 19. Schwartz, A.L., Bolognesi, A., and Fridovich, S.E. (1984). J. Cell Biol, 98,732. 20. Cresswell, P. (1977). Eur. J. ImmunoL, 7,636. 21. Springer, T.A., Kaufman, J.F., Siddoway, L.A., Mann, D.L.and Strominger, J.L. (1977). J. Biol. Chem., 252,6201. 22. Lanzavecchia, A., Reid, P.A., and Watts, C. (1992). Nature, 357,249. 23. Guy, K., van Heyningen, V., Cohen, B.B., Deane, D.L., and Steel, C.M. (1982). Eur. J. ImmunoL, 12, 942. 24. Barnstable, C.J., Bodmer, W.F., Brown, G., Galfre, G., Milstein, C., Williams, A.F., and Ziegler, A. (1978). Cell, 4,9.

32

3

Methods for detecting signalling via MHC class II molecules which can lead to either activation or programmed cell death N. MOONEY, J. P. TRUMAN and D. CHARRON

1. Introduction The transmission of signals by MHC class II molecules has been extensively examined. A number of studies have demonstrated either an augmentation or a decrease in cell proliferation (assessed by 3H-thymidine incorporation) after treatment of B lymphocytes with anti-MHC class II monoclonal antibodies (mAbs). Initial studies in the mouse described the activation and nuclear translocation of protein kinase C (PKC) and the production of cyclic adenosine monophosphate (cAMP) (1) after stimulation of B lymphocytes with anti-MHC antibodies. Later studies in humans revealed the activation of phospholipase C (PLC), increased intracellular calcium [Ca ++ ]j and production of diacylglycerol (DAG) (2, 3, 4). Interestingly, the involvement of PKC in signalling via HLA class II antigens did not necessarily involve translocation, since increased membrane and cytosol associated PKC activity was detected as well as increased levels of PKC a and PKC 3 mRNA (5, 6). A later study revealed that PKC a, 3 and 8 are all increased at the mRNA level following signalling via HLA class II (7). Signals via MHC class II influence proliferative responses to polyclonal activators of lymphocytes (PWM, PMA, anti-IgM, PHA, anti-CD3) as well as antigen presentation (8, 9). A novel feature of HLA class II mediated signalling has recently been described, that of induction of programmed cell death hi either resting (10) or activated (11) B lymphocytes. Some of the second messengers and signalling pathways have been described in mice and some in humans, and results which apply to one species do not necessarily apply to both. For example, increased levels of cAMP are induced via MHC class II in the mouse but have not been detected in humans. It should be remembered that the pattern of expression of MHC

N. Mooney, J. P. Truman and D. Charron class II antigens is quite different in the two species. For example, murine T lymphocytes do not express class II antigens even after activation. A further difference may arise from the use of allospecific mAbs in the mouse and heterospecific mAbs in the human.

2. Changes in intracellular calcium levels in signal transduction The measurement of intracellular calcium is of interest since increases can regulate diverse cellular processes. The pathway which has been studied using anti-HLA class II mAbs involves triggering via HLA-DR molecules leading to activation of phospholipase C. This activation results in hydrolysis of a membrane phospholipid phosphatidylinositol 4,5 biphosphate (PIPi) which provides inositol 1,4,5-triphosphate (IP3) and the lipid DAG. IP3 causes calcium release from intracellular stores and is required for the activation of the calcium dependent isoforms of PKC. Both mAbs and bacterial superantigens have been used to induce intracellular calcium fluctuations via HLA class II antigens (2,3,12). Lymphocytes are readily loaded with Indo-1, which permits detection of intracellular calcium by flow cytometry (Protocol 1), although other cell types may be less easily loaded due to compartmentalization of the Indo-1. More efficient loading is readily obtained by treating the cells with a low concentration of pluronic detergent (Pluronic F-127, Molecular Probes, P-1572). The optimal concentration may vary according to the cells under investigation but a final concentration of 0.01-0.05% pluronic detergent during the incubation with Indo-1 will improve uptake. Frozen cells should always be thawed on the eve of the experiment for maximum Indo-1 uptake. Pre-incubation of the cells with the anti-HLA class II mAb before addition of the cross-linker can also magnify the response. Typically one calcium measurement takes ten minutes, the first two minutes to measure the first baseline (cells alone), the next two minutes to measure the next baseline (cells and anti-HLA class II mAb) and the remaining six minutes to measure the intracellular calcium after addition of the cross-linking mAb. Protocol 1. Measurement of intracellular calcium fluxes Equipment and reagents • Flow cytometer with a UV light source and a heated sample chamber

• INDO-1 (molecular probes) . F(ab)2 rabbit anti-mouse mAb (Cappel)

A. Loading of cells with Indo-1 pentaacetoxymethyl ester (Indo-1 AM) 1. Wash the cellsa twice with complete mediumb and resuspend at a concentration of 2 x 107 cells/ml. 34

3: Methods for detecting signalling via MHC class II molecules 2. Resuspend the lndo-1c and add to the cells at a final concentration of 3 MM. 3. Incubate for 30 min in a 37°C water bath. 4. Wash the cells twice with complete medium and resuspend at 10 x 106 cells/ml and protect from light during the entire experiment. B. Verification of flow cytometer 1. Adjust the flow cytometer. The ratio of lndo-1 violet to blue will be the final measure. Use a violet bandpass filter centred at 385 ± 10 nm and a blue bandpass filter centred at 500 ± 20 nm. Set the light scatter gates and optimize the photomultiplier tube gain settings. Use linear rather than logarithmic measures and gate out the dead cells. 2. Check that the instrument is appropriately set by testing with ionomycin at a final concentration of 2 ug ml"1; the entire population of cells should respond rapidly. Thoroughly wash the loading chamber of the instrument in order to remove all traces of ionomycin before the tests. C. Analysis of intracellular calcium flux 1. Warm cells for 5-10 min at 37°C before analysis. Measure the basal level of calcium in unstimulated cells for at least 2 min preceding addition of monoclonal antibodies. 2. Anti-HLA class II mAbs require cross-linking before a calcium flux is detected. Another baseline therefore has to be established after addition of the anti-HLA class II antibody (10-50 ug mT1) and before addition of an F(ab)2 fragment of a rabbit anti-mouse mAb (10-30 ugmT1, Cappel). 4. Determine optimal doses of both the cross-linking mAb and of the anti-HLA class II mAb to ensure that the cross-linking mAb alone does not induce a calcium flux. 5. After addition of the cross-linking mAb to the cell suspension containing the anti-HLA class II mAb, continue measurement of the violet:blue ratio for at least 5 min in order to establish a clear maximum. "Thaw frozen cells on the eve of the experiment and store in complete medium overnight. ''Complete medium: RPMI 1640 (Flow laboratories) supplemented with 10% heat inactivated fetal calf serum (FCS), penicillin and streptomycin (100 ugml-1), pyruvate (1 mM), and glutamine (2 mM). c lndo-1 AM (Molecular Probes 1-1223) is diluted with DMSO to a final concentration of 1 mM before aliquoting (50 ulAube) in 0.5 ml Eppendorf tubes. The aliquots are sealed with parafilm and stored at -80°C in the presence of a desiccant.

This protocol can be used to measure intracellular calcium rather than to determine whether or not an intracellular calcium flux is induced, by resuspending cells in buffers containing known free calcium concentrations which can be used to standardize Indo-1 fluorescence ratio changes (13). Two 35

N. Mooney, J. P. Truman and D. Charron buffers are required, one containing an equimolar solution of calcium and the calcium chelator EGTA and an identical buffer lacking calcium, which are mixed in order to give the desired calcium concentrations.

3. Inositol phospholipid turnover during lymphocyte activation The generation of diacylglycerol and inositol phosphates is believed to be mediated by phospholipase C (PLC). The receptor-mediated activation of PLC may therefore be measured by quantifying the production of inositol phosphates (InsPj, InsP2, and InsP3) (Protocol 2). Both anti-HLA class II DR mAb and bacterial superantigens activate this pathway. Cells are labelled with myo-[2-3H] Inositol overnight and tritiated InsP1; InsP2 and InsP3 can be separated on Dowex columns using increasing concentrations of ammonium formate in 0.1 M formic acid before quantifying by liquid scintillation counting. The efficiency of this method depends largely on the uptake of myo-[2-3H] Inositol and depends on relatively high numbers of cells in each experiment to obtain clear and reproducible results. It is imperative that the tests are carried out at least in duplicate and that the count rate (c.p.m.) in the total lipid fraction is measured in every experiment. The total level of IPs in lymphocytes is very low (between 0.1 and 5% of the total lipid fraction) so the total lipid fraction should always be measured and the results considered in relation to this quantity. The actual count rate cannot be used as an absolute measure due to the variability of uptake from one experiment to another as well as variability from one cell type to another. Protocol 2.

Detection of PLC activation

Equipment and reagents PIPES
. Buffer 4:0.4 M ammonium formate / 0.1 M formic acid Buffer 5:1 M ammonium formate / 0.1 M formic acid i Calcium-free Tyrode's solution (prepare on the eve of the experiment): 20 ml of 1 M PIPES (0.02 M final), 50 ml of NaCI / KCI (0.012 M / 0.002 M final), 1 g glucose, 1 g Bovine serum albumin. Make up to 1000 ml at pH 6.8 with NaOH Staphylococcus aureus (Sigma, 52014) > Scintillation fluid (Baker)

Method 1. Prepare the columns on the eve of the experiment. Wash the Dowex 1-X8 resin five times in a large volume of water, leaving sufficient water from the final wash to provide a slurry that is possible to 36

3: Methods for detecting signalling via MHC class II molecules

2.

3.

4.

5.

6.

7.

8.

pipette. Mark the level of 0.8 ml on each column and break off the bottom of each column so that liquid can drain through. Use a 5 ml pipette to add the resin up to the 0.8 ml mark on each column. Wash lymphocytes in medium lacking inositol (e.g. Hanks solution buffered with 25 mM hepes to pH 7.4) and resuspend at 5 x I06ml-1. Add tritiated myo-inositol to a final concentration of 15-20 uCiml-1 and incubate the cells overnight in a humidified incubator at 37 °C with 5% C02. Wash the cells three times in calcium free Tyrode's solution (CFT) and resuspend at a concentration of 107mr-1 in CFT containing 10 mM lithium chloride before incubating for 15 min at 37°C in a water bath. Aliquot the anti-HLA class II mAb (e.g. to final concentrations of 0, 10, 20, 40, and 80 ugml-1) in a suitably small volume of CFT (e.g. 100 ul) in a 2 ml Eppendorf tube (add CFT to maintain a constant volume). Add an equal volume of the tritiated cells. (For a transformed cell line a cell suspension of 2 x 106ml-1 is adequate while up to 5x 106mr-1 may be required for resting human lymphocytes.) Incubate the cells with the mAb for 15 min and stop each test by adding 750 ul of an ice cold solution of chloroform/methanol (1:2). Immediately vortex the sample and place it on ice. Separate the aqueous from the organic phase by the sequential addition of 250 ul H2O and 250 ul of chloroform and vortex after each addition. Microfuge the mixture for 5 min in a pre-cooled Eppendorf at 12000 x g. Collect the aqueous (upper) phase with a Pasteur pipette and dilute immediately in 5 ml of water. Add the diluted aqueous phase to the Dowex column and wash twice with 3 ml of H2O. If the different IPs are to be collected separately, take a 1 ml sample of the eluate to verify that the count (c.p.m.) is at background level. Then sequentially wash the columns in order to remove the glycerophosphates (Buffer 2) followed by IP1 IP2 and IP3 using 3 ml of Buffers 3, 4, and 5 respectively. The fractions corresponding to IP 1 IP 2 and IP3 can usually be eluted using 3 ml of each buffer but if separation of each IP is required, the background counts should be verified between addition of each buffer and the eluates should be collected into separate tubes for each IP. Dry the samples overnight before addition of scintillation fluid and counting. A sample of non-treated cells should be included in each series of tests, and SAC 10-4 v/v also provides a good positive control for B lymphocytes in order to ensure reproducibility, while PHA provides a good positive control for T lymphocytes. Carry out a time course to optimize the incubation time with a given activator (10 min incubation of B lymphocytes with an anti-HLA class II mAb was optimal). 37

N. Mooney, J. P. Truman and D. Charron

4. Signal transduction by kinases and phosphatases The state of phosphorylation of a given cell is controlled by the action of a variety of kinases and phosphatases. These kinases and phosphatases are major contributors to the transmission of signals via various pathways. Before measuring the activity of a given kinase it is a useful first step simply to establish whether or not there is a difference in the overall state of phosphorylation after fixation of a ligand. Another preliminary step would be the use of inhibitors of either PKC or PKA or TPK. However, interpretation of the data obtained with various inhibitors may be difficult because of the possibility of interactions occurring between signalling pathways involving different kinases and because of the lack of absolute specificity of the inhibitors. With regard to PKC a specific inhibitor has been described (bisindolymameimide GF109203X, ref.15) and an inhibitor specific for PKC (31 has also been reported (16).Otherwise staurosporine (UBI #19-119) can be used to inhibit PKC activity, although the isoenzyme specificity is unknown, while depletion of PKC may be carried out by prolonged incubation with phorbol myristate acetate (PMA, 100 ngml'1 overnight), although neither PMA nor staurosporine affect all the PKC isoenzymes equally. A caveat for the interpretation of any results concerning the PKC family of isoforms is that at least three sub-groups have been described with different requirements for calcium and different affinities for phorbol esters (17). Okadaic acid (UBI #19-119) does not inhibit protein tyrosine phosphatases but can be used to investigate the roles of serine and threonine protein phosphatases PP1 and PP2A. TPKs can be inhibited by either herbimycin A (Calbiochem, ref. 375670) or by genestein (UBI #19-110).The tyrosine phosphatase CD45 is expressed on almost all haematopoietic cells and anti-CD45 mAbs may therefore be used to detect tyrosine kinase activity indirectly. mAbs stimulate this activity. However, the most precise way in which to examine TPK activity is by estimation of kinase activity. Metabolic labelling of cells with [32P] orthophosphoric acid and polyacrylamide gel electrophoresis of cell lysates provides a simple and rapid technique to determine whether or not phosphorylation is involved in a signal transduction pathway (Protocol 3). It is easier to detect phosphorylation events using resting rather than activated cells since the elevated background levels of phosphorylation in activated cells may mask phosphorylation induced by stimulation of a receptor. Metabolic labelling of resting cells may require longer incubations with the radiolabel. Measurement of the activity of either PKC or of PKA in cellular extracts is important if preliminary studies of phosphorylation (i.e. using inhibitors) indicate a role for one or either of these kinases. Protocols will not be given in this chapter as extremely efficient kits (with detailed protocols) for the measurement of their activity are now available. Use of these kits for measurement of either PKC or PKA activity is not an expensive option (e.g. 38

3: Methods for detecting signalling via MHC class II molecules Gibco-BRL or Amersham) by comparison with the cost of setting up either assay and they are particularly useful for the assay of limited numbers of samples. PKC isoform-specific antibodies have recently been produced and have allowed localization, translocation and activation studies by laser scanning confocal microscopy of cytospins (18). The regulation of PKC isoenzyme mRNA by HLA class II molecules has also been examined using a reverse transcriptase polymerase chain reaction (RT-PCR) with oligonucleotide primers derived from either human or rat PKC cDNA sequences (7). Protocol 3.

Detection of phosphorylation

Equipment and reagents K

P orthophosphoric acid (New England Nuclear, NEX-054) Minimum essential medium (MEM, Flow), phosphate free PCS previously dialysed against phosphate free medium 1 M phenylmethylsulfonyl fluoride (PMSF, Sigma; dilute in acetone)

Plexiglass screen and box NP40 containing extraction buffer: 10 mM Tris-HCL buffer, pH 7.4, 1% NP40 w/v, 150 mM NaCL, 1 mM diethylenediamine tetraacetic acid (EDTA) 2 ml eppendorf tubes 1 ml syringes and needles

Method 1. Incubate resting lymphocytes (5 x 106mr1 ) in phosphate free MEM containing 2.5% PCS which has previously been dialysed against phosphate-free medium. 2. Centrifuge the cells (5 min at 500 x g) at room temperature and discard the supernatant. Resuspend the cells in MEM containing 32P orthophosphoric acid at a final activity of 0.5 uCimT1 (32P was manipulated behind a Plexiglass screen throughout the entire experiment). Metabolic labelling was continued during 4 h at 37°C in a humidified CO2 incubator. 3. Add mAbs or other stimuli 15 min before the end of the experiment. Use PMA (20 ng/ml) as a positive control for induction of lymphocyte phosphorylation. 4. Wash the cells three times in a ten-fold volume of cold PBS and add the NP40 containing lysis buffer to the final pellet (1 ml per 107 cells) and keep the lysates on ice for 20 min before centrifugation at 13000 x g at 4°C for 15 min. Collect the supernatant and carry out electrophoresis as previously described (14). Radiolabelled proteins are readily detected by autoradiography after fixing and drying the gels. Incubation of the gels in 1 M KOH for 1 h at 55°C causes selective dephosphorylation of proteins phosphorylated on serine and threonine residues, although great care should be taken in drying the gels after this treatment as it greatly increases their fragility.

39

N. Mooney, J. P. Truman and D. Charron

5. Detection of tyrosine kinase activity Anti-phosphotyrosine western blotting (Protocol 4) is a straightforward technique for detecting tyrosine phosphorylated substrates and a number of antiphosphotyrosine mAbs are now available (Sigma, Boehringer-Mannheim, UBI, Santa-Cruz). Both the 4610 and the py20 mAbs recogize a wide variety of substrates. Preparation of cell lysates for antiphosphotyrosine blotting requires either an extraction buffer containing inhibitors of tyrosine phosphatases or a rapid extraction procedure using boiling SDS sample buffer which inactivates kinases and phosphatases. Alternatively, immunoprecipitation with an anti-phosphotyrosine mAb after extraction in the presence of tyrosine phosphatase inhibitors is a sensitive way of detecting tyrosine phosphorylated substrates (Protocol 5). Immunoprecipitation followed by immunoblotting is a highly sensitive means of detection. Protocol 4. Preparation of cell lysates to detect tyrosine phosphorylation by Western blotting Equipment and reagents Non-adherent cells suspended at 5 x lO'ml"1 for stimulation SDS 2x extraction buffer: Tris-CL 0.125 M, Glycerol 20%, p-mercaptoethanol 10%, SDS 160 mM; adjust to pH 6.8 with concentrated HCI

1 ml syringes and needles 2 ml eppendorf tubes

Method 1. Aliquot 1 ml of the cell suspension in a 2 ml eppendorf tube for each test. The number of cells required varies according to the cell type under study (e.g. fewer cells can be used for the study of EBVtransformed cell lines compared with resting human lymphocytes). 2. Leave the aliquots in a 37°C humidified CO2 incubator for at least 1 h, which allows the background level of phosphorylation to stabilize and results in clearer differences after stimulation. 3. Prepare the reagents for stimulation. Anti-HLA class II mAbs stimulates at a concentration of 1 ugmT1 without cross-linking, although an optimal dose should be established for each mAb or reagent. 4. Carry out a time course of 1, 2, 5, 10, 30 and 60 min. Tyrosine phosphorylation can be extremely rapid: the majority occurs within the first 60 min. (The minimum time should be as short as possible within the limits of manipulation; tyrosine phosphorylation via the TCP has been observed within 5 sec). 5. Add the anti-HLA mAbs and mix rapidly. Incubate at 37°C or microfuge to stop the stimulation (12000 X g for 1 min). Then tip the 40

3: Methods for detecting signalling via MHC class II molecules tube to remove the supernatant, and replace with an appropriate volume of boiling SDS 2x. 100 ul of SDS 2x is adequate for the lysis of 5 x 106 cells without causing excessive dilution. Vortex the lysate for 30 sec and immediately boil for 3-4 min; then fragment the DNA by repeatedly passing through a 26 G needle on a 1 ml syringe. Centrifuge at 13000 x g at 4°C for 15 min. 6. Prepare the samples for electrophoresis as previously described and migrate on a 10% SDS-PAGE gel (14).

Two factors largely determine the clarity of results of anti-phosphotyrosine blotting: the quantity of phosphotyrosine-containing material present and the prevention of dephosphorylation. Hot SDS buffer lysis is an efficient way to prevent dephosphorylation; an alternative is the addition of tyrosine phosphatase inhibitors (e.g. sodium orthovanadate, sodium fluoride) to a cold lysis buffer containing standard protease inhibitors (leupeptin, aprotinin, and PMSF). Finally, the choice of anti-phosphotyrosine mAb is important with regard to the range of substrates recognized. Protocol 5.

Immunoblotting of tyrosine phosphorylated proteins

Equipment and reagents Enhanced Chemiluminescence (ECL, Amersham) Lysis buffer: Tris 25 mM, NaCI 150 mM, EDTA 1 mM NaF 10 mM, Dithiothreitol 1 mM, 1% NP40, 100 uM sodium orthovanadate, 1% aprotimen, 20 urn leupeptir PVPP membrane (Millipore)

anti-phosphotensive mAb (Upstate Biotechnology Incorporation) BSA or skimmed milk powder Developing solution (to be prepared immediately before use) PBS, NBT substrate (0.4 mM, Promega) 66 ul /10 ml, BCIP substrate (0.4 mM, Promega) 33 ulml-1

Method 1. Carry out steps 1-3 of Protocol 4. 2. Add the anti-HLA mAbs and immediately mix. Incubate at 37°C and stop the stimulations by microfuging for 1 min at 12000 x g. Wash the pellets twice with cold PBS containing 100 pM sodium orthovanadate. Immediately add the lysis buffer containing inhibitors of tyrosine phosphatases (25 mM Tris, pH 7.5; 150 mM NaCL; 1 mM EDTA; 10 mM NaF; 1 mM dithiothreitol; 1% NP40, 100 pM sodium orthovanadate, 1% aprotinin, 20 uM leupeptin) and vortex. Keep the lysates on ice for 15 min, microfuge at 12000 x g for 10 min and collect the supernatants. 3. Transfer the proteins onto a PVDP (Millipore Immobilon P) membrane as previously described (14) and incubate the membrane in a blocking solution (2% BSA or 5% skimmed milk powder in PBS) for 1 h. 41

N. Mooney, J. P. Truman and D. Charron Protocol 5. Continued 4. Rinse the membrane with PBS and incubate with the anti-phosphotyrosine mAb diluted in the blocking solution overnight at 4°C with constant agitation. All incubations are carried out in sealed plastic bags in order to minimalize the amount of mAb used. As a general measure 1 ug ml'1 is adequate and a volume of 10 ml is sufficient for a membrane measuring 10 cm2; most mAb solutions can be recovered and used once again at this concentration if re-used within one week. 5. Incubate overnight and wash the membrane once with PBS containing 0.05% Tween before rinsing several times in PBS. The method of detection depends largely on the amount of protein transferred and the degree of tyrosine phosphorylation. We have found that detection with a second antibody labelled with alkaline phosphatase (UBI) gives a high level of sensitivity. After rinsing several times with PBS, the developing solution is added. The membrane should be checked every few minutes and a large volume of PBS added to limit the background colouring as soon as the bands are clearly visible. Dry the membrane on some Whatman paper and store in a sealed plastic pocket after photographing. A horseradish peroxidase labelled second antibody allows immunodetection using chemiluminescence and is clearly the most sensitive system.

The use of 1251-labelled protein A increases the sensitivity of detection, although appropriate precautions for working with radioactive material must be taken. A further advantage of using either 125 I or chemiluminescense (e.g. ECL system, Amersham) as a detection system is that a permanent record of the results is provided on film.

6. Immunoprecipitation of tyrosine-phosphorylated substrates The immunoprecipitation of tyrosine-phosphorylated substrates followed by SDS-PAGE (Protocol 6) allows better resolution of phosphorylated proteins and is a more useful technique if identification of the substrates is sought rather than simply determination of whether or not there is tyrosine kinase activity. Finally, if the background levels of tyrosine phosphorylation are elevated, the combination of immunoprecipitation with an anti-phosphotyrosine mAb followed by immunoblotting with an anti-phosphotyrosine mAb is extremely sensitive, particularly if known tyrosine kinase substrate-specific mAbs are used. 42

3: Methods for detecting signalling via MHC class II molecules Protocol 6. Immunoprecipitation of tyrosine phosphorylated proteins Equipment and reagents Lysis buffer: see Protocol 5 Wash buffer 1: 10 mM Tris-HCL pH 7.5, 140 nM NaCL, 0.05% NP40, 1% sodium deoxycholate, 0.1% SDS, 100 uM sodium orthovanadate

Wash buffer 2: 10 nM Tris-HCL pH 7.5, 100 uM sodium orthovanadate anti-phosphotyrosine mbs on agarose (UB1)

Method 1. Prepare extracts in ice cold lysis buffer as described for western blotting (Protocol 2, steps 1-3). 2. Check that the total protein concentration of the lysates is greater than 1 mg ml-1 and that it is equal in all the samples to be tested. 3. Centrifuge the lysates for 20 min at 12000 X g at 4°C and recover all but the pellet which contains the aggregated material. Pre-clear the lysate by incubating with an irrelevant Ab (of the same isotype as the anti-phosphotyrosine mAb) and 50 ul of a 50% solution of Protein G Sepharose. Centrifuge as above and recover the supernatant before adding 15 uJ of the anti-phosphotyrosine mAb conjugated to agarose beads (UBI #16-10101) to 100 ul of the extract. Incubate for 4 h at 4°C with continuous rotation. 4. Microfuge for 5 min at 4°C at 12000 x g and collect the supernatant which may be stored at -70°C for further immunoprecipitation. Wash the agarose-anti-phosphotyrosine twice with wash buffer 1 and twice with wash buffer 2, discarding the supernatant each time. Make sure that the pellet is well resuspended after each wash. 5. Add a small volume of SDS 1x (500 ul) and resuspend. Boil for 5 min to dissociate the immune complexes. 6. Centrifuge and remove the supernatant which contains the dissociated antigen and antibodies. 7. Add 10% glycerol containing 0.1% bromophenol blue and centrifuge at 12000 x gfor 3 min. Migrate on a 10% SDS polyacrylamide gel and either silver stain the gel or colour it with Coomassie Brilliant blue (11). 43

N. Mooney, J, P. Truman and D. Charron

7. Detection of programmed cell death via MHC class II molecules Programmed cell death via MHC class II antigens has recently been described in activated and resting lymphocytes in humans and in mice respectively (10, 11). The detection of programmed cell death via MHC class II molecules is complicated due to the rapid induction of aggregation. Disaggregation may result in damage to the cells, while many methods of detecting cell death are difficult to apply to multicellular aggregates. We have found that FACS analysis of isolated nuclei and examination of apoptotic nuclei after staining with a DNA binding dye are simple and reliable methods (Protocol 7).

8. A flow cytometric method for the quantification of cell death This is a method of choice for the quantification of cell death induced via HLA class II as it circumvents the problems encountered due to the rapid and strong aggregation induced via HLA class II after only 20-30 min incubation with anti-HLA class II mAbs. Disruption of these aggregates can lead to a high level of background noise, so care is taken not to dis-aggregate the cells before lysis. The technique is a modification of the method described by Nicoletti et al. (19)

Protocol 7. Quantification of cell death by FACS Equipment and reagents • Hypotonic fluorochrome solution, containing 0.1% sodium citrate, 0.1% Triton-X-100, and 50 ugml~1 propidium iodide diluted with double-distilled water (all reagents from Sigma).8 • Fluorescence activated cell sorter (FACS)

Method 1. Suspend the cells at 0.5 x 106ml-1, and to minimize the background level of events perform all the steps in the same Falcon 2054 tubes (Becton Dickinson, Mountain View, CA). The optimal time of incubation with anti-HLA class II mAbs can vary according to cell type; 24-72 h is recommended. The use of a secondary antibody to cross-link the first may be required to increase the level of apoptosis. 44

3: Methods for detecting signalling via MHC class II molecules 2. Centrifuge the cell sample at 1500 r.p.m. for 5 min, and remove the medium. Carefully resuspend the cell pellet in 1 ml of the hypotonic fluorochrome solution. Take care to ensure that the cells are properly resuspended using a vortex mixer (never by pipetting), as this can drastically increase the numbers of background events. Leave the samples overnight at 4°C in the dark. 3. Analyse the nuclei using the logarithmic scale of the red-sensitive FL2 channel of a FACScan. 10000 events per sample is the recommended minimum when acquiring data. Events with very small side and forward scatter, corresponding to cellular debris (mostly incompletelylysed membranes and necrotic cells), are removed either by gating or by raising the forward scatter threshold. 4. The DMA content of individual nuclei stained by propidium iodide should be concentrated around a narrow region. As the cell progresses through the cell cycle, from G1 to G2/M, the nuclei contain greater amounts of DMA and so stain brighter, and should be apparent in proliferating cells as two closely spaced peaks. The first contains the G1 nuclei, the second, usually smaller peak contains the G2/M nuclei. The events between the two peaks correspond to nuclei that were synthesizing DNA during their S phase. Any apoptotic nuclei are shown to the left of the normal nuclei, as these contain less than their full complement of DNA. "Appropriate care must be taken in the handling and disposal of propidium iodide. Keep the stock solution in the dark at 4°C.

9. Visualization of programmed cell death by fluorescence microscopy This rapid and simple technique relies on the use of a DNA binding dye, Hoescht 33342 (Protocol 8). Hoescht 33342 is a blue fluorescing dye, although if the microscope does not have the necessary filters, it will fluoresce at green wavelengths as well, at the expense of an increased background level. The nuclei of apoptotic cells fluoresce brighter than either normal or necrotic nuclei due to chromatin condensation, and are therefore readily distinguishable. Furthermore, apoptotic nuclei typically have altered morphology, appearing either 'bent' or crescent-shaped, smaller or fragmented into various small, highly fluorescent particles. 45

N. Mooney, J, P. Truman and D. Charron

Protocol 8. Visualization and identification of apoptotic cells by fluorescent microscopy Equipment and reagents Fluorescent microscope (e.g. Leica or Zeiss Axiophot)

Hoescht 33342 (bisBenzimide, Sigma). The stock solution is at 1 mgml1, made up with distilled water, and kept in the dark at 4°C. As with all DNA binding agents, care should be taken in the handling and disposal of Hoescht 33342

Method 1. During the treatments, the cells are at a concentration of 5 x 105 ml-1. For visualization, the cells are incubated in the dark for 5 min at room temperature with 0.5 pgmT -1 of Hoescht 33342. It is important to ensure that the aggregates caused by HLA class II signalling are not physically disrupted, as this can damage the cell membranes and increase the level of background fluorescence. Use of disposable wide-diameter pipettes is therefore highly recommended.

Summary The transmission of signals via HLA class II DR molecules has been examined using either mAbs or bacterial superantigens as ligands. Involvement of the following signalling pathways have been described: the activation of a tyrosine kinase, the activation of PLC, an intracellular calcium flux, production of DAG, activation of PKC, and transcriptional activation of c-myc (6). The second messengers described here are therefore common to a number of signalling pathways induced via other members of the Ig superfamily. The abundance of isoforms of both PLC and of PKC might explain the diversity of responses induced by different signal transducing molecules. It will be an important task to pinpoint the PKC/PLC isoform(s) involved in class II mediated responses. To date, there is no evidence for a haplotype-related diversity of signalling. DR transmits signals efficiently while both DQ and DP are more variable in their responses and this variability does not appear to be simply a consequence of different levels of expression. Given the diversity of cells which either constitutively express HLA class II molecules or can be induced to do so, a diversity of responses to signals transmitted via these molecules is to be expected. Furthermore, the state of activation of the cell, the presence or absence of other signalling molecules, and the environment in which the cell is located may well affect the outcome of 46

3: Methods for detecting signalling via MHC class II molecules signals via MHC class II molecules. The latter consideration may be particularly important with regard to the presence or absence of co-signalling cytokines as well as potential ligands present on neighbouring cells. The circumstances in which MHC class II signalling leads to programmed cell death rather than to activation have not been defined. Little is known of the cytosolic interactions leading to the generation of second messengers although two putative HLA class II associated proteins have recently been purified and characterized (20). The potential to influence antigen presentation via MHC class II signalling as well as the potential for signalling via 'aberrantly' expressed molecules, (e.g. in autoimmune conditions) underlines the importance of the elucidation of signalling pathways and of their role in physiological and in pathological conditions.

References 1. Cambier, J.C., Newell, M.K., Justement, L.B., McGuire, J.C., Leach, K.L., and Chen, Z.Z. (1987). Nature, 327, 629. 2. Mooney, N., Grillot-Courvalin, C., Hivroz, C., Ju, L, and Charron, D. (1990). / Immunol., 145,2070. 3. Lane, P.J., McConnell, P.M., Schieven, G.L., Clark, E.A., and Ledbetter, J.A. (1990). J. Immunol., 144,3684. 4. Charron, D., Brick-Ghannam, C., Ramirez, R., and Mooney, N. (1991). Res. Immunol., 142,467. 5. Brick-Ghannam, C., Mooney, N., Charron, D. (1989). Hum. Immunol., 30,202. 6. Brick-Ghannam, C., Huang, F.L., Temine, N., and Charron, D. (1991). /. Eiol Chem.,266,24169. 7. Brick-Ghannam, C., Ericson, M., Schelle, I., and Charron, D. (1994). Hum. Immunol., 41,216. 8. St-Pierre, Y. and Watts T. (1991). J Immunol., 147,2875. 9. Nabavi, N., Ghogawala, Z., Myer, A., Griffith, I.J., Wade, W., Chen, Z., Me Kean, D., and Glimcher, L. (1989). J. Immunol., 142,1444. 10. Newell, M., Vanderwall, J., Scott Beard, K., and Freed, J.H. (1993). Proc. Natl Acad. Sci.,9,10459. 11. Truman, J.P., Ericson, M., Seebold, C., Charron, D., and Mooney, N. (1994). Int. Immunol., 6, 887. 12. Kanner S.B., Odum N., Grosmaire L., Masewicz S., Svejgaard A., and Ledbetter J.A. (1992). J. Immunol, 149,3482. 13. Chused, T.M., Wilson, M.A., Greenblatt, D., Ishida, Y., Edison, L.J., Tsien, R.Y., and Finkelman, F.D. (1987). Cytometry, 8,396. 14. Gallagher, S. In Current protocols in immunology (ed. J.E. Coligan, A.M. Kruisbeek, P.H. Margulies, E.M. Shevach, and W. Strober), Vol. 1, Unit 8.4.1 (Electrophoretic separation of proteins), Wiley Interscience. 15. Touillec, D., Pianetti, P., Coste, H., Belleverge, P., Grand-Perret, Ajakane, M., Bandet, V., Boissin, P., Boursier, E., Loriolle, F., Duhamel, L., Charon, D., and Kirilovsky, J. (1991). J. Biol. Chem., 266,15 771. 47

N. Mooney, J. P. Truman and D. Charron 16. Ryves, W., Evans, A., Olivier, A., Parker, P., and Evans, F. (1991). FEBS Lett., 288,5. 17. Nishizuka, Y. 1992. Science, 258,607. 18. Pongracz, J., Johnson, G., Crocker, J., Burnett, D., and Lord, J. (1994). Biochem. Soc. Transactions, 22, 994. 19. Nicoletti, I., Migliorati, G., Pagliacci, M.C., Grignani, F., and Riccardi, C. (1991). J. Immunol. Methods, 139,271-279. 20. Vaesen, M., Barnikol-Watanabe, S., Gotz, H., Adil Awni, L., Cole, T., Zimmermann, B., Kratzin, H.D., and Hilschmann, N. (1994). Biol. Chem., 375,113.

48

4

Peptide translocation into the ER JACQUES NEEFJES, FRANK MOMBURG, GUNTER HAMMERLING and JOOST ROELSE

1. Introduction A crucial step in antigen presentation by MHC class I molecules is the translocation of peptide fragments of protein antigens from the cytosol to the lumen of the endoplasmic reticulum (ER). Here, peptides are bound with high affinity by freshly assembled MHC class I H-chain/B2m heterodimers and are then intracellularly transported to the cell surface and presented by class I molecules to CD8+ T cells (1,2). An ER-located peptide pump termed TAP (Transporter associated with Antigen Presentation) has been genetically defined and demonstrated to translocate peptides from the cytosol to the lumen of the ER (3,4). Essential to the characterization of TAP were mutant cell lines with inactivated genes for one or both of the two subunits of TAP. In the absence of TAP activity, MHC class I molecules are not loaded with peptides. This results in retention of the class I heterodimer in the ER and decreased cell surface expression of class I molecules. TAP is a heterodimer, each subunit consisting of a multimembrane spanning domain and a cytoplasmic domain with an ATP binding cassette. Hydrolysis of ATP is required for the translocation of peptides by TAP (3). Recently, TAP-dependent peptide translocation has been shown in streptolysin O-permeabilized cells (3, 7) and in microsomes (4); the assays are described in detail in Sections 2 and 3 respectively. Specificity controls are performed either by using permeabilized cells or microsomes lacking TAP or by performing the assay in the absence of ATP. Peptide translocation by TAP is influenced by the sequence of the peptide and the polymorphism of the TAP subunits (8, 9). Furthermore, TAP prefers peptides of 8-13 amino acids, covering the length of class I-binding peptides (8-11 amino acids) (10, 11). This indicates that TAP pre-selects the peptides that associate with MHC class I molecules in the ER. The precise definition of the specificity of TAP may thus allow a better prediction of the peptides that finally bind to class I molecules.

Jacques Neefjes et al.

2. TAP-dependent peptide translocation 2.1 Peptide translocation in streptolysin O-permeabilized cells TAP translocates peptides from the cytosol to the ER. Since peptides do not diffuse over the plasma membrane, one has to gain access to the cytosol in order to study TAP-dependent peptide translocation. This can be achieved by selective permeabilization of the plasma membrane in a way that allows free diffusion of peptides to the cytosol. To control for the integrity of the ER membrane, such an assay is preferably performed on cells that have a counterpart deficient for one or both of the TAP subunits, e.g. the murine cell lines RMA and RMA-S (12) or the human cell lines LCL721 (or Tl) and T2 (13). Any difference in peptide translocation between the TAP-expressing and TAP-deficient cell lines can then be attributed to the function of TAP. Peptide translocation from the cytosol into the ER can be analysed in different ways (see Section 2.1.2). The advantage of the use of the bacterial toxin streptolysin O for the permeabilization of cells in order to study peptide translocation is that it can be applied to almost every cell type from different species, in addition to which it is reproducible and easily applicable. 2.1.1 Permeabilization There are multiple methods for selectively permeabilizing the plasma membrane. Low concentrations of detergents like digitonin can render the cytosol accessible to large molecules but, in our hands, this approach could not be applied successfully to peptide translocation studies. Bacterial toxins have been widely used for the permeabilization of the plasma membrane. The atoxin from Staphylococcus aureus generates small pores (2 nm) that allow the passage of only small molecules like ATP (14). Peptides are around the exclusion size, which makes this toxin less than ideal for this assay. We and others (3, 7) have successfully used the bacterial toxin streptolysin O for the permeabilization of the plasma membrane of different cell types. Binding of streptolysin O to the cell surface can be competed for by exogenous cholesterol, suggesting that cholesterol is a receptor for streptolysin O. Streptolysin O can bind to the cell surface at any temperature between 0°C and 37 °C but pore formation requires temperatures higher than ~ 20 °C. Pore formation is irreversible and the pore size (> 15 nm) allows the passage of large molecules (15). Since a large portion of the cytosol will leak out of the cell after pore formation, a number of biological processes (e.g. vesicular transport) can continue only when cytosol is added back. However, peptide translocation by TAP works in the absence of soluble cytosolic factors. Peptide translocation can be performed in the continuous presence of streptolysin O. Although it can, in principle, also permeabilize the ER 50

4: Peptide translocation into the ER membrane, we have never observed streptolysin O treatment to result in significant TAP-independent entry of peptides into the ER. Another option is a protocol in which streptolysin O is first bound to the cell surface at 0°C, the non-bound fraction is removed by washing, and the cells are then permeabilized at 37 °C. The efficiency of permeabilization is measured by trypan blue exclusion or LDH release. 50-80% permeabilization is optimal for translocation studies. The efficiency of permeabilization is determined by (1) the cell type used, (2) the concentration (and activity) of streptolysin O, and (3) competing factors in the serum. The optimal concentration of streptolysin O required for 50-80% permeabilization has to be determined for each cell type or cell line. Concentrations which are too high will result in complete cell lysis, concentrations which are too low in inefficient permeabilization. The activity of streptolysin O varies, dependent on supplier and storage. Streptolysin O requires a reduced cysteine for activity, and can be activated with low (0.1 mM) concentrations of DTT. Oxidation of streptolysin O (e.g. due to storage in solution) results in reduced activity or inactivation. 2.1.2 Entry in the ER Entry of peptides into the ER of permeabilized cells has been assayed in three ways: • following peptide binding to MHC class I molecules; • by removing the non-transported cytosolic peptide fraction through washing; • by following a peptide modification in the ER. Peptide translocation can be observed by following binding of a peptide to MHC class I molecules and the resulting release of the class I-peptide complex from the ER (7). In this assay system, however, it is difficult to dissect peptide translocation by TAP from the molecular events resulting in loading of MHC class I molecules. It is therefore impossible to determine whether ATP is required for translocation or for peptide loading and at what level peptide competition is acting. Furthermore, binding of radiolabelled peptides to cell surface MHC class I molecules is possible and should be controlled for. Alternatively, radiolabelled peptides are translocated in cells permeabilized with streptolysin O (Protocol 1). After different periods, cells are extensively washed and the retained peptides are quantitated. Essential for the read-out of this assay is the inclusion of a proper negative control (e.g. a TAP-deficient cell line or translocation in the absence of ATP) to monitor for peptides, or fragments thereof, that non-specifically bind to the permeabilized cells. Furthermore, there is the problem that peptides imported into the ER by TAP are released again into the cytosol and degraded. This implies 51

Jacques Neefjes et al.

that peptide translocation or competition for peptide translocation with other peptides have to be assayed during the initial short period (5-10 min) after addition of radioiodinated input peptide when there is still enough peptide present to maintain a steady state level in the ER. The relatively high background and the time required for the extensive washings make this assay not the method of choice. Translocation of peptides that obtain an TV-linked glycan upon arrival in the lumen of the ER provide a convenient and reproducible method to follow TAP-dependent peptide translocation. This is usually performed with peptides that contain a tyrosine for radioiodination and an AMinked glycosylation consensus sequence (Asn-X-Thr/Ser; where X is every aminoacid except Pro). Proline residues directly adjacent to the consensus sequence also negatively influence the rate of glycosylation. Glycosylated peptides are stable and are not exported from the ER to the cytosol (11). Peptide translocation can be easily assayed after lysis of the streptolysin O-permeabilized cells at defined periods and recovery of the radiolabelled and glycosylated peptides with concanavalin A-Sepharose. However, most naturally processed peptides do not contain an AMinked glycan consensus sequence. By competing for TAP-dependent translocation of a glycosylatable (model) peptide, the relative affinities of these non-glycosylated peptides for the TAP translocator can be determined. Experiments should be performed over brief periods (5-10 min) to assure competition during the initial increasing phase of TAP-dependent peptide translocation (3). 2.1.3 Peptidase activity A problem associated with the use of streptolysin O-permeabilized cells is that they contain strong peptidase activity. Peptidase activity is found associated with the cell surface, the cytosol and possibly other membranes as well. A fraction of the peptidase activity can be eliminated by washing off the cytosol that leaks out of the permeabilized cells. N- and C-terminal exopeptidase activities and endopeptidase activities are observed, and we have been unable to inhibit these activities using the commercially available protease inhibitors. Interestingly, small peptides are far better substrates for these peptidases than longer peptides (11). As a result of the existing peptidase activity, the peptide that is actually translocated is not necessarily identical to the input peptide and therefore its nature has to be determined (11). Protocol 1. Peptide translocation by Streptolysin-0 permeabilized cells This protocol allows the study of peptide translocation in almost every cell type, provided that the conditions for permeabilization are carefully established. The efficiency of permeabilization should be controlled, e.g. by trypan blue exclusion.

52

4: Peptide translocation into the ER Equipment and reagents lodinated peptides Lysis mix: 1% (v/v) Triton X-100, 50 mM NaCI, 5 mM MgCI2, 50 mM Tris-HCI (pH 7.5) Concanavalin A-Sepharose (Con ASepharose)

< Incubation buffer: 130 mM KCI, 10 mM NaCI, 1 mM CaCI2, 2 mM EGTA, 2 mM MgCI2, 5 mM Hepes (pH 7.3) > Streptolysin O i ATP or an ATP-generating system

Method 1. 10 ug of peptide is dissolved in -100 ul PBS and 0.2-1 mCi Na125l and 10 ug (freshly dissolved) chloramine T are added. Incubate for 5 min and separate free from peptide-bound iodine by addition of Dowex OH". Free iodine binds to Dowex OH" and the iodinated peptide is recovered in the non-bound fraction. Some peptides do bind to DowexOH-1;such peptides can be separated from free iodine by Sephadex G10. To avoid peptide degradation, the iodinated peptides are routinely stored at -20°C. 2. Dissolve Streptolysin 0 in incubation buffer to the optimal concentration for the cells used. 3. Optional: activate Streptolysin 0 with 0.2 mM DTT (15 min on ice). 4. Harvest the cells and wash once with incubation buffer. 5. Permeabilize the cells (2-5 x 106 cells/translocation) with 50 ul activated Streptolysin 0 in activation buffer for 10 min at 37°C. 6. Prepare a 100 mM ATP solution in incubation buffer. ATP is acidic, therefore adjust to pH 7 with 15% (v/v) of a 1 M Tris-HCI solution. Store on ice. 7. Add 10 ul radioiodinated peptide (corresponding to approximately 100 ng) to a solution of 50 ul permeabilized cells, 10 ul 100 mM ATP, and 30 ul incubation buffer to a final 'reaction' volume of 100 ul. Alternatively, include competing peptides, but add them together with the radioiodinated peptide. 8. lncubateat37°C. 9. Stop translocation by adding 1 ml lysis mix. Keep the lysate on ice for at least 15 min. 10. Remove nuclei by centrifugation for 5 min at 14000 r.p.m. (4°C) in a table centrifuge. 11. Transfer supernatant to 50-100 ul Con A-Sepharose to recover glycosylated peptides. 12. Incubate for at least 0.5 h at 4°C, with shaking. 13. Wash Con A-Sepharose 5 times with 1 ml cold lysis mix. Pellet the beads by centrifugation for 2 min at 5000 r.p.m. during each wash. 14. Quantitate the Con A-associated radioactivity by gamma counting. 53

Jacques Neefjes et al. Protocol 1. Continued 15. Optional: release the glycosylated peptides with 100 mM a-methyl mannoside. 16. Optional: analyse the radiolabelled peptides (prior to or after endoglycosidase H treatment to remove the glycan; see ref. 11) on TLC. We routinely use Kieselgel 60 plates (Merck) with a running buffer containing n-butanol:pyridine:acetic acid:water = 97:75:15:60. The plates can be analysed by auto radiography.

2.2 Peptide translocation in microsomes 2.2.1 Generation Microsomes can be made from almost every cell type. In most protocols, cells are fractionated and microsomes isolated by different centrifugation steps (see Protocol 2). Since microsomes may be permeable to peptides and may therefore translocate peptides in a TAP-independent fashion (16, 17), a proper negative control is essential. Thus, peptide translocation by TAP should be ATP-dependent and should not occur in microsomes deficient for TAP. Extensive washes of microsomes to remove not-translocated peptides may result in loss of microsomes (and of the translocated peptide). Microsomes are most easily separated from cytosol and/or free peptides by a single centrifugation over a sucrose cushion (4). 2.2.2 Peptide entry into microsomes Entry of peptides into the ER can be assayed as described for permeabilized cells (see Section 2.1.2, ref. 4, and Protocol 3) and the advantages and disadvantages of the different protocols are similar. Translocated peptides are rapidly exported from the microsomes, unless the assay is performed at a temperature lower than the physiological temperature (23 °C) (4, 8, 10). However, whether peptide translocation at 23 °C and 37 °C give identical results remains to be established. 2.2.3 Peptidase activity Although the cytosol (and therefore a considerable amount of peptidase activity) is removed from the microsomal fraction during the isolation procedure, microsomes are certainly not free of peptidase activity. The extent of peptidase activity associated with the microsomal fraction can vary, depending on the source of microsomes and the protocol followed for the microsomal preparation. It is however unclear whether the peptidases are associated with ER membranes or whether they are derived from contaminating lysosomal, cell surface or other membranes. In general, however, peptidase activity associated with microsomes seems to be lower than in permeabilized cells. 54

4: Peptide translocation into the ER Protocol 2.

Preparation of microsomes

Equipment and reagents RM: 250 mM sucrose, 50 mM TEA-HCI (pH 7.5), 50 mM KAc, 2 mM MgAc2, 1 mM DTT.

• STKMM: 250 mM sucrose, 50 mM TEA-HCI (pH 7.5), 50 mM KAc, 5 mM MgAc2, 0.1% pmercaptoethanol. Add PMSF just before use to a final concentration of 10 ixg ml-1.

Method. 1. Harvest 1-3 x 109 cells and wash once in PBS at 4°C. 2. Resuspend the cells In 20 ml STKMM and centrifuge for 10 min at 1500 r.p.m. (4°C). 3. Resuspend the cells in 10 ml H2O and homogenize in a dounce (15 strokes). 4. Add 30 ml STKMM and centrifuge for 10 min at 7500 r.p.m. Continue with supernatant. 5. Spin the supernatant for 40 min at 18000 r.p.m. Continue with pellet. 6. Resuspend the pellet in 20 ml STKMM and dounce again. 7. Centrifuge for 40 min at 18000 r.p.m. 8. Resuspend the pellet in 1 ml RM, dounce, and freeze (the microsomes) in small aliquots in liquid nitrogen. 9. Store microsomes at -70°C.

Protocol 3.

