Roitt's Essential Immunology, 11th Edition

Essentiql Peter J. Delves PhD Department of Immunology and Molecular Pathology University College London London,U.K.

Seomus J.Mortin PhD,FTCD,MRIA TheSmurfit Institute of Genetics TrinityCollege Dublin2,Ireland

Dennis R.Burton PhD Departmentof Immunology and Molecular Biology TheScrippsResearchInstitute Califomia,USA

lvonM.Roitt MA, DSc(Oxon), FRDPath, Hon FRCP (Lond), FRS Emeritus Professor,Department of Immunology and Molecular Pathology UniversityCollege London London

Blackwell Publishing

Roittt Essentiol lmmunology

PelerJ. Delves Dr Delves obtained his PhD from the University of London in 1986 and is currently a Reader in Immunology at University College London. His research focuses on molecular aspects of antigen recognition. He has authored and edited a number of immunology books, and teaches the subject at a broad range of levels

Seomus J. Mofiin Professor Martin received his PhD from The National University of Ireland in 1990 and trained as a post-doctoral fellow at University College London (with Ivan Roitt) and The La Jolla Institute for Allergy and Immunology, California, USA (with Doug Green). Since 1999, he is the holder of the Smurfit Charr of Medical Genetics at Trinity College Dublin and is also a Science Foundation Ireland Principal Investigator His research is focused on various aspects of programmed cell death (apoptosis) in the immune system and in cancer and he has received several awards for his work in this area. He has previously edited two books on apoptosis and was elected as a Member of The Royal IrishAcademy in 2006

Dennis R.Burlon Professor Burton obtained his BAin Chemistry from the Universityof

OxfordinT9T4and his PhD in Physical Biochemistry from the University of Lund in Sweden in 1978 After a period at the University of Sheffield, he moved to the Scripps Research Institute in La Jolla, California in 1989 where he is Professor of Immunology

and Molecular Biology. His research

interests include antibodies, antibody responses to pathogens and vaccine design, particularly in relation toHIV

lvon M. Roitl Professor Roitt was born in 1927 and educated at King Edward's School, Birmingham and Balliol College, Oxford. In 1956, together with Deborah Doniach and Peter Campbell, he made the classic discovery of thyroglobulin autoantibodies in Hashirnoto's thyroiditis which helped to open the whole concept of a relationship between autoimmunity and human disease. The workwas extended to an intensive study of autoimmune phenomena in pernicious anaemia and primary biliary cirrhosis. In 1983 he was elected a Fellow of The Royal Society, and has been elected to Honorary Membership of the Royal College of Physicians and appointed Honorary Fellow of The Royal Society of Medicine.

@2005 Peter J Delves, Seamus J. Martin, Dennis R. Burton,Ivan M Roitt Published by Blackwell Publishing Ltd Blackwell Publishing,Inc.,350 Main Street, Malden, Massachusetts 02148-5020,USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Publishing Asia Pty Ltd,550 Swanston Street, Carlton, Victoria 3053,Austraha The right of theAuthors to be identified as the Authors of this Workhas been asserted in accordance with the Copyright, Designs and PatentsAct 1988 All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form orby any means, electronic, mechanical, photocopying, recording or otherwise, except as permittedby the UK Copyright, Designs and PatentsAct 1988,without the prior permission of the publisher. First published 1971 Reprinted 1972 (twice), 1973 (twice) Second edition 1974,Reprtnted, 1975 Third edition 1977,Reprinted 1978,1979 Fourth edition 1980,Reprinted 1982,7983 Fifth edition 1984 Sixth edition 1988,Reprinted 1988,1989 Seventhedition 1991 Eighth edition 1994,Reprinted 1996 Ninth edition 1997,Reprinted 1999 Tenth edition 2001, Reprinted 2003

Portugueseedittons1973,1979,7983 Frencheditions1975. 7979. 7990 Dutch editions1975, 1978, 1982 fapaneseeditions1976, 1978,1982,1986,1988 Germaneditions1977,1984,1988,1993 Polishedition 1977 Greekeditions1978,7989,7992 Turkishedition 1979 Slovakedition 1981 Indonesianeditions1985,1991 Russianedition1988 Koreanedition 1991

Spanish editions1972,1975,1978,7982, 1988,7989.1993 Italian editions 1973,7975,L979,1,986, 1988,1990,1.993,7995 Eleventh Edition 2006

ELBS editions 1977,1982,1.988,Reprinted 1991 Chinese (Taiwan) editions 1991,1994

Library of Congress Cataloging-in-Publication Data Data available

ISBN-13: 978-1- 4051,-3603-7 ISBN-10:1-4051-3603-0 Acatalogue record for this title is available from the British Library Set n9.75 / 12.5pt Palatinoby SNP Best-setTypesetter Ltd., Hong Kong Printed andbound by Rotolito Lombarda SpA, Turin,Italy Commissioning Editor: EditorialAssistant: DevelopmentEditor: ProductionController:

Martin Sugden CarolineBoyd MirjanaMisina KateCharman

Website: Meg Barton, Shaun Embury, fake Farr For further information on Blackwell Publishing, visit our website: http: / /www blackwellpublishing.com The publisher's policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the pubtsher ensures that the textpaper and coverboard used have met acceDtableenvironmental accreditation standards.

Gontents

Acknowledgements, x Preface,xi Abbreviations, xii User guide, xvi

1 Innate immunity, 1 Extemal barriers against infection, 1 Phagocytic cells kill microorganisms, 2 Complement facilitates phagocytosis, 10 Complement can mediate an acute inflammatory reaction, 13 Humoral mechanisms provide a second defensive strategy,l6 Extracellular killing, 18 2 Specific acquired immunity,2l Antibody-the specific adaptor, 2L Cellular basis of antibody production, 23 Acquired mernory,ZS Acquired immunity has antigen speci ficlty, 29 Vaccination depends on acquired memory,30 Cell-mediated immunity protects against intracellular organisms, 31 Immunopathology,32 3 Antibodies,3T The division of labor, 37 Five classesof immunoglobulin,3T The IgG molecule,38 The structure and function of the immunoglobulin classes,43 Genetics of antibody diversity and function,52 4 Membrane receptors for antigen,5l The B-cell surfacereceptor for antigen (BCR),61 The T-cell surface receptor for antigen (TCR),53

lui

CONTENTS

The generation of diversity for antigen recognition, 67 NK receptors, 72 The major histocompatibility complex (MHC), 75 5 The primary interaction with antigen,86 What antibodies see,86 Identifying B-cell epitopes on a protein, 89 Thermodynamics of antibody-antigen interactions, 91 Specificity and cross-reactivity of antibodies, 92 What the T-cell sees,95 Processing of intracellular antigen for presentation by Class I MHC, 95 Processing of antigen for Class II MHC presentation follows a different pathway, 97 Cross-presentation for activation of naive CD8+ T-cells, 100 The nature of the'groovy'peptide, 101 The ap T-cell receptor forms a ternary complex with MHC and antigenic peptide, 103 T-cells with a different outlook, 105 Superantigens stimulate whole families of lymphocyte receptors, 106 The recognition of different forms of antigen by B- and T-cells is advantageous to the host, 107 6 Immunological methods and applications, 111 Making antibodies to order, 111 Purification of antigens and antibodies by affinity chromatography, 117 Modulation of biological activity by antibodies, 118 Immunodetection of antigen in cells and tissues, 119 Detection and quantitation of antigenby antlbody,721 Epitope mapping,130 Estimation of antibody, 131 Detection of immune complex formation,I3T Isolation of leukocyte subpopulations, 137 Gene expression analysis, 139 Assessment of functional activity, 140 Cenetic engineering of cells,147 7 The anatomy of the immune response, 155 The need for organized lymphoid tissue, 155 Lymphocytes traffic between lymphoid tissues, 156 Lymphnodes, 160 Spleen,162 The skin immune system, 162 Mucosalimmunity,I63 Bone marrow can be a major site of antibody synthesis, 166 The enjoyment of privileged sites, 167 The handlin g of antigen, 167 8 Lymphocyte activation, L71 Clustering of membrane receptors leads to their activation, 171 T-lymphocytes and antigen-presenting cells interact through several pairs of accessory m olecules, 772 The activation of T-cells requires two signals,l73 Protein tyrosine phosphorylation is an early eventin T-cell signaling,773 Downstream events following TCR sign aling, 174

CONTENTS B-cells respond to three different types of antigen, 178 The nature of B-cell activation, 180 9 The production of effectors, 185 Cytokines act as intercellular messengers, 185 Different T-cell subsets can make different cytokine patterns, 191 Activated T-cells proliferate in response to cytokines,l94 T-cell effectors in cell-mediated immunity, 195 Proliferation and maturation of B-cell responses are mediated by cytokines, 200 \zVhatis going on in the germinal center?,200 The synthesis of antibody,202 Immunoglobulin class switching occurs in individual B-cells,202 Factors affecting antibody affinity in the immune response,204 Memorycells,206 10 Controlmechanisms,2l-L Antigens can interfere with each other,211 Complement and antibody also play a role,277 Activation-induced cell de ath, 273 T-cell regulation,2l3 Idiotype networks,218 The influence of genetic f actors,22} Regulatory immunoneuroendocrine n etw orks, 223 Effects of diet, exercise,trauma and age on immunity,226 11 Ontogeny and phylo geny, 229 Hematopoietic stem cells, 229 The thymus provides the environment for T-cell differentiation,229 T-cell ontogeny,233 T-cell tolerance,237 B-cells differentiate in the fetal liver and then inbonernanow,243 B-1 and B-2 cells represent two distinctpopulations,244 Development of B-cell spe cificity, 245 The induction of tolerance in B-lymphocytes,247 Natural killer (NK) cell ontogeny, 249 The overall responsein the neonate,250 The evolution of the immune response,250 The evolution of distinct B- and T-cell lineages was accompaniedby the development of separate sites for differentiation, 252 Cellular recognition molecules exploit the immunoglobulin gene superfamily,252 12 Adversarial strategies duringinfection,25S Inflammation revisited, 256 Extracellularbacteria susceptibleto killingbyphagocytosis and complement,260 Bacteria which grow in an intracellular habitat,268 Immunity to v ir al infection, 272 Immunityto fungi,277 Immunity to parasitic infe ctions, 278 13 Vaccines,287 Passively acquired immunity, 287 Vaccination,290 Killed organisms as vaccines,290

vii I

t...

I vill

CONIENTS

Live attenuated organisms have many advantagesas vaccines,291 Subunit vaccines containing individual protective anligens, 294 Epitope-specific vaccines may be nee ded, 297 Current vaccines,301 Vaccines under development, 301 Vaccines against parasitic diseaseshave proved particularly difficult to develop, 305 Vaccines for protection against bioterrorism, 307 Immunization against can cer,307 Other applications for vaccines,307 Adjuvants,307 14 Immunodeficiency,3l2 Deficienciesof innate immune mechanisms,3l2 Primary B-cell deficiency, 315 Primary T-cell deficiency, 316 Combined immunodeficiency, 318 Recognition of immunodeficiencies, 320 Treatment of primary immunodeficiencies, 320 Secondary immunodeficie ncy, 320 Acquired immunodeficiency sl,ndrome (AIDS), 321 15 Hypersensitivity,335 Anaphylactic hypersensitivity (Type I), 336 Antibody-dependent cytotoxic hypersensitivity (Type II), 347 Immune complex-mediated hypersensitivity (Type III), 350 Cell-mediated (delayed-type) hypersensitivity (Type IV), 356 Stimulatory hypersensitivity (Type V), 359 'Innate' hypersensitivity reactions,360 15 Tiansplantatiory364 Genetic control of transplantation anti gens, 364 Someother consequencesof MHC incompatibility,366 Mechanisms of graft rejection,367 The prevention of graft rejection,369 Is xenografting a practical proposition?, 375 Stem cell therapy,376 Clinical experience in gra fting, 377 The fetus is a potential allograft,380 1.7 Tumor immunology, 384 Cellular transformation and immune surveillance, 384 Tumor antigens,385 Spontaneousimmune responsesto tumors,389 Tumor escapemechanisms,391 Unregulated development gives rise to lymphoproliferative disorders, 392 Approaches to cancer immunotherapy, 398 Immunodiagnosis of solid tumors, 407 18 Autoimmunediseases,4l0 The scopeof autoimmune diseases,410 Nature and nurture,413 Autoreactivity comes na tur ally, 420 Is autoimmunity driven by antigen?, 422

Control of the T-helpercell is pivotal,423 Autoimmunity canarisethroughbypassof T-helpers,424 Autoimmunity canarisethrough bypassof regulatorymechanisms, 428 Autoimmune disordersaremultifactorial, 431, Pathogeniceffectsof humoral autoantibody,432 Pathogeniceffectsof complexeswith autoantigens,435 T-cell-mediatedhypersensitivityasapathogenicfactorin autoimmunedisease,440 Someother systemicvasculardisorderswithimmunopathological components,444 Diagnosticvalue of autoantibodytests,445 Treatmentof autoimmunedisorders,445

Appendix: Glossary,456 Index,467

Companionwebsite:www.roitt.com

Aclrnowledgements

The input of the editorial team of Martin Sugden, Mirjana Misina and Meg Barton at Blackwell Publishing, and the illustrators Anthea Carter and Graeme Chambers is warmly acknowledged. We are much indebted to the co-editors of lmmunology, ]. Brostoff and D. Male, together with the publishers, Mosby, and the following individuals for permission to utilize or modify their figures: J. Brostoff and A. Hall for figures 1.15 and 15.10;J. Horton for figure 77.20;G. Rook for figures 72.5 and72.11;and I. Thverne for fi gure12.22and table12.2. IMRwould like to acknowledge the indefatigable secretarial assistance of Christine Griffin. DRB wishes to particularly acknowledge the invaluable contributions of Amandeep Gakhal, Erin Scherer, Rena Astronomo and Wendelien Oswald. He is grateful to Jenny Woof,

Ann Feeney, Beatrice Hahn, fim Marks, Don Mosier, Paul Sharp, Robyn Stanfield, James Stevens and Mario Stevenson for many very helpful comments. PJD would particularly like to thank Per Brandtzaeg, Volker Brinkmann and Peterlydyard. Every effort has been made by the authors and the publisher to contact all the copyright holders to obtain their permission to reproduce copyright material. However, if any have been inadvertently overlooked, the publisher will be pleased to make the necessary arrangements at the first opportunity. A number of scientists very generously provided illustrations for inclusion in this edition, and we have acknowledged our gratitude to them in the relevant figure legends.

Prefoce

In view of the ever-increasing intensity and scope of the subject of Immunology and global differences in the curricula, it was decided that the authorship of the 10th edition be broadened to include notable scientists from other countries. Accordingly, Professor Seamus Martin of Trinity College, Dublin and Professor Dennis Burton of the Scripps Research Institute, California were invited to contribute to the current edition. We feel sure that the reader will find that their contributions will make a powerful impact and yet still maintain the reader-friendly character of earlier editions. Many subjects have been thoroughly updated and

sometimes expanded. These include sections on HIV AIDS, regulatory T-cells, chemokines, cell signaling, T-cell development, vaccines and therapies involving lymphocyte ablation. Dear reader, we hope you will enjoy this new edition and find the content attractive and rewarding.

PeterJ.Delves SeamusJ. Martin DennisR. Burton IvanM.Roitt

Abbreviqtions

AAV Ab ACh-R ACT ACTH ADA ADCC AEP Ag AID AIDS AIRE ANCA APC ARRE-1 ARRE-2 ART ASFV AZT

adeno-associatedvirus antibody acetylcholine receptor adoptive cell transfer adrenocorticotropic hormone adenosine deaminase antibody-dependent cellular cytotoxicity asparagine endopeptidase antigen activation-induced cytidine deaminase acquired immunodefi ciency syndrome autoimmune regulator antineutrophil cytoplasmic antibodies antigen-presenting cell antigen receptor responseelement-1 antigen receptor responseelement-2 antiretroviral therapy African swine fever virus zidovudine (3' -azido-3' -deoxythymidine)

BAFF

B-cell-activating factor of the tumor necrosis factor family lymphocyte which matures inbone marrow bacille Calmette-Gu6rin attenuated form of fuberculosis B-cell receptor bonemarrow bovine serum albumin bovine spongif orm encephalopathy Bruton's tyrosine kinase bromodeoxyuridine

B-cell BCG BCR BM BSA BSE Btk BUDR C cu(F/v/6) CALLA cAMP CCP CD CDR

complement constantpart ofTCR o(p/y/6) chain common acute lymphoblastic leukemia antigen cyclic adenosine monophosphate complement control protein repeat cluster of differentiation complementarity determining regions of Ig or TCR variable portion

CEA CFA cGMP CHIP Cn(r) CLA CLIP CMI CML CMV Cn Cn iCn Cna

cpG CR(n) CRP CsA CSF CSR CTLR D gene DAF DAG DC DMARD DNP DTH DTP

EAE EBV ELISA EM

EO EPO

carcinoembryonic antigen complete Freund's adjuvant cyclic guanosine monophosphate chemotaxis inhibitory protein constant part of Ig heavy (light) chain cutaneous lymphocyte antigen classIl-associated invariant chain peptide cell-mediated immunity cell-mediated lympholysis cytomegalovirus complement component'n' activated complement component'n' inactivated complement component'n' small peptide derived by proteolytic activation of Cn cytosine phosphate-guanosine dinucleotide motif complement receptor'n' C-reactive protein cyclosporinA cerebrospinal fluid classswitch recombination C-type lectin receptors diversity minigene joining V and J segmentsto form variable region decay accelerating factor diacylglycerol dendritic cells disease-modifying antirheumatic drug dinitrophenyl delayed-type hypersensitivity diphtheria, tetanus, pertussis triple vacclne experimental allergic encephalomyelitis Epstein-Barr virus enzyme-linked immunosorbent assay electron microscope eosinophil ervthroooietin

ABBREVIATIONS ER ES ET

endoplasmic reticulum embryonic stem (cell) exfoliative toxins

F(B) Fab

factor (B, etc.) monovalent Ig antigen-binding fragment after papain digestion divalent antigen-binding fragment after pepsin digestion Fas-ligand fluorescence-activatedcell sorter Ig crystallisable-fragment originally; now non-Fab part of Ig receptor for IgG Fc fragment follicular dendritic cell flk-2ligand (single chain) Vrr-V, antigenbinding fragment

F(ab'), FasL FACS Fc FcyR FDC flt-3 (sc)Fv

GADS gb m G-CSF GEFs GM-CSF gpn GRB2 GSK3 g.v.h. F{-2 H-2D /K/L (A/E) HAMA HATA HBsAg hCG HCMV HEL HEV

GRB2-related adapter protein glomerular basem*nt membrane granulocyte colony-stimulating factor guanine-nucleotideexchangefactors granulocyte-macrophage colony-stimulating factor nkDaglycoprotein growth factor receptor-binding protein 2 glycogen synthase kinase 3 graft versus host

Hi HIV-1(2) HLA HLA-A/B /C (DPIDQ/DR) HMG HR HRF HSA HSC hrp 5HT HTLV H-Y

the mouse major histocompatibility complex main loci for classicalclassI (classII) murineMHCmolecules human antimouse antibodies human anti-toxin antibody hepatitis B surface antigen human chorionic gonadotropin humancytomegalovirus hen egg lysozyme high-walled endothelium of post capillary venule high human immunodeficiency virus-1 (2) the human major histocompatibility complex main loci for classicalclassI (classII) humanMHCmolecules highmobilitygroup hypersensitive response hom*ologous restriction factor heat-stable antigen hematopoietic stem cell heat-shock protein S-hydroxytryptamine human T-cell leukemia virus male transplantation antigen

IBD ICAM-I Id (uld) IDC

inflammatorybowel disease intercellularadhesionmolecule-1 idiotype (anti-idiotype) interdigitating dendritic cells

IDDM IDO IEL IFNct Ig IgG slg Ig-a/Ig-B IgSF IL-1 iNOS IP, ISCOM ITAM ITIM ITP IVIg

insulin-dependent diabetes mellitus indoleamine 2,3-dioxygenase intraepithelial lymphocyte a-interferon (also IFNB,IFNI) immunoglobulin immunoglobulin G (also IgM ,IgA,lgD, IgE) surface immunoglobulin membrane peptide chains associatedwith slg B-cellreceptor immunoglobulin superfamily interleukin-1 (also IL-2, IL-3, etc ) inducible nitric oxide synthase inositol triphosphate immunostimulating complex immunoreceptor tyrosine-based activation motif immunoreceptor tyrosine-based inhibitory motlt idiopathic thrombocytopenic PurPura intravenous immunoglobulin

JAK J chain ,l gene

Januskinases peptide chain in IgAdimer and IgM joining gene linking V or D segment to constantregron

Ka(d)

association (dissociation) affinity constant (usually Ag-Ab reactions) units of molecular mass in kilo Daltons killer immunoglobulin-like receptors keyhole limpethemocyanin

kDa KIR KLH LAK LAMP LAT LATS LBP LCM y"a/b/x LFA-1 LGL LHRH LIF Lo LT(B) LPS MQ mAb MAC MAdCAM MALT MAM MAP kinase MAPKKK MBL MBP

lymphokine activated killer cell lysosomal-associatedmembrane Proteins linker for activation of T cells long-acting thyroid stimulator LPSbindingprotein lymphocytic choriomeningitis virus Lewisu/b/'blood group antigens lymphocyte functional antigen-1 large granular lymphocyte luteinizing hormone releasing hormone leukemia inhibiting factor low leukotriene(Betc.) lipopolysaccharide(endotoxin) macrophage monoclonal antibody membrane attack complex mucosal addressin cell adhesion molecule mucosa-associatedlymphoid tissue Mycoplasmaarthritidismitogen mitogen-activated protein kinase mitogen-associatedproteinkinasekinase kinase mannosebinding lectin majorbasicproteinof eosinophils(alsomyelin basic protein)

I xiv

ABBREVIATIONS

MCP MCP-1 M-CSF MDP MHC MICA MIDAS MIF MIIC MLA MLR MMTV MRSA MS MSC MSH MTP MULV

membrane cofactor protein (C' regulation) monocyte chemotactic protein-1 macropha ge colony-stimulating f actor muramyl dipeptide maj or histocompatibility complex MHC classI chain-related Achain metal ion-dependent adhesion site macrophage migration inhibitory factor MHC classIl-enriched compartments monophosphoryl lipid A mixed lymphocyte reaction mouse mammary tumor vrrus methicillin-res istant StaphyIococcus aLUeus multiple sclerosis mesenchymal stemcell meianocyte stimulating hormone microsomal triglyceride-transfer protein murine leukemia virus

NADP

nicotinamide adenine dinucleotide phosphate neutrophil activating peptide nitro blue tetrazolium neutrophil chemotactic factor nuclear factor of activated T-cells nuclear transcription factor natural killer cell nitric oxide Nonobese diabetic mouse New Zealand Blackmouse New Zealand Blackmousex NZ White F1 hybrid

NAP NBT NCF NFAT NFrB NK NO. NOD NZB NZBxW

.o-2 OD ORF OS Ova PAF(-R) PAGE PAMP PBSCs PCA PCR PERV PG(E) PHA phox PI3K PIAS PIgR PIP2 PKC PKR PLC PLC^/2 PMN PMT PNH

superoxide anion optical density open reading frame obesestrain chicken ovalbumin platelet activating factor (-receptor) polyacrylamide gel electrophoresis pathogen-associatedmolecular pattern peripheral blood stem cells passive cutaneous anaphylaxis polymerase chain reaction porcine endogenous retrovrruses prostaglandin (E etc.) phytohemagglutinin phagocyte oxidase phosphatidylinositol 3-kinase protein inhibitor of activated STAT poly-Ig receptor phosphatidylinositol diphosphate protein kinase C RN A-dependentprotein kinase phospholipaseC phospholipase C/2 polymorphonuclear neutrophil photomultiplier tube paroxysmal nocturnal hemo globinuria

PPAR PPD PRR PTFE PTK PWM RA RANTES RAST RF Rh(D) RIP RNAi ROI RSS SAP SAP SAR SARS SARS-CoV SC SCF scFv

SCG SCID SDF SDS SDS-PAGE SEA(Betc.) SEREX siRNA SIV SLE SLIT SLP76 SOCs SPE SRID SSA STAT

TACI

TAP T-ALL TB Tc T-cell TCF

peroxisome proliferator-activated receptor purified protein derivative from My cobacterium tuberculosis pattern recognition receptors polytetrafl uroethylene protein tyrosine kinase pokeweedmitogen rheumatoid arthritis regulated uponactivation normal T-cell expressedand secretedchemokine radioallergosorbenttest rheumatoid factor rhesus blood group (D) rat insulin promoter RNAinterference reactive oxygen intermediates recombination signal sequence serum amyloid P sphingolipid activator protein systemic acquired resistance severeacute respiratory syndrome SARS-associated coronav irus Ig secretory component stem cell factor single chain variable region antibody fragment (Vr, + V,-joined by a flexible linker) sodium cromoglycate severecombined immunodeficiency stromal-derived factor sodium dodecyl sulfate sodium dodecylsulf ate-polyacrylamide gel electrophoresis Staphylococcus aureusenterotoxin A (B etc ) serological analysis of recombinant cDNA expression libraries short-interfering RNA Simian immunodeficiency virus systemic lupus erythematosus sublingual allergen immunotherapy SH2-domain containing leukocyte protein of 76kDa suppressor of cytokine signaling streptococcalpyogenic exotoxrns single radial immunodiffusron streptococcalsuperantigen signal transducer and activator of transcription transmembrane activator and calcium modulator and cyclophilin ligand [CAML] interactor transporter for antigen processing T-acute lymphoblastic leukemia tubercle bacillus cytotoxic T-cell thymus-derived lymphocyte T-cell factor

TCR1(2) TdT TG-A-L TGFB rh(t/2) THF Thp TLI TLR TM TNF TNP TPO Ti"g Ts TSAb TSE TSH(R) TSLP TSST fum-

T<ell receptor with /6 drains (with a/ p chains) terminal deoxynucleotidyl transferase polylysine with polyalanyl side-chains randomly tipped with fyrosine and glutamic acid hansforming growth f actor-p T-helper cell (subset 1 or 2) thymic humoral factor T-helper precursor total lymphoid irradiation Toll-like receptor transmembrane fumor necrosis factor trinitrophenol thrombopoietin regulatoryT-cell suppressor T-cell thyroid stimulating antibodies transmissible spongiform encephalopathy thyroid stimulating hormone (receptor) thymic stromal lymphopoietin toxic shock symdrome toxin strongly immunogenic mutant tumors

TUNEL

TdT-mediated dUTP (deoxyuridine triphosphate) nick end labeling

vs(F/v/6)

V*/r VCAM VEGF VIMP VLA VLP VNTR VP1

variable part of TCR o(F/y/6) chain variant Creutzfeldt-Jakob disease valosin-containing protein variable region gene for immunoglobulin T-cellreceptor variable part oflg heavy chain vasoactive intestinal peptide variable part oflight chain variable part of r(1,) light chain vascular cell adhesion molecule vascular endothelial cell growth factor VCP-interacting membrane protein verylateantigen virusJike particle variable number of tandem repeats virus-specific peptide 1

X-linked

zAP-70

zeta chain associatedprotein of 70 kDa

vCJD VCP Vgene VH VIP VL

Userguide

Throughout the illustrations standard forms have been used for commonlyoccurring cells and pathways. Akey to these is given in the figure below.

SMALLLYMPHOCYTE

(MO) I/ACROPHAGE

POLYMORPHONUCLEAR (POLYMORPH) LEUCOCYTE

Website:www.roitt.com Visit the companion website for this book at:

Includes: . Interactive MCQs with feedback

www.roitt.com

r Animations showing keyconcepts o DatabaseofFigures

Innoteimmunity

INTRODUCTION We live ina potentiallyhostileworld filledwith abewildering array of infectious agents (figure 1.1) of diverse shape, size, composition and subversive character which would very happily use us as rich sanctuaries for propagating their 'selfish genes/ had we not also developed a series of defense mechanisms at least their equal in effectiveness and ingenuity (except in the caseof many parasitic infections where the situation is best described as an uneasy and often unsatisfactory truce). It is these defense mechanisms which can establish a state of immunity against infection (Latin immunitas, freedom from) and whose operation provides the basis for the delightful subject called 'Immunology'. Aside from ill-understood constitutional factors which make one species innately susceptible and another resistant to certain infections, a number of relatively nonspecific antimicrobial systems (e.9. phagocytosis) have been recognized which are innate in the sense that they are not intrinsically affected by prior contact with the infectious agent. We shall discuss these systems and examine how, in the state of specific acquired immunity, their effectivenesscan be greatly increased.

EXTERNA BTA R R I E RASG A I N SITN F E C I I O N The simplest way to avoid infection is to prevent the microorganisms from gaining accessto the body (figure 1.2). The major line of defense is of course the skin which, when intact, is impermeable to most infectious agents;when there is skin loss, as for example in burns, infection becomes a major problem. Additionally, most bacteria fail to survive for long on the skinbecause of the

direct inhibitory effects of lactic acid and fatty acids in sweat and sebaceous secretions and the low pH which they generate. An exception is Stnphylococcusaureus which often infects the relatively vulnerable hair folliclesand glands. Mucus, secreted by the membranes lining the inner surfacesof the body, acts as a protective barrier to block the adherence of bacteria to epithelial cells. Microbial and other foreign particles trapped within the adhesive mucus are removed by mechanical stratagems such as ciliary movement, coughing and sneezing. Among other mechanical factors which help protect the epithelial surfaces, one should also include the washing action of tears, saliva and urine. Many of the secreted body fluids contain bactericidal components, such as acid in gastric juice, spermine and zinc in sem*n, lactoperoxidase in milk and lysozyme in tears, nasal secretions and saliva. A totally different mechanism is that of microbial antagonism associatedwiththe normalbacterial flora of the body. This suppresses the growth of many potentially pathogenic bacteria and fungi at superficial sites by competition for essential nutrients or by production of inhibitory substances.To give one example, pathogen invasion is limited by lactic acid produced by particular species of commensal bacteria which metabolize glycogen secreted by the vagin*l epithelium. \Alhen protective commensals are disturbed by antibiotics, susceptibility to opportunistic infectionsby Candidaand Clostridium dfficile is increased. Gut commensals may also produce colicins, a classof bactericidins whichbind to the negatively charged surface of susceptible bacteria and insert a hydrophobic helical hairpin into the membrane; the molecule then undergoes a'Jekyll and Hyde' transformation to become completely hydrophobic and forms a voltage-dependent channel in the membrane

CHAPTER I -INNATEIMMUNITY

SIZE (mm) '103

MYCOBACTERIUM STAPHYLOCOCCUT-l o,r^rrr$o 'cHLAMyDTA MYCOPLASMA

Cilio

Mucus

Metchnikoff at the turn of the last century microphages and macrophages.

Susceptible hoirtollicle Skinborriel

Figure 1.1. The formidablerange of infectious agents which confronts the immune system. Although not normally classified as such because of their lack of a cell wall, the mycoplasmas are included under bacteria for convenience. Fungi adopt many forms and approximate values for s o m eo f t h e s m a l l e s tf o r m s a r eg i v e n range of sizes observed fo!the organism(s) indicated by the arrow; <1, the organisms listed have the size denoted bv the arrow

Normolbocteriol microfloro

Figure 1.2, The first lines of defense against infection: protection at the external body surfaces.

which kills by destroying the cell's energy potential. Even at this level, survival is a tough game. If microorganisms do penetrate the body, two main defensive operations come into play, the destructive effect of soluble chemical factors such as bactericidal enzymes and the mechanism of phagocytosis -literally 'eating' by the cell (Milestone 1.1).

P H A G O C Y I I C E T T SK I t T M I C R O O R G A N I S M S Neutrophils ondmocrophoges orededicoted 'professionol' phogocyles The engulfment and digestion of microorganisms are assigned to two major cell types recognized by

The p oly morphonucle ar neutrophil This cell, the smaller of the two, shares a common hematopoietic stem cell precursor with the other formed elementsof theblood and is the dominantwhite cell in the bloodstream. It is a nondividing short-lived cell with a multilobed nucleus and an array of granules (figure 1.3),which are virtually unstained by histologic dyes such as hematoxylin and eosin, unlike those structures in the closely related eosinophil and basophil (figure 1.4). These neutrophil granules are of two main types: (i) the primary azurophil granule which develops early (figure 1.4 e), has the typical lysosomal morphology and contains myeloperoxidase together with most of the nonoxidative antimicrobial effectors including defensins, bactericidal permeability increasing (BPI) protein and cathepsin G (figure 1.3), and (ii) the peroxidase-negative secondary specific granules containing lactoferrin, much of the Iysozyrne, alkaline phosphatase (figure 1.4 d) and membranebound cytochrome buu,(figure 1.3).The abundant glycogen storescanbe utilizedby glycolysis enabling the cells to function under anerobic conditions. Themacrophage These cells derive from bone marrow promonocytes which, after differentiation to blood monocytes, finally settle in the tissues as mature macrophages where they

CHAPTER I - I N N A T EI M M U N I T Y

The perceptive Russian zoologist, Elie Metchnikoff (18451916), recognized that certain specialized cells mediate defense against microbial infections, so fathering the whole concept of cellular immunity. He was intrigued by the motile cells of transparent starfish larvae and made the critical observation that, a few hours after the introduction of a rose thorn into these larvae, they became surrounded by these motile cells. A year later, in 1883,he observed that fungal spores can be attacked by the blood cells of Daphnia, a tiny metazoan which, also being transparent, can be studied directly under the microscope. He went on to extend his investigations to mammalian leukocytes, showing their ability to engulf microorganisms, a processwhich he termed phagocytosis. Becausehe found this process to be even more effective in animals recovering from infection, he came to a somewhat polarized view thatphagocytosis provided the main, if notthe only, defense against infection He went on to define the existence of two types of circulating phagocytes: the polymorphonuclear leukocyte, which he termed a'microphage', and the larger'macrophage'.

Figure M1.1.1. Caricature of Professor Metchnikoff fuorn Chanteclair,1908,No.4, p. 7 (Reproduction kindly provided by The Wellcome Institute Library London.)

Figure M1.1.2. Reproductions of some of the illushations in Metchnikoff's book, Cotnparatiae Pathology of Inflnmmation (1893). (a) Four leukocytes from the frog, enclosing anthrax bacilli; some are alive and unstained, others which have been killed have taken up the vesuvine dye and have been colored; (b) drawing of an anthrax bacilIus, stained by vesuvine, in a leukocyte of the frog; the two figures represent two phases of movement of the same frog leukocyte which

contains stained anthrax bacilli within its phagocytic vacuole; (c and d) a foreign body (colored) in a starfish larva surrounded by phago' cytes which have fused to form a multinucleate plasmodium shown athigherpower in (d); (e) this gives afeelfor the d1'namic attractionof the mobile mesenchymal phagocytes to a foreign intruder within a starfish larva.

I - I N N A T EI M M U N I T Y CHAPTER

Polternrecognilionleceplor (PRRs)on phogocyliccells tecognizeondore oclivoledby polhogen-0ss0cioled moleculorpollerns(PAMPS)

Figure 1.3. Ultrastructure of neutrophil. The multilobed nucleus and two main types of cytoplasmic granules are well displayed (Courtesy ofDrD Mcl,aren.)

constitute the mononuclear phagocyte system (figure 1.5). They are present throughout the connective tissue and around the basem*nt membrane of small blood vessels and are particularly concentrated in the lung (figure 1.4h; alveolar macrophages), liver (Kupffer cells) and lining of spleen sinusoids and lymph node medullary sinuseswhere they are strategicallyplaced to filter off foreign material. Other examples are mesangial cells in the kidney glomerulus, brain microglia and osteoclasts in bone. Unlike the polymorphs, they are long-lived cells with significant rough-surfaced endoplasmic reticulum and mitochondria (figure 1.8b) and, whereas the polymorphs provide the major defense against pyogenic (pus-forming) bacteria, as a rough generalization it may be said that macrophages are at their best in combating those bacteria (figure 1.4g), viruses and protozoa which are capable of living within the cells of the host.

It hardly needs to be said but the body provides a very complicated internal environment and the phagocytes continuously encounter an extraordinary variety of different cells and soluble molecules. They must have mechanisms to enable them to distinguish these friendly self components from unfriendly and potentially dangerous microbial agents-as Charlie ]aneway so aptly put it, they should be able to discriminate '. between'noninfectious self and infectious nonself Not only must the infection be recognized but it must also generate a signal which, as proposed by Polly Matzinger, betokens'danger'. In the interests of host survival, phagocytic cells have evolved a system of receptors capable of recognizing molecular patterns expressed by pathogens (PAMPs) which are conserved (i.e. unlikely to mutate), shared by a large group of infectious agents (sparing the need for too many receptors) and clearly distinguishable from self patterns. Several of these pattern recognition receptors (PRRs) are lectin-like and bind multivalently with considerable specificity to exposed microbial surface sugars with their characteristic rigid threedimensional geometric configurations (PAMPs). They do notbind appreciably to the array of galactoseor sialic acid groups which are commonly the penultimate and ultimate sugars which decorate mammalian surface polysaccharides so providing the molecular basis for discriminating between self and nonself microbial cells. Amajor subsetof thesePRRsbelongto the classof socalled Toll-like receptors (TLRs) becauseof their similarity to the ToIl receptor in the fruit fly, Drosophila, which in the adult triggers an intracellular cascadegenerating the expression of antimicrobial peptides in response to microbial infection. A series of cell surface TLRs acting as sensors for extracellular infections have been identified (table 1.1)which are activated by microbial elements such as peptidoglycan, lipoproteins, mycobacterial lipoarabinomannan, yeast zymosan and flagellin. Phagocytes also display another set of PRRs, the cell bound C-type (calcium-dependent) lectins, of which the macrophage mannose receptor is an example. These transmembrane proteins possess multiple carbohydrate recognition domains whose engagement with their cognate microbial PAMPs generates an intracellular activation signal. Scavengerreceptors representyet a further classof phagocytic receptorswhichrecognize a variety of anionic polymers and acetylated low density

CHAPTER I -INNATEIMMUNITY

Figure 1.4. Cells involved in innate immunity. (a) Monocyte, 'horseshoe-shaped' showing nucleus and moderateiy abundant pale cytoplasm Note the three multilobed polymorphonuciear neutrophils and the small lymphocyte (bottom left). Romanowsky stain (b) Two monocytes stained for nonspecific esterase with cr-naphthyl acetate. Note the vacuolated cytoplasm. The smallcellwithfocal staining atthe top is a TJymphocyte. (c) Four polymorphonuclear neutrophils and one eosinophil. The multilobed nuclei and the cytoplasmic granules are clearly shown, those of the eosinophil being heavily stained. (d) Polymorphonuclear neutrophil showing cytoplasmic granules stained for alkaline phosphatase. (e) Early neutrophils in bone marrow The primary azurophilic granules (PG), originally clustered near the nucleus, move towards the periphery where the neutrophilspecific granules are generated by the Golgi apparatus as the cell matures. The nucleus gradually becomes lobular (LN) Giemsa (f) Inflammatory cells from the site of a brain hemorrhage showing the large active macrophage in the center with phagocytosed red cells and prominent vacuoles To the right is a monocyte with horseshoe-shaped

nucleus and cytoplasmic bilirubin crystals (hematoidin). Several multilobed neutrophils are clearly delineated. Giemsa. (g) Macrophages in monolayer cultures after phagocytosis of mycobacteria (stained red) Carbol-Fuchsin counterstained with Malachite Green. (h) Numerous plump alveolar macrophages within air spacesin the lung (i) Basophil withheavily staining granules compared withaneutrophil (be1ow).( j) Mast cell from bone marrow. Round central nucleus surrounded by large darkly staining granules. Two small red cell precursors are shown at the bottom Romanowsky stain. (k) Tissue mast cells in skin stained with Toluidine Blue. The intracellular granules are metachromatic and stain reddish purple Note the clustering in relation to dermal capillaries. (The slides from which illustrations (a), (b), (d), (e), (f), (i) and (j) were reproduced were very kindly provided by Mr M. Watts of the Department of Haematology, Middlesex Hospital Medical School; (c) was kindly supplied by Professor J.J.Owen; (g) by Professors P Lydyard and G. Rook; (h) by Dr Meryl Griffiths; and (k) by Professor N. Woolf.)

I -INNATEIMMUNITY CHAPTER

In addition to activating phagocytosis, binding to PAMPs rapidly releases a group of diverse multifunctional host proteins which recruit and prime l\illCR0GLlA macrophages and dendritic cells for interaction with lymphocytes to initiate adaptive immune responses LY[/PH (which will be discussed in the following chapter) BLOOD NODEIUO IVONOCYTE through differentiation of immature dendritic cells and PRECURSORS upregulation of critical costimulatory molecules B7.1 ALVEOLAR M CHRONIC and 87 .2 (cf . p. 172). These potent immunostimulants, INFLA[/MATION: OSTEOCLASTS ACTIVATED MQ including defensins and cathelicidin, which are antimiPLEURAL EPITHELIOID & crobial in their own right, serve as early-warning signals CAVITY to alert innate and adaptive immune responses. SPLENIC KUPFFER CELLS Programed cell death (apoptosis; see below) is an SYNOVIAL RESIDENT essential component of embryonic development and CONNECTIVE GLOI\4ERULAR the maintenance of the normal physiologic state. The TISSUE [/ESANGIAL HISTIOCYTES dead cellsneed to be removed by phagocytosisbut since CELLS 'danger' this must be done they do not herald any silently without setting off the alarm bells. Accordingly, Figure 1.5. The mononuclear phagocyte system. Promonocyte precursors in the bone marrow develop into circulating blood monocytes recognition of apoptotic cells by macrophages directly which eventually become distributed throughout the body as mature through the CDl4 receptor and indirectly through the macrophages (MQ) as shown. The other major phagocytic cell, the binding of C1q to surfacenucleosome blebs (seep. a35) polymorphonuclearneutrophil, is largely confined to the bloodstream proceeds without provoking the release of proinflamexcept when recruited into sites of acute inflammation. matory mediators. In sharp contrast, cells which are injured by infection and become necrotic release endogenous heat-shock protein 60 which acts as a proteins. The role of the CD14 scavenger molecule in the danger signal to the phagocytic cells and establishes a handling of Cram-negative LPS (endotoxin) merits protective infl ammatory response. some attention, since failure to do so can result in septic shock. The biologically reactive lipid A moiety of LPS is Microbesore engulfedby qclivotedphogocyliccells recognized by a plasma LPS-binding protein, and the complex which is captured by the CD14 scavengermolAfter adherence of the microbe to the surface of the neuecule on the phagocytic cell then activates TLR4. Like trophil or macrophage through recognition of a PAMP (figure 7.6.2),the resulting signal (figure 1.6.3)initiates the engagement of the other cell surface TLRs with their cognate microbial PAMPs which alert the cell to the ingestion phase by activating an actin-myosin condanger and initiate the phagocytic process, this in turn tractile system which extends pseudopods around the unleashes a series of events culminating in the releaseof particle (figures L.6.4 and 1.7); as adjacent receptors NFrB from its inhibitor. The free NFrB translocates to sequentially attach to the surface of the microbe, the the nucleus, and together with interferon regulatory plasma membrane is pulled around the particle just like transcription factors, triggers phagocytosis accompaa'zipper' until it is completely enclosed in a vacuole (phagosome; figures 1.6.5 and 1.8). Events are now nied by the releaseof proinflammatory mediators (table 1.1).Theseinclude the anti-viral interferons (cf. p.275), moving smartly and, within l minute, the cytoplasmic the small protein cytokines interleukin-1B (IL-1B),IL-6, granules fuse with the phagosome and discharge their lL-72 and TNF (TNFa) (cf. p. 186) which activate contents around the imprisoned microorganism '1..6.7 (figures other cells through binding to specific receptors, and and 1.8)which is subject to a formidable chemokines such as IL-8 which represent a subset of battery of microbicidal mechanisms. chemoattractant cytokines. Turning now to the sensing of infectious agents that Thereis on qrroyof killingmechonisms have succeeded in gaining accessto the interior of a cell, microbial nucleotide breakdown products can be recogKiIIing by reactia e oxy gen intermediates nized by the so-called NOD proteins and the typical Trouble starts for the invader from the moment phagoCpG DNA motif binds to the endosomal TLR9. Other cytosis is initiated. There is a dramatic increase in activendosomal Toll-like receptors, TLR3 and TLRT /8, are ity of the hexose monophosphate shunt generating responsive to intracellular viral RNA sequences. reduced nicotinamide-adenine-dinucleotide phosphate

CHAPTER I - I N N A T EI M M U N I T Y

Table 1.1' Microbial PAMP'danger signals'activate macrophages and dendritic cells through Toll-like receptors. Engagement of the PAMP cornplex with the cellular receptor stimulates intracellular reaction sequencesleading to activation ofNFrB and other transcription factors. The effector molecules induce phagocytosis and recruit and prime antigen-presenting cells for initiation of adaptive immune responses (cf. Chapter 2).

Cellsurtoce TLRI TLRIfiLR2 TLR2fiLR6

peptidoglycon Grom+ve Lipoproteins

Inflommolory cyfokines

Mycobocteriol lipoorobinomonnon Yeost zymoson

lRF, interferon regulotory lronscription foctor

Ghemofoxis

Figure 1.6. Phagocytosis and killing of a bacterium. Stage 3/4, respiratory burst and activation of NADPH oxidase; stage 5, damage by reactive oxygen intermediates; slage 6/7, damaee by peroxidase, cationic proteins, antibiotic peptide defensins, lysozyme and lactoferrin

Phogosome lormollon

Adhe.ence lhrough PAiIPrecognlllon

Membrone oclivolion fhrough'donge/ slgnol

lniliofionof phogocyfosis

Fusion

Killingond digeslion

Releoseof degrodoflon producls

I - I N N A T EI M M U N I T Y CHAPTER

Figure 1.7. Adherence and phagocytosis. (a) PhagocytosisoI Candidaalbicansbya polymorphonuclear leukocyte (neutrophil)

a{ {

(b)

Adherence to the yeastwall surfacemannan rnitrates enclosure of the fungal particle within arms of cytoplasm Lysosomal granules are abundant but mrtochondria are rare (x15 000) (b) Phagocytosisof C albicans by a monocyte showing near completion of phagosome formation (arrowed) around one organism and complete ingestion of two others (x5000) (Courtesyof DrH Valdimarsson )

Figure 1.8. Phagolysosome formation. (a) Neutrophil 30 mrnutes after ingestron of C, albicons The cytoplasm is already partly degranulated and two lysosomal granules (arrowed) are fusing with the phagocytic vacuole. Two lobes of the nucleus are evident (x5000) (b) Higher magnification of (a) showing fusing granules discharging their contents into the phagocytic vacuole (arrowed) (x33 000) (Courtesy of Dr H (a)

Valdimarsson.)

(NADPH). Electronspass from the NADPH to a flavine adenine dinucleotide (FAD)-containing membrane flavoprotein and thence to a unique plasma membrane cytochrome (cyt b,,,).This has the very low midpoint redox potential of -245 mV which allows it to reduce molecular oxygen directly to superoxide anion (figure 1.9a).Thus the key reaction catalyzed by this NADPH oxidase, which initiates the formation of reactive oxygen intermediates (ROI), is: NADPH+O,

%45 NADP+ + O;

(superoxide anion)

The superoxide anion undergoes conversion to hydrogen peroxide under the influence of superoxide dismutase,and subsequently to hydroxyl radicals (.OH). Each of these products has remarkable chemical reactivity with a wide range of molecular targets, making them formidable microbicidal agents;.OH in particular is one

of the most reactive free radicals known. Furthermore, the combination of peroxide, myeloperoxidase and halide ions constitutes a potent halogenating system capableof killing both bacteria and viruses (figure 1.9a). Although HrO, and the halogenated compounds are not as active asthe free radicals,they aremore stableand therefore diffuse further, making them toxic to microorganisms in the extracellular vicinity. Killing by reectia e nitro gen interme di qt es Nitric oxide surfaced prominently as a physiologic mediator when it was shown to be identical with endothelium-derived relaxing factor.This has proved to be just one of its many roles (including the mediation of penile erection, would you believe it!), but of major interest in the present context is its formation by an inducible NO. synthase (iNOS) within most cells, but

CHAPTER I - I N N A T EI M M U N I T Y particularly macrophages and human neutrophils, thereby generating a powerful antimicrobial system (figure 1.9b).Whereasthe NADPH oxidase is dedicated to the killing of extracellular organisms taken up by phagocytosis and cornered within the phagocytic vacuole, the NO. mechanism can operate against microbes which invade the cytosol; so, it is not surprising that the majority of nonphagocytic cells which may be infected by viruses and other parasitesare endowed with an iNOS capability. The mechanism of action may be through degradation of the Fe-S prosthetic groups of certain electron transport enzymes, depletion of iron and production of toxic .ONOO radicals. The N-ramp gene linked with resistance to microbes such as bacille Calmette-Gu6rin (BCG), Salmonella and Leishmania, which can live within an intracellular habitat, is now known to express a protein forming a transmembrane channel which may be involved in transporting NO. acrosslysosome membranes. KiIIing by preformed antimitobials ( figure 1.9c) These molecules, contained within the neutrophil granules, contact the ingested microorganism when fusion with the phagosome occurs. The dismutation of superoxide consumes hydrogen ions and raises the pH of the vacuole gently, so allowing the family of cationic proteins and peptides to function optimally. The latter, known as defensins, are approximately 3.5-4 kDa and invariably rich in arginine, and reach incredibly high concentrations within the phagosome, of the order of

20-100 r g/ml Like the bacterial colicins described above, they have an amphipathic structure which allows them to insert into microbial membranes to form destabilizing voltage-regulated ion channels (who copied whom?). Theseantibiotic peptides/ at concentrations of 10-100pg/ml, act as disinfectants against a wide spectrum of Gram-positive and -negativebacteria, many fungi and a number of enveloped viruses. Many exhibit remarkable selectivity for prokaryotic and eukaryotic microbes relative to host cells, partly dependent upon differential membrane lipid composition. One must be impressed by the ability of this surprisingly simple tool to discriminate large classesof nonself cells.i.e. microbes,from self.

REAcTTVE oxycEnmrERnEDtATEs

NADPH

NADP+<-

r{rRrc oxrDE

Y Figure 1.9. Microbicidal mechanisms of phagocytic cells. (a) Production of reactive oxygen intermediates Electrons from NADPH are transferred by the flavocytochrome oxidase enzyme to molecular oxygen to form the microbicidal molecular species shown in the orange boxes (For the more studious The phagocytosis triggering agent binds to a classic G-protein-linked seven transmembrane domain receptor which activates an intracellular guanosine triphosphate (GTP)-binding protein This in tum activates an array of enzymes: phosphoinositol-3 kinase concerned in the cytoskeletal reorganization underlying chemotactic responses (p 10), phospholipaseC/2 mediating events leading to lysosome degranulation and phosphorylation of p47 phox through activation of protein kinase C, and the MEK and MAP kinase systems (cf figure 8 7) which oversee the assembly of the NADPH oxidase. This is composed of the membrane cytochrome brrr, consisting of a p21 heme protein linked to gp91 with binding sites for NADPH and FAD on its intracellular aspect, to which phosphorylated p47 and p67 translocate from the cytosol on activation of the oxidase.) (b) Generation of nitric oxide The enzyme, which structurally resembles the NADPH oxidase, can be inhibited by the arginine analog N-monomethyl-r--arginine (I--NMMA). The combination of NO with superoxide anion yields the highly toxic peroxynitrite radical .ONOO which cJ.eaveson protonation to form reactive OH and NO, molecules. NO can form rnononuclear iron dithioldinitroso complexes leading to iron depletion and inhibition of several enzymes (c) The basis of oxygen-independent antimicrobial systems

NOSYNTHASE

Fe(RS)2(N0)2

FelRSH CTTRULLTNE L-NMMA

MEcHAlusrils oxyGEl{-r1{DEpENDE}tr

Colheosin G Lowmol,wt defensins Highmol.wl cotionicproleins permeobilily Boctericidol prolein (BPl) increosing

Domoge to microbiol membrones

Lysozyme

in Splitsmucopepfide bocferiol cellwoll

Locfoferrin

withiron Complex

Proteolyfic enzymes Vorielyof other hydrolyfic enzymes

Digestion of killed orgonrsms

I to

E o

IUU

CHAPTER I - I N N A T EI M M U N I T Y

Loctolerrin Lysozyme pluslocloferrin Lysozyme

e =5- no

E co

2 Time(hours) Figure 1.10. The synergistic microbicidal action of lysozyme and lactoferrin. (Reproduced with permission from Singh PK et ol (2000) Anerican Journnl of Physiology279,L799 L805.)

As if this was not enough, further damage is inflicted on the bacterial membranes both by neutral proteinase (cathepsinG) action and by direct transfer to the microbial surfaceof BPI, which increasesbacterial permeability. Low pH, lysozyme and lactoferrin constitute bactericidal or bacteriostatic factors which are oxygen independent and can function under anerobic circ*mstances.Interestingly, lysozyme and lactoferrin are synergistic in their action (figure 1.10).Finally, the killed organisms are digested by hydrolytic enzymes and the degradation products released to the exterior (figure 1.6.8). By now, the reader may be excused a little smugness as sheor he sheltersbehind the impressive antimicrobial potential of the phagocytic cells. But there are snags to consider; our formidable array of weaponry is useless unless the phagocyte can: (i) 'home onto'the microorganism, (ii) adhere to it, and (iii) respond by the membrane activation which initiates engulfment. Some bacteria do produce chemical substances, such as the peptide formyl.Met.Leu.Phe, which directionally attract leukocytes, a process known as chemotaxis; many organisms do adhere to the phagocyte surface and many do spontaneously provide the appropriate membrane initiation signal. However, our teeming microbial adversaries are continually mutating to produce new specieswhich may outwit the defensesby doing none of these. What then? The body has solved these problems with the effortless easethat comes with a few million years of evolution by developing the complement system.

C O M P T EEM N TF A CTI I I A T E P SHAGOCYTOSIS Gomplemenl onditsoclivotion Complement is the name given to a complex series of

some 20 proteins which, along withblood clotting, fibrinolysis and kinin formation, forms one of the triggered enzyme systems found in plasma. These systems characteristically produce a rapid, highly amplified response to a trigger stimulus mediated by a cascade phenomenon where the product of one reaction is the enzymic catalyst of the next. Some of the complement components are designated 'C' by the letter followed by a number which is related more to the chronology of its discovery than to its position in the reaction sequence.The most abundant and the mostpivotalcomponentis C3 whichhas amolecular weight of 195kDa and is present in plasma at a concentration of around 1.2mg/ml. C3 undergoesslow spontaneous cleeaqge Under normal circ*mstances, an internal thiolester bond in C3 (figure 1.11) becomes activated spontaneously at a very slow rate, either through reaction with water or with trace amounts of a plasma proteolytic erl.zyrrre,to form a reactive intermediate, either the split product C3b, or a functionally similar molecule designated C3i or C3(HzO).ln the presenceof Mg2' this can complex with another complement component, factor B, which then undergoes cleavageby a normal plasma enzyme (factor D) to generate C3bBb.Note that, conventionally, a bar over a complex denotes enzymic activity and that, on cleavageof a complement component, the larger product is generally given the suffix 'b'and the smaller'a'. C3bBb has an important new enzymic activity: it is a C3 convertase which can split C3 to give C3a and C3b. We will shortly discuss the important biological consequencesof C3 cleavagein relation to microbial defenses, but under normal conditions there must be some mech'tick-over' Ievel anism to restrain this processto a sinceit rise more that is, we are can also give to C3bBb, dealing with a potentially runaway positive-feedback loop (figure 1.12).As with all potentially explosive triggered cascades,there are powerful regulatory mechanisms. C3b leaels arenormally tightly controlled In solution, the C3bBbconvertaseis unstable and factor B is readily displaced by another component/ factor H, to form C3bH which is susceptibleto attack by the C3b inactivator, factor I (figure 7.72; further discussed on p. 314). The inactivated iC3b is biologically inactive and undergoes further degradation by proteasesin the body fluids. Other regulatory mechanisms are discussed at a Iaterstage(seep. 314). C3 conaertase is stabilized on microbial surfaces A number of microorganisms can activate the C3bBb

ill

CHAPTER I - I N N A T EI M M U N I T Y

C3o

c3b

ic3b

c3f

C3c

C3dg

H0(NHr)

sbt

CELLSURFACE

Figure1.11. StructuralbasisforthecleavageofC3byC3convertaseanditscovalentbindingto OHor NHrgroupsatthecellsurfacethroughexposure of the internal thiolester bonds. Further cleavage leaves the progressively smaller fragments, C3dg and C3d, attached to the membrane. (Based e s s e n t i a l l y o n L a w S . H A & R e i d K B . M ( 1 9 8 8 )C o n r p l e m e n t , f i g u r e 2 I4R L P r e s s , O x f o r d . )

SURFACE PROTECTED MICROBIAL

Figure 1.12. Microbial activation of the alternative complement pathway by stabilization of the C3 convertase (C3bBb), and its control by factors H and I. When bound to the surface of a host cell or in the fluid phase, the C3b in the convertase is said to be 'unprotected'in that its affinity for factor H is much greater than for factor B and is therefore susceptible to breakdown by factors H and I On a microbial surface, C3b binds factor B more strongly than factor H and is therefore'prot e c t c d 'f r o m o r ' s t a b i l i z e d ' a g a i n s ct l e av age-even more so when subsequentlv bound by properdin Although in phylogenetic terms this is the oldest complement pathway, it was discor.ered after a separate pathway to be discussed in the next chapter, and so has the confusing designatron'alternative' .,'+ representsan activation process The horizontal bar above a component desisnates its activation

HOST ORFLUIDPHA.SE CELLSURFACE UNPROTECTED

I t,

I - I N N A T EI M M U N I T Y CHAPTER

convertase to generate large amounts of C3 cleavage products by stabilizing the enzyme on their (carbohydrate) surfaces, thereby protecting the C3b from factor H. Another protein, properdin, acts subsequently on this bound convertase to stabilize it even further. As C3 is split by the surface membrane-bound enzyme to nascent C3b, it undergoes conformational change and its potentially reactive internal thiolester bond becomes exposed. Since the halflife of nascent C3b is less than 100 psec, it can only diffuse a short distance before reacting covalently with local hydroxyl or amino groups available at the microbial cell surface (figure 1.11).Each catalytic site thereby leads to the clustering of large numbers of C3b molecules on the microorganism. This series of reactions leading to C3 breakdown provoked directly by microbes has been called the alternative pathway of complement activation (figure 1.12). The p o st- C3 p athw ay generates a membr ane attack complex Recruitment of a further C3b molecule into the C3bBb enzymic complex generates a C5 convertase which activates C5 by proteolytic cleavage releasing a small polypeptide, C5a, and leaving the large CSb fragment loosely bound to C3b. Sequential attachment of C6 andCT to C5b forms a complex with a transient membrane-binding site and an affinity for the B-peptide chain of C8. The C8a chain sits in the membrane and directs the conformational changes in C9 which transform it into an amphipathic molecule capable of insertion into the lipid bilayer (cf. the colicins, p. 1) and polymerization to an annular membrane attack complex (MAC; figures 1.1,3and2.4).This forms a transmembrane channel fully permeable to electrolytes and water, and due to the high internal colloid osmotic pressure of cells, there is a net influx of Na+ and water frequently leading to lysis.

hoso rongeof defensive biologicolfunclions Gomplemenl Thesecanbe grouped convenientlyunderthree headings.

Figure 1.13. Post-C3 pathway generating C5a and the C5b-9 membrane attack complex (MAC). (a) Cartoon of moiecular assembly. The conformational change in C9 protein structure which converts it from a hydrophilic to an amphipathic molecule (bearing both hydrophobic and hydrophilic regions) can be interrupted by an antibody raised against linear peptides derived from C9; since the antibody does not react with the soluble or membrane-bound forms of the molecule, it must be detecting an intermediate structure transiently revealed in a deep-seated structuraL rearrangement. (b) Electron micrograph of a membrane C5b-9 complex incorporated into liposomal membranes clearly showing the annular structure The cylindrical complex is seen from the side inserted into the membrane of the liposome on the left, and end-on in that on the right Although in itself a rather splendid structure, formation of the annular C9 cylinder is probably not essential for cytotoxic perturbation of the target cell membrane, since this can be achieved by insertion of amphipathic C9 molecules in numbers too few to form a clearly defined MAC (Courtesy of Professor J. Tlanum-Jensen and Dr S. Bhakdi )

1 C3b adheresto complement receptors Phagocytic cells have receptors for C3b (CR1) and iC3b (CR3) which facilitate the adherence of C3b-coated microorganisms to the cell surface (discussed more fully onp.265). 2 Biologically actioe fragments are released C3a and C5a, the small peptides split from the parent molecules during complement activation, have several

important actions. Both act directly on phagocytes, especially neutrophils, to stimulate the respiratory burst associated with the production of reactive oxygen intermediates and to enhance the expression of surface receptors for C3b and iC3b. Also, both are anaphylatoxins in that they are capable of triggering mediator

CHAPTER I - I N N A T EI M M U N I T Y

13 |

produce vasodilatation and increased permeability, an effect which seems to be prolonged by leukotriene Bn released from activated mast cells, neutrophils and macroPhages. 3 The terminal complex can induce membrane lesions As described above, the insertion of the membrane attack complex into a membrane may bring about cell lysis. Providentially, complement is relatively inefficient atlysing the cell membranes of the autologous host due to the presence of control proteins (cf. p . 374).

A NA C U T E C O M P T E M EC NA T NM E D I A I E INFTAMMAIORE YACIION We can now put together an effectively orchestrated defensive scenarioinitiated by activation of the alternative complement pathway (figure 1.16). In the first act, C3bBb is stabilized on the surface of the microbe and cleaveslarge amounts of C3. The C3a fragment is released but C3b molecules bind copiously to the microbe. Theseactivatethe next step in the sequence to generate CSa and the membrane attack complex (although many organisms will be resistant to its action). Themoslcell ployso centrolrole

Figure 1.14. The mast cell. (a) A resting cell with many membranebound granules containing preformed mediators. (b)A triggered mast cell Note that the granules have releascdtheir contents and are morphologically altered, beinp;larger and less electron dense Although most of the altered granules remain within the circumference of the cell, they are open to the extracellular space (Electron micrographs x5400 ) (Courtesy of Drs D. Lawson, C Fewtrell, B Gomperts and M C Raff from (1975)JournolofExpcrnnentalMedicittcl42,39l )

releasefrom mast cells (figures 1.4k and 1.14)and their circulating counterpart, the basophil (figure 7.41),a phenomenon of such relevance to our present discussion that we have presented details of the mediators and their actions in figure 1.15;note in particular the chemotactic properties of these mediators and their effects on blood vessels. In its own right, C3a is a chemoattractant for eosinophils whilst CSa is a potent neutrophil chemotactic agent and also has a striking ability to act directly on the capillary endothelium to

The next act seesC3a and C5a, together with the mediators they trigger from the mast cell, acting to recruit polymorphonuclear phagocytes and further plasma complement components to the site of microbial invasion. The relaxation induced in arteriolar walls causes increasedblood flow and dilatation of the small vessels, while contraction of capillary endothelial cells allows exudation of plasma proteins. Under the influence of the chemotaxins, neutrophils slow down and the surface adhesion moleculesthey are stimulated to expresscause them to marginate to the walls of the capillaries where they pass through gaps between the endothelial cells (diapedesis)and move up the concentration gradient of chemotactic factors until they come face to face with the C3b-coated microbe. Adherence to the neutrophil C3b receptors then takes place, C3a and CSa at relatively high concentrations in the chemotactic gradient activate the respiratory burst and, hey presto, the slaughter of the last act canbegin! The processesof capillary dilatation (redness),exudation of plasma proteins and also of fluid (edema) due to hydrostatic and osmotic pressurechanges,and accumulation of neutrophils are collectively termed the acute inflammatory response.

I to

I - I N N A T EI M M U N I T Y CHAPTER

(i)

NEUTMLPROTEASES B-GLUCOSAMINIDASE

PLATELET ACTIVATING FACTOR I N T E R L E U K I N S 3 , 4 , 5 & 6 Multiple, including mocrophoge GM-CSF, TNF octivoii0n, lrigger0cufephoseproteins, elc.(cf.Chopter 9) NEWTY SY}ITHESIZED LEUKoTRTENES C4,D4(SRS-A), 84

PROSTAGLANDINS THROMBOXANES

Affeclbronehiol muscle,plqtelel oggregotion ondvosodilof olion

Mocrophoges conolsodo if Although not yet establishedwith the same confidence that surrounds the role of the mast cell in acute inflammation, the concept seemstobe emerging that the tissue macrophage may mediate a parallel series of events with the same final end result. Nonspecific phagocytic events and certain bacterial toxins such as the lipopolysaccharides (LPSs) can activate macrophages, but the phagocytosis of C3b-opsonized microbes and the direct action of C5a generated through complement activation are guaranteed to goad the cell into copious

Figure 1.15. Mast cell triggering leading to releaseof mediators by two major pathways: (i) releaseof preformed mediators present i.n the granules,and (ii) the metabolism of arachidonic acid produced throup;h^actrvation of a phospholipase.Intracellular Ca'- and cyclic AMP are central to the initiation of these events but details are still unclear Mast cell triggering may occur through C3a, C5a and even by some microorganisms which can act directly on cell surfacereceptors Mast cell heterogeneityis discussedon p 337 ECF, eosinophil chemotactic factor; CM-CSF, granulocyte macrophage colony-stimulating factor; NCF, neutrophil chemotacticfactor Chemotaxis refers to directed migration of granulocytes up the pathway concentration sradient of the mediator.

secretion of soluble mediators of the acute inflammatory response(figure 1.17). These upregulate the expression of adhesion molecules for neutrophils on the surface of endothelial cells, increase capillary permeability and promote the chemotaxis and activation of the polymorphonuclear neutrophils themselves. Thus, under the stimulus of complement activation, the macrophage provides a pattern of cellular events which reinforcesthe mast cellmediated pathway leading to acute inflammation-yet another of the body's fail-safe redundancy systems (often known as the 'belt and braces'principle).

CHAPTER I - I N N A T EI M M U N I T Y

Figure 1.16. The defensive strategy of the acute inflammatory reaction initiated by bacterial activation of the alternative C pathway. Directions: rOstart with the activation of the C3bBb C3 convertase by the bacterium, @notice the generation of C3b (@ which binds to the bacterium), C3a and C5a, @ whrch recruit mast cell mediators; @ follow their effect on capillary dilatation and exudation of plasma proteins and @ their chemotactic attraction of neutrophils to the C3b-coated bacterium and triumph in @ the adherenceand final activation of neutrophils for the kill.

Figure 1.17. Stimulation by complement components and bacterial toxins such as LPS induces macrophage secretion of mediators of an acute inflammatory response.Blood neutrophils stick to the adhesion molecules on the endothelial cell and use this to provide traction as they force their way between the cells, through the basem*nt membrane (with the help of secreted elastase)and up the chemotactic sradient

15 |

l16

I -INNATEIMMUNITY CHAPTER

H U M O R AMTE C H A N I S MPSR O V I DAES E C O N D DEFENSIV SE IRATEGY Microbicid0l foclors inseclelions Tirrning now to those defense systems which are mediated entirely by soluble factors, we recollect that many microbes activate the complement system and may be lysed by the insertion of the membrane attack complex. The spread of infection may be limited by enzymes released through tissue injury which activate the clotting system. Of the soluble bactericidal substanceselaborated by the body, perhaps the most abundant and widespread is the enzyme lysozyme, a muramidase which splits the exposed peptidoglycan wall of susceptiblebacteria (cf. figure 12.5). Like the cr-defensins of the neutrophil granules, the human B-defensins are peptides derived by proteolytic cleavage from larger precursors; they have p-sheet structures, 2940 arnino acids and three intramolecular disulfide bonds, although they differ from the crdefensins in the placement of their six cysteines. The main human B-defensin, hDB-1, is produced abundantly in the kidney, the female reproductive tract, the oral gingiva and especially the lung airways. Since the word has it that we are all infected every day by tens of thousands of airborne bacteria, this must be an important defense mechanism. This being so, inhibition of hDB-1 and of a second pulmonary defensin, hDB-2,by high ionic strength could account for the susceptibility of cystic fibrosis patients to infection since they have an ion channel mutation which results in an elevated chloride concentration in airway surface fluids. Another airway antimicrobial active against Gram-negative and -positive bacteria is LL-37, a 37-residue cr-helical peptide releasedby proteolysis of a cathelicidin (cathepsin L-inhibitor) precursor. This theme surfaces again in the stomach where a peptide split from lactoferrin by pepsin could provide the gastric and intestinal secretions with some antimicrobial policing. A rather longer two-domain peptide with 108 residues, termed secretory leukoprotease inhibitor (SLPD, is found in many human secretions. The C-terminal domain is anti-protease but the Nterminal domain is distinctly unpleasant to metabolically active fungal cells and to various skin-associated microorganisms, which makes its production by human keratinocytes particularly appropriate. In passing, it is worth pointing out that many D-amino acid analogs of peptide antibiotics form left-handed helices which retain the ability to induce membrane ion channels and hence their antimicrobial powers and, given their resistance to catabolism within the body, should be attractive candidates for a new breed of synthetic antibiotics.

Lastly, we may mention the two lung surfactant proteins SP-A and SP-D which, in conjunction with various lipids, lower the surface tension of the epithelial lining cells of the lung to keep the airways patent. They belong to a totally different structural group of molecules termed collectins (seebelow) which contribute to innate immunity through binding of their lectin-like domains to carbohydrates on microbes, and their collagenous stem to cognate receptors on phagocytic cells-thereby facilitating the ingestion and killing of the infectious agents. Aculephoseproleinsincleosein responselo infeclion A number of plasma proteins collectively termed acute phase proteins show a dramatic increase in concentra'alarm' mediators such as tion in response to early macrophage-derived interleukin-1 (IL-1) released as a result of infection or tissue injury. These include Creactive protein (CRP), mannose-binding lectin (MBL) and serum amyloid P component (table 1.2).Other acute phase proteins showing a more modest rise in concentration include ar-antichymotrypsin, fibrinogen, ceruloplasmin, C9 and factor B. Overall, it seems likely that the acute phase response achieves a beneficial effect through enhancing host resistance, minimizing tissue injury and promoting the resolution and repair of the inflammatory lesion. To take an example, during an infection, microbial products such as endotoxins stimulate the release of

Table 1.2 Acute phaseproteins.

increoses Dromolic in concenlrolion: prolein C-reoctive

opsonizes Fixescomplemenl,

leclin Monnose binding 0r-ocidglycoprotein

opsonizes Fixescomplement, prolein Tronsport

P componenl omyloid Serum

precursor Amyloid comp0nenl

increoses Moderofe in concenirotion: inhibitors cr,-proteinose cxr-ontichymolrypsin

proleoses Inhibilbocleriol proleoses Inhibitbocleriol

C3,C9,toctorB

function complement Increose .0; scovenger

Ceruloplosmin Fibrinogen Angiolensin

Coogulotion Bloodpressure

Hoptoglobin

Bindhemoglobin

Fibroneclin

Celloftochmenl

C H A P T E RI - I N N A T E I M M U N I T Y IL-1, which is an endogenous pyrogen (incidentally capable of improving our general defenses by raising the body temperature), and IL-6. These in turn act on the liver to increase the synthesis and secretion of CRP to such an extent that its plasma concentration may rise 100O-fold. Human CRP is composed of five identical polypeptide units noncovalently arranged as a cyclic pentamer around a Ca-binding cavity. These protein pentraxins havebeen around in the animalkingdom for some time, since a closely related hom*olog,limulin, is present in the hemolymph of the horseshoe crab, not exactly a close relative of hom*o sapiens.A major property of CRP is its ability to bind in a Ca-dependent fashion, as a pattern recognition molecule, to a number of microorganisms which contain phosphorylcholine in their membranes, the complex having the useful property of activating complement (by the classical and not the alternative pathway with which we are at present familiar). This results in the deposition of C3b on the surface of the microbe which thus becomes opsonized (i.e. 'made ready for the table') for adherence to phagocytes. Yet another member of this pentameric family is the serum amyloid P (SAP) component. This protein can complex with chondroitin sulfate, a cell matrix glycosaminoglycan, and subsequently bind lysosomal enzymes such as cathepsin B released within a focus of inflammation. The degraded SAP becomes a component of the amyloid fibrillar deposits which accompany chronic infections-it might even be a key initiator of amyloid deposition (cf. p. 395). A most important acute phase opsonin is the Cadependent mannose-binding lectin (MBL) which can react not only with mannose but several other sugars, so enabling it to bind with an exceptionally wide variety of Gram-negative and -positive bacteria, yeasts, viruses and parasites; its subsequent ability to trigger the classical C3 convertase through two novel associated serine proteases(MASP-1 and MASP-2) is the basis of what is known as the lectin pathway of complementactivation. (Pleaserelax, we unravel the secretsof the classicaland lectin pathways in the next chapter.) MBL is a multiple of trimeric complexes, each unit of which contains a collagen-like region joined to a globular lectin-binding domain. This structure places it in the family of collectins (collagen + lectin) which have the ability to recognize'foreign' carbohydrate patterns differing from 'self ' surfacepolysaccharides,normally terminal galactose and sialic acid groups, whilst the collagen region can bind to and activate phagocytic cells through complementary receptors on their surface. The collectins, especially MBL and the alveolar surfactant molecules SP-A and SP-D mentioned earlier, have many attributes that qualify them for a first-line role in innate immunity.