Peptide translocation by microsomes

Method 1. Thaw the microsomes quickly at 37°C and store immediately on ice. 2. Add 30 M.I incubation buffer, 10 ul radioiodinated peptide and 5 ul 100 mM ATP (pH 7) to 5 ul microsomes. 3. Incubate for 10 min at 37°C. 4. Stop translocation by adding 1 ml lysis mix. 5. Centrifuge for 5 min at 14000 r.p.m. and transfer the supernatant to 100 uJ Con A-Sepharose. 6. Incubate, wash, and quantitate the Con A-bound radioactivity (see Protocol 1). 55

Jacques Neefjes et al.

3. Peptide substrates 3.1 Peptide size MHC class I molecules usually bind peptides of 9 amino acids, although association with longer and shorter peptides (8-11 amino acids) is observed. This corresponds with the size selectivity of TAP, which has the highest relative affinity for peptides between 9 and 11 amino acids long, although longer and shorter peptides are found to be translocated albeit with lower affinity (11). Therefore, the results of competition experiments for TAP-dependent peptide translocation will be influenced by the size of the competing peptide.

3.2 The sequence of peptides Not only the size, but also the sequence, of a peptide determines its affinity for TAP. Notably, species-specific differences have been observed for the Cterminal residue: rat TAP2a and human TAP have no profound selectivity whereas murine TAP and another rat TAP allele (TAP2U) prefer peptides with aliphatic or aromatic C-terminal residues (8, 9). No major effects on the rate of translocation have been observed for aminoacid substitutions at other positions in the peptide, with the exception of proline at denned positions. To study the translocation of hydrophobic peptides is technically difficult. These peptides dissolve in water only with the help of DMSO, ethanol or NaOH. All of these solvents inhibit TAP activity. Furthermore, these peptides may bind to cell membranes or even to the reaction tube and the amount of free peptide in solution cannot be determined with certainty. These peptides therefore cannot be used in competition experiments.

3.3 Peptide modifications Stable association of peptides with MHC class I molecules requires unmodified N- and C-termini (9). Blocking the extremities of a peptide also reduces the efficiency of peptide translocation by TAP. C-terminal modifications are less harmfull for peptide translocation than N-terminal modifications (9). Also, introduction of D-amino acids into a peptide can affect the efficiency of peptide translocation. Therefore, any major peptide modification may influence the efficiency of TAP-dependent peptide translocation and may even abolish it.

3.4 Competition for peptide translocation Most peptides contain neither a tyrosine for radioiodination nor a glycosylation consensus sequence to allow recovery of the translocated peptide by Concanavalin A-Sepharose. TAP-dependent peptide translocation of these peptides can only be analysed by competition for translocation of a model peptide (5-11) (Protocol 4). If translocation of the model peptide is followed by binding to MHC class I molecules, then it should be ruled out that com56

4: Peptide translocation into the ER petition occurs at the level of the loading of MHC class I molecules or during any process, other than peptide translocation by TAP, preceding loading. Similarly, if the radioiodinated peptide obtains an JV-linked glycan, it should be concluded that the actual competition is not at the level of glycan addition. Such competition is obviously not the case when the competing peptides lack the Af-linked glycosylation consensus sequence. Paradoxically, when peptide translocation is performed in the presence of high concentrations of competing peptides, the efficiency of the translocated peptide may increase because peptidase activity is inhibited by the competitor. This effect can be overcome by decreasing the time of the competition experiment, thereby reducing the contribution of peptide breakdown. Protocol 4. Competition experiments Method 1. Prepare different concentrations of competing peptides, in the range 0.1-25 nmol, in 30 |xl incubation buffer and add 10 ul of radiolabelled peptide (100 ng). Usually around 1% of the peptide is actually iodinated. The radiolabelled peptide should contain an N-linked glycan consensus sequence. 2. Permeabilize cells (see Protocol 7). 3. Wash the permeabilized cells twice with incubation buffer to remove cytosol and unbound streptolysin 0. 4. Add ATP (10 M,! 100 mM) and permeabilized cells (50 ul) to the different peptide mixtures (100 ul final volume). 5. Incubate for 5 min at 37°C 6. Stop translocation by adding 1 ml lysis mix, isolate the glycosylated peptides with Con A-Sepharose and quantitate (see Protocol 7). 7. Instead of permeabilized cells, microsomes can be used.

References 1. Rammensee, H.-G., Falk, K., and Rotschke, O. (1993). Ann. Rev. ImmunoL, 11, 213. 2. Neefjes, J.J. and Momburg, P.M. (1993). Curr. Opinion ImmunoL, 5,27. 3. Neefjes, J.J., Momburg, P.M., and Hammerling, GJ. (1993). Science, 261, 769. 4. Shepherd, J.C., Schumacher, T.N.M., Ashton-Rickardt, P.G., Imaeda, S., Ploegh, H.L., Janeway, C.A., and Tonegawa, S (1993). Cell, 74,577. 5. Townsend, A., Ohlen, C., Bastin, J., Ljunggren, H.-G., Foster, L., and Ka'rre, K. (1989). Nature, 340,443.

57

Jacques Neefjes et al. 6. Neefjes, J.J., Hammerling, G.J., and Momburg, P.M. (1993). J. Exp. Med., 178, 1971. 7. Androlewicz, M.J., Anderson, K.S., and Cresswell, P. (1993). Proc. Natl Acad. Sci. USA, 90,9130. 8. Heemels, M.-T., Schumacher, T.N.M., Wonigeit, K., and Ploegh, H.L. (1993). Science, 262,2059. 9. Momburg, P.M., Roelse, J., Howard, J.C., Butcher, G.W., Hammerling, G.J., and Neefjes, JJ. (1994). Nature, 367,684. 10. Schumacher, T.N.M., Kantesaria, D.V., Heemels, M.-T., Ashton-Rickardt, P.G., Shepherd, J.C., Fruh, K., Peterson, P.A., Janeway, C.A., Tonegawa, S., and Ploegh, H.L. (1994).J. Exp. Med., 179,533. 11. Momburg, P.M., Roelse, J., Hammerling, G.J., and Neefjes, J.J. (1994). J. Exp. Med., 179,1613. 12. Ljunggren, H.-G. and Karre, K. (1985). J. Exp. Med., 162,1745. 13. Salter, R.D., Howell, D.N., and Cresswell, P. (1985). Immunogenetics, 21,235. 14. GIBCO BRL, product sheet. 15. Buckingham, L. and Duncan, J.L. (1983). Biochim. Biophys. Ada., 729,115. 16. Levy, P., Gabathuler, R., Larsson, R., and Kvist, S. (1991). Cell, 67, 25. 17. Bijlmakers, M.J.E., Neefjes, J.J., Wojcik-Jacobs, E.H.M., and Ploegh, H.L. (1993). Ear. J. Immunol, 23,1305.

58

5 The 20S proteasome and antigen presentation ANGELA SEELIG and PETER M.-KLOETZEL

1. Introduction The 20S protesome is the major cytosolic proteinase complex of eukaryotic cells. This 700 kDa multisubunit enzyme complex is involved in ubiquitinindependent proteolysis and is also central to the ATP/ubiquitin-dependent intracellular protein degradation pathway where it represents the proteolytic core of the '26S' proteinase complex (1, 2). Thus, proteasomes are involved in the turnover or activation of transcription factors (3, 4), in cell cycle control (5), and the generation of antigenic peptides which are presented at the cell surface by MHC class I molecules (for review see refs 6 and 7). The enzyme complex consists of 14 different subunits ranging in molecular mass from 21 to 31 kDa. The subunits can be classified as a and B-type, based on their homology to the two different subunits, a and /?, of the 20S proteasome complex of the archeon Thermoplasma acidophilum (8). Each ring of the eukaryotic proteasome contains seven different proteins (9). Kopp and coworkers (10, 11) demonstrated that the outer rings are formed by a subunits whereas the two inner rings contain B subunits. The same authors also provided evidence that each a and each B subunit occurs twice in a proteasome molecule. Thus, the 20S proteasome complex represents a homodimer complex in which seven a and seven B subunits each form two rings stacked in order a7, B7, B7, a7 to build the cylinder-shaped enzyme complex. As deduced from mutational analysis in yeast (12, 13), as well as in Thermoplasma (14), the proteolytic activity of the complex resides in its /3type subunits. During the assembly of the 20S proteasome, the N-terminal prosequences of the various j3 subunits, which differ in their primary sequence and length (15), are cleaved from several of the ft subunits. X-ray crystallography of the Thermoplasma proteasome (16), as well as inhibitor studies using lactacystin (17), provided evidence that the N-terminal threonine may function as a nucleophile in the active sites which are oriented towards the lumen of the cylinder.

Angela Seelig and Peter M.-Kloetzel The MHC class I restricted pathway of antigen presentation allows the presentation of intracellular viral antigens to cytototoxic T lymphocytes on the cell surface. Three components of this pathway are encoded completely or in part in the MHC locus (6, 7). These include two subunits, LMP2 and LMP7, of the proteasome, the two subunits of the ER peptide transporter, TAP1 and TAP2, and the MHC class I heavy chain itself, which binds octa or nonameric peptides at defined C-terminal and internal anchor residues in its peptide binding cleft. All three of these components are inducible by iterferon-y, an antiviral cytokine. During the assembly of the 20S proteasomes, LMP2 and LMP7 replace two homologous and so-called constitutive subunits 6 (or Y) and MB1 (or X) after stimulation of the cells by interferon-y. Recently, a third subunit exchange has been discovered in the 20S proteasome, the IFN-y inducible subunit MECL1, which is encoded outside the MHC and replaces the homologous subunit MC14 (in mouse) or Z (in human) (18). It is important to note that all three interferon-y inducible subunits, as well as their constitutive counterparts, possess a free N-terminal active site thronine as well as a conserved Lys33, and thus represent the subunits which most likely determine the proteolytic activity of the 20S proteasome. This suggests, that by cytokine induced change of the subunit composition the cell also alters the proteolytic properties of the 20S proteasome and renders the cell more flexible to immunomodulation.

2. Biochemical characterization of the 20S proteasome 2.1 Purification The 20S proteasome is a complex enzyme composed of 28 subunits. It is an abundant enzyme accounting for approximately 1% of the soluble cytoplasmic protein. Proteolytic activity, subunit composition, total amount, and cellular localization can vary between different cell lines. This variation is relative to the proteasome content of the tissue of origin. The highest concentration is found in the liver (1% of soluble cytoplasmic protein) (19). In order to analyse proteolytic activities and subunit composition it is necessary to purify the 20S proteasome to homogeneity. A consequence of purification is the loss of loosely associated protein components. There is an ongoing discussion about the assembly of the 20S proteasome with other components to form the ATP-dependent 26S proteasome (20-23). The 26S proteasome is the proteinase responsible for the ATP-dependent degradation of ubiquitinated protein. The 20S proteasome is the proteolytic core of the 26S complex, but further work is necessary to analyse the complex relation between the 20S proteasome and the 26S proteinase. Despite its function in the ubiquitin-dependent pathway of protein degradation, it has been shown that the 20S proteasome degrades non-ubiquitinated, denatured, and misfolded pro60

5: The 20S proteasome and antigen presentation teins in vitro. All methods described here are for purification of the proteolytically active and very stable 20S proteasome complex rather than the purification of the 26S proteasome. An overview of the purification procedure follows. For more detailed information refer to Protocol 1. The purification starts with 5-20 freshly isolated mouse livers (approximately 10-40 g). After homogenization a soluble cytoplasmic supernatant (Sol-80) is generated (24). Most of the 20S proteasomes can be found in this fraction. However, the remaining pellet does contain some 20S proteasomes. By washing the pellet in extraction buffer containing 1% Triton X-100 a membrane-associated 20S proteasome subpopulation has been isolated (25). The Sol-80 supernatant is applied to a weak anion exchange column (DEAE-Sephacel, Pharmacia) and washed extensively to elute the unbound material. It is possible to isolate three different proteasomal subpopulations by elution with a 200 mM, 300 mM, and a 500 mM salt step-gradient (25, 26). For a more general approach the total 20S proteasome population can be eluted from the column with a 500 mM KAc buffer. The main OD2go peak is collected, concentrated, and applied to a gel filtration column (S400Sephacryl, Pharmacia). Proteasome-containing fractions are identified by enzyme assay using the fluorogenic peptide Suc-Leu-Leu-Val-Tyr-Mec as a substrate (Protocol 2). The fluorogenic peptide Suc-Leu-Leu-Val-Tyr-Mec (chymotrypsin-like proteolytic activity) is a specific substrate for mouse liver proteasomes. There are other fluorogenic peptide substrates available, but for routine identification during purification of the 20S proteasome the above substrate is generally used. The proteolytically active fractions are combined, concentrated (Amicon Ultrafiltration), and loaded onto a 10-40% sucrose gradient (24) (Protocol 1). The proteasome sediments at approximatly 20S, which can be monitored by the enzyme assay (Protocol 2). The final purification step utilizes a Mono Q anion exchange column. Compared to DEAE, Mono Q is a strong anion exchanger. After Mono Q column elution, the proteasome-containing fractions can be visualized as an OD2g0 peak and by enzyme assay. Proteasome purity is checked by 12% SDS-PAGE followed by Coomassie staining. If there are contaminating proteins > 40 kDa, the Mono Q purification step is repeated. Purification of 10 mouse livers should yield approximately 300 ug of highly purified 20S proteasomes. This purified proteasome population can be analysed by two-dimensional gel electrophoresis (Protocol 5) to determine the subunit composition, and it can be used for more detailed analysis of its proteolytic activity. For purification of 20S proteasomes from tissue culture cells, highly purified material is obtained by preparation of a Sol-80 extract, DEAE ion exchange chromatography, 10-40% sucrose gradients, and Mono Q ion exchange chromatography. However, if the amount of cells available for purification is a limiting factor, lyse the cells in Sol-80 containing 0.5% Triton 61

Angela Seelig and Peter M.-Kloetzel X-100 and apply directly to the sucrose gradient. The proteolytically active fractions are then subjected to Mono Q chromatography. The amount of proteasome obtained should be enough for a 2D gel and several protease assays. Protocol 1. Purification of 20S proteasome Method All purification steps are performed at 4°C. For the chromatography steps use FPLC (Pharmacia). 1. Equilibrate 10 ml DEAE-Sephacel matrix using 80 mM KAc buffer (80 mM KAc, 5 mM MgAc2,10 mM Hepes, pH 7.4) for at least 45 min. 2. Homogenize 10 mouse livers (approximately 25 g) in 50 ml 80 mM KAc buffer, first with an electric mixer and then with a Dounce Homogenizer (equipped with tightly fitting pestle). Using tissue culture cell lines, homogenize approximately 7 x 108 cells with a Dounce Homogenizer. Alternatively, cells can be lysed by repeating cycles of freeze-thrawing (10 min at -70°C) in 80 mM KAc buffer followed by vortexing. 3. Centrifuge the homogenate at 40000 g for 20 min (SS34 rotor, Sorvall, 18000 r.p.m.). 4. Incubate the resulting Sol-80 supernatant for at least 1 h while shaking with the equilibrated DEAE-Sephacel matrix.* 5. Pour the matrix into a column (Pharmacia) and wash with 80 mM KAc buffer until OD28o is < 0.1 The flow rate should be 2 ml/min. 6. Elute bound material with 500 mM KAc buffer." Collect 8 ml fractions. When analysing cell lines, collect 2 ml fractions. 7. Pool and concentrate peak fractions revealing Suc-Leu-Leu-Val-TyrMec (10 uM) hydrolysing activity (Protocol 2) to 10-20 ml by a 50 ml Amicon Ultrafiltration (Amicon), using filter XM 300 (Amicon). This filter retains only molecules with a molecular weight greater than 300 kDa. 8. Subject the 10-20 ml sample to S-400 Sephacryl chromatography (column C28/100, adaptor AC28, Pharmacia), using a flow rate of 0.5 ml/min. Collect 8 ml fractions. Identify proteasome-containing fractions by enzyme assay (Protocol 2). 9. Use the proteasome-containing fractions that have been concentrated by Amicon filtration (step 7) for sucrose gradient centrifugation. Gradients are poured (6.4 ml 10%, 6.4 ml 40% sucrose in 80 mM KAc buffer) into SW40 polyallomer tubes (Beckman) using a gradient mixer (Pharmacia). Each 10-40% sucrose gradient can be loaded with up to 1.2 ml of sample. If the starting material was 10 mouse livers then two or three 10-40% sucrose gradients are used. For cell lines,

62

5: The 20S proteasome and antigen presentation one or two sucrose gradients are used. Centrifuge the 10-40% sucrose gradients in a Beckman ultracentrifuge with the SW40 Ti Rotor at 40000 r.p.m. until an integral w2t value of 1.0 x 1012 is reached. Collect 600 ul fractions by hand or with an ISCO gradient analyser. The proteasome peak sediments at approximately 208, which corresponds to fractions 11-13 if the gradient is collected from the top. Using an ISCO gradient analyser, there should be an OD254 peak around fraction 11-13. Check proteolytic activity by using the enzyme assay (Protocol 2). 10. The final purification step is a Mono Q anion exchange column (Pharmacia). Sucrose gradient fractions containing mouse liver proteasomes are dialysed against Mono Q running buffer (100 mM KCI, 5 mM MgCI2, 10 mM Hepes pH 7.2). The sample is applied to a FPLC Mono Q column using a 50 ml loop (flow rate 1 ml/min). Elute proteasomes with a salt gradient between buffer A (100 mM KCI, 5 mM MgCI2/ 10 mM Hepes pH 7.2) and buffer B (1 M KCI, 5 mM MgCI2, 10 mM Hepes pH 7.2) according to the following program: 6 ml buffer A, 6.5 ml 0-25% buffer B, 10 ml 25-35% buffer B, 3.5 ml 35-45% buffer B, 4 ml 45-100% buffer B, 4 ml 100% buffer B, 4 ml buffer A. The proteasome elutes between 25-35% buffer B and can be identified by enzyme assay. 11. Proteasome purity is checked by 12% SDS-PAGE followed by Coomassie blue R-200 (Sigma) staining. If there are contaminating proteins (> 40 kDa) repeat step 10 by diluting the proteasome fraction with buffer A lacking the KCI and reapply it to the column. "If the amount of cells is limited, skip DEAE and gel filtration chromatography and apply the Sol-80 extracts directly to the sucrose gradient. 6 It is possible to elute the DEAE-Sephacel with a salt step-gradient of 200 mM, 300 mM, and 500 mM KAc buffers. Each salt step will contain a proteasome subpopulation.

2.2 Analysis of proteolytic activities of the 20S proteasome Small peptide substrates with a chromophore or fluorophore reporter group are often used for characterization of the catalytic activity of the 20S proteasome. The enzyme is capable of hydrolysing peptide bonds on the C-terminal side of basic (trypsin-like), hydrophobic (chymotrypsin-like), and acidic (peptidyl-glutamyl proteolytic activity) amino acids. Detailed analyses using a large variety of fluorogenic peptides to determine proteolytic activities of 20S proteasomes, isolated from different species have been published (for reviews see refs 2, 27, and 28). It is impossible to analyse cleavage specificities this way. To address the question of the ability of the proteasome to generate specific peptides which are capable of binding to MHC class I molecules it is necessary to study in vitro the degradation of whole proteins or synthesized larger peptides. 63

Angela Seelig and Peter M.-KIoetzel 2.2.1 Enzyme assay with fluorogenic peptides The use of small fluorogenic peptides is a very helpful tool for generating information about proteolytic activity and cleavage specificity of different amino acids at the C-terminus. In combination with the use of certain proteinase inhibitors it is possible to analyse catalytic centres in more detail. 20S proteasome subpopulations isolated from different sources representing different subpopulations can show different proteolytic cleavage capacity when using fluorogenic peptides (24, 25). Protocol 2 describes a method using fluorogenic peptides. Protocol 2.

Protease assay using small fluorogenic peptides

Equipment and reagents Fluorogenic peptides (Bachem Feinchemikalien)a Substrate buffer: 50 mM Tris pH 7.5, 25 mM KCI, 10 mM NaCI, 10 mM MgCI2 1 mM DTT.0.1 mMEDTA

Stop buffer: 30 mM NaAc, 70 mM acetic acid, 100 mM Monochlor acetic acid*

Method 1. Remove a small aliquot of the fraction to be analysed (10-50 pi) and add it to 1 ml of substrate buffer containing 20 uM fluorogenic peptide. Mix and incubate for 30 min at 37°C. 2. Stop the reaction by adding 100 pi stop buffer (29). Mix the sample well. 3. Measure the release of fluorogenic groups after hydrolysis of the peptide bond in a spectrofluorometer (Shimadzu RF-5000). Record at the excitation wavelength of 366 nm and emission wavelength of 480 nm (alternatively use a spectrofluorometer with fixed wavelengths). aKeep stock solutions of 100 mM in DMF. 6 Caution: Monochlor acetic acid is toxic! Do not inhale and wear gloves while making the solution.

2.2.2 Peptide degradation The 20S proteasome is the main proteinase which degrades polypeptides to small peptides in the cytoplasm. These peptides are transported into the ER where they can bind to MHC class I molecules. After peptides binding the complex migrates to the cell surface where the antigenic peptide is presented to cytotoxic T lymphocytes. To address the question can the proteasome, indeed, produce antigenic peptides, it is possible to use synthetic polypeptides containing sequences known to be presented by MHC class I molecules for in vitro analysis (Protocol 3). The cleavage products can be analysed by HPLC chromatography followed by amino acid sequencing or cytotoxic T-cell assays. 64

5: The 20S proteasome and antigen presentation Protocol 3. Degradation of a 25-mer polypeptide and HPLC analysis of the generated fragments Equipment and reagents Assay buffer: 20 mM-KOH Hepes pH 7.8, 2 mM MgAc2, 1 mM DTT

Buffer A: 0.1%TFA Buffer B: 0.1% TFA, 70% acetonitrile

Method 1. To assay the polypeptide cleavage properties of the highly purified 20S proteasome (Protocol 1) a 25-mer polypeptide was chosen and synthesized. The sequence was derived from the murine cytomegalovirus IE pp89 protein: RLMYDMYPHFMPTNLGPSEKRVW-MS (30). 2. Incubate 20 ug polypeptide and 1 ug proteasome in assay buffer at 37°C overnight in a total volume of 300 ul. 3. For analysis of the cleavage products, apply 50 ul of the assay mix to an uRPC PC 3.2/3 (Pharmacia) for separation by HPLC chromatography (SMART, Pharmacia). 4. Elute polypeptides with a gradient of buffer A and buffer B. 5. The eluted peptides are further analysed by peptide sequencing or cytotoxic T cell assays. More detailed descriptions of methods for analysing and sequencing peptides, as well as methods for cytotoxic T cell assays, are found elsewhere in this book.

3. Isolation and analysis of proteasomes from cell lines before and after IFN-g induction Each given cell line or tissue has its proteasome pool. These proteasome pools may vary in subunit composition, proteolytic activity, and function, in relation to different metabolic states of the cell line or tissue analysed. Therefore, it is possible that proteasome subunit composition might be altered after certain cell stimuli, like cytokine treatment, e.g. IFN--y. IFN--y treatment induces strong expression of MHC class I molecules and other proteins related to antigen expression. It has been shown that IFN-"y induction has effects on the expression of proteasome subunits, especially LMP2 and LMP7 whose genes are found in the MHC (31-34). IFN--y influences not only LMP2 and LMP7, but also other proteasomal subunits, e.g. MECL1 (18). For analysis of IFN--Y effects, proteasomes from mouse RMA cells (35) before and after induction are compared. Care must be taken in choosing the IFN--y concentration, because a high concentration might effect cell viability. Therefore 65

Angela Seelig and Peter M.-Kloetzel

Figure 1. Immunoprecipitation of [35S] methionine labelled RMA cells (A) before and (B) after IFN--y (20 U/ml) stimulation with anti-proteasome antibody. The unlabelled arrow in (B) shows the INF-y inducible subunit MECL1 (18).

the use of a low concentration which is close to in vivo concentrations after virus infection is recommended. Monitoring of subunit composition changes or modification of a particular proteasomal subunit can best be achieved by two-dimensional gel electrophoresis, because several subunits have the same molecular weights but different isoelectric points. A quick and simple method of analysing subunit composition changes after certain stimuli is radioactive labelling of cells and immunoprecipitation of proteasomes from control and stimulated cells. Precipitation of proteasomes from unfractionated cell extracts is not recommended since those contain Table 1.

Proteasome 650-700 kDa

Size Morphology

Four staggered rings, cylinder-like

Total number of subunils

28

MW of subunits

20-35 kDa

Total number of different subunits 14 Proteasome genes

Two gene families encoding subunits of a-and /3-type

Proteolytic activities

Trypsin/chymotrypsin like, peptidyl-glutamyl hydrolysing activity, proteinolytic activity

pH optimum

Neutral/alkaline

Inhibitors

Peptide aldehydes, factacystein

Inducers

Interferon-)y PA28 activator, SDS

Cellular localization

Cytoplasm and nucleus

66

5: The 20S proteasome and antigen presentation proteasome precursor complexes which differ in their complexity and subunit composition from the 20S proteasome (38). Figure 1 shows the changes in proteasome subunit composition resulting from IFN--y stimulation of RMA cells (18). More detailed information about proteasomal subunit composition can be achieved by using subunit-specific antibodies for immunoprecipitation. Protocol 4. Radioactive labelling, IFIM--/ treatment, and immunoprecipitation of proteasomes from tissue culture cells Method 1. Stimulate mouse RMA cells with 20 U/ml recombinant marine IFN-'y (recombinant IFN--y, Boehringer) for 72 h. 2. For efficient radioactive labelling of proteasome subunits, grow 3 x 106 mouse RMA cells for 1 h in medium lacking methionine and FCS before labelling and then incubate for about 8 h with 250 uCi/ml[35S] methionine (specific activity > 1000 Ci/mmol, Amersham). 3. After labelling and IFN--y incubation, centrifuge control and induced cells (1500 r.p.m. for 5 min) and wash in 5 ml PBS. 4. For immunoprecipitation, lyse the cells in 1 ml 80 mM KAc buffer containing 1% Triton X-100 for 30 min. 5. Spin down lysed cells at 40000 g (18000 r.p.m. in a Sorval SS34 rotor). 6. To immunoprecipitate fully assembled 20S proteasomes, apply cell extracts to sucrose gradients or gel filtration (Protocol 1). For identification of the 20S proteasome use the enzyme assay (Protocol 2).* 7. Incubate protein A Sepharose beads (Pharmacia) in 80 mM KAc buffer (Protocol 1) for at least 3 h. 8. Wash the beads once with 80 mM KAc buffer (Protocol 1) and remove 50-100 ul protein A beads for incubation with 50-100 ul preimmune serum (for pre-clearing of cell extracts) and the same amount of beads for incubation with anti-proteasome serum. The amount of anti-proteasome serum used is dependent on the quality of the serum. 9. Incubate protein A beads and antiserum for 1 h or up to overnight at 4°C on an electric roller. 10. Discard the supernatant and wash the protein A beads three times with 1 ml 80 mM KAc buffer. 11. Incubate proteolytically active fractions (obtained in step 6) or lysis supernatants (obtainted in step 5)s with protein A beads covered with pre-immune antibodies for 1 h to pre-clear the immunoprecipitation reaction. 67

Angela Seelig and Peter M.-Kloetzel Protocol 4. Continued 12. Take the supernatant and incubate at 4°C for 1-4 h with the protein A beads covered with anti-proteasome antibody. 13. Wash the protein A beads 5-10 times with 80 mM K Ac buffer + 0.5% Triton X-100.b The protein A conjugates are now ready for further analysis by SDS-PAGE or, for a more qualitative analysis, on twodimensional isoelectric focusing gels (Protocol 5}. "By using crude cell lysates, the immunoprecipitation results in both fully assembled 20S proteasomes and all precursor proteasomes. If analysis of only fully assembled 20S proteasomes is desired, purify further by, e.g. sucrose gradients or gel filtration (Protocol 1). 'The 20S proteasome is a very stable complex. The immunoprecipitation can be made more stringent by using higher salt concentrations, up to 300-500 mM KAc.

An introduction two-dimensional gel electrophoresis can be found in Gel electrophoresis of proteins: a practical approach (36) and also O'Farrell, 1975 (37). The equipment as well as the methods used for two-dimensional gel electrophoresis are described there in great detail. Protocol 5 describes only the variations used for optimal separation of 20S proteasome subunits. Table 2. Solutions for the first dimensional isoelectric focusing Tube gel mix solution

Sample buffer

5.5 g urea 1.33 ml 28.4% acrylamide, 1.6% bisacrylamide 2 ml (10%) Nonidet P-40 2 ml water 350 u,l Ampholine pH 5-7 150 (JL! Ampholine pH 3-10

5.7 g (9.5 M) urea 2ml (10%) Nonidet P-40 400 u,l Ampholine pH 5-7 100 uJ Ampholine pH 3-10 0.5 ml (5%) p-mercaptoethanol 150(jLl(20%)SDS 2 ml water Store in aliquots at -20°C

27 \L\ APS 19 IJL! TEMED

Good for six tube gels of 10cm x 1.5 mm Overlay 5.4 g (9 M) urea 500 ml Ampholine pH 3-10 7 ml water Store in aliquots at -20°C

Running buffer: 0.02 M NaOH upper reservoir 0.01 M H3PO4 lower reservoir

Protocol 5. Two-dimensional isoelectric focusing of 20S proteasomes Method 1. Sample preparation for the first dimension of isoelectric focusing: (a) Using purified enzyme (Protocol 7): precipitate 25-50 ug of purified 20S proteasome with 2.5 vol. of EtOH for 1 h to overnight at —20°C. Centrifuge the samples in an Eppendorf centrifuge (5 min 68

5: The 20S proteasome and antigen presentation at 13000 r.p.m.), discard the supernatant, and wash the pellet with 70% EtOH. After drying the pellet, add 100 \s\ of sample buffer (Table 2) and dissolve by shaking for at least 4 h (Eppendorf shaker) at room temperature. (b) Using immunoprecipitated samples (Protocol 4)'. add 2 vol. of sample buffer (Table 2) to the washed protein A beads. Incubate with shaking for at least 4 h (Eppendorf shaker) at room temperature. 2. While shaking prepare the first dimension isoelectric focusing gels. Use the gel mix shown in Table 2 for optimal isoelectric focusing of proteasomal subunits. For technical advice on preparing tube gels, see Gel electrophoresis of proteins: a practical approach (36). 3. Add 20 ul sample buffer to the tube gels and pre-run the gels for 15 min at 200 V, followed by 30 min at 300 V, and finally 60 min at 600 V. Remove the NaOH in the upper reservoir and the fluid from the tops of the tube gels. 4. Spin the sample for 5 min at 13000 r.p.m. to precipitate insoluble aggregates. Apply approximately 100 ul of the supernatant to the prerun tube gel. Overlay the sample with 20 ul overlay buffer (Table 2) and run at 400 V overnight (12-16 h). Increase the voltage to 800 V for 1 h at the end of the run to sharpen the bands. 5. Blow out the first dimension tube gel, equilibrate, and run on a 12% SDS-polyacrylamide gel as second dimension. Methods, equipment, and handling of the second dimension is described in great detail in Gel electrphoresis of proteins: a practical approach (36), and also in O'Farrell (37). 6. After two-dimensional gel electrophoretic separation, stain the gels with Coomassie R-250 and destain until proteasomal subunits are visible. If the radiolabelled, immunoprecipitated proteasomal subunits are being analysed, soak the gel after fixing for 20 min in amplifier solution (Amersham), dry onto Whatman filter paper, and expose to Kodak X-ray film.

References 1. 2. 3. 4.

Goldberg, A.L. (1992). Eur. J. Biochem., 203,9. Rivett, A.J. (1993). Biochem. J., 291,1. Kornitzer, D., Raboy, B., Kulka, R.G., and Fink, G.R. (1994). EMBO J., 13,6021. Palombella, V.J., Rando, O.J., Goldberg, A.L, and Maniatis, T. (1994). Cell, 78, 773. 5. Hershko, A. (19%). Adv. Exp. Med. Biol., 389,221. 6. Groettrup, M, Soza, A, Kuckelkorn, U, and Kloetzel, P.-M. (1996). Immunol. today, 17,429. 7. Lehner, P.J. and Cresswell, P. (19%). Curr. Opin. Immunol., 8,59. 69

Angela Seelig and Peter M.-Kloetzel 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Zwickl, P., Kleinz, J., and Baumeister, W. (1994). Struct. Biol, 1,765. Hendil, K.B., Kristensen, P, and Uerkvitz, W. (1995). Biochem. J., 305,245. Kopp, F., Dahlmann, B., and Hendil, K.B. (1993). J. Mol. Biol., 229,14. Kopp, F., Kristensen, P., Hendil, K.B., Johnsen, A., Sobeck, A., and Dahlmann, B. (1995). J. Mol. Biol., 248,264. Heinemeyer, W., Kleinschmidt, J.A., Saidowsky, J., Escher, C., and Wolf, D.H. (1991). £MfiOy.,10,555. Enenkel, C., Lehmann, H., Kipper, J., Giickel, R, Hilt, W., and Wolf, D.H. (1994). FEES Lett., 341,193. Seemuller, E., Lupas, A., Stock, D., Lowe, J., Huber, R., and Baumeister, W. (1995). Science, 268,579. Schmidtke, G., Kraft, R., Kostka, S., Henklein, P., Frommel, C., Lowe, J., et al, (1996). EM BO J., 15,6887. Lowe, J., Stock, D., Jap, B., Zwickl, P., Baumeister, W., and Huber, R. (1995). Science, 268,533. Fenteany, G., Standaert, R.F., Lane, W.S., Choi, S., Corey, E.J., and Schreiber, S.L. (1995). Science, 268,726. Groettrup, M, Kraft, R., Kostka, S., Standera, S., Stohwasser, R., and Kloetzel, P.M. (1996). Ear. J. Immunol., 26,863. Tanaka, K., Li, K., Ichihara, A., Waxman, L., and Goldberg, A.L. (1986). J. Biol Chem., 261,15197. Hough, R., Pratt, G., and Rechsteiner, M. (1987). J. Biol. Chem., 262, 8303. Eytan, E., Ganoth, D., Armon, T., and Hershko, A. (1989). Proc. Natl. Acad. Sci. USA,S6,715l. Driscoll, J. and Goldberg, A.L. (1990). J. Biol. Chem., 265,4789. Orino, E., Tanaka, K., Tamura, T., Sone, S., Ogura, T., and Ichihara, A. (1991). FEBS Lett., 284,206. Kloetzel, P.-M., Falkenburg, P.E., Hossel, P., and Glatzer, K.H. (1987). Exp. Cell Res., 170,204. Seelig, A., Boes, B., and Kloetzel, P.-M. (1993). Enzyme Protein, 47,330. Falkenburg, P.E. and Kloetzel, P.-M. (1989). J. Biol. Chem., 264, 6660. Orlowski, M. (1990). Biochemistry, 29,10289. Orlowski, M. (1993). J. Lab. Clin. Med., 121,187. Barret, AJ. (1980). Biochem. J., 187, 909. DeVal, M., Munch, K., Reddehase, M.J., and Koszinowski, U.H. (1989). Cell, 58;, 3641. Kelly, A., Powis, S.H., Glynne, R., Radley, E., Beck, S., and Trowsdale, J. (1991). Nature, 353,667. Ortiz-Navarrete, V., Seelig, A., Gernold, M., Frentzel, S., Kloetzel, P.-M., and Ha'mmerling, G. (1991). Nature, 353,662. Glynne, R., Powis, S., Beck, S., Kelly, A., Kerr, L.A., and Trowsdale, J. (1991). Nature, 353,357. Yang, Y., Waters, J.B., Friih, K., and Peterson, P.P. (1992). Proc. Natl. Acad. Sci. USA, 89,4928. Karre, K., Ljunggenren, H.-G., Piontek, G., and Kiessling, R. (1986). Nature, 308, 641. Hames, B.D. and Rickwood, D. (ed.) (1990). Gel electrophoresis of proteins: a practical approach, 2nd edn. IRL Press, Oxford. O'Farrell, P.H. (1975). J. Biol. Chem., 250,4007. Frentzel, S., Pesold-Hurt, B., Seelig, A., and Kloetzel, P.M., (1994). J. Mol. Biol., 236,975. 70

6

Analysis of MHC class II-specific T cell clones GRAHAM PAWELEC, FUMIYA OBATA, DAVID SANSOM, HILKE FRICCIUS, THOMAS DAIKELER, MEDI ADIBZADEH, KURT SCHAUDT and HEIKE POHLA

1. Introduction The application of T cell cloning in immunogenetics at the end of the 1970s (1) enhanced an early awareness of the complexity of functionally relevant lymphocyte stimulating determinants, mostly associated with MHC class II, and cytotoxic lymphocyte targets, mostly associated with MHC class I, and proved instrumental in establishing the mechanism of T cell recognition of antigen and alloantigen. Techniques for the generation, propagation and utilization of monoclonal strains of lymphocytes-T have become established as vital tools in diverse areas of immunology, but, with the exception of some early necessity-driven studies on the optimization of culture conditions (e.g. see ref. 2), there are relatively few published data on the practical details of establishing and maintaining human T cell clones (TCC). In fact, standard techniques work well enough, if not perfectly, that they have changed little since they were first established (1, 3). However, there seems to remain a widespread assumption, at least amongst those cellular and molecular immunologists with less personal experience in the area, that once T cell clones have been established, and provided that culture conditions are favourable, cultures can be maintained as permanent cell lines. While this may be true of a relatively small number of T cell clones, probably most experienced cloners will agree that, at least in humans, this is not the case for the majority of normal T cell lines (which should therefore be referred to as strains, not lines, according to tissue culturists' standard nomenclature). Thus, the majority of T cells maintained in culture are found to have finite lifespans, and those examples of immortal lines in the literature may represent rare, not completely normal, variants. Where the matter has been specifically studied, T cells retaining antigen recognition and effector function, yet apparently in a post-mitotic senescent or presenescent state, have been described (4). These investigators also demonstrated that aged human T cells paralled the senescent

Graham Pawelec et al. phenotype of fibroblasts in that on restimulation, fewer cells responded by entering the cell cycle, the remainder being arrested prior to S-phase. As in senescent fibroblasts, the cell cycle was also prolonged in those —20% of senescent cells which still could be restimulated (5). Therefore there is probably no good reason to assume that cultured T cells are immune to the 'normal' processes of ageing in vitro which seem to apply to other types of somatic cells. Our own experience has been that established T cell clones tend not to survive for a period greater than an estimated maximum of ca. 100 population doublings (PD), which is at the upper end of the range expected for normal somatic cells (6). Of course, the figure of 100 PD represents in practise a very large number of cells, more than enough for most experimental applications and probably even for therapeutic applications. The generation and analysis of human alloreactive or antigen-specific TCC requires: • a source of activated T cells for cloning and feeder/antigen-presenting cells on which to grow the clones; • culture medium containing all essential nutrients for cell growth and supplemented with T cell growth factors; • cryopreservation of TCC; • T cell restimulation measured by proliferation, cytotoxicity, or cytokine production.

2. Source of cells 2.1 Isolation and activation of T cells The usual source of human T cells for cloning is peripheral blood mononuclear cells (PBMC) separated from blood by density gradient centrifugation. PBMC isolated from normal donors commonly consist of a majority of resting T cells with some monocytes, fewer B cells and dendritic cells, and very few activated T cells. The proportion of 'null' cells (mostly natural killer cells) varies, but can be considerable, particularly if using blood buffy coats as starting material. Other sources of T cells which have been successfully used to obtain clones for specific applications include cells infiltrating tumours, allografts, autoimmune lesions or graft-versus-host disease lesions. Before they can be cloned, resting T cells have to be activated by specific antigen, alloantigen, or non-specific agents; in the case of infiltrating cells, one usualy wishes to obtain only those T cells already activated in vivo. However, the present discussion will limit itself to a consideration of MHC-specific TCC, which are usually obtained after in vitro activation of resting T cells. Mixed lymphocyte cultures (MLCs) are common and convenient sources of activated T cells for studies on the recognition of MHC products. Alternatively, antigen-specific MHC-restricted TCC can also be used. In its simplest form, the MLC is obtained by mixing together PBMC from two genetically 72

6: Analysis ofMHC class II-specific T cell clones different individuals and culturing for a week (Protocol 1). This results in T cell activation by alloantigen, where MHC class II is predominantly recognized by CD4+ T cells which undergo autocrine proliferation and clonal expansion, whereas MHC class I is predominantly recognized by CD8+ cytotoxic T cells with lower or no autocrine proliferative capacity. In MLC between donors' PBMC mismatched for HLA-class I and class II antigens, most proliferative reactivity is stimulated by HLA-DR and most cytotoxicity is directed towards HLA-A and HLA-B. By selecting donors matched at HLA-DR, reactivity to HLA-DQ and HLA-DP can also be detected, particularly if the cells are cultured for 10 days and then restimulated with the same donor cells (secondary MLC). Protocol 1. Performance of MLC Method 1. Select donors on the basis of their HLA types and prepare PBMC under sterile conditions. Irradiate stimulator cells at 20 Gy to prevent cell division but retain viability. Alternatively, in the absence of irradiation capability, mitomycin-C can be used at 25 ug ml-1 for 30 min at 37°C, followed by washing three times. Do not use chemically fixed PBMC. 2. Mix stimulator cells with the unirradiated responders at a ratio of 1:1 at 1 x 106/ml each and culture in 200 ul aliquots in flat-bottom microtiter plate wells (7 mm diameter, tissue culture-treated), or in 2 ml aliquots in 16 mm-diameter cluster plate wells. The cultures can be scaled up to larger culture vessels, but so many cells from the MLC are rarely required. 3. For alloactivation across an HLA-DR disparity, incubate at 37 °C for 7 days and then harvest cells by resuspending gently with a pasteur pipette. For alloactivation aimed at generating HLA-DQ- or -DP-reactive cells, incubate for 10 days, then restimulate with the same donor's cells by removing half the culture medium in the well and adding new stimulator cells in fresh medium. Continue the cultures for another 4 days, then harvest the cells. 4. Assess the proportion of lymphoblastic (large) cells in the harvested populations, and accept for cloning those cultures containing a high proportion of blasts. If necessary, enrich the latter by centrifugation over 15-38% Percoll gradients or by positive enrichment of T cells expressing IL-2 receptors. The cells are now ready for cloning.

T cells for cloning can also be activated in variants of the MLC using different types of stimulator cells, for example tumour cells, for specific purposes. 73

Graham Pawelec et al. For MHC analysis, useful stimulators might be mutant B-lymphoblastoid B cell lines (B-LCL) expressing restricted MHC antigens, or cells transfected with human MHC genes. In the latter case, most host cells are xenogeneic, either mouse L cells or hamster CHO cells (both fibroblastoid). These have the advantage of focusing the immune response onto a preselected highly limited human target antigen, but at the same time have the disadvantages (1) that they may lack human-specific peptide antigens required for alloreactivity, (2) that they may lack co-stimulatory structures required for activation, and (3) reciprocally, they may express potentially stimulatory xenoantigens. However, these apparent disadvantages can also be viewed as advantageous in providing a useful system for studying these very questions, i.e. which component of the alloimmune response is dependent on specific peptides bound to MHC (see Chapter 11), which co-stimulatory molecules are required for activation? Transfectants expressing human class II molecules alone are commonly poor stimuli of primary alloreactivity. There have been isolated reports in the literature suggesting that stimulation may occur, but specificity was not rigorously demonstrated (7). Using CHO transfectants, it has been our experience that cells expressing HLA-DR fail to stimulate primary MLC, but that double transfectants expressing the accessory molecule B7 in addition to HLA-DR are quite potent stimulators. However, the level of stimulation is drastically reduced when using purified T cells instead of PBMC as responders, suggesting that the majority of the activation of PBMC is via the so-called indirect pathway for alloreactivity. The indirect pathway is the term for presentation of foreign MHC peptides by autologous antigen presenting cells (APC), which are present in PBMC but not purified T cell populations. The existence of the alternative pathway was first formally demonstrated in an important paper from Termijtelen's group, who showed that synthetic peptides representing the third hypervariable region of HLA-DR could be recognized in the context of autologous HLA-DP, and that peptide-primed T cells could recognize allogeneic stimulators in the absence of exogenous peptide if they expressed both the relevant DR allele and the appropriate HLA-DPencoded restriction element (8). The use of purified T cells as responders and double transfected CHO cells as stimulators in contrast enables the direct pathway of alloreactivity to be investigated in isolation from the confounding factors of the indirect pathway. Synthetic peptides or isolated fractions of cells or endogenous peptides previously bound to MHC molecules can be used to reconstitute the response of purified T cells to such transfectants. A reliable method for separating T cells to high enough purity is magnetic separation involving coupling of magnetic particles to cells via antibody and the removal of these cells by a magnet. Of the several approaches available, negative selection of T cells by 'Dynabeads' (Dynal, Oslo) has in our hands proved itself best thus far. Negative selection, in which non-T cells are selectively depleted from the mixed population, 74

6: Analysis ofMHC class II-specific T cell clones rather than positive selection with T cell antibody, is essential to obtain pure populations of resting T cells which have not been pre-activated in any way by antibody binding (see Protocol 2). Protocol 2. Negative selection of T cells Method 1. Choose antibodies for selectively depleting non-T cells. A cocktail of CD14 (expressed by monocytic cells, macrophages, dendritic cells), CD16 (on natural killer cells), CD19 (B cells and B cell precursors), and HLA-DR (monocytes, B cells, activated T cells, and dendritic cells) has proved appropriate, but other combinations could certainly be employed. 2. Incubate 107 cells/ml at 4°C for 30 min with ~10 |j,g/ml of each antibody, centrifuge, wash twice, and resuspend in 1.5 ml phosphate buffered saline with 0.1% BSA. 3. Add ~108 washed Dynabeads in 0.5 ml and incubate at room temperature for 1 h with occasional gentle shaking. Add 2 ml of PBS and slide the tube gently into the magnetic field for 1-2 min. Gently aspirate the supernatant containing negatively selected cells not held by the magnet, wash twice, and check for purity by indirect immunofluorescence with anti-T cell antibody. 4. To remove any remaining contamination by functional accessory cells, treat the population with L-leucyl-L-leucine methyl esther (LME, Sigma). Incubate the cells at 2.5 x 106/ml in 10 mM LME for 45 min at room temperature in culture medium without serum. Wash twice thereafter. Check absence of functional accessory cells by demonstration of lack of proliferative response to mitogens such as phytohaemagglutinin in the absence of added cells, and reconstitution of the response in the presence of B-LCL or CHO double transfectants expressing DR and B7. Note that CHO cells are extremely radioresistant and may require fixation instead of irradiation. Fixation in 0.025% glutaraldehyde for 3 min is suitable (fix after peptide pulsing).

3. T cell cloning Popular ways of isolating single T cells to propagate as monoclonal populations include single cell deposition by a sorter, or, more commonly, limiting dilution plating of highly diluted cell suspensions (see Protocol 3). In the latter case, control plates must always be set up to check that expected dilutions have in fact been made. In common with most other normal somatic cells,

75

Graham Pawelec et al. isolated T cells will not grow in culture wells, even with supposedly optimal culture media unless they are offered a 'feeder cell bed'. The function of the feeder cells is not completely clear, but presumably includes the provision of a three-dimensional correlate of the in vivo situation where cells interact and function normally only when in contact with other cells. In addition, the feeder cells may provide nutrients and factors lacking in the culture media, as well as containing a source of (allo)antigen presenting cells required for T cell restimulation. The latter is necessary, because normal T cells cannot be maintained in culture without intermittent reactivation via their receptors for antigen (or alternative activation pathways). The presence of T cell growth factors alone does not suffice to maintain the T cells in a constantly proliferating state. Protocol 3. Cloning human T cells by limiting dilution Method 1. Dilute cells to be cloned in appropriate culture medium (see section 4), such that 10 ^l contain 45, 4.5, or 0.45 cells. Introduce this amount of the 0.45 suspension to 60 x 1 mm-diameter wells of culture trays (Terasaki plates') and leave in a vibration-free area for 1 h. Check the distribution of cells in the wells using an inverted microscope and being careful to look around the edges of the wells. According to the Poisson distribution, only a maximum of 37% of the wells should contain cells. Re-adjust dilutions if necessary, and re-check. 2. Plate at least 5 trays with the 0.45 cells/10 ul suspension, one with 4.5 and one with 45, and add a constant number of feeder cells to each well. Irradiated PBMC are commonly used as feeder cells at 1 x 104/ well. Use PBMC from either the donor of the original MLC stimulating cells, or cells from a donor known to express the particular MHC specificity against which clones are sought. Alternatively, a standard stimulator pool of irradiated PBMC from 20 donors can be used. The latter is convenient because large batches of pre-irradiated feeder cells can be prepared from buffy coats and frozen (see Section 5) aliquotted in appropriate amounts. Such a pool works well for clonings with fairly common MHC types (whether antigen-specific or allo-specific), but of course may not be appropriate for rare specificities. 3. Stack plates and wrap in aluminium foil for ease of handling and as a precaution against contamination. Incubate for about a week and then examine the plates using an inverted microscope. Transfer contents of positive wells (more than one-third full) to 7 mm-diameter flatbottom microtiter plate wells with fresh medium and 1 x 105 feeder cells. Check Terasaki plates again at intervals of a few days up to 2-3 weeks of age to identify late developers and transfer these also. Check 76

6: Analysis ofMHC class II-specific T cell clones microtiter plates every few days, and identify wells becoming crowded before a week has passed since transfer. These must be split 1:1 into new culture wells and refed with medium (not feeder cells). After one week in microtiter plates, contents of wells with growing cells are transferred to 16 mm-diameter cluster plate wells with 2-5 x 105 of the same feeder cells, and fresh medium. Observe after 3-4 days and establish which wells are already full or nearly full. The former should be divided into four, the latter into two, with fresh media, but no more feeders. After a total of one week in cluster plates, count the number of cells in each clone and split to 2 x 105/well, again with 2-5 x 105 feeders/well and fresh medium. Feed after 3-4 days with fresh medium, and split again if necessary. Clones successfully propagated in cluster plate wells for this second week are taken to be established. At this point, some (or all) can be cryopreserved at a young age (see Section 5) and the remainder cultured under different conditions to establish optimal parameters for each particular clone. Having a frozen stock enables one to test different culture conditions in order to optimize growth, without the fear of losing the whole clone. 4. Propagate established clones with feeder cells consisting of 80 Gy irradiated B-lymphoblastoid cell lines (B-LCL) instead of PBMC feeders. Most TCC flourish on B-LCL alone, but some appear for unknown reasons to benefit from the presence of PBMC as well (this is especially true during cloning). Of course, propagation of the TCC on PBMC feeders can be continued, but many laboratories may find it easier to grow large amounts of B-LCL than to isolate the PBMC. The international availability of well characterized MHC homozygous B-LCL makes it possible to match the feeder cell to the specificity of the TCC being propagated and enhance the antigen-presentation function of the feeders. 5. As a matter of convenience, it is easier to grow TCC in scaled-up culture vessels than in cluster plates, but not all clones can be adapted to growth in flasks. This must also be tested for each clone, using 1 x 105 and 5 x 105 /ml TCC with an equal number of feeders in tissue culture flasks. Those clones not growing under these conditions can be rarely adapted to flask growth by altering the amounts or concentrations of TCC or feeders seeded or by increasing or decreasing the frequency of stimulation and/or feeding. It remains unknown why some TCC fail to flourish in flasks. 6. Establish restimulation parameters for each clone. T cells require periodic reactivation in order to retain responsiveness to growth factors, and this can be accomplished specifically or non-specifically. All clones can be propagated with weekly restimulation; some but not all can be propagated with restimulation only every 2 weeks. The

77

Graham Pawelec et al. Protocol 3. Continued former tend to become anergic earlier than the latter (9). Therefore, for applications requiring measurement of autocrine proliferation, establish which clones can be cultured with fortnightly restimulation and concentrate on these.