1 7I

Figure 1.18. A major defensive strategy by soluble factors. The pattern recognition elements (PRRs) Iink the microorganism to a microbicidal system through the adaptor region. PAMP, pathogenassociated molecular pattern

These include the ability to differentiate self from nonself, to bind to a variety of microbes, to generate secondary effector mechanisms, and to be widely distributed throughout the body including mucosal secretions. They are of course the soluble counterparts to the cell surface C-type lectins and other pattern recognition receptors described earlier. Interest in the collectin conglutinin has perked up recently with the demonstration, first, that it is found in humans and notiust in cows, and second, thatit canbind to N-acetylglucosamine; being polyvalent, this implies an ability to coat bacteria with C3b by cross-linking the available sugar residue in the complement fragment with the bacterial proteoglycan. Although it is not clear whether conglutinin is a member of the acute phase protein family, we mention ithere becauseit embellishes the general idea that the evolution of lectin-like molecules which bind to microbial rather than self polysaccharides, and which can then hitch themselves to the complement system or to phagocytic cells, has proved to be such a useful form of protection for the host (figure 1.18). inhibilvir0lreplicoti0n Inlerferons These are a family of broad-spectrum antiviral agents present inbirds, reptiles and fishes as well as the higher animals, and first recognized by the phenomenon of viral interference in which an animal infected with one virus resists superinfection by a second unrelated virus. Different molecular forms of interferon have been identified, all of which have been gene cloned. There are at least 14 different c,-interferons (IFNa) produced by leukocytes, while fibroblasts, and probably all cell types, synthesize IFNB. We will keep a third type (IFNy), which is not directly induced by viruses, up our sleeves for the moment.

| 't

I -INNAIE IMMUNITY CHAPTER

Cells synthesize interferon when infected by a virus and secreteit into the extracellular fluid where it binds to specific receptors on uninfected neighboring cells. The bound interferon now exerts its antiviral effect in the followingway. At least two genesare thought to be derepressedin the interferon-treated cell allowing the synthesis of two new enzymes. The first, a protein kinase, catalyzes the phosphorylation of a ribosomal protein and an initiation factor necessary for protein synthesis, so greatly reducing mRNA translation. The other catalyzes the formation of a short polymer of adenylic acid which activatesa latent endonuclease;this in turn degradesboth viral and host mRNA. Whatever the precise mechanism of action ultimately proves to be, the net result is to establish a cordon of uninfectable cells around the site of virus infection so restraining its spread. The effectivenessof interferon ln aiao rnay be inferred from experiments in which mice injected with an antiserum to murine interferons could be killed by several hundred times less virus than was needed to kill the controls. However, it must be presumed that interferon plays a significant role in the recovery from, as distinct from the prevention of, viral infections. As a group, the interferons may prove to have a wider biological role than the control of viral infection. It will be clear, for example, that the induced enzymes described above would act to inhibit host cell division just as effectively as viral replication. The interferons may also modulate the activity of other cells,such as the natural killer cells, to be discussed in the following sectlon.

E X T R A C E t t U t AKR IttING (NK)cells Nofurol killer Viruses lack the apparatus for self renewal and so it is essential for them to penetrate the cells of the infected host in order to take over its replicative machinery. It is clearly in the interest of the host to find a way to kill such infected cellsbefore the virus has had a chanceto reproduce. NK cells appear to do just that when studied ln altro. They are large granular leukocytes (figure 2.6a) with a characteristic morphology (figure 2.7b). Killer and target are brought into close opposition (figure 1.19) through recognition by lectin-like (i.e. carbohydratebinding) and other receptors on the NK cell (c1.p.27) of structures on high molecular weight glycoproteins on the surfaceof virally infected cells.Activation of the NK cell ensues and leads to polarization of granules between nucleus and target within minutes and extra-

cellular releaseof their contents into the spacebetween the two cells followed by target cell death. One of the most importantof the granule components is a perforin or cytolysin bearing some structural hom*ology to C9; like that protein, but without any help other than from Caz*, it can insert itself into the membrane of the target, apparently by binding to phosphorylcholine through its central amphipathic domain. It then polymerizes to form a transmembrane pore with an annular structure, comparable to the complement membrane attack complex (figure 1.19).

TRIGGER

t I

/l n

o TNF structure -Perforin induced Virolly s Gronzyme v TNFreceplor receptor NKoclivoting

Figure 1.19. Extracellular killing of virally infected cell by natural kilter (NK) cell. Binding of the NK recePtors to the surface of the virally infected cell triggers the extracellular release of perforin molecules from the granules; these polymerize to form transmembrane channels which may facilitate lysis of the target by permitting entry of granzymes which induce apoptotic cell death through activation of the caspase protease cascade and ultimate fragmentation of nuclear DNA. (Model resembling that proposed by Hudig D , Ewoldt G R & Woodward S.L (i993) Current Opinion in Immunology 5,90.) Another granule component, TNR activates caspase-dependent aPoP'death domains' of the surface TNF receptors on the tosis through the target cell Engagement of the NK receptor also activates a parallel killing mechanism mediated through the binding of the FasJigand (FasL) on the effector to the target cell Fas receptor whose cytoplasmic death domains activate procaspase-8. Because apoptosis is such a fun'default' mechanism in every cel1,it is crucial for there to be damental heavy regulation: thus a large group of regulatory proteins, the Bcl-2 subfamily, inhibit apoptosis while the Bax and BH3 subfamilies 'apoptosis' in ancient Greek describes the falling promote it. The word trees or ofpetals from flowers and aptly illustrates apopfrom ofleaves tosis in ce1lswhere they detach from their extracellular matrix support structures. (Seefigure 11.8for morphological appearance of apoptotic cells )

CHAPTER I _ I N N A T EI M M U N I I Y

T0rgelcellsorefoldlo commilsuicide WhereasC9-induced cell lysis is brought about through damage to outer membranes followed later by nuclear changes, NK cells kill by activating apoptosis (programed cell death), a mechanism present in every cell which leads to self immolation. Apoptosis is mediated by a cascadeof proteolytic enzymes termed caspases. Like other multicomponent cascades,such as the blood clotting and complement systems, it depends upon the activation by proteolytic cleavageof aproenzyme next in the chain, and so on. The sequence terminates with very rapid nuclear fragmentation effected by a Cadependent endonuclease which acts on the vulnerable DNA between nucleosomes to produce the 200 kb 'nucleosome ladder' fragments; only afterwards can one detect releaseof slCr-labeled cytoplasmic proteins through defective cell surface membranes. These nuclear changesarenotproducedbyC9. Thus, although perforin and C9 appear to produce comparable mem'pores', brane there is a dramatic difference in their killing mechanisms. In addition to perforin, the granules contain tumor necrosis factor (TNFcx), lymphotoxin-B, IFNy and a family of serine proteases termed granzymes, one of which, granzyme B, can function as an NK cytotoxic factor by passing through the perforin membrane pore into the cytoplasm where it can split procaspase-8and activatethe apoptotic process.Tumor necrosisfactor can induce apoptotic cell death through reaction with cell surface TNF receptors whose cytoplasmic 'death domains' can also activate procaspase-8.Chondroitin sulfate A, a protease-resistant highly negatively charged proteoglycan present in the granules, may subserve the function of protecting the NK cell from autolysis by its own lethal agents. Killing by NK cells can still occur in perforin-deficient mice, probably through a parallel mechanism involving

re I

Fas receptor molecules on the target cell surface. Engagement of Fas by the so-called Fas-ligand (FasL) on the effector cell provides yet another pathway for the induction of an apoptotic signal in the unlucky target. The various interferons augment NK cytotoxicity and, since interferons are produced by virally infected cells, we have a nicely integrated feedback defense system.

Eosinophils Large parasites such as helminths cannot physically be phagocytosed and extracellular killing by eosinophils would seem to have evolved to help cope with this situ'cousins' of the neuation. These polymorphonuclear trophil have distinctive granules which stain avidly with acid dyes (figure 1.4c) and have a characteristic appearance in the electron microscope (figure 12.23).A major basic protein is localized in the core of the granules while an eosinophilic cationic protein together with a peroxidase have been identified in the granule matrix. Other enzymes include arylsulfatase B, phospholipase D and histaminase. They have surface receptors for C3b and on activation produce a particularly impressive respiratory burst with concomitant generation of active oxygen metabolites. Not satisfied with that, nature has also armed the cell with granule proteins capable of producing a transmembrane plug in the target membrane like C9 and the NK perforin. Quite a nasty cell. Most helminths can activate the alternative complement pathway, but although resistant to C9 attack, their coating with C3b allows adherence of eosinophils through their C3b receptors.If this contact should lead to activation, the eosinophil will launch its extracellular attack which includes the release of the major basic protein and especially the cationic protein which damagesthe parasite membrane.

Awide range of innate immune mechanisms operate which do not improve with repeated exposure to infection.

factors such as lysozyme andbyphagocytosis withintracellular digestion.

Borriers ogoinsl infeclion

Phogocylic cellsldllmicroorgonisms

o Microorganisms are kept out of the body by the skin, the secretion of mucus, ciliary action, the lavaging action of bactericidal fluids (e.g. tears), gastric acid and microbial antagonism. . If penetration occurs, bacteria are destroyed by soluble

. The main phagocytic cells are polymorphonuclear neutrophils and macrophages. . The phagocytic cells use their pattern recognition receptors (PRRs) to recognize and adhere to pathogenassociatedmolecularpattems(PAMPs)onthemicrobesurface (Continuedp 20)

lro

CHAPTER I _ I N N A T EI M M U N I T Y

o PRRs include Toll-like, C-type and scavenger receptors. . Organisms adhering to the phagocyte surface activate the engulfment process and are taken inside the cell where they fuse with cytoplasmic granules. . A formidable array of microbicidal mechanisms then come into play: the conversion of O, to reactive oxygen intermediates, the synthesis of nitric oxide and the releaseof multiple oxygen-independent factors from the granules. . Adherence to PRRs on dendritic cells initiates adaptive immune processes (seeChapter 2). Complemenlf0cilitolesph0gocylosis . The complement system, a multicomponent

triggered erl.zyrr.ecascade, is used to attract phagocytic cells to the microbes and engulf them. o In what is known as the alternative complement pathway, the most abundant component, C3, is split by a convertase enzyme formed from its own cleavage product C3b and factor B and stabilized against breakdown caused by factors H and I, through association with the microbial surface. As it is formed, C3b becomes linked covalently to the microorganism and acts as an opsonin. o The next component, C5, is activated yielding a small peptide, C5a; the residual CSb binds to the surface and assembles the terminal components C6-9 into a membrane attack complex which is freely permeable to solutes and can lead to osmotic lysis. . C5a is a potent chemotactic agent for neutrophils and greatly increases capillary permeability. . C3a and C5a act on mast cells causing the release of further mediators, such as histamine, leukotriene Bn and tumor necrosis factor (TNF), with effects on capillary permeability and adhesiveness, and neutrophil chemotaxis; they also activate neutrophiis Ihe complemenl-medioled 0cuteinfl0mm0loryteoclion . Following the activation of complement with the ensuing

attraction and stimulation of neutrophils, the activated phagocytes bind to the C3b-coated microbes by their surface C3b receptors and may then ingest them. The influx of polymorphs and the increase in vascular permeability constitute the potent antimicrobial acute inflammatory response (figure 2.18). . Inflammation can also be initiated by tissue macrophages which subserve a similar role to the mast cell, since signaling by bacterial toxins, C5a or iC3b-coated bacteria adhering to surface complement receptors causes release of neutrophil chemotactic and activating factors.

provide Humorol mechonisms osecond defensive strolegy r In addition to lysozyme, peptide defensins and the complement system, other humoral defenses involve the acute phase proteins, such as C-reactive and mannose-binding proteins, whose synthesis is greatly augmented by infection. Mannose-binding lectin generates a complement pathway which is distinct from the alternative pathway in its early reactions, as will be discussed in chapter 2. It is a member of the collectin family which includes conglutinin and surfactants SP-A and SP-D, notable for their ability to 'self' distinguish microbial from surface carbohydrate groups by their pattern recognition molecules. . Recovery from viral infections can be effected by the interferons which block viral replication. Exlrocellulor killing . Virally infected cells can be killed by natural killer (NK) cells through a perforin/granzyrne and a separate Fasmediated pathway, leading to programed cell death (apoptosis) mediated by activation of the caspase protease cascadewhich fragments the nuclear DNA. . Extracellular killing by C3b-bound eosinophils may be responsible for the failure of many large parasites to establish a foothold in potential hosts.

Specificocquiredimmunity

INTRODUCTION Our microbial adversaries have tremendous opportunities through mutation to evolve strategies which evade our innate immune defenses. For example, most of the successfulparasites activate the altemative complement pathway and bind C3b, yet eosinophils which adhere are somehow not triggered into offensive action. The same holds true for many bacteria, while some may so shape their exteriors as to avoid complement activation 'devise' completely. The body obviously needed to defense mechanisms which could be dovetailed individually to each of these organisms no matter how many there were. In other words auery largenumberof specific immune defenses needed to be at the body's disposal. Quite a tall order!

I H E S P E C I F IACD A P I O R A N T I B O D_Y Evolutionary processescame up with what can only be described as a brilliant solution. This was to fashion an adaptor molecule which was intrinsically capable not only of activating the complement systernand of stimulating phagocytic cells, but also of sticking to the offending microbe. The adaptor thus had three main regions, two concerned with communicating with complement and the phagocytes (the biological functions) and one devoted to binding to an individual microorganism (the external recognition function). In most biological systems like hormones and receptors, and enzymes and substrates, recognition usually occurs through fairly accurate complementarity in shape allowing the ligands to approach so close to each other as to permit the normal intermolecular forces to become relatively strong. In the present case,each adaptor would have a recognition portion complementary in shape to some

microorganism to which it could then bind reasonably firmly. The part of the adaptor with biological function would be constant, but for each of hundreds of thousands of different organisms, a special recognition portion would be needed. Thus the body has to make hundreds of thousands, or even millions, of adaptors with different recognition sites. The adaptor is of course the molecule we know affectionatelyas antibody (figure 2.1). Anlibodyinilioteso newcomplemenlpothwoy('clossic0l') Antibody, whenbound to a microbe, will link to the first molecule in the so-called classical complement sequence, C1q, and trigger the latent proteolytic activity of the C1 complex (figure 2.2). This then dutifully plays its role in the amplifying cascade by acting on cornponents C4 and C2 to generate many molecules of C4b24 a new C3-splitting enzyme (figure 2.3). The molecular events responsible for this seem to be rather clear. C1q is polyvalent with respect to antibody binding and consists of a central collagen-like stem branching into six peptide chains each tipped by an antibody-binding subunit (resembling the blooms on a bouquet of flowers). C1q is associatedwith two further subunits, C1r and C1s,in a Ca2*-stabilizedtrimolecular complex (figure 2.2). Both these molecules contain repeats of a 60-amino acid unit folded as a globular domain and referred to as a complement controlprotein (CCP) repeat since it is a characteristic structural feature of several proteins involved in control of the complement system. Changes in C1q consequent upon binding the antigen-antibody complex bring about the sequential activation of proteolytic activity in C1r and then C1s. The next component in the chain, C4 (unfortunately components were numbered before the sequence was

t,,

C H A P T E 2R_ S P E C I F I A CC Q U I R E D IMMUNITY

INCREASED VASCUTAR PERMEABILITY

< RECOGNITION SITE

I I

PHAGOCYTOSIS

CHEMOTAXIS V Aclivotion PHAGOCYTE Figure 2.1. The antibody adaptor molecule. The constant part with biological function (BIOL) activates complement and the phagocyte. The portion with the recognition unit for the foreign microbe (REC) varies from one antibody to another.

established),now binds to C1 through theseCCPsand is cleaved enzymically by C1s. As expected in a multienzyme cascade,several molecules of C4 undergo cleavage,eachreleasinga small C4a fragment and revealing a nascent labile internal thiolester bond in the residual C4b like that in C3 (cf. figure 1.11)which may thenbind either to the antibody-C1 complex or the surface of the microbe itself. Note thatC4a,like C5a and C3a,has anaphylatoxin activity, although feeble, and C4b resembles C3b in its opsonic activity. In the presenceof Mg2*, C2 can complex with the C4b to become a new substrate for the C1s, the resulting product C4b2a now has the vital C3 convertaseactivity required to cleaveC3. This classical pathway C3 convertase has the same specificity as the C3bBb generated by the alternative pathway, likewise producing the same C3a and C3b fragments. Activation of a single C1 complex can bring about the proteolysis of literally thousands of C3 molecules.From then on things march along exactly in parallel to the post-C3 pathway with one molecule of C3b added to the C4b2a to make it into a C5-splitting enzyme with eventual production of the membrane attack complex (figures 1.73and2.4).Just as the alternative pathway C3 convertase is controlled by factors H and I, so the breakdown of C4b2a is brought about by either a C4-binding protein (C4bp) or a cell surface C3b receptor (CR1)in the presenceof factor L

Themonnose-binding leclinondclossic0lcomplement

pothwoys merge

CI COIVIPLEX Figute 2.2. Activation of the classical complement pathway. C1 is composed of C1q associated with the flexible rodlike Ca-dependent complex, Clrr-C1s, (H@; sandrindicatepotential serine protease active sites), which interdigitates with the six arms of C1q, either as rndicated or as 'W' shapes on the outer srde of these arms. The C1-inhibitor normally prevents spontaneous activation of Clrr-C1sr. If the complex of a microbe or autigen with antibodies attachestwo or more of the globular Ab-binding siteson C1q, the mo1ecule presumably undergoes conformational change which releases the Cl-Inh and activatesC1r, C1s"

It is appropriate at this stage to recall the activation of complement by innate immune mechanisms involving mannose-binding lectin (MBL) (cf. p . 77). On complexing with a microbe, MBL associateswith and activates the latent proteolytic activity of the serum proteases, MASP-1 and 2, which structurally resembleC1r and C1s respectively. In an entirely parallel fashion, the complex splits C4 and C2 to generate the classicalpathway C3 convertase. The similarities between the pathways are set out in figure 2.3 and show how antibody can supplement and even improve on the ability of the innate immune system to initiate acute inflammatory reactions through its ability to recognize a multitude of different microbes. Human antibodies are divided into five main classes:immunoglobulin M (shortened to IgM), IgG, IgA,IgE and IgD, which differ in the specialization of their 'rear ends' for different biological functions such as complement activation or mast cell sensitization. The ability of immunoglobulin E antibodies to sensitize mast cells through binding to their specific surface receptors so that combination with antigen triggers

C H A P T E 2R_ S P E C I F I A CC O U I R E D IMMUNITY

23 1

Figure 2.3. Comparison of the alternative, classical and mannosebinding lectin complement pathways. The classicalpathway is activated by antibody whereas the alternative and MBL pathways are not The molecular units with protease activity are highlighted, the enzymic domains showing considerable hom*ology. Beware confusion with nomenclature; the large C2 fragment which forms the C3 conver-

tase is designated as C2a, but to be consistentwith C4b, C3b and CSb, it would have been more logical to call it C2b Note that C-reactive protein (p 17), onbinding to microbial phosphorylcholine, can trigger the classical pathway. Mannose-binding lectin, when combined with microbial surface carbohydrate, associates with the serine proteases MASP-1 and 2 (p 17) which splitC4 and C2

mediator releaseindependentiy of C3a or Csa (cf. figure 1.15),adds yet more flexibility to our mechanisms for generating infl ammatory responses.

bacterium could be attracted t o a m a c r o p h a g e a t h o u sand times more strongly than a single antibody molecule (figure 2.5).

Complexedontibodyoctivotesphogocyticcells

CEttUtAR BASISOFANIIBODY PRODUCIION

We drew attention to the fact that some C3b-coated organisms may adhere to phagocytic cells and yet avoid provoking their uptake. If small amounts of antibody are added the phagocyte springs into action. It does so through the recognition of two or more antibody molecules bound to the microbe, using specialized receptorson the cell surface. A single antibody molecule complexed to the microorganism is not enough because it cannot cause the cross-linking of antibody receptorson the phagocyte surface membrane which is required to activate the cell. There is a further consideration connected with what is often called the bonus effect of multivalency; for thermodynamic reasons, which will be touched on in Chapter 5, the associationconstantof ligands, which use severalrather than a single bond to reactwith receptors, is increased geometrically rather than arithmetically. For example, three antibodies bound closetogether on a

Anlibodiesore modeby lymphocyles The majorityof restinglymphocytes are small cellswith a darkly staining nucleus due to condensed chromatin and relatively little cytoplasm containing the odd mitochondrion required for basic energy provision. Figures 2.6 and2.7 compare the morphology of these cells with that of the minority population of large granular leukocytes which includes the natural killer (NK) set referred to in Chapter 1. The central role of the small lymphocyte in the production of antibody was established largely by the work of Gowans. He depleted rats of their lymphocytes by chronic drainage of lymph from the thoracic duct by an indwelling cannula, and showed that they had a grossly impaired ability to mount an antibody response to microbial challenge. The abilityto form antibody could be restored by injecting thoracic duct

lro

IMMUNITY C H A P I E R2 - S P E C I F I CA C A U I R E D lymphocytes obtained from another rat. The same effect could be obtained if, before injection, the thoracic duct cellswere first incubated at37oC for24hours under conditions which kill off large- and medium-sized cells and leave only the small lymphocytes. This shows that the small lymphocyte is necessary for the antibody resPonse. The small lymphocytes can be labeled if the donor rat is previously injected with tritiated thymidine; it then becomes possible to follow the fate of these lymphocytes when transferred to another rat of the same strain which is then injected with microorganisms to produce an antibody response(figure 2.8).It transpires that after contact with the injected microbes, some of the transferred labeled lymphocytes develop into plasma cells (figures 2.6d and 2.9) which can be shown to contain (figure 2.6e)and secreteantibody.

Figure 2.4.Multiple lesions in cell wall of Escherichia colibaclefism caused by interaction with IgM antibody and complement. Each 'dark prt' due Lesionis caused by a single IgM molecule and shows as a 'negative to penetration by the stain' This is somewhat of an illusion sinceinreality these'pits'are like volcano cratersstandingproud ofthe 'membrane attack' complexes Comparasurface, and are each single ble results may be obtained in the absence of antibody since the cel1 wall endotoxin can activate the alternative pathway in the presence of higher concentration of serum (xa00 000) (Courtesy of Drs R Dourmashkin and J H Humphrey.)

low offinifysinglebond

whichmoke0ntibody Anligenseleclslhe lymphocytes The molecules in the microorganisms which evoke and react with antibodies are called antigens (generates antibodies). We now know that antibodies are formed before antigen is ever seen and that they are selected for by antigen. It works in the following way. Each lymphocyte of a subsetcalled the B-lymphocytes, becausethey differentiate in thebonemarrow,is programed to make one, and only one, antibody and it places this antibody on its outer surfaceto act as a receptor.This can be detectedby

bond Highovidilymulfiple

Anlibody recepror ilo frlggerlng

Idggerphogocylosis

Figure 2.5. Binding of bacterium to phagocyte by multiple antibodies gives strong association forces and triggers phagocytosis by cross-linking the surface receptors for antibody.

25 1

C H A P T E 2R- S P E C I F I C IMMUNITY ACQUIRED

FLUORESCENT ANTI-Ig

SURFACE rg

B-CELL PATCH FORMATION (b)

r$ (d)

(e)

(f)

Figure 2.6. Cells involved in the acquired immune response. (a) Sma1llymphocytes Condensed chromatin gives rise to heavy staininp; of the nucleus The cell on the bottom is a typical resting agranular Tcell with a thin rim of cytoplasm The upper nucleated cell is a large granular lymphocyte; it has more cytoplasm and azurophilic granules are evident lsolated platelets are visible B-lymphocytes rangc from small to intermediate in size and lack granules Ciemsa stain. (b) Transformed T-lymphocytes (lymphoblasts) following stimulation of lymphocvtes in culture with a polyclonal activator, such as the lectins phytohemagglutinin, concanar.alin A and pokeweed mrtogen which stimulate a wide range of cells rndependently of their specifrcity for antigen The large lymphoblastsn iththeir relativelyhighratio of cytoplasm to nucleus may be compared in size with the isolated small lymphocyte One cell is in mitosis May Gninwald-Giemsa (c) Immunofluorescent staining of BJymphocyte surface immunoglobulin using fluorescein-conjugated( ) anti-Ig Provided the reaction is carried out in the cold to prevent prnocytosis, the labeled antrbodv cannot penetrate to the interior of the viable lymphocytes and reacts

only with surface components Patches of aggregated surface Ig are seen which are beginning to form a cap in the right-hand lymphocyte During cap formation, submembranous myosin becomes redistributed in association with the surface Ig and induces locomotion of the previouslv sessilecell in a drrection away from the cap (d) Plasma cells. The nucleus is eccentric The cytoplasm is strongly basophilic due to high RNA content The juxtanuclear lightly stained zone corre(e) Plasma sponds with the Colgi region May-Gninwald-Giemsa cells stained to show intracellular immunoglobulin using a fluorescern-labeled anti-IgG (green) and a rhodamine-conjugated anti-IgM (rcd) (f) Langerhans' cells in human epidermrs rn leprosy, increased in the subepidermal zone, possibly as a consequence of the disease process Stained red by the immunoperoxidase method with 5-100 antibodies (Material for (a)was kindly supplied by Mr M Watts of the Department of Haematology, Middlesex Hospital Medical School;(b) and (c) by ProfessorP Lydyard; (d) and (e)by ProfessorC Grossr;and (f) bv Dr Marian Ridley.)

using fluorescent probes and, in frgwre2.6c,one can see the molecules of antibody on the surface of a human BIymphocyte stained with a fluorescentrabbit antiserum raised against a preparation of human antibodies. Each lymphocyte has of the order of 105identical antibody molecules on its surface. When an antigen enters the body, it is confronted by a dazzling array of lymphocytes all bearing different anti-

bodies each with its own individual recognition site. The antigenwill onlybind to those receptorswithwhich it makes a good fit. Lymphocytes whose receptorshave bound antigen receive a triggering signal and develop into antibody-forming plasma cells and, since the lymphocytes areprogramed to make only one antibody, that secretedby the plasma cell will be identical with that originally acting as the lymphocyte receptor,i.e. it will

lru

CHAPTE2 R- S P E C I F I C ACOUIRED IMMUNITY

Figure 2.7. Lymphocyte ultrastructure. (a) Small agranular TJymphocyte Indented nucleus with condensed chromatin, sparse cytoplasm: single mitochondrion shown and many free ribosomes but otherwise few organelles (x13 000) BJymphocytes are essentially srmilar with slightly more cytoplasm and occasional elements of rough-surfaced endoplasmic reticulum. (b) Large granular lymphocyte (x7500). The more abundant cytoplasm contains several mitochondria (M), free ribosomes (R) with some minor elements of rough-surfaced endoplasmrc reticulum (ER),prominentGolgi apparatus (Go) and characteristic membrane-bound electron-dense granules (Gr). The nuclear chromatin is lesscondensed than that of the agranular T-cel1Certain types of T lymphocyte passessa large granular appearance However, many large granular /eukocytesare classical NK cells whrch, because they do not express antigen-specific receptors, are not lymphocytes (Courtesy ofProfessorsA ZiccaandC E Crossi )

DONOR RAT

31o/24hrincubolion

RECIPIENT RA]

PLASIVACELL

bind well to the antigen. In this way, antigen selects for the antibodies which recognize it effectively (figure 2.10).

Theneedfot clonolexponsion meonshumotolimmunity muslbe ocquired Becausewe can make hundreds of thousands, maybe even millions, of different antibody molecules, it is not feasiblefor us to have too many lymphocytes producing each type of antibody; there just would not be enough room in the body to accommodatethem. To compensate for this, lymphocytes which are triggered by contact

Figure 2.8.Labeled small lymphocytes become antibody-f orming plasma cells when transferred to a recipient rat which is immunized with a bactenum Transferred cell with radioactive nucleus shown by autoradiography. Intracelluiar antibody revealed by staining with fluorescent probes (cf figure 2.6e).

with antigen undergo successivewaves of proliferation to build up a large clone of plasma cells which will be making antibody of the kind for which the parent lymphocyte was programed. By this system of clonal selection, large enough concentrations of antibody can be produced to combat infection effectively (Milestone2.7; figure 2.11). The importance of proliferation for the development of a significant antibody response is highlighted by the ability of antimitotic drugs to abolish antibody production to a given antigen stimulus completely. Becauseit takes time for the proliferating clone to build up its numbers sufficiently, it is usually several

C H A P T E2 R- S P E C I F I C A C A U I R E DI M M U N I T Y

A tinyfroction of thetotol populolion lymphoc!4e

27 1 Recognition site of surfoce ontibody receptol V

o.r,uot,on

Secreted onlibody combrnes wilhontigen ANTIBODY SECRETING PLASIVA CELL Figure 2.9.Plasma cell (x10 000) Prominent rough-surfacedendoplasmic reticulum associated with the synthesisand secretionof Ig

Figure 2.11.The cellular basis for the generation of effector and memory cells by clonal selection after primary contact with antigen. Thc cell selectedby antigen undergoes many divisions durrng the clonal pro[feration and tlre progeny mature to give an expanded population of antibody-forming cells The antibody responseis particularly vulnerable tcr antimitotic agentsat the proliferation stage A fraction of the progeny of the original antigenreactivelymphocytes become nondivid ing memory cells and others the effectorcells of either humoral, i e antibody-mediated, or, as we shall seesubsequently,cell-mediated immunity Memory cellsrequire fewer cycles before they develop into effectorsand this shortens the reaction time for the secondary response The expanded clone of cellswith memory for the original antigen provides the basis for the greater secondaryrelative to the primary immune response Primrng with 1ow dosesof antigen can oftcn stimulate effective memory without prclducing very adequate antibody synthesis

Figure 2.10. Antigen activates those B-cells whose surface antibody receptors it can combine with firmly.

Clonol proliterotion

EFFECIOR CELTSGIVII{GIMMUI{ERESPONSE

lrt

ACQUIRED IMMUNITY CHAPTE2 R- S P E C I F I C

Antibodyproduclion0ccordinglo Ehrlich ln 1894,well in advance of his time as usual, the remarkable Paul Ehrlich proposed the side-chain theory of antibody production. Each cell would make a large variety of surfacereceptors which bound foreign antigens by complementary shape 'lock and key' fit. Exposure to antigen would provoke overproduction of receptors (antibodies) which would then be shed into the circulation (figure M2.1.1). Templolelheories Ehrlich's hypothesis implied that antibodies were preformed prior to antigen exposure. However, this view was difficult to

which in some way provokes further synthesis of thatparticular antibody. Buthow? Mac Burnet now brilliantly conceived of a cellular basis for this selection process. Let each lymphocyte be programed to make its own singular antibody which is inserted like an Ehrlich'side-chain' into its surface membrane. Antigen will now form the complex envisaged by Jerne, on the surface of the lymphocyte, and by triggering its activation and clonal proliferation, large amounts of the specific antibody will be synthesized (figure 2.11).Bow graciously to that soothsayer Ehrlich- he came so close in 1894!

accept when later work showed that antibodies could be formed to almost any organic structure synthesized in the chemist's laboratory (e.g. azobenzene arsonate; figure 5.6) despite the fact that such molecules would never be encountered in the natural environment. Thus was bom the idea that antibodies were synthesized by using the antigen as a template. Twenty years passed before this idea was 'blown out of the water'by the observation that, after an antibody molecule is unfolded by guanidinium salts in the absenceof antigen, it spontaneously refolds to regenerate its original specificity. It became clear that each antibody has a different amino acid sequencewhich governs its final folded shape and hence its ability to recognize antigen.

Seleclion theorles The wheel turns full circle and we oncemore live with the idea tha! since different antibodies must be encoded by separate genes,the information for making these antibodies must preexist in the host DNA. In 1955,Nils Jerne perceived that this could form thebasis for a selective theory ofantibody production. He suggested that the complete antibody repertoire is expressed at a low level and that, when antigen enters the body, it selectsits complementary antibody to form a complex

days before antibodies are detectable in the serum following primary contact with antigen. The newly formed antibodies are a consequence of antigen exposure and it is for this reason that we speak of the acquired immune response.

A C O U I R EM DE M O R Y \Alhen we make an antibody response to a given infectious agent, by definition that microorganism must exist in our environment and we are likely to meet it again. It would make sense then for the immune mechanisms alerted by the first contact with antigen to leave behind some memory system which would enable the response

Figure M2.1.1. Ehrlich's side-chain theory of Ab production. (ReproducedfromProceedings oftheRoyalSocietyB (1900\,56,424.)

to any subsequent exposure to be faster and greater in magnitude. Our experience of many common infections tells us thatthis mustbe so. We rarely suffer twice from such diseasesas measles,mumps, chickenpox,whooping cough and so forth. The first contact clearly imprints some information, imparts some memory, so that the body is effectively prepared to repel any later invasion by that organism and a state of immunity is established. Secondory ontibodylesponsesore betler By following the production of antibody on the first and second contacts with antigen, we can seethebasis for the

C H A P T E 2R- S P E C I F I C A C S U I R E DI M M U N I T Y

PRIMARY RESPONSE

SECONDARY RESPONSE

I st injeclion of ontigen

2ndinjeclion of onligen

2sl

that this antigen is capable of giving a secondary booster response. We have explained the design of the experiment at some length to call attention to the need for careful selection of controls. The higher response given by a primed lymphocyte population can be ascribed mainly to an expansion of the numbers of cells capable of being stimulated by the antigen (figure 2.11), although we shall see later that there are some qualitative differences in these memory cells as well (pp.206-207).

M M U N I TH YA S A C O U I R EID A N T I G ES NP E C I F I C I T Y

u0ys Figure 2.12.Primary and secondary response.A rabbit is injected on two separate occasions with tetanus toxoid. The antibody response on the second contact with antigen is more rapid and more intense

development of immunity. For example, when we inject a bacterial product such as tetanus toxoid into a rabbit, for the reasons already discussed, several days elapse before antibodies can be detected in the blood; these reach a peak and then fall (figure 2.12).If wenow allow the animal to rest and then give a second injection of toxoid, the course of events is dramatically altered. Within 2-3 days the antibody level in the blood rises steeply to reachmuch higher values than were observed in the primary response.This secondary response then is characterizedby a more rapid and more abundant production of antibodyresulting from the'tuning up'or priming of the antibody-forming system. With our knowledge of lymphocyte function, it is perhaps not surprising to realize that these are the cells which provide memory. This can be demonstrated by adoptive transfer of lymphocytes to another animal, an experimental system frequently employed in immunology (cf. figure 2.8).In the present case,the immunological potential of the transferred cells is expressedin a recipient treated with X-rays which destroy its own lymphocyte population; thus any immune response will be of donor not recipient origin. In the experiment described in figure 2.13,small lymphocytes are taken from an animal given a primary injection of tetanus toxoid and transferred to an irradiated host which is then boosted with the antigen; a rapid, intense production of antibody characteristic of a secondary responseis seen.To exclude the possibility that the first antigen injection might exert anonspecificstimulatory effect on the lymphocytes, the boosting injection includes influenza hemagglutinin as a control antigen. Furthermore, a 'criss-cross' control group primed with influenza hemagglutinin must also be included to ensure

Discrimin0li0n belween differenl 0nligens The establishment of memory or immunity by one organism does not confer protection against another unrelated organism. After an attack of measles we are immune to further infection but are susceptible to other agents such as the chickenpox or mumps viruses. Acquired immunity then shows specificity and the immune system can differentiate specifically between the two organisms. Amore formal experimental demonstration of this discriminatory power was seen in figure 2.13where priming with tetanus toxoid evoked memory for that antigen but not for influenza and vice versa. The basis for this lies of course in the ability of the recognition sites of the antibody molecules to distinguish between antigens; antibodies which react with the toxoid do not bind to influenza and, mutatis mutandisas they say, anti-influenza is not particularly smitten with the toxoid. Discriminolion belweenselfqndn0nself This ability to recognize one antigen and distinguish it from another goes even further. The individual must also 'nonself'. recognize what is foreign, i.e. what is The failure to discriminate between self and nonself could lead to the slmthesis of antibodies directed against components of the subject's own body (autoantibodies), which in principle could prove to be highly embarrassi.g. Otr purely theoretical grounds it seemed to Burnet and Fenner that the body must develop some mechanism 'and'nonself 'could whereby'self be distinguished, and they postulated that those circulating body components which were able to reach the developing lymphoid system in the perinatal period could in some way be 'learnt' as 'self '. Apermanent unresponsiveness or tolerance would then be created so that as imrnunological maturity was reached there would normally be an inabil'self ' ity to respond to components. At this stage it is salu-

lso

IMMUNITY ACQUIRED C H A P I E R2 - S P E C I F I C

lsl injection ol ontigen>

Tronsfer smoll to lymphocytes irrodioled recipient

to: response onlibody BooslwilhMixlute Meosure .; of bothontigens ^*** _o,o-1

TETANUS TOXOID V SECONDARY

PRIiIARY

Figure 2.13, Memory for a prirnary response can be transferred by srnall lymphocytes. Recipients are treated with a dose of X-rays which directly kiil lymphocytes (highly sensitive to radiation) but only affect other body cells when they divide; the recipieni thus functions as a living'test-tube'whichpermits the function of the donor cellstobe fol-

lowed. The reasons for the design of the experiment are given in the text In practice, because of the possibility of interference between the two antigens, it would be wiser to split each of the primary antigeninjected groups into two, giving a separate boosting antigen to each to avoid using a mixture.

tory to note that Bumet had the sagacity to realize that his clonal selection theory could readily provide the cellular basis for such a mechanism to operate. He argued that if each lymphocytewere preoccupied withmaking its own individual antibody, those cells programed to express antibodies reacting with circulating self components could be rendered unresponsive without affecting those lymphocytes specific for foreign antigens. In other words, self-reacting lymphocytes could be selectively suppressed or tolerized withoutundermining the ability of the host to respond immunologically to infectious agents. As we shall see in Chapter 11, these predictions have been amply verified, although we will learn that, as new lymphocytes differentiate throughout life, they will all go through this self-tolerizing screening process. However, self tolerance is not absolute and normally innocuous but potentially harmful anti-self lymphocytes exist in all of us.

Doys

DN E P E N DOSNA C S U I R E D VACCINATIO MEM O R Y Some 200 years ago, Edward jenner carried out the remarkable studies which mark the beginning of immunology as a systematic subject.Noting the pretty pox-free skinof the milkmaids, he reasoned that deliberate exposure to the pox virus of the cow, which is not virulent for the human, might confer protection against the related human smallpox organism. Accordingly, he inoculated a small boy with cowpox and was delighted- and presumably breathed a sigh of relief to observethat the boy was now protected against a subsequent exposure to smallpox (what would today's ethical committees have said about that?!). By injecting a harmless form of a disease organism, |enner had

Figwe 2.14. The basis of vaccination illustrated by the response to tetanus toxoid Tieatment of the bacterial toxin with formaldehyde destroys its toxicity (associatedwith aa ) but retains antigenicity (assocrated with l). Exposure to toxin in a subsequent nafural infection boosts the memory cells, producing high levels of neutralizing antibody which are protective

utilized the specificity and memory of the acquired immune response to lay the foundations for modern vaccination (Latin uacca, cow). The essentialstrategy is to prepare aninnocuousforfir of the infectious organism or its toxins which still substantially retains the antigens responsible for establishing protective immunity. This has been done by using killed or live attenuated organisms, purified microbial components or chemically modified antigens (figure 2.14).

3ll

C H A P T E 2R- S P E C I F I C A C A U I R E DI M M U N I T Y

C E t T - M E D I A I EI M D M U N I TP YR O I E C I S A G A I N SITN T R A Ct T EU I . A R ORGANISMS

Cytokine-producing T-cellshelpmocrophoges l0 kill porosiles inlrocellulor

Many microorganisms live inside host cells where it is impossible for humoral antibody to reach them. Obligate intracellular parasites like viruses have to replicate inside cells; facultative intracellular parasites llke Mycobacteriaand Leishmoniacan replicate within cells,particularly macrophages,but do not have to; they like the intracellular life because of the protection it affords. A totally separate acquired immunity system has evolved to deal with this situation based on a distinct lymphocyte subpopulation made up of T:cells, designated thus because, unlike the B-lymphocytes, they differentiate within the milieu of the thymus gland. Becausethey are specialized to operate against cellsbearing intracellular organisms,T-cellsonly recognize antigen when it is on the surface of a body cell. Accordingly, the T-cell surface receptors,which are different from the antibody molecules used by B-lymphocytes, recognize antigen plus a surface marker which informs the T-lymphocyte that it is making contact with another cell. Thesecell markers belong to an important group of molecules known as the major histocompatibility complex (MHC), identified originally through their ability to evoke powerful transplantation reactions in other members of the same species.Now naive or virgin T-cells must be introduced to the antigen and MHC by a special dendritic antigen-presenting cell (figures 2.6f and 7.15)before they can be initiated into the rites of a primary response.However, once primed, they are activated by antigen and MHC present on the surface of other cell types such as macrophages as we shall now see.

These organisms only survive inside macrophages through their ability to subvert the innate killing mechanisms of these cells. Nonetheless, they mostly cannot prevent the macrophage from processing small antigenic fragments (possibly of organisms which have spontaneously died) and placing them on the host cell surface. A subpopulation of T-lymphocytes called Thelper cells, if primed to that antigen, will recognize and bind to the combination of antigen with so-called classII MHC molecules on the macrophage surfaceand produce a variety of soluble factors termed cytokines which include the interleukins IL-2, etc. (p. 185).Differentcytokines canbe madebyvarious cell types and generally act at a short range on neighboring cells. Some T-cell cytokines help B-cells to make antibodies, while others such as y-interferon (IFNy) act as macrophage activating factors which switch on the previously subverted microbicidal mechanisms of the macrophage and bring about the death of the intracellular microorganisms (figure 2.15).

Virollyinfecledcellsconbe killed by cyloloxicT-cells ondADCC We have already discussedthe advantage to the host of killing virally infected cells before the virus begins to replicate and have seen that natural killer (NK) cells (p. 18) can subserve a cytotoxic function. However, NK cells have a limited range of specificities and, in order to improve their efficacy, this range needs to be expanded.

Figure 2.15.Intracellular killing of rnicroorganisms by macrophages. (1) Surfaceantigen (l ) derived from the intracellular mrcrobesis compiexed with class II MHC molecules (ffi ) (2) The primecl T-helper cell binds to this surfacecomplex and is triggered to release the cytokrne yinterferon (IFNy) This activatesmicrobicidal mechanisms in the macrophage (3) The infectious agent meets a timely death

lnfection by inlrocellul0r f0cullolive orgonisms

[/ocrophoge oclivolion

Deolhof inlrocellulor microbes

lsz

C H A P I E R2 - S P E C I F I CA C S U I R E DI M M U N I T Y

One way in which this can be achieved is by coating the target cell with antibodies specific for the virally coded surfaceantigens becauseNK cells have receptors for the constant part of the antibody molecule, rather like phagocytic cells. Thus antibodies will bring the NK cell very close to the target by forming a bridge, and the NK cell being activated by the complexed antibody molecules is able to kill the virally infected cell by its extracellular mechanisms (figure 2.16).This system, termed antibody-dependent cellular cytotoxicity (ADCC), is very impressive when studied inaitrobutit has proved difficult to establish to what extent it operates within the body. On the other hand, a subset of T-cells with cytotoxic potential has evolved for which there is clear evidence of in aiao activity. Like the T-helpers, these cells have a very wide range of antigen specificities because they clonally express a large number of different surface receptors similar to, but not identical with, the surface antibody receptors on the B-lymphocytes. Again, each lymphocyte is programed to make only one receptor and, again like the T-helper cell, recognizes antigen only in associationwith a cell marker, in this casethe class I MHC molecule (figure 2.16).Through this recognition of surface antigen, the cytotoxic cell comes into intimate

'kiss of apopcontact with its target and administers the totic death'. It also releasesIFNywhich would help to reduce the spread of virus to adjacent cells, particularly in caseswhere the virus itself may prove to be a weak inducer of IFNcr or P. In an entirely analogous fashion to the B-cell, T-cells are selected and activatedby combinationwith antigen, expanded by clonal proliferation and mature to give T-helpers and cytotoxic T-effectors, together with an enlarged population of memory cells. Thus both Tand B-cellsprovide specific acquired immunity with a variety of mechanisms, which in most casesoperate to extend the range of effectiveness of innate immunity and confer the valuable advantage that a first infection prepares us to withstand further contact with the same microorganism.

IMMUNOPATHOTOGY The immune system is clearly'a good thing', but like mercenary armies, it can turn to bite the hand that feeds it, and causedamage to the host (figure 2.17). Thus where there is an especially heightened response or persistent exposure to exogenous antigens, tissue damaging or hypersensitivity reactions may

, Speciflc l-cell cyloloxlclty

Inoppropriofe onligen e.g.groft

ooo oo lnfectious 0genI Figure 2.16. Killing virally infected cells. The killing mechanism oI the NK cell can be focused on the target by antibody to produce antibody-dependent cellular cytotoxicity (ADCC) The cytotoxic T-cell homes onto its target specifically through recePtor recognition o{ surface antigen in association with MHC class I molecuLes

Figwe 2.lT.Inappropriate and suboptimal immune responses can produce damaging reactions such as the hypersensitivity response to destructionof self tissueby inhaledotherwiseinnocuousallergens'the autoimmune attack, the reiection of tissue transplants, and the susceptibilitv to infection in immunodeficient individuals.

C H A P T E 2R- S P E C I F I C IMMUNITY ACQUIRED result. Examples are allergy to grass pollens, blood dyscrasias associated with certain drugs, immune complex glomerulonephritis occurring after streptococcal infection, and chronic granulomas produced during tuberculosis or schistosomiasis. In other cases, hypersensitivity to autoantigens may arise through a breakdown in the mechanisms which control self tolerance, and a wide variety of autoimmune diseases, such as insulin-dependent diabetes and multiple sclerosis and many of the rheumatologic disorders, have now been recognized.

-the speclflc Antibody orloptor . The antibody molecule evolved as a specific adaptor to attach to microorganisms which either fail to acfivate the altemative complement pathway or prevent activation of thephagocytic cells. r The antibody fixes to the antigen by its specific recognition site and its constant structure regions activate complement through the classical pathway (binding C1 and generating aC4b2a convertase to split C3) and phagocytes through their antibody receptors. . This supplementary route into the acute inflammatory reaction is enhanced by antibodies which sensitize mast

33 1

Another immunopathologic reaction of some consequence is transplant rejection, where the MHC antigens on the donor grafl may well provoke a fierce reaction. Lastly, one should consider the by no means infrequent occurrence of inadequate functioning of the immune system-immunodeficiency. We would like to think that at this stage the reader would have no difficulty in predicting that the major problems in this condition relate to persistent infection, the type of infection being related to the elements of the immune system which are defective.

cells and by immune complexes which stimulate mediator releasefrom tissue macrophages (figure 2.18). o The innate immune reaction of mannose-binding lectin with microbes activates the MASP-1 and MASP-2 proteases which join the classical complement pathway by splitting C4andC2.

producti0n of0nllbody Gellulor b0sis . Antibodies are made by plasma cells derived from Blymphocytes, each of which is programed to make antibody of a single specificity which is placed on the cell surface as a receDtor.

Figure 2.1E.Production of a protective acute inflammatory reaction by microbes either: (i) through tissue injury (e.g bacterial toxin) or direct activation of the altemative complement pathway, or (ii) by antibody-dependent triggering of the classical complement pathway or mast cell degranulation (a speciai class of antibody, IgE, does this) (Continuedp 34)

lro

IMMUNITY C H A P I E R2 - S P E C I F I C ACQUIRED

o Antigen binds to the cell with a complementary antibody, activates it and causesclonal proliferation and finally maturation to antibody-forming cells and memory cells. Thus the antigen brings about clonal selection of the cells making antibody to itself.

Acquired memory ondvoccinolion r The increase in memory cells after priming means that the acquired secondary response is faster and greater, providing the basis for vaccination using a harmless form of the infective agent for the initial injection. Acquiredimmunilyh0s 0nligenspecificity . Antibodies differentiate between antigens because recognition is based on molecular shape complementarity. Thus memory induced by one antigen will not extend to another unrelated antigen. o The immune system differentiates self components from foreign antigens by making immature self-reacting lymphocytes unresponsive through contact with host molecules; lymphocytes reacting with foreign antigens are unaffected since they only make contact after reaching maturity. Cell-medioledimmunilyprolecls0g0insl

inlrocellulor orgonisms . Another class of lymphocyte, the T-cell, is concerned with control of intracellular infections. Like the B-cell, each T-cell has its individual antigen receptor (although it differs structurally from antibody) which recognizes antigen and the cell

INTRACELLULAR PARASITES - r < t

I I

then undergoes clonal expansion to form effector and memory cells providing specific acquired immunity. o The T-cell recognizes cell surface antigens in association with molecules of the MHC. Naive T-cells are only stimulated to undergo a primary response by specialized dendritic antigen-presenting cells. . Primed T-helper cells, which see antigen with class II MHC on the surface of macrophages, release cytokines which in some casescan help B-cells to make antibody and in others activate macrophages and enable them to kill intracellular parasites. . Cytotoxic T-cells have the ability to recognize specific antigen plus class I MHC on the surface of virally infected cells which are killed before the virus replicates. They also release y-interferon which can make surrounding cells resistant to viral spread (figure2.79). . NK cells have lectinlike 'nonspecific' receptors for cells infected by viruses but do not have antigen-specific receptors; however, they can recognize antibody-coated virally infected cells through their Fcyreceptors and kill the targetby antibody-dependent cellular cytotoxicity (ADCC). . Although the innate mechanisms do not improve with repeated exposure to infection as do the acquired, they play a vital role since they are intimately linked to the acquired systems by two different pathways which all but encapsulate the whole of immunology. Antibody, complement and polymorphs give protection against most extracellular organisms, while T-cells, soluble cytokines, macrophages and NK cells deal with intracellular infections (figure 2.20)

VIRUS.INFECTED UNINFECTED CELL CELL

Figure 2.19. T-cells link with the innate immune system to resist intracellular infection. ClassI (kl) and classII (kl)

Mocrophoge octivolion ond chemoloxis

RISE GIVES T0

^ -> ACTIVATES- +

INHIBITS

majorhistocompatibility molecules are important for T-cell recognition of surface antigen. The T-helper cells (Th) cooperate in the development of cytotoxic T-cells (Tc) from precursors The macrophage (MQ) microbicidal mechanisms are switched onbymacrophage activating cytokines Interferoninhibits viral replication and stimulates NK cells w h i c h , t o g e t h e rw i t h T c ,k i l l v i r u s infected cells

C H A P T E 2R- S p E C t F l CA C a U I R E Dt M M U N t T y

lmmunopothology o Immunopathologically mediated tissue damage to the host can occur as a result of: inappropriate hypersensitivity reactions to exogenous antigens;

35 |

loss of tolerance to self giving rise to autoimmune disease; reaction to foreign grafts. o Immunodeficiency leaves the individual susceptible to infection.

Figure 2.20.The two pathways linking innate and acquired immunity which provide the basis for humoral and cellmediated immunity, respectively.

F U R T H ERRE A D I N G General reading Alt F & MarrackP (eds)(2001) CurrentOpinion in lntnrunology 13

(Appears bimonthly; Issue No. 1 deals with 'Innate Immunity, ) Very valuable critical current reviews Kim T. & Kim YJ (2005) Overview of innate immunity in Drosophila I Biochen Mol Biol 38,12I-727 Matzinger P (2005) The danger model: a renewed sense of self Science296,301-305 Parker L C et al (2005)The expression and roles of Toll-like receptors in the biology of the human neutrophil I LetLkocBiol. 77, 886-992 Reid K B.M (1995) The complement system - a major effector mechanism in humoral immunity. Thelmmunologist 3,206 Segal A W. (2005) How neutrophils kill microbes Ann. Rea lmmunol 'r? 147-))2,

Srnkovics J.G & Horvath J C. (2005) Human natural killer cells: a comprehensive review. Inf. I Oncol.27,547. Sitaram N & Nagaraj R (1999) Interaction of antimicrobial peptides with biological and model membranes: structural and charge requirements for activity. Biochimicoet BiophysicaActa-Biontembranes1462,29 Stuart L M. & Ezekowitz R A (2005)Phagocytosis:elegant complexity. lwnunity 22, 539-550 Worthley D.L., Bardy P.G.& Mulligan C.G. (2005)Mannose binding lectin: biology and clinical implications. Intern. Med. l. 35, 538-SSE.

Historical Clarke W.R (7997) TheEtperimental Fountiationsof Modern lmmunology,4th edn fohn Wiley & Sons, New York (Important for those wishing to appreciate the experiments leading up to many of the maior discoveries ) Ehrlich P (1890) On immunity with special reference to cell life. In Melchers F ef al. (eds)Progressln lmmunologyylL springer-Verlag, Berlin. (Translation of a lecture to the Royal Society (London) on the side-chain theory of antibody formation, showing this man,s perceptlve genius Amust!)

Landsteiner K. (7946) The Specit'city of SerologicnlReactions Harvard University Press (reprinted 7962 by Dover Publications, New York). Mazumdar P.M M (ed.) (7989) lmmunology 1930-1980. Wall & Thompson, Toronto. Metchnikoff E. (1893) Comparatioe Pathology of InJTammation Kegan Paul, Trench, tubner, London (translated by F.A & E H. Starling). Palmer R. (ed ) (1993) Outstanding Papersin Biology Current Biology, London (A delight to browse through some of the seminal papers which have shaped modern biology; wonderful material for teaching ) SilversteinA.M (7989) AHistory of lmmunology. Academic Press, San Diego Tauber A.I. (1991) Metchnikoff and the Origins of lmmunology. Oxtord University Press, Oxford

In-depth seriesfor the adaanced reader Adoances in lmmunology Elsevier Science Publications AtluancesinNeuroimmunology (edited by G B. Stefano & E.M. Smith). Pergamon, Oxford Annual Reaiewof lmmunology Annual Reviews Inc., California lmmunological Reaiews (edited by P Parham) Munksgaard, Copenhagen. (Specialized, authoritative and thoughtful ) Nature Reuiews,Nature Publishing Group, London Progressin Allergy Karger, Basle. Seminars in lmmunology. Elsevier Science Publications. (In-depth treatment of single subjects.)

Currentinformation Current BioLogy.Current Biology, London (What the complete biologist needs to know about significant current advances ) Current Opinion in lmmunology. Current Science, London. (Important personal opinions on focused highlights of the advances made in the previous year; most valuable for the serious immunologist.)

C H A P I E R2 - S P E C I F I CA C A U I R E DI M M U N I T Y

136 Table 2.1. The major immunologic factors.

journals and their impact

Cell

28,4

EMBOJ

l05

L0ncel

2 11

Nolure

32,2

Nolure Medicine

3t,2

NewEnglond Journol ol Medicine

386

Proceedings 0l lhe N0ti0n0lAcodemyol Sciencesof lhe USA

10,5

Science

3lI

AIDS

Allergy

TrendsinMolecularMedicine. ElsevierSciencePublications,Amsterdam. (Frequent articles of interest to immunologists with very good perspective ) The lmmunologlsf. Hogrefe & Huber Publishers, Seattle. (Official organ of the International Union of Immunological SocietiesIUIS. Excellent, didactic and compact articles on current trends in irnmunology.) Trends in Immunology. Elsevier Science Publications, Amsterdam. (The immunologist's'newspaper'. Excellent )

Multiple choicequestions Roitt I.M. & Delves P.J 400 MCQs each with five annotated learning responses. See Website.

Website (linked to Roitt's Essential lmmunology) http :I I wzuw.roitt.coffi The website contains: . 400 multiple-choice revision questions with feedback on each answer selected . Key concePts illustrated with animations o Further reading and reference archive . Image archive with over 400 illustrations from Essential Immunology

Auloimmunity

Concer lmmunology lmmunolheropy

Cellulor lmmunology

CIinicol ondExperimentol Allergy

CIinicol ondExperimentol lmmunology

2.3

Major journals

Clinicol lmmunology

Europ€on Journol of lmmunology

5,0

The major journals of interest and their impact factors are noted in table 2.1.

Humon lmmunology

lmmunity

t55

|mmunobiology

lmmun0genelics

lmmunologic Reseorch

rmmun0rogy

3,0

lmmunology ondCellBiology

lmmunology Lelters

hleclionondlmmunily

4.0

lnlernotionol Archives ofAllergy ondlmmunology

2.5

Inlernolionol lmmunology

Inlernolionol Journol of lmmunopothology ondPhormocology

J,O

Journol 0fAllergy ondClinicol Immunology

1.2

Journol ol Autoimmunity

JoumolofClinicol lmmunology

2.4

Journol ot ExDerimentol Mediclne

14,6

Joumol of lmmunology

6,5

Journol of lmmunologicol Melhods

Journol 0l lmmunolheropy

3.5

Journol of bukocyte Biology Journol 0l Reproduclive lmmunology

Moleculor lmmunology

J.Z

Noturelmmunology

276

Porosile lmmunology

Scondinovion Journol of lmmunology

t.9

Tissue Antigens

Tronsplonlotion

Voccine

*lltPACTfa*gTOR = relotive fcquency withwhichtheJournol's 'overoge orticle' hosbeencitedin otherpublicolions

Antibodies

INTRODUCTION

IHEDIVISIOO N FT A B O R

In essence,antibody molecules carry out two principal functions in immune defence. The first function is to recognize and bind to foreign material (antigen). This generally means binding to molecular structures on the surface of the foreign material (antigenic determinants) that differ from molecular structures made by the cells of the host. These antigenic determinants are usually expressed in multiple copies on the foreign material, e.g. proteins or carbohydrates on a bacterial cell surface or envelope spikes on the surface of a virus. Host antibodies can recognize a huge variety of different molecular structures-a human is capable of producing antibodies against billions of different molecular structures. This is described as antibody diversity and is necessary to respond to the huge diversity of molecular structures associated with (often highly mutable) pathogens. The simple act of antibody binding may be sufficient to inactivate a pathogen or render harmless a toxin. For instance, antibody coating of a virus can prevent it entering target cells and thereby'neutralize' the virus. However, in many instances, a second function of the antibody molecule is deployed to trigger the elimination of foreign material. In molecular terms, this involves the binding of certain molecules (effector molecules) to antibody-coated foreign material to trigger complex elimination mechanisms, e.g. the complement system of proteins, phagocytosis by host immune cells such as neutrophils and macrophages. The powerful effector systems are generally triggered only by antibody molecules clustered together as on a foreign cell surface and not by free unliganded antibody. This is crucial considering the typically high serum concentrations of antibodies.

The requirements imposed on the antibody molecule by the two functions are in a sensequite opposite. The first function requires great antibody diversity. The second function requires that many different antibody molecules share common features, i.e. it is not practical for Nature to devise a different molecular solution for the problem of elimination for each different antibody molecule. The conflicting requirements are elegantly metby the antibody structure shown diagramatically in figure. 3.1. The structure consists of three units. TWo of the units are identical and involved in binding to antigen-the Fab (fragment antigen binding) arms of the molecule. These units contain regions of sequence which vary greatly from one antibody to another and confer on a given antibody its unique binding specificity. The presence of two identical Fab arms enhances the binding of antibody to antigen in the typical situation where multiple copies of antigenic determinants are presented on foreign material. The third unit-Fc (fragment crystalline) is involved inbinding to effector molecules. As shown in figure. 3.1, the antibody molecule has a four-chain structure consisting of two identical heavy chains sparuring Fab and Fc and two identical light chains associated only with Fab. The relationship between antigen binding, the different units and the four-chain structure of the antibody molecule were revealed by a series of key experiments summarized in Milestone 3.1.

F I V EC I A S S E S O FI M M U N O G I O B U t I N Antibodies are often referred to as immunoglobulins (immune proteins). There are five classesof antibodies or immunoglobulins termed immunoglobulin G (IgG),

l.t

C H A P T E3 R- A N T I B O D I E S

Antigenbinding

Anligeb ninding

Figure 3.1. Simplified overall layout of the antibody molecule. The structure consists of four polypeptide chains, two identrcal heavy (H) chains and two identical light (L) chains, arranged to span three structurai units as shown The two identical Fab units bind antigen and the third unit (Fc) binds effector molecules to trigger antigen elimination and to mediate functions such as maternal -fetal transoort

IgM, IgA, IgD and IgE. All these classeshave the basic four-chain antibody structure but they differ in their heavy chains termed y, p, cx,, 6 and e,respectively. The differences are most pronounced in the Fc regions of the antibody classesand this leads to the triggering of different effector functions on binding to antigen, e.g. IgM recognition of antigen might lead to complement activation whereas IgE recognition (possibly of the same antigen) might lead to mast cell degranulation and anaphylaxis (increased vascular permeability and smooth muscle contraction). These differencesare discussed in greater detail below. Structural differences also lead to differences in the polymerization state of the monomer unitshownin figure3.1. Thus, IgG and IgE are generally monomeric whereas IgM occurs as a pentamer. IgA occurs predominantly as a monomer in serum and as a dimer in seromucoussecretions. The major antibody in the serum is IgG and, as this is the best-understood antibody in terms of structure and function, we shall consider it first. The other antibody classeswill be considered in relation to IgG.

I H E l g GM 0 t E C U t E In IgG, the Fab arms are linked to the Fc by an extended region of polypeptide chain known as the hinge. This region tends to be exposed and sensitive to attack by proteases that cleave the molecule in to its distinct func-

tional units arranged around the four-chain structure (Milestone 3.1). This structure is represented in greater detail in figure 3.2a.The light chains exist in two forms knownas kappa (r) and lambda (1,).Inhumans, rchains are somewhat more prevalent than )"; in mice, l. chains are rare. The heavy chains can also be grouped into different forms or subclasses,the number depending upon the species under consideration. In humans there are four subclasses having heavy chains labeled ^17,"/2,"'13 and y4 which give rise to the IgG1, IgG2, IgG3 and IgG4 subclasses.In mice, there are again four subclasses denoted IgG1,IgG2a,IgC2b and IgG3. The subclassesparticularly in humans-have very similar primary sequences, the greatest differences being observed in the hinge region. The existence of subclasses is an important feature as they show marked differences in their ability to trigger effector functions. In a single molecule, the two heavy chains are identical as are the two light chains;hybrid molecules have notbeen described. The amino acid sequencesof heavy and light chains of antibodies have revealed much about their structure and function. However, obtaining the sequencesof antibodies is much more challenging than for many other proteinsbecausethe population of antibodies in an individual is so incredibly heterogeneous.The opportunity to do this first came from the study of myeloma proteins. In the human disease known as multiple myeloma, one cell making one particular individual antibody divides over and over again in the uncontrolled way a cancer cell does, without regard for the overall requirement of the host. The patient then possessesenormous numbers of identical cells derived as a clone from the original cell and they all synthesize myeloma proteinthe same immunoglobulin-the which appears in the serum, sometimes in very high concentrations. By purification of myeloma proteins, preparations of a single antibody for sequencing and many other applications canbe obtained. An alternative route to single or monoclonal antibodies arrived with the development of hybridoma technology. Here, fusing individual antibody-forming cells with a B-cell tumor produces a constantly dividing clone of cells dedicated to making the one antibody. Finally, recombinant antibody technologies, developed most recently, provide an excellentsourceof monoclonal antibodies. Sequence comparison of monoclonal IgG proteins indicates that the carboxy-terminal half of the light chain and roughly three quarters of the heavy chain, again carboxy-terminal, show little sequence variation between different IgG molecules. By contrast, the amino-terminal regions of about 100 amino acid residues show considerable sequencevariability in both chains. Within these variable regions there are relatively

3eI

C H A P T E3 R- A N T I B O D I E S

Early studies showed the bulk of the antibody activity in serum to be in the slow electrophoretic fraction termed y-globulin (subsequently immunoglobulin) The most abundant antibodies were divalent, i.e. had two combining sites for antigen and could thus form a precipitating complex (cf. figure 6.2). To Rodney Porter and Gerald Edelman must go the credit for unlocking the secrets of the basic structure of the immunoglobulin molecule. If the internal disulfide bonds are reduced, the component polypeptide chains still hang together by strong noncovalent attractions. However, if the reduced molecule is held under acid conditions, these attractive forces are lost as the chains become positively charged and can now be separatedby gel filtration into larger so-called heavy chains of approximately 55 000 Da (for IgG, IgA and IgD) or 70 000 Da (for IgM and IgE) and smaller light chains of

The clues to how the chains are assembled to form the IgG molecule came from selective cleavage using proteolytic enzymes. Papain destroyed the precipitating power of the intact molecule but produced two univalent Fab fragrnents still capable of binding to antigen (Fab = fragment antigen binding); the remaining fragment had no affinity for antigen and was termed Fc by Porter (fragment crystnllizable).AIter digestion with pepsin a molecule called F(ab'), was isolated; it still precipitated antigen and so retained both binding sites, but the Fc portion was further degraded. The structural basis for these observations is clearly evident from figure M3.1.1.In essence,with minor changes,all immunoglobulin molecules are constructed from one or more of the basic four-chain monomer unlts.

abott24 000 Da.

II,4MUNoGLoBULIN

IMMUNOGLOBULIN

Antigen combining SITE

TMMUNoGLoBULTN

o 'tr

REDUCE ANDACIDIFY

PEPSIN

PAPAIN

W ISOLATED CHAINS

[4 PEPSINFRAGMENTS

PAPAIN FRAGMENTS

Figure M3.1.1. The antibody basic unit (IgG is represented), consisting of two identical heavy and two identical light chains held togetherby interchain disulfide bonds, can be broken down into its constituent polypeptide chains and to proteolytic fragments, the pepsin F(ab'),retaining two binding sitesfor antigen

and the papain Fab with one. After pepsin digestion the pFc'fragment representing

the C-terminal

half of the Fc region is formed

and is held together by noncovalent bonds. The portion of the heavy chain in the Fab fragment is given the symbol Fd. The Nterminal residue is on the left for each chain.

loo

C H A P T E3 R- A N T I B O D I E S

short sequenceswhich show extreme variation and are designated hypervariable regions. There are three of these regions or 'hot spots' on the light chain and three on the heavy chain. Since the different IgGs in the comparison recognize different antigens, these hypervariable regions are expected to be associatedwith antigen recognition and indeed are often referred to as complementarity determining regions (CDRs). The structural setting for the involvement of the hypervariable regions in antigen recognition and the genetic origins of the constant and variable regions will be discussedshortly. The comparison of immunoglobulin sequencesalso reveals the organization of IgG into 12 hom*ology regions or domains each possessingan internal disulfide bond. The basic domain structure is central to an understanding of the relation between structure and function in the antibody molecule and will shortly be taken up below. However, the structure in outline form is shown in figure 3.2b,c.It is seen that the light chain consistsof two domains, one corresponding to the variable sequenceregion discussed above and designated the V. (variable-light) domain and the other corresponding to a constant region and designated the C, (constant-light) domain. The IgC heavy chain consists of four domains, the V' and Crrl domains of the Fab arms being joined to the Crr2 and Crr3 domains of Fc via the hinge. Antigenbindingis a combinedproperty of the V. and V" domains at the extremities of the Fab arms and effector molecule binding a property of the Crr2 and/ or C"3 domains of Fc.

Figure 3.2. The four-chain structure of IgG. (a) Linear representation Disulfide bridges Lnk the two heavy chains and the light and heavy chains A regular arrangement of intrachain disulfide bonds is also found Fragments generated by proteolytic cleavage at the indicated sites are represented. (b) Domain representation. Each heavy chain (shaded) is folded into two domains in the Fab arms, forms a region of extended polypeptide chain in the hinge and is then folded into two domains in the Fc region The light chain forms two domains associated only with a Fab arm Domain pairing leads to close interacti.onof heavy and light chains in the Fab arms supplemented by a disulfide bridge. The two heavy chains are disulfide bridged in the hinge (the number of bridges depending on IgG subclass) and are in close domain-paired interaction at their carboxy-termini (c) Domain nomenclature The hear.y chain is composed of V*,, C*,1,C*,2 and Cr,3 domains The light chain is composed of V, and C, domains All the domains are paired except for the Cr,2 domains, which have two branched Nlinked carbohydrate chains interposed between them. Each domain has a molecular weight of approximately 12,000leading to a molecular weight of -50,000 for Fc and Fab and 150,000 for the whole IgC molecule Antlgen recognition involves residues from the V' and V. domains, complement triggering the Cr,2 domain, leukocyte Fc receptor binding the Cr,2 domain and the neonatal Fc receptor the Cr,2 and C*,3 domains (see text) (Adapted from Burton D R Structure and function of antibodies. In: New Comprehensive Biochemistry series, Vol 17: Molecular genetic of immunoglobulin, F Calabi and M.S. Neuberger (eds) Elsevier,pp 1-50,1.987)

It is also clear (figure 3.2b,c) that all of the domains except for Crr2 are in closelateral or'sideways'association with another domain: a phenomenon described as domain pairing. The Crr2 domains have two sugar chains interposed between them. The domains also exhibit weaker cls-interactions with neighboring domains on the same polypeptide chain. Human IgGl is shown in figure3.2 as a Y-shapedconformation with the Fab arms roughly in the same plane as the Fc. This is the classicalview of the antibody mole-

frogmenl F(ob')2

/7 ?// Yt

LIGHTCHAIN

Popoin cleovoge Pepsin cleovoge

cu2 cH0

HEAVY CHAIN

cs3

Fcfrogmenl

Io*' )

Disulphide bridge belween heovy ondlighlchoins

Disulphide bridges between 2 heowchoins <- HEAVY CHAIN C lermini

CHAPTER 3-ANTIBODIES cule that has adorned countless meetings ads and appears in many company logos. In reality, this is likely just one of many shapesthat the IgG molecule can adopt since it is very flexible as illustrated in figure 3.3. It is believed that this flexibility may help IgG function. Thus Fab-Fab flexibility gives the antibody a 'variable reach' allowing it to grasp antigenic determinants of different spacings on a foreign cell surface or to form intricate immune complexes with a toxin (imagine a Y to T shape change). Fc-Fab flexibility may help antibodies in different environments, on foreign cells for example, to interact productively with common effector molecules. Figure 3.4 shows the complete structure of a human IgGl antibody molecule determined by crystallography. The structure is guite removed from the classical symmetrical Y shape. The Fc is closer to one Fab arm than another and is rotated relative to the Fab arms. This 'snapshot' is simply a of one of the many conformations that the antibody can adoptby virtue of its flexibility. The structural organization of IgG into domains is clearly evident from figures 3.2-3.4. Each of these domains has a common pattern of polypeptide chain folding (figure 3.5). This pattern, the 'immunoglobulin fold', consists of two twisted stacked B-sheetsenclosing an internal volume of tightly packed hydrophobic residues. The arrangement is stabilized by an internal disulfide bond linking the two sheets in a central position (this internalbond is seen in figure3.2a). One sheet has four and the other three anti-parallel B-strands. These strands or framework regions are joined by bends or loops which generally show little secondary structure. Residues involved in the p-sheets tend to be conserved while there is a greater diversity of residues in the loops. The chain folding illustrated in figure 3.5 is for a constant domain. The B-sheets of the variable

Foborm rololion

41 |

Figure 3.4. The structure of a human IgG molecule. This antibody recognizes the gp120 surface glycoproteinofHlV. Theheavychains are shown in blue and the light chains in green. Relative to the classical cartoon of an IgG molecule as aY shape, this'snapshot'of the molecule finds the Fc (bottom)'side on'to theviewer and muchcloser to one Fab arm than the other. (Courtesy of Erica Ollmann Saphire.)

Fobelbow bend

Fc wogging

Figure 3.3. Modes of flexibility in the IgG molecule. These modes have been described from electron microscopic studies (seefigure 3 10) and biophysical techniques in solution. Flexibility in structure probably facilitates flexibility in antigen recognition and effector function triggering.

fold. An anti-parallel threeFigure 3.5. The immunoglobulin stranded p-sheet (red) interacts with a four-stranded sheet (blue) The arrangement is stabilized by a disulfide bond linking the two sheets. The B-strands are connected by helices, bends and other structures. This overall structure is the basis of all Ig and Iglike domatns.

lo,

CHAPTER 3-ANIIBODIES

domain are more distorted than those of the C domain and the V domain possessesan extra loop. Slruclureof Fobfrogmenl The four individual domains are paired in two ways (figure 3.6).The V' and Vrdomains are paired through the two respective three-strand p-sheet layers (red in figure 3.5). The Crrl and C, domains are paired by contact between the two four-strand layers (blue in figure 3.5).The interacting facesof the domains are predominantly hydrophobic and the driving force for domain pairing is thus the removal of these residues from the aqueous environment. The arrangement is further stabilized by a disulfide bond between C.,1 and Crdomains. In contrast to the 'sideways' interactions, the 'longwise' or cls interactions between V' and C*r1 domains and between V, and C, domains are very limited and 'elbow allow bending'. Elbow angles seen in crystal structures varybetween about 137oand 180'.