4. Culture media Since TCC have of necessity to be maintained for extended periods in culture, it is essential to ensure the optimal constitution of culture media. Practically all laboratories use either RPMI 1640 or Dulbecco's MEM as basal media, to which various supplements are added. Unfortunately, the most important of these supplements is serum, most usually human or foetal calf. Both have the severe drawback that absolutely undefined amounts of unknown substances are being added to the chemically-defined culture media. However, human T cells will grow only to a limited extent in serumfree media formulations, and we have not been able to obtain clones using such media. We have, however, developed a recipe which does allow for cultivation of established clones over several PD (see Protocol 4). This is a further modification of a formulation previously published in a book of this series (10). Protocol 4. Serum-free medium for cultivation of human TCC over limited periods Method 1. Make stock solutions of the basic additives with which to supplement Iscove's modified Dulbecco's MEM: • Fatty acid-free bovine serum albumin (Conn Fraction V) at 2.5 mgml-1 in IDMEM • Human insulin at 5 mgml"1 in 0.01 M HCI • Human transferrin (with iron) at 35 mg mT1 in PBS • Ethanolamine at 20 mM in IDMEM • Linoleic acid at 1 mg ml-1 in ethanol • Palmitic acid at 1 mg ml"1 in ethanol • Oleic acid at 1 mg mT1 in ethanol 2. Add 1 ml of each of the above to 1 litre of IDMEM, being sure to add the BSA first This is the standard SFM, to which should be added (freshly, for long-term culture): • 250 (ug ml-1 of p-cyclodextrin • 20 ngml-1 low-density lipoprotein 78

6: Analysis of MHC class II-specific T cell clones 2 uM L-putrescine 2 mM L-glutamine 2 (j.g ml-1 zinc acetate 2000 U ml-1 bovine liver catalase 1 mg ml-1 N-acetyl-L-cysteine and, if desired, 20 mM HEPES and antibiotics. 3. All media must be sterile filtered through membranes selected for low protein- absorbing capacity.

The last component of vital importance in the culture medium is the exogenous growth factor. Most T cell cloning and propagation protocols limit the TCGF to purified IL-2, usually recombinant, and utilization of supernatants from activated T cells has more or less fallen into disuse. The latter, of course, contain a great variety of cytokines in addition to IL-2. Some of these may be beneficial in generating clones from some types of T cell that may require factors other than IL-2 alone, at least initially. For example, for generating a particular kind of autoreactive TCC (11), it was found that factors present in conditioned media other than IL-2 or IL-2 plus IL-3, IL-4 or interferon--/ were required (12). Since a number of other additional T cell growth factors, including IL-7, IL-9 and IL-12 have now been identified, it would not be surprising if some T cells required these factors for growth and function.

5. Cryopreservation As mentioned in the introduction, human T cell clones generated as described manifest finite lifespans and cannot be maintained in culture indefinitely. In experiments involving the concurrent use of a large number of different TCC, it would also be very impractical to maintain all the necessary lines continuously in culture. For these reasons, human TCC are routinely cryopreserved, so that supplies of young clones can be grown up when required, and so that many clones cultured at different times can be used concurrently (see Protocols). Protocol 5. Cryopreservation of human TCC Method 1. Resuspend washed cells in RPMI 1640 with 40% PCS at twice the concentration desired for freezing (between 2 and 10 x 106 per ml) at room temperature. 2. Make up cryoprotectant stock solution by mixing cold protein-free RPMI with sterile dimethylsulphoxide (DMSO) to give 20% final concentration.

79

Graham Pawelec et al. Protocol 5.

Continued

3. To one volume of cell suspension, add one-half volume of DMSO stock solution in one aliquot at room temperature and mix rapidly. Wait 5 min for the DMSO to equilibrate across the cell membrane (to avoid osmotic damage), and then add a second one-half volume aliquot of DMSO stock to give a final concentration of 10%. Transfer cells to freezing vials, put in a cardboard honeycomb box and place in -70°C freezer as fast as possible. 4. Allow to freeze for at least 4 h, but do not leave at -70°C for longer than a week. For long-term storage, transfer to -196°C liquid nitrogen tanks. 5. To thaw, place in a 37°C water bath, transfer vial contents to centrifugation tubes and add an equal volume of RPMI 1640 at room temperature. Wait 5 min to equilibrate, then add another equal volume of RPMI, immediately centrifuge and resuspend in culture medium.

6. Restimulation of MHC class II-specific TCC and data analysis 6.1 Proliferation assay Most class II-specific T cells are capable of lymphokine secretion and autocrine proliferation. Measuring proliferation by means of incorporation of radioactive DNA precursors remains the most convenient and widespread assay (Protocol 6). Alternatively, activation can be measured by assay of secreted cytokines or by assay of accumulation of cytokine mRNA. Protocol 6. Measurement of proliferation and data analysis Method 1. Dispense responder TCC cells into U-wells of microtiter plates containing 50 ul of 30% human serum at 1 X 104 cells/well in 50 ul of culture medium (generally RPMI 1640). Add irradiated stimulator cells or antigen-presenting cells plus antigen in appropriate amounts (usually 2.5 x 104B-LCL, 1 X 106PBMC, or established optimal numbers of other types of stimulator). Include negative controls: responder cells with medium alone, irradiated stimulator cells with medium alone; and positive control: stimulator cell with non-specific mitogen (eg. 1% PHA) to establish autocrine proliferative capacity of responders. For each responder + stimulator combination, set up cultures in triplicate. 2. Set up replicate sets of plates according to experimental design, to

80

6: Analysis ofMHC class II-specific T cell clones assess the kinetics of the proliferative response. This generally peaks after 1 or 2 days for MHC-specific TCC. Pulse wells with 37 kBq tritiated thymidine (specific activity 185 GBqmmor1) for 16 h and harvest nuclei onto glass fibre filter paper. 3. Assess radioactivity of each sample in an appropriate counter, e.g. by liquid scintillation spectroscopy. Transfer and analyse data using suitable computer programs (see e.g. ref. 13), or by hand. 4. When establishing specificity of e.g. MHC-specific alloreactive TCC, and analysing responses against large panels of defined stimulators (e.g. B-LCL of known HLA type), inspection of the data may not always result in satisfactory assignment of specificities since the data may not group themselves into obviously positive and negative populations. There are several mathematical methods which may help in approaching this. For example, 'centroid cluster analysis' can be employed, which in the simplest case assumes that there are only two clusters of values, i.e. positives and negatives (whether or not more than two clusters are actually possible). Identifying these clusters is done as follows: • First calculate the mean of each possible cluster • From each value within the proposed positive cluster, subtract the mean value and square the difference • Sum these differences • Repeat for the proposed negative cluster • Add these two sums • Move to second possible cluster and repeat, then repeat for all possible clusters • Select proposed clusters with the lowest value of the summed sums In each case, one must bear in mind whether these assumptions can actually be made: there are only two groups, and variations of the means are due only to experimental variables; and that these groups really can be identified by minimizing the sums of the squares (SSQ) of the deviations within the clusters. For example, in some cases there might well be more than two clusters of values, even when working with monoclonal populations. The cluster analysis can be conveniently carried out using a computer program with the following steps: • Assign an ascending number to each of the responder clones • Sort values of clones (CPM) in ascending order, store together with the appropriate responder number • For I = 1 to last number-1 do • mean_1 = 0 :a for Jb = 1 to I: m=tn + value(J): next J : mean_1 = m/l • mean_2 = 0 : for J = I + 1 to last number: n = n + value(J): next J : mean_2 = n/(last number — I) 81

Graham Pawelec et al. Protocol 6.

Continued

• SSQ_1 = 0 : for J = 1 to I : SSQ_1 = SSQ_1 + (mean_1 - value(J)) 2 : next J • SSQ_2 = 0 : for J = I + 1 to last number : SSQ_2 = SSQ _2 + (mean_2 - value(J)) 2 : next J . SSSQ(I) = SSQ_1 + SSQ_2 • next I • Search for the minimum of the SSSQs; its index defines the cluster boundary • Print out all clones with a responding level <= the responding level of SSSQ(I) to the group of negatives • Print out all other clones to the group of positives In practice, the levels of responses can be spread over the whole range from background to maximal. Therefore, the analysis can be perfomed twice, once with arithmetic and once with log-transformed data. Only those values assigned to either positive or negative groups in both analyses are considered unequivocal. The reasons for the existence of equivocal values could be that subgroups actually exist: further experiments would have to establish this, eg. by titrating the amount of stimulating antigen or allogeneic stimulating cells. "':' is a delimiter between commands. b J refers to the field index, not to the numbers.

6.2 Assay of cytokine mRNA by polymerase chain reaction (PCR) On stimulation, TCC incapable of autocrine proliferation may nonetheless secrete certain cytokines, which can be detected by appropriate bioassays, immunoassays, or molecular techniques. Therefore, analysing cytokine production provides a means for investigating activation of a wider variety of T cells than proliferation assays. Bioassays or immunoassays are inconvenient and are not necessary for assessing T cell activation. Detecting the accumulation of mRNA is easier and more sensitive and PCR techniques have become established as the method of choice for this purpose (see Protocol T). There are certain drawbacks, however, due to the very sensitivity of PCR: • T cells can only be analysed after extended periods of resting without exposure to stimulating antigen (2-3 weeks in our hands); • TCC may only be activated by stimulators which are themselves incapable of expressing the cytokines to be analysed. In practice these restrictions are not limiting, since resting the T cells is not problematic (as long as they have sufficient IL-2) and activation can easily be 82

6: Analysis of MHC class II-specific T cell clones assessed by choosing T cell-specific cytokines not expressed by the non-T stimulator cells. Protocol 7. Cytokine PCR assay for assessing activation of alloreactive TCC Method a 1. Rest the cells and then restimulate with B-LCL for 5 h. Isolate RNA thus: • Lyse 5 x 106 cells in 500 ul guanidinium isothiocyanate (Gl) buffer (4 M Gl, 20 mM sodium acetate pH 5.2, 0.1 mM dithiothreitol and 0.5% N-laurylsarcosine). • Add 50(ul of 2 M sodium acetate at pH 4.0 and shake. • Add 500 ul of phenol and then 200 ul of chloroform:isoamyl alcohol (49:1), vortex, then chill on ice for 15 min. • Centrifuge at 10000 X g for 20 min at 4°C. • Aspirate the aqueous upper fraction and mix 1:1 with isopropanol, then allow to precipitate for 30 min at -80°C. • Centrifuge down the precipitate at 10000 x g for 20 min at 4°C and wash with ice-cold 70% ethanol, then dissolve in 20 ul of autoclaved distilled water. • Quantify the amount of RNA by using a photometer to measure absorption at 260 nm. 2. Reverse transcriptase reaction: use 0.5 ug RNA in 13.5 ul distilled water for each reaction, and linearize at 95°C for 5 min. Place on ice. Then add: • 0.5 ul of each deoxynucleotide at 10 mM • 2 ul of X10 PCR buffer (100 mM Tris-HCI at pH 8.3, 500 mM KCI, 15 mM MgCI2, and 0.01% gelatin) • 1 ul of Oligo-dT-Primer (0.5 ugul-1) • 0.5 ul of RNAsin RNase inhibitor (20 U ul-1) • 1 ul of Moloney murine leukemia virus reverse transcriptase (200 Uul-1) 3. Leave for 10 min at room temperature to allow the oligo-dt-primer to bind the RNA, then maintain at 42°C in a water bath for 1 h. Heat to 95°C for 10 min to inactivate the reverse transcriptase. 4. For the PCR reaction, add: • • • •

8 ul of X10 PCR buffer 20 pmol sense primer 20 pmol antisense primer 1 U of Taq DNA polymerase 83

Graham Pawelec et al. Protocol 7. Continued Make up to 100 ul with water and overlay with 80 ul of mineral oil. Put into a thermocycler and denature at 95°C for 3 min, then run 35 cycles as follows: • • • •

1 min denature at 95°C (denaturation) 1 min oligonucleotide binding at 55°C (annealing) 1.5 min polymerization at 72°C (extension) last cycle 5 min at 72°C

5. Examples of sequences of primer pairs useful for measuring T cell activation: Length of product • IL 2 sense: CAGTGTCTAGAAGAAGAACTCAAACC antisense: AAGGCCTGATATGTTTTAAGTGGG • IL 3 sense: ATGAGCCGCCTGCCCGTCCTG antisense: GCGAGGCTCAAAGTCGTCTGTTG • L 4 sense: ATGGGTCTCACCTCCCAACTGCT antisense: CGAACACTTTGAATATTTCTCTCTCAT • IL 5 sense: GCTTCTGCATTTGAGTTTGCTAGCT antisense: TGGCCGTCAATGTATTTCTTTATTAAG • GM-CSF sense: ACACTGCTGAGATGAATGAAACAGTAG antisense: TGGACTGGCTCCCAGCAGTCAAAGGGGATG • IFN-g sense: ATGAAATATACAAGTTATATCTTGGCTTT antisense: GATGCTCTTCGACCTCGAAACAGCAT

276bp 449 bp 456 bp 291 bp 286 bp 501 bp

These cytokines are transcribed in essentially all human T cells (at least in alloreactive TCC). 6. Carry out DNA gel electrophoresis to analyse size of amplified product: • Dissolve 4 g of agarose in 200 ml of TBE buffer (90 mM Tris-HCI, 90 mM boric acid, 2 mM Na2-EDTA, pH 8), boil briefly in microwave oven, cool to 60°C in a magnetic stirrer, and pour into gel chambers • Electrophorese for 2 h at 70 mV using 20 (jJ of PCR-product and 5 ul of 5x loading buffer (50 mM Tris-HCI at pH 7.6, 50 mM EDTA at pH 7.6, 0.5% SDS, 0.1% bromophenol blue, 40% sucrose) compared to molecular weight standards • Stain gel with 0.5 ugml-1 ethidium bromide for 20 min and photograph under UV 7. Confirmation of amplification specificity by Southern blot: • Denature DNA in agarose gels by 30 min incubation in 0.5 M NaOH and 1.5 M NaCI, followed by washing in water and neutralization in 0.5 M Tris-HCI, 1.5 M NaCI and 0.001 M EDTA. Make sure that the gel remains in two changes of neutralization buffer at least as long as in denaturation buffer 84

6: Analysis of MHC class II-specific T cell clones • Wash the gel with water and lay on Whatman 3 mm paper soaked in 20x SCC (3 M NaCI and 0.3 M sodium citrate at pH 7.0) • Cover with Hybond N+ membrane, and then absorbent paper to a depth of about 5 cm, and leave for 3 h under 1 kg pressure • Fix membrane by laying Whatman paper soaked in 0.4 M NaOH on it for 20 min, then wash briefly with 5x SSC and hybridize with the appropriate cDNA probe • Visualize hybridization using standard radioactive methods or with non-radioactive methods a For any procedure involving PCR, all solutions and containers must be rigorously RNasefree, and gloves must be worn at all times.

7. Sequencing T cell receptors for antigen An important aspect of the analysis of MHC-specific T cells in modern immunology concerns identification of the receptors involved in the alloreactive repertoire. Until recently, this was approached using specific oligonucleotide sequences from known variable region families, using protocols similar to the one described above for cytokine mRNA analysis. This method is cumbersome and can only detect known receptors. Recently, protocols have been established for direct sequencing of TCR in monoclonal populations. One convenient method has been applied to alloreactive TCC in a collaborative project (14, 15), and is given here (Protocol 8): Protocol 8. TCR variable region sequencing utilising a Vb universal PCR primer Equipment and reagents • 10x PCR buffer: 100 mM Tris-HCI (pH 8.8 at 25°C), 500 mM KCI, 25 mM MgCI2, 1 mg ml-1 gelatin • PCR mixture 1: double stranded cDNA, 2.5 ul; 100 uM VBUN (sense primer), 2.5 ul; 25 uM FPR27 (antisense), 0.5 ul; 5 mM of each dNTP, 1 ul; 10x PCR buffer, 2.5 ul; water to 24.6 ul • PCR mixture 2: double stranded DNA, 1 ul; 25 uM FPR19 (sense), 0.5 ul; 25 uM FPR15 (antisense), 0.5 ul; 5 mM of each dNTP, 1 ul; 10x PCR buffer, 1 ul; water to 24.6 ul

• Oligo(dT) bead suspension (Dynabeads Oligo (dT)25 mRNA isolation kit, from Dynal, Oslo) • 5x first strand buffer: 250 mM Tris-HCI (pH 8.3 at 42°C), 40 mM MgCI2, 150 mM KCB, 50 mM dithiothreitol « 5x second strand buffer: 100 mM Tris-HCI (pH 7.5), 25 mM MgCI2, 50 mM (NH4)2 S04, 500 mM KCI, 250 (ug/ml BSA, 200 uM each of the dNTP « 5x kination buffer: 250 mM Tris-HCI, pH 8.0, with 50 mM MgCl2, 25 mM dithiothreitol and 25% glycerol)

A. Isolation of mRNA 1. Centrifuge down 1-3 x 106 TCC and lyse by adding 100 ul of lysis buffer (provided in kit) along with RNase inhibitor (1 Uml"1 RNasin), mix well and vortex, then cool on ice for 1 min. 85

Graham Pawelec et al. Protocol 8. Continued 2. Centrifuge at 15000 x g for 30 sec at 4°C 3. Add lysate to 100 ul of oligo(dT) bead suspension (350 ug, prewashed with 2x binding buffer (supplied in bead kit). 4. Suspend the beads and allow the RNA to bind to oligo(dT) at room temperature for 5 min, then wash twice with 200 ul of washing buffer, holding the beads with a magnet. 5. Add 32 ul of diethylpyrocarbonate (DEPC)-treated water and 8 ul of elution buffer (from the bead kit) and incubate at 65°C for 2 min, then recover the eluate containing the mRNA and transfer to new tube. B. Synthesis of TCRb cDNA 1. Prepare first strand mixture for synthesis of first strand: RNA solution, 10-20 ul; 40 Uml-1 RNasin, 0.5 ul (containing 20 U); 25 uM FPR15, 4 ul (2 uM); 5x 1st strand buffer, 10 |xl; water to bring the volume of the mixture to 44.5 (jJ. 2. Heat at 65°C for 3 min, then cool on ice and add 4 JJL! each of 25 mM dNTP (dATP + dGTP + dCTP + dTTP), 0.5 jJ of RNasin (20 U) and 1 pil of reverse transcriptase (25 U of RAV2). 3. Incubate at 42 °C for 60 min. 4. Prepare second strand mixture: first strand DNA mix, 50 pJ; 5X 2nd strand buffer, 20 jxl; RNase H (in 1x 2nd strand buffer), 1 |o,l (1 U); £. co//DNA poymerase, 4 jtl (24 U); water, 25 p\. 5. Incubate at 12°C for 60 min, then at 22°C for a further 60 min. 6. Stop reaction by adding 11 |o.l of 0.1 M EDTA (pH 7.5) to give final concentration of 10 mM. 7. Extract once with phenol:chloroform:isoamyl alcohol (25:24:1) with 8-hydroxy-quincline as antioxidant at 0.1% w/v of phenol, then centrifuge at 10000 X gfor 5 min. 8. Add 10 M.g glycogen as a carrier, 75 (xl of 5 M NH4-acetate, and 470 n.l of ethanol. Mix and leave at -20°C for 1 h. 9. Centrifuge 15000 x g for 10 min. Remove supernatant, rinse precipitate with 800 p.1 of 70% ethanol and dry under vacuum. 10. Dissolve precipitate in 25 p. I of T^o.i (1 mM Tris-HCI, pH 7.5 and 0.1 mM EDTA, pH 7.5). C. Polymerase chain reaction 1. Prepare two separate PCR mixtures with two different combinations of sense and antisense primers. The PCR mixture 1 is for amplification of Vp, Dp + N, Jp and the beginning of Cp for direct sequencing. The Vp universal primer designated VBUN anneals with a conserved

86

6: Analysis of MHC class II-specific T cell clones sequence of all known Vp genes. The PCR mixture 2 is for amplification of TCR Cp regions to confirm successful cDNA synthesis. 2. Incubate the PCR mixtures for 5 min at 98°C, then add 0.4 |jj (2 U) of Taq DNA polymerase and overlay with 40 (J of mineral oil 3. Start PCR. First cycle: annealing (55°C for 3 min); extension (72°C for 3 min). Following 29 cycles: denaturation (97°C for 1 min); annealing (55°C for 1 min); extension (72°C for 3 min). Last cycle: extension (72°C for 3 min). 4. Check amplification by running 2.5 JJL! of PCR mixtures on a 2% agarose gel. The PCR mixture 1 must give a major DNA fragment around 320 bp. The product size varies (300-340 bp) depending on the length of Vp and Dp + N. In addition to the main product, other minor DNA fragments or smears can be expected in some cases. The PCR mixture 2 must give a 383 bp Cp DNA fragment. D. Purification of the PCR product 1. Apply the remaining PCR product from mixture 1 on 1.8% low melting point agarose gel (SeaPlaque, FMC Bioproducts, Rockland, ME). 2. Cut out the region of the gel containing the major PCR product and transfer to Eppendorf tube. 3. Melt the gel by heating at 70°C for 10 min, vortexing occassionally. 4. Recover DNA from the melted gel. This can be accomplished in several ways. A convenient way is to use a commercial kit, e.g. Magic PCR Preps DNA Purification System, Promega, Madison): (a) Add 1 ml of resin solution (from the kit) to the melted gel. Vortex for 1 min. Transfer slurry to a 2.5 ml syringe barrel equipped with a mini column (from the kit) and remove liquid by depressing syringe plunger. (b) Wash twice with the solution provided in the kit and elute DNA from the resin with 50 (xl of TiE0.i. (c) Add 10 n,g of glycogen, NH4-acetate (final concentration, 2 M), and a 2.5-fold volume of ethanol. Leave tube for 1 h at -20°C. (d) Centrifuge at 15000 x gfor 10 min. Remove supernatant. Rinse with 70% ethanol and dry. (e) Dissolve the precipitate in 10 p.l T^Q.I. Do not be concerned about the large amount of insoluble material remaining. E. Sequencing the product 1. Prepare a labelling mixture consisting of: 2 |j,M FPR27 (0.5 ^1); 222 kBq 32P--y ATP (ca. 220 Tbqmmol-1) (0.6 M-D; 5x kination buffer (0.5 nJ); T4 polynucleotide kinase (0.1 nl) (2 U); water (0.8 |xl). 87

Graham Pawelec et al. Protocol 8.

Continued

2. Incubate at 37°C for 30 min, then boil for 2 min to inactivate the T4 polynucleotide kinase. 3. Prepare the sequence reaction mixture: DNA sample (2—8.7 |xl); 32Plabelled sequence primer (2.5 fJ); 10x PCR buffer (1.3 nJ); Taq DNA polymerase (0.5 JJL!) (2.5 U); water (0-6.7 |J). 4. Mix and centrifuge down. 5. Add 3 (j,l of the mixture to each of four tubes containing 2 |o,l of either A, G, C or T sequence mix (where A = 25 uM dNTP and 835 uM ddATP in 1 x PCR buffer; G = 25 uM dNTP and 53 uM ddGTP in 1x PCR buffer; C = 25 txM dNTP and 308 |xM ddCTPin 1 x PCR buffer; T = 25 uM dNTP and 1,250 uM ddTTP in 1 x PCR buffer). 6. Mix, centrifuge down and overlay with 8 ul of mineral oil. 7. Start cycle sequencing. First 20 cycles: 97°C for 30 sec; 55°C for 30 sec; 72°C for 1 min. Last 10 cycles: 97°C for 30 sec; 72°C for 1 min. 8. Add 5 nl of stop solution: 95% formamide, 20 mM EDTA (pH 7.5), 0.05% bromophenol blue and 0.05% xylene cyanol. 9. Incubate at 80°C for 3 min, then on ice for 5 min. 10. Analyse on urea/polyacrylamide (6%) sequence gel. T cell clones expressing single TCRp transcripts will give readable sequence ladders. TCC expressing more than one major TCRp transcript will give unreadable mixed prolfiles. However, in this case, it is usually possible to identify two candidate Jp segments. Two TCRp sequences can be determined separately using squence primers specific for each of the two Jp segments instead of FPR27. Sequences for the primers used are as follows: VBUN: 5'- GGGG (AGCT) (AGCT) (AGCT) (AGC) T (ACT) T (ATC) (CT) TGGTA-3' FPR15: 5'-AAGCCACAGTCTGCTCTACC-3' FPR19: 5'-TGTTCCCACCCGAGGTCGCT-3' FPR2 7: 5' -GGAGATCTCTGCTTCTGATG-3'

Acknowledgements This work was supported by the European Union Concerted Action on the Molecular Biology of Immunosenescence, EUCAMBIS; grants SFB 120 A1 and Pa 361-1/1 from the Deutsche Forschungsgemeinschaft; grant W/2/91/Pal from the Mildred Scheel-Stiftung; award EU-93-1-002 from the Sandoz Foundation for Gerontological Research; grant S117 from the Arthritis and Rheumatism Council of the UK; an ARC grant from the Deutsche 88

6: Analysis of MHC class Il-specific T cell clones Akademische Austauschdienst and the British Council and by grant 02670718 from the Japanese Ministry of Education.

References 1. Bach, F.H., Inouye, H., Hank, J.A., and Alter, B.J. (1979). Nature, 281, 307. 2. Kahle, P., Wernet, P., Rehbein, A., Kumbier, I., and Pawelec, G. (1981). Scand. J. Immunol., 14, 493. 3. Pawelec, G. and Wernet, P. (1980). Immunogenetics, 11, 507. 4. Perillo, N.L., Naeim, F, Walford, R.L., and Effros, R.B. (1993). Mech. Ageing. Dev., 67, 173. 5. Perillo, N.L., Naeim, F., Walford, R.L., and Effros, R.B. (1993). Exp. Cell. Res., 207, 131. 6. Hayflick, L. (1992). Exp. Gerontol., 27, 363. 7. Nakatsuji, T., Inoko, H., Ando, A., Sato, T., Koide, Y., Tadakuma, T., Yoshida, T.O., and Tsuji, K. (1987). Immunogenetics, 25, 16. 8. Saskia De Koster, H., Anderson, D.C., and Termijtelen, A. (1989). J. Exp. Med., 169, 1191. 9. Pawelec, G., Brocker, T., Busch, F.W., Buehring, H.-J., Fernandez, N., Schneider, E.M., and Wernet, P. (1988). J. Mol. Cell. Immunol, 4,21. 10. Pawelec, G. (1993). Cloning and propagation of human T lymphocytes. In Tumour immunobiology. a practical approach (ed. G. Gallagher, R. C. Rees and C. W. Reynolds), pp. 131-41. IRL Press, Oxford. 11. Pawelec, G., Fernandez, N., Brocker, T., Schneider, E.M., Festenstein, H., and Wernet, P. (1988). J. Exp. Med., 167, 243. 12. Pawelec, G. (1991). Cell. Immunol, 134, 265. 13. Schaudt, K. and Pawelec, G. (1991). J. Immunol. Methods., 138, 155. 14. Obata, F., Tsunoda, M., Ito, K., Ito, L, Kaneko, T., Pawelec, G., and Kashiwagi, N. (1993). Hum. Immunol, 36, 163. 15. Obata, F., Tsunoda, M., Kaneko, T., Ito, K., Ito, L, Masewicz, S., Mickelson, E.M., Oilier, W.E.R., Pawelec, G., Cella, M., Ferrara, G.B., and Kashiwagi, N. (1993). Immunogenetics., 38, 67.

89

This page intentionally left blank

a Methods for studying the role of the major histocompatibility complex in NK cell cytotoxicity J. PENA, R. SOLANA, F. BORREGO, and B. ROLSTAD

1. Introduction Our knowledge about how T cells recognize foreign cells has been increased through studies of the interactions between the T cell receptors (TCR) and antigenic peptides presented by class I or class II molecules. The role of the MHC is now equally challenging in NK cell research, as it has been shown that NK cells can recognize certain MHC class I specificities. The receptors that NK cells utilize to recognize class I antigens have recently been identified, and named KiR (killer inhibitory receptors). An attractive idea supported by experimental data is the 'missing self hypothesis (1, 2) which envisages NK cells as a surveillance mechanism against cells showing reduced or aberrant expression of certain MHC class I molecules regarded as 'self by the immune system (for reviews see refs. 3 and 4). Much of our knowledge about the role of class I molecules in NK cell recognition has been based on studies of tumour target cells with loss of class I molecules. More recently class I transfected tumour target cells, or cells from class I knockout or transgenic mice, have shed important light on NK recognition mechanism (5, 6). Although these studies have been informative with regard to the central role of class I molecules, a limitation is that they have employed either neoplastic cells, which may expose a number of antigens other than class I, or effector or target cells from genetically manipulated animals. Three experimental systems have used normal lymphoblasts or bone marrow cells as target cells for NK cells in vitro, namely allo-reactive NK cell clones by Moretta et al. (7), allogeneic lymphocyte cytotoxicity (ALC) in rats (8), and the recently described in vitro assay for F1 hybrid resistance in the mouse (9). These systems have enabled immunogenetic mapping of target cell specificities, and some of these data have challenged the 'missing self hypothesis.

J. Pena, R. Solana, F. Borrego and B. Rolstad Here we describe how NK cell reactivity against autologous or allogeneic target cells can be detected and measured by in vitro tests. We include some information on NK cells markers and biodistribution, experimental protocols for isolation and purification of NK cells, and how the reactivity of these cells can be measured in simple cytolytic assays using either normal or neoplastic cells as target. The most extensive analysis of the immunogenetics of NK reactivity against parental or allogeneic bone marrow cells has been carried out in the mouse. However, since no in vitro assay system for measuring NK reactivity against such target cells has been available in this species until recently, and since the authors' personal experience is with human and rat NK systems, we will focus our description on these latter species.

2. Definition of NK cells NK cells are lymphocytes that do not rearrange immunoglobulin (Ig) or TCR genes. Thus, neither Ig nor the TCR/CD3 complex is expressed at the cell surface, except for the zeta (£) chain. The most characteristic cell surface markers are CD16 (Fc_RIIIA) and H CD56 (N-CAM) in human NK cells (10), NKR-P1 in the rat (11), and NK1.1 in the mouse. The latter two molecules are members of the NKR-P1 gene family (12), which belongs to the NK complex and encodes for membrane proteins with C-type lectin domains (13). These latter molecules are mainly expressed in NK cells, although sub-populations of T cells may also show weak expression.

2.1 Cell surface markers Depending on the species, NK cells share cell surface markers with T cells (CD2, CD8, and CD43 in the rat and man) and also express members of the b2 integrin family shared with myelomonocytic cells (e.g. CDllb/CD18) (14). Human NK cells are CD3e negative, but the £ chain, which in T cells is associated with the TCR, is coupled to the CD16 receptor in NK cells (10). Other surface molecules expressed in human NK cells or NK subpopulations are: CD7, CD26, CD27, CD29, CD43, CD45, CD56, CD57, CD69, CD81, CD94 (Kp43) and CD122 (interleukin-2 (IL-2) receptor p-chain), p38 (C1.7.1), and PEN5.

2.2 Tissue distribution and turnover NK cells represent 5-15% of lymphocytes in human blood, from where they can be isolated and their functional activity tested (Protocols 1 and 2). In the rat and mouse the spleen is a good source of NK cells, and NK cells can also be isolated from the liver (15). Unlike T and B cells, NK cells do not recirculate between blood and lymph nodes, and a majority of them have a relatively short physiological life span in the periphery, in the order of one to two weeks. 92

7: Methods to study the role of the major histocompatibility complex Protocol 1. Equipment and reagents • Na51CrO, (CEA-ORIS, Saclay, France) • 96-well round bottom microtiter plates (Costar, Cambridge, MA) • Gamma counter (Ultrogamma LKB, Bromma, Sweden) • Cell culture medium: RPMI 1640 (Bio-Whittaker, Walkersville, Maryland) with 25 mM HEPES buffer (Sigma, St. Louis, MO), 2 mM L-glutamine (Bio-Whittaker), penicillin/ streptomycin (Bio-Whittaker) and 5% fetal calf serum (FCS) (Bio-Whittaker France)

. Triton X-100 • Target cells: PHA (human) or Con A blasts (rat), or cultured K562, Raji, C1R, P815, YAC-1, etc. • Effector NK cells (freshly isolated human, see Protocol 2) or IL-2 activated NK cells (rat, see Protocol 2) or alloreactive NK cells lines or clones (human, see Protocol 3) • Titertek microtiter plate harvesting system (Skatron, Lyerbyen, Norway)

A. NK cytotoxic assay in the human and the rat 1. Prepare effector cells according to Protocol 2 or 3. 2. Incubate target cells (1-10 X 106 cells/ml) with 100-200 u,Ci Na51CrO4 for 1-1.5 h in culture medium at 37°C in 5% C02. 3. After washing the target cells three times in medium resuspend the cells in culture medium to 5 x 104 cells/ml (human) or 105 cells/ml (rat). 4. Perform the cytotoxic assay in triplicate at each effector/target cell ratio by mixing 5 x 103 (human) or 104(rat) target cells with two fold dilutions of effector cells producing different effectortarget cell ratios in a final volume of 200 |xl. Centrifuge the plates for 1 min at 300 x g. 5. For spontaneous release samples of target cells resuspend in culture medium alone. 6. Obtain maximum 51Cr release (human) or total 51Cr (rat) incorporated into target cells by mixing 100 p,l 5 x 103 labelled target cells with 100 uJ 2% Triton X (human) or counting the total radioactivity the target cells (rat). 7. After 4 h incubation at 37A°C in 5% C02, remove 100 |xl of supernatant from the wells after centrifugation of the plate at 900 r.p.m. for 3 min. Alternatively, measure the radioactivity in approx. 85% of the supernatant collected in filters with the Skatron microplate hervesting system. Count the radioactivity in a gamma-counter. 8. Calculate specific 51Cr-release with the following equation: Exp. release - Spontaneous release (Max or Total) - Spontaneous release 9. 'Max' in the above equation refers to the radioactivity in 100 p.! supernatant from Triton X treated cells. Total radioactivity in the cells by counting. The latter method is preferred when the radioactivity in the cells is low, e.g. when using small lymphocytes as target cells. 93

/. Pena, R. Solana, F. Borrego and B. Rolstad Protocol 1. Continued B. The generation of target lymphoblasts 1. Rat Con A blasts are generated from spleen cells separated on Lymphoprep for 30 min at 400 x g (Nycomed, Pharma, Oslo, Norway). 2. Adjust the mononuclear cells to 2 x 106 cells/ml and culture for 2-3 days in RPMI 1640 with 5% normal rat serum, 2 mM L-glutamine, 1 mM Na-pyruvate, 5 x 10-5 M 2-mercaptoethanol and 5 ugml-1 of Con A. Before use, remove dead cells by Lymphoprep separation.

2.3 Cytokine production and response to cytokines NK cells produce and secrete a variety of cytokines, such as several of the interleukins, interferon-a (IFN-a), interferon-y (IFN--y), tumour necrosis factor a and (3 (TNF-a and TNF-0), as well as various colony stimulating factors (G-CSF, GM-CSF) and TGF-pl (15). The production of this plethora of cytokines indicates that NK cells have important regulatory and/or developmental roles in both T- and B-cell mediated immune responses and in other parts of the haemopoietic system. They also respond to a variety of cytokines such as IFN-a, IFN-p, IL-2, IL-4, IL-7, and IL-12 by a change in function (e.g. proliferation, cytolytic activitity, or cytokine production). This can be exploited experimentally, e.g. in the generation of large numbers of activated NK cells in vivo or in vitro with IL-2 or in the generation allospecific NK clones or lines.

2.4 Functions Apart from the production of cytokines, the most salient feature of NK cells is their ability to spontaneously kill, without previous sensitization, a variety of target cells, including both neoplastic and virally infected cells. They very likely play a role as first line defence against both viral infections and neoplasms. One effector function of NK cells related to the adaptive immune system is that NK cells can recognize antibody-coated target cells with their Fc receptors (CD 16) and are therefore the effector cells of antibody-dependent cellular cytotoxicity (ADCC). NK cells activated with IL-2 increase their lytic potential and the repertoire of target cells recognized also becomes broader, to include most all neoplastic cells. However, it has also been shown that NK cells can spontaneously recognize and kill MHC mismatched normal differentiated cells, especially within the haemopoietic system (9), which indicates a role in bone marrow allograft rejection. It has also been shown that a given NK cell population can discriminate between different allospecificities in MHC mismatched target cells, which probably reflects a division of NK cells into clones or subsets with defined allospecificities. 94

7: Methods to study the role of the major histocompatibility complex

2.5 Receptors 2.5.1 Activation receptors Some surface molecules expressed on NK cells may act as activation receptors, e.g. CD16 capable of triggering ADCC (15). In humans NK cells the binding of CD16 to Fc of immunoglobulin on the target cell leads to a rise in intracellular Ca2+, increased phosphoinositide turnover, and activation of the lytic machinery (16). However, a relevant role in natural killing has not been demonstrated for this molecule, since NK cells lacking CD16 can still perform natural cytotoxicity. In the mouse and rat NK1.1 and NKR-P1 are also activation molecules since binding of specific antibodies to them leads to similar activation signals (11) and also stimulates redirected lysis. It has been recently shown that p50, an isoform of p58, can act as a NK activatory receptor. 2.5.2 Inhibitory receptors NK cells have inhibitory receptors for MHC-controlled antigens, and these receptors divide NK cells into functional subsets. In the human, Moretta and colleagues have described a family of clonally distributed receptors on NK cells (p58) using the mAbs GL183 y EB6, which define subsets of NK cells differing in their allospecificities (7). Other recently described receptor on human NK cell is NKB1 (p70), which is the receptor for the HLA-B5101, B5801 ANA B2705 (17,18). Other receptor for MHC molecules are disulfidelinked heterodimers of CD94 and NKG2 subunit (23). A common feature of these NK inhibitory receptors is the presence of ITIM sequences in the cytoplasmic domain. In the mouse the Ly-49 gene family, genetically coupled to NKR-P1 region within NK complex, codes for membrane proteins which, like the NKR-P1 family, are C-type lectin membrane proteins (13). Two antibodies, anti Ly-49 (19) and 5E6 (20) define members of this family expressed on subsets of mouse NK cells and associated with certain MHC class I specificities. The function of Ly-49, p58, Kp43 (CD94) and NKB1 as membrane receptors remains elusive, but there are indications that they are inhibitory receptors when there is interaction between those receptors and specific MHC alleles on target cells. In mouse Ly-49+ cells bind to purified H-2Dd or H-2Dk molecules and this binding is prevented by anti-Ly-49 antibodies (21, 22). Table 1 shows some of these inhibitory receptors described in humans and mice.

2.6 Generation of IL-2 activated NK cells, lines and clones in vitro The relative paucity of NK cells within the blood and spleen makes it difficult to obtain these cells in pure form and in a number large enough to perform functional tests. Since alloreactive NK cells can be kept functionally active in culture for several days with IL-2, a simple procedure in the rat is to isolate mononuclear cells from blood or spleen, remove the TCR+ T cells with anti95

/. Pena, R. Solana, F. Borrego and B. Rolstad Table 1. Examples of human and mice receptors for MHC antigens: structure Receptor

mAbs

Structure

MHC recognition

References

p58.1 p58.2 Kp43 NKB1 Ly-49"

EB6 GL183 HP3B1 DX9/Z27 A1

Ig superfamily. 58 kDa Ig superfamily. 58 kDa C-type lectin. 43 kDa 70 KDa C-type lectin, 44 KDa

HLA-Cw2,4,5,6 HLA-Cw1, 3,7,8 HLA-B7,8,14 HLA-B51,58,27 H-2Dd, H-2Dk

7,25,26 7,25,26 17 18 13,19

a

ln mouse

CD3 antibody and culture the cells for several days with IL-2. These cells are all CD3~ and positive for NK-markers (NKR-P1 in the rat). In the human system NK cells can be propagated with IL-2 for prolonged periods of time as lines or clones using appropiate stimulator or feeder cells. Protocol 2. NK cell purification* Equipment and reagents • Lymphocyte separation medium (LSM): Lymphoprep (Nycomed) . Culture medium (CM): RPMI 1640 with Hepes buffer supplemented with 10% fetal calf serum (FCS), antibiotics, and 2 mM Lglutamine • Phosphate buffered saline (PBS) (Unipath, Hampshire, England) • Laminar flow hoods • Petri dishes (Costar) • Goat anti-mouse (GAM) coupled magnetic beads (Dynal, Oslo, Norway)

• Nylon wool • Monoclonal antibodies: anti-CD3 (anti-Leu 4, Becto-Dickinson, San Jose, CA), antiCD56 (NKH-1, Coulter, Hialeah, FL), antiCD19 (J4.119, Immunotech, Marseille, France), anti-CD14 (RM052, Immunotech) and anti-CD16 (3G8, Immunotech). • Magnetic particle concentrator (MPC1, Dynal) • Flow cytometer (EPICS Profile from Coulter or FACS Sort from Becton-Dickinson)

A. Preparation of freshly human NK cells 1. Isolate peripheral blood mononuclear cells (PBMC) by centrifugation (2500 r.p.m. for 15-20 min) of peripheral blood diluted 1:1 with PBS over LSM. 2. Remove the PBMC from the LSM interface and dilute with PBS, centrifuge at 1500 r.p.m. for 5-10 min. Decant and repeat wash steps two more times. 3. Perform a white cell count. Dilute the cell suspension to a final concentration of 5 X 106 PBMC/ml in CM and incubate in Petri dish for 60-90 min at 37"C in 5% C02, for monocyte depletion. 4. Recover non-adherent cells, wash twice with PBS, and dilute in 2 ml of CM. 5. Incubate the cells in pre-washed nylon wool (with CM) for 45-60 min at 37°C in 5% CO2, for B cell depletion. 96

7: Methods to study the role of the major histocompatibility complex 6. Recover the non-adherent (T/NK cells) by washing the column with 20 ml of CM, resuspend in 500 u,l and incubate with anti-CD3 (10-25 (jig) for 30 min at 4°C. 7. Wash the cells twice with PBS. 8. 0.7-0.9 x 108 of GAM coupled magnetic beads is added for 30 min at 4°C under gentle shaking for T cell depletion. 9. Add 9 ml of PBS and use a magnetic particle concentrator to remove the cells resetting with the beads. Collect the supernatant containing CDS' cells. 10. Centrifuge the cells, resuspend in CM and use for the cytotoxic assay. 11. The cell population obtained is 75-95% CD56+ and CD16+, < 10% CD3+ and < 5% CD14+ and CD19+ as routinely studied by flow cytometry. B. Preparation of IL-2 activated rat NK cells 1. Remove spleen cells aseptically and wash twice in PBS. 2. Centrifuge the cells for 30 min at 400 x g on Lymphoprep (Nycomed; density: 1.077 gml"1), and collect mononuclear cells from the interface. 3. Wash the cells twice in PBS + 2% FCS and resuspend in RPMI 1640 with HEPES buffer (25 mM), 2% FCS and standard antibiotic-antimycotic solution (all reagents from Gibco, Paisley, Scotland). To remove CD3+ cells, add 1F4 or 64.1 (anti-CD3) and sterile normal rabbit serum as a source of complement (diluted 1:20) to the cells in equal volumes and the cells are incubated at 37°C for 75 min. 4. Wash the cells three times in PBS + 2% FCS and resuspend in cold PBS + 2% FCS. Dynabeads (M450 coated with rat-anti-mouse, provided by Dynal) and further coated with 3.2.3 antibody are added at a Dynabeadxell ratio of 1:5 (at -30 x 106 cells/ml). 5. Incubate the cells and Dynabeads under continuous movement for 45 min at 4°C. Then wash the cells binding Dynabeads 8-10 times in PBS+2%FCS using Dynal's Magnetic Particle Concentrator. Resuspend the cells bound to the Dynabeads in LAK-medium. (i.e. RPMI 1640 with 10% FCS, 5 x 10"5 M 2-ME, 1 mM Na-Pyruvate, 2 mM Lglutamine and 1000 ID ml"1 of rat or human rlL-2, the latter obtained from Cetus, California). 6. Culture the cells in 5% CO2/95 % H2O at 37°C for about 7 days. Most of the cells will by then have released the beads, will be more than 99% 3.2.3+ and less than 1% CD3+, and can be used directly for cytotoxicity testing (Protocol 1). •Protocol reproduced after C. Naper er a/., (34) and with the permission of C. Naper.

97

/. Pena, R. Solana, F. Borrego and B. Rolstad Protocol 3.

Generation of NK clones

Equipment and reagents • Recombinant interleukin 2 (rlL-2) (Jansen, Sweden)

• Culture medium (CM) with 10% human AB serum

Method 1. Isolate NK cells as described in Protocol 1. 2. Incubate 105 NK cells with 5 x 104 irradiated (4000 rad) autologous mononuclear cells and irradiated (10000 rad) allogeneic EBVtransformed cells in a total volume of 2 ml CM containing 250 Uml"1 IL-2 in a 24-well flat-bottom microtiter plate. Incubate the plate for 4 days in a humidified, 37°C, 5% C02 incubator. 3. Remove 1 ml medium and replace with 1 ml containing 500 U rlL-2 (final concentration 250 U ml"1). Incubate the plate for 3 days. 4. Collect the cells and separate over a lymphocyte separation medium (LSM). 5. Remove cells from the interface. Wash the cells by centrifuging for 5-10 min at 1500 r.p.m. Count the cells. 6. Place the cells (at limiting dilution) into each well of a 96-well roundbottom microtiter plate containing 10s irradiated autologous mononuclear cells and 5 x 104 irradiated allogeneic EBV-transformed cells and 250 U ml"1 rlL-2 in a final volume of 200 |il. Incubate the plates at 37 °C, 5% C02. 7. Remove 100 u,l of medium and replace with 100 |J CM with 500 U ml"1 rlL-2. Incubate for an additional 3 days. 8. Remove 100 (jJ medium. Add 5 x 104 irradiated allogeneic EBVtransformed cells in CM and 50 U ml"1 rlL-2. Incubate for 4 days. 9. Remove 100 p,l medium and replace with 100 n,l CM containing 500 U ml'1 rlL-2. Incubate for 3 days. 10. Identify growing wells. 11. Enrich clones by LSM density gradient centrifugation. 12. Wash the interface twice in serum-free RPMI. Resuspend at 106 cells/ml in CM. 13. Place 1 ml of this suspension in each well of a 24-well microtiter plate.

98

7: Methods to study the role of the major histocompatibility complex 14. Add 5 x 105 irradiated allogeneic EBV-transformed cells in 1 ml CM containing 25 U ml"1 rlL-2 to each well. Incubate for 4 days. 15. Remove 1 ml CM and add fresh CM containing rlL-2 only. Incubate for 3 days. 16. Repeat steps 14 and 15 weekly.

3. In vitro assay for NK cell cytotoxicity against normal allogeneic cells and tumor cell lines A standard 51Cr release assay for measuring NK cell cytotoxicity in vitro is given in Protocol 4. The methods for measuring both human and rat NK cell cytotoxicity are in principle identical, and resemble standard cytotoxic T cell assays. Graded numbers of effector cells are added to a fixed number of 51Crlabelled target cells in microwell plates and the percentage specific 51Cr release calculated. Tumour or normal lymphoblast target cells can be used without further modification of the assay. When using small, non-dividing thoracic duct lymphocytes or bone marrow cells from the rat as target cells, it is important to add phytohaemagglutinin (PHA) at 10 M-gml"1 directly to the cytotoxic cultures to obtain efficient lysis of the allogeneic cells (8). This assay can also be used to determine allospecificity in cold target inhibition assays. In these assays unlabelled inhibitor cells are added in increasing numbers to effector and target cells at a given ratio. This has enabled the definition of at least four different MHC controlled allospecificities in the rat (24). Among the human NK clones a repertoire of at least five different allospecificities has similarly been discerned (7). When working with rat NK cells one should be aware of the problem that the alloreactivity repertoire may differ widely between different rat strains, and in some strains alloreactivity may not be present against any of the known rat MHC haplotypes. This is illustrated in Table 2a, which shows the alloreactivity patterns of NK cells from six different rat strains against Con-A blasts from a panel of MHC congenic target cells. It is evident that an MHC incompatibility between effector and target cells is required, since MHC compatible, but otherwise disparate target cells are not killed. However, not all MHC haplotypes (as defined by classical class I RT1.A typing) are recognized as foreign. Some strains like the PVG and AO respond to all the foreign RT1 haplotypes tested, while NK cells from e.g. the DA strain do not recognize any of the RT1 incompatible cells. The reason for this discrepancy between known MHC incompatibilities and NK allorecognition is not known but may related to the MHC-controlled gene products recognized by the NK cells, as described below, or to regulatory effects on the responder genes that map outside the MHC. 99

/. Pena, R. Solana, F. Borrego and B. Rolstad Table 2a. NK cell allorecognition patterns in the rata Source of NK cellsb

Target Con A blastsc

PVG (c) A0(u) LOU/C (u) BN(n) LEW (1) DA(av1)

LEW. 1C (c) ++ ++ -

LEW.1W

LEW.1N

LEW

LEW.AVI

(u)

(n)

(I)

++

++ ++ ++

++

(av1) ++ ++ ++

+ +

-

-

+ + _ -

-

a With permission fromLovikefa/. (33) 6 NK cells were from the rat strains as indicated, with their MHC (RT1) haplotype indicated brackets. "Target Con A blasts were from MHC congenic strains on the LEW background with the RT1 haplotype in brackets.

Table 2b. Predominant role of the second, nonclassical class i cluster, the RT1.C region, in controlling target cell sensitivity to alloreactive NK cellsa Target cell sensitivity

RT1 regions

Effector cells: PVG.1U Target cells: LEW.1U LEW.1A LEW.1WR2 LEW.1WR1 LEW.1AR2 PVG.R8

A

B/D

u

u

C u

u a u u a a

u a a u a u

u a a a u u

++ ++ ++ -

a Target cells were Con A blasts from a panel of intra-MHC recombinant rat strains in which crossing over between the u and a haplotypes had taken place between the RT1.A and B/D region or between the B/D and C region. Note how strong lysis correlates with MHC incompatibility between the effector and target cells in the C, but not the A or B/D region (after Vaage et al. 24).

Protocol 4. Generation of allospecific NK cell lines a Method 1. Some NK cells can specifically recognize allogeneic cells (13). The alloantigen NK-1 on target cells is controlled by HLA-C gene. Target cells are susceptible to allospecific lysis are homozygous for a two aminoacids (Ser77 and Asn80) in the a1 domain of HLA-Cw1, -Cw3, -Cw7, -Cw8 and by some of the -Cw blank alleles. The alloantigen NK-2 correlates with homozygosity for a different pair of amino acids at the same positions (Asn77 and Lys80) shared by HLA-Cw2, -Cw4, -Cw5, -Cw6 and some other -Cw blank alleles. 100

7: Methods to study the role of the major histocompatibility complex 2. To generate an NK cell line with specificity for NK-1, NK cells from the peripherical blood lymphocytes (PBLs) of an HLA-C (Asn77, Lys80) homozygous donor are co-cultured with irradiated PBLs from an HLAC (Ser77, Asn80) homozygous donor. NK cells specific for NK-2 are also generated by stimulation of NK cells from donor homozygous for HLA-C (Ser77, Asn80) with irradiated PBLs from an HLA-C (Asn77, Lys80) homozygous donor. 3. Isolate NK cells according to the method described in the Protocol 1. 4. Culture NK cells in 96-well round-bottom plates (5 x 104 per well) in the presence of irradiated (4.000 rad) allogeneic PBLs (104 cells per well) in a final volume of 200 \L\. Incubate for 3 days. 5. Remove 100 jxl of medium and replace with 100 jxl CM with rlL-2 and IL-2-containing supernatant at a final concentration of 250 UmT1 and 10%, respectively. Incubate for 7 days. 6. Before the cytotoxicity assay, stain effector cells for the surface expression of CDS, CD16 and CD56. 7. For the cytotoxicity assay target cells are PHA blasts obtained of PBLs isolated from blood samples or buffy coats. Adjust the PBLs to 1 x 106 cells/ml and culture in CM with PHA (1 n-gmr1) and rlL-2 (100 Urnl'1) for 2-3 days. "Adapted from ref. 26.

4. Description of the human, rat and mouse MHC regions relevant for NK alorecognition A description of the mouse, rat and human MHC is found elsewhere in this book and we will here summarize only certain features.

4.1 Rat and mouse MHC The mouse and rat MHC resemble each other in that a central class II and class III region is flanked on each side by class I regions. The left side of the MHC shows extensive homology between the mouse (H-2K) and rat (RT1.A) in that both regions code for class I molecules that are polymorphic and function as T cell restriction elements. Both these regions are therefore regarded as classical class I regions. A major functional difference lies in the right hand side of the MHC. In the mouse this region consists of both classical and non-classical class I genes. The H-2 D/L region codes for the polymorphic class I molecules that also restrict T cell responses to environmental antigens. No functional equivalent is present in the rat. Although the 101

/. Pena, R. Solana, F. Borrego and B. Rolstad majority of rat class I genes reside within the RT1.C region, they usually do not restrict T cell responses. Some of these genes are polymorphic and are expressed at low density on the cell surface, but their function is not known. The genetic elements controlling target cell susceptibility to alloreactive NK cells reside both within the left and the right hand side of the MHC in the mouse (H-2K vs H-2D) and the rat (RT1.A vs RT1.C) (4). In the mouse certain genes termed haemopoietic histocompatibility genes (Hh), close to the H-2D region, affect target cell susceptibility (9). However, there are indications that the H-2D class I molecules themselves are important in determining target cell sensitivity. According to the 'missing self model for NK cell recognition the expression of certain K and D alleles efficiently protects target cells against NK lysis (5). This was originally shown with class I loss mutant cells, but later also with class I transgenic and knockout mice (5,6). In the rat target cell susceptibility to alloreactive NK cells is also usually controlled by genes within the right part of MHC (4), i.e. the RT1.C region (Table 2b), but the mode of inheritance of RT1.C controlled target cell specificities is usually nonrecessive (24). However, also classical class I molecules, (RT1.A) can be recogized by rat NK cells.

4.2 Human MHC In the human, all the class I genes (HLA-A, B and C) are located in the right flank of the MHC. The best-characterized restriction elements for cytotoxic T cells are the HLA-A and B molecules. The HLA-C region is less well characterized, is expressed in lower density at the cell surface, and only a few examples of T cell restriction are known. Human target cells were found to be protected from lysis by a certain alloreactive NK clones (NK-2) if they expressed a certain HLA-C allele (HLA-Cwl, -Cw3, -Cw7, -Cw8, and by some of -Cw blank alleles) (25), and it has also been shown that target cells susceptible to allospecific lysis have to be homozygous for a two-amino acid polymorphism of HLA-Cw, namely Ser77 and Asn80 in the al domain. In addition, another alloantigen (NK-1) correlates with homozygosity for a different pair of amino acids at the same positions, Asn77 and Lys80, shared by HLA-Cw2, -Cw4, -Cw5, -Cw6, and some other -Cw blank alleles. Cells heterozygous for these pairs of amino acids can not be lysed by NK cells that recognize NK-1 or NK-2. The requirement for both HLA-C alleles of the target provides a rationale for the reproducible generation of NK cells lines with NK-1 and NK-2 allospecificities that have so far only been generated by stimulation between random donors (see Protocol 4) (26). Another NK allospecificity (NK-3) has recently been described, achieved by stimulation of a Bw4 homozygous responder with a Bw6 stimulator, and two NK clones were found which were inhibited by HLA-Bw4, but not by HLA-Bw6, allotypes. Inhibition of NK-cell mediated lysis by these molecules correlated in particular with the presence of an isoleucine residue at position 80 of the protective 102

7: Methods to study the role of the major histocompatibility complex Table 3. HLA polymorphisms recognized by NK clones NK clone specificity

HLA recognition

NK-1 NK-2 NK-3 NK-4

Cw2, 4, 5, 6 Cw1, 3, 7, 8 Bw4 B2705

Aminoacids 77 and 80 associated to NK resistance 77 80 Asn Ser

Lys Asn

lie Thr

allele (27). Thr 80 in B27 subtypes can also act as an inhibitory residue of NK cytotoxicity (31). The different human NK allospecificities described up to now are summarized in Table 3. The control of target cell sensitivity in humans is not confined to the HLAC and B regions: the expression of certain alleles within the HLA-A locus may also confer resistance to NK cells (28). Moreover-, site-directed mutagenesis studies of amino acid residues within the peptide binding cleft have shown that the target cell sensitivity is critically dependent on the configuration of this cleft (29). These and other data have led to the hypothesis that NK cells recognize a complex of MHC I molecule and associated peptide (30). In conclusion, certain class I encoding regions of the MHC play a major role in determining target cell sensitivity to NK cells in mice, rats and humans. Not only one class I region may be involved, and the expression of certain allospecificities shows both recessive and nonrecessive inheritance, the latter only shown in the rat. The combined interpretation of mouse rat and human data therefore indicates that MHC class I alleles may have complex up- and down regulatory effects on target cell sensitivity to NK cells, depending on the MHC specificities studied. The knowledge of NK recognition of MHC molecules has grown very quickly in the last year; for a comprehensive review of the subject see reference (32).

5. Concluding remarks NK cell recognition of MHC-mismatched cells is a complex and still not fully characterized phenomenon, in which the expression of certain MHC class I alleles on the target cell often plays a central role in determining sensitivity. The development of reliable in vitro assays for studying NK reactivity against MHC-matched and mis-matched, tumour and normal cells as described here, and the identification in mouse and man of new subset-specific NK cell surface markers associated with given allospecificities have provided important new tools for studying the basis for allospecificity within the non-

103

/. Pena, R. Solana, F. Borrego and B. Rolstad adaptive immune system. With the identification of new NK surface molecules and the availability of transfection techniques to study the effect of individual NK surface molecules or MHC class I molecules in NK-target interactions, the enigmas surrounding the role of MHC molecules in activation or inhibition of NK cells should come to a resolution.

References 1. Karre, K., Ljunggren, H.G., Piontek, G., and Kiessling, R. (1986). Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. Nature, 319, 675. 2. Ljunggren, H.G. and Karre, K. (1990). In search of the 'missing self: MHC molecules and NK cell recognition. Immunol. Today, 11, 237. 3. Pena, J. and Solana, R. (1994). Effect of target cell histocompatibility antigens in NK-cell-mediated recognition and lysis. In MHC antigens and NK cells (ed. R. Solana and J. Pena), p. 35-49. R.G. Landes Company, Austin. 4. Rolstad, B., Wonigeit, K., and Vaage, J.T. (1993). Alloreactive fat natural killer (NK) cells in vivo and in vitro: the role of the major histocompatibility complex (MHC). In Natural immunity to normal hemopoietic cells (ed. B. Rolstad), pp.99-149. CRC Press, Inc., Boca Raton. 5. Hoglund, P.I., Junggren, H.G., Ohlen, C., Arhrlund-Richter, L., Scangos, S., Bieberich, C., Jay, G., Klein, G., and Karre, K. (1988). Natural resistance against lymphoma grafts conveyed by H-2Dd transgene to C57BL mice. J. Exp. Med., 168, 1469. 6. Liao, N.S., Bix, M., Zijlstra, M., Jaenisch, R., and Raulet, D. (1991). MHC class I deficiency: susceptibility to natural killer (NK) cells and impaired NK activity. Science, 253, 199. 7. Ciccone, E., Pende, D., Viale, O., Di Donato, C., Tripodi, G., Orengo, A.M., Guardiola, J., Moretta, A., and Moretta, L. (1992). Evidence of a natural killer (NK) cell repertoire for (allo) antigen recognition: definition of five distinct NKdetermined allospecificities in humans. J. Exp. Med., 175, 709. 8. Rolstad, B. and Fossum, S. (1987). Allogeneic lymphocyte cytotoxicity (ALC) in rats: establishment of an in vitro assay, and direct evidence that cells with natural killer (NK) activity are involved in ALC. Immunology, 60, 151. 9. Yu, Y.Y.L., Kumar, V., and Bennett, M. (1992). Murine natural killer cells and marrow graft rejection. Annu. Rev. Immunol., 10, 189. 10. Solana, R. and Pena, J. (1994). Natural killer (NK) cell receptors: Activation and inhibition of NK-cell-mediated cytotoxicity. In MHC antigens and NK cells (ed. R. Solana and J. Pena), pp. 16-34. R.G. Landes Company, Austin. 11. Giorda, R., Rudert, W.A., Vavassori, C., Chambers, W.H., Hiserodt, J.C., and Trucco, M. (1990). NKR-P1, a signal transduction molecule on natural killer cells. Science, 249, 1298. 12. Ryan, J.C., Turck, J., Niemi, E.G., Yokoyama, W.M, and Seaman, W.E. (1992). Molecular cloning of the NK1.1 antigen, a member of the NKR-P1 family of natural killer cell activation molecules. J. Immunol., 149, 1631. 104

7: Methods to study the role of the major histocompatibility complex 13. Yokoyama, W.M. and Seaman, W.E. (1993). The Ly-49 and NKR-P1 gene families encoding lectin like receptors on natural killer cells: the NK gene complex. Annu. Rev. ImmunoL, 11, 613. 14. Borrego, F., Galiani, M.D., Madueno, J.A., Perez-Bermejo, L., Garcia-Cozar, F., Pena, J., and Solana, R. (1995). Phenotypic and functional characterization of NK mAb using resting NK cell and the NK cell line YT. In Leucocyte typing V (ed. S.F. Schlossman et al). Oxford University Press (pp:1427-1430). 15. Trinchieri, G. (1989). Biology of natural killer cells. Adv. ImmunoL, 47, 187. 16. Cassatella, M.A., Aneg6n, I., Cuturi, M.C., Griskey, P., Trinchieri, G., and Perussia, B. (1989). FcyR (CD16) interaction with ligand induces Ca2+ mobilitation and phosphoinositide turnover in human natural killer cells: role of Ca2+ in FcyR (CD16)-induced transcription and expression of lymphokines genes. J. Exp. Med., 169, 549. 17. Moretta, A., Vitale, M., Sivori, S., Borrino, G., Morelli, L., Augugliaro, R., Barbaresi, M., Pende, D., Ciccone, E., Lopez-Botet, M., and Moretta, L. (1994). Human NK cells receptors for HLA-1 class molecules. Evidence that Kp43 (CD94) molecule functions as receptor for HLA-B alleles. J. Exp. Med., 180, 545. 18. Litwin, V., Gumperz, J., Parham, P., Phillips, J.H., and Lanier, L.L. (1994). NKB1: an NK cell receptor involved in the recognition of HLA-B. J. Exp. Med., 180, 537. 19. Karlhofer, F.M., Ribaudo, R.K., and Yokoyama, W.M. (1992). MHC class I alloantigen specificity of Ly-49+ IL-2-activated natural killer cells. Nature, 358, 66. 20. Sentman, C.L., Hackett, J.Jr., Kumar, V., and Bennett, M. (1989). Identification of a subset of murine natural killer cells that mediates rejection of Hh-ld but not Hh-lb bone marrow grafts. J. Exp. Med., 170, 191. 21. Kane, K.P. (1994). Ly-49 mediates EL4 lymphoma adhesion to isolated class I major histocompatibility complex molecules. J. Exp. Med., 179, 1011. 22. Daniels, B.F., Karlhofer, F.M., Seaman, W.E., and Yokoyama, W.M. (1994). A natural killer cell receptor specific for a major histocompatibility complex class I molecule. J. Exp. Med., 180, 687. 23. Brooks, A.G., Posch, P.E., Scorzelli, C.J., Borrego, F. and Coligan, J.E. (1997). NKG2A complexed with CD94 defines a novel inhibitory natural killer cell receptor. J. Exp. Med., 17, 1. 24. Vaage, J.T., Naper, C., Lovik, G., Lambracht, D., Rehm, A., Hedrich, H.J., Wonigeit, K., and Rolstad, B. (1994). Control of natural killer cell-mediated allorecognition by a major histocompatibility complex region encoding nonclassical class I antigens. J. Exp. Med., 180, 641. 25. Colonna, M., Borsellino, G., Falco, M., Ferrara, G.B., and Strominger, J.L. (1993). HLA-C is the inhibitory ligand that determines dominant resistance to lysis by NK1- and NK2-specific natural killer cells. Proc.Natl.Acad. Sci. USA, 90, 1200. 26. Colonna, M., Brooks, E.G., Falco, M., Ferrara, G.B., and Strominger, J.L. (1993). Generation of allospecific natural killer cells by stimulation across a polymorphism of HLA-C. Science, 260, 1121. 27. Cella, M., Longo, A., Ferrara, G.B., Strominger, J.L., and Colonna, C. (1994). NK3-specific natural killer cells are selectively inhibited by Bw4-positive HLA alleles with Isoleucine 80. J. Exp. Med., 180, 1235. 28. Storkus, W.J., Alexander, J., Payne, J.A., Cresswell, P., and Dawson, J.R. (1989). The <xl/a2 domains of class I HLA molecules confer resistance to natural killing. J. ImmunoL, 143, 3853. 105

/. Pena, R. Solana, F. Borrego and B. Rolstad 29. Storkus, W.J., Salter, R.D., Alexander, J., Ward, F.E., Ruiz, R.E., Cresswell, P., and Dawson, J.R. (1991). Class-I induced resistance to natural killing: identification of nonpermissive residues in HLA-A2. Proc. Natl. Acad. Sci. USA, 88, 5989. 30. Storkus, W.J., Salter, R.D., Cresswell, P., and Dawson, J.F. (1992). Peptideinduction modulation of target cell sensivity to natural killing. J. Immunol., 149, 1185. 31. Luque, I., Solana, R., Galiani, M.D., Gonzalez, R., Garcia, F., Lopez de Castro, J. and Pena, J. (1996). Threonine 80 on HLA-B27 confers protection against lysis by a group of natural killer clones. Eur. J. Immunol., 26, 1974. 32. Moretta, A., Biassoni, R., Bottino, C., Pende, D., Vitale, M., Poggi, A., Mingari, M.C. and Moretta, L. (1997). Major histocompatibility complex class I-specific receptors on human natural killer and lymphocytes. Immunological Reviews, 155, 105. 33. Lovik, G., Vaage, J.T., Naper, C., Benestard, H.B. and Rolstad, B. (1995). Recruitment of alloreactive natural killer cells to the rat peritoneum by a transfected cell line secreting rat recombinant interleukine-2. /. Immunol. Methods, 179, 59. 34. Naper, C., Vaage, J.T., Lambracht, D., Lovik, G., Butcher, G.W., Wonigeit, K., Rolstad, B. (1995). Alloreactive natural killer cells in the rat: complex genetics of major histocompatibility complex control. Eur. J. Immunol, 25, 1249.

106

8

The interaction of superantigens with MHC class II molecules CAROL HORGAN and JOHN D. FRASER

1. Introduction Superantigens are specific viral or bacterial products that simultaneously bind to class II MHC molecules and T cell receptors (TCR) without requiring proteolytic processing. They can activate up to 20% of peripheral T lymphocytes, as compared to conventional antigens which activate only about 0.001% of peripheral T cells. The binding occurs at sites on the MHC outside of the peptide binding groove and on the (3 chain of the TCR, away from the conventional antigen binding site. Viral superantigens, first referred to as MLS (minor lymphocyte stimulating) antigens are products of the mouse mammary tumour virus (MMTV) and perhaps other retroviruses as well. Bacterial superantigens are bacterial exotoxins secreted by several types of bacteria including Staphylococcus aureus, Streptococcus pyogenes, and Mycoplasma arithritidus. The best characterized bacterial superantigens are the staphylococcal enterotoxins (SE), SEA, SEB, SEC(l-3), SED, SEE, and toxic shock syndrome toxin (TSST) (Table 1). These will be the subject of this chapter. The SEs can be divided into three groups, type I, II, and III, on the basis of sequence homology (1) and their abilities to compete with one another for binding to MHC class II. The SEs bind to numerous isotypes and alleles of class II molecules. They bind human class II better than mouse class II, and bind HLA-DR better than HLA-DQ or HLA-DP (2). Type I SEs (SEA, SEE, and SED) require the presence of Zn2+ cations to bind MHC, whereas type II (SEB, SEC1, SEC2, and SEC3) and type III (TSST) do not (3). The binding of the toxins to MHC class II molecules is readily detectable. Competitive binding studies show that SEA competes equally well for HLA-DR1 binding with SEB and TSST, but SEB and TSST are non-competitive (3-7). This could suggest separate binding sites on the MHC for the different types of toxins. However, alternative explanations for the non-competitive nature of SEB and TSST, such as conformational changes in MHC after toxin binding, cannot be ruled out.

Carol Morgan and John D. Fraser Table 1. Aligned sequences of staphylococcal superantigens. SEA SEE SED SEC2 SEC 3 SECl SEB TSST

1 SEKSEEINEK DLRKKSELQG . . .SEEINEKDLRKKSELQR SVKEK ELHKKSELSS . ESQPDPTPD ELHKSSEFTG . ESQPDPMPD DLHKSSEFTG . ESQPDPTPD ELHKASKFTG . ESQPDPKPD ELHKSSKFTG S

SEA SEE SED SEC2 SEC 3 SECl SEB TSST

TILFKGFFTD TLLFKGFFTG TLLYKKFFTD DLIYNISDKK DLIYNISDKK DLIYNISDKK DLIYSIKDTK SLGS.MRIKN

50 TALGNLKNIY YYNEKAKTEN KESHDQFLQH NALSNLRQIY YYNEKAITEN KESDDQFLEN TALNNMKHSY ADKNPIIGEN KSTGDQFLEN T . MGNMKYLY . DDHYVSATK VMSVDKFLAH T.MGNMKYLY . DDHYVSATK VKSVDKFLAH L . MENMKVLY . DDHYVSATK VKSVDKFLAH L . MENMKVLY . DDNHVSAIN VKSIDQFLYF T . NDNIKDLL DWYS . SGSDT F . SNSEVLDN

51 HSWYNDLLVD HPWYNDLLVD LINFEDLLIN LKNYDKVKTE LKNYDKVKTE LKNYDKVKTE LGNYDNVRVE .TDGSISLII

FDSKDIVDKY LGSKDATNKY FNSKEMAQHF LLNEDLAKKY LLNEDLAKKY LLNEGLAKKY FKNKDLADKY FPSPYYSPAF

100 K . GKKVDLYG AYYGYQCA. . K . GKKVDLYG AYYGYQCA. . K . SKNVDVYA IRYSINCY. . K . DEWDVYG SNYYVNCYFS K . DEWDVYG SNYYVNCYFS K . DEWDVYG SNYYVNCYFS K . DKYVDVFG ANYYYQCYFS K . GEKVDLNT KRTKKSQH. .

101 TPNKTACMYG GVTLHDNNRL TEEKKVPINL TPNKTACMYG GVTLHDNNRL TEEKKVPINL EIDRTACTYG GVTPHEGNKL KERKKIPINL G . . . KTCMYG GITKHEGNHF DNGNLQNVLI G. . .KTCMYG GITKHEGNHF DNGNLQNVLV G . . . KTCMYG GITKHEGNHF DNGNLQNVLI TDKRKTCMYG GVTEHNGNQL D. .KYRSITV EGTYIHFQIS GVTNTE. . KL PTPIELPLKV

SEA SEE SED SEC2 SEC3 SECl SEB TSST

GG GG GG SKDNVGKVTG SKDNVGKVTG SKDNVGKVTG KKTNDINSHQ TS

SEA SEE SED SEC2 SEC 3 SECl SEB TSST

151 PLETVKTNKK NVTVQELDLQ PIDKVKTSKK EVTVQELDLQ SLDKVQTDKK NVTVQELDAQ SFE . VQTDKK SVTAQELDIK SFE . VQTDKK SVTAQELDIK SFE . VQTDKK SVTAQELDIK SFD.VQTNKK KVTAQELDYL KYG . PKFDKK QLAISTLDFE

SEA SEE SED SEC2 SEC 3 SECl SEB TSST

201 STEPSVNYDL SEGSTVSYDL SDGSKVSYDL NNGNTFWYDM NNGNTFWYDM NNGNTFWYDM .NENSFWYDM NDGSTYQSDL

ARRYLQEKYN ARHYLHGKFG ARRYLQKDLK ARNFLINKKN ARNFLINKKN ARNFLINKKN TRHYLVKNKK IRHQLTQIHG

150 WL . DGKQNTV WI . DGKQTTV WI . NGVQKEV RVYENKRNTI RVYENKRNTI RVYENKRNTI RVFEDGKNLL KV.HGKDSPL

200 LYNSDVFDGK VQRGLIVFHT LYNSDSFGGK VQRGLIVFHS LYNNDTLGGK IQRGKIEFDS LYEFN. .SSP YETGYIKFIE LYEFN. .SSP YETGYIKFIE LYEFN. .SSP YETGYIKFIE LYEFN. .NSPYETGYIKFIE LY. . . .RSSD KTGGYWKITM

250 FGAQGQY . . SNT . LLRIYRDN KTINSENM.H IDIYLYTS . . FDAQGQY . . PDT.LLRIYRDN KTINSENL.H IDLYLYTT . . EK.QLRIYSDN KTLSTEHL.H IDIYLYEK. . FDVKGDF . . P MPAPGDKFDQ SK . YLMMYNDN KTVDSKSV.K IEVHLTTKNG MPAPGDKFDQ SK . YLMMYNDN KTVDSKSV.K IEVHLTTKNG MPAPGDKFDQ SK . YLMMYNDN KTVDSKSV.K IEVHLTTKNG MPAPGDKFDQ SK . YLMMYNDN KMVDSKDV . K IEVYLTTKKK SKKF.EY. N TEKPPINIDEI KTIEAEIN

Sequences were aligned using the Pileup program (GCG Version 7, Genetics Computer Group, Wisconsin, USA). Residues are represented by the one letter amino acid code.

8: The interaction of superantigens with MHC class II molecules While superantigen binding to MHC class II is not restricted to specific alleles, the binding to TCR is specific. Each toxin stimulates a characteristic profile of T cells, depending on the TCR p chain expressed. The binding site on the TCR p chain is the same for all superantigens (bacterial and viral) and involves the fourth hypervariable region (HV4) of the Vp domain (8-11). Individual amino acid residues in this loop probably contact residues of the superantigen, but this binding is weak and only occurs after the superantigen has been bound to a class II molecule. This suggests that superantigen activation of T cells requires SEA:MHC binding, SEA:TCR binding, and MHCTCR interaction. Much can be learned of MHC involvement in T cell activation by examination of superantigen interactions with MHC and TCR. These important tools have allowed the positioning of SEA:MHC structures with the SEA:TCR contact residues, and these alignments have generated hypotheses concerning MHC:TCR interactions (12). The recent elucidation of the structure of the MHC class II molecule HLA-DR1 (13, 14) provides much information about MHC structure and function. The co-crystallization of SEB plus MHC class II (15) will further our knowledge of the interaction of superantigens with class II MHC. In this chapter we endeavour to provide the basic techniques involved in studying the interaction of bacterial superantigens with class II MHC. In some cases, space constraints prohibit inclusion of all relevant protocols; in these cases, we have referenced the original papers so that the reader can easily obtain the information required to utilize the techniques.

2. Production of recombinant bacterial superantigens A serious limitation of using purified toxins from S. aureus to examine superantigen:MHC interactions, is the possibility of contamination by one or more toxins from a single bacterial isolate. The SEs are biochemically similar, and are extremely difficult to separate completely. To prepare pure isolates of SEs, expression of the individual genes in E. coli is recommended. This method also allows for mutational analysis of the toxins. PCR is used to amplify the toxin genes from the genomic DNA. The amplified genes are then cloned into the pGEX-2T vector (Pharmacia) using standard molecular biology techniques (16). The pGEX-2T vector, when transformed into E. coli, allows the expression of the introduced gene as a fusion product with glutathione S transferase (GST). Fusion protein can be produced in amounts of up to 30 mg per litre of culture, and which is then affinity purified from bacterial lysates in a single step by affinity chromatography on a glutathione agarose column. Cleavage of the fusion product with trypsin, while still bound to the affinity column, results in the retention of the GST moiety and the cleaved, recombinant toxin is washed from the column with PBS. PCR primer pairs can be designed using the known nucleotide sequences of the toxins (Table 2) (17-22). The primers should contain 5' BamHI and 3' 109

Carol Horgan and John D. Eraser Table 2. 5' and 3' nucleotide sequences of the staphylococcal enterotoxin genes. N

TERMINAL +1 E S SEA: AGC GAA E S SEE: AGC GAA V S SED: TCA GTA E S SEE: GAG AGT S E SEC3 :GAG AGT T S TSST :TCT ACA

C

TERMINAL

N AAC N SEE: AAT S SED: TCC K SEB: AAA K SEC 3 :AAA I TSST :ATA SEA:

S TCT S TCA T ACA D GAT S AGT N AAT

REGION

E GAA E GAA K AAA Q CAA

I ATA I ATA E GAG P CCA

N AAT N AAT K AAA D GAT D Q P CAA CCA GAC N N D AAC GAT AAT

E GAA E GAA E GAA P CCT P CCT I ATA

K AAA K AAA L TTG K AAA M ATG K AAG

D GAT D GAT H CAT P CCA P CCA D GAT

L TTG L TTG K AAA D GAT D GAT L TTG

R CGA R CGA K AAA E GAG D GAT L CTA

K AAA K AAA S TCT L TTG L TTG D GAC

K AAG K AAG E GAA H CAC H CAC W TGG

S TCT S TCT L TTA K AAA K AAA Y TAT

E GAA E GAA S AGT S TCG S TCA S AGT

L TTG L TTA S AGT S AGT S AGT S AGT

H CAT H CAT H CAT E GAA E GAA I ATA

I ATT I ATT I ATT V GTT V GTC K AAA

D GAT D GAT D GAC Y TAT H CAC T ACT

I ATA L TTG I ATC L CTT L CTT I ATA

Y TAT Y TAT Y TAT T ACG T ACA E GAA

L TTA L TTA L TTA T ACA T ACA A GCA

Y TAT Y TAC Y TAT K AAG K AAG E GAA

T ACA T ACA E GAA K AAA N AAR I ATT

T AGT S ACT K AAG K AAG G GGA N AAT

END TAA END TGA END TAG END TGA END TAA END TAA

REGION E GAA E GAG E GAG V GTG V GTG I ATT

N AAC N AAC H CAC K AAG K AAG D GAT

M ATG L CTC L CTT I ATT I ATA E GAA

EcoRI cloning sites to allow easy insertion of the amplified fragment into the pGEX-2T vector. Since the fusion proteins will be cleaved by trypsin to release the recombinant toxin, it is best to design the 5' PCR primers using the sequences of the mature toxin protein, rather than sequences at the start of the toxin leader peptides. Protocol 1. Cloning and expression of staphylococcal enterotoxins Reagents • Staphylococcal growth media: 3% (w/v) peptone, 0.2% yeast extract, 0.2% glucose, pH 7.0 . 50 mM MgCI2 • Lyphostaphin « STE: 100 mM NaCI, 10 mM Tris-HCI pH 8.0, 1 mM EDTA

« . • • • •

110

SDS: 10% in H20

TE: 10 mM Tris-HCI pH 7.4, 1 mM EDTA Proteinase K 3 M sodium acetate pH 5.2 Tris-saturated phenol CHCI3:isoamyl alcohol (24:1)

8: The interaction of superantigens with MHC class II molecules • Ethanol • PCR primers (12.5 (iM in H2O) • Tap polymerase • EcoRI and BamHI restriction endonucleases • pGEX-2T (Pharmacia) • Low melting point agarose . IPTG • Competent E. coli strain DH5a . Triton X-100 « TPCK treated trypsin (Sigma) • 0.1 MCaCI2

• Sonicator • Glutathione agarose (Sigma) or glutathione sepharose (Pharmacia) . 10x PCR buffer: 100 mM Tris-HCI pH 8.3, 500 mM KCI, 0.1 mg ml"' gelatin . 5 mM dNTPs (mix of dATP, dGTP, dCTP, dTTP, 5 mM each) • Luria broth (LB) 1% (w/v) bacto tryptone, 0.5% yeast extract, 1% NaCI, pH 7.0, supplemented with 50 fj,g ml"1 ampicillin . PBS: 120 mM NaCI, 2.7 mM KCI, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.0

A. Cloning of staphylococcal enterotoxins from genomic staphylococcal DNA 1. Grow up a 100 ml overnight culture of the staphylococcal strain in staphylococcal growth media. 2. Pellet the bacteria at 5000 r.p.m. for 10 min. 3. Wash twice with STE. 4. Resuspend the pellet in 5 ml TE. 5. Add 0.5 mg lyphostaphin. 6. Incubate at 37°C until lysis occurs (usually within 2 min, lysis can be detected by a change in viscosity of the mixture). 7. Add 0.5 ml 10% SDS and bring the volume to 10 ml with TE. 8. Add 100 M,| 3 M NaOAc pH 5.2. 9. Add Proteinase K to 50 u,g/ml final concentration. 10. Incubate at 60°C for 1 h. 11. Extract with an equal volume of tris-saturated phenol:CHCI3:isoamyl alcohol (25:24:1), followed by extraction with CHCI3:isoamyl alcohol (24:1). 12. Add 2 volumes of EtOH to precipitate the DNA. 13. DNA can be recovered by spooling on to sterile toothpicks or by centrifugation. 14. Resuspend the precipitated DNA in TE and repeat steps 11-13. 15. Resuspend the DNA in TE for PCR amplification of the toxin genes. B. Amplification of SE genes 1. Prepare a 50 uJ reaction containing: 5 jj-l 10x PCR buffer (minus Mg2+ 1 fil each of primers (12.5 uJvl), 1 uJ 5 mM dNTPs, 2.5 jil 50 mM MgCI2, 2.5 U Tag polymerase, 250-500 ng template DNA. 2. Overlay the reaction mixture with 50-100 uJ mineral oil and cycle for 30 rounds at: 95°C for 30 seconds; 55°C for 2 minutes; 72°C for 1 minute.

111

Carol Horgan and John D. Fraser Protocol 1. Continued C. Cloning of the amplified genes into pGEX-2T 1. Digest pGEX-2T with BamHI and EcoRI in the appropriate digestion buffer. 2. Run the digested sample on a 1% low melting point agarose gel and cut out the linear band. 3. Purify the DNA by phenol:CHCI3 extraction or other method of choice. 4. Remove excess primers from the PCR sample by ammonium acetate precipitation of high MW DNA, agarose gel purification, or some other method. 5. Digest the PCR product with BamHI and EcoRI. 6. Gel purify the product on low melting point agarose. 7. Ligate the insert into the vector and use to transform competent E. coli strain DH5a. D. Expression of the amplified genes 1. Grow up a 100 ml overnight culture of E. coli harbouring the plasmid in LB supplemented with 50 ^g ml"1 ampicillin. 2. Dilute the overnight culture 10-fold into fresh LB/ampicillin, and grow up for one hour at 37°C, with shaking. 3. Induce the expression of the fusion protein by the addition of IPTG to 0.1 mM final concentration and continue growing with shaking for a furthers h. 4. Harvest the culture by centrifugation (5000 r.p.m., 10 min) and resuspend the pellet in 4 ml STE. 5. Add Triton X-100 to 1%. 6. Lyse the bacteria by sonication, three separate 15 sec bursts with 1 min on ice between bursts. 7. Spin the samples for 10 min in a microfuge to pellet the bacterial debris and transfer the supernatant containing the fusion protein to a clean tube. E. Affinity purification of the fusion product 1. Prepare a 1 ml glutathione-agarose column in the barrel of a 2 ml disposable syringe. 2. Wash the column with PBS containing 1% Triton X-100, and rinse with PBS before loading the sample. 3. Load the sonicated bacterial lysate on to the column, and wash with 4 ml PBS/1% Triton X-100, followed by 12 ml PBS.

112

8: The interaction of superantigens with MHC class II molecules 4. To cleave the fusion protein, add 500 n,l of TPCK-treated trypsin (2 (ig ml"1 in 0.1 M CaCI2) to the column. Seal the column ends and incubate overnight at 37 °C. 5. Elute the cleaved product with sequential 1 ml aliquots of PBS. 6. Assay the aliquots for protein concentration by OD280 (e = 1.4). 7. Pool the fractions with the highest protein concentrations.

3. Binding of staphylococcal enterotoxins to MHC class II Superantigens must first associate with class II MHC molecules before they can exert their stimulatory effect on T cells. The binding of SE to MHC class II-expressing cells reveals that SEA has the highest affinity of the SEs for HLA-DR1; Kd = 36 nM. The Kd for SEE is 125 nM (3, 23). TSST has a Kd of 120 nM, and SEB has the lowest affinity; Kd = 1 jxM (6). Other studies show the same order of affinities (2). The binding of type I SEs involve the zinccoordinating residues H187, H225 and D227 in SEA (D187, H225 and D227 in SEE), along with the H81 of the p chain of DR1 (23-25). SEA binds to all DR alleles except DRw53, because in DRw53, the p chain has a tyrosine instead of a histidine at residue 81. SEB binds to all DR alleles thus far tested. SEB residue F44 has been shown to be involved in the binding of DR1 (26) and probably interacts with residues on the a helix of DR1 (15). Further, TSST binding involves residues M36 and K39 of the DR1 a chain (27). All the toxins bind to DQ and DP but with much lower affinity. Direct binding of the SEs to class II MHC on fixed class II-expressing cells can be demonstrated by radioimmunoassay. Competitive inhibition of binding by unlabelled toxins or peptides can easily be assessed by this protocol. The staphylococcal enterotoxins are readily radio-iodinated on tyrosine residues either by the chloramine-T or iodogen method (Pierce Laboratories). Class II expressing cells are grown in RPMI/10% FCS in a 37 °C incubator with 5% CO2. These are fixed in paraformaldehyde prior to use in the cell binding assays. 10% fetal calf serum (FCS) is included in the assay to minimize non-specific binding of the toxin to the cells. Since the type I toxins (SEA, SEE, SED) require zinc for binding, at least 10 uM Zn must be present to obtain saturating conditions. Standard culture media contains sufficient zinc as does PBS supplemented with 10% FCS. Type II and III toxins bind MHC in the absence of zinc. Type I toxins (SEA, SEE, SED) all require physiological temperatures to bind with maximal efficiency, so incubations must be carried out at temperatures greater than 20°C. Type II and III toxins (SEB, SEC, TSST) bind as well at 4°C as at 37°C. Direct binding assays can also be performed in the absence of cells by utilizing purified class II molecules. Human DR1 has been successfully purified 113

Carol Horgan and John D. Fraser from human lymphoblastoid cells (28) and soluble forms of human DR1 have been expressed in insect cells (29). The soluble DR1 can be used with 125Ilabelled toxins to examine binding kinetics. Protocol 2. Radioimmunoassay of SE binding to MHC class II Reagents • Purified toxin • 10x phosphate buffered saline (PBS): 1.2 M NaCI, 27 mM KCI, 100 mM Na2HP04, 20 mM KH2P04, pH 7.4 • Chloramine T: make 100x stock solution in H2O; 50 mg chloramine T dissolved in 0.5 ml H20. Dilute 1:100 into H2O just before use. • Sodium metabisulfite (Na2S205): make 20x stock solution in H2O; 50 mg Na2S2O5 dissolved in 0.5 ml H20. Dilute 1:20 into H20 just before use. • G-25 Sephadex column: equilibrate Sephadex G-25 in 1x PBS and prepare a 1-2 ml column in a 2 ml disposable syringe barrel.

. PBS/1% BSA: 1x PBS supplemented with 1% bovine serum albumin « Trichloroacetic acid (TCA): 10% in H2O • lodogen (Pierce Laboratories): 100 jig ml"1 in chloroform • Paraformaldehyde: 1% in 1x PBS • LG-2 cells: (human B lymphoblastoid line homozygous for DR1) (ATCC) • PBS/10% FCS: 1x PBS supplemented with 1% fetal calf serum • RPMI/10% FCS: RPMI culture medium (Gibco), supplemented with 10% fetal calf serum « Gamma counter

A. Radio-iodination of toxin using chloramine T 1. In a total of 20 uJ reaction, combine 2 u,g toxin, 5 ^l 10x PBX, 2 pJ 1 mg ml"1 chloramine T, and 300-500 uCi 125l-Na. 2. Incubate for 1 min at room temperature, then stop the reaction by adding 2 uJ 5 mg ml"1 sodium metabisulfite. 3. Bring the volume up to 100 p,l with PBS and desalt on a 2 ml G-25 column. Wash with PBS and collect 100 u,l fractions. Monitor the c.p.m. per fraction with a Geiger counter, or count them in a gamma counter. 4. Pool the protein fractions having the highest c.p.m., and test for unincorporated 125I by TCA precipitation as follows: (a) To 99 |J PBS/1% BSA add 1 uJ 125l-toxin. (b) Add 100 u,l 10% trichloroacetic acid. (c) Incubate for 2-5 min at room temperature, then spin in a microfuge for 5 min. (d) Separate the supernatant (SN) from the pellet and count each separately in a gamma counter. (e) Percent precipitation = c.p.m. in pellet/(c.p.m. in pellet + c.p.m. in SN). (f)

If your 125l-sample is less than 90% precipitable, repeat the desalting procedure to remove unincorporated 125I.

114

8: The interaction of superantigens with MHC class II molecules B. Alternate labelling protocol (lodogen)* 1. Coat 1 ml glass tubes or vials with 100 |o,l iodogen at 100 p.g ml'1 in chloroform. 2. Evaporate the chloroform under a stream of N2, and store the tubes at -20°C (in a desiccator) until ready for use. 3. In 50-100 |xl volume, add 2 |o.g toxin to the iodogen coated tube. 4. Add 300-500 u,Ci 125l-Na. 5. Incubate for 5 min at room temperature. 6. Remove the sample from the reaction tube and desalt as above. 7. Test TCA precipitability as above. C. Paraformaldehyde fixation of class II expressing cells 1. Wash cells once in PBS and resuspend in 1% paraformaldehyde at approximately 106 cells/ml. 2. Incubate for 15 min at room temperature. 3. Wash twice in PBS/10% FCS. 4. Resuspend in PBS/10% FCS (or RPMI/10% FCS) at 4 x 106/ml. 5. Fixed cells can be stored at 4°C for several weeks if 0.2% NaN3 is added. D. Binding of 125l-toxin to fixed cells 1. In a total reaction volume of 1 ml in PBS/10% FCS combine: 1 ng 125ISEA, SEE, or TSST or 10 ng 125I-SEB, 100 (jil fixed LG-2 cells (4 x 105 cells), 0-10 ixg unlabelled competitive toxin or peptide. 2. Incubate for 1 h at 37°C. 3. Wash twice in PBS/10% FCS. 4. Pellet the cells and count pellet in a gamma counter. a SEA and SEE should normally be labelled to a specific activity of about 50 (iCi/pig. SEB and TSST can be labelled as hot as 120 M-Ci/^g, but this should be avoided as radiolytic degradation is rapid, and leads to inactivation of the toxins. For best results aim for 50 M.Ci/|ig for all toxins.

4. Zinc binding by type I staphylococcal enterotoxins, SEA, SEE,and SED The binding of SEA, SEE, and SED to class Il-expressing MHC cells is completely abolished by low levels of EDTA, and the binding can be restored by addition of trace amounts of Zn2+ (in excess over the EDTA). The binding of type II (SEB, SEC) and type III (TSST) toxins are not affected by EDTA. 115

Carol Horgan and John D. Fraser The Kd of zinc for SEA is 1 u,M, as determined by Scatchard analysis, and there is a 1:1 stoichiometry of zinc to SEA (23). Zinc binding can be demonstrated by the binding of radioactive 65Zn to immobilized toxins in the presence or absence of class II molecules. The zinc binding site in SEA was mapped to residues H187, H225, and D227 using this technique (30). Protocol 3.

65

Zn binding assay

Reagents • . • • • • •

Purified toxin EDTA (100 |iM) Hybond C nylon membrane (Amersham) Chelex resin (Sigma) Dot blot apparatus Coomassie Blue R stain Destain solution: 45% methanol, 7% acetic acid • X ray film

•

65

ZnCI2 (New England Nuclear). Low specific activity will do, diluted to 300 p.M with zinc free H20 • Gamma counter • Zinc-free H2O: incubate H20 overnight with Chelex resin to remove free zinc . 10X zinc free buffer": 1.5 M NaCI, 0.1 M MgCI2, 0.2 M Tris-HCI, pH 7.3. Incubate overnight with Chelex resin. Prepare 1x zinc free buffer with zinc free H20

Method 1. Prepare a reaction volume of 50 (jj in zinc free H2O with: 15 pJ toxin at 1 mg ml"1, 5 p.1 100 jxM EDTA, 5 pJ 300 u,M 65ZnCI2, and 5 |xl 10x zinc free buffer. 2. Incubate for 2 h at 37°C. 3. Wet a Hybond C nylon membrane with zinc-free water. 4. Blot 16 |xl of ^Zn-toxin reactions in triplicate to the membrane in a 96well dot blot apparatus. 5. Rinse 5 times under vacuum with 200 u,l per well 1x zinc free buffer. 6. Remove membrane from apparatus and wash twice for 2 min in 1x zinc free buffer. (To distinguish between zinc binding and zinc entrapment, replace the final two washes with two washes in 10 n-M EDTA (or if peptides are included in the assay, 2 washes with 0.75 |j,M EDTA followed by 2 more washes with 1.5 ^M EDTA). This will strip off zinc which is not bound by the zinc-coordinating residues of the toxin and/or class II.) 7. Air dry the filter, and autoradiograph. 8. To quantify the zinc binding, cut out the dots from the filter and count them in a gamma counter. 9. To quantitate protein concentration, stain a duplicate membrane with Coomassie Blue R, and destain in 45% MeOH/7% acetic acid. "To determine the effects of MHC on the coordination of zinc, we have included purified class II molecules or peptides based on class II sequences in the initial reaction mixtures. In this case, we used 40 mM HEPES buffer in place of the Tris and a 20-fold molar excess of peptides to toxin (30).

116

8: The interaction of superantigens with MHC class II molecules

5. T cell stimulation by superantigens T cell stimulation by superantigens requires the prior association of superantigen with molecules of class II MHC. All bacterial as well as viral superantigens are believed to bind to a common site, residues 70-74 on the TCR p chain. This region is hypervariable, and according to molecular models of the TCR is a loop structure, termed HV4 (31-34). The SEA residues involved in TCR binding are S206 and SN207 (35). SEE residues N23 and Y61 are also involved (26) and according to the SEB crystal structure (36) these residues are all found in a groove between the two domains of the toxins. This region conceivably acts as a pocket into which the TCR 3 chain HV4 loop can fit (12). To assess the appropriate interaction of the superantigen with MHC, T cell stimulation of peripheral blood lymphocytes or T cell clones can be performed by incubating the toxins, antigen presenting cells, and T cells together. Interleukin-2, produced by activated T cells, can subsequently be measured in an assay which measures the incorporation of 3H-thymidine into an IL-2 dependent cell line. Table 3 shows the TCR Vp reactivity of the different SEs. T cell hybridomas expressing known Vp molecules are used in these assays to examine the T cell reactivity of the recombinant toxins. The IL-2 dependent cell lines, HT-2 or CTLL-2 (37-39), can be used as indicator cells for detecting IL-2 in the culture supernatants of the stimulated T cells. The IL-2 dependent cells should be cultured in the presence of 20-40 ng mr1 human IL-2 in RPMI/10% PCS. These are washed free of IL-2 before use in the assay. Because of the extreme potency of all the bacterial superantigens, the only way to compare activity is by serial dilution. Thus all stimulation assays must be performed with serial dilutions of the toxin across a range of 1 fg ml"1 to 10 jig ml"1. Potency is defined as the concentration required to induce 50% maximal proliferation (P50). Most superantigens interact with more than one TCR P chain, and there is Table 3. V|3 reactivity of staphylococcal enterotoxins Superantigen

Murine Vps

Human Vps

SEA

1, 3, 10, 11, 12, 17

1.1, 5.3, 6.3, 6.4, 6.9, 7.3, 7.4, 9.1, 23.1

SEE

11, 15, 17

5.1, 6.3, 6.4, 6.9, 8.1, 18

SED

SEC-2

3, 7, 8.3, 11 7, 8.2, 8.3, 1 1 8.2, 10

1.1, 5.3, 6.9, 7.4, 8.1, 12 3.2, 6.4, 6.9, 15.1 12, 13, 14, 15, 17,20

SEC-3

7, 8.2

5, 1

TSST

15, 16

2.1,8.1

SEC-1

The pattern of TCR Vp usage was derived from a recent review (40) and Hundson et al. (35).

117

Carol Horgan and John D. Fraser a characteristic profile of T cells stimulated by a given SAg, defined by the Vp which is expressed. The Vp profile of specific SEs can be identified by examination of the Vp sequences of T cells which have been enriched in the population after in vitro stimulation by SE. The technique employs the polymerase chain reaction to amplify all TCR 3 chain cDNAs from a sample of reverse transcribed RNA after SE stimulation of resting PBLs. The cDNAs are radiolabelled, and hybridized to immobilized Vp genes on nylon filters. Those Vp genes which have been enriched in the stimulated population 'light up' and define the range of Vp reactivity of the particular SAg. This technique has been described in detail (35). Protocol 4. T cell stimulation assay Reagents Purified toxin H-thymidine (New England Nuclear) (20 tid mr1) Heparinized whole blood Phorbol myristate acetate (PMA) (Sigma) Ficoll (Sigma) Human IL-2 Cell harvester Liquid scintillation counter 3

. RPMI/10% PCS (supplement with 0.5 (xM 2mercaptoethanol for growing T cell hybridomas) • T cell line expressing an appropriate V(3 (see Table 3) • LG-2 cells (ATCC): human DR1 expressing B lymphoblastoid line « HT-2 cells or CTLL-2 cells (ATCC), or other IL-2 dependent cell line

A. Isolation of PBLs from human blood 1. Obtain 10 ml heparinized whole blood. 2. Layer the blood over an equal volume of Ficoll (Histopaque 1077) and centrifuge at 1500 r.p.m. for 30 min at room temperature. 3. With a pipette, collect the white lymphocyte layer, and wash twice in PBS to remove all the Ficoll. 4. Resuspend the cells in RPMI/10% FCS at 2 x 106/ml. B. Stimulation of PBLs with superantigen 1. In a 96-we 11 culture plate prepare in duplicate, 5-fold serial dilutions starting with 10 p,g ml"1 of toxin and working down to 1 fg ml'1 in 150 |J RPMI/10% FCS. 2. Add 50 ixl PBLs (1 x 105 PBLs) per well. 3. Incubate at 37 °C, 5% CO2 for three days. 4. Pulse the cells by addition of 20 uJ 3H-thymidine per well and grow for 16 hat37°C. 5. After 16 h, harvest the cells on a cell harvester, and count the pellets in a liquid scintillation counter.