Theonlibodycombining site Comparison of antibody sequenceand structural data shows how antibodies are able to recognize an enormously diverse range of molecules. Sequence data shows that the variable domains have six hypervariable regions that display great variation in amino acids between different antibody molecules (figure 3.7). Structural data of antibody-antigen complexes reveals that these hypervariable regions, or complementarity determining regions, come together in 3D space to form the antigenbinding site, often also termed the antibody combining site (figure 3.8). (Courtesy of Robyn Stanfield)

Slruclureof Fc For the Fc of IgG (figure 3.9), the two Crr3 domains are classically paired whereas the two C*r2 domains show no closeinteraction, but have interposed between them two branched N-linked carbohydrate chains that have limited contact with one another. The carbohydrate chains are very heterogeneous. The Crr2 domains contain the binding sites for several important effector molecules,complement C 1q and Fc receptorsin particular, as shown. The neonatal Fc receptor, which is important in binding to IgG and maintaining its long half-life in serum, binds to a site formed between Crr2 and Crr3 domains. Protein A, much used in purifying IgGs, also binds to this site.

Thehinge region ondlgGsubclosses

Figure 3.6. The structure of Fab. The heavy chain is shown in yellow and the lightchain in green. The V,, and V. domains (top) are paired by contact between their three-strand faces (red in figure 3.5) and the Cr,1 and C. domains between the four-strand faces (blue in figure 3 5) (Courtsey of Robyn Stanfield)

The term 'hinge' arose from electron micrographs of rabbit IgG, which showed Fab arms assuming different angles relative to one another from nearly 0' (acute Yshaped) to 180' (T-shaped). The Fab was specific for a small chemical group dinitrophenyl (DNP) that could be attached to either end of a hydrocarbon chain. As shown in figures 3.10 and 3.11, different shapes were observed as the Fab arms linked together the bivalent antigen molecule using different Fab-Fab arm angles. Other biophysical techniques have demonstrated hinge flexibility in solution. The function of this flexibility has generally been seen as allowing divalent recognition of variably spaced antigenic determinants. The IgG class of antibody in humans exists as four subclasses and the biggest difference between the subclassesis in the nature and length of the hinge. IgGl has been shown above. IgG3 has a hinge which if fully extended would be about twice the length of the Fc, thereby potentially placing the Fab arms far removed from the Fc. On the other hand, IgG2 and IgG4 have

43 1

C H A P T E3 R- A N T I B O D I E S short compact hinges which probably lead to close approach of Fab and Fc. Interestingly, IgGl and IgG3 are generally superior at mediating effector functions such as complement activation and ADCC relative to IgG2 andIgC4. ALLHUMAN LIGHT CHAINS

80 70 60 850 -3 ao o'

30 20 t0 n

20 30 40 50 60 70 80 90 100 ll0 Residue + ALLHUMAN HEAVY CHAINS

110

r00 90 80 70 _: ou -e 50 >40 30 20 l0

r 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 1 0 0l t o t 2 0 Residue *

Figure 3.7. Amino acid variability in the V domains of human Ig heavy and light chains. Variability, for a given position, is defined as the ratio ofthenumber of differentresidues found atthatoosition compared to the frequency of the most common amino acid ihe CDRs are apparent as peaks in the plot and the frameworks as intervening regions of low variability. (After Dr E.A Kabat )

IHESTRUCTUR AN N FI H E E DF U N C I I O O I M M U N O G T O B UC TT I NA S S E S The immunoglobulin classes (table 3.1) fulfill different roles in immune defence and this can be correlated with differences in their structures organized around the four-chain Ig domain arrangement (figure 3.12).IgG is monomeric and the major antibody in serum and nonmucosal tissues, where it inactivates pathogens directly and through interaction with effector triggering molecules such as complement and Fc receptors.IgM is pentameric, is found in serum and is highly efficient at complement triggering. A monomeric form of IgM with a membrane-tethering sequence is the major antibody receptor used by B lymphocytes to recognize antigen (cf . figure2.7O). IgM differs from IgG in having an extra pair of constant domains instead of the hinge region. IgA exists in three soluble forms. Monomeric and small amounts of dimeric IgA (formed from two monomers linked by an extra polypeptide called ] chain) are found in the serum where they can help link pathogens to effector cells via Fc receptors specific for IgA. Secretory IgA (see below) is formed of dimeric IgA and an extra protein known as secretory component and is crucial in protecting the mucosal surfaces of the body against attack by microorganisms. IgA exists as two subclasses in humans. IgA2 has a much shorter hinge than IgAl and is more resistant to attack by bacterially secretedproteases.IgE is a monomeric antibody typically found at very low concentrations in serum. In fact most IgE is probably bound to IgE Fc receptors on mast cells. Antigen binding to IgE cross-links IgE Fc receptors and triggers an acute inflammatory reaction that can assistin immune defence. This can also lead to unwanted allergic symptoms for certain antigens (allergens). IgE, like IgM, has an extra pair of constant domains instead of the hinge region. Finally IgD is an antibody primarily found on the surface of B cells as an antigen receptor together with IgM, where it likely serves in the control of lymphocyte activation

(b)

Figure 3.8. The proximity of CDRs (variable loops) at the tip of the Fab arms creates the antibody combining site. The V' and V,_ domains are shown from the side (a) and from above (b) The sixCDRs (cf figure3.7) are numbered 1-3 as belonging to the heavy (H) or light (L) chain. (CourtesyofRobyn Stanfield)

loo

C H A P T E3 R- A N T I B O D I E S

Figure 3.9. Structure of Fc of human IgG. The C,.,3domains (bottom) are paired. The Cr,2 domains arenot and have two carbohydrate chains filling some of the space between them Binding sites for the leukocyte as Fc^yRIIIreceptor (red), complement C1q (green) and neonatal Fc receptor FcRn (yellow) are shown. The FcyRIII and FcRn sites were determined by crystallographic studies (Sondermann et al (2000) N ature 406, 267; Martrn et al (2001) Molecular Cell 7, 867) and the C1 q site by mutation analysis (Idusogie e/ al (2000) lournnl of lmmunology L64,4778) (Courtesy of Robyn Stanfield).

Figure 3.10. (A,B) Electron micrograph (x1000000) of complexes f ormed on mixing the divalent DNP hapten with rabbit anti-DNP anti'negative bodies The stain'phosphotungstic acid is an electron-dense penetrates into the spaces between the protein molewhich solution 'light' structure in the electron cules. Thus the protein stands out as a links the Y-shaped antibody molecules to beam. The hapten together form trimers (A) and pentamers (B). The flexibility of the molecule at thehinge regionis evident from the variationin angle of the arms of the 'Y'. (C) As in (A), but the trimers were formed using the F(ab')r antibody fragment from which the Fc structures have been digested by pepsin (x500 000). The trimers can be seen to lack the Fc projections at each corner evident in (A) (After Valentine R.C. & Green N.M. (1967) lournal of Molecular Biology 27,615; courtesy of Dr Green and with the permission of Academic Press,New York )

and suppression. It is monomeric and has a long hinge reSron. The structures of the Fc regions of human IgAl and IgE (C-terminal domains) have been determined and are compared with IgGl in figure 3.13.In all three cases, the penultimate domains are unpaired and have carbohydrate chains interposed between them. Anlibodies ondcomplement The clustering together of IgG molecules, typically on the surface of a pathogen such as a bacterium, leads to the binding of the complement C1 molecule via the hexavalent C1q subcomponent (cf. figure 2.2). This triggers the classical pathway of complement and a number of processesthat can lead to pathogen elimination. The subclasses of IgC trigger with different efficiences. IgCl and IgG3 trigger best; IgG2 is only triggered by antigens at high density such as carbohydrate antigens on a bacterium; and IgG4 does not trigger. Cenerally, the nature of the antigen and its environment seems able to influence how well complement is triggered. IgM triggers by a different mechanism. It is already

Figure 3.11. Three DNP antibody molecules held together as a trirner by the divalent antigen (H). Compare figure 310A When the Fc fragments are first removed by pepsin, the corner pieces are no longer visible (figure 3.10C)

45 1

3-ANTIBODIES CHAPTER

Figure 3.12. Schematic structures of the antibody classes. The two heavy chains are shown in dark and pale biue (two colors to highlight chain pairing; the chains are identical) and the light chains in grey. Nlinked carbohydrate chains (branched structures) are shown in blue and Olinked carbohydrates (linear structures) in green The heavy chain domains are designated according to the class ofthe heavy chain, e g C/2 for the Cr,2 domain of IgG, etc For IgG, IgAand IgD, the Fc is connected to the Fab arms via a hinge region; for IgM and IgE an extra pair of domains replaces the hinge IgA, IgM and IgD have tailpieces at the C termini of the heavy chains IgA occurs in monomer and dimer forms. IgM occurs as a pentamer (a) IgG 1. The other human IgG subclasses(and IgGs ofmost other species) have this same basic structure but differ particularly in the nature and length of the hinge. (b) IgA1. The structure resembles IgGl but with a relatively long hinge bearing O-linked sugar chains The Fc also shows some differences from IgGl (seefigure 3 13) In IgA2, the hinge is very short and, in the predominant allotype, the light chains are disulfide linked not to the heavy chainbut to one another (c) IgM rnonomeric unit. This representation relies greatly on comparison of the amino acid sequences of p and yheavy chains (d) IgE Themolecule issimilarto the monomeric unit of IgM (e) IgD. The hinge canbe divided into a region rich in charge (possibly helical) and one rich in O-linked sugars. The structure of the hinge may be much less extended in solution than represented schematically here. It is however very sensitive to proteolytic attack so that serum IgD is unstable. Mouse IgD has a structure very different to that of human IgD in contrast to the general similarity in structures for human and mouse Igs. (f) Secretory IgA (seealso figure 3 19). (g) Pentameric IgM. The molecule is represented as a planar star shape. One monomer unit is shown shaded as in (c) For clarity the carbohydrate structures have been omitted in (f) and (g). The Fab arms can likely rotate out of the plane about their twofold axis (seealso figure 3 14)

lgGl

lgAl

lgM

lgE

lgD

lgA secretory

multivalent (pentavalent) but occurs in an inactive form. Binding to multivalent antigen appears to alter the conformation of the IgM molecule to expose binding sitesthat allow C1q to bind and the classicalpathway of complement to be triggered. Electron microscopy studies suggest the conformational change is a'star' to 'staple' transition, in which the Fab arms move out of the plane of the Fcs (figure 3.1a).IgM antibodies tend to be

lgM Penfomeric

of low affinity as measured in a univalent interaction, e.g. binding of IgM to a soluble monomeric molecule or binding of an isolated Fab from an IgM to an antigen. However, their functional affinity (avidity) can be enhanced by multivalent antibody-antigen interaction (see p. 92) and it is precisely under such circ*mstances that they are most effective at activating comDlement.

lou

CHAPTER 3-ANTIBODIES

Table 3.1. The human immunoglobulins.

lsc(v)

Monomer

v2,L2

(-12 mg/mi), Serum lissues

lgG3> lgcI >>lgc2>>lgc4 (clossicol)

Monomer

u2,L2

Yes (monnose-binding

Dimer

(u2,L2)r,J

(-3 mg/ml): Serum 90%monomer, l 0 %d i m e r

Secrelory

(u2,L2)r,J,SC

secretions, Seromucous teors milk,colosfrum,

ISM(rr)

Penfomer

J Qt2,L2)u,

(-l .5mg/ml) Serum

Yes(clossicol)

lgE(e)

Monomer

e 2 ,L 2

(0 05 pg/ml) Serum

lgD(6)

Monomer

62,L2

(30pg/ml) Serum

lgGl lgG2 lgG3 lgG4

lgA(c)

lgAl lgA2

Figure 3.13. The structures of the Fc regions of human IgGl, IgE and IgA1. The structures shown were determined by crystallographic analysis of Fcs in complex with Fc receptors. One heavy chainis shown in red, the other in yellow and the N-linked carbohydrate chains that are interposed between the penultimate domains are shown in blue. For IgE, the structure does not include the Ce2 domains, which form

Anlibodiesond humonleukocyteFcreceplols Specific human Fc receptors have been described for IgC, IgA and IgE (table 3.2).The receptorsdiffer in their specificities for antibody classesand subclasses,their affinities for different association states of antibodies (monomer vs associated antigen-complexed antibody), their distributions on different leukocyte cell types and their cellular signaling mechanisms.Most of the leukocyte Fc receptors are structurally related, having evolved as members of the Ig gene superfamily. Each comprises a unique ligand binding chain (cr chain),

lectin)

part of the Fc region ofthis antibody. For IgA1, the Nlinked sugars are attached at a position quite distinct from that for IgGl and IgE Also the tips of the Ccr2 domain are joined by a disulfide bridge (Courtesy of Jenny Woof; after Woof J M. and Burton D.R. (2004) Nature Reaiews lmmunology4,89-99 )

which is often complexed via its transmembrane region with a dimer of the common FcR-1 chain. The latter plays a key role in the signaling functions of many of the receptors. FcR-ychains carry immunoreceptor tyrosine-based activation motifs (ITAMs) in their cytoplasmic regions, critical for initiation of activatory signals. Some receptor o(chains carry their own ITAMs in their cytoplasmic regions, while others bear the immunoreceptor tyrosine-based inhibitory motifs (ITIMs). For IgG, three different classes of human leukocyte FcyRs have been characterized, most with several variant forms. In addition, the neonatal Fc receptor

CHAPTER 3-ANTIBODIES

Figure 3.14. Structural changes in IgM associatedwith complement activation. (a) The'star'conformation EM of an uncomplexed IgM protein shows a'star'shaped conformation (cf figure 3.12g).(b) The 'staple' conformation EM of a specificsheepIgMboundto a Solmonello paratyphi flagellum as antigen suggests that the 5 F(ab'), units and Cp2 domains have been dislocated relative to the plane of the Fcs to produce a 'staple'or'crab-like'conformation Complement C1 is activated onbinding to antigen-complexedIgM (staple),but interactsonly very weakly, yielding no sigmficant activation, with free IgM (star), implying that the dislocation processplays an important role in complement activation It is suggestedthat movement of the Fabs exposes a C1q binding site on the Cp3 domains of IgM. This is supported by observations that an Fc5 molecule, obtained by papain digestion of IgM, can activate complement directly in the absenceof antigen (Electronmicrographsarenegativelystainedpreparations of magnificatron ^ A . ^2 110o, i.e. 1 mm represents0.5nm; kindly provided by DrsA Feinstein andEA Munn)

FcRn also binds IgG and will be dealt with later. FcyRI (CD64) is characterized by its high affinity for monomeric IgG. It is also unusual in that it has three extracellular Ig-like domains in its ligand-binding chain, while all other Fc receptorshave two. FcIRI is constitutively present on monocytes, macrophages and dendritic cells, and is induced on neutrophils and eosinophils following their activation by IFNy and GCSF @ranulocyte colony-stimulating /actor). Conversely,FcyRIcan be downregulated in responseto IL-4 and IL-13. Structurally, it consists of an IgG-binding o chain and a y chain hom*odimer containing ITAMs. It binds monomeric IgG avidly to the surface of the cell thus sensitizing it for subsequent encounter with antigen. Its main roles are probably in facilitating phagocytosis, antigen presentation and in mediating extracellular killing of target cells coated with IgG antibody, a processreferred to as antibody-dependent cellular cytotoxicity (ADCC; p. 32). FcftII (CD32)binds very weakly to monomeric IgG but with considerably enhanced affinity to associated IgC as in immune complexes or on an antibody-coated target cell. Thereforecellsbearing FcyRIIare able to bind antibody-coated targets in the presenceof high serum concentrations of monomeric IgG. Unlike the single

47 1

isoform of FcyRI, there are multiple expressed isoforms of FcyRII which collectively are present on the surface of most types of leukocyte (table 3.2). The binding of IgG complexes to Fc^yRII triggers phagocytic cells and may provoke thrombosis through their reaction with platelets. The FcyRIIa mediates phagocytosis and ADCC whilst the FcyRIIb2 (and FcyRIII) efficiently mediates endocytosis leading to antigen presentation. FcyRIIbl on B-cells does not endocytose immune complexes and therefore B-cells principally present only their cognate antigen following ligation and endocytosis of the B-cell receptor (BCR). In fact, the FcyRIIb molecules have cytoplasmic domains which contain ITIMs and their occupation leads to downregulatior of cellular responsiveness.In the caseof the B-cell this mediates the negative-feedback effect of IgG on antibody production (cf. p.272). Thus, whereas the isoforms on phagocytic cells are associated with ligand internalization, that on the B-cell fails to internalize but concentrates instead on lymphocyte regulation. FcyItIII (CD16) also binds rather poorly to monomer IgG but has low to medium affinity for aggregated IgG. The two FcyRIIl genes encode the isoforms FcyRIIIa and FcyRIIIb whichhave a medium and low affinity forIgG, respectively. FcyRIIIa is found on most types of leukocyte, whereas FcyRIIIb is restricted mainly to neutrophils and is unique amongst the Fc receptors in being attached to the cell membrane by a glycosylphosphatidylinositol (GPI) anchor rather than a transmembrane segment. FcyRIIIa is known to be associated with the y chain signaling dimer on monocytes and macrophages,and with either ( and/or ychain signaling molecules in NK cells, and its expression is upregulated by transforming growth factor B (TGFp) and downregulated by IL-4. With respect to their functions, FcyRIIIa is largely responsible for mediating ADCC by NK cells and the clearance of immune complexes from the circulation by macrophages. For example, the clearance of IgG-coated erythrocytes from the blood of chimpanzees was essentially inhibited by the monovalent Fab fragment of a monoclonal anti-FcyRIII. FcyRIIIb cross-linking stimulates the production of superoxide byneutrophils. For IgE, two different FceRshave been described. The binding of IgE to its receptor FceRI is characterized by the remarkablyhigh affinity of the interaction, reflecting a very slow dissociation rate (half-life of complex is -20 hours). FceRI is a complex comprising a ligand-binding a chain structurally related to those of FcyR, a B chain, and the FcR-ychain dimer. Contact with antigen leads to degranulation of the mast cells with release of preformed vasoactive amines and cytokines, and the synthesis of a variety of inflammatory mediators derived

lot

CHAPTER 3-ANIIBODIES

Table3.2. HumanleukocyteFcreceptors.(FromWoofJ.M.&BurtonD.R.(2004)NatureReaiewsImmunology4,S9.)

MW(kDo)

50-70

50-80

Mojor Fc'yRlo isoforms expressed

FsEllo

Allolypes

FqRllb

bind

Afiinilyfor

High

Low(<107)

Low(

monomer l g( M ' )

(108- r0s)

Signoling

Ycn0rn ITAM

IgG2doesn'l

humonlg*

motif

Fc'yRlllb

Low(

45-50

FcrRl

Fc€R llo

50-70 FceRllb

FccRlo

NAI OndNM lgc3> I >>2 >4

for

FsyRlllo

lgc3> I =2 l g c 3 > l > > > 2 lgG4doesn'i lgG4doesn'l bind bind

Sp€cificity l g G = l 3>4

FsEllc

45-65

lgGl = 3 >>>

lgE

Serum rgA'l= 2, =SlgA2 SlgAl

Veryhigh

Low (

Medium(l 07)

crype

C-type leclin

ychoinITAM

a_A

Low(

(l 07) l\4edrum

Low(

(t 0r0)

0 choinITAM

schoinlTlM

qcnotn

ych0ln

ITA[4

ITAM

Nosignoling TchoinITAM pchoinolso molit in Anchored

presenlbul

membrone

Ioleuncleor

leclin

vr0grycon phosphotidylinosilol(GPl) linkoge Cellulor

Monocyles,

Monocytes, mocrophoges,

plot€lets, dislribulion mocrophoges, neulrophils, DC,neulTophils Long€rhons cells (lFNyslim),

TB-cells,

Neutrophils,

mocrophoges, mocrophoges, NKcells,

eosinophils

bosophils,

celts,

monocytes,

B-cells

(lFNyslim)

L0ngerh0ns

monocyies,

some

cells,octivoled

eosinophils,

mocrophoges,

monocylos

mocrophoges eosinophils,

l\4onocytes,

l\.,lonocytes, neulrophils, B-cells

eosinophils

Mocrophoges, Neulrophils, Moslcells,

T6T-cells, somg monocyles

B-cells

Kuplfffcells,

(lFNystim)

someDC

from arachidonic acid (cf. figure 1.15).This process is responsible for the symptoms of hay fever and of extrinsic asthma when patients with atopic allergy come into contact with the allergen, e.g. grass pollen. The main physiologicalrole of IgE would appear to be protection of anatomical sites susceptible to trauma and pathogen entry by local recruitment of plasma factors and effector cells through the triggering of an acute inflammatory reaction. Infectious agents penetrating the IgAdefenses would combine with specific IgE on the mast cell surface and trigger the release of vasoactive agents and factors chemotactic for polymorphs, so leading to an influx of plasma IgG, complement, neutrophils and eosinophils (cf. p. 33). In such a context, the ability of eosinophils to damage IgG-coated helminths and the generous IgE response to such parasites would constitute an effective defense. The low affinity IgE receptor FceRII (CD23) is a Ctype (calcium-dependent) lectin. It is present on many different types of hematopoietic cells (table 3.2). Its primary function appears to be in the regulation of IgE

synthesis by B-cells, with a stimulatory role at low concentrations of IgE and an inhibitory role at high concentrations. It can also facilitate phagocytosis of IgE opsonized antigens. For IgA, FcoRI (CD89), is the only well characterized Fc receptor. Its ligand-binding cr chain is structurally related to those of the FcyRsand FceRI but represents a more distantly related member of the family. In fact, it shares closer hom*ology with members of a family including natural killer cell immunoglobulin-like receptors (KIRs), leukocyte Ig-like receptors (LIR/ LILR/ILIs) and the platelet-specific collagen recePtor (GPVD. FccRI is present on monocytes, macrophages, neutrophils, eosinophils and Kupffer cells' The crosslinking of FccrRI by antigen can activate endocytosis, phagocytosis, inflammatory mediator release and ADCC. Expression of FccrRI on monocytes is strongly upregulated by bacterial polysaccharide. Crystal structures are available for FcyRIIa, FcyRIIb, FcyRIIIb, FceRI and FcoRI (figure 3.15). In all cases, the structures represent the two Ig-like extracellular

C H A P T E3 R- A N T I B O D I E S domains of the receptor crchain, termed D1 (N-terminal, membrane distal) and D2 (C-terminal, membraneproximal). No structure is yet available for the cytoplasmic portions of any receptor. The extracellular regions of FcyRIIa/b, FcIRIII and FceRI are seen to share the same overall heart-shaped structure and are so similar that they can be readily superimposed. Despite the basic sequence similarity between FccrRIand these receptors, the IgA receptor turns out to have a strikingly different structure. While the two individual domains of the FcsRI extracellular portion fold up in a similar manner to those of the other receptors, the arrangement of the domains relative to each other is very different. The domains are rotated through -180' from the positions adopted in the other Fc receptors, essentially inverting the D1-D2 orientations. Crystallographic studies of antibody-Fc receptor complexes have revealed how antibodies interact with leukocyte Fc receptors (figure 3.16).For the IgG-FcIRIII interaction, the D2 membrane-proximal domain of FcyRIII interacts with the top of the Crr2 domains and the bottom of the hinge. This requires the antibody adopt a

Figure 3.15. Structures of human leukocyte Fc receptors, In each case,a similar view of the receptor is shown, in its uncomplexed state D1, membrane distal; D2, membrane proximal domain For the FcyRsand FceRI, the Fc binding site is present at the 'top' of the D2 domain For FccrRI,the D1-D2 domain arrangement is reversed and the Fc interaction site is present at the top of the D1 domain. (Courtesy of Jenny Woof.)

4el

'dislocated' conformation in which the Fab arms are rotated out of the plane of the Fc. One consequence of this mode of interaction, recognized many years ago/ is that it promotes close approach of the target cell membrane (upwards on the page) to the effector cell membrane. This may favor effector cell activity against the target cell. Given the similiarities between FcyRI, FcyRII and FcyRIII, it is likely that all three FcRs share a common mode of binding to IgG. Indeed, this mode of binding seems also to be shared by IgE binding to the FceRI receptor although the Ce2-Ce3 domain linker region replaces the hinge contribution to receptor binding. By contrast, IgAbinds to the FcoRI receptor at a site between Ccr2 and Cu3 domains. This mode of binding permits an IgA:FcR stoichiome try of 2 : 1 whilst the stoichiometry for IgG and IgE in the above complexes is 1 : 1. The significance of these differences in the modes of binding is not understood at this time. Antibodies0ndlhe neonololFcreceptor An important Fc receptor for IgG is the neonatal recep-

luo

C H A P T E3 R- A N T I B O D I E S

tor, FcRn.This receptor mediates transport of IgG from mother to child acrossthe placenta (figure 3.17).Such antibody, surviving for some time in the blood of the newborn child, is believed to be important in directly protecting the child from pathogens. Furthermore, the presence of maternal antibody has been proposed to help the developmentof cellular immunity inthe young child by attenuating pathogen challenge rather than stopping it completely. FcRn may also be important in transporting maternal IgG from mother's milk across the intestinal cells of the young infant to the blood. Equally, FcRn is crucial in maintaining the long halflife of IgG in serum in adults and children. The receptor binds IgG in acidic vesicles (pH . 6.5) protecting the molecule from degradation, and then releasing the IgG at the higher pH of 7.4inblood. Structural studies have revealed the molecularbasis for FcRn activity. FcRn is unlike leukocyte Fc receptors and

instead has structural similarity to MHC classI molecules. It is a heterodimer composed of a pr-microglobulin chain noncovalently attached to a membrane-bound chain that includes three extracellular domains. One of these domains, including a carbohydrate chain, together with Br-microglobulin interacts with a site between the Crr2 and Cr,3 domains of Fc (figure 3.18). The interaction includes three salt bridges made to histidine (His) residues on IgG that are positively charged at pH < 6.5.At higher pH, the His residues lose theirpositive charges,the FcRn-IgC interaction is weakened and IgG dissociates.

lgA Secretory IgA appears selectively in the seromucous secretions, such as saliva, tears, nasal fluids, sweat, colostrum, milk, and secretions of the lung, genitourinary and gastrointestinal tracts, where it clearly has the job of

Figure 3.16. Structures of antibody-Ieukocyte Fc receptor intetactions. The left hand side and middle columns show views of the crystal structures of the complexes of the FcRs with their respective Fc ligands. The extracellular domains of the receptors are shown in blue while one heavy chain of each Fc region is shown in red and the other in gold In the left hand column, each Fc region is viewed face on The similarity between the IgG-FcyRIII and IgE-FceRI interaction is striking while the IgA-FcuRI interaction is quite different in terms of the sites involved and the stoichiometry. The middle column shows a view where the D2 domains of each of the receptors are positioned so that their Ctermini face downwards. Here the Fc r e g i o n so f I g C a n d l g E a r es e e ni n a horizontal position from the side. For the IgA interaction only one receptor molecule is shown The right hand column shows a schematic representation of the receptors and their intact ligands from the same viewpoint as the images in the middle column Light chains are shown in pale yellow The necessity for dislocation of IgG and IgE to allow positioning of the Fab tips away from the receptor-bearing cell surface is apparent (Courtesy of Jenny Woof )

5rl

C H A P T E3 R- A N T I B O D I E S

9./

FETAL ACQUISITIoN 0F N/ATERNAL lgG

lvloternol )irculolion

PLACENTA

NEONATAL ACOUISITION OF MILK IqG

Y-

TRANSPORT OFIgG BIDIRECTIONAL

I tso

I lsc

Figure 3.17. Function of the neonatal receptor for IgG (FcRn) on epithelial cells. (a) The FcRn receptor is present in the placenta where it fulfills the important task of transferring maternal IgG to the fetal circulation This will provide protection prior to the generation of immunocompetence in the fetus Furthermore, ii is self-evident that any infectious agent which might reach the fetus ln ufero will have had to have passed through the mother first, and the fetus will rely upon the mother's immune system to have produced IgG with appropriate binding specificities This maternal IgG also provides protection for the neonate, because it takes some weeks following birth before the transferred IgG is eventually all catabolized (b) It has been clearly demonstrated in rodents, although remains speculative in humans, that there is epithelial transport of IgG from maternal milk across the intestinal cells of the newborn IgG binds to FcRn at pH 6 0, is taken into the cell within a clathrin-coated vesicle and released at the pH of the basal surface The directional movement of IgG is achieved by the

defending the exposed external surfaces of the body against attack by microorganisms. This responsibility is clearly taken seriously since approximately 40 mg of secretory IgA/kg body weight is transported daily through the human intestinal crypt epithelium to the mucosal surface as compared with a total dally prodwction of IgG of 30 mg/kg. The IgA is synthesized locally by plasma cells and dimerized intracellularly together with a cysteine-rich polypeptide called J chain, of molecular weight 15 000. Dimeric IgA binds strongly to a receptor for polymeric Ig (poly-Ig receptor/ pIgR, which also binds polymeric IgM) present in the membrane of mucosal epithelial cells. The complex is then actively endocytosed, transported across the cytoplasm and secreted into the external body fluids after cleavage of the pIgR peptide chain. The fragment of the receptor remaining bound to the IgA is termed secretory component and the whole molecule, secretory IgA (figure 3.19).

lsolypes,0llolypesond idiotypes:0nlibodyvorionls The variability of antibodies is often conveniently divided into three types. Isotypes arevariants presentin all healthy members of a species: immunoglobulin

Y Y lgG

:Y Y-' ) Releosed M i t k m0Tern0l l g G

BASAL INTESTINAL CELLOFADULT SURFACE pH1.4 Acidicendosome

.-Irwi

.-I|;WI

l\4o1ernol

BASAL Fetol LUMEN LUMEN INTESTINAL CELLOFNEONATE SURFACE circulolion p H 7 . 4 p H6 , 0 p H6 , 0

Antigen

a--->

) Releosed ( milk

Releosed complex

asymmetri.c pH effect on Ig-receptor interaction Knockout mice lacking FcRn are rncapable of acqurring maternal Ig as neonates Furthermore, they have a grossly shortened IgG half-life, consistent wrth the role of FcRn as a protective recePtor that prevents degradation of IgG and then recycles it to the circulation The IgG half-life is unusually long compared with that of IgA and IgM and this enables the response to antigen to be sustained for many months following infection (c) An additional role of FcRn may be as a bidirectional shuttle receptor IgG brnding on the nonluminal side of the epithelial cell may occur, follow ing endocytosis, within the more favorable pH of acidic endosomes This receptor may thus provide a mechanism for mucosal immunosurveillance, traveling back and forth across the epithelial cell, delivering IgG to the intestinal lumen and then returning the same antibodies in the form of immune complexes for the stimulation of B-cells by follicular dendritic cells

Figure 3.18. Structure of the rat neonatal Fc receptor binding to the Fc of IgG. A heterodimeric Fc (hdFc) is shown with the FcRn binding chain in red and the nonbinding chain in orange The orange chain has been mutated at several positions to eliminate FcRn binding. If the normal hom*odimeric molecule is used then oligomeric ribbon structures are created in which FcRn dimers are bridged by Fcs thereby preventing crystallizatron The ihree domains of FcRn are shown in dark blue (two are close together at the bottom of the picture in this view) and Br-microglobulin in light blue A portion of the cr, domain, an Nlinked carbohydrate attached to this domain and the C-terminus of Brmicroglobulin form the FcRn side of the interaction site Residues at the C,,2/Cn3 domain interface form the Fc side of the interaction site (After Martin W L. et aI (2001,) MolecularCellT,867 )

lu,

C H A P T E3 R_ A N T I B O D I E S

classesand subclassesare examplesof isotypic variation involving the constant region of the heavy chain. Allotypes are variants that are inherited as alternatives (alleles) and therefore not all healthy members of a species inherit a particular allotype. Allotypes occur mostly as variants of heavy-chain constant-region genes/in man in all four IgG subclasses,IgA2and IgM. The nomenclature of human immunoglobulin allotypes is based on the isotype on which it is found (e.g. G1m defines allotypes on an IgGl heavy chain, Km defines allotypes on r light chains) followed by an accepted WHO numbering system. The variable region of an antibody can act as an antigen, and the unique determinants of this region that distinguish it from most other antibodies of that species are termed its idiotypic determinants. The idiotype of an antibody, therefore, consists of a set of idiotypic determinants which individually are called idiotopes. Polyclonal anti-idiotypic antibodies generally recognize a set of idiotopes whilst a monoclonal anti-idiotype recognizesa single idiotope. Idiotypes are usually specific for an individual antibody clone (private idiotypes) but are sometimes shared between different antibody clones (public, recurrent or cross-reacting idiotypes). An anti-idiotype may react with determinants distant from the antigen binding site, it may fit the binding site and expressthe image of the antigen or it may react with determinants close to the binding site and interfere with antigen binding. Sequencing of an anti-idiotypic

antibody generated against an antibody specific for the polypeptide GAI antigen in mice revealed a CDR3 with an amino acid sequence identical to that of the antigen epitope, i.e. the anti-idiotype contains a true image of the antigenbut this is probably the exception rather than the rule.

Y IVERSIIY G E N E I I CO S FA N I I B O D D A N DF U N C I I O N genes Anlibodies oleproduced recombinolion bysomolic The immunoglobulin repertoire is encoded for by multiple germline gene segments that undergo somatic diversificationindevelopingB-cells. Hence, althoughthe basic components needed to generate an immunoglobulin repertoire are hherited, an individual's mature antibody repertoire is essentially formed during their lifetime by alteration of the inherited germline genes. The first evidence that immunoglobulin genes rearrange by somatic recombination was reported by Hozumi and Tonegawalnl976 (Milestone 3.2).Because somatic recombination involves rearrangement of DNA in somatic, in contrast to gamete cells, the newly recombined genes are not inherited. As a result, the primary immunoglobulin repertoire will differ slightly from one individual to the next, and will be further modified during an individual's lifetime by their exposure to different antigens.

GLANDULAR EPITHELIAL CELL

Figure 3.19. IgA secretion at the mucosal surface. The mucosal cell sl,nthesizes a receptor for polymeric Ig (pIgR) which is inserted into the basal membrane. Dimeric IgAbinds to this receptor and is transported via an endocytic vacuole to the apical surface. Cleavage of the receptor releases secretory IgAstill attached to part of the receptor termed the secretory component. Note how the receptor cleavage introduces an asymmetry, which drives the transport of IgA dimers to the mucosal surface (in quite the opposite direction to the transcytosis of milk IgG in figure3.77).

53 I

CHAPTER 3-ANTIBODIES

Susumu Tonegawa was awarded the 1987 Nobel Prize in Physiology or Medicine for 'his discovery of the genetic prin-

probes specific for (i) both variable and constant regions and (ii) only the constant region, whereas both probes localized to a single band when hybridized to DNA from an antibody-

ciple for generation of antibody diversity.'ln his 1976paper, Tonegawa used southern blot analysis of restriction enzyme digested DNA from lymphoid and nonlymphoid cells to show that the immunoglobulin variable and constant genes are distant from each other in the germline genome. Embryo DNA showed two components when hybridized to RNA

producing plasmacytoma cell. He proposed that the differential hybridization pattems could be explained if the variable and constant genes were distant from each other in germline DNA, but came together to encode the complete immr.rnoglobulin gene during lymphocyte differentiation.

lgH,chromosome l4:

Vil

Vr,

-Loo -Vr*

Dr,

Jr,-Jffi

DrD

lgl,,chromosome 22

---l L, V^,

L, V,

^-Lo -V*o

J^, Cr,

J*u C*

lgr, chromosome 2:

--l L' V"

L, V,

-Lo

-V*oo

Figure 3.20. The human immunoglobulin loci. Schematics of the human heavy chain (top) and light chain lambda (middle) and kappa (bottom) loci are shown The human heavy chain locus on chromosome 14 consists of approximately 38-46 functional V" genes,27 D' genes and six J' genes, which are organized into clusters upstream of the constant regions The human larnbda locus on chromosome 22 con-

Theimmunoglobulin votioblegenesegments ondloci The variable light and heavy chain loci in humans contain multiple gene segments which are joined, using somatic recombination, to produce the final V region exon. The human heavy chain variable region is constructed from the joining of three gene segments, V (variable), D (diversity), and J (joining), whereas the light chain variable gene is constructed by the joining of two gene segments, V and ]. There are multiple V, D and j segments at the heavy chain and light chain loci, as illustrated in figure 3.20. The human V" genes have been mapped to chromosome 14, although orphan IgH genes have also been identified on chromosomes 15 and 16. The human V' locus is highly polymorphic, and may have evolved through the repeated duplication, deletion and recombination of DNA. Polymorphisms found within the germline repertoire are due to the insertion or deletion of gene segments or the occurrence of different alleles of

J.-J*u

sists of approximately 30 functional V^ genes and five functional J^ gene segments, with each J segment followed by a constant segment. The human kappa locus on chromosome 2 consists of 3tl-40 functional V* genes and five functional J* genes, with the J segments clustered upstream of the constant region. (L, leader sequence )

the same segment. A number of pseudogenes, ranging from those that are more conserved and contain a few point mutations to those that are more divergent with extensive mutations, are also present in immunoglobulin loci. There are approximately one hundred human V' genes, which can be grouped into seven families based on sequence hom*ology, and these families canbe further grouped into three clans. Members of a given family show approximately 80% sequence hom*ology at the nucleotide level. The functional heavy chain repertoire is formed from approximately38-46 functionalVrt genes,27 D Hgenesand six Is genes. The human lambda locus maps to chromosome22, with approximately 30 functional V^ genes and five functional 11 gene segments. The V^ genes can be grouped into 10 families, which are further divided into seven clans. The human kappa locus on chromosome 2 is composed of a total of approximately 34-a0 functional VK genes and five functional j* genes. However, the kappa locus contains a large duplication of most of the V* genes, and most of

luo

C H A P T E3 R- A N T I B O D I E S

the V* genes in this distal cluster, although functional are seldom used. The immunoglobulin loci also contain regulatory elements (figure 3.21), such as a conserved octamer motif and TATA box in the promoter regions. Leader sequencesare also found upstream of the variable segments, and enhancer elements are also present within the loci to facilitate productive transcription.

V(D)Jrecombinolion ondcombinoloriol diversity The joining of these gene segments, illustrated in figure 3.22,is known as V(D)J recombination. V(D)j recombination is a highly regulated and ordered event. The light chain exon is constructed from a single V-to-J gene segment join. However, at the heavy chain locus, a D-

PROMOTER

segment is first joined to a j-segment, and then the Vsegment is joined to the combined DJ sequence.The rearranged DNA is transcribed, the RNA transcript is spliced to bring together the V region exon and the Cregion exon, and lastly the spliced mRNA is translated to produce the final immunoglobulin protein. Numerous unique immunoglobulin genes can be made by joining different combinations of the V D and J segmentsat the heavy and light chain loci. The creation of diversity in the immunoglobulin repertoire through this joining of various gene segmentsis known as combinatorial diversity. Additional diversity is createdby the pairing of different heavy chains with different lambda or kappa light chains.For example, the potential heavy chain repertoire is approximately 40Y nx27 Dnx 6Jn= 6.5 x 103different combinations. Similarly, there

lg VARIABLE REGI0N

Figure 3.21 Regulatory elements of immunoglobulin loci. Each VD/ segment encoding the variable region is associated with a leader sequence Closely upstream is the TATA box of the promoter, which binds RNA polymerase II and the octamer motif which is one of a number of short sequences which bind transacting regulatory transcription factors The V region promoters are relatively inactive and only association with enhancers, which are also composites of short sequence motifs capable of binding nuclear proteins, will increase the

transcription rate to levels typical of actively secreting B-cells The enhancers are near to the regions that control switching from one Ig class constant region to another, e.g IgM to IgG (figure 3 27) Prirnary transcripts ate imtiated 20 nucleotides downstream of the TATA box and extend beyond the end of the constant region. These are spliced, cleaved at the 3' end and polyadenylated to generate the translatable mRNA

DNA Germlrne

Somoticolly DNA recombined

v Tronscription produces 0 pflmory RNAmessoge

polyA toil

A toil PolY

Figwe 3.22. Overview of V(D)j recombination. Diversity (D) and joining (J) gene segments in the germline DNA are joined together through somatic recombination at the heavy chain locus The variable (V) gene segment is then ioined to the recombined D-J gene, to produce the fully recombined heavy chain exon At the light chain ]oci, somatic recombination occurs with V and J segments only. The recombined DNA is transcribed, and the primary RNAtranscript is then spliced, bringing together the variable and constant regions The spliced nRNAmoiecule is translated to produce the immunoglobulin orotein The contribution of the different gene segments to the polypeptide sequence is illustrated for one of the heavy chains H, hinee

C H A P T E3 R- A N T I B O D I E S are approximately 165(33 V^ x 5 J^)and 200 (40V* x 5 ]*) different combinations, for a total of 365light chain combinations. If we consider that each heavy chain could potentially pair with each light chain, then the diversity of the immunoglobulin repertoire is quite large, on the order of 106possible combinations.Additional diversity is also generated during gene segment recombination and by somatic hypermutation, as explained in the following sections. In this manner, although the number of germline gene segments appears limited in size, an incredibly diverse immunoglobulin repertoire can be generated.

Recombinotion signol sequences The recombination signal sequence (RSS) helps to guide recombination between appropriate gene segments. The RSS (figure 3.23) is a noncoding sequence which flanks coding gene segments.It is made up of a conservedheptamer and nonamer sequences,which are separated by an unconserved 72- or 23-nucleotide spacer. Efficient recombination occurs between segments with a 12-nucleotide spacer and a 23-nucleotide spacer.This'12123'rule helps make certain that appropriate gene segmentsarejoined together. At the V' locus, the V and J segments are flanked by RSSs with a 23-nucleotide spacer, whereas the D segments are flanked by RSSswith a 12-nucleotide spacer. At light chain loci, the V* segments are flanked by RSSs with 12-nucleotide spacers,I* segments are flanked by RSSswith 23-nucleotide spacers,and this arrangement is reversedin the lambda locus. - unconserved - nonomer - 3' RSS: 5' - heplomer spocer n - I 2 bosepoirspocer - ACAAAMCC - 3' p 5'- CACAGTG - 23 bosepoirspocer - ACAMMCC - 3' 5'- CACAGTG

W Figure 3.23. The recombination signal sequence. The recombination signalsequence(RSS)ismadeup of conservedheptamerandnonarner sequences, separated by an unconserved 12- or 23-nucleotide spacer. Efficient recombination occurs between segments with a 12-nucleotide spacer and a 23-nucleotide spacer. RSSs with 23-nucleotide spacers flank the V and J segments of the heavy chain locus, the J segments of the kappa locus and the V segments of the lambda locus, whereas RSSs with l2-nucleotide spacers flank the D segments of the heavy chain locus, the V segments of the kappa locus and the J segments of the lambda locus

55 I

mochinery Ihe lecombinose The V(D)J recombinase is a complex of enzymes that mediates somatic recombination of immunoglobulin gene segments (figure 3.24). The gene products of recombination-activating genes 7 and 2, RAG-1 and RAG-2, are lymphocyte-specific enzymes essential for V(D)| recombination. In the initial steps of V(D)] recombination, the RAG complex binds the recombination signal sequencesand, in association with high mobility group (HMG) proteins that are involved in DNA bending, the two recombination signal sequences are brought together. In contrast to the lymphoid-specific RAG enzymes, HMG proteins are ubiquitously expressed. Next, a single-stranded nick is introduced between the 5' heptameric end of the recombination signal sequence and the coding segment. This nick results in a free 3'OH group, that attacks the opposite, anti-parallel DNA strand in a transesterification reaction. This attack gives rise to a double strand DNAbreak that leads to the formation of covalently sealed hairpins at the two coding ends and the formation of blunt signal ends. At this stage a post-cleavage complex is formed, in which the RAG recombinase remains associated with the DNAends. The DNAbreak is finallyrepairedbynonhom*ologous end-joining machinery. The recombination signal sequences are joined precisely to generate the signal joint. By contrast, nucleotides can be lost or added during repair of the coding ends (figure 3.25). Junctional diversity is the diversification of variable region exons due to this imprecise joining of the coding ends. First, a small number of nucleotides are often deleted from the coding end by an unknown exonuclease.Also, junctional diversify involves the potential addition of two types of nucleotides, N-nucleotides and Pnucleotides. N-nucleotides are generated by the nontemplated addition of nucleotides to the coding ends, which is mediated by the enzyme terminal deoxynucleotidyl transferase (TdT). The palindromic sequences that result from the asymmetric cleavage and template mediated fill-in of the coding hairpins are referred to as P-nucleotides. Although N- and P- nucleotides and deletion of the coding end and nucleotides serve to greatly diversify the immunoglobulin repertoire/ the addition of these nucleotides may also result in the generation of receptor genes that are out of frame. Similar to the RAG recombinase complex/ the DNA repair machinery works as a protein complex. However, unlike the RAG recombinase, the nonhom*ologous end-joining proteins are ubiquitously expressed. In the first steps of DNA repair, the Ku70 and

lu.

CHAPTER 3-ANIIBODIES

complex binding ondsynopsis I RAG v (RAG-'l, proteins) RAG-2, HNiIG Thedouble strondDNA is benlto bringtogether thetworecombinotion srgn0r sequences RSS

I uicting

RAGcomplex (shownin red) Hoirpin formolion (lronseslerif reoction) icotion

Post-cleovoge complex

Hoirpin opening ondcleovoge (NHEJ mochinery)

joint Coding joint Signol

Ku80 proteins form a heterodimer which bind the brokenDNAends. The Ku complex recruits the catalytic subunit of DNA-dependent protein kinase, DNA-PKcs, a serine-threonine protein kinase. The activated DNAPKcs then recruits and phosphorylates XRCC4 and Artemis. Artemis is an endonuclease that opens the hairpin coding ends. Finally, DNA ligase IV binds XRCC4 to form an end-ligation complex, and this complex mediates the final ligation and fill-in steps needed to form the coding and signal joints. Reguloting V(D)Jrecombinolion V(D)j recombination and the recombinase machinery must be carefully regulated to avoid wreaking havoc on the cellular genome. For instance, aberrant V(D)J recombination is implicated in certain B-cell lymphomas. V(D)J recombination is largely regulated by controlling expression of the recombination machinery and the accessibility of gene segments and nearby enhancers

Figure 3.24. The V(D)J recombinase. In the initial steps of V(D)J recombination, the RAG-1 and RAG-2 proteins associatewith the recombination signal sequences A single-stranded nick is then introduced between the 5' heptameric end of the recombination signal sequence and the coding segment, giving rise to a free 3'OH group that mediates a transesterification reaction. This reaction leads to the formation of DNAhairpins at the coding ends. Hairprn cleavage and resolution of the post-cleavage complex by nonhom*ologous end-joining (NHEJ) proteins results in the formation of separate coding and signaljoints, in the final steps of V(D)J recombination

and promoters. As previously mentioned, RAG-1 and RAG-2 activity is specific to lymphoid cells, and further regulation is imposed by downregulating RAC activity during appropriate stages of B-cell development. Differential accessibility of gene segments to the recombinase machinery, which can be achieved by altering chromatin structure, also plays a role in making certain that appropriate gene segments are recombined in an appropriate order. Cls-actingtranscriptional control elements, such as enhancers and promoters, also help regulate recombination. Although it is not a hard and fast rule, transcription from certain regulatory elements seems to correlate with rearrangement of the adjacent genes. This sterile, or nonproductive, transcription may somehow help target required proteins or modulate gene accessibility. Finally, in addition to directing recombinationbetween appropriate gene segments, the precise sequences of the RSS itself, as well as the sequences of the gene segments themselves, can influence the efficiency of the recombination reaction.

CHAPTER 3-ANTIBODIES

-----_T------__leT, I

/oM \crc

lTcca/

Single-stronded DNAhoirpins

cleovoge I Hoirpin

----l----lffiTrccr

(P) Polindromic nucleolides

-f-_-]* a@rcc

I ror

Deletion of nucleotides from coding end

---l----lmrccnarc

N_nucleolides

----f---__lrerccn^ucrcs I rc^T@rrrcffi

lmprecisejoin 0l voriobleexons

Figure 3.25. Junctional diversity further diversifies the immune repertoire. The immunoglobuhn reperioire is further diversified during cleavage and resolution of the coding-end hairpins by deletion of a variable number of coding end nucleotides, the addition of Nnucleotrdes by terminal deoxynucleotidyl transferase (TdT), and palindromic (P)nucleotides that arise due to template-mediated fill-in of the asymmetrically cleaved coding hairpins. TdT randomly adds nucleotides to the DNA ends (N-nucleotides), and the single-stranded ends pair, possibiy but not necessarily, through complementary nucleotides (TG on top strand and AC on bottom strand) Exonuclease tdmming, to remove unpaired nucleotides, and the DNA repair machinery act to repair the DNA joint

57 1

and preferentially targeted to appropriate V regions, while constant regions of the immunoglobulin loci remain protected, is not clearly understood and is the subject of current research. Transcription through the target V region seems required, but is not necessarily sufficient, for somatic hypermutation. Additionally, the enzyme activation-induced cytidine deaminase (AID) has been demonstrated to be essential for both somatic hypermutation and class-switch recombination. AID is a cytidine deaminase capable of carrying out targeted deamination of C to U, and shows strong hom*ology with the RNA-editing enzyme APOBEC-1. TWo current hypotheses have been proposed to explain the mechanism by which AID acts, one favoring RNA editing while the second favors DNA deamination. It is possible thatAID recognizesand acts on an mRNAprecursor, or more likely that AID directly deaminates DNA to produce U : G mismatches. The exact mechanism by which AID can differentially regulate somatic hypermutation and class switch recombination is currently being studied, and may depend on interactions of specific cofactors with specific domains of AID. Therefore, diversity within the immunoglobulin repertoire is generated by: (i) the combinatorial joining of gene segments; (ii) junctional diversity; (iii) combinatorial pairing of heavy and light chains; and (iv) somatic hypermutation of V regions.

Somolichypermutotion

Geneconvelsi0n0ndrepenoirediversificolion

Following antigen activation, the variable regions of immunoglobulin heavy and light chains are further diversified by somatic hypermutation. Somatic hypermutation involves the introduction of nontemplated point mutations into V regions of rapidly proliferating B-cells in the germinal centers of lymphoid follicles. Antigen-driven somatic hypermutation of variable immunoglobulin genes can result in an increase rn binding affinity of the B-cell receptor for its cognate ligand. As B-cells with higher affinity immunoglobulins can more successfully compete for limited amounts of antigen present, an increase in the average affinity of the antibodies produced during an immune response is observed. This increase in the average affinity of immunoglobulins is known as affinity maturation. Somatic hypermutation occurs at a high rate, thought to be on the order of about 1 x 10-3mutations per base pair per generation, which is approximately 106times higher than the mutation rate of cellular housekeeping genes. There is a bias for transition mutations, and the 'mutation hotspots' in variable regions map to RGWY motifs (R = purine, ] = pyrimidine, W = A or T). The exact mechanisms by which mutations are introduced

Although mice and humans use combinatorial and junctional diversity as a mechanism to generate a diverse repertoire, in many species, including birds, cattle, swine, sheep, horses and rabbits, V(D)j recombination results in assembly and expression of a single functional gene. Repertoire diversification is then achieved by gene conversion, a process in which pseudo-V genes are used as templates to be copied into the assembled variable region exon. Further diversification maybe achievedby somatic hypermutation. The process of gene conversion was originally identified in chickens, in which immature B-cells have the same variable region exon. During B-cell development in the bursa of Fabricius, rapidly proliferating B-cells undergo gene conversion to diversify the immunoglobulin repertoire (figure 3.26). Stretches of sequences from germline variable region pseudogenes, located upstream of the functional V genes, are introduced into the V, and V' regions. This process takes place in the ileal Peyer's patches of cattle, swine and horses, and in the appendix of rabbits. These gut-associated lymphoid tissues are the mammalian equivalent of the bursa in thesespecies.

CHAPTER 3-ANTIBODIES

158

IgE isotypes are located downstream of the IgM (Cp) exon, and CSR occurs between switch or S regions. S regions are repetitive sequences,which are often G-rich on the nontemplate strand, that are found upstream of each C' exon. Breaks are introduced into the DNA of two S regions and fusion of the S regions leads to a rearranged Crr locus, inwhich the variable exon is joined to an exon for a new constant region. The DNAbetween the two switch regions is excised and forms an episomal circle. Finally, alternative splicing of the primary RNA transcript generated from the rearranged DNA gives rise to either membrane-bound or secreted forms of the immunoglobulin. Prior to recombination between switch regions, tran-

Clossswitchrecombinolion Antigen-stimulated IgM expressing B-cells in germinal centers of secondary lymphoid organs, such as the spleen and lymphnodes, undergo classswitchrecombination. Class switch recombination (CSR) allows the IgH constant region exon of a given antibody to be exchanged for an alternative exon, giving rise to the expression of antibodies with the same antigen specificity but of differing isotypes, and therefore of differing effector functions as described above. CSR occurs through a deletional DNA recombination event at the IgH locus (figure 3.27), which has been extensively studied inmice. Constantregionexons for IgG, IgA, and

Wzs

Wgo

DH JH

Ci,

V(D)Jrecombinolion vDJ

Gene conversion I Figure 3.25. Immunoglobulin diversifi cation using gene conversion. V(D)J recombination in chicken B-cells results in assembly of a single variable region exon In theprocess of gene conversion, sequences of pseudogenes, located upstream of the functional gene segments, are copied into the recombined variable exons at the light and heavy chain loci in rapidly proliferating B-cells in the bursa ofFabricius This results in a diversified antibody repertoire.

I hypermutolion Somotic

Germline tronscription

3'regulotory

CY,

Figure 3.27. Class switch recombination allows expression of different antibody isotypes. Class switch recombination invoives DNA recombination at repetitive sequences termed switch or S regions, and is illustrated here for an IgM to IgE switch, at the mouse heavy chain locus. Switching to an IgE isotype begins with germline transcription from the promoter upstream of the constant region exon

and recombination between the Sp and Se regions This DNA recombination reaction brings the IgE constant region exon downstream of the variable region exon The remaining switch regions and constant region exons are deleted and form an episomal circle Transcription of the rearranged DNA yields IgE mRNA, which can be translated to give rise to the IgE immunoglobulin protein

CHAPTER 3-ANTIBODIES scription is initiated from a promoter found upstream of an exon that precedesall C' genescapable of undergoing CSR, the intervening (I) exon. These germline transcripts include I, S and C region exons, and do not appear to code for any functional protein. However, this germline transcription is required, although not sufficient, to stimulate CSR.The precisemechanism responsible for CSR is the subject of current study, but work indicates thatAID, described previouslv to be involved

Antibody slructure ondfuncfion . Antibodies recognize foreign material and trigger its elimination. o They are Y- or T-shaped molecules in which the arms of the molecule (Fab) recognize foreign material and the stem (Fc) interacts with immune molecules that lead to the elimination of the antibody-decorated foreign material. r Antibodies are based on a four-chain structure consisting of two identical heavy chains and two identical light chains. o The N-terminal parts of the heavy chains and the light chains form the two identical Fab arms that are linked to the Fc stem of the molecule consisting of the C-terminal parts of theheavychains. o The extremities of the Fab arms consist of regions of variable amino acid sequences that are involved in binding antigen and thereby give each antibody its unique specificity. The human antibody repertoire is vast, allowing the recognition of essentially any molecular shape. . The Fc stem of the molecule has a more conserved sequence and is involved in binding effector molecules such as complement and Fc receptors. . Differences in the Fc regions lead to different classesand subclassesof antibodies or immunoglobulins (Igs). . There are five different classesof Ig-IgG, IgM, IgA, IgD and IgE -which fulfill differentroles in immune protection. They also have different polymerization states. o The structure of antibodies is organized into domains based on a B-sheetarrangement called the immunoglobulin fold. . For IgG, the Fab arms, consisting of two variable domains and two constant domains, are linked via a flexible hinge region to the Fc, which consists of four constant domains. . Flexibility is an important feature of antibody structure allowing interaction with antigens and effector molecules in a variety of environments. Anlibodyinferocfionwilh effeclormolecules . IgG triggers complement by binding C1q when clustered onanantigensuch asa pathogen. IgMis alreadymultivalent

5sl

in somatic hypermutation, helps mediate CSR, along with some components of the nonhom*ologous endjoining pathway and several other DNA repair pathways. The joining of S regions may be mediated by association with transcriptional promoters, enhancers, chromatin factors, DNA repair proteins, AlD-associated factors or by interactions involving S region sequences themselves.

but, on binding antigen, it undergoes a conformational change to bind C1q. . Leukocyte receptors have been described for IgG, IgA and IgE that, on binding antigen-associated antibody, trigger effector mechanisms such as phagocytosis, antibody-dependent cellular cytoxicity and acute inflammatory responses. Interaction between antibody and Fc receptors can also be immunoregulatory. o The structures of IgG Fc receptors and the mast cell IgE Fc receptor and the mode of interaction of the receptors with Ig appear to be quite similar. The IgA receptor has however a distinct structure and mode of interactionwith IgA. . IgG interacts with the neonatal receptor FcRn to promote transport of IgG from mother to child and to maintain the longhalf-life of IgG in serum.

oveNiew ollhelgclosses . IgG is monomeric and the major antibody in serum and nonmucosal tissues, where it inactivates pathogens directly and through interaction with triggering molecules such as complement and Fc receptors. . IgA exists mainly as a monomer in plasma, but in the seromucous secretions, where it is the major Ig concerned in the defense of the external body surfaces, it is present as a dimer linked to a secretory component. . IgM is most commonly a pentameric molecule although a minor fraction is hexameric. It is essentially intravascular and is produced early in the immune response. Because of its high valency it is a very effective bacterial agglutinator and mediator of complement-dependent cytolysis and is therefore a powerful firstline defense against bacteremia. . IgD is largely present on the lymphocyte and functions together with IgM as the antigen receptor onnaive B-cells. . IgE binds very tightly to mast cells and contact with antigen leads to local recruitment of antimicrobial agents through degranulation of the mast cells and release of inflammatory mediators. IgE is of importance in certain parasitic infections and is responsible for the symptoms of atopic allergy. (Continuetlp 60)

lro

CHAPTER 3-ANTIBODIES

Thegenerolionof 0ntibodydiversily o The antibody repertoire of an individual is generated through somatic recombination events from a limited set of germline gene segments. . The human heavy chain variable region is generated by joining of Vrr, D and J gene segments and the light chain variable regions (r and l,) by joining of V'- and I segments. Joining is imprecise, leading to the generation of further diversitv.

F U R I H ER E A D I N G Arakawa H. & Buerstedde |. (200a) Immunoglobulin gene conversion: rnsights from bursal B cells and the DT40 cell line Deaelopmen t al D y nomi cs 229, 458464. Chaudhuri J. & Alt F W. (2004) Class-switch recombination: interplay of transcription, DNA deamination, and DNA repair. Naf ure ReaiewsImmun ology 4, 5 41-552. Cook G P & Tomlinson I M (1995) The human immunoglobulin V' repertoire lmmunology Today16, 237-242 Dudley D.D., Chaudhuri, J.,Bassing, C.H. & Alt, F.W. (2005)Mechanism and control of V(D)J recombinati.on verus class switch recombination: similarities and differences Adaances in lmmunology 85,43-112. Groner B., Hartmann C. & Wels W. (2004) Therapeutic antibodies. Current Molecular Medicine 4, 539-547 Honjo T., Nagaoka H, Shinkura R. & Muramatsu M. (2005)AID to overcome the limitations Naflre of genomic information. Immunology 6,655-661. Hozumi N & Tonegaw a S (7976) Evidence for somatic rearrangement of immunoglobulin genes coding for variable and constant regions. Proceedingsof the National Academy of Sciencesof the USA 73,36281632. Hudson PJ & Souriau C (2003) Engineered antibodies Naftre Medicine 9,\29-734. IMGT database: http: / /imgt.cines.fr Jung D. & Alt F W. (2004) Unraveling V(D)J recombination: insights into gene regul atron. CelI 716, 299-311, Maizels N (2005) Immunoglobulin gene diversification Annual Reaiew of Genetics39, 2346 Maki R, Traunecker, A., Sakano, H, Roeder, W & Tonegawa, S (1980) Exon shuffling generates an immunoglobulin heavy chain gene P roceedingsof the N ational Academy of Sciencesof tlrc U SA 77, 2138-2142. Martin A. & Scharff M.D (2002) AID and mismatch repair in antibody diversification. Naf rre Reaiewslmmunology 2,605-674 Martin W.L., WestA.P. Jr,Gan L. & Bjorkman P.J.(2001) Crystal structure at 2.8 A of an FcRn/heterodimeric Fc complex: mechanism of pH-dependent bindlng. Molecular Cell 7, 867-877.

. Still further diversification results from somatic mutation events targeted to the variable regions. Somatic mutation and selection allows affinitymaturation of antibodies. o Somespeciesusegeneconversionratherthancombinatorial and junctional diversity to achieve antibody diversification. . Class switch recombination events allow the same antibody specificity (variable regions) to be associated with different antibody classes and subclasses (constant regions) and therefore with different functions.

Matsuda F. & Honjo T (7996) Organization of the Human Heavy-Chain Locus. Adaances in lmmunology Immunoglobulin 62,7-29 McCormack W.T., Tjoelker L.W. & Thompson C B. (1991) Avian Bcell development: generationof animmunoglobulinrepertoireby gene conversion. AnnuslReaiew of lmmunology 9,219-241,. Metzger H. (2004) The high affinity receptor for IgE, FceRI . Noztartis F oundn t ion Symposium 257, 57-59. Min I.M. & Selsing E (2005) Antibody class switch recombination: roles for switch sequences and mismatch repair p roleins. Adaances in lmmunology 87,297-328. Neuberger M.S., Harris R.S., Di Noia J & Petersen-Mahrt S K. (2003) Immunity through deamination.Trends in BiochemicnlSciences28, 305-312. Padlan E.A. (1994) Anatomy of the antibody molecule. Molecular Immunology 37,769-277 Padlan E A (1996) X-ray crystallography of antibodies . Adaances in P rotein Chemistry 49, 57 -1,33. Parren P.W & Burton D.R (2001) The antiviral activity of antibodies in vitro and in vivo Adaancesin lmmunology 77,795-262. Ravetch J.V. & Bolland S (2001) IgG Fc receptors. Annuql Reztiewof Immunolo gy 19, 27 5-290. Roth D B (2003) Restraining the V(D)J recombinase. Naf ure Reaiews Immunology3,656-666. Swanson P.C. (2004) The bounty of RAGs: recombination signal complexes and complex outcomes. Immunological Reaiezas2oo, 90-11.4. Ward E.S. (2004) Acquiring matemal immunoglobulin; different receptors, similar functions lmmunity 20, 507-508. Woof I M & Burton D.R. (2004) Human antibody-Fc receptor interactions illuminated by crystal structures. Nature Reaiews lmmunology4,89-99. Woof J.M & Kerr M A (2004) IgA function -variations on a theme. Immunolo gy l1.3, 175-77 7 Yoo E M & Morrison S.L. (2005) IgA: an immune glycoprotein. Clinical lmmunology 116,3-70. Zach.ar H.G. (2000) The immunoglobulin kappa gene families of human and mouse: a cottage industry approach. Biological Chemistru 381. 951-954.

Membrone forontigen reoeptors

INIRODUCIION The interaction of lymphocytes with antigen takes place through binding to specialized cell surface antigenspecific receptors functioning as recognition units. In the case of B-cells, the situation is straightforward as membrane-bound immunoglobulin serves as the receptor for antigen. T-cells use distinct antigen receptors, which are also expressed at the plasma membrane, but T-cell receptors (TCRs) differ from B-cell receptors (BCRs) in a very fundamental way; TCRs cannot recognize free antigen as immunoglobulin can. The majority of T-cells can only recognize antigen when presented within the peptide-binding groove of an MHC molecule. While this may seem rather cumbersome, a major advantage that T-cells have over their B-cell brethren is that they can inspect antigens that are largely confined within cells and are therefore inaccessible to Ig. Another leukocyte class,natural killer (NK) cells, can also detect trouble brewing within. NK cells possess their own unique receptors that check for appropriate levels of MHC molecules, as these are normally expressed on practically all nucleated cells within the body; NK receptors can also detect signs of abnormality such as increases in the expression of stress proteins by cells. Here we will focus mainly on the structural aspects of thesevarious receptor types.

T H EB . C E t tS U R F A CREE C E P I OFRO R A N i l G E N( B C R ) TheB-cell disploys o lronsmembrone immunoglobulin on ilssurfoce In Chapter 2 we discussedthe curming systembywhich an antigen can be led inexorably to its doom by activat-

ing B-cells that are capable of making antibodies complementary in shape to itself through interacting with a copy of the antibody molecule on the lymphocyte surface. It will be recalled that binding of antigen to membrane antibody can activate the B-cell and cause it to proliferate followed by maturation into a clone of plasma cells secreting antibody specific for the inciting antigen(cf.figure 2.11). Immunofluorescent staining of live B-cells with labeled anti-immunoglobulin (anti-Ig) (e.g. figure 2.6c) reveals the earliest membrane Ig to be of the IgM class. Each individual B-cell is committed to the production of just one antibody specificity and so transcribes its individual rearranged VlCrc(or L) andvDlCltgenes. Ig can be either secreted or displayed on the B-cell surface through differential splicing of the pre-mRNA transcript encoding a particular immunoglobulin. The initial nuclear p chain RNA transcript includes sequences coding for hydrophobic transmembrane regions which enable the IgM to sit in the membrane where it acts as the BCR, but if these are spliced out, the antibody molecules canbe secretedin a soluble form (figure 4.1). As the B-cell matures, it coexpresses a BCR utilizing surface IgD of the same specificity. This surface IgM+surface IgD B-cell phenotype is abundant in the mantle zone lymphocytes of secondary lymphoid follicles (cf. figure 7.8c) and is achieved by differential splicing of a single transcript containing VD], Cp and C6 segments producing either membrane IgM or IgD (figure 4.2).As the B-cell matures further, other isotypes such as IgG m ay be utllizedin the BCR (cf . p. 2al. is complexed with ossocioled Suffoceimmunoglobulin

proleins membrone Because secreted immunoglobulin

is no longer physi-

lu,

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

Figure 4.1. Splicing mechanism for the switch from the membrane to the secreted f o r m o f l g M . A l t e r n d t i v ep r o c e s s i n g determines whether a secreted or membranebound form of the p heavy chain is produced If transcription termination or cleavage occurs in the intron between Cp oand Mr, the Cltopoly-A addition signal (AAUAAA) is used and the secreted form is produced. If transcription continues through the membrane exons, then C,u4can be spliced to the M sequencesresulting in the M, poly-A addition signal being utilized The hydrophobic sequence encoded by the exons M, and M, then anchors the receptor IgM to the membrane For simplicity, the leader = lntrons sequence has been omitted. *

ffi FoTEn'l @

rrt,

IIG@[emailprotected] cally connected to the B-cell that generated it, there is no way for the B-cell to know when the secreted Ig has found its target antigen. In the case of membraneanchored immunoglobulin however, there is a direct Iink between antibody and the cell making it and this canbe exploited to instructthe B-cell to scale-upproduction. As any budding industrialist knows, one way of increasing production is to open up more manufacturing plants, and another is to increase the rate of productivity in each one. When faced with the prospect of a sudden increase in demand for their particular product, B-cellsdoboth of thesethings, through clonal expansion and differentiation to plasma cells. So how does the BCR spur the B-cell into action upon encounter with antigen? Unlike many plasma membrane receptorsthat boast all manner of signaling motifs within their cytoplasmic tails, the corresponding tail region of a membrane-

Figure 4.2. Surface membrane IgM and IgD receptors of identical specificity appear on the same cell through differential splicing of the composite primary RNA transcript (leader sequences again omitted for simplicity)

anchored IgM is a miserable three amino acids long. In no way could this accommodate the structural motifs required for interaction with either adaptor proteins, intracellular protein kinases or phosphatases that typically initiate signal transduction cascades.With some difficulty, it should be said, it eventually proved possible to isolate a disulfide-linked heterodimer, Ig-c (CD79 a')and Ig-p (CD79b), which copurifies with membrane Ig and is responsible for transmitting signals from the BCR to the cell interior (figure 4.3).Both Ig-o and IgB have an extracellular immunoglobulin-type domain, but it is their C-terminal cytoplasmic domains that are obligatory for signaling and which become phosphorylated upon crosslinking of the BCR by antigen, an event also associatedwith rapid Ca2+mobilization. Ig-a and Ig-B each contain a single ITAM (immunoreceptor fyrosine-based activation motifl within their cytoplas-

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

63 1

manner that is quite distinct from the way in which Bcells do; the receptors that most T-cells are equipped with cannot directly engage soluble antigens but 'see' instead short fragments of antigen that are immobilized within a narrow groove on the surface of MHC molecules. Moreover, T-cells cannot secretetheir receptor molecules in the way that B-cellscan switch production of Ig from a membrane-bound form to a secreted form. These differences aside, T:cell receptors are structurally quite similar to antibody as they are also built from modules that are based upon the immunoglobulin fold. S-SI

FS-S

helerodimer foronligen iso lronsmembrone Thereceptor

Figure 4.3. Model of B-cell receptor (BCR) complex. The Ig-crllg-B heterodimer is encoded by the B-cell-specific genes mb-7 and 829, respectively. Two of these heterodimers are shown with the Ig-cr associating with the membrane-spanning region of the IgM F chain The IgJike extracellular domains are colored blue. Each tyrosine (Y)-containing box possesses a sequence of general structure Tyr X, Leu X, Tyr.Xr.Ile (where X is not a conserved residue), referred to as the immunoreceptor tyrosine-based activation motif (ITAM) On activation of the B-cell, these ITAM sequencesact as signal transducers through their ability to associate with and be phosphorylated by a series of tyrosine kinases Note that whilst a r light chain is illustrated for the surface IgM, some B-celis utilize a ),light chain

mic tails and this motif contains two precisely spaced tyrosine residues that are central to their signaling role (figure 4.3).Engagement of the BCR with antigen leads to rapid phosphorylation of the tyrosines within each ITAM, by kinases associated with the BCR, and this has the effect of creating binding sites for proteins that have an affinity for phosphorylated tyrosine residues. In this case, a protein kinase called Syk becomes associated with the phosphorylated Ig-alp heterodimer and is instrumental in coordinating events that culminate in entry of the activated B-cell into the cell cycle to commence clonal expansion. We will revisit this topic in Chapter 8 where the details of the BCR signal transduction cascadewill be elaborated upon in greater detail (seep. 180).