118

8: The interaction of supemntigens with MHC class II molecules 6. Calculate the P5o as the concentration of toxin required to induce 50% maximal proliferation. C. Tcell hybridoma stimulation 1. In a 96-well culture plate prepare duplicate 200 ^l samples of RPMI/10% FCS containing 5 X 105 T cells, 1 x 105 LG-2 cells (fixed or live), 10 fg-10 (jig toxin (serial dilutions as above), and 10 ng ml"1 PMA 2. Incubate at 37°C for 24 h. 3. At this point the plate can be frozen or the supernatant can be used immediately in the interleukin-2 bioassay. D. IL-2 bioassay 1. Wash the CTLL-2 or HT-2 cells 2-3 times in RPMI/10% FCS and resuspend at 106 ml~1. 2. In duplicate wells of a 96-well culture plate make 2-fold serial dilutions of the supernatant from the stimulated cells down the plate from undiluted to 1:128. Do the same for an IL-2 standard of known concentration. 3. Add 100 |il CTLL-2 or HT-2 cells per well (105 cells/well). 4. Incubate 24 hat 37 °C. 5. Pulse with 3H-thymidine (add 20 IA|3H-thymidine) for 6 hours, then harvest the cells and count in a liquid scintillation counter. 6. Calculate the amount of IL-2 produced by comparison with the IL-2 standard curve.

Acknowledgements We gratefully acknowledge the contributions of Dr Keith Hudson to this work. We also wish to acknowledge the generous support of the Wellcome Trust (UK), the Health Research Council of New Zealand and The Auckland Medical Research Foundation.

References 1. Bergdoll, M. S. (1983). In Staphylococcal and streptococcal infections (ed. C. S. Easom and C. Aslam), p. 559. Academic Press, New York. 2. Mollick, J. A., Chintagumpala, M., Cook, R. G., and Rich, R. R. (1991). /. Immunol., 146,463. 3. Fraser, J. D. (1989). Nature, 339,221. 4. Pontzer, C. H., Russel, J. K., and Johnson, H. M. (1991). Proc. Natl Acad. Sci. USA, 88,125. 119

Carol Horgan and John D. Eraser 5. Purdie, K., Hudson, K. R., and Fraser, J. D. (1991). In Antigen Processing and Presentation (ed. J. MacClusky), pp. 193. CRC Press, Boca Raton. 6. Scholl, P. R., Diez, A., and Geha, R. S. (1989). J. Immunol., 143, 2583. 7. Chintagumpala, M. M, Mollick, J. A., and Rich, R. R. (1991). J. Immunol., 147, 3876. 8. Choi, Y., Herman, A., DiGiusto, D., Wade, T., Marrack, P., and Kappler, J. W. (1990). Nature, 346, 471. 9. Pontzer, C. H., Irwin, M. J., Gascoigne, N. R. J., and Johnson, H. M. (1992). Proc. Natl Acad. Sci. USA, 89, 7727. 10. Pullen, A. M., Wade, T., Marrack, P., and Kappler, J. (1990). Cell, 61,1365. 11. White, J., Herman, A., Pullen, A. M., Kubo, R., Kappler, J. W., and Marrack, P. (1989). Cell, 56, 27. 12. Fraser, J. D. and Hudson, K. R. (1993). In T cell receptor genes (ed. J. Bell, E. Simpson, and M. Owen). Oxford University Press, Oxford (in press). 13. Brown, J. H., Jardetzky, T. S., Gorga, J. C., Stern, L. J., Urban, R. G., Strominger, J. L., and Wiley, D. C. (1993). Nature, 364,33. 14. Stern, L. J., Brown, J. H., Jardetzky, T. S., Gorga, J. C., Urban, R. G., Strominger, J. L., and Wiley, D. C. (1994). Nature, 368,215. 15. Jardetzky, T. S., Brown, J. H., Gorga, J. C., Stern, L. J., Urban, R. G., Chi, Y., Stauffacher, C., Strominger, J. L., and Wiley, D. C. Nature, 368,711. 16. Maniatis, T., Fritsch, E. F., and Sambrook, J. (ed.) (1982). Molecular Cloning: A laboratory manual. Cold Spring Harbor Press, New York. 17. Blomster-Hautamaa, D. A., Kreiswirth, B. N., Kornblum, J. S., Novick, R. P., and Schlievert, P. M. (1986). J. Biol Chem., 261, 15 783. 18. Bayles, K. W. and landolo, J. J. (1989). J. Bact., 171, 4799. 19. Betley, M. J. and Mekalanos, J. (1988). J. Bact., 170, 34. 20. Couch, J. L., Soltis, M. T., and Betley, M. J. (1988). J. Bact., 170, 2954. 21. Hovde, C. J., Hackett, S. P., and Bohach, G. A. (1990). Mol. Gen. Genet., 220, 329. 22. Jones, C. L. and Khan, S. A. (1986). J. Bact., 166, 29. 23. Fraser, J. D., Urban, R. G., Strominger, J. L., and Robinson, H. (1992). Proc. Natl Acad. Sci. USA, 89,5507. 24. Herman, A., Labrecque, N., Thibodeau, J., Marrack, P., Kappler, J. W., and Sekaly, R. P. (1991). Proc. Natl Acad. Sci. USA, 88, 9954. 25. Karp, D. R. and Long, R. O. (1992). J. Exp. Med., 175, 415. 26. Kappler, J. W., Herman, A., Clements, J., and Marrack, P. (1992). J. Exp. Med., 175, 387. 27. Panina-Bordignon, P., Fu, X., Lanzavecchia, A., and Karr, R. W. (1992). J. Exp. Med., 176, 1779. 28. Gorga, J. C., Knudsen, P. J., Foran, J. A., and Strominger, J. L. (1987). /. Biol. Chem., 262,16087. 29. Stern, L. J. and Wiley, D. C. (1992). Cell, 68,465. 30. Hudson, K.R., Tiedemann, R.E., Urban, R.G., Lowe, S.C., Strominger, J.L. and Fraser, J.D. (1995). J. Exp. Med., 182,711. 31. Claverie, J. M., Prochnicka-Chalufour, A., and Bougueleret, L. (1989). Immunol. Today, 1,10. 32. Chothia, C., Boswell, D. R., and Lesk, A. M. (1988). EMBO J, 7, 3745. 33. Davis, M. M. and Bjorkman, P. J. (1988). Nature, 334,395. 120

8: The interaction of superantigens with MHC class II molecules 34. 35. 36. 37. 38. 39. 40.

lores, R., Alzari, P. M., and Meo, T. (1990). Proc. Natl Acad. Sci. USA, 87, 9138. Hudson, K. R., Robinson, H., and Fraser, J. D. (1993). J. Exp. Med., 177, 175. Swaminathan, S., Furey, W., Fletcher, J., and Sax, M. (1992). Nature, 359, 801. Watson, J. (1979). J. Exp. Med., 150, 1510. Gillis, S., Perm, M. M., Ou, W., and Smith, K. A. (1978). /. Immunol., 120, 2027. Gillis, S. and Smith, K. A. (1977). Nature, 268, 154. Herman, A., Kappler, J. W., Marrack, P., and Pullen, A. M. (1991). Ann. Rev. Immunol., 9, 745.

121

This page intentionally left blank

a Methods for the generation of T cell clones and epithelial cell lines from excised human biopsies or needle aspirates G. DE LIBERO

1. Introduction The effector mechanisms and the cellular interactions that regulate the immune response can be studied by taking advantage of denned in vitro systems. These studies can be performed using cloned reagents, such as T cell clones and monoclonal antibodies (mAbs). Studies performed with T cell clones have allowed the identification of surface molecules involved in cellular interactions (1-7), of lymphokines responsible for immunoregulation (8), and of different functional properties of T cells (9, 10). T cell clones have also facilitated the analysis of the molecular requirements for antigen presentation (11-13), antigen recognition (14), and T cell activation (15), and are a unique source of nucleic acids for the characterization of the genes that are expressed only in rare lymphocytes (16, 17). Medical studies have also benefited greatly from the use of T cell clones. Experiments performed with clones revealed the complex variety of T cell populations in different organs (18), and permit the identification of the antigens and restriction molecules involved in diseases (19). Another important application of T cell clones is in the study of the TCR V gene repertoire, TCR chain pairing, and T cell function and antigen specificities during immune responses in tissues (20, 21).

2. Establishment of T cell clones It is possible to establish human T cell clones which can be maintained in vitro for long periods. In most cases the cells retain their antigen specificity and continue to release the same pattern of lymphokines upon activation. These characteristics make them the ideal tool to study the biology of human

G. De Libero T cells. Human T cell clones can be established from many tissues, within the first few hours after sample collection. The techniques for expanding and cloning human T cells are based on the capacity of T lymphocytes to perform several cycles of replication upon proper stimulation. T cells proliferate for 4-6 weeks if the important factors are present in the culture medium. The protocols described in this chapter allow the expansion and cloning of T lymphocytes and take advantage of polyclonal stimulation of T cells by using phytohaemagglutinin (PHA). This lectin interacts with many surface glycoproteins and therefore cross-links surface receptors, activates T cells, and drives them into the cell cycle. However, PHA alone is not sufficient to obtain optimal proliferation of T cells because other factors are required. For this reason interleukin-2, which is a growth factor for T cells, is added to the medium. In addition, irradiated feeder cells are also required. We use peripheral blood mononuclear cells as a source of feeder cells, since they contain several cell populations that secrete many different factors and express co-stimulatory molecules on their membranes which are important for maximal activation of T cells. Furthermore, feeder cells avoid the problem that T cells kill each other when crosslinked by PHA. Using these protocols it is possible to activate most T cells as shown by evaluation of plating efficiencies after T cell cloning in limit dilution. A simple and consistent method for establishing T cell clones is the stimulation of T cells in limiting numbers and the expansion of single cells in individual wells (Protocol 1). It is possible to seed single cells in 96-well plates using a cell sorter. We prefer to dispense the cells manually, and routinely make several groups with different cell numbers per well. This method of cloning permits immediate evaluation of the plating efficiency, and the use of Terasaki plates where single cells grow better because of the small volume of the wells (20 JJL!). It is also cheaper than other methods. Protocol 1. Establishment of T cell clones Equipment and reagents Sterile flow bench Inverted microscope Haemocytometer CO2, 37°C incubator Bunsen burner Pipettor Cloning mixture: complete medium plus PHA at Vgmr1 (Wellcome, Purified PHA, HA16) and irradiated (3000 Rad) peripheral blood mononuclear cells (PBMC) at 5 x 106 cells/ml"

• Gilson-like pipettes • Complete medium: RPMI 1640 with NaHC03 without Hepes (Gibco 21875-034); 2 mM L-glutamine (Gibco, 21051-024); 1 mM Na pyruvate (Gibco, 11840-048); 1% non-essential amino acids (Gibco, 11140-035); 100 ngml"1 kanamycin (Gibco, 15160-021); 5% AB human serum, not heat inactivated (Swiss Red Cross, Bern); 100 Uml~1 human recombinant IL-2 (HoffmannLa Roche, Nutley)"

Method 1. Prepare a cell suspension in complete medium and vortex.

124

9: Methods for the generation of T cell clones 2. Count the cells to be cloned. 3. Add the cells to the cloning mixture to have a final concentration of 1500 cells/ml. Using this concentration and distributing 20 JJL! per well you will plate 30 cells per well. From this concentration make four 1:3 dilutions using the cloning mixture. In this manner a concentration of 18.5 cells/ml (0.3 cells/well) is reached. 4. Plate 20 p-l of each dilution mixture in Terasaki plates. 5. After 8-12 days screen the plates using the microscope and mark the positive wells. 6. Calculate the plating efficiency using the Poisson distribution and identify wells containing clones. 7. Transfer proliferating cells in flat bottom 96-well plates and add 200 |j.l of fresh complete medium. 8. Clones are diluted 1:1 and split into two wells in the same flat bottom 96-well plate using complete medium after 2-5 days. The cells may be split into a 24-well plate if they grow very fast. About 3-4 weeks after cloning they must be restimulated (Protocol 2). 'This medium is kept at 4°C for a maximum period of three weeks, away from light. 'Prepare on day of use.

Established T cell clones can be expanded in culture for several weeks (4-8), then they stop proliferating and start to rest. At this stage they have to be restimulated to begin proliferating again. Protocol 2 describes the conditions for optimal restimulation of cloned T cells. Using these procedures it is possible to keep T cell clones in vitro for long periods, even years. Care is essential to avoid any microbial contamination, and it is very important to avoid clumps when cells are counted and plated. Therefore the cell suspensions should be vortexed both before counting and before plating them. Depending on which tissue the T cells are derived from, there may be large variations in plating efficiencies. We recommend starting with concentrations of 300 or 100 cells per well. With cells from peripheral blood it is sufficient to plate 3,1 and 0.3 cells per well. The number of Terasaki plates to be seeded depends (i) on the plating efficiency of the population that has to be cloned, and (ii) on the number of T cell clones you wish to recover. For example, if cells previously expanded in a bulk line are cloned then plating efficiencies close to 100% can be expected. If cells sorted using anti-TCR monoclonal antibodies are cloned, the plating efficiency drops to 10-50%. If immature and fragile double positive sorted thymocytes are cloned then plating efficiencies may be as low as 1-5%. It is important to use Terasaki plates where wells do not dry during the 8-12 days of incubation. We use Nunc plates (Nunc, 1-63118). Be sure to properly close the lids! 125

G. De Libero We use a multisyringe Hamilton device, that allows delivery of 5 jil per step in six wells at the same time. These syringes are sterilized with ethanol immediately before use, rinsed several times with PBS, and finally the metal tips are heated until glowing in the Bunsen flame. Wait long enough to let the tips cool before using! It is important to clean the syringes properly immediately after use. We routinely first rinse them in PBS and then in ethanol. Keep the piston closed when the syringes are stored. Do not open the incubator very frequently, and only examine the cells under the microscope one day after plating (to check for eventual contamination) and again after eight days (to check for cell growth). These operations enormously increase the evaporation, thus favouring drying of the plates and dramatically affecting the plating efficiency. It is also important to have an incubator that has 100% H2O-saturated air. Only cells grown in wells where they were plated at two dilutions below the plating efficiency are considered clones. For example, with a plating efficiency of 1/2, we only consider cells derived from 0.3 cell/well clones and not from wells where one cell was plated. We only transfer 48 clones in each 96-well plate. We fill six columns containing eight wells each and leave the adjacent wells empty. This allows the use of a multichannel pipette to feed, restimulate, and split the clones, thus reducing the time for manipulation of many clones and the risk of cell contamination. The empty well is used for the first splitting before transfer into 24-well plates.

Protocol 2. Restimulation of T cell clones Equipment and reagents • Restimulation mixture: complete medium plus PHA at 2 (igrnr1; irradiated (3000 Rad) PBMC at 106 cells/ml

• Complete medium: see Protocol 1

Method 1. Seed T cell clones to be restimulated in 1 ml of complete medium at a 2-5 X 105 cells/ml in 24-well plates. 2. Add 1 ml of restimulation mixture. 3. Check the cells after 2-3 days and split 1:2-4 if large blasts that are starting to cover the bottom of the well are visible. Ensure that the medium does not become too acidic. Keep the cells in complete medium. 4. When cells grow very rapidly cultures should be checked every 2-3 days.

126

9: Methods for the generation of T cell clones Since we use PHA, feeder cells need not be from the T cell donor. We use as feeders PBMC from buffy coats. PBMC can be frozen after irradiation and used for restimulation. For cloning use only fresh PBMC. We have found that purified PHA gives more consistent results and is not toxic for T cells. Remember to plate the restimulation mixture alone in some wells, to check that all feeder cells die after 10-15 days. Do not over-dilute the cloned cells (< 105 cells/ml), or they will grow more slowly. We usually dilute 1:1 and split the cells into two wells when the culture is dense or when the medium starts to become acidic. 100 Uml-1 of IL-2 will support the growth of quite dense cultures (up to 2 X 106ml-1). However, never leave cultures with a high cell density (> 106 cells/ml) for more than one day because at high concentration T lymphocytes quickly consume all the IL-2 and then die. In addition, do not keep the cultures in acidic medium for longer than 1 day because the cells stop growing. After 2-3 weeks the cells start to divide more slowly and must be fed less often. Usually after restimulation the cells grow for 3-4 weeks in IL-2. When the cells are in IL-2 they can be harvested at any time and used for experiments. We suggest not using them until at least 10 days after restimulation.

3. Preparation of cells from tissue biopsies The problem which more often is met with cell cultures established from excised tissue is bacterial and fungal contamination. We try to dimmish this risk, by collecting the sample immediately after the surgical intervention in a sterile tube containing complete medium with gentamycin, cyprofloxacyn and fungizone added. Using this cocktail of antibiotics we have enormously reduced the number of contaminated cultures without affecting viability and cloning efficiency of the cells. Once the sample is collected it has to be processed in the shortest possible time. Usually our laboratory receives the samples 2-4 h after excision, and this delay does not affect the recovery of the cells.

Protocol 3. Isolation of lymphocytes from tissue biopsies Equipment and reagents • • • •

Water bath at 37°C Vortex Complete medium: see Protocol 1 Restimulation mixture: see Protocol 2

• PCS medium: composition is like complete medium (see Protocol 1) without human recombinant IL-2 and human serum. Substitute human serum with 10% Fetal Calf serum and add Hepes.

127

G. De Libero Protocol 3.

Continued

Method 1. Collect the tissue directly in a tube containing complete medium added with gentamycin (50 jigml"1), cyprofloxacyn (10 jj,gmr1), and fungizone (2.5 n,gml~1). 2. Transfer the biopsy in a petri dish and wash with a Pasteur pipette to eliminate contaminating blood lymphocytes. Add 5 ml of medium and use a scalpel to mince the tissue. 3. Transfer all the tissue to a tube with FCS-medium with 200 U ml"1 collagenase type IV (Sigma, C 5138), and 2 mgrnl'1 deoxyribonuclease type I (Sigma, D 4263) added. 4. Incubate for 5 h at 37°C in a water bath and mix periodically. 5. Vortex the tube, allow the tissue debris to settle for 1-2 min and transfer the cells into a second tube. 6. Wash once, resuspend the pellet in complete medium and count the cells. 7. Seed the cells at 1-5 x 105ml"1. If a smaller number of cells is recovered they can be seeded in 100 |xl in a flat bottom 96-we 11 plate. 8. Stimulate the cells by adding the same volume of restimulation mixture.

T cells can be easily expanded from many different tissues, provided that the biopsies are collected only a few hours before cell culture is started. If it is not possible to process the sample rapidly, we have found that cells survive best at room temperature in complete medium. When the biopsy is transferred to the petri dish add medium containing 10 mM Hepes. Different amounts of cells will be recovered from different tissues. For example, from a thymic biopsy 1-5 XlO8 cells can easily be recovered, while from a liver or gut biopsy < 105 cells may be recovered. If the recovered cells will not be used immediately, freeze them in a solution of 10% DMSO in heat-inactivated PCS. Cells are frozen at a rate of approximately 1°C per minute by placing them in a styrofoam box in a — 70°C deep freezer. The day after freezing transfer them to liquid nitrogen. When we freeze thymocytes, we put high numbers of cells/tube since many of them die when thawed. In addition, when we thaw thymocytes we make a Ficoll gradient immediately after thawing, to reduce the number of clumped (and definitively lost) cells. We routinely use 15 ml V-bottom tubes because they allow more efficient recovery of cells after washing. Sometimes the presence of cells that might alter the growth capabilities 128

9: Methods for the generation of T cell clones of T cells must be avoided, in which case a Ficoll-Hypaque gradient can be prepared and lymphocytes isolated at the interface. We have also used 2 ml Eppendorf tubes with excellent recovery. Using this technique we were able to expand T cells from thymus, stomach, duodenum, liver, and lymph nodes. Protocol 4. Isolation of lymphocytes from synovial biopsies Equipment and reagents • Complete medium: see Protocol 1 • Restimulation mixture: see Protocol 2

• PCS medium: see Protocol 3

Method 1. Perform steps 1 to 4 of Protocol 3. 2. After incubation there will be much debris and a viscous cartilage containing many cells. To liberate these cells pass the debris through a 20 G needle several times, and then repeat the operation several times using a 22 G needle, until the tissue is finely minced. 3. Isolate lymphocytes using a Ficoll-Hypaque gradient and centrifuge for 20 min at 800 x g at room temperature. 4. Wash the cells once, resuspend in complete medium and count. 5. Dispense the cells at 1-5 x 10s ml"1 in 24-well plates. 6. Stimulate by mixing 1 volume of restimulation mixture with 1 volume of cells.

Protocol 5. Isolation of lymphocytes from tonsillar biopsies Equipment and reagents • Steril scalpel • Steril sieve • Phosphate Buffered Saline solution (PBS)

• Complete medium: see Protocol 1 • Restimulation mixture: see Protocol 2 • PCS medium: see Protocol 3

Method 1. Transfer the tonsil in a 50 ml tube containing pure ethanol and mix the biopsy for 2 minutes. 2. Discard the ethanol and add sterile PBS. Wash twice with PBS to remove all ethanol. 3. Mince the tissue using a scalpel and prepare small pieces of tissue. 4. Using a sieve and the sterile piston of a syringe gently dissociate the cells from the tissue.

129

G. De Libero Protocol 5. Continued 5. Resuspend the tissue and cells in a tube, allow the debris to settle and transfer the supernatant containing the cells to a second tube. 6. Purify the mononuclear cells on a Ficoll-Hypaque gradient. 7. Wash the cells once, resuspend in complete medium and count. 8. Dispense the cells at 1-5 x 105mr1 in 24-well plates. 9. Stimulate using 1 volume of restimulation mixture and 1 volume of cells.

The protocols which have been described allow the expansion of bulk T cell lines and do not enrich for antigen-specific populations. They also allow expansion of most T cells, even of those cells that are resting in the tissue. To enrich for cells which were recently activated in vivo, first cultivate the cells isolated from the biopsy with IL-2 alone, to favour selective expansion of those T lymphocytes expressing IL-2 receptors, which are markers of activated state. We have found that this protocol may sometimes be misleading if the expanded cells are later used for analysis of the TCR repertoire. Indeed, some sub-populations of a(3 and -y8 cells very often express the p55 IL-2 receptor and respond to IL-2 even when they are resting according to the lack of other surface activation markers (CD71, MHC class II, or CD69). Therefore we prefer to stimulate all the cells isolated from the biopsies with PHA and consider these bulk T cell lines as representative of the entire population. Protocol 6. Isolation of lymphocytes from fluid aspirates (cerebrospinal fluid and synovia! fluid) Equipment and reagents • Complete medium: see Protocol 1

• Restimulation mixture: see Protocol 2

Method 1. Collect the fluid in a sterile V-bottom tube. 2. Centrifuge the cells and resuspend the pellet in complete medium. 3. Count and dispense the cells at 1-5 x 105mr1 in 24-well plates. 4. Add the same volume of restimulation mixture.

A possible constraint to the studies of TCR repertoire which use bulk T cell lines is that the individual T cells when stimulated in vitro can expand with different rates, thus altering their relative frequency in the line. We have found that this happens with lines which have been in culture for 4-5 weeks, 130

9: Methods for the generation of T cell clones while there is a limited variation in fresh lines that have been in culture for only 2-4 weeks. Thus, we recommend the use of fresh lines to perform TCR repertoire studies. When antigen-specific T cell clones have to be isolated from the tissue biopsies, we use Protocol 7 for stimulating T cells. Protocol 7. Isolation of antigen-specific T cell clones from biopsies Equipment and reagents • Complete medium: see Protocol 1

Method 1. Prepare a cell suspension from tissues as described in the previous protocols. 2. Wash the cells and resuspend in complete medium without IL-2. 3. Seed the cells in flat bottom 96-we 11 plates or 24-well plates in case you recover many cells. We suggest plating a maximum number of 106 cells/well in the bigger plates, while the minimum number is dictated by the recovery. 4. Add irradiated PBMC isolated from the same donor at a concentration of 2 x 106 cells/ml in the same volume as the responder cells. 5. Add the proper amount of antigen and incubate. 6. After 3-6 days add 5 U mM of IL-2. 7. Change half the volume of the medium with fresh medium containing 5 U ml"1 IL-2 every 3-4 days. Split the cultures in a second well if you see a vigorous proliferation. 8. After three weeks you can restimulate the cultures again with irradiated feeder cells from the same donor and antigen, or clone T cells using Protocol 1. 9. When clones are transferred into 96-well plates, test their specificity, possibly before the next restimulation, to avoid the expansion of useless clones.

It is important to use the proper dose of antigen. This quantity can be determined by establishing an antigen-specific peripheral blood cell line and performing activation experiments testing different amounts of antigen. We use the lowest dose that gives maximal activation to establish cell lines from tissue biopsies. Use PBMC from the same donor as the biopsy as feeder cells. With 20 ml 131

G. De Libero of blood you can isolate ~2 x 107 PBMC that can be frozen and used throughout the experiments. We routinely establish EBV-transformed cell lines from the same donor, which we later use as APC in the experiments where MHC restriction is important. These lines represent an infinite source of cells with the proper HLA, thus avoiding the continuous bleeding of the donor (who is very often a hospitalized patient). We suggest cloning the cells at least 3-4 weeks after the last stimulation, when they start to become resting. When we test the specificity of the clones the first time, we perform proliferation or killing experiments and when the number of the clones to be tested is high (> 100) we only screen single wells of each clone without counting the cell number. After selection of the responding clones we repeat the assay (in triplicate) with only the positive clones, to confirm their antigenspecificity. When antigen-specific clones have been isolated expand the cells immediately (using Protocol 2) and freeze several aliquots.

4. Isolation of epithelial cells from human biopsies Epithelial cells have important roles in the immune response. They release soluble factors that interact with T cells, and participate in immune responses or inflammatory events that precede activation of T lymphocytes. In addition, epithelial cells can express surface MHC molecules, thus playing a role as antigen-presenting cells in the tissue (22). Recently, it has been found that class I-like MHC molecules are expressed in the thymus or intestine in mouse (23), thus making it possible that the antigen-specific responses in these micro environments may be different from other tissues. Therefore, isolation of epithelial cell lines that can be expanded in vitro for prolonged periods may allow studies of the interactions which occur in tissues between lymphocytes and epithelial cells. In our laboratory we have been interested in the epithelial cells isolated from thymus and intestine. We have used these cell lines, isolated from normal donors or patients, to stimulate T cell clones. In several cases we could isolate both epithelial and T cells from the same biopsy, thus performing the experiments with syngeneic stimulating and responder cells. In this section the conditions to establish epithelial cell lines are described. Protocol 8. Expansion of epithelial cells Equipment and reagents • Water bath at 37°C • Vortex

• Medium for expansion of epithelial cells: see Paragraph 4.1 • PCS medium: see Protocol 3

132

9: Methods for the generation of T cell clones Method 1. Prepare the tissue as described in Protocol 3, without digesting with enzymes. 2. Wash once in PBS and resuspend the pellet in medium for epithelial cells (for a detailed description of the preparation of this medium see the end of this section). 3. Inoculate a flask or a 24 well plate with the tissue debris. Epithelial cells are expanded on a feeder layer of 3T3 cells (for preparation see Protocol 9). 4. After 1-2 weeks several clusters of cells with epithelioid shape can be seen starting to invade the well. Replace half the volume with fresh medium every 3-4 days. 5. When the epithelial cells are confluent discard the medium, wash gently with PBS without Ca2+ and Mg2+ (to eliminate floating dead cells and FCS) and trypsinize for 10 min at 37°C. We use trypsin-EDTA from Gibco (45300-019). 6. Wash once and inoculate 2 x 10s cells in a 75 cm2 flask on irradiated 3T3 cells. 7. Split cells every five days.

In some experiments we have started the culture without feeder cells and have used them only after the first trypsinization. In this manner it is easily possible to see the epithelial cells starting to expand after a few days. We have used several plastics and found that the plastic common for cell culture gives good results. We also tried plastics from several companies and we did not notice major differences in cell growth. It is important to freeze aliquots of the epithelial cells quickly to avoid risks of microbial or cellular contamination. Indeed, in some cases, after many passages fibroblasts which grow faster than epithelial cells may be selected. It is important to split the cells at least once a week, because if they are kept in the same dish for a long time the proliferation of fibroblasts is favoured. Protocol 9.

Preparation of supporting 3T3 cells

Equipment and reagents • Medium for expansion of 3T3 cells: see Paragraph 4.2

• Medium for expansion of epithelial cells: see Paragraph 4.1

Method 1. Keep frozen ampoules of 3T3 cells which are used to generate fresh stocks every 2-3 months. Plate 3T3 cells in 75 cm2 flasks at the concentration of 10s cells/ml in 5 ml of DMEM medium containing 10% FCS (see the list of media at the end of this section). 133

G. De Libero Protocol 9. Continued 2. Split the cells every 5-7 days and avoid keeping the same cells in culture longer than 8 weeks. 3. To prepare 3T3 cells for epithelial cell expansion irradiate (6000 Rad) the flask when the cells become half-confluent, add epithelial cell medium and transfer the trypsinized epithelial cells. Alternatively, irradiate the flask, trypsinize the 3T3 cells and inoculate new flasks with 1.8 x 106 3T3 cells in epithelial cell medium. Leave for 4 h at 37°C (to allow attachment to the plastic) and inoculate with trypsinized epithelial cells.

4.1 Medium for expansion of epithelial cells The medium used for epithelial cell cultures is bicarbonate-buffered DMEM/F12 (1:1) (Gibco 21331-020) supplemented by different constituents. Stock solutions of medium constituents are prepared as follows. Constituents kept at — 20 °C are stored as aliquots to avoid continuous freezing and thawing. • Insulin stock (1000 X) Prepare a stock solution 5 rngml"1 of bovine insulin (Sigma, 1-6634) in 0.005 N HC1, and keep at — 20 °C. The final concentration in the medium is SjjLgml-1. • Transferrin/T3 (T/T) stock (1000X) Dissolve 250 mg of human transferrin (Behring Diagnostic 616397) in 30 ml PBS. Dissolve 13.6 mg of T3 (3,3',5 triiodo-L-thyronine sodium salt, Sigma, T-6397) by adding drops of 0.02 N NaOH. Add PBS up to 100 ml (concentrated stock solution 2 x 10"4 M). Add 0.5 ml of T3 concentrated stock solution to the 30 ml of transferrin solution and bring the volume to 50 ml with PBS. T/T 1000 x stock solution contains transferrin at a concentration of 5 rngmr1 and T3 at 2 X 10-6 M. Store at -20°C. The final concentrations in the medium are: transferrin 5 (xgrnr1, T3 2 X 10-9 M. • Hydrocortisone stock (500 X) Dissolve 25 mg of hydrocortisone (Sigma H 0888) in 5 ml of 95% ethanol (concentrated stock 5 mgmr1). Keep at 4°C. Dilute 1:25 in serum-free medium (200 (igml"1), and store at -20°C. The final concentration in the medium is 0.4 ixgrnl"1. • Cholera Toxin stock (lOOOx) Add 1.18 ml distilled H2O to 1 mg vial (Schwarz-Mann, 856011). This is a stock solution of 10-5 M. Do not freeze this solution. Dilute this stock 100X with medium containing 10% fetal calf serum and store at —20°C. This stock has a concentration 10-7 M. The final concentration of the medium is 10-10 M. 134

9: Methods for the generation of T cell clones • Adenine (Sigma A 2786) Prepare fresh every time when medium is prepared. For 500 ml of medium dissolve by stirring 12.16 mg in 10 ml of 10 mM HC1. The final concentration in the medium is 1.8 X 10-4 M. • EGF stock (1000X) Dissolve lyophilized EGF (Chiron Corporation) at 10 p-gnir1 m sterile distilled H2O. Store frozen at -70°C. The final concentration in the medium islOngmT 1 . • Fetal calf serum (PCS) Use at a final concentration of 10%. We use serum tested for its ability to support efficient colony formation. Once you have a good serum, keep some aliquots as a standard against which to evaluate new batches of sera. Each batch should be tested on cultures inoculated with 100 and 1000 cells of a standard epithelial cell line. At day 14 compare the number and size of colonies obtained with the test serum with those obtained with the standard serum. • L-glutamine (Gibco, 21051-024) Store aliquots of 200 mM at -20°C. When thawing ensure that all the precipitate dissolves. The final concentration in the medium is 2 mM. • Kanamycin (Gibco, 15160-021) Store aliquots of 10 mgmr1 at — 20 °C. The final concentration in the medium is 100 jigml"1. • Non-essential amino acids (MEM AA, Gibco, 11140-035) Store at 4 °C • Sodium pyruvate (Gibco, 11840-048) Store aliquots of 100 mM at -20°C. The final concentration in the medium is 1 mM. Table 1. Constituents of epithelial cell medium Constituent

Volume

DMEM/F12 Adenine L-glutamine (100x) MEM AA (100x) Sodium pyruvate (100x) Kanamycin (100X) Insulin (X1000) T/T (1000X)

417 ml 10 ml 5 ml 5 ml 5 ml 5 ml 0.5 ml 0.5 ml

Hydrocortisone (500x) Cholera toxin (1000x) EGF (1000X) FCS

1 ml 0.5 ml 0.5 ml 50 ml

135

Final concentration 1.8 x 10 -4M 2 mM

1 mM 50 M,g/ml 5 ^g/ml Transferrin 5 |o,g/ml T 3 2 X IQ-'M 0.4 ji-g/ml 10'10 M 10 ng/ml 10%

G. De Ldbero Table 1 summarizes the constituents and relative volumes of stock solutions to obtain 500 ml of medium. Very often we need smaller volumes and therefore we aliquot the complete medium in 50 ml tubes and store at -20°C up to six months. In that case we add EGF only when medium is thawed before sterilization.

4.2 Medium for expansion of 3T3 cells • • • • • •

DMEM with NaHCO3 without Hepes (Gibco 31885-023 ) 2 mM L-glutamine (Gibco, 21051-024) 1 mM Na pyruvate (Gibco, 11840-048) 1% non-essential amino acids (Gibco, 11140-035) 100 fig ml'1 kanamycin (Gibco, 15160-021) 5% fetal calf serum

5. Use of T cell clones to study tissue compartmentalization of human TCR y§ cells A second T cell receptor (TCR) for the antigen was identified in 1986 (4). This TCR consists of the product of the previously discovered -y gene (2) associated with a chain that was called 8 (24, 25). The gene organization of both 7 and 8 loci has been clarified, as have the tissue distribution and lymphokines released by -y8 cells. However, their antigen specificities and the mode of antigen recognition are still not known (26). Several hypotheses have been put forward to explain their function during immune responses. It has been suggested that -yS cells, which are only a minority of the circulating T lymphocytes, represent a first line of defence against microbial attack since they are located mainly in the epithelia or in the mucosa (27). This hypothesis was supported by the finding that some mouse -y8 hybridomas recognize peptides from heat shock proteins shared by various bacteria (28). These peptides might be expressed preferentially in cells invaded by intracellular pathogens (29). However, only a small number of -y8 hybridomas showed such a reactivity and in the human system these specificities were never found. To study the function of human -y8 cells our approach was to establish large numbers of T cell clones and analyse their antigen specificities, restriction, and TCR gene usage. In addition, the clones were used to study the lymphokine patterns after TCR triggering. We took advantage of our technique to generate T cell clones from small biopsies to study the expression of TCR variable -y and 8 genes and the presence of particular V8-V-y pairs in various organs. We thought that the analysis of the TCR -y& repertoire might disclose some features of the behaviour of "y8 cells in vivo, such as local expansion in response to antigens, enrichment in some organ-specific dis136

Table 2. TCR -y8 phenotype of T cell clones isolated from human organs VS-V-y pairs (% of analysed -yS clones) 51 -y9

S1-y4

81 -yx

82-y9

82-^4

82-yx

8x-Y9

8x-y4

8x--yx

Clones analysed

3 9 7 5 15 19 8 18

8 20 8 3 18 24 14 26

12 38 12 2 26 10 28 21

70 2 60 85 11 5 30 14

0 1 0 0 0 0 0 0

0 1 0 0 0 0 0 0

2 4 0 0 5 2 0 0

3 10 5 2 10 25 7 9

2 15 8 3 15 15 13 12

750 281 63 75 105 452 80 253

Organ Peripheral blood Thymus Tonsils Bone Marrow Stomach Duodenum Synovial fluid Cerebrospinal fluid

The following mAbs were used: SI (pan--yS), A13 (anti-VS1), 4G6 (anti-VB2), B3 (anti-V-y9), 4A11 (anti-V-y4). All the clones positive with 81 and not with 4G6 and A13 are defined as VSx, those negative with B3 and 4A11 are defined as V-yx.

G. De Libero eases, or compartmentalization as a result of a tissue-specific activation. We analysed clones isolated from peripheral blood, thymus, tonsils, bone marrow, stomach, duodenum, synovial fluid and cerebrospinal fluid of normal donors and patients. The TCR repertoire was compared in these tissues by immunofluorescence analyses using a panel of monoclonal antibodies specific for human V8 and ¥7 chains. In Table 2 the percentage of clones displaying different VS-V-y pairs is reported, together with the total number of clones analysed. The results show that there is high variability in the usage of different V genes. The analysis of 78 pairing in various organs showed a preferential usage of V81 chain. In the duodenum and in the cerebrospinal fluid this chain is very often associated with ¥74 (more often than expected, considering that 6 different ¥7 chains can be used). A second major V8 chain is V83, which pairs with various V-y chains without apparent bias. Clones derived from the stomach also use V81 and V83 more often than V82, and the ¥7 chains of the ¥7! family are found more frequently. No preferential V8-V-y pair was found in this organ. In the synovial fluid the presence of a majority of V81 bearing cells was found in some donors, while in others V82-V-y9 cells were more frequent. These results might be due to the presence of contaminating peripheral blood lymphocytes in the samples. In peripheral blood, as well as in bone marrow and tonsils, the ¥79-¥82 population represents the majority of the 78 cells. This is true in most of the analysed donors and confirms the results obtained in other laboratories (30,31). The ¥79-¥82 population is a minor percentage in the thymus (< 3% of total 78 thymocytes), while cells expressing ¥79 chains associated with V8 chains other than V82, or cells expressing V82 chains associated with ¥7 chains other than ¥79, can also be found. In the postnatal thymus a different population predominates, which frequently uses the V81 gene (32). This chain can associate with various ¥7 partners, without the formation of preferential pairs. There is also a relevant population that uses V8 chains other than V81 and V82, as well as ¥7 chains other than ¥79 and ¥74. Taken together, these results suggest (i) that in the thymus the TCR 78 repertoire is random, without bias for particular pairs; and (ii) that the preferential association of ¥79 with ¥82 in the peripheral blood is not explained by protein constraints and probably reflects peripheral selection, perhaps antigeninduced expansion (33). In some cases to distinguish between olygoclonal and polyclonal populations we performed TCR gene sequence analyses of the ¥(D)J junctions (unpublished results), and always found polyclonal populations. Thus, the compartmentalization of 78 cells in these tissues does not reflect local expansion of a few precursor cells that would lead to oligoclonal populations. If they were expanded by antigen stimulation, two hypotheses can be suggested: (i) these 78 pairs are preferentially stimulated by tissue-specific anti-

138

9: Methods for the generation of T cell clones gens, or (ii) in different organs there are tissue-specific antigen presenting molecules which may interact with particular VS-V-y pairs, thus favouring their expansion. The first hypothesis can be tested by stimulating T cell clones derived from various organs with tissue-specific antigens or exogenous antigens which are preferentially localized in those tissues. The second hypothesis can be tested by using antigen presenting cells isolated from the various tissues to stimulate T cell clones with known antigen specificities. Preliminary results suggest that classical MHC molecules are not important in antigen presentation to human 78 cells (33), and perhaps new molecules play such a role. What these new antigen-presenting molecules are, and whether they differ in various tissues, is not clear yet. Epithelial cells isolated from different organs and used as antigen-presenting cells might facilitate attempts to address these questions.

Acknowledgements I thank M. Biirk and L. Mori for suggestions and careful reading of the manuscript. This work was founded by Swiss National Funds (grant N. 31-36450.92) and by the Swiss Multiple Sclerosis Society.

References 1. Hedrick, S.M., Cohen, D.I., Nielsen, E.A., and Davis, M.M. (1984). Nature, 308, 149. 2. Saito, H., Kranz, D., Takagaki, Y., Hayday, A., Eisen, H., and Tonegawa, S. (1984). Nature. 312, 36. 3. Chien, Y.-H., Becker, D.M., Lindsten, T. et al. (1984). Nature, 312, 31. 4. Brenner, M.B., McLean, J., Dialynas, D.P., Strominger, J.L., Smith, J.A., Owen, F.L., Seidman, J.G., Ip, S., Rosen, F., and Krangel, M.S. (1986). Nature, 322, 145. 5. Meuer, S.C., Schlossman, S.F., and Reinherz, E.L. (1982). Proc. Natl. Acad. Sci. USA, 79 ,4395. 6. Springer, T.A. (1990). Nature, 346, 425. 7. Lanzavecchia, A. (1985). Nature, 314, 537. 8. Scott, P. and Kaufmann, S.H. (1991). Immunol. Today, 12, 346. 9. Romagnani, S. (1992). Immunol. Today, 13, 379. 10. Bloom, B.R., Modlin, R.L., and Salgame, P. (1992). Annu. Rev. Immunol., 10, 453. 11. Monaco, J. (1992). Immunol. Today, 13, 173. 12. Neefjes, J.J. and Ploegh, H.L. (1992). Immunol. Today, 13, 179. 13. Lanzavecchia, A., Reid, P.A., and Watts, C. (1992). Nature, 357, 249. 14. Townsend, A., Rothbard, J., Gotch, F. et al. (1986). Cell, 44, 959. 15. Mueller, D.L., Jenkins, M.K., and Schwartz, R.H. (1989). Annu. Rev. Immunol., 7, 445. 16. Hofer, E., Duchler, M., Fuad, S.A., Houchins, J.P., Yabe, T., and Bach, F.H. (1992). Immunol. Today, 13, 429. 139

G. De Libero 17. Dellabona, P., Casorati, G., Friedli, B., Angman, L., Sallusto, F., Tunnacliffe, A., Roosneck, E., and Lanzavecchia, A. (1993). J. Exp. Med., 177, 1763. 18. Kapsenberg, M.L., Wierega, E.A., Bos, J.D., and Jansen, H.M. (1991). Immunol. Today, 12, 392. 19. Harrison, L.C. (1992). Immunol. Today, 13, 348. 20. De Libero, G., Rocci, M.P., Casorati, G., Giachino, C., Oderda, G., Tavassoli, K., and Migone, N. (1993). Eur. J. Immunol., 23, 499. 21. Minutello, M.A., Pileri, P., Unumatz, D., Censini, S., Kuo, G., Houghton, M., Brunetto, M.R., Bonino, F., and Abrignani, S. (1993). J. Exp. Med., 178, 17. 22. Mayer, L. and Shlien, R. (1987). J. Exp. Med., 166, 1471. 23. Hershberg, R., Eghtesady, P., Sydora, B., Brorson, K., Cheroutre, H., Modlin, R., and Kronenberg, M. (1990). Proc. Natl. Acad. Sci. USA, 87, 9727. 24. Chien, Y.H., Iwashima, M., Kaplan, K.B., Elliott, J.F., and Davis, M.M. (1987). Nature, 327, 677. 25. Band, H., Hochstenbach, F., McLean, J., Hata, S., Krangel, M.S., and Brenner, M.B. (1987). Science, 238, 682. 26. Born, W.K., O'Brien, R.L., and Modlin, R.L. (1991). FASEB J., 5,2699. 27. Janeway, C.J., Jones, B., and Hayday, A. (1988). Immunol. Today, 9,73. 28. Born, W., Hall, L., Dallas, A., Boymel, J., Shinnick, T., Young, D., Brennan, P., and O'Brien, R. (1990). Science, 249,67. 29. Janis, E.M., Kaufmann, S.H., Schwartz, R.H., and Pardoll, D.M. (1989). Science, 244, 713. 30. Triebel, F. and Hercend, T. (1989). Immunol. Today, 10, 186. 31. Falini, B., Flenghi, L., Pileri, S., Pelicci, P., Fagioli, M., Martelli, M.F., Moretta, L., and Ciccone, E. (1989). J. Immunol., 143, 2480. 32. Casorati, G., De Libero, G., Lanzavecchia, A., and Migone, N. (1989). /. Exp. Med., 170, 1521. 33. De Libero, G., Casorati, G., Giachino, C., Carbonara, C., Migone, N., Matzinger, P., and Lanzavecchia, A. (1991). /. Exp. Med., 173, 1311.

140

lo] The application of mass spectrometry to the analysis of peptides bound to MHC molecules ANDREA L. COX, ERIC L. HUCZKO, VICTOR H. ENGELHARD, JEFFREY SHABANOWITZ and DONALD F. HUNT

1. Introduction Cellular 'self and foreign proteins are constantly being degraded and the resulting peptides presented by MHC molecules on the cell surface. Cytotoxic T-lymphocytes recognize cells that express new antigens as a result of viral infection, cellular transformation, or tissue transplantation. Sequence analysis of peptides presented to the immune system in association with class I MHC molecules is an important initial step in the design and development of peptide vaccines or other immune regulators effective against such disease states as AIDS, cancer, and autoimmune disorders. Identification of peptides associated with a particular disease is challenging because these antigens are usually present in a complex mixture dominated by peptides derived from normal cellular proteins. In the case of HLA-A2.1, extraction of peptides from the MHC cleft results in over 10000 distinct peptides (6). Tandem mass spectrometry is unique among other methods for direct peptide sequencing in its ability to derive sequence information from individual peptides present at low levels in complex mixtures. Information obtained from mass spectra has been used to estimate the number of peptides presented, to determine a binding motif for particular MHC molecules, to identify variations in peptide length and sequence which result from antigen processing or presentation defects, and to sequence the peptide portions of T-cell epitopes (1-6). Tandem mass spectrometry provides the mass value and the relative abundance of peptides present in a mixture with as little as 5 fmol and sequence information about peptides presented on the cell surface with as little as 30 fmol of material (6). Microcapillary HPLC in combination with electrospray ionization/tandem mass spectrometry has been used to sequence peptides derived from both class I and class II MHC molecules. This chapter will focus on class I peptide

Andrea L. Cox et al. analysis. The protocols given for peptide isolation by immunoaffinity purification, fractionation by HPLC, and analysis by mass spectrometry are generally applicable to identification of peptides relevant in a number of different disease states.

2. Preparing samples for analysis by electrospray ionization mass spectrometry 2.1 Immunoaffinity purification of MHC-associated peptides 2.1.1 Construction of immunoaffinity columns The first step in analysis of class IMHC peptides by tandem mass spectrometry involves immunoaffinity chromatography. Immunoaffinity purification can be accomplished either by the addition of affinity resin directly to a detergent solubilized cell lysate (batch purification) or by passing the lysate through a column containing the resin. We estimate that the column method is up to three times more efficient than the batch procedure. Since the amount of expressed class I differs among cell lines and the affinity of each antibody varies, some adjustment of quantities may also be necessary for purification of sufficient quantities of material. As a guideline, we use approximately 5 mg of murine monoclonal antibodies to purify 500 (xg-700 u,g of class I heavy chain from 1 X 1010 lymphoblastoid cells with an estimated copy number of 0.8-1 x 106 per cell. Protocol 1A describes construction of an immunoaffinity column for antibodies that have a high affinity for protein A (7). Hydrazide coupling of antibody to a solid matrix (Protocol 1B) is more effective for antibodies with low or negligible affinity for protein A. This method utilizes periodate oxidation of the antibody carbohydrate to produce aldehyde groups that react with immobilized hydrazides on agarose beads to form stable covalent hydrazone linkages. Because the antibody is bound via the carbohydrate in the Fc region, the combining sites are believed to be consistently oriented away from the matrix. This alleviates the problem of blocked combining sites encountered using primary amino acid coupling. Protocol 1. Construction of an immunoaffinity column Reagents . 50 mM Tris-HCI, pH 8.0, 150 mM NaCI (Tris • 25 mg NalCV1.2 ml deionized water. Make immediately before use buffered saline) at 4°C • Hydrazide agarose beads in isopropanol • Protein A-sepharose (PAS, Sigma Chemi(BioRad Laboratories, Inc. Hercules, CA) cals), 50% slurry in Tris buffered saline • 5 mL of 1.5 mg ml"1 antibody in 100 mM • 100 mM sodium acetate, pH 5.5, 150 mM sodium acetate, pH 5.5, 150 mM NaCI NaCI

142

10: The application of mass spectrometry • Phosphate-buffered saline (PBS): 4.3 mM NaHP04 1.4mMKH2PO4 2.7 mM KCI 140 mM NaCI

• 100 mM sodium acetate, pH 5.5, 0.5 M NaCI . PBS with 0.5 M NaCI . PBS with 0.04% azide

A. IgG antibodies that bind well to protein A 1. Pour 10 ml of the PAS slurry into an Econo column (1.5 cm diameter, 10 cm long, BioRad Laboratories, Inc., Hercules, CA). Allow to settle briefly without flow, and then drain down to the top of the resin bed. Wash with five volumes of Tris buffered saline. 2. Pass 15 ml of a solution of 0.5 mg ml"1 of PAS-purified antibody in Tris buffered saline over the column twice. Approximately 1 mg of mouse monoclonal antibody of the lgG2a or lgG2b subclasses is bound by 1 ml of PAS. Measure the optical density at 280 nm of the antibody solution before and after passage to determine the extent of antibody binding. 3. Equilibrate the column with four volumes of Tris buffered saline at 4°C. B. Antibodies that do not bind well to protein A 1. Add 0.5 ml of freshly prepared NalO4 solution to 5 ml of the antibody solution in a 17 x 125 mm snap-capped tube. Cover the tube with foil and turn end over end for 1 h at room temperature. 2. Remove the unreacted NalO4 from the antibody by concentrating the solution using an Amicon Ultrafiltration device with a YM100 membrane. Dilute with 100 mM sodium acetate, pH 5.5, 150 mM NaCI, and reconcentrate to the original volume several times to achieve at least a 1:1000 dilution of the Nal04. 3. De-gas the antibody solution by placing it under vacuum for 10-15 min. Avoid prolonged storage or dialysis, which may cause formation of a precipitate that must be removed before proceeding. 4. Centrifuge the hydrazide agarose beads at less than 200 x g and remove the isopropanol by pipette. Wash the beads five times by centrifuging with 10 volumes of 100 mM sodium acetate, pH 5.5, 150 mM NaCI. 5. Make a 50% slurry of beads in 100 mM sodium acetate, pH 5.5, 150 mM NaCI and pour into an Econo column (1.5 cm diameter, 10 cm long, BioRad Laboratories, Inc. Hercules, CA). 6. Allow the buffer to drain to the top of the column bed. 7. Add the oxidized antibody solution to the top of the column and recirculate for 5a-12 h using a peristaltic pump. This method minimizes precipitation of the antibody from solution. 143

Andrea L. Cox et al. Protocol 1. Continued 8. Wash the column with one volume of 100 mM sodium acetate, pH 5.5, 0.5 M NaCI. 9. Measure the optical density at 280 nm in the eluate to determine how much antibody was bound. 10. Wash the column first with three volumes of PBS with 0.5 M NaCI, then with one volume of PBS with 0.04% azide. Transfer the beads as a 50% slurry to a new Econo column prior to use. Store at 4°C. 11. Gel electrophoresis performed on 5-15 u.l of beads treated with sodium dodecyl sulfate (SDS) sample buffer may be used to confirm the amount of antibody bound based on the amount of light chain released (see step 9 of Protocol 2).