I H E T - C E t tS U R F A CREE C E P T OFR OR TCR) A N T T G E( N As alluded to earlier, T-cells interact with antigen in a

Identification of the TCR proved more difficult than initially anticipated (Milestone 4.1), but eventually the receptor was found to be a membrane-bound molecule composed of two disulfide-linked chains, a and B. Each chain folds into two Ig-like domains, one having a relatively invariant structure and the other exhibiting a high degree of variability, so that the aB TCR has a structure really quite closely resembling an Ig Fab fragment. This analogy stretches even further -each of the two variable regions has three hypervariable regions, regions (o. complementarity-determining data have defined as which X-ray diffraction CDRs) incorporating the amino acids which make contact with the peptide-major histocompatibility complex (MHC) ligand. Although the manner in which the TCR makes contact with peptide-MHC is still not fully understood, it appears that CDRs 1 and 2 of the TCRbear much of the responsibility for making contact with the MHC molecule itself, while CDR3 makes contact with the peptide; thus it is here that much of the variability is seen between TCRs, as we shall discuss later. Both s and B chains are required for antigen specificity as shown by transfection of the T-receptor genes from a cytotoxic T-cell clone specific for fluorescein to another clone of different specificity; when it expressed the new u and B genes, the transfected clone acquired the ability to lyse the fluoresceinated target cells. Another type of experiment utilized T-cell hybridomas formed by fusing single antigen-specific T-cells with 'immortality'. One hybriT-cell tumors to achieve doma recognizing chicken ovalbumin, presented by a macrophage, gave rise spontaneously to two variants, one of whichlost the chromosome encoding the crchain, and the other, the B chain. Neither variant recognized antigenbut, when they were physically fused together, each supplied the complementary receptor chain and reactivity with antigen was restored.

loo

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I E E N

Since T-Iymphocytes respond by activation and proliferation when they contact antigen presented by cells such as macrophages, it seemed reasonable to postulate that they do so by receptors on their surface. In any case,it would be difficult to fit T-cells into the clonal selection club if they lacked such receptors. Guided by Ockam's razor (the Law of Parsimony,whichcontends that it is the aim of scienceto present the facts of nature in the simplest and most economical conceptual formulations), most investigators plumped for the hypothesis that nature would not indulge in the extravagance of evolving two utterly separate molecular recognition speciesfor B- and T-cells,and many fruitless years were spent looking for the Holy Grail of the T-cell receptor with antiimmunoglobulin serums or monoclonal antibodies (cf. p. 112). Success only came when a monoclonal antibody directed to the idiotype of a T-cell was used to block the response to antigen. This was identified by its ability to block one individual T-cell clone out of a large number, and it was

correctly assumed that the structure permitting this selectivity would be the combining site for antigen on the T-cell receptor. Immunoprecipitation with this antibody brought down a disulfide-linked heterodimer composed of 4044 kDa subunits (figureM4.1 1). The other approach went directly for the genes,arguing as follows. The T-cell receptor should be an integral membrane protein not present in B-cells. Hence, T-cell polysomal mRNA from the endoplasmic reticulum, which should provide an abundant sourceofthe appropriate transcript, was used toprepare cDNAfromwhich genescommon to B- and Tcells were subtracted by hybridization to B-cell mRNA. The resulting T-specific clones were used to probe for a T-cell gene which is rearranged in all functionally mature T-cellsbut is in its germ-line configuration in all other cell types (figure M4.1.2).Insuchawaywere the genesencoding the B-subunit of the T-cell receptor uncovered.

isololed byB-cellsubtrociion T-cellg€nes T-cellswifhdifiering specificity

FusewithT-celllumor to produce hybridomo clones

@@ ttt

qv qv qv

Inhibn. Agrecognilion bymonoclonol onlihybT2 idiotype j

Ablo T-cell receplor immunoprecipitoles 2 choins

Surfoce lobeled T-hybridomo

Solubilize Pptwifh onli-ld, RunonSDSPAGE

30 45 69 97 tttl tttt I

:ll Reduced

Unreduced

Figure M4.1.1. Ab to T-cell receptor (anti-idiotype) blocks Ag recognition. (Based on Haskins K., Kubo R., \alhite J., Pigeon M., Kappler f . & Marack P. (1983) lournal of ExperimentalMedicine 157, 1149;simplified a little.)

for TCRS CD4ondCD8moleculesocl os co-receptors In addition to the TCR, the maiority of peripheral T-cells also expressone or other of the membrane proteins CD4 or CD8 that act as co-receptors for MHC molecules (figure 4.4). CD4 is a single chain polypeptide containing four Ig-like domains packed tightly together to form

Figure M4.1.2. Isolation of T-cell recePtor genes. Different sized DNA fragments produced by a resfriction enzyme are separated by electrophoresis and probed with the T-cell gene. The T<ells show rearrangement of one of the two germ-line genes found inliver or Bcells. (Based on Hendrick S.M., Cohen D.I., Nielsen E.A. & Davis M.M. (1984)Nature 3O8,149.)

an extended rod that projects from the T-cell surface. The cytoplasmic tail of the CD4 molecule is important for TCR signaling as this region is constitutively bound by a protein tyrosine kinase, Lck, that initiates the signal transduction cascade that follows uPon encounter of a T-cell with antigen. CD8 plays a similar role to CD4, as it also binds Lck and recruits this kinase to the TCR

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

65 1

MHC molecules that obtain their peptide antigens primarily from intracellular (MHC classI), or extracellular (MHC class II), sources. This has major functional implications for the T-cell, as those lymphocytes that become activated upon encounter with antigen Presented within MHC class I molecules (CD8* T-cells) invariably become cytotoxic T-cells, and those that are activated by peptides presented by MHC class II molecules (CD4* T-cells) become helper T-cells (see figure 8.1).

I receplors 0fT-cel There oretwoclosses Not long after the breakthrough in identifying the cxp TCR, came reports of the existenceof a second type of receptor composed of y and 6 chains. Since it appears earlier in thymic ontogeny, the y6 receptor is sometimes referred to as TCR1 and the crBreceptor as TCR2 (cf.

Figure 4.4. CD4 and CD8 act as co-receptorsfor MHC molecules and define functional subsets of T-cells. (a) Schematic representation of CD4 and CDS rnolecules CD4 is composed of four Ig-like domains (D.' to Dn, as indicated) and prolects from the T-cell surface to interact with MHC class II molecules CD8 is a disulfidelinked heterodimer composed of IgJike o and p subunits connected to a heavily glycosylated rodJike region that extends from the plasma membrane CD8 interacts with MHC class I molecules The cytoplasmic tails of CD4 and CD8 are associated with the tyrosine kinase Lck (b) Ribbon diagram representations of the extracellular portions of CD4 and CD8 The IgJike domains (D, to D,,) of CD4 are colored blue, green, yellow and red respectively. A CD8 hom*odimer of two cr subunits is shown (Structures were kindly providedby Dr Dan Leahy and are based upon coordinates reported in Leahy ef a1 (7992) Cell 68, 7745 and W r et al. (1,997) Nature 387,527.)

complex/ but is structurally quite distinct; CD8 is a disulfide-linked heterodimer of o(and B chains, each of which contains a single Ig-like domain connected to an extended and heavily glycosylated polypeptide projecting from the T-cell surface (figure 4.4). CD4 and CDS molecules play important roles in antigen recognitionby T-cellsas thesemolecules dictate whether a T-cell can recognize antigen presented by

p.235). y6 cells make up only 1-5% of the T-cells that circulate in blood and peripheral organs of most adult animals; however these cells are much more common in epithelial-rich tissues such as the skin, intestine, reproductive tract and the lungs where they can comprise almost50% of the T-cell population. It cannot be denied that y6 Tcells are somewhat of an oddity among T-cells; unlike uB T-cells, yD cells do not appear to require antigen to be presented within the context of MHC molecules and are thought to be able to recognize soluble antigen akin to Bcells. Perhaps because of this lack of dependence on MHC for antigen presentation, the majority of yb T-cells do not express either of the MHC co-receptors, CD4 or CDS (table4.1). The mechanism of antigen recognitionby y6 T-cells is still somewhat mysterious but these cells are known to be able to interact with MHC-related molecules, such as the mouse T10 and T22 proteins, in a manner that does not require antigen. Becausethe latter MHC-like molecules are upregulated upon activation of aB T-cells,this has led to the view thaty6T-cells mayhave animportant immunoregulatory function; by becoming activated by molecules that appear on activated T-cells, y6 T-cells may help to regulate immune responses in a positive or negative manner. y6 T-cells can also recognize lipids, organic phosphoesters, pathogen-derived nucleotide conjugates and other nonpeptide ligands. of T-cellreceplolsis simil0rto Ihe encoding lhol of immunoglobulins The gene segments encoding the TCR p chains follow a broadly similar arrangemerrtof V, D,/and constant segments to that described for the immunoglobulins (figure

luu

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

4.5).In a parallel fashion, as an immunocompetentT-cell is formed, rearrangement of V , D and / genes occurs to form a continuous 7D/ sequence. The firmest evidence that B- and T-cells use similar recombination mechanisms comes from mice with severe combined immunodeficiency (SCID) which have a single autosomal recessivedefect preventing successful recombination of V, D and / segments (cf. p. 5a). hom*ozygous mutants fail to develop immunocompetent B- and T-cells and identical sequence defects in VDI joint formation are seeninboth pre-B- and pre-T-celllines.

Table 4.1. Comparison between eB and y6 T-cells.

Anligen receptor

opTCR+CD3 comprex

+ CD3 16TCR comptex

Formofontigen recognrzeo

MHC+ peptide

MHC-like molecules plusnonprofein ligonds

CD4/CD8 expression Yes

Moinlv no

Frequencyin blood

60-75%

l-5%

MHCreslricted

Yes

Mostlvno

Funclion

Helpforlymphocyte ond mocrophoge octivotion Cytoloxic killing

lmmunoregulotory function? Cylotoxic octivity

Looking first at the B chain cluster, one of the two DB genes rearranges next to one of the /B genes. Note that,because of the way the genesare organized, the first Dp gene, Dp' can utilize any of the 13 lB genes,but Dp, can only choose from the seven /B, genes (figure 4.5). Next one of the 50 or so I/p genesis rearranged to the preformed DBIB segment. Variability in iunction formation and the random insertion of nucleotides to create N-region diversity either side of the D segment mirror the same phenomenon seen with Ig gene rearrangements. Sequence analysis emphasizes the analogy with the antibody molecule; each V segment contains two hypervariable regions, while the D/ junctional sequence provides the very hypervariable CDR3 structure, making a total of six potential complementarity determining regions for antigen binding in each TCR (figure 4.6).As in the synthesis of antibody, the intron between VDJ and C is spliced out of the mRNA before translation with the restriction that rearrangements involving genes in the D Brl Brcluster can only link

toCp,. All the other chains of the TCRs are encoded by genes formed through similar translocations. The achain gene pool lacks D segments but possesses a prodigious number of / segments. The number of VTand Vd genesis small in comparison with V a and V B. Like the a chain pool, the y chain cluster has no D segments. The awkward location of the dlocus embedded within the a gene cluster results in T-cells which have undergone V a-l ucombination having no dgenes on the rearranged chromosome; in other words, the dgenes are completely excised.

NN VV

D I J

C )TM

N V ,vd l+/C

,JA l-ou

NN IV

,l + JD 6 ^

,I -----:> J6^

JTM

N V

L IV

) J

Figure4.5. Genesencodingcrpandy6T-cellreceptors.Genesencodingthe6chainsliebetweentheVaand/aclustersandsomeVsegmentsinthis region canbe used in either Eor crchains, i e. as either Vaor Vd TCR genes rearrange in a manner analogous to that seen with immunoglobulin genes, includingN-regiondiversityattheV(D)/junctions Oneof theV6genesisfounddownstream(3,)oftheC6geneandrearrangesbyaninversional mechanism

67 1

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

Figure 4.6. The T-cell receptor/CD3 complex. The TCR resembles the i m m u n o g l o b u l i n F a ba n t ig e n - b i n di n g fragment in structure. The variable and constant segments of the TCR s and B chains (VaCu/VpCB), and of the corresponding y and 6 chains of the y6 TCR, belong structurally to the immunoglobulin-type domain family. (a) In the model the a chain CDRs are coiored magenta (CDR1),purple (CDR2) and yellow (CDR3), whilst the p chain CDRs are cyan (CDR1), navy blue (CDR2) and green (CDR3) The fourth hypervariable region of the B chain (CDR4), which constitutes part of the binding site for some superantigens(cf. p 7), is colored orange (Reproduced from Garcia,K. et al (7998)Science 279,1166,with permission ) The TCR crand p CDR3 loops encoded by (D)/ genes are both short; the TCR yCDR3 is also short with a narrow length distribution, but the 6 loop is long with a broad length distribution, resembling the Ig light and heavy chain CDR3s, respectively. (b) The TCRs may be expressed in pairs linked to t h e C D 3 c o m p l e x N e g a t i v ec h a r g e so n transmembrane segments of the invariant chains of the CD3 complex contact the opposite chargeson the TCR Ca and Cp chai.nsconceivably as depicted (c) The cytoplasmic domains of the CD3 peptide chains contain immunoreceptor fyrosinebasedactivation notifs (ITAM; cf. BCR, figure 4 3) which contact src protein tyrosine kinases. Try not to confuse the TCR y5 and the CD3 yE chains

vo(y)

vp(6)

cp(6)

TCRslruclure -+ l+-l l-+l+-

Cross-section through tronsmembrone segment

TheCD3complexis 0n integrolpon of theT-cellreceptor The T-cell antigen recognition complex and its B-cell counterpart can be likened to army scouts whose job is to let the mainbattalion knowwhen the enemyhasbeen sighted. When the TCR'sights the enemy', i.e. ligates antigen, it relays a signal through an associatedcomplex of transmembrane polypeptides (CD3) to the interior of the TJymphocyte, instructing it to awaken from its slumbering G0 state and do something useful-like becoming an effector cell. In all immunocompetent Tcells, the TCR is noncovalently but still intimately linked with CD3 in a complex which, as current wisdom has it, may contain two heterodimeric TCR cB or Tb recognition units closely apposed to one molecule of the invariant CD3 polypeptide chains y and 6, two molecules of CD3e, plus the disulfidelinked (-( dimer. The total complex therefore has the structure TCRr-CD3y6er-(, (figure 4.6b). Similar to the BCR-associatedIg-a/9 heterodimer, the CD3 chains also contain one or more ITAMs and these motifs, once again, are instrumental in the propa-

CD3C0MPLEX

€

(r,( - 11or4,

gation of activation signals into the lymphocyte. Upon encounter of the TCR with peptide-MHc, the ITAMs within the CD3 complex become phosphorylated at tyrosine residues; these then act as a platform for the recruitment of a veritable multitude of phosphotyrosine-binding proteins that further disseminate the signal throughout the T-cell. It is here that the role of the CD4 and CD8 co-receptors becomes apparenti phosphorylation of the ITAMs within the CD3 ( chain is accomplished by the Lck tyrosine kinase which, you may recall, is associated with the cytoplasmic tails of CD4 and CD8 (figure 4.4). In mice, either or both of the ( chains can be replaced by a splice variant from the ( gene termed 11.The ( chain also associateswith the FcyRIIIA receptor in natural killer (NK) cells where it functions as part of the signal transduction mechanism in that context also.

FD I V E R S I TFYO R IHEGENERATION RECOGNIIION ANIIGEN We know that the immune system has to be capable of

lut

FS O RA N T I G E N C H A P T E 4R- M E M B R A N ER E C E P T O R

recognizing virtually any pathogen that has arisen or might arise. The awesome genetic solution to this problem of anticipating an unpredictable future involves the generation of millions of different specific antigen receptors, probably vastly more than the lifetime needs of the individual. Since this greatly exceeds the estimated number of 25,000-30,000genes in the human body, there are some clever ways to generateall this diversity, particularly sincethe total number of V, D, / and C genes in an individual human coding for antibodies and TCRs is only around 400. Let's revisit the genetics of antibody diversity, and explore the enormous similarities, and occasionaldifferences,seenwith the mechanisms employed to generateTCR diversity.

Inlrochoin 0mplilicoti0n0f diversily Random VDI combination increasesdia ersity geometrically We saw in Chapter 3 that, just as we can use a relatively small number of different building units in a child's construction set such as Lego to createa rich variety of architectural masterpieces,so the individual receptor gene segments can be viewed as building blocks to fashion a multiplicity of antigen specific receptors for both Band T-cells. The immunoglobulin light chain variable regions are created from V and / segments, and the heavy chain variable regions fromV, D and/segments. Likewise, for both the op and y6 T-cellreceptorsthe variable region of one of the chains (cror y) is encodedby a V and a / segment, whereas the variable region of the other chain (B or 6) is additionally encodedby a D segment.As

for immunoglobulin genes, the enzymes RAG-1 and RAG-2 recognize recombination signal sequences (RSSs)adjacentto the coding sequencesof the TCR y, D and ,l gene segments. The RSSs again consist of conserved heptamers and nonamers separated by spacers of either 72 or 23 base pairs (cf. p. 55) and are found at the 3' of each V segment, on both the 5' and 3' sides of eachD segment,and at the 5' of each/ segment.Incorporation of a D segment is always included in the rearrangement; VB cannot join directly to jp, nor V6 directly to J6.To seehow sequencediversity is generated for TCR, let us take the aB TCR as an example (table 4.2). Although the precise number of gene segments varies from one individual to another, there are typically around 75Vcxgenesegmentsand 60/cgene segments.If there were entirely random joining of any one V to any one / segment, we would have the possibility of generating 4500 V/ combinations (75 x 60). Regarding the TCR B-chain,there are approximately 50 VB geneswhich lie upstream of two clusters of DBIB geneseach of which is associatedwith a CB gene (figure 4.7).The first cluster, that associatedwith CB1,has a single DB1 gene and 6 jB1 genes,whereas the second cluster associatedwith CB2 again has a single DB gene (DB2) with 7 ]82 genes.The Dp1 segment can combine with any of the 50 7B genes and with any of the 13/81 andl$2genes (figure 4.7).DP2 behaves similarly but can only combine with one of the sevendownstream/B2 genes.This provides 1,000different possible VDj combinations for the TCR B-chain' Therefore, although the TCR cr and B chainV, D and J genes add up arithmetically to just 200,they produce a vast number of different s and B variable regions by

Table 4.2. Calculations of human V gene diversity. It is known that the precise number of gene segments varies from one individual to another, perhaps tlOor so in the case of the V' genes for example, so that these calculations represent'typical' numbers. The number of specifrcitres generated by straightforward random combination of germ-line segments is calculated These will be increased by the further mechanisms listed: *minimal assumption of approximately 10 variants for chains lacking D segments and 100 for chains with D segments. The calculation for the T-cell receptor B chain requires further explanation. The first of the two D segments, Dp,,,can combine with 50 7 genes and with all 13/pt andl Btgenes D Bt behavessimilarly but can only combine with the seven downstream /p, genes

v genesegmenrs Dgenesegmenls J genesegments joining Rondom combinoloriol junctionol (wilhout diversity) Totol Combinotoriol helerodimers (ounded) Totol Ds in 3 reoding fromes, olhermechonisms: junctionol N region insertion;* x 103 diversity, Somotic mutolion

y6TCR (ICRI)

apTCR(TCR2)

.-8 a

VxDxJ VxJ 12x5 8x3x3 60 72 6 0x 1 2 43 x l03 4 , 3x 1 0 6

15 50 l,t 6,7 60 VxDxJ VxJ 7 5x 6 0 5 0 ( 1 3 + 7 ) 1000 4500 4500x 1000 45x'l06 4 5 x lOe

). 40 40 27 655 VxJ VxDxJ 40x21xO 40x5 200 6480 6480x200 l3x106 L3 x lOs

VxJ 30x4 I 50 6480x'l50 l0xlo6 L0 x lOs

6el

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

TCRB-choingermline DNA:

Jpr(r-6)

cpr

//./\\

vp2 DBrJp2.2cp2 Figure 4.7. Rearrangement of the T-cell receptor p-chain gene locus. In this example DpL has rearranged to J82.2,and then the Vp2 gene selected out o f t h e 5 0 o r s o ( V B n ) V B g e n e s I f t h e s a m e V a n d D s e g m e n t s h a d b e e n u s e d , b u t t h i s t i r n e J p l4 h a d b e e n e m p l o y e d , t h e n t h e C p l g e n e s e g m e n t w o u l d have been utilized instead of C82.

geometric recombination of the basic elements.But, as with immunoglobulin gene rearrangement, that is only thebeginning. PI ay i ng with th e j unct i ons Anotherploy to squeezemore variationout of the germline repertoire that is used by both the T-cell receptor and the immunoglobulin genes (cf. 3.25)involves variable boundary recombinations of V, D and / to produce different junctional sequences(figure 4.8). As discussed in Chapter 3, further diversity results from the generation of palindromic sequences (Pelements) arising from the formation of hairpin structures during the recombination process and from the insertion of nucleotides at the N region between the 7, D and/segments, a processassociatedwith the expression of terminal deoxynucleotidyl transferase.Whilst these mechanisms add nucleotides to the sequence,yet more diversity can be created by nucleaseschewing away at the exposed strand ends to remove nucleotides. These maneuvers again greatly increasethe repertoire, especially important for the TCR y and 6 genes which are otherwise rather limited in number. Additional mechanisms relate specifically to the Dregion sequence: particularly in the case of the TCR 6 genes,where the D segment can be read in three different reading frames and two D segments can join together,such DD combinations produce a longer third complementarity determining region (CDR3) than is found in other TCR or antibody molecules. Sincethe CDR3 in the various receptor chainsis essentially composed of the regions between tll.e V(D)I segments, where junctional diversity mechanisms can introduce a very high degree of amino acid variability, one can seewhy it is that this hypervariable loop usually contributes the most to determining the fine antigenbinding specificity of thesemolecules.

GERM-uNE_RECoMB|I{ED-PRoIEIN DNA-DilA-SEqUEilCE V"

J.,

r*T;;;l

l-;;l

luuuluuvl

lrvel ----T------

lccclTGGl

- ProTrp-

iccclcccl@

;;T;;;-t

- ProArg-

tc6Tc-tcl@

[;T;;;l

- ProPro-

lvuuluuel

Figure 4.8. Junctional diversity between a TCR Va and Jclgermline segment producing three variant protein sequences. The nucleotide triplet which is spliced out is colored the darker blue For TCR B-chain and Ig heavy chain genes junctional diversity can apply to V, D and I segments

Receptor editing Recent observations have established that lymphocytes are not necessarily stuck with the antigen receptor they initiallymake; if they don'tlike itthey can changeit. The replacement of an undesired receptor with one which has more acceptable characteristics is referred to as receptor editing. This process has been described for both immunoglobulins and for TCR, allowing the replacement of either nonfunctional rearrangements or autoreactive specificities. Furthermore, receptor editing in the periphery may rescue low affinity B-cells from apoptotic cell death by replacing a low affinity recePtor with a selectable one of higher affinity. That this does indeed occur in the periphery is strongly supported by the finding that mature B-cells in germinal centers can express RAG-1 and RAG-2 which mediate the rearransem*ntprocess.

Iro

C H A P I E R4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

But how does this receptor editing work? Well, in the caseof the receptor chains which lack D gene segments, namely the immunoglobulin light chain and the TCR cr chain, a secondary rearrangement may occur by a I/ gene segment upstream of the previously rearranged V/ segment recombining to a 3'/ gene sequence,both of these segments having intact RSSsthat are compatible (figure 4.9a). However, for immunoglobulin heavy chains and TCR B chains the process of VDI rearrangement deletes all of the D segment-associated RSSs (figure 4.9b). BecauseVn and /" both have 23 base pair spacers in their RSSs, they cannot recombine: that would break the 72/23 rule. This apparent obstacle to receptor editing of these chains may be overcome by the presence of a sequencenear the 3'end of the l/ coding sequences that can function as a surrogate RSS,such that the new V segment would simply replace the previously rearranged V, maintaining the same D and / sequence(figure 4.9b). This is probably a relatively inefficient process and receptor editing may therefore occur more readily in immunoglobulin light chains and TCR s chains than in immunoglobulin heavy chains and TCR B chains. Indeed, it has been suggested that the TCR crchain may undergo a series of rearrangements, continuously deleting previously functionally rearranged V/ segments until a selectable TCR is produced.

Inlerchoin omplificotion The immune system took an ingenious step forward when two different types of chain were utilized for the recognition molecules because the combination produces not only a larger combining site with potentially greater affinity, but also new variability. Heavy-light chain pairing amongst immunoglobulins appears to be largely random and therefore two B-cells can employ the same heavy chain but different light chains. This route to producing antibodies of differing specificity is easily seen in ztitro where shuffling different recombinant light chains against the same heavy chain can be used to either fine tune or sometimes even alter the specificity of the final antibody. In general, the available evidence suggests that in uiao the major contribution to diversity and specificity comes from the heavy chain, perhaps not unrelated to the fact that the heavy chain CDR3 gets off to a head start in the race for diversity being, as it is, encoded by the junctions between three gene segments: V, D and l. This random association between TCR yand 6 chains, TCRaand p chains,and Igheavyand light chainsyields a further geometric increase in diversity. From table 4.2 it can be seen that approximately 230 functional TCR

and 153 functional Ig germ-line segments can give rise to 4.5 million and 2.3 million different combinations, respectively, by straightforward associations without taking into account all of the fancy junctional mechanisms described above. Hats off to evolution!

Somolichypelmulotion As discussed in Chapter 3, there is inescapable evidence that immunoglobulin V-region genes can undergo significant somatic hypermutation. Analysis of 18 murine l" myelomas revealed 12 with identical structure, four showing just one amino acid change, one with two changes and one with four changes, all within the hypervariable regions and indicative of somatic hypermutation of the single mouse l, germ-line gene. In another study, following immunization with pneumococcal antigen, a single germ-line T15 7" gene gave rise by mutation to several different V" genes all encoding phosphorylcholine antibodies (figure 4.10). A number of features of this somatic diversification phenomenon are worth revisiting. The mutations are the result of single nucleotide substitutions, they are restricted to the variable as distinct from the constant region and occur in both framework and hypervariable regions. The mutation rate is remarkably high, approximately 1 x 10-3per base pair per generation, which is approximately a million times higher than for other mammalian genes. In addition, the mutational mechanism is bound up in some way with class switch recombination since the enzyme activation-induced cytidine deaminase (AID) is required for both processes and hypermutation is more frequent in IgG and IgAthan in IgM antibodies, affecting both heavy (figure 4.10) and light chains. However, 7n genes are on average more mutated than V. 8enes. This might be a consequence of receptor editing acting more frequently on light chains, as this would have the effect of wiping the slate clean with respect to light chain 7 gene mutations whilst maintaining already accumulated heavy chain V gene pointmutations. Somatic hypermutation does not appear to add significantly to the repertoire available in the early phases of the primary response, but occurs during the generation of memory and is responsible for tuning the response towards higher affinity. Recently data have been put forward suggesting that there is yet another mechanism for creating further diversity. This involves the insertion or deletion of short stretches of nucleotides within the immunoglobulin 7 gene sequence of both heavy and light chains. This mechanism would have an intermediate effect on antigen recognition, being more dramatic than single

C H A P T E 4R- M E M B R A N ER E C E P T O R FS O RA N T I G E N

D* so 3

VrooQrsJHz

------fg vH3sDH3JH2 Figure 4.9. Receptor editing. (a) For imrnunoglobulin light chain or TCR cx chain the recombination signal sequences (RSSs; heptamer-nonamer motifs) at the 3' of each variable (li) segment and the 5' of each joining (/) segment are compatible with each other and therefore an entirely new rearrangement can potentially occur as shown. This would result in a receptor with a different light chainvariable sequence (in this example Vk rrlkoreplactngVkrf) together with the original heavy chain. (b) With respect to the immunoglobulin heavy chain or TCR p chain the organization of theheptamer-nonamer sequencesin the RSSprecludes a Vsegment directly recombining with the / segment. This is the so-called 72/23 rule whereby the hep-

tamer-nonamer sequences associated with a 23 base pair spacer (colored violet) can only base pair withheptamer-nonamer sequences containing a 12 base pair spacer (colored red). The heavy chain Vand/ both have an RSSwith a 23 base pair spacer and so this is a nonstarter. Furthermore, all the unrearranged D segments have been deleted so that there are no 12 base pair spacers remaining This apparent bar to secondary rearrangement is probably overcome by the presence of an RSS-like sequence near the 3'end of the V gene coding sequences, so that only the V gene segment is replaced (in the example shown, the sequence VrrrrD,rl nrreplacesVnooDnJn ).

point mutation, but considerably more subtle than receptor editing. In one study, a reverse franscriptasepolymerase chain reaction (R|-PCR) was employed to amplify the expressed V" and 7, genes from 365 IgG+ Bcells and it was shown that 6.5% of the cells contained

nucleotide insertions or deletions. The transcriPts were left in-frame and no stop codons were introduced by these modifications. The percentage of cells containing these alterations is likely to be an underestimate. All the insertions and deletions were in, or near to, CDR1

lrt

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

Sequenliol num0eflng Hypervorio regron: 2 l---l HPCM I IHPCM3 I M llgMlHPO I IHPCM6 I IHPCM4 8 [--l HPCM t3 I IHPCM I E G I H P Cl 4M I I H P C lMl I r H P C 1M2 Figure 4.10. Mutations in a gerrn-line gene. The amino acid sequencesoftheVrrregionsoffivelgMandfivelgGmonoclonalphosphorylcholine antibodies generated during an antipneumococcal responseinasinglemousearecomparedwiththeprimarystructureof the T15 germ-line sequence Aline rndicates identity with the T15 prototype and an orange circle a single amino acid difference Mutations

have only occurred in the IgG molecules and are seen in both hypervariableandframeworksegments (AfterGearhartP.J.(1982)ImmunologyToday3,l0T )Whilstinsomeotherstudiessomatichypermutatron has been seen in IgM antibodies, the amount of mutation usually greatly increases following class switching.

and/or CDR2. N-region diversity of the CDR3 meant that it was not possible to analyse the third hypervariable region for insertions/deletions of this type and therefore these would be missed in the analysis. The fact that the alterations were associated with CDRs does suggest that the B-cells had been subjected to selection by antigen. It was also notable that the insertions/deletions occurred at known hot spots for somatic point mutation, and the same error-prone DNA polymerase responsible for somatic hypermutation may also be involved here. The sequenceswere often a duplication of an adjacent sequence in the case of insertions or a deletion of a known repeated sequence. This type of modification may, like receptor editing, play a major role in eliminating autoreactivity and also in enhancing antibody affinity. T-cell receptor genes, on the other hand, do not generally undergo somatic hypermutation. It has been argued that this would be a useful safety measure since T-cells are positively selected in the thymus for weak reactions with self MHC (cf. p.234), so that mutations could readily lead to the emergence of high affinity autoreactive receptors and autoimmunity. One may ask how it is that this array of germ-line genes is protected from genetic drift. With a library of 390or so functional V ,D andl genes,selectionwould act only weakly on any single gene which had been functionally crippled by mutation and this implies that a major part of the library could be lost before evolutionary forces operated. One idea is that each subfamily of related V genes contains a prototype coding for an antibody indispensable for protection against some commonpathogen, so that mutation in this gene would put the host at a disadvantage and would therefore be

selectedagainst.If any of the other closely related genes in its set became defective through mutation, this indispensable gene could repair them by gene conversion, a mechanism in which it will be remembered that two genes interact in such a way that the nucleotide sequenceof part or all of one becomesidentical to that of the other. Although gene conversion has been invoked to account for the diversification of MHC genes, it can also act on other families of genesto maintain a degree of sequence hom*ogeneity. Certainly it is used extensively by, for example, chickens and rabbits, in order to generate immunoglobulin diversity. In the rabbit only a single germ-line Vn gene is rearranged in the majority of B-cells; this then becomes a substrate for gene conversion by one of the large number of V" pseudogenes. There are also large numbers of V" pseudogenes and orphan genes (genes located outside the gene locus, often on a completely different chromosome) in humans which actually outnumber the functional genes, although there is no evidence to date that these are used in gene conversion processes.

N KR E C E P T O R S Natural killer (NK) cells are a population of leukocytes that, like T- and B-cells, employ receptors that can provoke their activation, the consequencesof which are the secretion of cytokines, most notably IFN-y, and/ or cytotoxic granules that are capable of killing the cells they target. Unlike the antigen receptors of T- and B'hard-wired' lymphocytes, NK receptors are and do not undergo VDJ recombination to generate diversity. NK cells, unlike aB T-cells, are not MHC-restricted in the sensethat they do not see antigen only when presented

7sl

C H A P T E 4R- M E M B R A N ER E C E P T O R FS O RA N T I G E N within the groove of MHC classI or MHC classII molecules. On the contrary, one of the main functions of NK cells is to patrol the body looking for cells that have lost expression of the normally ubiquitous MHC class I molecules; a situation that is known as 'missing-self' recognition (figure 4.11). Such abnormal cells are usually either malignant or infected with a microorganism that interfereswith classI expression.Becauseof the central role that MHC classI molecules play in presenting peptides derived from intracellular pathogensto the immune system/ it is relatively easy to understand why these molecules may attract the unwelcome attentions of viruses or other uninvited guests planning to gatecrash their cellular hosts. It is probably for this reason that NK cells co-evolved alongside MHC-restricted Tcells to ensure that pathogens, or other conditions that may interefere with MHC class I expression and hence antigen presentation to crBT-cells, are given short shrift. Cells that end up in this unfortunate position are likely to soon find themselves looking down the barrel of an activated NK cell. Such an encounter typically results in death of the errant cell as a result of attack by cytotoxic granules, containing a battery of proteases and other destructive enzymes releasedby the activated NK cell. NKreceplorsconbe oclivotingor inhibitory NK cells play an important role in the ongoing battle against viral infection and tumor development and carry out their task using two setsof receptors;activating receptors, that recognize molecules collectively present on all cell surfaces, and inhibitory receptors that recognize MHC classI molecules. It is the balance between inhibitory and activating stimuli that will dictate whether NK-mediated killing will occur (figure 4.71). TWo structurally distinct families of NK receptors have been identified: the C-type lectin receptors (CTLRs) and the lg-like receptors. Both receptor types include inhibitory and activating receptors.Those that are inhibitory contain ITIMs (lmmunoreceptor fyrosine-based lnhibitory motifs) within their cytoplasmic tails that exert an inhibitory function within the cell by recruiting phosphatases,such as SHP-1, that can antagonize signal transduction events that would otherwise lead to release of NK cytotoxic granules or cytokines (figure 4.12).Activating receptors, on the other hand, are associated with accessory proteins, such as DAP-12, that contain positively acting ITAMs within their cytoplasmic tails that can promote events leading to NKmediated attack. Upon engagement with their cognate ligands (MHC class I molecules), inhibitory receptors

Inhibiting receptor MHCClossI

Self

Noresponse

receptor Activoling Missing self

Aclivoling ligond ,,//

MHCclossI

lnduced self

NKottocks lorgetcell ligonds Activoling

Benefit of doubt?

Figure 4.11. Natural killer (NK) cell-mediated killing and the 'missing-self'hypothesis. (a) Upon encounter with a normal autologous MHC classI-expressingcell, NK inhibitoryreceptors are engaged and activating NK receptors remain unoccupied because no activating ligands are expressed on the target cell The NK cell does not become activatedinthis situation (b)Lossof MHCclasslexpression('missingself'), as well as expression of one or more ligands for activating NK receptors, provokes NK-mediated attack of the cell via NK cytotoxic granules. (c) Upon encountering a target cell expressing MHC class I, but also expressing one or more ligands for activating NK receptors ('induced-self'), the outcome will be determined by the relative strength of the inhibitory and activating signals received by the NK cell (d) In some cases,cells may not express MHC class I moiecules or activating ligands and may be ignored by NK cells possibly due to expression of alternative ligands for inhibitory NK receptors

suppress signals that would otherwise lead to NK cell activation. Cells that lack MHC class I molecules are therefore unable to engage the inhibitory receptors and are likely to suffer the consequences(figure 4.11). The main class of MHC class I-monitoring receptors in the mouse is represented by the Ly49 multigene family of receptors that contains approximately 23 distinct genes;Ly49Ato W. These receptors are expressed as disulfide-linked hom*odimers, with each monomer composed of a C-type lectin domain connected to the cell membrane via an g-helical stalk of -40 amino acids (figure 4.72a); each NK cell expresses from one to four different Ly49 genes.Rather remarkably, humans do not use Ly49-based receptors to carry out the same task, but instead emplov a functionally equivalent, but struc-

lro

CHAPTER 4 _ M E M B R A N ER E C E P T O RFSO RA N T I G E N

Ly49inhibiforyreceplor CTLD

Ly49Arecepiol CTLD

DAP-I 2

ITIM

KIRrecepfor

Figure 4.12. NK receptors. (a) Schematic representation of an inhibitory Ly49 receptor dimer composed of two C-type lectin domains (CTLD) The cytoplasmic tails of inhibitory Ly49 receptors contain lmrnunoreceptor fyrosine-based lnhibitory motifs (ITIMs) that can recruit phosphatases, such as SHP-1, capable of antagomzing NK activation. Activating Ly49 receptors lackITIMs and can associatewith ITAM-containing accessoryproteins such as DAP-12 that can promote NK cell activation (b) C-type lectin-like domain of the Ly49 NK cell receptors The three-dimensional structure shown is the dimeric Ly49A(Protein Data Bank entry code 1QO3), the monomerAis colored blue and the monomer B is colored green For clarity, secondary structural elements o-helices, p-strands, disulfide bonds and N and C termini are labeled only on one monomer (Kindly provided by Dr Nazzareno Dimasi.) (c) The human KIRs (killer lmmunoglobulin{ike

KIRreceplor

receptors) are functionally equivalentto the murine Ly49 receptorsbut remain structurally distinct. These receptors contain two or three Iglike extracellular domains and can also be inhibitory or activating depending on the presence of an ITIM motif in their cytoplasmic domains, as shown Activating receptors can associatewith the ITAMbearing DAP-12 accessory complex to propagate activating signals into the NK cell that result in NK-mediated attack (d) Structure of the extracellular Ig-like domains (D1 and D2) of a KIR receptor (Kindly provided by Dr Peter Sun and based upon coordinates originally published in Boyington ef al (2000)Nature 405, 537 ) (e) Ribbon diagram of the crystal structure of theLy4gC/H-2Kb complex Ly49C, the H-2Kb heavy chain, and BrM are shown in red, gold and green, respectively. The MHC-bound peptide (gray) is drawn inball-and-stick representation. (Kindly provided by Dr Lu Deng and Professor Roy A Mariuzza.)