2.1.2 Purification of MHC-associated peptides Since the mass spectrometer is a universal detector for almost all organic and inorganic compounds regardless of whether they absorb UV light in an HPLC instrument, all the washing steps listed in Protocol 2 must be performed as described to remove all contaminants. All of the filters used should be washed multiple times with the buffer or solvent used in the following step of the procedure to ensure that any contaminants are removed before the peptide is added. Protocol 2. Isolation of peptides bound to class I molecules Reagents • Lysis buffer (use 0.2 n-m filtered deionized water and very clean containers): 150 mM NaCI, 20 mM Tris HCI (pH 8.0), 1% CHAPS • PBS (see Protocol 1B) • 100 \iM iodoacetamide, 18.5 mg ml"1 in lysis buffera,1:100 (v/v) • 5 jig/mL aprotinin, 5 mg ml"1 in lysis buffera,1:1000 (v/v) • 10 ng/mL leupeptin, 10 mg ml"1 in lysis buffer8,1:1000 (v/v) • 10 tig/mL pepstatin A, 10 mg ml"1 in methanol8,1:1000(v/v)

• 5 mM EDTA, 500 mM6, 1:100 (v/v) • 0.04% azide, 20% in deionized water", 1:500 (v/v) • 1 mM PMSFd, 17.4 rng ml"1 in isopropanol", 1:100 (v/v) • Immunoaffinity columns from Protocol 1 • 20 mM Tris, pH 8.0, 150 mM NaCI . 20 mM Tris, pH 8.0, 1 M NaCI . 20 mM Tris, pH 8.0 • 0.2 N acetic acid • Glacial acetic acid

A. Lysis of cells 1. Harvest cells from culture by centrifugation and pool. Disrupt the pellet by vortexing, and wash the cells two times in ice-cold PBS. Cell pellets may be frozen at -80°C at this point, so that several successive cell harvests may be pooled and extracted at one time. 2. ALL SUBSEQUENT STEPS SHOULD BE CONDUCTED AT 4°C. Resus144

10: The application of mass spectrometry pend the fresh or frozen cell pellets in lysis buffer containing freshly added protease inhibitors. Use 1 ml of buffer per 1 x 10s cells. Stir or rock gently for 1 h. 3. Centrifuge at 100 x kg for 1 h to remove cellular debris. 4. To the end of a 60 ml syringe, connect in series a 0.8 |xm low protein binding prefilter (Gelman Sciences no. 4187) and then a 0.2 jtnri low protein binding filter (Gelman Sciences no. 4192). Pass the lysed cell supernatant through both filters, changing filters if they become clogged. B. Immunoaffinity isolation of class I molecules 1. Assemble in series the immunoaffinity columns prepared using Protocol 1. Use a short 200-300 |xL Sepharose CL-4B column first to capture any material remaining in the lysate which may clog the antibody columns. This column may be replaced during the procedure if the flow rate diminishes significantly. The second column in the series should contain an irrelevant antibody that can be used to derive a negative control peptide extract. Place one column, or more if several class I molecules are being isolated, containing class l-specific antibodies last in the series. Equilibrate all of the columns with lysis buffer. Run the filtered lysate through this series at no more than 1 ml min~1. 2. Separate the columns and wash in parallel each of the negative control and experimental affinity columns with: • 2 column volumes of lysis buffer; • 30 column volumes of 20 mM Tris, pH 8.0,150 mM NaCI; • 30 column volumes of 20 mM Tris, pH 8.0,1.0 M NaCI; • 30 column volumes of 20 mM Tris, pH 8.0. 3. Allow the liquid to drain down to the top of the gel bed. Add one column volume of 20 mM Tris, pH 8.0, cover the top of the column with Parafilm, and resuspend the beads by gentle rotation. Remove 20-50 H.L of the suspension for later quantitation of bound class I by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE). Allow the beads to settle and the excess liquid to drain. 4. (a) PAS columns: Elute with four volumes of 0.2 N acetic acid, collecting 1.5 ml fractions in 1.7 ml Eppendorf tubes. Add 100 ^L of glacial acetic acid to each fraction, and place in a boiling water bath for 5 min. (b) Hydrazide agarose columns: Elute with three volumes of 10% acetic acid, collecting 1.5 ml fractions in 1.7 ml Eppendorf plastic tubes (National Scientific Supply Company, Inc. San Rafael, CA) tubes. Place each fraction in a boiling water bath for 5 min, and then chill on ice. 145

Andrea L. Cox et al. Protocol 2.

Continued

5. Transfer the contents of each Eppendorf tube into the top reservoir of a 5000 Da cut-off filter units (Millipore Corporation, Marlborough, MA, UFC4LCC25). Centrifuge the units at 3 500 x g for approximately 5 h at 4°C. Save both the filtrate and retentate. (Prior to use, the unit should be prewet with 1 ml of 10% acetic acid, spun for 1 h, and all liquid in both reservoirs discarded.) 6. Transfer the filtrate to 1.7 ml Eppendorf tubes and freeze at -80°C. 7. Concentrate the peptide extracts to a final volume of ~ 100 (il per tube using vacuum centrifugation without heat. Do not allow the samples to thaw completely or to dry- Pool the concentrates in one Eppendorf tube, and wash each original tube with 50 ^.l of 10% acetic acid. Pool the washes (with the concentrate if possible), and vacuum centrifuge again. Repeat as necessary until the total extract and washes have been concentrated to a final volume of no more than 250 jit. Refreeze at -80°C. 8. Use SDS-PAGE to estimate the amount of class I heavy chain present in the beads removed in step 3. Remove the excess liquid and add 40 |xl of Laemmli sample buffer to the gel portion removed for class I quantitation, mix and boil for 5 min, and then centrifuge lightly to pellet the beads. Analyse 40 |o.l of the supernatant on a 12% SDS-PAGE gel, stained with Coomassie blue. Compare the intensity of the class I heavy chain band to those of a series of ovalbumin standards in the range of 0.5-5 M.Q each. "Store at -20°C. "Store at room temperature. Store at 4°C. "Add PMSF last to lysis buffer.

c

3. Reverse-phase HPLC (RP-HPLC) of eluted peptides 3.1 First dimension chromatography After concentration by vacuum centrifugation of the eluted peptides, the sample should be checked by electrospray ionization tandem mass spectrometry before RP-HPLC to verify that the 5000 Da filter units successfully removed all traces of protein, especially p2-microglobulin. If there is a large amount of |32-microglobulin present it will elute across the gradient upon fractionation by RP-HPLC, making detection of lower abundance peptides more difficult. Filtration of individual HPLC fractions results in large losses of material, but refiltration is possible if the |32-microglobulin is detected before fractionation. Fractionation of the mixture by RP-HPLC is the next 146

10: The application of mass spectrometry step. Protocol 3 describes first dimension RP-HPLC separation conditions for 0.1-15 nmol of peptide extract. In general, sample contact surfaces such as sample loops and column inlet and outlet tubing should be composed of PEEK rather than stainless steel to minimize adsorption. For all HPLC separations, avoid collecting fractions in siliconized or 'slick' tubes because organic solvents leach phthalates from the coatings of these tubes. Phthalates are cytotoxic and interfere with analysis by certain types of mass spectrometry. If the peptides are being used in identification of the peptide epitope of a cytotoxic T-lymphocyte (CTL) of interest, use a standard 51Cr release assay to identify the fraction containing the peptide capable of reconstituting that CTL epitope (8). Protocol 3. RP-HPLC fractionation of 0.1-15 nmol of peptide Reagents • Solvent A: 0.1% trifluoroacetic acid (TFA) in deionized water

. Solvent B: 0.085% TFA and 60% HPLC grade acetonitrile in deionized water

Method 1. Inject the peptides on to a Brownlee narrow-bore C-18 Aquapore column (2.1 mm x 3 cm, 300 A, 7 jtm) and elute with a 65 min gradient of v/v 0-15% (5 min), 15-60% (50 min), and 60-100% (10 min) solvent B in solvent A with a flow rate of 200 \t,L min"1. 2. Collect fractions at one minute intervals into 1.5 ml screw cap polypropylene tubes (Sarstedt, Inc. Newton, NC). If the peptides are being used in identification of the peptide epitope of a CTL of interest, vortex, transfer the portion required for biological analysis to another polypropylene tube, and immediately freeze both portions on dry ice. Otherwise, freeze the entire fraction on dry ice.

3.2 Second dimension chromatography There are usually too many peptides present in the active fraction to identify the peptide capable of reconstituting the CTL epitope, so some of the candidates are eliminated by rechromatographing the biologically active first dimension fraction. Two solvent systems that are particularly effective for second dimension chromatography following separation using TFA are (Protocol 4) heptafluorobutyric acid (HFBA) and hexafluoroacetone (HFA). Using HFBA as an ionic modifier produces longer retention times for most peptides, while HFA causes elution at lower concentrations of acetonitrile. HFA provides a separation that differs more from the separation achieved with TFA than does HFBA, but the basic pH of the HFA solvent shortens both narrow-bore and microcapillary column life. The fraction may be 147

Andrea L. Cox et al. divided into two portions and one half separated using HFA while the other half is separated using HFBA as the ionic modifier. The biologically active fractions from each half should contain a unique peptide profile because of the different solvents used, but the peptide responsible for reconstituting the epitope should be in common to active fractions from both solvent systems. If the amount of material is too small to permit division of the sample into two portions, the entire fraction may be rechromatographed using either of the two organic modifiers. Protocol 4. Second dimension RP-HPLC fractionation Reagents • Solvent A: 0.1% HFA (Aldrich) in deionized water, pH adjusted to 8.1 with 14.8 M ammonium hydroxide (Protocol 4B) • Solvent B: 100% HPLC grade acetonitrile (Protocol 4B)

. Solvent A: 0.1% HFBA (Aldrich) in deionized water (Protocol 4A) • Solvent B: 0.1% HFBA in 100% HPLC grade acetonitrile (Protocol 4A)

A. Heptafluorobutyric acid (HFBA) separation 1. Concentrate first dimension fractions by about 4-fold in order to remove acetonitrile from the previous separation. 2. The gradient used depends on the elution time of the first dimension fraction. As a general rule, use a sharp gradient (1.5%B mirr1) until the concentration of acetonitrile is approximately 10% acetonitrile lower than the percentage at which the first dimension fraction eluted and a shallow gradient (0.7%B min'1) thereafter. For example, if the peptide eluted at 30% acetonitrile in the first dimension of separation, use a sharp gradient until the solvent is 20% acetonitrile and then a shallow gradient from 20% acetonitrile until 60% acetonitrile is reached. Use a flow rate of 200 |xL min"1. 3. Collect fractions at one minute intervals into screw-capped (size) polypropylene tubes (Sarstedt, Inc. Newton, NC). If the peptides are being used in identification of the peptide epitope of a CTL of interest, vortex, transfer the portion required for biological analysis to another polypropylene tube, and freeze immediately both portions using dry ice. Otherwise, freeze the entire fraction on dry ice. B. Hexafluoroacetone (HFA) separations 1. Concentrate first dimension fractions by about 4-fold in order to remove acetonitrile from the previous separation. 2. The gradient used depends on the elution time of the first dimension fraction. As a general rule, use a shallow gradient (0.7%D/min) until the concentration of acetonitrile is approximately 10% acetonitrile

148

10: The application of mass spectrometry higher than the percentage at which the first dimension fraction eluted and a sharp gradient (1.5% D/min) thereafter. For example, if the peptide eluted at 30% acetonitrile in the first dimension of separation, use a shallow gradient until the solvent is 40% acetonitrile and then a sharp gradient from 40% acetonitrile until 60% acetonitrile is reached. Use a flow rate of 200 pi mirf1. 3. Collect fractions at one minute intervals into screw-capped (size) polypropylene tubes (Sarstedt, Inc. Newton, NC). If the peptides are being used in identification of the peptide epitope of a CTL of interest, vortex, transfer the portion required for biological analysis to another polypropylene tube, and freeze immediately both portions using dry ice. Otherwise, freeze the entire fraction on dry ice.

3.3 Column effluent splitting device If there are still so many peptides present in the active fraction that more material is required for sequencing than is available, some of the peptides can be eliminated as candidates by rechromatographing the biologically active second dimension fraction. In several cases, the peptides of interest have not been present in large enough quantities to survive three dimensions of chromatography. For this reason, we constructed an apparatus to minimize sample loss in further fractionation (see Figure 1). The apparatus splits the effluent from a microcapillary column, directing 5/6 of the sample into the mass spectrometer and the remaining 1/6 of the material into the wells of a microtitre plate. It was previously impossible to split such a low volume of effluent without significantly reducing chromatographic resolution, but there is one design that splits the sample and maintains resolution. The end result is a record of the masses of the peptides deposited into each well. A 51Cr release assay may then be used to determine which wells contain the peptide responsible for biological activity (8). The number of candidates for sequencing may thus be reduced with minimal sample loss because the peptide does not have to be transferred from fraction collection tubes to the mass spectrometer or to assay plates.

4. Mass spectrometry for peptide analysis 4.1 Electrospray ionization mass spectrometry for peptide mass analysis Advantages to the electrospray ionization method include direct introduction of HPLC column effluent into the mass spectrometer and high sensitivity and ionization efficiencies for peptides that permit detection at the femtomole level. We record mass spectra on Finnigan (San Jose, CA) TSQ-70, TSQ700, and TSQ-7000 triple quadrupole mass spectrometers equipped with the

149

10: The application of mass spectrometry Figure 1. Electrospray ionization tandem mass spectrometry system with online microcapillary column effluent splitter to direct the effluent simultaneously to the mass spectrometer and to the wells of a microtitre plate. If the sample is being split so that a portion is used for a 51Cr release assay, then a microcapillary HPLC column (100 |o,m by 22 cm) is butt connected using a zero dead volume union (Valco) to two small capillaries of different lengths and interior diameters (25 ixm and 40 ixm ID, Polymicro Technologies). The column is eluted into the union with a gradient of acetonitrile. The larger of the two capillaries directs 5/6 of the material into the electrospray ionization source and mass spectra are recorded on the material using a triple quadrupole mass spectrometer. The range of m/z values from 300 to 1500 is scanned every 1.5 s and the resultant spectra are summed. The 20 ixm capillary deposited the remaining 1/6 of the material into 50 p.L of media in microtitre plate wells. Timing of the splitting is adjusted so that m/z values of the peptides are recorded at the instant in which they are deposited into the well, providing a record of the peptides present in each well. A chromium release assay is then used to determine which well contains the peptide portion of the epitope. If the sample isn't being split, a microcapillary HPLC column (75 jj,m by 12 cm) is used to introduce sample into the ionization source. The acidic solution of protonated peptides eluted from the microcapillary column is sprayed as a fine mist from the tip of the electrospray needle, which is operated at a voltage differential of 5000 volts and with a coaxial sheath flow of 2-4 jiL/min of a 3:1 mixture of methanol:0.5% acetic acid. The ion sampling capillary is operated at ground and is heated to 150°C. The ions produced are focused into the first quadrupole (Q1) using an octapole and a pair of electrostatic lens. Differential pumping and dc offset voltages direct the peptide ions through the instrument quadrupoles. If the mass of the peptide is being determined, rf and dc potential are applied to the quadrupole rods to cause the first quadrupole (Q1) to function as a mass filter, thereby separating ions according to their mass-to-charge ratios (m/z). The standard mass range scanned is m/z 300 to 1500 every 1.5 seconds. In order to effect transmission of all ions from Q1 through the second (Q2) and third (Q3) quadrupoles, only rf potential is applied. The ions are then detected by a high voltage conversion dynode (15 keV) electron multiplier and the resultant spectra are summed. The resulting spectrum contains ions of (M+nH)+n, where n is greater than or equal to one, that are characteristic of the molecular masses of each peptide in the mixture. If sequence information is required, Q1 is set to pass all ions within a 2-3 mass unit window around the peptide ion of interest. Q2 is used as a collision chamber (rf only) in which the peptide ions undergo multiple low energy (10-40 eV) inelastic collisions with argon atoms (2-3 mtorr present). Translational energy is transformed into vibrational energy, causing virtually random fragmentation at the amide bonds of the peptide backbone. Collision offset helps determine the degree of fragmentation and varies as a function of the size and charge state of the ion of interest. For singly charged class I peptides, the offset ranges from -30 eV for smaller peptides to -38 eV for larger peptides. Similarly, the ranges of collision energy for doubly and triply charged peptides are -16 to -26 eV and -15 to -20 eV, respectively. The neutral species resulting from fragmentation are pumped away by the vacuum system while fragment ions are transmitted to Q3 for mass analysis using a scan rate of 500 Da/sec. Q3 is operated at less than unit resolution to maximize ion transmission by passing all ions within a 3-4 mass window.

Finnigan APCI/ESI electrospray ionization source (see Figure 1). Samples are introduced into the instrument using a microcapillary HPLC column constructed as described in Protocol 5. PRP-1 packing material is chosen for hydrophilic samples that are not retained by C-18 on silica or by POROS R/H II supports. POROS R/H II is less prone to clogging and peptide losses are lower than those obtained with either C-18 on silica or PRP-1. LC Pack151

Andrea L. Cox et al. ing International, San Francisco, CA is a source of packed fused silica capillaries if purchasing prepared columns is preferred to packing. Samples are loaded onto the column and eluted into the mass spectrometer using a gradient of 0-80% acetonitrile in 0.5% acetic acid over 12 min (Protocol 6). The ion current is reduced by a factor of five if trifluoroacetic acid is used as the eluant instead of acetic acid. Protonated peptides are sprayed as a fine mist from the electrospray needle, resulting in charged droplets. The charged droplets are then desolvated using a stream of nitrogen gas and a heated ion sampling capillary, producing positively charged peptide ions in the gas phase. Electrospray ionization places positive charges on the N-terminus and on the side chains of basic residues present in peptides, thus the charge state is usually +1 or +2 and less frequently +3 for the nonamers derived from class I molecules. The ions produced are then directed into the first quadrupole (Ql) by an octapole and a pair of electrostatic lenses. If the mass of the peptide is being determined, ions are separated in Ql on the basis of mass-to-charge (m/z) value and the second and third quadrupoles are set to pass all ions. Ions are detected by a conversion dynode and electron multiplier. Total ion chromatograms, plots of total ion current mass spectrum number are obtained by recording mass spectra every 1.5 sec over the range of m/i from 300 to 1500. Since the peptides contained in a single HPLC fraction typically elute from the column for a period of about 3 min, approximately 100 scans are acquired under data system control. Individual spectra can then be replotted to show the masses of the peptides eluting at any particular time interval. Under the experimental conditions employed in these experiments, no fragmentation of the peptides is observed so that all of the observed signals correspond to the whole peptide. This mass analysis requires loading a minimum of 5 fmol of peptide on to the column. Protocol 5. Preparation of microcapillary HPLC columns Method 1. Cut with a capillary cleaving tool (Supelco, Inc., Bellefonte, PA) a piece (40-60 cm) of fused silica capillary (190 jjim O.D., 75 urn I.D., SGE, Ltd. Ringwood, Australia or 360 |a,m O.D., 100 jjim I.D., Polymicro Technologies, Phoenix, Az.). 2. Burn off about 3 mm of the polymer coating at one end using an orange flame from a microtorch (Microflame, Minnetonka, MN) and inspect the end under a stereomicroscope to ascertain that the end is square. 3. Tap the exposed end of the capillary into Lichrosorb Si 60 5 |j,m spherical silica particles (EM Science, Giggstown, NJ) 50-60 times until a small plug of about 0.1 mm of particles is formed. The plug will appear under the microscope as grey, densely packed material.

152

10: The application of mass spectrometry 4. Quickly pass the plug through a blue flame on the microtorch so that it glows red for a fraction of a second to sinter the particles and form a frit. Repeat 2-3 times. Under the microscope, the plug will appear to be more condensed and to be adhering to the walls of the capillary. 5. To test that the frit will hold under running conditions, flush propanol through the capillary at 1000 psi of pressure using a stainless steel high pressure vessel (see Figure 2). The frit will appear translucent. If the frit fails, dry out the capillary, cut off 0.5 cm of the previously plugged end, and repeat the protocol. 6. Make a slurry of packing material by adding it to 0.5 ml of 2-propanol in a 1.5 ml glass vial. For packing material, use 10-20 mg of 10 m spherical C-18 (YMC Corp., Morris Plains, NJ), 2-4 mg of POROS II R/H polymer reversed-phase (PerSeptive BioSystems, Cambridge, MA), or 2-4 mg PRP-110 m particles (Hamilton Corp., Reno, NV). 7. Add a magnetic stir bar (flea) to the slurry and place the vial inside a stainless steel pressure vessel resting upon a magnetic stir plate. Turn on the stir plate so the flea spins. 8. Push the end of the fused silica tubing without the frit through the lid of the vessel and into the slurry. 9. Pressurize the vessel with helium at a pressure of 200-500 p.s.i. to cause continuous flow of the packing material into the tubing. When the packing material has filled the first 10-15 cm (for C-18 and PRP-1) or the first 20-30 cm (for POROS R/H II) of the tubing, stop the flow of slurry by stopping the flow of helium. 10. Compress the bed and wash the remaining beads in the tubing to the packed region by pumping 5% acetic acid in water over the column for 15-25 min at 200-500 p.s.i. 11. Prior to use, connect the column to the solvent delivery system and run one blank gradient and one gradient with 10 pmol of angiotensin loaded on to the column. The flow of the column should be at least 0.5 JJL.I min"1 at 600 p.s.i. 12. Store columns with both ends immersed in deionized water in a test tube.

Protocol 6. Sample introduction into the electrospray source Method 1. Load samples on to the columns using the same type of bomb used for column packing in step 4 of Protocol 5. 153

Andrea L, Cox et al. Protocol 6. Continued 2. Determine the amount of sample loaded by measuring with a 1-5 u,l graduated disposable glass pipette the amount of solvent displaced from the column. 3. Wash the column with 0.1 M acetic acid for five minutes prior to beginning the gradient. 4. Elute the sample from the rnicrocapillary column by using a linear gradient of 0-80% acetonitrile in 0.1 M acetic acid in 15 min. We use an Applied Biosystems (Foster City, CA) Model 140A dual syringe pump to deliver the solvents. To allow proper mixing of the solvents and quick introduction of the mobile phase to the column, set the solvent delivery system to deliver between 180 and 300 uJ of solvent per min. 5. The ideal flow rate is approximately 0.5 u,l per min through 75 u,m I.D. microcapillary columns and 1 ul per min through 100 ^.m I.D. columns. Effect a pre-column split of the mobile phases by using a restriction capillary (50 u.m I.D. fused silica) which can be adjusted to change the back pressure on the system (see Figure 3).

Figure 2. Pressure vessel used for packing and for loading samples on to rnicrocapillary columns. The end of the microcapillary column is inserted through the lid of the vessel into the slurry of packing material. Pressurizaiion of the vessel to 250-500 p.s.i. occurs when the flow of helium is begun, inducing the contents of the microvial to flow into the column.

154

10: The application of mass spectrometry

Figure 3. Microcapillary HPLC chromatography system. A flow rate of 180-500 \L\ mirf1 from the solvent delivery system is used to ensure that the solvents reach the column as rapidly as possible and that proper mixing occurs. Since the flow rate through the column should be 0.5 (J min"1 for a 75 (im I.D. column and 1 p.I min~1 for a 100 n,m column, pre-column splitting is necessary. The column end without the frit is connected to the solvent delivery system using a 1/16 inch nut and either a teflon 0.3 mm ferrule (for 190 M-m O.D. columns) or a vespel 0.3 mm ferrule (for 360 n,m O.D. columns). The length of the 50 ixm capillary used as a restrictor controls the amount of liquid that is directed through the column.

4.2 Sequencing peptides with tandem mass spectrometry Three quadrupoles are connected in series to produce a triple quadrupole or tandem mass spectrometer. By using a triple quadrupole instrument, it is possible to obtain sequence information for individual peptides (9). After the mass of the peptide of interest has been determined using the conditions described in Section 4.1, the data system is used to select that particular ion for transmission through the first (Ql) of the three quadrupoles. The ions of the selected m/z are transmitted to the second quadrupole (Q2), which is operated as a collision cell. In Q2, the ions undergo 10-100 low energy collisions with argon atoms, causing virtually random fragmentation along the peptide backbone. Fragments containing varying numbers of amino acids linked together are produced in Q2 and transmitted to Q3 for mass analysis. The fragment ions are detected by a high voltage conversion dynode electron multiplier. The result is a plot of relative ion abundance versus mass, called collisional activation decomposition spectra (CAD), for all of the fragments of the peptide selected. Information on up to ten peptides can be obtained in a single HPLC run, requiring a total of 30 min of instrument time. Analysis of CAD spectra allows the use of as little as 15 fmol of material to obtain peptide sequence information, but more routinely 100-300 fmol of material are 155

Andrea L. Cox et al.

Figure 4. CAD mass spectrum of peptide (M+2H)++ ions at m/z 519 using material derived from JY cells after fractionation by HPLC. Predicted masses for fragment ions of types -b and -y are written above and below the deduced sequence, respectively. Ions of type-b contain the N-terminus of the peptide and have the form H(NHCH(R)CO)n+, where n is any number between one and the total number of amino acid residues. Ions of type-y contain the C-terminus and have the composition H2(+NHCH(R)CO)nOH. Subtraction of m/z values for any two fragments that differ by a single amino acid generates a value that specifies the mass and thus the identity of the extra amino acid in the larger fragment (9). For example, subtraction of y, from y8 (875-762) provides the residue mass for leucine: 113. This indicates that the extra residue present in y8 is a leucine. This procedure may be repeated using both the b- and the y-ions to generate the rest of the sequence. Residues 1-8 were identified using y ions, residue 9 was deduced from ions of fragment type-b. Ions of type-a differ from b-ions by 28 Da, which corresponds to a loss of CO. Although they are not observed in this spectrum, ions produced as a result of loss of water and ammonia may also be observed. Loss of water occurs in fragments that contain the amino acids serine and threonine at any position and glutamic acid at the Nterminus. Loss of ammonia occurs if the fragments contain the amino acids arginine, lysine, glutamine, or asparagine. Low mass ions of the type +NH2=CHR that are characteristic of the amino acid composition of the peptide are observed for some but not for all amino acids. Since lie and Leu are of identical mass, they cannot be distinguished in CAD spectra recorded on a triple quadrupole mass spectrometer. Since these residues do have different retention times on microcapillary columns, coelution experiments often allow determination of the naturally occurring sequence. Ala = alanine. Leu = leucine, lle = isoleucine, His = histidine, Pro = proline, Tyr = tyrosine, and Val = valine.

required. A sample of a CAD spectrum generated on a nonamer peptide from HLA-A2.1 is shown in Figure 4. In the ideal experiment, all the amide bonds in a particular peptide would fragment at identical rates and the charge would be evenly distributed over the N- and C-terminal fragments. Unfortunately, fragmentation occurs at different rates in the mass spectrometer and the relative abundances of fragments can differ by more than an order of magnitude. When the level of material used approaches the instrument's lower limit of detection, some of the less abundant fragments are not discernable above the background noise. Difficulty in assigning fragments as being derived from one end of the peptide rather than from the other and unequal fragmentation make sequence 156

10: The application of mass spectrometry analysis difficult and time-consuming. Spectra recorded on additional aliquots of chemically modified peptide supplement information obtained with the unmodified peptides.

5. Chemical modification of peptides Chemical modifications are performed on the peptide of interest to facilitate sequencing. These modifications can provide information helpful in determining the peptide sequence by indicating the number and location of acidic amino acids and cystines, by identifying ions of type-b and -y, and by providing the identity of the N-terminal amino acid. Protocols 7A-7D describe four chemical modifications that are particularly useful in analysis of class I peptides. Protocol 7A describes esterification of the free carboxylic acid moieties of peptides. Esterification using methanolic-HCl results in an increase in mass of 14 Da for the C-terminal COOH and for each acidic (glutamic or aspartic acid) residue present in the peptide. Esterification using ethanolic HC1 results in an increase of 28 Da for each free COOH. Prolonged exposure to the esterification reagent results in shifts of 15 Da with methanolicHCl and 29 Da with ethanolic-HCl because of hydrolysis and esterification of asparagine and glutamine. A shift of 14 in the parent mass is harder to detect with triply charged peptides because the shift in observed mass is 14 divided by three or 4.7 Da for these peptides. Therefore triply charged peptides should be converted to ethyl esters instead of methyl esters. For singly and doubly charged peptides, methyl esters are preferable because either modification can be detected easily and the reaction proceeds more completely with methanol than with ethanol. Every y-ion in the spectrum can be identified by a shift of 14 or 28 because y-ions contain the C-terminus. In addition, every fragment ion of any type containing an acidic residue is also shifted. Ions that fail to shift must not contain an acidic residue or the Cterminus. Protocol 7B describes N-acetylation, which causes an increase of 42 Da for the N-terminus as well as for each lysine and cysteine residue present in the peptide. After prolonged exposure to acetylating reagent, tyrosine, threonine, histidine, and serine may be acetylated. In the absence of base catalysis, acetylation occurs only on the N-terminus. This is a less commonly used procedure for modification of class I peptides because all sites for charging are removed unless an arginine or histidine is present. If all the possible charge sites are acetylated, the signal is dramatically reduced because less of the peptide is ionized. The more highly charged a peptide, the less likely it is that all possible charge sites will be acetylated. Since most class I nonamers are not highly charged and do not contain many histidine or arginine residues, acetylation is only performed with triply charged peptides or those known to contain one of those residues. The presence of one or more histidine residues 157

Andrea L. Cox et al. is indicated by a low mass peak at m/z 110. Acetylation was very useful in analysis of peptides extracted from the class I molecule B7 because they contain arginine as part of the sequence motif (2). A single cycle of manual Edman degradation can be performed as described in Protocol 7C to remove the N-terminal amino acid of the peptide of interest. Since all b-ions in the spectrum contain the N-terminus, each b-ion decreases in mass by the mass of the residue lost from the peptide. This modification provides the identity of that first amino acid, which is helpful because the bx ion that would indicate the identity of the N-terminal residue is rarely detected. On-column carboxyamidation (Protocol 7D) is useful for verification of the presence of a cystine residue and its location within a peptide. Protocol 7. Chemical modification of peptides A. Esterification 1. Prepare the esterification solution by adding dropwise with stirring 80 (jil of acetyl chloride to 500 p.1 of the anhydrous alcohol. Allow reagent to sit at room temperature for ten minutes. 2. Add 20-50 JJL! of 3N methanolic- or ethanolic-HCI prepared in the preceding step to lyophilized HPLC fractions containing the peptides. 3. Allow the reaction to proceed at room temperature for 1-2 h for methyl esters and for 3 h for ethyl esters. 4. Remove solvent by vacuum centrifugation. 5. Resuspend the modified peptides in 15-20 |il of 5% acetic acid. B. On-column N-acetylation 1. Load the peptides on to a microcapillary column using the bomb described in step 4 of Protocol 1 and wash with distilled water for 5 minutes. 2. Immediately prior to use, add 1 ^l of acetic anhydride to 99 |J of a 200 mM ammonium acetate solution adjusted to pH 8. 3. Wash 3 |J to 4 (xl of the acetylating reagent through the column. 4. Wash the column with water for 5 min at a flow rate of 1 IJL! per min. 5. Start the gradient and elute the acetylated peptide from the column. C. Manual Edman degradation 1. Prepare a fresh solution of 5% phenylisothiocyanate (PITC) in pyridine. 2. Concentrate the HPLC fraction containing the peptide of interest to 1 n-l total volume by vacuum centrifugation.

158

10: The application of mass spectrometry 3. Add 2 M-L to 10 jj-L of the PITC solution made in step 1 Protocol 7C and vortex, overlay the solution with argon, and incubate the mixture at 45°C for 30-45 min. 4. Remove the solvent by vacuum centrifugation. 5. Treat the residue with 15 pJ of concentrated trifluoroacetic acid (sequencing grade) to release the newly formed phenylthiohydantoin (PTH) derivative. Overlay the solution with argon and incubate for 10 min at37°C. 6. Remove the solvent by vacuum centrifugation. 7. Add 15 jjj of water to the residue and vortex. 8. Extract the sample with butyl acetate (2 x 15 n,l of CH3CO2(CH2)3CH3) to remove the truncated peptides from the PTH derivatives and organic salts. 9. Remove the solvents by vacuum centrifugation of the aqueous layer. 10. Save the organic layer because some hydrophobia peptides containing lysine may be found in this layer instead of in the aqueous layer. This layer should be checked if the peptide of interest is not found in the aqueous layer. D. On-column carboxyamidation 1. Load sample on to microcapillary packed with POROS R/H II using the vessel described in step 4 of Protocol 1. Wash the column with water for 5 min. 2. Flow a solution of 10 mM iodoacetamide in 100 mM Tris-HCI, pH 8.5 through the column at the rate of 0.25 (J min"1 (75 (j,m column) to 0.75 |il min~1 (100 ^.m column) for 15 min. 3. Wash the column with 0.1 mM acetic acid for 5 min before beginning the gradient to ensure that the Tris buffer is completely removed.

Acknowledgements This work was supported by grants AI20963 (to VHE) and GM37537 and AI33993 (to DFH) from the United States Public Health Service. ELH was supported by the Medical Scientist Training Program of the University of Virginia.

References 1. Hunt, D. F., Henderson, R. A., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., Cox, A. L., Appella, E., and Engelhard, V. H. (1992). Science, 255, 1261. 159

Andrea L. Cox et al. 2. Huczko, E. L., Bodnar, W. M., Benjamin, D., Sakaguchi, K., Zhu, N. Z., Shabanowitz, J., Henderson, R. A., Appella, E., Hunt, D. F., and Engelhard, V. H. (1993). /. ImmunoL, 151, 2572. 3. Chicz, R. M., Urban, R. G., Gorga, J. C, Vignali, O. A. A., Lane, W. S., and Strominger, J. L. (1993). J. Exp. Med., 178, 27-47. 4. Henderson, R. A., Michel, H., Sakaguchi, K., Shabanowitz, J., Appella, E., Hunt, D. F., and Engelhard, V. H. (1992). Science, 255,1264. 5. Henderson, R. A., Cox, A. L., Sakaguchi, K., Appella, E., Shabanowitz, J., Hunt, D. F., and Engelhard, V. H. (1993). Proc. Natl Acad. Sci. USA, 90, 10275. 6. Cox, A. L., Skipper, J., Chen, Y., Henderson, R. A., Darrow, T. L., Shabanowitz, J., Engelhard, V. H., Hunt, D. F., and Slingluff, C. L., Jr. (1994). Science, 264, 716-719. 7. Langone, J. J. (1978). J. ImmunoL Methods, 24, 269. 8. Slingluff, C. L., Jr., Cox, A. L., Henderson, R. A., Hunt, D. F., and Engelhard, V. H. (1993). J. ImmunoL, 150, 2955. 9. Hunt, D. F., Yates, J. R., Shabanowitz, J., Winston, S., and Hauer, C. R. (1986). Proc. Natl Acad. Sci. USA, 83, 6233.

160

11 The molecular basis of alloreactivity LIZ LIGHTSTONE, SARAH DEACOCK, TRICIA HEATON and ROBERT LECHLER

1. Introduction Allogeneic major histocompatibility (MHC) molecules are capable of stimulating uniquely strong primary immune responses. This is illustrated in vitro in the mixed lymphocyte reaction, and in vivo leads to organ graft rejection, and graft versus host disease in recipients of allogeneic bone marrow transplants. The strength of the allo-response is due to the very high frequency of T cells in the 'normal' T cell repertoire which are able to react with MHC alloantigens. Whilst this phenomenon was observed many years ago (1) it is only relatively recently that the tools have become available to characterize the molecular basis of allo-recognition. In this chapter, the methods used by ourselves and others to elucidate the structure-function relationships between allo-reactive T cells and their ligands are described. Studies of the structural basis of allo-recognition have revealed that the interaction between T cells and antigen presenting cells (APC) in an alloresponse involves the tri-molecular complex of T cell receptor (TCR), MHC complex and peptide, i.e. the same elements used in antigen-specific interactions. How then is the high frequency of allo-reactive T cells explained? Two main hypotheses have been put forward: (a) the high determinant density hypothesis and (b) the multiple binary complex hypothesis. The former envisages that the allo-reactive T cell's receptor is specific for the foreign MHC molecule itself, in particular the a-helical residues predicted to point up towards the TCR, irrespective of what, if any, peptides are bound and displayed with the allogeneic MHC molecules at the cell surface. If this is the case, all the MHC molecules of a given type on the surface of an allogeneic cell (approximately 105 class II molecules of a given isotype are expressed on the surface of a B cell) could act as ligands for the allo-reactive T cell (2). This represents at least a 100-fold increase in 'determinant density' compared to that available for an antigen-specific, self MHC-restricted T cell, since it is unlikely that more than 1% of the MHC molecules of a particular

Liz Lightstone et al. type are occupied with any one individual peptide derived from a processed protein antigen (3). Such a high density of ligand would allow T cells bearing low affinity receptors to be recruited into the allo-response, hence increasing the precursor frequency. The alternative multiple binary complex hypothesis (4) proposes that anti-MHC allo-responses are closely analogous to antigenspecific responses in that they are mediated by T cells that are specific for peptide:MHC complexes. In this case the peptide is a naturally processed peptide derived from a serum or cellular protein, and the MHC molecule is allogeneic. This would account for the high frequency of anti-MHC allo-reactive T cells because of the wide diversity of naturally processed peptides that are displayed at the cell surface by each type of MHC molecule. Indeed, recent data from peptide elution studies suggest that at least 2000 different peptides may be bound to a single type of MHC molecule (5). Thus, a single MHC allo-antigen could stimulate a large number of different T cell clones, each specific for a different peptide-MHC complex. However, whilst both these hypotheses account for the high frequency of allo-reactive T cells, at first sight they are apparently incompatible with the notion of positive selection, namely that the normal peripheral T cell repertoire has been selected in the thymus for self-MHC restriction. If positive selection for self-MHC restriction does occur, and there is much data tosupport it (6), then why can so many T cells recognize allogeneic-MHC? Detailed knowledge of the sequences of MHC alleles and elucidation of the crystallographic structure of MHC molecules supply possible explanations (7, 8). The TCR-contacting surface of class I and class II molecules contributes to MHC restriction and comprises the upper surfaces of the a-helices of either the al and a2 domains in class I and the al and (31 domains in class II molecules (Figure 1). The peptides which bind are affected by polymorphisms in the floor of the groove and in the inward facing residues of the a helices. Comparison of a-helical sequences of different alleles shows that several different allelic forms of MHC molecules share conserved sequences in their TCR-contacting, or 'histotopic', surfaces (9). However, the alleles that share TCR-contacting, MHC restriction-determining surfaces have multiple amino acid differences in the peptide-binding groove (Figure 2). Hence certain allo-responses would be focused on non-self peptides presented by an allogeneic MHC molecule that appears to the T cell to resemble self-MHC (10, 11). Alternatively, when the MHC allo-antigen displays multiple amino acid differences in the TCR-contacting regions, it is reasonable to assume that the allo-response is focused largely on the allo-MHC molecule itself. How does this possibility fit with the concept of positive selection for self MHC restriction? We would postulate that since TCRs which undergo positive selection show relatively low affinity for self MHC, with high affinity interactions resulting in negative selection in order to avoid auto-reactivity, it may be by chance that some TCRs are selected which co-incidentally have high affinity for a particular allo-MHC molecule. Since such an MHC molecule is not 162

11: The molecular basis of alloreactivity

Figure 1. Schematic representation of MHC class I and class II molecules. Class II molecules comprise a heterodimer of a and b chains whilst the class I heavy chain is non-covalently linked with (J2 migroglobulin 02M). The membrane distal portion of each molecule (formed by al+fji domains in class II and a1 and a2 in class I) contains hypervariable regions which are focused on the upper surfaces of the a-helices and the inner aspect of the peptide binding groove (formed by the a-helices and p-pleated floor of both distal domains).

present during thymic selection, those T cells survive and are able to recognize foreign MHC structures encountered on allogeneic cells. A third and novel possibility, recently shown to occur in vivo, is that one T cell may express two TCR's (12). Providing neither can trigger negative selection and one can be positively selected, the other TCR may recognize foreign MHC. The hypotheses addressed above are entirely testable and indeed we and others have provided evidence for direct recognition of exposed polymorphisms on foreign MHC molecules as well as for the contribution of bound peptide in allo-responses. In the following sections we will outline the experimental procedures used to elucidate the nature of allo-reactivity, all of which are based on an understanding of the sequences and crystal structure of MHC molecules (7,8,12,13).

2. Allo-reactive T cells Clonal populations of allo-specific T cells are a central requirement of any approach to understanding the structure function relationships in allo-recognition. T cell clones have the advantage that their antigen- and allo163

Liz Lightstone et al.

Figure 2. Schematic representation of similarity between 'histotopic' regions of two different class II MHC molecules. The histotopic regions are those purported to come in contact with the TCR directly and are represented by the light grey shaded areas. As is clear these regions are identical for the two alleles shown, which differ only at position 71 which faces into the peptide binding groove and is likely to influence the peptides bound.

specificity can be readily determined and reproduced. The generation of T cell clones and (he principles of T cell assays have been fully described in chapter 6 by Pawelec et al. and in much of our published work, e.g. ref. 14. Hence, we will not describe T cell cloning or proliferation assays here unless specific adaptations are required.

3. Cell lines for use as allo-stimulator cells In order to determine the allelic specificity of an allo-reactive T cell, comprehensively typed stimulator cells are required. Such cells have been made available by the assembly of a panel of EBV-transformed B cell lines (B-LCL) which are homozygous for specific class I and class II alleles (15). This panel of cells was assembled for and characterized in detail during the 164

11: The molecular basis of alloreactivity XI International Histocompatibility Workshop held in Japan. Such lines are efficient allo-stimulator cells in that they express both high levels of MHC class I and class II molecules as well as the costimulatory molecules necessary for full T cell activation. In particular, they can be used to determine the allospecificity of T cell clones if used in conjunction with anti-MHC antibodies which, for instance, may block class I or class II mediated stimulation. Such assays are addressed in detail in chapters 1 and 6.

4. Transfection of class II molecules into adherent and non-adherent cells A technology that has proved invaluable in analysing the phenomenon of allorecognition has been transfection of wild type or mutated MHC genes into mouse, or human, fibroblast lines or into mutant B-LCL lacking expression of particular locus products. Such an approach allows very specific questions to be asked about the nature of the allo-ligand and raises the possibility of molecular manipulation of MHC molecules to allow the various hypotheses outlined earlier to be tested.

4.1 Preparation of DNA MHC class II molecules comprise an a^ heterodimer and the genes encoding both chains need to be transfected in order to achieve cell surface expression in MHC class II negative cells. This is normally achieved by co-transfection of two vectors containing a and (3 chain encoding genes or cDNAs. This can be done simultaneously on separate eukaryotic expression vectors. The vectors should be chosen on the basis of promoter, cloning site and drug selection agent if selection is to be covalently linked. An alternative strategy involves triple transfection, using a and (3 gene-encoding vectors together with a third plasmid containing a drug-resistance gene. This may be simpler than sub-cloning one or both MHC genes into vectors that contain the drugresistance gene. In order to maximize the likelihood of high-level expression of the MHC genes, molar ratios of 10, 20 and 50:1 of MHC gene-containing: drug-resistance plasmids should be used. However, since both chains need to be expressed in the same cell it is preferable to transfect each in a vector with different covalently linked selection markers. Addition of both selection agents to the subsequent cell culture should promote survival of only those cells expressing both MHC genes and hence likely to express mature a[3 heterodimers. The DNA used should be purified using either a standard plasmid preparation (48) or a commercially available kit, e.g. Wizard Prep (Pharmacia). The DNA should be checked for quantity and quality by running on a 1% agarose gel stained with ethidium bromide. DNA need not be linearized prior to transfection.

165

Liz Lightstone et al.

4.2 Which cells to use for transfection? In general, if transfecting class II MHC genes it is preferable to use a cell which does not express class II or which expresses a locus product that can be readily distinguished from the transfected product with allele-specific monoclonal antibodies. As already mentioned, B-LCL have some distinct advantages as allo-stimulators as they co-express a variety of co-stimulatory molecules. RJ2.2.5 is a class II negative B-LCL which has defective class II expression due to a regulatory defect rather than a deletion of the class II locus (16). In any cell line to be transfected it is necessary to establish whether it is not only constitutively negative for class II expression but also whether cytokines such as gamma interferon (-ylFN) can induce expression of class II. Ml is a human fibroblast line derived from a patient with the disease xeroderma pigmentosum (17). Ml is constitutively negative for class II and invariant chain and is refractory to induction with -ylFN as judged by northern and western blotting (Lightstone, L. et al., submitted). It is important to remember that the phenotype of the class II expressed may not be identical to that on a 'professional' antigen presenting cell or even on a different cell lineage. For example, although class II can be readily expressed in Ml following transfection with DRA and DRB1 0101, the DR1 molecules on the transfectants are less stable than DR1 expressed on B-LCL, as judged by the ease with which dimers dissociate into monomers at room temperature on a western blot. Preliminary studies also suggest that the peptides eluted from affinity purified DR1 from B-LCL show some striking similarities and differences from those eluted from DR1 on Ml (Lightstone, Bobek, Stauss, and Lechler, unpublished). Despite these differences, DR1expressing Ml transfectants can present peptide to peptide-specific T cell clones very well and can also stimulate a subset of allo-specific T cell clones (63 and Lightstone et al., submitted). In contrast, the product of the same DR1 genes transfected into a myocyte line, TE671, fails to stimulate either antigen- or allo- specific T cells and indeed, anergizes them (18). A cell line that has proved extraordinarily useful for transfection of human MHC genes, is the mouse fibroblast DAP.3. It is readily transfectable, provides excellent co-stimulation and is able to process antigens. Importantly for subsequent functional assays, DAP.3 express B7.1 and mouse B7.1 can interact with human CD28. P815, a mouse mastocytoma line, has also been used extensively for transfection of mouse class I and class II genes (19).

4.3 Choice and preparation of selection agents In choosing which selection agent to use the following have to be considered: 166

11: The molecular basis of alloreactivity (a) Will the cells to be transfected die in selection media in the absence of the resistance gene? There is little point transfecting in a resistance gene if the cells are already resistant! (b) Does the cell line to be transfected already express another transfected resistance gene? If so, a different selection agent will have to be chosen. (c) Is a suitable vector available which expresses the appropriate resistance gene and a friendly cloning site for inserting the gene of interest, e.g. DRB? (d) Cost is a not insignificant factor as many selection agents are very expensive, and if it is predicted that large volumes of media will be needed in the future then it may reasonable to try to use the least expensive agent. 4.3.1 Preparation of selection agents Standard stock solutions of commonly used selection agents are given in Table 1. Note that when making up concentrations of MXH above 1X, only the concentration of the mycophenolic acid should be increased as the xanthine will precipitate out at higher concentrations at neutral pH. Hence, it is useful to make stocks of 200X mycophenolic acid and 200X XH to be mixed to the appropriate final concentrations. G418 and histidinol are acidic and MXH extremely alkaline. However, it is rarely necessary to buffer them as small amounts are usually added to cultures and they are buffered rapidly in

Table 1. Standard stock solutions of commonly used selection agents. Stock x 1x

Selection agent

Gene

Ingredients

Stock concentrations + volume of stock

G418

neomycin resistance hygromycin resistance gpt

G418

20 mg/ml 100X make up 10 mis 10 mg/ml 100x make up 10 mls 8 ml 100x 6 mg 250 mg 15 mg 1.5 ml mix at 37°C until dissolved, make up to 10 mis with dH20 250 mM 100X buffered with 100 mM Hepes

(geneticin) Hygromycin MXH

hygromycin dH20

mycophenolic acid xanthine hypoxanthine 4 M NaOH

Histidinol

histidine histidinol dehydrogenase dehydrogenase

167

200 |xg/ml 100 jig/ml

6 (ig/ml 250 |ig/ml 15 jig/ml

2.5 mM

Liz Lightstone et al. medium containing bicarbonate incubated in 5-7% CO2. Also note that the active moiety in G418 is only a fraction of any given amount and this needs to be factored in when calculating the weight required to make a 100X solution. All solutions should be filter sterilized (0.2-0.45 JJL) prior to addition to medium or cells. They should be stored at 4°C.

4.3.2 Assessing sensitivity to selection agents Prior to transfection it is necessary to check the amount of selection agent required to kill cells not expressing the selectable marker. Plate out cells in identical numbers in a 24-well plate using normal medium, e.g. RPMI1640 buffered with sodium bicarbonate (0.24% final concentration) and supplemented with glutamine (2 mM), penicillin (50 iumT1), streptomycin (50 (igmr1) and 10% FCS. Add the selection agent in a variety of concentrations such as 0, IX, 2X, 4X, 8x and 16X. Estimate the percentage cell death at 48 h intervals by eye or by counting. If using dual selection when transfecting, determine the concentrations needed when both drugs are added. The optimal concentration of selective drug is that which kills > 50% at 3 days following which the concentration should be reduced so that > 80% cells are dead at one week. Some cells are extremely resistant to selection agents and it may be preferable to allow longer for cells to die rather than to use high concentrations of expensive agents.

4.3.3 Preparation of cells for transfection Cells should be in the exponential phase of cell growth. 4.3.4 Choice of transfection method For adherent cells and the mouse mastocytoma cell line, P815-HTR, calcium phosphate transfection may be the first line approach (Protocol 1). Alternative strategies include electroporation, protoplast fusion, lipofection and micro-injection. For non-adherent cells, electroporation is normally the first method to try (Protocol 2). Alternatives include lipofection or protoplast fusion. Electroporation conditions need to be determined for each cell line used and a useful preliminary guide is to determine the conditions which lead to approximately 50% cell death 24 h after sham transfection by electroporation. In general, BLCL should be growing exponentially on the day of transfection and short pulse times should be used when electroporating. The efficiency of transfection is generally low, of the order of 1:104 for calcium phosphate. However, this yields approximately 50 colonies of cells growing in drug selection. For any cells which have a very low transfection efficiency it is worthwhile considering micro-injection of the desired DNA. 168

11: The molecular basis of alloreactivity Protocol 1. Calcium phosphate transfection of adherent cells Equipment and reagents* • 2x UBS: 280 mM NaCI, 50 mM Hepes buffer, 1.5 mM disodium hydrogen phosphate dihydrate, adjusted to pH 7.1 (important to be accurate) • Calcium chloride, 0.5 M

. TE: 10 mM Tris/HCI pH 7.5, 1 mM EDTA pH 7.5 • 20% glycerol, made up with dH20 • Sterile (autoclaved) Eppendorf tubes

Method

1. Day1 (i) (ii)

(iii)

Plate out cells at 0.5 x 106/100 mm tissure culture plate, in 10 ml DMEM, NOT RPMI. Precipitate the DNA. 5 fig is usually adequate though this may be varied. Remember to use molar ratios of DNA, i.e. take into account the size of the vector, Precipitation. Make up volume of DNA (all the vectors to be used at a particular ratio should be precipitated together in the same tube) to 90 u,l with sterile TE, add 10 jil of 3 M sodium acetate pH 5.6 followed by 250 |o,l of ethanol. Freeze at -20°C overnight or for a minimum of 2 h.

2. Day 2 (i) (ii) (iii) (iv) (v) (vi) (vii) (viii) (ix) (x)

(xi)

Aspirate media from tissue culture dishes and add 10 ml of warm DMEM (not RPMI). Spin DNA for 15 min in a cold microfuge, remove the ethanol, and allow pellets to air dry in tissue culture hood. Resuspend each DNA pellet in 250 uJ of sterile TE. Dissolve well, Add 250 uJ of 0.5 M calcium chloride. Put 500 nl of 2x HBS into a sterile 35 mm non-tissue culture grade dish. Add the DNA/calcium chloride mix in a dropwise manner whilst gently swirling the dish. Leave for 30 min at room temperature to allow precipitate to form. Check under a microscope that the precipitate is fine; if clumps have formed break them up with a pastette. Use a sterile pastette to pick up the DNA precipitate; scrape the surface of the dish as DNA is sticky. Add each DNA precipitate to the appropriate plate of cells in a dropwise manner to cover as much of the surface area as possible. Leave to stand for 2 min before gently swirling the plate and returning to the incubator (37°C, 5-7% C02).

169

Liz Lightstone et al. Protocol 1. Continued (xii) Always do a mock transfection calcium phosphate precipitation, i.e. with no added DNA, to which selection will later be added. This is the control that gives information on the rate of dying in the absence of the transfected resistance gene. 3. Day 3: Hypertonic shock (i)

4.

5. 6. 7. 8.

Aspirate off the medium and add 10 ml 20% sterile glycerol in DMEM culture medium for 30 sec; do one plate at a time. (ii) Remove glycerol and wash once with fresh media before putting back into culture. Day 4 (i.e. 48 h after transfection): aspirate off media and add selection agent, for 24 h at 5x (providing not using MXH) then reduce to 'normal' concentration. Change the medium twice per week for first couple of weeks and thereafter once per week. The untransfected control plate will give an indication as to how quickly cells not expressing the resistance gene will die in selection. Within a week, small discreet colonies of transfected, drug-resistant cells should be growing up. When sufficient numbers are available, analyse the cells for expression of the transfected genes.

a All buffers need to be autoclaved so prepare in advance.

Protocol 2. Transfection of non-adherent cells using electroporation Equipment and reagents • Sheared salmon sperm DNA (carrier DNA) (shear using sonicator or pulling repeatedly through an 18 gauge needle) at 10 mg ml-1 • Electroporation cuvettes: sterile, 0.4 cm width

• HeBs buffer: 20 mM Hepes pH 7.05, 4.76 g/litre; 137 mM NaCI, 8.006 g/litre; 5 mM KCI, 0.37 g/litre; 0.7 mM Na2HP04, 0.11 g/litre; 6 mM dextrose 0.108 g/litre • Electroporator

Method

1. Day 1 (i) (ii)

Split cells with aim to be growing exponentially on day of electroporation. Precipitate plasmid DNA (25 - 100 p,g/sample) as described in Protocol 1, Remember to also precipitate the salmon sperm (ss) DNA (200 u,g per sample)

170

11: The molecular basis of alloreactivity 2. Day 2 For each sample, use 5-10 x 106 cells. Wash once in media/ 10%FCS and once in HeBs and resuspend so that each sample is in 0.5 ml of HeBs. (ii) Spin the precipitated DNA and resuspend in sterile water or TE at 1 M-9(jJ~1 for the plasmid DNA and 200 tigml"1 for the ssDNA. The latter is often difficult to dissolve and may need heating to 65°C. (iii) Add the sheared salmon sperm DNA to the cells, (iv) Split the cells into 0.5 ml aliquots and add each to numbered sterile electroporation cuvettes. (v) Add plasmid DNA (1 (xg^r1) to appropriate cuvettes and mix by gently tapping the sides of the cuvette. Put cover on cuvette prior to removing from tissue culture hood. (vi) Remember to include a sham transfection as a control. (vii) Electroporate. In general more gentle settings are preferable and aim for a short time constant, say 3-6 msec. (viii) Typical settings: resistance = infinite; capacitance = 250 p.f; voltage = 250 V. (ix) Place the cuvette in the electroporator and pulse the cells. (x) Leave 10 min at room temperature. (xi) Transfer electroporated cells into 10 ml of medium in a T75 flask. (i)

3. Day 3: Add selection agent.

4.3.5 Maintenance of newly transfected cells Monitor controls (no plasmid DNA transfected, grown in selection agent) for death. If transfection is successful, by the time most of the controls have died, discrete colonies of adherent cells or clumps of healthy B-LCL in suspension should be visible in the transfected cultures. Cells should be maintained in selection and split when appropriately dense. Primary cultures should be stained for expression of the transfected genes using appropriate anti-MHC antibodies. If cultures are positive, then aliquots should be frozen down for safekeeping at the earliest opportunity. Protocol 3. Staining of transfected cells for analysis on a fluorescence activated cell sorter (FACS) 1. 105 cells per point washed and resuspended in optimal (saturating) concentration of mAb supernatant in V bottomed 96-well plate in 100 til. 2. Leave for 30 min on ice.

171

Liz Lightstone et al. Protocol 3.

Continued

3. Wash twice in PBS 1% BSA; vortex plate after throwing off supernatant and prior to adding any further wash buffer, between each wash to resuspend cells. 4. Resuspend in 100 |J of saturating second layer antibody e.g. sheep anti-mouse FITC. 5. Leave for 30 min on ice. 6. Wash twice in PBS 1% BSA, vortexing between each wash to resuspend cells 7. Resuspend in 1% paraformaldehyde, mix well and run on FACS. 8. Control stainings: (i) an irrelevant antibody of the same isotype and subclass at the same concentration as the relevant antibody; (ii) a known positive e.g. a B-LCL expressing the appropriate MHC; (iii) an untransfected parent cell.

Protocol 4. Cryopreservation of transfected cells 1. In general, a T25 of nearly confluent adherent cells or optimal density (< 106mr1) B-LCL can be frozen in 1 freezing vial. 2. Wash cells which should be in exponential phase of growth, resuspend in 0.5 ml of freezing mix (90% FCS, 10% DMSO), put in freezing vial and freeze slowly overnight in a polystyrene container in -70°C freezer. The following day cells can be transferred to liquid nitrogen.

4.3.6 Enriching for positive transfectants using FACS sorting or bead sorting Occasionally, primary transfectants are largely positive for expression of the transfected genes. More often, particularly when two genes are transfected, only a proportion express the proteins at a significant level. As soon as a positive population is identified, some of the culture should be frozen so that primary transfectant stocks are available. The positive cells should be enriched for as early as cell numbers permit, otherwise the cells expressing perhaps only a single gene, which may well grow in selection, may overgrow the double positives. When transfecting genes encoding MHC molecules which are expressed at the cell surface it is simple to enrich primary cultures by sorting for positive cells using fluorescent staining and flow cytometric cell sorting or antibody staining combined with anti-antibody coated magnetic beads. Stain the cells using sterile reagents (filter dilutions of antibodies) and keep 172

11: The molecular basis of alloreactivity the cells on ice. Scale up the amount of antibody appropriately, compared to staining as described in Protocol 3, e.g. if sorting 107 cells, stain in 1 ml of antibody supernatant. At the end of staining resuspend in a small volume (500-750 (Jil) in a sterile FACS tube. Since contamination is a major risk in sorting it is worth adding gentamicin to the cells (10 ^gml"1) and to the medium the sorted cells are collected into. Controls for sorting (far fewer cells needed) are (i) second layer only and (ii) an un-transfected parent cell line. The proportion of cells to sort depends on the end requirements. In general the brightest population (e.g. the top 10%) are chosen but sometimes duller populations are required in order to get different APC expressing comparable levels of MHC molecules. After sorting, wash the cells and plate out in a 24-well or 12-well plate depending on number. Add selection agent. Check for viability and exclude infection. Grow as usual and restain when numbers permit. If the sort is successful, the first priority again is to establish frozen stocks. If a FACS is not available, an alternative strategy is to incubate the transfectants with unlabelled antibody, wash away free antibody, and then incubate with magnetic beads e.g. Dynal, coated with the appropriate second layer. After 30 min incubation on ice, the cells are put into a magnetic field and the bead-bound cells will be retained whereas the negative cells can be aspirated off. After a couple of such washes to remove all non- bead-bound cells, the magnet is withdrawn and the bead-bound fraction put into culture. The beads are non-toxic, providing they are adequately pre-washed, and will detach in culture. 4.3.7 Cloning of transfected cells Ideally, transfectants should be cloned so that stable populations can be assessed (Protocol 5). The advantage of great stability of clonal populations is that they can often grow without drug selection. Clone in the presence of irradiated feeder cells. Protocol 5. Cloning of transfected cells Method 1. Preparation of feeder cells (PBMC) 40 ml of heparinized blood (preservative free heparin), passed over Ficoll (Lymphoprep, Pharmacia) (i.e. density gradient centrifugation). (i)

Overlay 15 ml of blood on 10 ml Ficoll. Immediately spin at 770 g for 20 min at 20°C with slow break and acceleration.

(ii) Remove PBMC from interface using a pasteur pipette, wash with medium for a further 20 min at 20°C, spin at 770 g, and wash a second time for 10 min at 20°C at 400 g.

173

Liz Lightstone et al. Protocol 5. Continued

2.

3. 4. 5.

6.

(iii) Irradiate at 3000 gy. Wash cells. Count cells. Resuspend at 5 x 106mr1 and plate out at 100 ^I/well (5 x 105/well) in 96-well flatbottomed culture plates. Transfectants should be cloned by limiting dilution at 3,1 and 0.3 cells per well, using a whole plate for each cell concentration. (i) Serially dilute cells in medium containing selection agent so that the appropriate numbers of cells are added in 100 |o.l. Distribute in wells containing feeders. Score plates at 10-14 days looking for single colonies indicating that cells have grown from one cell per well. Feed the plates with fresh medium containing selection agent when necessary. When the positive colonies are confluent they can be split into 24-well plates and subsequently into 6-well plates at which point there should be enough cells to take an aliquot for staining. Clones expressing the genes at the desired level can then be grown up, stocks frozen and used in functional assays.

If the primary population contains sufficient positively staining cells (> 20%), it is possible to opt to clone immediately with the expectation that approximately 1:5 clones will be positive. This approach has the advantage that it should yield clones with a variety of levels of expression which might subsequently allow matching of different transfectants for levels of transfected MHC molecules. Alternatively, cloning can be initiated once a relatively homogenous positive population has been achieved by repeated sorting. 4.3.8 Use of transfected cell lines as allo-stimulator cells The principles of these functional assays have been addressed in Chapter 6. To summarize: (i) Adherent APC should be treated with mitomycin C. Add 50 n-gmT1 to 1 ml cells at < 5 x 106 per ml; for more cells, double the volume and double the amount of mitomycin C added. Incubate with mitomycin C for 45 min at 37 °C, wash three times prior to careful counting, dilute to the required concentration, in RPMI 1640 with 10% AB serum for human cells and not FCS, and add to plates. (ii) B-LCL can be mitomycin C treated as well but if available, irradiation is simpler (100 gy approx.). Irradiation or mitomycin C are necessary to prevent proliferation of the stimulator cells in the T cell assay. Such treatments impair presentation of intact antigen but not peptide and hence can be done prior to peptide binding but must be done after incubation with intact antigen. 174

11: The molecular basis of alloreactivity (iii) Always stain APC to compare levels of MHC expression, and adhesion molecules if relevant. (iv) Plan controls carefully, e.g. for an L-cell transfectant expressing the human HLA molecule DR1, the positive control would be a DR1 expressing B-LCL or PBMC from a DR1 positive individual. The negative control would be untransfected L-cells or L-cells expressing an irrelevant DR allele. If the specificity of the T cell being assayed is in question then an anti class I or class II antibody can be added in at a variety of concentrations. Always include controls of APC without T cells, T cells alone and T cells + APC if a peptide is being assayed. If allo-recognition of the transfectants is not predicted but the MHC allele can present a known peptide, then it is possible to assess its APC capacity by pulsing with peptide at a range of concentrations (e.g. 30, 10, 3 and 1 jjugmT1) for 4 h or overnight, washing the cells and then adding antigen-specific T cells. (v) Titrate the APC e.g. 105, 3 X 104 and 104 per well. (vi) Assay for the purity of T cells. Pure T cells, whether derived from peripheral blood or clones, cannot support a PHA response in the absence of added APC. When assessing a transfectant for antigen presentation two aspects have to be considered: the allo- (or antigen) stimulatory capacity and the co-stimulatory capacity of the transfectant. If truly pure T cells are used the latter can be assessed. Purity of T cells can be tested by having a control plate with triplicates of T cells alone, T cells + optimal concentration of PHA (in humans, 2 (igml"1), T cells + APC e.g. adherent cells from purification of PBMC (see Chapter 6) + PHA. If they are really pure, only the T cells in the latter triplicate will proliferate. (vii) Allo-reactivity against class I can be tested using cytotoxicity assays which have the advantage of being much more sparing of the stimulators but the disadvantage of being greedy for T cells. It is worth remembering that many CD4+class II-restricted T cell clones are also cytotoxic and can be assayed by cytotoxicity as well as proliferation. (viii) Activated human T cells express class II at relatively high levels so when testing peptide responses it is worthwhile pre-pulsing the APC either overnight or for at least 4 h and then washing the cells, to avoid both inappropriate capture of peptide by the T cells and also T:T presentation which may inhibit antigen-specific proliferation (20). (ix) Exclude mycoplasma infection of all immortalized cell lines.

5. Measurement of frequencies of alloreactive cells Limiting dilution analysis (LDA) is a technique which allows measurement of the frequencies of cells involved in immune responses. Depending on 175

Liz Lightstone et al. the particular experimental protocol used, frequencies of different types of effector cells may be measured, such as alloreactive helper T cells (Th) or cytotoxic T cell precursors (CTLp), antigen-specific Th or cytotoxic T cells (Tc), B cells or natural killer (NK) cells. The statistical theory on which the technique of LDA is based is known as the Poisson distribution. If the cell of interest is randomly distributed amongst many different replicate culture wells, and if the presence of a single cell of this type results in a positive culture (so called 'single-hit' kinetics), then it is possible to calculate the frequencies of this particular type of cell within the total responder cell population, using the 'zero term' of the Poisson distribution.

5.1 Theoretical basis of LDA Limiting dilution assays are quantal dose/response assays that detect an 'all or nothing' (positive or negative) response in each individual culture. Experiments are designed so that stimulator cells and any other components necessary for the generation of a positive culture, such as feeder cells or interleukin-2 (IL-2), are present in excess. Limiting numbers of responder cells are added to the culture wells. If assay wells receive one or more specific responder cells then a positive culture will result. However, since the presence of one or more specific responder cells generates a positive result, it is not possible to infer from the presence of a positive well how many specific responder cells were actually present in that well. In contrast, the presence of a negative well indicates that no specific responder cells were present in the particular well. According to the zero-term of the Poisson distribution it can be shown that the proportion of negative wells at each sample size of responder cells is linearly related to the frequency of responder cells: where, Pneg is the fraction of negative wells, / is the frequency of responder cells, and X is the sample size of responder cells/well. When fX is one, i.e. there is on average only one responder cell/culture well, Therefore, cell frequencies may be interpolated from a graph where the fraction of negative wells (on a log scale on the ordinate/'y' axis) is plotted against the number of responder cells/well (on a linear scale on the abscissa/'x' axis). At the point where the fraction of negative wells is 0.37, that sample size 'Z', will contain one responder cell, i.e. frequency = 1 in 'Z' (see Figure 3). Derivation of cell frequencies using the zero-term of the Poisson distribution is valid only when 'single-hit' kinetics apply to the experimental system, i.e. a positive culture depends only on the presence of a single responder cell, and all other components necessary for the generation of a positive response are present in excess. In some experimental systems, more complicated kinetics, e.g. 'multi hit' or multitarget' may apply (reviewed in ref. 21). It is therefore 176

11: The molecular basis of alloreactivity

Figure 3. IDA graphs. The fraction of negative wells (log scale) is plotted on the ordinate/ V axis, against the number of responder cells/well on the abscissa/'x' axis. Actual experimental data are plotted as asterisks. The line of best fit, generated using a 'maximum likelihood' statistical programme, is shown as a solid line. At the point where the fraction of negative wells is 0.37, that sample size (17021), contains one responder cell, i.e. frequency = 1 in 17021. 95% confidence limits are indicated by dotted lines. From the x2 value and the number of degrees of freedom (the number of responder dilutions minus one), p values may be derived. A p value greater than 0.05 is consistent with single-hit kinetics.

important that the experimental protocol is designed to conform to a singlehit model of kinetics.

5.2 Outline of LDA In practice, all LDA protocols involve plating out many replicate wells of serially diluted responder cells. A minimum number of 24 replicate cells of each dilution of responder cells is usual. A constant, excess number of stimulator cells is added and, after an incubation stage, some effector cell function is measured. The nature of this effector function will depend on the type of responder cell being assayed. For example, IL-2 production after stimulation by allogeneic cells may be used to measure frequencies of alloreactive Th cells, whilst cytolysis of allogeneic target cells may be used to measure frequencies of alloreactive Tc cells. It is important to appreciate that in all LDA assays the frequencies measured will depend on the exact experimental protocol used, so that valid comparisons of frequencies may be made only if the same protocol is used throughout. Standardization of LDA protocols should 177

Liz Lightstone et al. allow comparison of frequency data generated by different investigators and one of the aims of the XII International Histocompatibility Workshop is standardization of an LDA assay for measurement of alloreactive CTLp in bone marrow transplant donor and recipient pairs. The following protocols describe LDA for measurement of alloreactive Th and CTLp frequencies. The numbers of responder and stimulator cells used, and the incubation times of the assays have been designed to allow measurement of frequencies of alloreactive cells found between bone marrow donor and recipient individuals. Throughout the protocols, RPMI 1640 medium, buffered with sodium bicarbonate (0.24% final concentration) and supplemented with 2 mM Lglutamine, 50 lUml"1 penicillin and 50 (Jigml"1 streptomycin is used for washing cells. For all incubation stages of the assays, this medium is supplemented with 10% AB serum. Incubations are performed at 37 °C, in a humidified incubator with 6% CO2.

5.3 Measurement of alloreactive Th cells The read-out for this assay is IL-2 production by alloreactive Th cells, which is measured by proliferation of an IL-2-dependent 'indicator' cell line 'CTLL-2'.3 X 106 responder PBMC and 1.1 X 107 stimulator PBMC are required for the assay. Protocol 6. Measurement of alloreactive Th cells A. Responder cells 1. Prepare peripheral blood mononuclear cells (PBMC) from the responder individual by density gradient centrifugation. 2. Make seven serial dilutions of these responder cells. 3. Plate our 24 wells of each dilution, in 50 \L\, in U-bottom 96-we 11 plates, with responder cell numbers of 5, 4, 3, 2,1, 0.5, 0.25X104 cells/well. B. Stimulator cells 1. Prepare PBMC from the stimulator individual and -y-irradiate (35 Gy). 2. Dilute to a concentration of 5 X 105ml"1 and add 5 x 10* cells, in 100 (jil, to each assay well. 3. Also plate our 24 wells of the stimulator PBMC with 50 |o.l of medium only, i.e. no responder cells added. These are the negative control wells. 4. Incubate the assay for 64 h.

178

11: The molecular basis of alloreactivity C. Growth and preparation of indicator CTLL-2 Line The 'indicator' cell line is a murine cytotoxic T cell line 'CTLL-2' (European Collection of Animal Cell Cultures, Salisbury UK) which proliferates in the presence of human IL-2 (22). 7X105 CTLL-2 cells are required for the assay. 1. The CTLL-2 line may be cultured in standard medium, supplemented with 5% human recombinant IL-2 (rlL-2)(10 Umr1). 2. The cells are sub-cultured every 2-3 days. 3. Prior to use in an LDA, the CTLL-2 are washed twice and cultured overnight in the usual medium, but without added rlL-2. (If the CTLL-2 are used directly from culture in medium containing rlL-2, spontaneous background proliferation is very high). D. Measurement of IL-2 production 1. After 64 h incubation, -/-irradiate the culture plates (25 Gy) to prevent further proliferation of the responder PBMC. 2. Wash the CTLL-2, which have been cultured overnight without rlL-2, and dilute to a concentration of 1.2 x 10Bmr1. (A small percentage (< 5%) of the CTLL-2 will be dead, and there should be very few dividing cells). 3. Add 3 x 103 indicator CTLL-2, in 25 n-l, to each well. 4. After 8 h label the plates with 1 ^Ci/well for tritiated thymidine (3H-TdR). 5. Assess proliferation of the CTLL-2 by 3H-TdR incorporation after a further 18 h incubation. (If the facility of plate -y-irradiation is not available, supernatant may be harvested from the individual assay wells, and CTLL-2 proliferation in this supernatant measured). Negative control wells contain stimulator PBMC and indicator CTLL-2 cells, but no responder cells. Assay wells are considered positive if proliferation exceeds the average plus 3 standard deviations (SD) of these control wells. In all experiments, measure CTLL-2 proliferation to a range of rlL-2 concentrations (e.g. serial doubling dilutions from 2-0.003 U ml"1 rlL-2) in parallel with the LDA to ensure that the CTLL-2 are responding to IL-2 in a dose dependent manner.

Protocol 7. Measurement of alloreactive CTLp frequencies The read-out for this assay is 51chromium (51Cr) release from radiolabelled target cells. 4.5 x 106 responder PBMC, 1.3 x 107 stimulator PBMC, and 3.5 x 106 third-party PBMC are required for the assay.

179

Liz Lightstone et al. Protocol 7. Continued Method A. Preparation of responder cells 1. Prepare eight serial dilutions of PBMC from the responder individual. 2. Plate out 24 wells of each dilution, in 50 (J, in 96-well U-bottom plates, with responder cell numbers of 5, 4, 3, 2, 1, 0.5, 0.25 and 0.125 x 104/well. B. Preparation of stimulator cells 1. Prepare stimulator PBMC at a concentration of 5 x 105mr1 and "/-irradiate (35 Gy). 2. Add 5 x 104 cells, in 100 jil, to each of the assay wells. 3. Also plate out 24 wells of stimulator PBMC plus 50 |xl of medium only, as the negative control wells. C. Positive ('third-party') controls 1. To establish that the responder cells are able to mount an alloresponse and that the stimulator cells are able to provide an allostimulus, positive controls are necessary. In these controls PBMC from an HLA-mismatched 'third-party' individual are used. For the 'third-party stimulator' positive control, (which establishes that the responder cells are responsive to an allogeneic stimulus), plate out 24 wells with 5 x 104, y-irradiated (35 Gy), third-party PBMC in 100 |xl. 2. Add 5 x 104 responder cells in 50 |j,l. 3. In addition, it is necessary to plate out 12 negative control wells, containing 5 x 104, -/-irradiated third-party stimulator cells only, which will provide the background control for the 'third-party stimulator' control itself. 4. For the 'third-party responder' positive control, (which establishes that the stimulator cells are providing an allostimulus), plate out 24 wells with 5 x 10" third-party PBMC and add 5 X 104, -/-irradiated stimulator cells in 100 \L\. (Additional negative control wells for this 'third-party responder' control are not required, since the negative control wells for the complete assay will provide the relevant background control). 5. Incubate the assay for 10 days, feeding the culture wells on days 3 and 6 with 25 jil of medium containing 35 Urnl"1 of rlL-2 (i.e. corresponding to a final additional concentration of rlL-2/well of 5 U mr1).

180

11: The molecular basis of alloreactivity D. Preparation of targets for the CTLp assay 3 x 106 stimulator blasts and 0.7 x 106 third-party blasts are required for the assay. 1. On day 5 of the assay, prepare 5 x 106 stimulator PBMC and 1.5 x 106 third-party PBMC, each at a concentration of 1 x lO'ml"1. 2. Plate out in 24-we 11 culture plates at 1 x 106cells/well. 3. Add phytohaemagglutinin (PHA) at a concentration of 2 ^gmT1 to each well and incubate for 5 days. 4. Split the cultures every 2-3 days as necessary, feeding with fresh medium containing rlL-2 (20 U mr1) as required. The PHA-blasts produced by this protocol will act as the targets for the chromium release assay. E. Chromium release assay The chromium release assay is performed under non-sterile conditions. 1. On day 10 of the assay, harvest the stimulator and third-party PHA blasts, wash three times and count. 2. Pellet the blasts in conical tubes, add 100 n,l of medium and label with 51Cr (200 |iCi 51Cr/107 PHA blasts) by incubating at 37°C for 1 h. (Resuspend the blasts gently every 15 min to ensure even labelling.) 3. Wash the labelled blasts/targets three times and dilute to a concentration of 1 x 105 targets/ml. 4. Remove 100 (xl from each well of the CTLp assay plates, taking care not to disturb the cell pellet in the bottom of each well. 5. Shake the plates on a microplate shaker for 1-2 min to disperse the cell pellet. 6. Add 1 x 104 stimulator targets, in 100 (jj to all the assay wells, the negative control wells and the 'third-party responded controls wells. (Resuspend the stimulator target cell suspension frequently during this step). 7. Also, add 1 x 104 third-party targets, in 100 jil, to the 'third-party stimulator' control wells and to the 12 wells containing third-party stimulator cells only, which will provide the background control for the 'third-party stimulator' control itself. 8. In addition, plate out controls for: • Spontaneous release—10 wells with stimulator targets plus 100 n-l of medium and 10 wells with third-party targets plus 100 ^l of medium. • wells with stimulator targets plus 1% Triton X detergent and 10 wells with third-party targets plus 1% Triton X

181

Liz Lightstone et al. Protocol 7. Continued detergent. (The detergent lyses the targets causing release of all the 51Cr radiolabel). 9. Incubate with 51Cr target in plates for 4 h. 10. Set up LP2 tubes in 96-well plates so that the pattern of these tubes corresponds to the pattern of wells on the assay plates, and label these plates accordingly. 11. Pellet the cells in the assay plates by centrifugation in a plate centrifuge, at 1000 r.p.m. for 5 min. 12. Transfer 100 ul of supernatant from all the assay wells into the corresponding LP2 tubes, taking care not to disturb the cell pellet. 13. Seal the LP2 tubes with molten wax and count the amount of 51Cr released into the supernatant using a y-spectrometer. 14. Check 51Cr incorporation into the stimulator and third-party targets by comparing spontaneous and total release. Spontaneous release should not be greater than 15-20% of the total release. • Negative control wells contain stimulator PBMC and 51Cr stimulator targets, but no responder cells. Assay wells are considered positive if 51Cr release exceeds the average plus 3 SD of these control wells. • Check that the positive control wells (i.e. 'third-party stimulator' and 'third-party responder' controls) are, in fact, positive confirming that the responder cells and stimulator cells have functioned as required.

5.4 Technical notes There is a theoretical risk in the CTLp assay (Protocol 7) that the addition of IL-2 may activate NK cells, leading to non-specific lysis of targets. To investigate this possibility, additional controls were included in preliminary CTLp experiments: lysis of stimulator targets by responder PBMC which had been cultured without any stimulator PBMC, or by responder PBMC which had been cultured with third-party stimulator PBMC, was measured. In all these experiments non-specific (NK-mediated) lysis was not detected, therefore these specificity controls are no longer routinely included in the experimental protocol. • PBMC from responder and stimulator individuals may be cryo-preserved and stored at -70°C, prior to being used in these LDA assays. • Epstein-Barr Virus (EBV)-transformed B-lymphoblastoid cell lines (BLCL) may be used as stimulator cells. All B-LCL express high levels of MHC class II molecules, unlike PMBC which are only approximately 35% 182

11: The molecular basis of alloreactivity MHC class II positive, therefore lower numbers of B-LCL stimulator cells are required. (Most adults are EBV-immune and there is, therefore, a theoretical risk that use of B-LCL stimulator cells will result in detection of EBV-specific cells, rather than allo-specific cells, via processing and presentation of EBV antigens by responder cells. In practice, this does not seem to be a significant problem since the numbers of EBV-specific, self-MHC-restricted cells have been found, in general, to be low (23) or undetectable (24) • The incubation tunes of the assays have been designed to allow measurement of frequencies of alloreactive cells found between bone marrow donor and recipient individuals. Measurement of alloreactive Th and CTLp frequencies in the donor (responder) versus recipient (stimulator) direction is used to help predict occurrence of graft versus host disease (GVHD) in the bone marrow recipient. Modifications of these protocols may be made, depending on the particular requirements of the investigator. For example, in the case of fully allogeneic responder and stimulator individuals, it is appropriate to use fewer responder cells in each particular responder cell dilution; a high frequency of allospecific cells would be predicted and therefore most assay wells would be positive if large numbers of responder cells were added. It has been shown the maximum statistical information is provided when the numbers of responder cells generate a fraction of negative wells between 0.1 and 0.37 (25).

5.5 Statistical analysis of LDA Once again it should be stressed that the experimental protocol used in the (LDA) must be designed to conform to a single-hit model of kinetics, since calculation of cell frequencies from the zero-term of the Poisson distribution is valid only if single-hit kinetics apply. Mathematical techniques suggest that the 'maximum likelihood' method, based on the methods of Finney (26), should be used to evaluate the experimental data. In addition to cell frequencies, 95% confidence limits of the frequencies and x2 estimates of probability are calculated. Frequencies may be regarded as different if the 95% confidence limits, (approximately 2 SD), do not overlap. From the x2 values and degrees of freedom, (the number of responder dilutions minus one), probability estimates p may be derived from statistical tables. p values > 0.05 are consistent with single-hit kinetics. It should be noted that whenever data from limiting dilution assays is presented, then 95% confidence limits for the cell frequencies and p values should also be shown. It is then possible to assess whether frequencies are significantly different and whether the assay has conformed to a single hit kinetics model. Frequency data derived from assays which do not conform to single hit kinetics must be viewed with caution. 183

Liz Lightstone et al. 6. Applications Using such transfection protocols we and others have established many MHC expressing mouse L-cells and human fibroblasts as well as reconstituting class I and class II mutant B-LCL. In the following sections, we describe how such tools have been used to analyse various aspects of allo-recognition.

6.1 Generation of allo-specific T cell clones using mouse L-cells expressing human class II molecules The advantage of using mouse fibroblast transfectants expressing human HLA molecules to generate allo-specific T cell clones is that the entire alloresponse will be focused on the particular molecule encoded by the transfected genes and the peptide it presents rather than being diluted by the presence of other MHC molecules as would be normally present on a human antigen presenting cell such as an EBV-transformed B cell line. This has been done successfully in our group and opens the way for generating clones against mixed isotype transfected MHC molecules as well as against hybrid molecules encoded by recombinant MHC genes. Interestingly, most of the clones generated by this method also recognize the same MHC molecules on human cell lines and PBMC, suggesting that the peptides presented by mouse cells are not so different from those presented by human cells (27).

6.2 Detecting precursor frequencies of alloreactive T cells—a tool in transplantation In the context of alloreactivity, LDA has been used to show that frequencies of alloreactive cells in MHC-incompatible responder/stimulator combinations are generally high (28, 29, 30). For example, in the mouse it has been shown that 1-4% of the total peripheral T cell pool can give rise to alloreactive CTLp (31). Variation in alloreactive cell frequencies occur between individuals and within individuals against different stimulator alloantigens (32, 33, 34, 35). Indeed, variations in alloreactive CTLp frequencies against the same allogeneic stimulator cell have been reported for identical twins (36). In general, as the degree of MHC disparity between responder and stimulator cells increases, so too does the frequency of alloreactive cells (37, 38). In the absence of allogeneic stimulation, for example by transplantation or transfusion, frequencies of alloreactive cells within an individual remain approximately constant over time (33, 36, 39, 40). LDAs measuring alloreactive cell frequencies are used clinically to help predict occurrence of acute GVHD in BMT patients (41, 42, 43) and have been used to predict episodes of rejection in cardiac (44) and renal (45) allograft patients. 184

11: The molecular basis of alloreactivity

6.3 The nature of the allo-ligand 6.3.1. The contribution of direct contact between a-helical residues and TCR The a chains of the DR class II molecules are non-polymorphic, so the new nomenclature for different DR alleles describes the p chain gene as follows: DR4DwlO is now known as DRB1:0402 and DR4Dwl4 as DRB1:0404. These two alleles differ in only three residues: 70, 67 and 71, in the B1 domain of the p chain, with residues 67 and 71 facing into the peptide-binding groove and residue 70 facing up towards the TCR. T cell clones from a DRB1:0402 individual recognize DRB1:0404 as an allo-molecule. Using sitedirected mutagenesis to mutate the single TCR contacting residue back to that found in the responder, five of six T cell clones no longer recognized DRB1:0404 as an allo-stimulator (46). Hence, such studies confirm that for some T cell clones the allo-response is directed against the allo-MHC molecule itself. Similar conclusions have been reached using the alternative strategy of hemi-exon or exon shuffling of MHC molecules to generate hybrid molecules with the a-helices and B-pleated floor or B1 and B2 domains derived from different alleles or species. Substituting the sequence encoding either a helix, or indeed the floor, of e.g. the B1 domain of the DRB gene with that from the mouse H-2E class II molecule led to abrogation of either MHC or peptide dependent allo-responses respectively. Confirmation of the role of direct recognition of the a-helical region has been obtained by showing that peptides corresponding to the relevant region of the a-helix can block allo-recognition (46). 6.3.2 Site-directed mutagenesis This allows mutation of a single residue and the principle is outlined in Figure 4. We will not describe a detailed protocol as kits are available which are entirely self-explanatory. We routinely use one from Amersham International (UK). The essential first steps are cloning of the template DNA into M13 and the subsequent extraction of single stranded DNA. Template DNA needs to be of high quality and free from contaminating fragments as these can act as primers during mutagenesis. Hence, template DNA needs to be sequenced prior to use. An oligonucleotide containing the mutagenized residue needs to be synthesized and available. The mutated single stranded DNA has to be re-polymerized and transformed back into M13 for subsequent sequence analysis. Once the correct sequence is achieved the mutated DNA can be cloned into a eukaryotic expression vector and transfected as described earlier. 6.3.3 Exon or hemi-exon shuffling An important method for determining the relative contribution of the ahelices and B-pleated floor has been that of hemi-exon shuffling. We have

185

Liz Lightstone et al.

Figure 4. Schematic representation of site-directed mutagenesis. The thiolated second strand containing the mutagenized site is protected from nicking and digestion and hence is available as template under repolymerizing conditions.

shuffled hemi-exons between mouse MHC class II molecules (H2-E being the homologue of DR) and human class II molecules. Two main approaches can be used: cutting and ligating, and splicing by overlap extension. i. Cutting and ligating to create shuffles The first involves taking advantage of naturally occurring restriction endonuclease sites or introducing matching cutting sites using site-directed mutagenesis, in the exons to be shuffled. The exons of interest are cut and ligated. The construct is then sub-cloned into M13 phage vector for checking the final product by sequencing and is subsequently cloned into a eukaryotic expression vector for transfection together with the gene for the partner MHC chain. This method is well described in earlier papers (47) and uses basic DNA cloning techniques to be found in (48). ii. Splicing by overlap extension to create shuffles Splicing by overlap extension (SOE) (Protocol 8) is advantageous if matched 186

11: The molecular basis of alloreactivity

Figure 5. Generation of a hybrid class II B chain using splicing by overlap extension. See text for explanation.

sites are not readily introducable into the hemi-exons. SOE is a sequenceindependent method for engineering recombinant DNA which makes use of the polymerase chain reaction (PCR). Compared with conventional methods of DNA engineering which rely on the presence of specific restriction enzyme recognition sequences, SOE can be used to generate constructs with very precise recombination points, thereby joining two coding sequences in frame. The basic method of SOE is shown schematically in Figure 5. SOE is a two step process. The first step is to generate the two fragments to be joined (represented in Figure 5 by BB1 and AB2). These will be modified to contain regions of complementarity, using specially designed hybrid oligonucleotide primers. In the second step, under denaturing and re-annealing conditions, the isolated fragments act as primers for each other by virtue of the overlap regions, which can then be extended using DNA polymerase to generate the recombinant product. This method can also be used for site-directed mutagenesis. The design of the primers is essential to the success of this method. It is important to standardize the template DNA strands and the orientation. By convention the reading frame is written from left to right, with the 5' end on the left and the 3' end on the right. Where only one strand is shown it will be the top strand. The easiest way to design the primers is to write out the double stranded sequence of the recombinant DNA product required and work out the primers accordingly, as shown in the example. The point of recombination of the template is marked by ', so for gene B the base-pairs flanking the point of recombination are: 187

Liz Lightstone et al.

for gene A:

The bases underlined and in italics are those which will appear in the final recombinant product, therefore the recombinant product will be as follows:

Since DNA polymerases synthesize DNA in a 5' to 3' direction, the primer necessary to synthesize the B gene fragment will be:

such that

and the primer necessary to synthesize the A gene fragment will be: such that

In order to achieve hybridization of the two fragments generated from the first step of SOE there has to be a region of complementarity between the fragments. The optimum length of homology is between 16 and 20 bp. If possible this region should contain equivalent numbers of each nucleotide and should avoid homology with internal sequences. In the example given (Figure 5), the oligo primers a and d were designed to anneal to the plasmid vector, so that the restriction enzymes used to insert the A and B genes into the plasmid vector could be used to liberate the hybrid gene and allow for ligation into an expression vector, and into M13 bacteriophage for sequencing. It should be noted that in vitro synthesis of DNA by PCR is prone to error, especially using Taq DNA polymerase at high magnesium concentrations, and therefore the products must be completely sequenced after synthesis. The error rate is also dependent on the DNA polymerase used, and it is advisable to seek advice on the best one to use. 188

11: The molecular basis of alloreactivity Protocol 8. Splicing by overlap extension to generate shuffled exons or hemi-exons A. Step 7 1. For each fragment, pipette into a 0.5 ml polypropylene tube: • • • • • • •

Template DNA, 100-500 ng (A or B gene) Oligo primer, 1 uM (a or d) Hybrid oligo primer, 1 uM (b or c) Buffer 1x (standard PCR buffer with no Mg2+) Mg2+, 0.5-4 mM dNTPs 200 uM (Pharmacia, ultra dNTP mix) Water to 95 ul

Add 2-3 drops of light mineral oil to the tube and heat to 98°C for 5 min. Then put on ice for 5 min before adding 5 ul Tag DNA polymerase, 0.5 U ul-1 (Perkin Elmer-Cetus), through the oil. 2. Load the reaction mixes into a DNA thermal cycler to cycle 25 times: denaturation at 94°C for 1 min; annealing at 50°C for 1 min; extension at72°C for 1.5 min. 3. Purify the products of this first step by running them on a 1% agarose gel in TBE (Tris-borate, EDTA (ethylenediaminetetraacetic acid) 89 mM Tris, 89 mM orthoboric acid, 2 mM EDTA) then onto DEAE paper and eluting with 2.5 M NaCI in TE (10 mM Tris-HCI, pH7.2, 1 mM EDTA) at 72°C for 2 h before precipitation with 10 ug yeast tRNA as a carrier and ethanol. Step 2 1. Using the two products from Step 1 as templates: • • • • • • • • • •

Template 1, 20-25% of product from Step 1 (AB2) Template 2, 20-25% of product from Step 1 (BB1) Oligo primer, 1uM(a) Oligo primer, 1uM(d) Buffer 1 x Mg2+, 0.5-4 uM dNTPs 200 mM Water to 95 ul Mineral oil Taq DNA polymerase, 2.5 U in 5 ul

2. Repeat the 25 cycles as before, and run the product on a 1% agarose gel to quantify.

189

Liz Lightstone et al. Any dilutions of reagents used in Protocol 8 should be carried out in low EDTA TE buffer (0.1 mM EDTA), as EDTA will chelate the Mg2+ in the PCR buffer and alter the Mg2+ concentration. The annealing temperature may need to be varied according to the machine used. Generally a lower temperature will decrease the specificity of the annealing. It may also be necessary to increase the extension time in some cases. The template concentration for both steps is not critical, and a range of concentrations will work. The more template there is the fewer rounds of replication will be necessary and therefore the risk of introducing errors into the newly synthesized DNA is lowered. However, if too much template is present it can, in some cases, inhibit the reaction. Having generated appropriately engineered MHC molecules checked by nucleotide sequencing and transfected them into the cells of choice, it is necessary to check expression in a similar manner to that described in Section 4.3.5. Such constructs incidentally provide a means of mapping anti-MHC antibody epitopes. When assessing the functional capabilities of the mutated MHC molecules it is always necessary to include cells (ideally the same cell type) expressing the appropriate wild type MHC molecules as a control. Furthermore, discrepancies between different transfectants in the level of expression of the allo-MHC molecules need to be taken into account when attributing functional consequences to altered structure. Ideally, transfectants with matched levels of expression should always be used in functional assays. This approach, namely the use of transfected mutated MHC molecules, has confirmed that some allo-responses involve direct recognition of highly dissimilar foreign MHC molecules (49). 6.3.4 Definition of the contribution of peptide in allo-responses Some allo-responses are not supported by DR1 molecules expressed in L-cells whereas they can be by DR1-bearing B cells. Lack of co-stimulation can be excluded by co-transfection of molecules such as LFA3 and the ability to support antigen-specific responses. The attractive explanation of failure to stimulate an allo-response is that the allo-TCR is recognising a peptide presented by DR1 on human cells but not by DR1 on mouse cells, i.e. a species-specific peptide. This has been confirmed by restoring allostimulation when the same L-cell transfectants are pulsed with a lysate from human B-LCL. Similarly the contribution of peptide to some allo-responses has been shown using cross reactive T cell clones which are specific for 'flu' peptide 306-321 presented by DRB1:0101 but allo-reactive to DRB1:0401 on B-LCL. These clones failed to recognize DR4 on mouse L-cells but the alloresponse could be restored by addition of the HA peptide (50). This confirms not only that allo-reactivity involves recognition of peptide and MHC but that structural similarity between the a-helices of DRB1:0101 and DRB1:0401 allows recognition of antigen presented by non-self MHC. An alternative approach has been to mutate residues of the MHC either in TCR

190

11: The molecular basis of alloreactivity interacting residues in the a-helices or in the B-sheet forming the floor of the peptide binding groove. Alterations in the floor of the groove would be predicted to lead to presentation of a different array of peptides and alter allo-stimulation if peptide contributes to the allo-ligand. This approach has clearly shown both the contribution of peptide and TCR contact residues in different allo-responses (see e.g. ref. 51). 6.3.5 Definition of the peptides involved in allo-recognition A more recent approach has developed out of the technology for eluting class I and class II specific peptides. Peptides separated by HPLC from either whole cell lysates or from affinity purified MHC molecules can be used to reconstitute allo-responses of non-stimulating APC bearing the appropriate MHC molecules. Details of preparation of peptides are given in chapters 10 and 12. In general, peptides should be speed vacuumed to dryness and re-suspended in PBS if to be used in assays. However, if the volumes to be added are small, at least for cytotoxicity assays, peptides from purified class I can be kept in the acetonitrile/TFA mix in which they are eluted off the HPLC column. Protocol 9. Identifying peptides recognized by allo-reactive T cells: fractionation of whole cell lysates of class I or class II bearing cells Equipment and reagents • 0.1% TFA and 0.7% TFA each made up in dH2O* • Acetonitrile for HPLC = 2.25 ml TFA into 2.5 litres • Dounce Homogenizer with 2 pestles • Sonicator with variable amplification

Ultracentrifuge Centricon 10 filter units Speed vac Sonicating water bath Eppendorf eppendorfs (these appear to be the least 'sticky')

Method 1. Grow up 5 x 108 - 109 cells, harvest, wash 3 times in cold PBS and either use the dry cell pellet immediately, or freeze the cell pellet rapidly by rotating the tube in which it is to be frozen in a mix of dry ice and ethanol. Such 'shell' drying ensures both rapid freezing and enables rapid thawing (to be done at 37°C). Both factors are critical in minimizing protease activity prior to the addition of acid. If the freezing the pellet, do so at -70°C. 2. Resuspend the cell pellet in 3 ml 0.7% TFA. 3. Homogenize the cells in a 7 ml dounce homogenizer using the less tight fitting pestle first, followed by a tight fitting pestle. 4. Transfer lysate into a universal tube and keep on ice for ultrasonification. Keep the homogenizer on ice as it will be used again.

191

Liz Lightstone et al. Protocol 9. Continued 5. Sonicate the lysate. We use a sonicator set at 50% maximum amplitude and sonicate for 10 bursts of 5 sec each. It is vital to ensure that the tip is immersed in the lysate, the universal is immersed in ice and that the tube does not move around. 6. Vortex tube for 3 min. 7. Transfer sample into 2 Eppendorf microfuge tubes 8. Spin in microfuge at 15000 r.p.m., 4°C for 30 min. Use a refrigerated table top microfuge or one that is kept in a cold room or cabinet. 9. Collect supernatant and transfer into a universal tube. 10. Resuspend pellet in 2 ml 0.1% TFA and transfer into dounce homogenizer. 11. Homogenize the sample and transfer into the universal containing the supernatant. 12. Sonicate 5 ml sample as before. 13. Transfer sample into polyallomer tubes which fit an ultracentrifuge swing-out rotor which has been pre-cooled to 4°C for 2-3 h. 14. Load the rotor with the 5 ml sample and a balancing tube. 15. Spin at 150000 x g for 30 min at 4°C. 16. Remove the supernatant carefully. At this stage it can be frozen at -20°C or filtered immediately. 17. Add the supernatant to 2 C.10 filter units which have been pretreated with 1 ml 0.1%TFA spun for 30 min at 5000 x g. Add the supernatant to 2 pre-wetted Centricon 10 filter units. Spin at 5000 x g for 30 min at 4°C in a fixed angle rotor. The retentate should be no more than 100 ul. 18. The flow through can either be frozen at —70°C or loaded directly on to an HPLC column. a Neat TFA is highly toxic and volatile and should only ever be handled in a fume hood and gloves should be worn.

Fractionation using HPLC is described in chapters 10 and 12. The relevant fractions from the HPLC column are in a mix of TFA and acetonitrile and eluted into Eppendorf microfuge tubes. This is likely to be toxic to cells but has been used directly in some stimulation assays (52). We have tended to speed vacum the samples down overnight to dryness and then resuspend very carefully in sterile PBS. Good re suspension is critical for preservation of the peptides and indeed, if not being used straight away, it is not unreasonable to freeze the dried peptides at -70°C and resuspend on the day of use. 192

11: The molecular basis of alloreactivity

Protocol 10. Re-suspension of dried peptides Methods Keep everything on ice as far as possible; wear gloves. 1. Add 100-200 ul sterile PBS. 2. Pipette up and down a number of times.

3. Vortex for 30 sec. 4. Sonicate in a water bath for 4 min avoiding overheating by adding ice to the water. 5. Immediately use or freeze at -70°C.

HPLC fractions from whole cell lysates or from peptides eluted from purified class I, have been used in CTL assays to determine the contribution of peptides to allo-recognition. The advantage of such an approach is that T cells can detect picomolar concentrations of peptide and identify relevant fractions which may be barely visible on an HPLC profile. These studies suggest that some T cell clones are peptide dependent, some are peptide independent, and some appear to recognize preferentially an MHC-peptide complex which is dependent upon conformation which cannot be restored by pulsing with peptide (53). Such studies have shown that it is a much harder task actually to determine the sequence of allo-peptides since not only are large amounts of sample required to get sequence data (e.g. starting with cells from 150 spleens or 1010 B-LCL!) but the stimulating samples need to be repeatedly fractionated as a single primary HPLC peak may contain up to 150 peptides. The advantage of this approach however is that the T cell is undoubtedly the most sensitive read-out currently available and T cell stimulation by subsequent fractions allows identification of a much purer fraction of physiologically processed peptides.

6.4 The indirect pathway of allo-recognition The previous sections have all dealt with direct allo-recognition by T cells. It is also apparent that foreign MHC molecules on transplanted tissue are taken up by host APC and presented by host MHC in a 'conventional' antigen presentation pathway (54). Recognition of foreign MHC peptides is thought to be important in chronic rejection of allo-grafts. The techniques already described can be used to further understand this pathway. For instance, the finding that some alloreactive-T cell clones can only be stimulated by a particular class II product if a particular class I molecule is also present suggests that the relevant peptide recognized is derived from the class I molecule (55). This can be directly tested by co-transfecting the class I and class II molecules 193

Liz Lightstone et al. into mouse L-cells. Furthermore, it should be possible to elute class I peptides from affinity purified class II molecules and show that these can reconstitute the allo-response.

6.5. Determining the nature of co-stimulatory molecules required for allo-reactivity Lack of allo-stimulation may not be due to absence of the appropriate ligand but to absence of appropriate co-stimulatory molecules on transfected APC. This may be due to an absolute lack or failure to recognize xeno-molecules. APC should be stained for presence of LFA3 (ligand for CD2), ICAM1 (ligand for LFA1) and B7 (ligand for CTLA4 and CD28). Bear in mind that there are several isoforms of B7. Exon-shuffling experiments as described earlier have shown that human CD4 interacts poorly with the mouse MHC class II molecule H-2E and this can be greatly augmented by substituting the H2EB2 domain with the DR32 domain (56). Others have confirmed that the CD4:class II interaction site is on the B2 domain of class II. Similarly, species matching, by exon shuffling, of the class I a3 domain augments class I:CD8 interactions (57). Different T-cell clones are more or less dependent on co-stimulation and in general, super-transfection of mouse L-cells with LFA3 optimizes allo- and antigen-specific stimulation. The transfection protocol is identical to that described in Section 4. B7 appears to be an absolute requirement for the induction of primary, but not secondary, allo-responses (58). It is worth noting that human T cells can be stimulated by mouse B7 which is expressed to varying levels on L-cells (59). B7 is much less stringently required for secondary responses but these are often augmented by the addition of LFA3 or ICAM1 or both (58). It is also worth noting that the presence or absence of other molecules may critically affect allo-recognition, largely by affecting the peptides which are displayed. The class II associated invariant chain, Ii, is involved both in routing of class II and loading of class II with peptides (60). Absence of Ii has been shown to both augment or abrogate allo-recognition depending on the system used (61, 62, 63). Mouse L-cells are usually negative for Ii but express it if grown densely. It is not yet clear however, whether mouse Ii can associate efficiently with human molecules. The peptide transporters TAP1 and TAP2 are essential for class I peptide loading, and deficiency of either results in impaired expression of class I and failure of allo-stimulation (64). Examples of TAP mutants include the mouse cell line RMA-S and the human B-LCL, 174.134 (65).

6.6 Investigating the allo-specific T cell repertoire The frequency of primary allo-reactive cells is similar among activated and resting T cells, and among T cells recognising DR or DQ MHC molecules 194

11: The molecular basis of alloreactivity (66, 67). The high frequency of allo-T cells makes it relatively easy to characterize and to generate T cell clones. The hypotheses stated at the beginning of this chapter also allow testable predictions about VB usage of allo-reactive T cell receptors. A computer-generated model of how the TCR interacts with an MHC molecule (68) suggests that the portions of the TCR that make direct contact with the MHC a-helices are those that are encoded by the germline variable segments. Hence, in an allo-response between two individuals whose MHC molecules differ extensively in their TCR contacting regions, biased usage of TCR variable domains may be expected. In contrast, in an allo-response between a responder and stimulator pair which share identity in the B1 domain a-helix, and hence one which presumably is dependent on peptide recognition, no biased VB usage would be expected. The most important aspect of this approach is identifying suitable responder stimulator pairs to test the hypothesis. We have done this by generating two panels of anti DR11 T cell clones, one from a DR17 and the other from a DR15 responder. The TCR contacting surfaces of DR17 and DR11 have multiple differences whilst those of DR15 and DR11 are very similar. Sequence analysis of the TCR VB genes expressed by the clones show that, as predicted, the clones from the dissimilar responder:stimulator pair have biased VB gene usage whilst those from the other panel do not (69).

7. Summary In summary we have described the approaches used by ourselves and others to show that the molecular basis of allo-reactivity is heterogeneous, involving either direct recognition of allo-MHC molecules or recognition of allo-peptides or both. Other molecules also contribute to the strength of allo-stimulation. It is now clear from such studies that both original hypotheses used to explain allo-reactivity can be supported by these data and furthermore that the phenomenon of allo-reactivity is not incompatible with the existence of positive selection.

References 1. Simonsen, M. (1967). Cold Spring Harbor Symp. Quant. Biol., 32, 517. 2. Bevan, M. (1984). Immunol. Today, 5, 128. 3. Chicz, R.M., Urban, R.G., Gorga, J.C., Vignali, D.A.A., Lane, W.S., and Strominger, J.L. (1993). J. Exp. Med., 178, 27. 4. Matzinger, P. and Bevan, M.J. (1977). Cell. Immunol., 29, 1. 5. Rudensky, A.Y., Preston-Hurlburt, P., Hong, S.-C., Barlow, A., and Janeway Jr, C.A. (1991). Nature, 353, 622. 6. von Boehmer, H., Teh, H.S., and Kisielow, P. (1989). Immunol. Today, 10, 57. 7. Bjorkman, P.J., Saper, M.A., Samraoui, B., Bennett, W.S., Strominger, J.L., and Wiley, D.C. (1987). Nature, 329, 506. 195

Liz Lightstone et al. 8. Brown, J.H., Jardetzky, T.S., Gorga, J.C., Stern, L.J., Urban, R.G., Strominger, J.L, and Wiley, D.C. (1993). Nature, 364, 33. 9. Hansburg, D., Heber, K.E., Fairwell, T., and Appella, E. (1983). J. Exp. Med., 158, 25. 10. Lechler, R.I., Lombardi, G., Batchelor, J.R., Reinsmoen, N., and Bach, F.H. (1990). Immunol. Today, 11, 83. 11. Lombardi, G., Sidhu, S., Batchelor, J.R., and Lechler, R.I. (1989). Proc. Nat. Acad. Sci. USA, 86, 4190. 12. Padovan, E., Casorati, G., Dellabona, P., Meyer, S., Brockhaus, M., and Lanzavecchia, A. (1993). Science, 262, 422. 13. Brown, J.H., Jardetzky, T., Saper, M.A., Samraoui, B., Bjorkman, P.J., and Wiley, D.C. (1988). Nature, 332, 845. 14. Lombardi, G., Sidhu, S., Lamb, J.R., Batchelor, J.R., and Lechler, R.I. (1989). J. Immunol., 142, 753. 15. Tsuji, K., Aizawa, M., and Sasazuki, T. (1992). Proc. Eleventh International Histocompatibility Workshop and Conference. 16. Koch, W., Candeias, S., Guardiola, J., Accolla, R., Benoist, C., and Mathis, D. (1988). J. Exp. Med., 167, 1781. 17. Royer-Pokora, A.B., Peterson, W.J.J., and Haseltine, W.A. (1984). Exp. Cell. Res., 151, 408. 18. Warrens, A.N., Zhang, J.Y., Sidhu, S., Watt, D.J., Lombardi, G., Sewry, C.A., and Lechler, R.I. (1994). Int. Immunol., 6, 847. 19. Van, P.A., De, P.E., and Boon, T. (1985). Somatic Cell Mol. Genet., 11, 467. 20. Sidhu, S., Deacock, S., Bal, V., Batchelor, J.R., Lombardi, G., and Lechler, R.I. (1992). J. Exp. Med., 176, 875. 21. Lefkovits, I. and Waldman, H (1979). Limiting dilution analysis of cells in the immune system. Cambridge University Press, Cambridge. 22. Gillis, S. and Smith, K.A. (1977). Nature, 268, 154. 23. Vie, H. and Miller, R.A. (1986). J. Immunol., 136, 3292. 24. Theobald, M., Hoffman, T., and Heit, W. (1989). J. Immunol.Methods, 121, 19. 25. Groth, F. d S. (1982). J. Immunol. Methods, 49, 11. 26. Finney, D. (1978). Statistical methods in biological assay. Griffin, London. 27. Warrens, A.N., Heaton, T., Sidhu, S., Lombardi, G., and Lechler, R.I. (1994). J. Immunol. Methods, 169, 25. 28. Wilson, D. and Nowell, P. (1971). /. Exp. Methods, 133, 442. 29. Van Oers, M., Pinkster, J., and Zeijlemaker, W. (1978). Eur. J. Immunol., 8, 477. 30. Fischer-Lindahl, K. and Wilson, D. (1977). J. Exp. Methods, 145, 500. 31. Ryser, J. and MacDonald, H. (1979). J. Immunol., 122, 1691. 32. Orosz, C., Adams, P., and Ferguson, R. (1987). Transplantation, 43, 718. 33. Sharrock, C., Kaminski, E., and Man, S. (1987). Immunol. Today, 11, 281. 34. Mans, S., Lechler, R., Batchelor, J., and Sharrock, C. (1990). Eur. J. Immunol. 20, 847. 35. Breur-Vriesendorp, B., Vingerhoed, J., van Twuyver, E., de Waal, L., and Ivanyi, P. (1991). Transplantation, 51, 1096. 36. Zhang, L., Li, S., Vandekerckhove, B., Termijtelen, A., van Rood, J., and Claas, F. (1989). J. Immunol. Methods, 121, 39. 37. Kaminski, E., Sharrock, C., Hows, J., Ritter, M., Arthur, C., McKinnon, S., and Batchelor, J. (1988). Bone Marrow Transplantation, 3, 149. 196

11: The molecular basis of alloreactivity 38. Deacock, S., Schwarer, A., Batchelor, R., Glodman, J., and Lechler, R. (1992). J. Immunol. Methods, 147, 83. 39. Jooss, J., Zanker, B., Wagner, H., and Kabelitz, D. (1988). J. Immunol. Methods, 112,85. 40. Theobald, M., Hoffman, T., Bunjes, D., and Heit, W. (1990). Transplantation, 50, 850. 41. Kaminski, E., Hows, J., Man, S., Brookes, P., Mackinnon, S., Hughes, T., Avakian, O., Goldman, J., and Batchelor, J. (1989). Transplantation, 48, 608. 42. Irschick, E.U., Hladik, F., Niederwieser, D., Nussbaumer, W., Holler, E., Kaminski, E., and Huber, C. (1992). Blood, 79, 1622. 43. Schwarer, A., Jiang, Y., Brookes, P., Barrett, A., Batchelor, J., Goldman, J., and Lechler, R. (1993). Lancet, 341, 203. 44. DeBruyne, L., Ensley, R., Olsen, S. et al. (1993). Transplantation, 56, 722. 45. Herzog, W., Zanker, B., Irschick, E., Huber, C., Franz, H., Wagner, H., and Kabelitz, D. (1987). Transplantation, 43, 384. 46. Lombardi, G., Barber, L., Sidhu, S., Batchelor, J.R., and Lechler, R.I. (1991). Int. Immunol., 3, 769. 47. Barber, L.D. and Lechler, R.I. (1991). J. Immunol., 147, 2346. 48. Sambrook, J., Fritsch, E.F., and Maniatis, T. (ed.) (1989). Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, NY. 49. Barber, L.D., Bal, V., Lamb, J.R., O'Hehir, R.E., Yendle, J., Hancock, R.J.T., and Lechler, R.I. (1991). Human Immunol., 32, 110. 50. Lechler, R.I., Heaton, T., Barber, L., Bal, V., Batchelor, J.R., and Lombardi, G. (1992). Immunol. Letters, 34, 63. 51. Lombardi, G., Matsui, M., Moots, R., Aichinger, G., Sidhu, S., Batchelor, R., Frelinger, J., and Lechler, R. (1991). Immunogenetics, 34, 149. 52. Vignali, D.A.A., Urban, R.G., Chicz, R.M., and Strominger, J.L. (1993). Eur. J. Immunol., 23, 1602. 53. Aosai, F., Ohlen, C., Ljunggren, H.G., Hoglund, P., Franksson, L., Ploegh, H., Townsend, A., Karre, K., and Stauss, H.J. (1991). Eur. J. Immunol., 21, 2767. 54. Liu, Z., Sun, Y.-K., Xi, Y.-P., Maffei, A., Reed, E., Harris, P., and Suciu-Foca, N. (1993). J. Exp. Med., 177, 1643. 55. Fangmann, J., Dalchau, R., and Fabre, J.W. (1992). J. Exp. Med., 175, 1521. 56. Ramesh, P., Barber, L., Batchelor, J.R., and Lechler, R.I. (1992). Int. Immunol., 4, 935. 57. Samberg, N.L., Scarlett, E.C., and Stauss, H.J. (1989). Eur. J. Immunol., 19, 2349. 58. Hargreaves, R., Logiou, V., and Lechler, R. (1995). Int. Immunol., 7, 1505. 59. Lombardi, G., Sidhu, S., Dodi, T., Batchelor, R., and Lechler, R. (1994). Eur. J. Immunol., 24, 523. 60. Roche, P.A. and Cresswell, P. (1990). Nature, 345, 615. 61. Rath, S., Lin, R.H., Rudensky, A., and Janeway, C.J. (1992). Eur. J. Immunol., 22, 2121. 62. Demotz, S. (1993). Eur. J. Immunol., 23, 2100. 63. Dodi, A.I., Brett, S., Nordeng, T., Sidhu, S., Batchelor, R.J., Lombardi, G., Bakke, O., and Lechler, R. (1994). Eur. J. Immunol., 24, 1632. 64. Powis, S.J., Townsend, A.R., Deverson, E.V., Bastin, J., Butcher, G.W., and Howard, J.C. (1991). Nature, 354, 528. 65. Spies, T. and DeMars, R. (1991). Nature, 351, 323.

197

Liz Lightstone et al. 66. Lechler, R. and Lombardi, G. (1990). Immunol. Res., 9, 135. 67. Merkenschlager, M., Ikeda, H., Wilkinson, D., Beverly, P.C., Trowsdale, J., Fisher, A.G., and Altmann, D.M. (1991). Eur. J. Immunol., 21, 79. 68. Bjorkman, P.J. and Davis, M.M. (1989). Cold Spring Harbor Symp. Quant. Biol., 1, 365. 69. George, A., Dazzi, F., Lynch, J., Sidhu, S., Marelli, F., Batchelor, R.J., Lombardi, G., and Lechler, R.I. (1994). Int. Immunol., 6, 1785.

198

12 Elution and analysis of peptide pools from MHC class I molecules SIMON J. POWIS and GEOFFREY W. BUTCHER

1. Introduction Class I major histocompatibility complex (MHC) molecules are widely distributed polymorphic cell surface glycoproteins involved in the presentation of antigens to CD8+ T-lymphocytes. They function by binding short peptides in the environment of the endoplasmic reticulum (ER): these peptide-class I complexes are then transported to the cell surface where intercellular interaction with T-cell receptors can occur. It is a requirement for recognition by T-lymphocytes that protein antigens are degraded into small fragments prior to MHC binding. The ability to predict what fragment of any given protein antigen would bind to a class I MHC molecule and become a T-cell epitope would therefore be of great use in the development of novel vaccine strategies. In recent years several important features of the class I MHC antigen presentation pathway have been elucidated which make identifying putative epitopes a viable experimental goal. • Crystallography data for class I molecules shows a binding cleft on the upper surface, containing a short peptide in extended conformation (1, 2). • The peptides bound by class I molecules are derived mostly from proteins resident in the cytoplasm or nucleus, i.e. from endogenous sources (3). • The peptides are usually 8-10 amino acids in length, and those peptides binding to a particular type of MHC class I molecule often contain common sequence features (4, 5). • Peptides are specifically transported into the ER by the MHC-encoded TAP transporter (6, 7, 8). The class I MHC binding cleft, unlike that of class II MHC molecules, is sealed off at both ends. The resultant length restrictions that this imposes causes bound peptides to share particular amino acids at certain positions in their sequences depending on the amino acid side chain and its ability to be accommodated in pockets lining the binding cleft of the class I MHC

Simon J. Powis and Geoffrey W. Butcher Table 1. Summary of MHC class I peptide binding motifs. Class 1 molecule Motif 1

H-2Kb H-2Kd H-2Kk H-2Db Qa-2 HLA-A2.1 HLA-B53 HLA-B35 HLA-A3 HLA-B8 HLA-B27

Refs 2

3

4

5

6

7

F/Y

N H

K/R

K/R

R

9

L I/L I M

F/Y E

L P P L

8

L

18 4 19 4 11

V

4,17

Y Y/K L7R

12 12 5 13 3

molecule (9). For example, the human class I molecule HLA-A2.1 binds peptides which usually contain leucine at position 2 and valine at position 9 of the peptide. This pattern of preferred amino acid residues at specific positions in the sequence of peptides bound to a single type of MHC class I molecule is known as the 'peptide anchor motif of the molecule. Table 1 summarizes the basic motifs of several class I molecules that have been determined. The majority of class I motifs appear to be composed of two anchors, one close to the amino terminus of the peptide at position 2 or 3, and the second occurring at the peptide's carboxy terminus. It becomes obvious that class I MHC alleles therefore have the potential to bind many different peptides for subsequent scrutiny by T-lymphocytes. And, since the peptides are usually of endogenous origin, the immune system thereby has the ability to be constantly on the alert for non-self or pathogen derived proteins within cells. Such cells can be efficiently targeted for elimination by cytotoxic T lymphocytes. The process by which a class I binding motif is currently obtained relies on either one, or a combination of the following methods. • Aligning sequences of T cell epitopes known to be presented by a particular class I molecule. • Screening a large panel of random synthetic peptides for class I MHC binding. • Direct sequencing of peptide pools eluted from class I MHC molecules. In circumstances where a class I MHC molecule is being studied for which few or no known epitope sequences currently exist, and where synthesis of a large panel of random peptides would prove prohibitively expensive, pooled peptide sequencing becomes an obvious choice, and can be a valuable step in 200

12: Elution and analysis of peptide pools from MHC class I molecules directing the course of further experiments. Direct sequencing of individual peptides is addressed in the accompanying chapter by Hunt and colleagues.

2. Affinity purification of class I MHC molecules Class I MHC polypeptides are receptor molecules, differing from many other receptor systems only in their promiscuity for bound ligand. Their isolation and purification can be straightforward, consisting of a process of detergent solubilization of cell membranes expressing the class I molecule of interest and monoclonal antibody (mAb) based affinity chromatography, followed by extraction and recovery of peptides (4). Wherever possible, high affinity mAbs recognizing the class I molecule under investigation should be used, though specific polyclonal antisera may also work. Reagents should also be tested to preclude cross-reactivity with other class I molecules which may be present in the cell membrane preparation. We routinely couple preparations of 10-20 mg of purified mAbs to CnBr-activated Sepharose 4B (Pharmacia), which allows a high protein-togel ratio (see Protocol 1), and provides enough material for several separate experiments. Subclasses of immunoglobulin which normally bind to Staphylococcal Protein A, and also those which will bind Protein A under high salt concentrations, can also be coupled to Protein A-Sepharose (or Protein A-Agarose matrices) using the cross-linker dimethylpimelimidate (DMP) as described by Harlow and Lane (10). Protocol 1. Preparation of immunoaffinity reagents. Equipment and reagents • Benchtop centrifuge (e.g. MSE Centaur 2) with 10 ml or 25 ml buckets

• Rotator or rocking platform

Method 1. Weigh out 1 g of CnBr-activated Sepharose 4B (Pharmacia) and place in 10 ml round-bottomed Sterilin tube.a 2. Wash the beads extensively by repeated centrifugation at 250 g for 3 min with freshly prepared 1 mM HCI. Aspirate and discard the supernatant each time. The washing volume should exceed 50 ml. Final swollen gel volume should be 3 to 3.5 ml. 3. Resuspend the beads and wash twice with 7 ml BCB buffer (0.1 M NaHCO3, 0.5 M NaCI, pH 8.3). 4. Immediately resuspend the beads with 4 ml of mAb solution (dialysed against BCB) containing between 10 and 20 mg of protein 3-6 mg ml-1 of beads). 201

Simon J. Powis and Geoffrey W. Butcher Protocol 1. Continued 5. Rotate or rock ensuring full mixing of contents for 60 min at room temperature. 6. Spin down as above, aspirate and retain supernatant (for assessment of coupling efficiencyb), and resuspend beads in 7 ml of 0.2 M glycine pH 8.0. Rotate for 1 h at room temperature, or overnight at 4°C to block remaining sites. 7. Spin down and wash beads six times alternatively in 7 ml ACB buffer (0.1 M NaOOCCH3, 0.5 M NaCI pH 4.0) and BCB buffer. 8. Wash beads twice in PBS containing 0.1% (w/v) sodium azide. Store beads in the same buffer at 4°C.C a For larger volumes of Sepharose beads use 25 ml universal tubes and increase buffer volumes accordingly. b Read the optical density of the supernatant at 280 nm to determine coupling efficiency, which should be above 80%. c Beads can safely be stored and used in excess of six months after preparation.

In addition to the specific immunoadsorbent, it is also advisable to prepare a non-specific adsorbent for use as a pre-column. This can be prepared using Protocol 1 with a non-specific (i.e. not recognizing class I) mAb, or by blocking the beads with glycine by omission of steps 4 and 5. It is recommended that a volume of at least 5 ml of non-specific beads be prepared, which can be regenerated for use several times by elution with 0.1 M glycine pH 2.5. A variety of cell types and tissues have been utilized for the isolation of MHC bound peptides. The most commonly reported however are spleen cells and in vitro cell lines (4, 5,11).If in vitro cells are to be used prior fluorescence activated cell sorting to select for high expression of class I molecules may be considered. For both cell lines and spleens the number of cells per experiment should be between 5 and 20 X 109 in total. Protocol 2. Isolation of class I MHC molecules and acid extraction of peptides. Equipment and reagents • Ultracentrifuge • Peristaltic pump

• Vacuum centrifuge

Method 1. For in vitro cell lines wash the required number of cells and resuspend in 20 ml of ice cold PBS. For spleens, cut up the spleens in 20 ml PBS with a scalpel blade, then tease into smaller fragments with fine forceps. Push the cell suspension through a nylon sieve with a

202

12: Elution and analysis of peptide pools from MHC class I molecules 10 ml syringe plunger, adding more PBS to rinse as required. Wash the cells once and resuspend in 20 ml ice cold PBS. 2. Lyse the cells by addition of 80 ml ice cold PBS containing 1 ml NP40 (1% final concentration) and 2 mM phenylmethylsulfonyl fluoride. Mix at 4°C for 30 min. 3. Spin the lysate at 100 000 x g for 60 min at 4°C. 4. Decant the supernatant and pass in series, at a flow rate of 20 ml h-1, through the non-specific column (volume 5 ml), then through the specific column (1-1.5 ml), all at 4°C. 5. After the lysate has passed through the columns, disconnect and wash the specific column for 10-15 mins with PBS 0.5% NP-40, then for 10-15 mins with PBS 0.1% NP-40, and finally for 5-10 mins with PBS alone. 6. Dismantle the column and extract a small sample of the beads (approx 10 ul) with a Pasteur pipette for SDS-PAGE analysis. Recover the remainder of the beads by careful resuspension with 1 ml of 0.1% trifluoroacetic acid (TFA) in H20. Transfer the beads to a 10 ml centrifuge tube. Rinse out the column with a further 1 ml of 0.1% TFA and transfer to the 10 ml tube. 7.

Heat the bead suspension to at least 90°C for 5 min with frequent resuspension of the beads by tapping the tube.

8. Separate the fluid from the beads using two P1000 Gilson tips containing a 2 mm glass ball (BDH). Transfer 1 ml of bead suspension into each Gilson tip and place the tip in a 1.5 ml Eppendorf tube. Attach a P1000 Gilson on to the tip and depress the plunger to express the fluid into the Eppendorf tube. Visually inspect the contents of the tube, if any beads have leaked through repeat using a fresh Gilson tip complete with glass ball.a 9.

Pool samples and reduce the volume to approximately 200 ul by vacuum centrifugation.

10. Adjust the sample volume to 550 ul with 0.05% TFA and spin through a Centricon 3 or 10 microconcentrator (Amicon) at 5000 x g for 60 min. 11. Store filtrate, which contains the peptides, at -20°C prior to HPLC analysis. a Simple centrifugation and aspiration of the supernatant can be used in place of this step, but less fluid is recovered.

The quality of class I purification can be assessed visually by performing reducing SDS-PAGE analysis of the sample extracted in step 6 of Protocol 2. An adjacent control lane should contain a sample of Mab alone (see Figure 7). 203

Simon J, Powis and Geoffrey W. Butcher

Figure 1. SDS-PAGE analysis of the rat class I molecule RTI.Aa purified from a spleen cell lysate. Samples were boiled in reducing buffer for 5 min, and analysed on a 13% gel followed by staining with Coornassie Blue. Lane 1 contains a sample of immunoaffinity beads that have not encountered the spleen lysate and indicates the heavy (IgH) and light chains (IgL) of the mAb attached to the beads. In lane 2 the sample has been extracted from the immunoaffinity column after lysate has been passed through and non-specific material washed away. Additional bands represent the presence of class I MHC heavy chain and B2-microglobulin (B2m).

3, HPLC separation and sequencing of peptide pools The low molecular weight pool obtained at the end of Protocol 2. representing either material below 3 000 or 10 000 Da depending on the microconcentrator used, requires analysis by HPLC to further isolate the peptides. Any dominant peaks, either of peptides or fragments of class I heavy chain, B2-microglobulin or immunoglobulin can then be excluded from the pool sequence to prevent bias occurring, or indeed can be considered for individual sequence analysis (3). Reverse-phase HPLC separation of MHC-derived peptide pools can be performed on most HPLC apparatus, with a variety of columns, including C8, C18 or C2/C18 analytical or preparative columns. No special conditions apply to the separation methodology; buffer A should be H2O containing 0.0250.1% TFA, and buffer B should be 90-100% acetonitrile plus TFA as above. Gradient conditions can be a simple 0-100% gradient of buffer B, but with most of the peptides of interest eluting in the range between 8-45% buffer B this region is worth extending to a shallow gradient with time, followed by a rapid rise to 100% buffer B (see Figure 2). Absorbance should be measured in 204

12: Elution and analysis of peptide pools from MHC class I molecules

Figure 2. HPLC analysis of peptides eluted from the RT1.Aa molecule immunoaffinity purified from spleens of PVG-RT1a rats. The peptide containing filtrate, as obtained in Protocol 2, was run on an Applied Biosystems Brown lee Aquapore C18 (25 cm) column, attached to a Pharmacia Smart System. Buffer A was 0.025% TFA in H20, Buffer B was 0.025% TFA in 90% acetonitrile, with the gradient of Buffer B indicated on the figure and on the right-hand axis. Absorbance was measured at 214 nm. Fractions of interest eluting between approximately 8% and 45% Buffer B with this C18 column, as indicated between the arrows, were pooled, dried down, and sequenced by Edman degradation chemistry.

the range 210-214 nm. Fractions of interest, as indicated in Figure 2, are then pooled and dried by vacuum centrifugation, followed by resuspension in a small volume (30 ul) of 0.05% TFA with vigorous pipetting and vortexing, and submission for sequencing by Edman degradation. The sample support should be glass fibre precycled with polybrene. PVDF membranes can also be used, but consistently return significantly lower signals compared to glass fibre. Both pulsed liquid and gas phase sequencing have been employed with success, and pulsed liquid may be further enhanced by the use of shortened cycle times. Sequencing of peptide pools from class I molecules, as first described by Falk et al. (4), works because the population of peptides are of similar length and are all related by the presence of a shared amino acid at one or more positions in the sequence. Thus as sequencing proceeds from the amino terminii of the peptides, non-anchor-containing cycles appear as a random mix of many different amino acids with relatively low pmol yields. Anchor positions, on the other hand, are evident as large increases in the pmol yield of a 205

Table 2. Sequencinga of the natural peptides eluted from the rat class I MHC molecule RT1.Aa

A

R

N

D

Q

E

G

H

1

L

K

1

143.7

66.6

17.9

24.5

26.4

33.1 345.2 6.9

22.4

15.5

37.5

2

43.6

23.0

7.5

15.6

86.5

34.2

17.5

94.0

5.2

120.3 4.3

M

F

7.4 18.7 33.2

5.5

P

S

T

16.5 104.3 81.2

W

Y

V

16.8

31.8

62.6

13.5

24.4 27.2

7.1

8.2

24.5

3

41.5

20.4 1SJ.

16.3

23.6

17.1

73.9

5.0

12.2

38.2

5.4

9.9 32.5

13.4

15.3

15.3

9.1

17.1

13.2

4 5

36.6

23.6

9.7

35,7

14.8

26.5

78.3

4.1

4.6

12.3

7.1

5.2

8.1

31.9

10.5

12.7

6.8

10.2

9.6

28.5

20.9

9.1

25.5

12.5

20.6

77.4 3.9

5.4

10.4

5.8

2.3

6.9

26.5

9.6

13.9

6.3

8.0

9.6

6

21.2

15.5

6.0

18.7

9.3

14.4

79.2 2.8

&2

7.5

4.9

2.5

5.9

21.2

8.1

10.7

6.0

7.9

7.6

7

17.9

12.4

5.3

11.6

8.7

12.3

60.4

2.1

5.5

5.3

3.6

2.4

4.2

10.7

6.9

9.1

5.7

4.3

6.7

8

16.0

15.Q '

5.8

11.2

8.3

11.9

53.9

2.7

3.7

4.0

3.9

2.0

2.7

8.0

6.2

7.3

5.8

4.0

5.1

9

12.2

14.1

4.6

8.5

6.5

8.5

45.6

1.8

2.7

3.3

2.1

1.3

1.9

5.4

5.9

6.5

5.6

2.9

3.3

9.2

12.8

4.3

7.8

5.6

7.0

41.2

1.5

2.2

2.9

1.4

1.0

2.5

4.8

5.9

5.5

5.1

2.8

3.3

10

y.'V

"Sequencing was performed on an Applied Biosystems 470A gas-phase instrument with 610A-01 data analysis software.

12: Elution and analysis of peptide pools from MHC class I molecules particular amino acid, to the virtual exclusion of all others. The analysis of results and derivation of a motif is greatly aided by the use of sequencing software such as the Applied Biosystems 610A-01 data analysis system. This, and other equivalent software, permits the construction of a table of pmol values (see Table 2). Increases of 50% or more in the pmol value compared to the previous or pre-previous cycle are considered significant. As can be seen in Table 2 for the analysis of peptides eluted from the rat class I molecule RT1.Aa, purified from PVG.R19 rat spleens, cycle 2 yields large increases in the yields of leucine, glutamine and also methionine. Position 2 therefore probably represents an anchor position. In cycle 3 phenylalanine, asparagine and tyrosine show increases, but in repeat experiments containing more material the additional presence of aspartic acid, arginine, histidine and proline suggest that Phe-3, Tyr-3 and Asn-3 are not anchors, but remain strongly preferred residues, often referred to as secondary anchor positions. The remaining cycles are similarly analysed. Cycle-to-cycle signal loss is often high when sequencing short peptides. Thus in later cycles, if yields are low, the '50% increase rule' may be relaxed. Apparently minor increases would justify further analysis with more material. In the data presented in Table 2, where the yields of material were fairly low, there is a suggestion that arginine is present in cycle 8. Further experiments confirmed this is an anchor residue. The data is then compiled as a completed motif, as in Table 3, which can then be used as a guide to recognizing and isolating antigenic epitopes.

4. Use of peptide motifs Peptide motifs can be a significant aid in the recognition of likely class I restricted epitopes (5, 12, 13). The accuracy of their use is considerably improved by inclusion of the secondary anchor residues when conducting epitope searches. However, identification of a likely ligand sequence may not Table 3. Assembled peptide motif for RT1.Aa. Position

1

3

4

Frequent residues

F Y N

D E P

Also observed

D W H K

R G K

Anchor residues

2 L Q M

207

5

6

I T

G

7

8/9 R

Simon J. Powis and Geoffrey W. Butcher necessarily lead to a naturally processed epitope. Why should this be? There is increasing evidence for selectivity both in the proteolytic events contributing to antigen processing and in the transport of peptides into the ER, so some epitopes may never be generated or gain access to class I molecules (14, 15, 16). Furthermore, even if a peptide is generated and then bound and presented by a MHC class I molecule there is no guarantee that a responsive T lymphocyte will exist in the host repertoire (e.g. for reasons of tolerance to identical or related 'self' peptides). These points reinforce the necessity to test putative epitopes further in other experimental systems. In addition to pooled sequencing of peptides, if the amount of material and HPLC resolution permit, Edman degradation can be performed on single peaks, as has been reported for the natural ligands of HLA-B2705 (3), or further analysed by mass spectrometry (see ref. 17 and the chapter by Hunt). In conclusion, pooled peptide sequencing can provide valuable information leading to the prediction of novel MHC class I presented T-lymphocyte epitopes.

Acknowledgements We thank Pat Barker and Howard Smith for advice on HPLC and peptide sequencing. This work is supported by the UK Biotechnology and Biological Sciences Research Council, the UK Arthritis and Rheumatism Research Council, and the Wellcome Trust.

References 1. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C. (1987). Nature, 329, 506. 2. Madden, D. R., Gorga, J. C., Strominger, J. L., and Wiley, D. C. (1991). Nature, 353, 321. 3. Jardetzky, T. S., Lane, W. S., Robinson, R. A., Madden, D. R., and Wiley, D. C. (1991). Nature, 353, 326. 4. Falk, K., Rotzschke, O., Stevanovic, S., Jung, G., and Rammensee, H. G. (1991). Nature, 351, 290. 5. DiBrino, M., Parker, K. C., Shiloach, J., Knierman, M., Lukszo, J., Turner, R. V., Biddison, W. E., and Coligan, J. E. (1993). Proc. Natl Acad. Sci. USA, 90, 1508. 6. Neefjes, J. J., Momburg, F., and Hammerling, G. J. (1993). Science, 261, 769. 7. Sheperd, J. C., Schumacher, T. N. M., Ashton-Rickardt, P. G., Imaeda, S., Ploegh, H. L., Janeway, C. A., and Tonegawa, S. (1993). Cell, 74, 577. 8. Androlewicz, M. J., Anderson, K. S., and Cresswell, P. (1993). Proc. Natl Acad. Sci. USA, 90, 9130. 9. Garrett, T. P. J., Saper, M. A., Bjorkman, P. J., Strominger, J. L., and Wiley, D. C. (1989). Nature, 342, 692. 10. Harlow, E., and Lane, D. (1988). Antibodies: A laboratory manual. Cold Spring Harbor Press, Cold Spring Harbor, NY.

208

12: Elution and analysis of peptide pools from MHC class I molecules 11. Rotzschke, O., Falk, K., Stevanovic, S., Grahovac, B., Soloski, M. J., Jung, G., and Rammensee, H. G. (1993) Nature, 361, 642. 12. Hill, A. V. S., Elvin, J., Willis, A. C., Aidoo, M., Allsop, C. E. M., Gotch, F. M., Gao, X. M., Takiguchi, M., Greenwood, B. M., Townsend, A. R. M., McMichael, A. J., and Whittle, H. C. (1992). Nature, 360, 434. 13. Sutton, J., Rowland-Jones, S., Rosenberg, W., Nixon, D., Gotch, F., Gao, X. M., Murray, N., Spoonas, A., Driscoll, P., Smith, M., Willis, A., and McMichael, A. (1993). Eur. J. Immunol., 23, 447. 14. Driscoll, J., Brown, M. G., Finley, D., and Monaco, J. J. (1993). Nature, 365, 262. 15. Gaczynska, M, Rock, K. L., and Goldberg, A. L. (1993). Nature, 365, 264. 16. Powis, S. J., Deverson, E. V., Coadwell, W. J., Ciruela, A., Huskisson, N. S., Smith, H., Butcher, G. W., and Howard, J. C. (1992). Nature, 357, 211. 17. Hunt, D. F., Henderson, R. A., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., Cox, A. L., Apella, E., and Engelhard, V. H. (1992). Science, 255, 1261. 18. Van Bleek, G. M. and Nathenson, S. (1990). Nature, 348, 213. 19. Rammensee, H. G., Falk, K., and Rotzschke, O. (1993). Ann. Rev. Immunol., 11, 213.

209

This page intentionally left blank

Al List of suppliers Amersham Amersham International plc., Lincoln Place, Green End, Aylesbury, Buckinghamshire HP20 2TP, UK. Amersham Corporation, 2636 South Clearbrodk Drive, Arlington Heights, IL 60005, USA. Anderman Anderman and Co. Ltd., 145 London Road, Kingston-Upon-Thames, Surrey KT17 7NH, UK. Beckman Instruments Beckman Instruments UK Ltd., Progress Road, Sands Industrial Estate, High Wycombe, Buckinghamshire HP12 4JL, UK. Beckman Instruments Inc., PO Box 3100, 2500 Harbor Boulevard, Fullerton, CA 92634, USA. Becton Dickinson Becton Dickinson and Co., Between Towns Road, Cowley, Oxford OX4 3LY, UK. Becton Dickinson and Co., 2 Bridgewater Lane, Lincoln Park, NJ 07035, USA. Bio Bio 101 Inc., do Statech Scientific Ltd, 61-63 Dudley Street, Luton, Bedfordshire LU2 0HP, UK. Bio 101 Inc., PO Box 2284, La-Jolla, CA 92038-2284, USA. Bio-Rad Laboratories Bio-Rad Laboratories Ltd., Bio-Rad House, Maylands Avenue, Hemel Hempstead HP2 7TD, UK. Bio-Rad Laboratories Division Headquarters, 3300 Regatta Boulevard, Richmond, CA 94804, USA. Boehringer Mannheim Boehringer Mannheim UK (Diagnostics and Biochemicals) Ltd., Bell Lane, Lewes, East Sussex BN17 1LG, UK. Boehringer Mannheim Corporation, Biohemical Products, 9115 Hague Road, P.O. Box 504 Indianapolis, IN 462-1414, USA. Boehringer Mannheim Biochemica, GmbH, Sandhofer Str. 116, Postfach 310120 D-6800 Ma 31, Germany.

List of suppliers British Drag Houses (BDH) Ltd., Poole, Dorset, UK. Difco Laboratories Difco Laboratories Ltd., P.O. Box 14B, Central Avenue, West Molesey, Surrey KT8 2SE, UK. Difco Laboratories, P.O. Box 331058, Detroit, MI 48232-7058, USA. Du Pont Dupont (UK) Ltd., Industrial Products Division, Wedgwood Way, Stevenage, Herts, SG1 4Q, UK. Du Pont Co, (Biotechnology Systems Division), P.O. Box 80024, Wilmington, DE 19880-002, USA. European Collection of Animal Cell Culture, Division of Biologies, PHLS Centre for Applied Microbiology and Research, Porton Down, Salisbury, Wilts. SP4 0JG, UK. Falcon (Falcon is a registered trademark of Becton Dickinson and Co.) Fisher Scientific Co., 711 Forbest Avenue, Pittsburgh, PA 15219-4785, USA. Flow Laboratories, Woodcock Hill, Haresfield Road, Rickmansworth, Herts WD3 1PQ, UK. Fluka Fluka-Chemie AG, CH-9470, Buchs, Switzerland. Fluka Chemicals Ltd., The Old Brickyard, New Road, Gillingham, Dorset SP8 4JL, UK. Gibco BRL Gibco BRL (Life Technologies Ltd.), Trident House, Renfrew Road, Paisley PA3 4EF, UK. Gibco BRL (Life Technologies Inc.), 3175 Staler Road, Grand Island, NY 14072-0068, USA. Arnold R. Horwell, 73 Maygrove Road, West Hampstead, London NW6 2BP, UK. Hybaid Hybaid Ltd., 111-113 Waldegrave Road, Teddington, Middlesex TW11 8LL, UK. Hybaid, National Labnet Corporation, P.O. Box 841, Woodbridge, N.J. 07095, USA. HyClone Laboratories 1725 South HyClone Road, Logan, UT 84321, USA. International Biotechnologies Inc., 25 Science Park, New Haven, Connecticut 06535, USA. Invitrogen Corporation Invitrogen Corporation 3985 B Sorrenton Valley Building, San Diego, CA 92121, USA. Invitrogen Corporation do British Biotechnology Products Ltd., 4-10 The Quadrant, Barton Lane, Abingdon, OX14 3YS, UK. Kodak: Eastman Fine Chemicals 343 State Street, Rochester, NY, USA. Life Technologies Inc., 8451 Helgerman Court, Gaithersburg, MN 20877, USA. 212

List of suppliers Merck Merck Industries Inc., 5 Skyline Drive, Nawthorne, NY 10532, USA. Merck, Frankfurter Strasse, 250, Postfach 4119, D-64293, Germany. Millipore Millipore (UK) Ltd., The Boulevard, Blackmoor Lane, Watford, Herts WD1 8YW, UK. Millipore Corp./Biosearch, P.O. Box 255, 80 Ashby Road, Bedford, MA 01730, USA. New England Biolabs (NBL) New England Biolabs (NBL), 32 Tozer Road, Beverley, MA 01915-5510, USA. New England Biolabs (NBL), c/o CP Labs Ltd., P.O. Box 22, Bishops Stortford, Herts CM23 3DH, UK. Nikon Corporation, Fuji Building, 2-3 Marunouchi 3-chome, Chiyoda-ku, Tokyo, Japan. Perkin-Elmer Perkin-Elmer Ltd., Maxwell Road, Beaconsfield, Bucks HP9 1QA, UK. Perkin-Elmer Ltd., Post Office Lane, Beaconsfield, Bucks HP9 1QA, UK. Perkin-Elmer-Cetus (The Perkin-Elmer Corporation), 761 Main Avenue, Norwalk, CT 0689, USA. Pharmacia Biotech Europe Procordia EuroCentre, Rue de la Fuse-e 62, Bll 30 Brussels, Belgium. Pharmacia Biosystems Pharmacia Biosystems Ltd., (Biotechnology Division), Davy Avenue, Knowlhill, Milton Keynes MK5 8PH, UK. Pharmacia LKB Biotechnology AB, Bjorngatan 30, S-75182 Uppsala, Sweden. Promega Promega Ltd., Delta House, Enterprise Road, Chilworth Research Centre, Southampton, UK. Promega Corporation, 2800 Woods Hollow Road, Madison, WI 53711-5399, USA. Qiagen Qiagen Inc., c/o Hybaid, 111-113 Waldegrave Road, Teddington, Middlesex TW11 8LL, UK. Qiagen Inc., 9359 Eton Avenue, Chatsworth, CA 91311, USA. Schleicher and Schuell Schleicher and Schuell Inc., Keene, NH 03431A, USA. Schleicher and Schuell Inc., D-3354 Dassel, Germany. Schleicher and Schuell Inc., c/o Andermann and Company Ltd. Shandon Scientific Ltd., Chadwick Road, Astmoor, Runcorn, Cheshire WA7 1PR, UK. Sigma Chemical Company Sigma Chemical Company (UK), Fancy Road, Poole, Dorset BH17 7NH, UK. 213

List of suppliers Sigma Chemical Company, 3050 Spruce Street, P.O. Box 14508, St. Louis, MO 63178-9916. Sorvall DuPont Company, Biotechnology Division, P.O. Box 80022, Wilmington, DE 19880-0022, USA. Stratagene Stratagene Ltd., Unit 140, Cambridge Innovation Centre, Milton Road, Cambridge CB4 4FG, UK. Stratagene Inc., 11011 North Torrey Pines Road, La Jolla, CA 92037, USA. United States Biochemical, P.O. Box 22400, Cleveland, OH 44122, USA. Wellcome Reagents, Langley Court, Beckenham Kent BR3 3BS, UK.

214

Index ADCC 94-5 antigen presentation 1-12, 33, 49, 74, 123, 141 antigen presenting cells 1-3, 11, 74, 117, 132, 139, 161, 174-75, 191-95 isolation 2-10 see also dendritic cells, macrophages, B cells APC's see antigen presenting cells apoptosis 13, 33, 44-7 detection 44-7 ATP 49-51, 60 B7 see CD80 B cells 9-12, 27, 33, 182-84 activation 9 lymphoblastoid 24, 74, 182-84, 190, 194 calcium, intracellular 34-6 cAMP 33 CD80 3, 9, 194 chloramine T 21 complement 5, 11-12 cryopreservation 79-80 CTLp see T cells—cytotoxic T lymphocyte precursors dendritic cells 2-8, 11-12 interdigitating cells 2 isolation 2-5 Langerhans cells 2-3 veiled cells 2 diacylglycerol 33, 36 endocytosis 18, 22-7 primaquine, effect of 26-7 endoplasmic reticulum see peptide translation epithelial cells, isolation 132-36 exon shuffling 185-90 splicing by overlap extension 186-90 flourescence activated cell sorter 171-73 glycoproteins, MHC 17, 22-9 GM-CSF 3, 94 hemi-exon shuffling see exon shuffling

HPLC 146-55, 191-93, 204-8 reverse phase 146-48 chromatography, first dimension 146-47 chromatography, second dimension 147-49 human leukocyte antigen (HLA) HLA-A 73,102,141,156 HLA-B 73, 95-6, 102, 208 HLA-C 96,102 HLA-DP 46, 73-4,107 HLA-DQ 46,73,107,194 HLA-DR 34,46,73-4,107,113-4,166,185, 190,194-5 ICAM-1 194 immunoaffinity chromatography 142-46 indo-1 34-5 inositol phospholipid turnover 36-7 interferon y 7-8,12-14,60, 65-8,79,94,166 interleukins IL-2 12-13,79,94-6,117,124,127,130, 178-79 IL-3 12-13,79 IL-4 9,12-13,79,94 IL-7 79, 94 IL-9 79 IL-12 8,79,94 invariant chain 1 labelling, radioactive 18-31,66-8,113-15 of cells 66-8 of enterotoxins 113-15 of MHC 18-30 lactoperoxidase 18-19 LDA see limiting dilution analysis limiting dilution analysis 175-83 macroglobulin, a2 7 macrophages 3,7-9,12 isolation 8-9 MHC II a chain 29-31,162 MHC II p chain 29-31,162 MHC glycoproteins see glycoproteins, MHC MHC labelling 18-21 see also labelling, radioactive MHC-peptide complexes 1-3,9,17,162 with superantigens 113-15 isolation of peptides 144-46 MHC purification 201-204 MHC, transfection into adherent cells 165-75 MHC turnover 17-31

Index microglobulin, a2 7 microglobulin, (32 24 mixed lymphocyte cultures 72-3 mixed lymphocyte reactions 2,161 MLC see mixed lymphocyte cultures MLRs see mixed lymphocyte reactions

SDS-PAGE 22,27-30,40-42,61-63, 68-69, 202-204 signal transduction 38-9 spectrometry,mass 141-59 electrospray ionization 142-44,149-55 tandem 155-57 streptolysin O 50-4 superantigens, bacterial 107-19 MHC binding 113-15 production 109-113 T cell stimulation 117-19 Zn binding 115-16

NK cells 91-104,176,182 allorecognition 99-101 cytokine production 94 cytotoxicity 99-101 functions 94 purification 96-9 receptors 95-6 PBMC see peripheral blood mononuclear cells PCR see polymerase chain reaction peptide modification, chemical 157-59 peptide sequencing 155-57,204-207 peptide translocation 49-57 into endoplasmic reticulum 49-54,199 in microsomes 49,54-5 TAP-dependent 49-54,199 see also TAP peripheral blood mononuclear cells 72-4,124, 127,131,173 phagocytosis 7 phagolysosomes 7 phospholipase C 33-7,46 pinocytosis, fluid phase 7 PKC see protein kinase C poisson distribution 176-77 see also limiting dilution analysis polymerase chain reaction 82-8,109-13, 187-90 protein kinase C 33,38-9,46 proteosome 20S 59-69 26S 60-1

TAP 49-57,60,199 T cells 71-88,91,123-39,161-95,199-201 alloreactive 11,174-95 cloning 75-9,123-31 cytotoxic 13-15,49,71-3,102,141,176, 200 cytotoxic T lymphocyte precursors 176-84 helper 73,176 isolation 72-5 proliferation 80-2 stimulation by superantigen 117-19 see also superantigens, bacterial activation 17,82-8,109,123-24,131 T cell receptor 9, 85-8,91-2,107-9,117-18, 123,130,136-39,161-65,185,190-91,195, 199 activation 11-12 sequencing 85-8 thymocytes 5 TNF see tumour necrosis factor tolerance peripheral 8 self 56-7 tumour necrosis factor 94 turnover time of MHC 31 tyrosine kinase 40-3,46 tyrosine phosphorylation 38-43

216