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N turally distinct, set of receptors for this purpose, the killer immunoglobulin-like receptors (KIRs). This is a good example of convergent evolution where unrelated genes have evolved to fufill the same functional role. IndividuaILy49 receptors recognize MHC class I molecules in a manner that is, in most cases,independent of bound peptide. Ly49 dimers make contact with MHC class I molecules at two distinct sites that do not significantly overlap with the TCR binding area on the MHC (figure 4.72e). By contrast, the KIRs make contact with MHC class I molecules in an orientation that resembles the docking mode of the TCR where contact withbound peptide is part of the interaction. However, it is worth emphasizing that although KIRs do make contact with peptide within the MHC class I groove, thesereceptorsdo not distinguishbetween self and nonself peptides as TCRs do. In addition to recognizing 'missing-self', NK cells also use their receptors to directly recognize pathogen components or MHC class I-like proteins, such as the MHC class I chain-related A chain (MICA), that are normally poorly expressed on normal healthy cells. MICA, and related ligands, have a complex pattern of expression but are often upregulated on transformed or infected cells and this may be sufficient to activate NK receptors that are capable of delivering activating signals; a phenomenon that has been termed'inducedself'recognition (figure 4.11).Upon ligation, the activating receptors signal the NK cell to kill the target cell and/or to secrete cytokines. The potentially anarchic situation in which the NK cells would attack all cells in the body is normally prevented due to the recognition of MHC class I by the inhibitory receptors. Another example of an activating NK receptor is CD16, the low-affinity Fc receptor for IgG that is responsible for antlbody-dependent cellular cytotoxicity (ADCC, cf . p . 32).In this case,the receptor ligand is IgG bound to antigen present on a target cell which is clearly an abnormal situation. The ligands for many of the other NK activating receptors remain obscure at present but this area is one of active investigation and is sure to yield interesting insights in the near future.

Cell stressond DNAdomogetesponsesconoclivoleNKcells Cellular stresscan also lead to NK cell activation. The CD94UNKG2 gene family have been found in human, rat and mouse genomesand belong to the CTLR classof NK receptors.Thesereceptorscan exist asCD94/CD94 hom*odimers or as CD94INKG2 heterodimers and are expressed on most NK cells as well as y6 T-cells. CD94INKC2A heterodimers are inhibitory receptors that recognize the MHC class I-related molecules, HLA-

75 1

E (human) and Qalb (mouse), which are notable for the fact that they mainly bind peptides that are found in the leader sequencesof the classicalMHC classI molecules. CD94INKG2 receptors appear to monitor the expression status of MHC classI molecules indirectly because in the absence of the leader sequences from these peptides, HLA-E and Qalb are not expressed on the cell surface, thereby triggering NK attack. In the context of cellular stress, heat-shock proteins such as HSP-60 are induced and peptides derived from this heat-shock protein can displace MHC class I-derived peptides; this results in NK activation because CD94INKG2 heterodimers cannot bind to HLA-E molecules that contain HSP-60peptides. Very recent studies also suggest that checkpoint kinases, such as Chk1, that are involved in the DNA damage response can induce expression of NKG2D receptor ligands when a cell is damaged by y-irradiation, or after treatment with DNA-damaging drugs. This suggests that cells that have suffered DNA damage may, in addition to activating their DNA repair machinery, also upregulate NK receptor ligands to alert the immune system; such cells are dangerous as they have the potential to escapenormal growth controls due to faulty or incomplete DNArepair.

MPTEX TY T H EM A J O RH I S T O C O M P A I I B I TCIO (MHC) Molecules within this complex were originally defined by their ability to provoke vigorous rejection of grafts exchanged between different members of a species (Milestone 4.2).We have already referred to the necessity for antigens to be associated with class I or class II MHC molecules in order that they may be recognized by T-lymphocytes. Let us now look at these molecules in greater detail.

ore ClossI ondclossll molecules hetelodimels membrone-bound MHCclassl Class I molecules consist of a heavy polypeptide chain of 44kDa noncovalently linked to a smaller The 72kDa polypeptide called B,-microglobulin. Iargest part of the heavy chain is organized into three globular domains (ay azand cru;figure 4.13) which protrude from the cell surface; a hydrophobic section anchors the molecule in the membrane and a short hydrophilic sequence carries the C-terminus into the cytoplasm. The solution of the crystal structure of a human class I molecule provided an exciting leap forwards in our

lru

4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N CHAPTER

Peter Gorer raised rabbit antiserums to erythrocytes from pure strain mice (resulting from >20 brother-sister matings) and, by careful cross-absorption with red cells from different strains, he identified the strain-specific antigen II, now known asH-2 (tableM4.2.1).

single gene locus, H-2 proved to be a large complex of multiple genes,many of which were highly polymorphic, hence the term major histocornpatibility complex (MHC). The major components of the current genetic maps of the human HLA and mouse H-2 MHC are drawn in figure M4.2.1 to give the

He next showed that the rejection of an albino (A) tumor by black (C57) mice was closely linked to the presence of the antigen II (table M4.2.2) and that tumor rejection was associated with the development of antibodies to this antigen.

reader an overall grasp of the complex make-up of this important region (to immunologists we mean!-presumably all highly transcribed regions are important to the host in some

Subsequently,George Snell introduced the term histocompatibility (H) antigen to describe antigens provoking graft

*uY)' TabIe M4.2.2. Relationship of antigen II to tumor rejection.

rejection and demonstrated that, of all the potential H antigens, differences at the H-2 (i.e. antigen II) locus provoked the strongest graft rejection seenbetween various mouse strains. Pocoa poco,the painstaking studies gradually uncovered a

Antigen ll phenolype ot recipienl slroin

remarkably complicated situation. Far from representing a TableM4.2.L. Identification of H-2 (antigenII).

(Aslroin)by: Rejeclion oflumorinoculum *Pureslroin --(A x C57)Fl bockcross lo C57

Ag ll +ve (A)

Ag ll -ve (C57)

17(r93) 0

17(r95) 44 (3e)

+Alumorinoculum derived fromA stroinbeoring ontigen ll is rejecled bylhe - = occeptonce). C57host(+ = rejeclion; **Offspring to lheC57porenlondthe of A x C57motingwerebockcrossed progenytesledforontigen resulfing ll (Agll) ondtheirobililytorejectlhetumor. = numberexpected Thefigures inbrockets iftumorgrowlh isinfluenced bylwo genes, oneofwhichdetermines lhepresence of0nligenll dominonl

Figure M4.2.1. Main genetic regions of the majorhistocompatibility

understanding of MHC function. Both Br-microglobulinand the o, regionresemble classic Ig domains in their folding pattern (cf. figure 4.13c).However, the cr1and cr2 domains, which are most distal to the membrane, form two extended cr-helicesabove a floor created by strands held together in a B-pleated sheet, the whole forming an undeniable groove (figure 4.73b,c).The appearance of these domains is so striking, we doubt whether the reader needs the help of gastronomic analogies such as 'two sausageson abarbecue'to prevent any classI structural amnesia. Another curious feature emerged. The groove was occupied by a linear molecule, now known to be a peptide, which had cocrystallized with the class I protein (figure 4.14). How antigenic peptides are processed and selected for presentation within MHC

complex.

molecules and how the TCR sees this complex will be revealed in the following chapter. MHCclasslI Class II MHC molecules are also transmembrane glycoproteins, in this case consisting of c and p polypeptide chains of molecular weight 34 kDa and 29 kDa, respectively. There is considerable sequencehom*ology with class I and structural studies have shown that the u, and B, domains, the ones nearestto the cell membrane, assume the characteristic Ig fold, while the o, and F1 domains mimic the classI o, and c,, informing a groovebounded by two cr-helices and a B-pleated sheet floor (figures 4.13aand4.14\.

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

77 1

c PEPTIDE BINDING CLEFT

Figure 4.13. Class I and class II MHC molecules. (a) Diagram showing domains and transmembrane segments; the cr-helicesand psheets are viewed end on (b) Schematic bird's eye representation of the top surface of human class I molecule (HLA-A2) based on the X-ray crystallographic structure. The strands making the p-pleated sheet are shown as thickgray arrows in the amino to carboxy direction; a-helices are represented as dark red helical ribbons. The inside-facing surfaces

of the two helices and the upper surface of the B-sheetform a cleft. The two black spheres represent an intrachain disulfide bond (c) Side view of the same molecule clearly showing the anatomy of the cleft and the typical lg-type folding of the c.- and pr-microglobulin domains (four antiparallel p-strands on one face and three on the other) (Reproduced from Bjorkman P.J.et al. (1987)Nature329,506,with permission )

lrt

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N H-2Kb SEVgpeplide

l-As7 GAD65peplide

The organization of the genes encoding the a chain of the human class II molecule HLA-DR and the main regulatory sequences which control their transcription are shown in figu re 4.15. MHGclossI ondclossll molecules orepolygenic

Figure 4.14. Surface view of mouse class I and class II MHC molecules in cornplex with peptide. Surface solvent-accessible areas of the mouse class I molecule (H-2Kb) in complex with a virus-derived peptide and the mouse class II molecule I-A87 in complex with an endogenous peptide. The views shown here are similar to that schematicallydepicted infigure4.13b and lookdownupon the surface of the MHC molecules. Note that the peptide-binding cleft of class I molecules is more restricted than that of class II molecules with the result that class I-binding peptides are typicaliy shorter than those that bind to class II molecules. (Kindly provided by Dr Robyn Stanfield and Dr Ian Wilson, Department of Molecular Biology, The Scripps Research Institute, La Jolla, California, USA )

LPSlFNc/0

Several different flavors of MHC class I and class II proteins are expressed by most cells. There are three different class I o-chain genes, referred to as HLA-A, HLA-B andHLA-C inman andH-2K,H-2D andH-2Lin the mouse, which can result in the expression of at least three different class I proteins in every cell. This number is doubled if an individual is heterozygotrs for the class I alleles expressed at each locus; indeed, this is often the casedue to the polymorphic nature of classI genesaswe shall discuss later in this chapter. There are also three different types of MHC class II oand B-chain genes expressed in humans, HLA-DQ, HLA-DP, aladHLA-DR, and two pairs in mice, H2-A (IA) and H2-E (I-E). Thus, humans can express a minimum of three different class II molecules, with this number increasing significantly when polymorphisms are considered; this is because different a and B chain combinations can be generated when an individual is heterozygous for a particular classII gene. The different types of class I and class II molecules all exhibit the same basic structure as depicted in figure 4.13aand all participate in presenting peptides to T-cells but, because of significant differences in their peptidebinding grooves, each presents a different range of peptides to the immune system. This has the highly desirable effect of increasing the range of peptides that

HLA-DRcr

(in mocrophoges) DOWNREGULATIoN Figure 4.15. Genes encoding human HLA-DRcn chain (darker blue) and their controlling elements (regulatory sequences in light blue and TATA box promoterin yellow). crr/cr, encode the two extracellular domains; TM and CYT encode the transmembrane and cytoplasmic segments, respectively. 3'UT represents the 3' untranslated sequence Octamer motifs are also found in virtually all heavy and light chain immunoglobulin V gene promoters(cf figure3.21) andinthepromotersofotherB-cell-specificgenessuchasB29andCD2?.

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N

7sl

can be presented to T-cells and reduces the likelihood that peptides derived from pathogen proteins will fail to be presented. Class I and classII MHC molecules probably evolved from a single ancestral gene that underwent serial gene duplications, followed by diversification due to selective pressure, to generate the different class I and class II genesthatwe seetoday (figure4.16).Genesthatfailed to confer any selective advantage or that suffered deleterious mutations were either deleted from the genome or are still present as pseudogenes (genes that fail to express a functional protein); indeed many pseudogenes are present within the MHC region. This type of gene evolution pattern has been termed the birth-anddeath model or the accordion model due to the way in which this generegionexpanded and contractedduring evolution.

factor B. The cytokines, tumor necrosis factor (TNF, sometimes referred to as TNFa) and lymphotoxin (LTu and LTp), are encoded under the classIII umbrella as are three members of the human 70 kDa heat-shock proteins. As ever, things don't quite fit into the nice little boxes we would like to put them in. Even if it were crystal clear where one region of the MHC ends and another begins (and it isn't), some genes located in the 'classical' (cf. figure 4.17) middle of the class I or II regions should more correctly be classified as part of the class III cohort. For example, the LMP and 7XP genes concerned with the intracellular processing and transport of T-cell epitope peptides are found in the class II region (seebelow), but do not have the classicalclassII structure nor are they expressed on the cell surface.

genes Severol immune response-reloted conlribule to lheremoining cl0ss lll region oftheMHc

The complete sequenceof a human MHC was published at thevery end of the lastmillennium after a gargantuan collaborative effort involving groups in England, France, Japan and the USA. The entire sequence,which represents a composite of several MHC haplotypes, comprises 224 gene loci. Of the 128 of these genes which are predicted to be expressed,it is estimated that about 40"/" of them have functions related to the immune system. It is not clear why so many immune responserelated genes are clustered within this relatively small region, although this phenomenon has also been observed with housekeeping genes that share related functions. Because the location of a gene within chromatin can profoundly influence its transcriptional activity, perhaps it has something to do with ensuring that the geneswithin this region are expressed at similar levels. Genes found within condensed regions of chromatin are often expressed at relatively low levels and in some cases may not be expressed at all. The region between class II and class I in the human contains 60 or so classIII genes.An overall view of the main clustersof class I, II and III genes in the MHC of the mouse and human may be gained from figure M4.2.7 in Milestone 4.2. More detailed maps of each region are provided in figures 4.174.79. A number of pseudogenes have been omitted from these gene maps in the interest of simplicity. The cell surface class I molecule, based on a transmembrane chain with three extracellular domains associated with Br-microglobulin, has clearly proved to be a highly useful structure judging by the number of variants on this theme that have arisen during evolution. It is helpful to subdivide them, first into the classical class I molecules (sometimesreferred to as classIa), HLA-A, -B and -C in the human and H-2K. -D and -L in the

A variety of other genes which congregate within the MHC chromosome region are grouped under the heading of classIII. Broadly, one could say that many are directly or indirectly related to immune defense functions. A notable cluster involves four genes coding for complement components, two of which are for the C4 isotypes C4A and C4B and the other two for C2 and

BIRTHAT{DDEAIHEVOTUTIOI{ iIODEtOF]{HC ANCESTRAL GENE

DUPLICATION GENE EVENTS

GENEDIVERSIFICATION THROUGH MUTATION ANDSELECTION

FURTHER DIVERSIFICATION

GENE LOSS Figure 4.16. Birth and death model of MHC evolution. Different MHC genes most likely arose though duplication events that resulted in diversification of the duplicated genes as a result of selective pressure Genes that confer no selective advantage can suffer deleterious mutations resulting in pseudogenes or may be deleted from the genome altogether Different environments rmpose distinct selective pressures/ due to different pathogens for exarnple, resulting in a high degree of polymorphism within this gene family. MHC polymorphism is seen primarily within the peptide-binding regions of MHC class I and class II molecules.

Genemopof lhe MHC

the MHC locus itself ('nonclassical'MHC molecules, for example the human HLA-E, -F and -G, HFE, MICA and MICB, the murine H-2T, -Q and -M), or elsewhere in the genome ('class I chain-related', including the CD1 family and FcRn). Nonclassical MHC genes are far less

mouse. These were defined serologically by the antibodies arising in grafted individuals using methods developed from Gorer's pioneering studies (Milestone 4.2). Other molecules, sometimes referred to as class Ib, have related structures and are either encoded within

MOUSE

H.2 GENE

IAPASIN

H.zL

GENEPRODUCT 'classical'polymorphic

M H-2[/

Figure 4.17. MHC class I gene map. The class I genes, HLA-4, -8, -C in humans and H-2K, -D, -L rn mice, are highlighted with orange shading and encode peptide chains which, together with B2-microglobulin, form the compiete class I molecules originally identified in earlier studies as antigens by the antibodies theyevokedon graftingintoanothermemberof the samespecies Note that only some strains of mice possess an H-2L gene The genes expressed most abundantly are HLA-A and -B tn the human and H-2K and -D in the mouse. The other class I genes ('class Ib') are termed 'non-

classical'or'class I chain-related'. They are oligo- rather thanpolymorphic or sometimes invariant, and many are silent or pseudogenes In the mouse there are approximately 15 Q (also referred to as Qa) genes, 25 T (also referred to as TL or Tla) genes and 10 M genes. MICA and MICB are ligands for NK cell receptors Tapasin is involved in peptide transport (cf. p.97). The gene encoding this molecule is at the centromeric end of the MHC region and therefore is shown in this gene map with respect to the mouse, but in figure 4 18, the class II gene map with respect to the human -look at figure M4.2.1 to seewhy.

Figure 4.18. MHC class II gene map with 'classical' HLA-DP, -DQ, -DR inthe humanandH-2A(I-A) andH-2E (/-E) in mice moreheavily shaded Both a and B chains of the class II heterodimer are transcribed from closely located genes. There are usually two expressed DRB genes, DRB1 and one of either DRB3, DRB4 or DRB5. A similar situation of a single o chain pairing with different B chains is found in the mouse I-E molecule. The IMP2 andLMPT genesencodepartof theproteasome complex which cleaves cytosolic proteins into small peptides which are transported by the ?hP gene products into the endoplasmic reticulum. HLA -DMA and -DMB (morse H-2D Ma, -DMb1 and -D Mb2) encode the DM crB heterodimer which removes class Il-associated

invariant chain peptide (CLIP) from classical class II molecules to permit the binding of high affinity peptides. The mouse H-2DM molecules are often referred to as H-2M1 and H-2M2, although this is a horribly confusing designation because the terrn H-2M rs also used for a completely different set of genes which lie distal to the H-2 T region and encode members of the class Ib family (cf. figure417).The HLA-DOA (alternatively called HLA-DNA) and -DOB genes (H-2Oa and -Ob inthe mouse) also encode an oB heterodimer which may play a role in peptide selection or exchange with classical class II molecules (Reproduced with permission from Nature Reviews Immunology VoI. 5, No 10, pp. 783-792 (2005),Macmillan Magazines Ltd.)

HUMAN UP2IB

UP2IA

MOUSE CYNIAI

CW2IM

Figure 4.19. MHC class III gene map. This region is something of a 'rag bag'. Aside from immunologically'respectable' products like C2, C4, factor B (encoded by the BF gene), tumor necrosis factor (TNF), lymphotoxin-cx, and lymphotoxin-B (encoded by LTA and LTB, rcspectively) and three 70 kDa heat-shock proteins (the HSPAIA, HSPAI.B andHSPA1Lgenes inhumans, HSP70-1,H5P70-3 and Hsc7Ofgenes in mice), genes not shown in this figure but nonetheless present in this locus include those encoding valyl tRNA synthetase (G7a),NOTCH4, which has a number of regulatory activities, and tenascin, an extra-

HSPAIB

HSPTGI

HSPAIA

HSPAIL

HSP70-3 Hs70t

INF

LTA

ITVF

LTA

cellular matrix protein. Of course many genes may have drifted to this location during the long passage of evolutionary time without necessarily having to act in concert with their neighbors to subserve some integrated defensive function. The 21-hydroxylases (21OHA and B, encoded by CYP2I-A and, CYP21B, respectively) are concemed with the hydroxylation of steroids such as cortisone Slp (sex-limited protein) encodes a murine allele of C4, expressed under the influence of testosterone

8rl

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N I I G E N polymorphic than the classical MHC, are often invariant, and many are pseudogenes.Many of thesenonclassical MHC class I molecules form structures that are very similar to class I molecules and have also been found to serve as antigen-presenting molecules in particular contexts. polymorphism Thegenes0f theMHCdisployremorkoble Unlike the immunoglobulin system where, as we have seen, variability is achieved in each individual by a multigenic system, the MHC has evolved in terms of variabilitybetween individuals with a highly polymorphic (literally 'many shaped') systembasedon multiple alleles (i.e. alternative genes at each locus). This has likely arisen through pathogen-driven selection to 'fitness' form new allelles that may offer increased for the individual; in this context, fitness could mean increased protection from an infectious organism. The classI and classII genesare the most polymorphic genes in the human genome; for some of these genes over 600 allelic variants have been identified (figure 4.20).This implies that there has been intense selective pressure on the MHC gene region and that genes within this region are mutating at rates much faster than at other gene loci. As is amply illustrated in figure 4.20,classI HLA-A, -B and -C molecules are highly polymorphic and so are the class II B chains (HLA-DRP most, -DPB next and -DQp third) and, albeit to a lesser extent than the B chains, the cr chains of -DP and -DQ. HLA-DRcr and Br-microglobulin are invariant in structure. The amino acid changes responsible for this polymorphism are restricted to the cr,and u, domains of classI and to the u, and B, domains of classII. It is of enormous significance that they occur essentially in the B-sheetfloor and on the inner surfaces of the cr-helices which line the central cavity (figure4.13a)and also ontheupper surfacesof the helices; these are the very surfaces that make contact with the peptides which theseMHC molecules offer up for inspectionby TCRs (figure 4.14).The ongoing drive towards creating new MHC molecules, with slightly altered peptide-binding grooves, is akin to a genetic arms race where the immune system is constantly trying to keep one step ahead of its foe. This genetic oneupmanship has been termed pathogen-driven balancing selection because heterozygotes typically have a selectiveadvantage over hom*ozygotes at a given locus. The MHC region represents an outstanding hotspot with mutation rates two orders of magnitude higher than non-MHC loci. These multiple allelic forms can be generated by a variety of mechanisms: point mutations, recombination, hom*ologous but unequal crossing over and gene conversion.

700 600 o o

Fnn

E 400 E 3oo E 2 2oo 100 0

200

r00 DRADRBDOADOBDPADPB HLAlocus

Figure 4.20. Polymorphism within hurnan HLA class I and class II genes. Number of distinct human HLA class I (A, B, C) and class II (DRA, DRB, DQA, DQB, DPA,DPB) alleles at each locus as of January 2005. (Based upon data gathered by the WHO Nornenclature Committee for Factors of the HLA system and published within Marsh el a/ (2005) TissueAnt igens 65, 307 )

The degree of sequence hom*ology and an increased occurrenceof the dinucleotide motif 5'-cytosine-guanine3' (to produce what are referred to asCpG islands) seemto be important for gene conversion, and it has been suggested that this might involve a DNA-nicking activity which targets CpG-rich DNAsequences. MHC genesthat lackthese sequences,forexample H-2EadandHLA-DRA, do notappear to undergo gene conversion, whereas those that possessCpG islands act as either donors (e'g. H-2Ebb, H-2Q2k,H-2Q10b),acceptors (e.g.H-2Ab; orboth (e.g.H2Kk, HLA-DQB1). The large number of pseudogenes within the MHC may represent a stockpile of genetic information for the generation of polymorphic diversity 'working'class I and classII molecules. in the

Nomenclolure Since much of the experimental work relating to the MHC is based on experiments in our little laboratory friend, the mouse, it may be helpful to explain the nomenclature used to describe the allelic genes and their products. If someone says to you in an obscure lan'we are having free elections', you fail to underguage stand, not becausethe idea is complicated but because you do not comprehend the language. It is much the same with the shorthand used to describe the H-2 system which looks unnecessarily frightening to the uninitiated. In order to identify and compare allelic genes within the H-2 complex in different strains, it is usual to start with certain pure hom*ozygous inbred strains, obtained by successive brother-sister matings, to provide the prototypes. The collection of genesin the H-2 complex is called the haplotype and the haplotype

l',

CHAPTER 4 _ M E M B R A N ER E C E P T O RFSO RA N T I G E N

of each prototypic inbred strain will be allotted a given superscript. For example, the DBA strain haplotype is designated H-2'l and the genes constituting the complex are therefore H-2K, H-2Aad, H-2Ab'1, H-2D,1and so on; their products will be H-2Kd, H-2Ad and H-2Dd and so forth (figure 4.21).When new strains are derived from these by genetic recombination during breeding, they are assigned new haplotypes, but the individual genes are designated by the haplotype of the prototype strain from which they were derived. Thus the A/J strain produced by genetic cross-over during interbreeding between (H-2kx H-2d)F7 mice (figure 4.22)is arbitrarily assigned the haplotype H-2n, but table 4.3 shows that individual genes in the complex are identified by the haplotype symbol of the original parents. Inheritonce of the MHc Pure strain mice derived by prolonged brother-sister mating are hom*ozygous for each pair of hom*ologous chromosomes. Thus, in the present context, the haplotype of the MHC derived from the mother will be identical to that from the father; animals of the C57BL strain,

Figure 4.21. How the definition of I{-2 haplotype works. Pure strain mice hom*ozygous for the whole H-2 region through prolonged brother-sister mating for at least 20 generations are each arbitrarily assigned a haplotype designated by a superscript Thus the particular set of alleles which happens to occur in the strain named C57BL is assigned the haplotype H-2b and the particular nucleotide sequence of

STRAIN

Figure 4.22. Inheritance and codominant expression of MHC genes.Eachhom*ozygous (pure) parental strain animal has two rdentical chromosomes bearing the H-2 haplotype, one paternal and the other maternal Thus in the present example we designate a strain which rs H-2k ask/k The first familial generation (F1) obtained by crossing the pure parental strains CBA (H-2k) and DBA / 2 (H-2\ has the H-2 genotypek/d Since 100% ofFl lymphocytes are killed in the presence of complement by antibodies to H-2k or to H-2d (raised by injecting H-2k lymphocytes into an H-2d animal and vice versa), the MHC molecules encoded byboth parental genes mustbe expressed on every lymphocyte. The same holds true for other tissues in the bodv.

for example, will each bear two chromosomes with the H-2b haplotype (cf . table 4.3). Let us seehow the MHC behaves when we cross two pure strains of haplotypes H-2k and H-2d, respectively. We find that the lymphocytes of the offspring (the F1 generation) all display bothH-2k andH-2d molecules on their surface,i.e.there is codominant expression (figure 4.22). If we go further and breed F1s together, the progeny have the genotypes k,k/d and d in the proportions to be expected if the haplotype segregates as a single Mendelian trait. This happens becausethe H-2 complex spans 0.5 centimorgans, equivalent to a recombination frequency between the K and D ends of 0.5"/', and the haplotype tends to be inherite d en bloc.Only the relatively infrequent recombinations caused by meiotic cross-overevents,as described for the A/ j strain above, reveal the complexity of the system. Thetissuedistributi0n of MHCmolecules Essentially, all nucleated cells carry classical class I molecules. These are abundantly expressed on both Iymphoid and myeloid cells, less so on liver, lung and

each allele in its MHC is labeled as geneb,e g H-214, eIc.It rs obviously more convenient to describe a given allele by the haplotype than to trot out its whole nucleotide sequence, and it is easier to follow the reactions of cells of known H-2 make-up by using the haplotype terminology-see, for example, the interpretation of the experiment in ftgure422

CBA

Fr HYBRTD

DBA/2

kxd I

H-2 GENOTYPE

LYMPHOCYTES (H-2 PHENoTYPE)

ANTI-H-2f, ANTI-H-2d

killing

killlng killing

killing

CHAPTER 4 - M E M B R A N ER E C E P T O RFSO RA N T I G E N kidney and only sparsely on brain and skeletal muscle. In the human, the surface of the placental extravillous cytotrophoblast lacks HLA-A and -B, although there is now some evidence that it may expressHLA-C. What is well established is that the extravillous cytotrophoblast and other placental tissues bear HLA-G, a molecule which generally lacks allodeterminants and which does not appear on most other body cells, except for medullary and subcapsular epithelium in the thymus, and on blood monocytes following activation with yinterferon. The role of HLA-G in the placenta is unclear, but it may function as a replacement for allodeterminant-bearing classical class I molecules and/or may play a role in shifting potentially harmful Th1 responses towards a Th2-type response.ClassII molecules,on the

Table 4.3. The haplotypes of the II-2 complex of some comrnonly used mouse strains and recombinants derived frorn them, A/J was derived by interbreeding (kxd) F1 mice, recombination occurring between E (classII) and S (c1assIII) regions*

b K 0 0 h4

Figure 4.23. Comparison of the crystal structures of CD1 and MHC class I. (a) Backbone ribbon diagram of mouse CD1d1 (red, a-helices; blue, p-strands). (b) Ribbon diagram of the mouse MHC class I molecule H-2Kb (cyan, o-helices; green, p-strands) (c) Superposition using alignment of Br-microglobulin

highlights some of the differences

other hand, are highly restricted in their exPression, being present only on B-cells, dendritic cells, macrophages and thymic epithelium. However, when activated by agents such as y-interferon, capillary endothelia and many epithelial cells in tissues other than the thymus express surface class II and increased levels of classI. molecules MHCondclosslchoin-reloled Thenonclossicol These molecules include the CD1 family which utilize Br-microglobulin and have a similar overall structure to the classicalclassI molecules (figure 4.23).They ate, however, encoded by a set of genes on a different chromosome to the MHC, namely on chromosome 1 in humans and chromosome 3 in the mouse. Like its true MHC counterparts, CD1 is involved in the presentation of antigens to T-cells, but the antigen-binding groove is to some extent covered over, contains mainly hydrophobic amino acids, and is accessibleonly through a narrow entrance. Instead of binding peptide antigens, the CD1 molecules generally present lipids or glycolipids. At least four different CD1 molecules are found expressed on human cells; CD1a, b and c are Present on cortical thymocytes, dendritic cells and a subset of B-cells, whilst CDld is expressed on intestinal epithelium, hepatocytes and all lymphoid and myeloid cells. Mice appear to only express two different CD1 molecules

between CD1d1 and H-2Kb. Note in particular the shifting of the crhelices. This produces a deeper and more voluminous Sroove in CD1d1, which is narrower at its entrance compared with H-2Kb' (Reprinted with permission from Porcelli S A et al (7998) Immunology Today79,362)

Ito

CHAPTER 4 _ M E M B R A N ER E C E P T O RFSO RA N T I G E N

which are both similar to the human CDld in structure and tissuedistribution and are referred to as CD1d1 and CD7d2 (or CD1.1and CD1.2). Cenes in the MHC itself which encode nonclassical MHC molecules include the H-2T, -Q and -M loci in mice, eachof whichencodes anumber of differentmolecules. The T22 and T10 molecules, for example, are induced by cellular activation and are recognized directly by y6 TCR without a requirement for antigen, possibly suggesting that they are involved in triggering immunoregulatory y6 T-cells. Other nonclassical class I molecules do bind peptides, such as H-2M3 which presents N-formylated peptides produced either in mitochondria or by bacteria. In the human, HLA-E binds a nine-amino acid peptide derived from the signal sequence of HLA-A, -B, -C and -G molecules, and is recognized by the CD94INKG2 receptors on NK cells and cytotoxic Tcells,as well as by the oB TCR on some cytotoxic T-cells. HLA-E is upregulated when other HLA alleles provide the appropriate leader peptides, thereby perhaps allowing NK cells to monitor the expression of polymorphic

TheB-cellsurtoc€ receptor torontlgen . The B-cell inserts its Ig gene product containing a transmembrane segment into its surface where it acts as a specific receptor for antigen. r The surfacelg is complexed with themembraneproteins Iga and Ig-B which become phosphorylated on cell activation and transduce signals received through the Ig antigen receptor. o The cytoplasmic tails of the Ig-cx,and Ig-B lmmunoreceptor fyrosine-based activation motifs (ITAMs) that, upon phosphorylation, can recruit phosphotyrosine-binding proteins that play important roles in signal transduction from theBCR.

TheT-cell surfqce recepfor loronligen

class I molecules using a single receptor. The murine hom*olog, Qa-1,has a similar function. The stress-inducible MICA and MICB (MHC class I chain-related molecules) have the same domain structure as classicalclassI and display a relatively high level of polymorphism. They are present on epithelial cells, mainly in the gastrointestinal tract and in the thymic cortex, and are recognized by the NKG2D activating molecule. One possible role for this interaction is in the promotion of NK and T-cell antitumor responses. The function of HLA-F is unclear. In contrast, although HLA-G shows extremely limited polymorphism, it is known tobind a range of self peptides with a defined binding motif and there is evidence for HLA-G restricted T-cells. HFE, previously referred to as HLA-H, possessesan extremely narrow groove which is unable to bind peptides, and it may serve no role in immune defense. However, it binds to the transferrin receptor and appears to be involved in iron uptake. A point mutation (C282Y) in HFE is found in 70-90"/, of patients with hereditary hemochromatosis.

bulins. The variable region coding sequence in the differentiating T-cell is formed by random translocation from clusters of V,D (for Band 6chains) and/segments togivea single recombinant V(D)/ sequence for each chain. . Like the Ig chains, each variable region has three hypervariable sequenceswhich functionin antigen recognition. o The CD3 complex, composed of 7 6, e and either (, (r1 or q, covalently linked dimers, forms an intimate part of the receptor and has a signal transducing role following ligand binding by the TCR. Thegenerolionol onlibodydiversilylor 0ntigenrecognlflon o Ig heavy and light chains and TCR crand p chains generally are represented in the germ-line by between 33 and 75 variable region genes, between 2 and27 D segment minigenes (Ig heavy and TCR p and 6 only) and 3-60 short/ segments. . TCR yand 6 chains are encoded by far fewer genes. . Random recombination of any single V, D and/from each

r The receptor for antigen is a transmembrane dimer, each chain consisting of two Ig-like domains. o The outer domains are variable in strucfure, the inner ones constant, rather like a membrane-bound Fab. . Both chains are required for antigen recognition. . CD4 and CD8 act as co-receptors, along with the TCR, for MHC molecules. CD4 acts as a co-receptor for MHC class II molecules and CD8 recognizes MHC class I molecules. . Most T-cells express a receptor (TCR) with o and chains B (TCR2). A separate lineage (TCR1) bearing y6 receptors is

gene cluster generates approximately 6.5 x 103 Ig heavy chain VD/sequences, 350 light chains, 4.5 x 103TCR s, 1 x 103 TCR B,butonly60 TCRyand 72 TCR 6. o Random interchain combination produces roughly 2.4x106lg, 4.5 x 106TCR cP and 4.3x 103TCR y6 receptors. . Further diversity is introduced at the junctions between

transcribed strongly in early thymic ontogeny but is associated mainly with epithelial tissues in the adult. . The encoding of the TCR is similar to that of immunoglo-

V, D and / segments by variable combination as they are spliced together by recombinase enzymes and by the Nregion insertion of random nontemplated nucleotide

C H A P T E 4R- M E M B R A N ER E C E P T O R FS O RA N T I G E N

sequences. These mechanisms may be particularly important in augmenting the number of specificities which canbe squeezed out of the relatively small y6 pool. . Useless or self-reactive receptors can be replaced by receptor editing. o In addition, after a primary response,B-cellsbut not T-cells undergo high rate somatic mutation affecting the V regions.

NKreceplors . NK cells bear a number of receptors with Ig-type domains and otherreceptors withC-type lectindomains. Members of both types of receptor family can function as inhibitory or activating receptors to determine whether the target cell should be killed. . Loss of MHC classI molecules can provoke attackbyNK cells. . NK cells can also recognize ligands that are upregulated by cells that suffer stressor DNA damage. MHC . MHC molecules act as receptors for antigen and present antigen-derived peptides to T-cells. o Each vertebrate specieshas an MHC identified originally through its ability to evoke very powerful transplantation rejection. . Each contains three classes of genes. Class I encodes 44 kDa transmembrane polypeptides associated at the cell surfacewith pr-microglobulin. ClassII molecules are transmembrane heterodimers Class III products are heteroge-

F U R I H ER E A D I N G BraudVM,AllanD S J & McMichaeJ A.J (1999)Functions of nonclassicalMHC and non-MHC-encoded class I molecules Currettt Opinion in lmrnunology71,100-108 Call M E & Wucherpfennig K W. (2005) The T cell receptor: critical role of the membrane envj.ronment in receptor assembly and function An n wtl Reztiezu of I m m unology 25, 707-725 Carding S R & Egan PJ. (2002) y5 T cells: functional plasticity and heterogeneity. Na tu r e ReaiezusImttu t tology 2, 336-345 Clark D A (1999) Human leukocyte antigen-G: new roles for old? Anrcricnn Jour nnl of Reproductiae I m m u nology 47, 117-720. Carcia K.C & Adams E J (2005) How the T ce1l receptor sees antigen-a structural view Cell 122,333-336 Gleimer M. & Parham P (2003)Stressmanagement: MHC classI and class II molecules as receptors of cellular stress lnmtunity 79, 469477 Crawunder U & Harfst E (2001) How to make ends meet in V(D)J recombination Ctrrcnt Opinion in lnm unology13, 786-794 Kelsoe G (1999)V(D)J hypermutation and receptor revision: coloring outside the lines Current Opinion in Immwtology 71,70-75 Kumanovics A ef a1 (2003)Genomic organization of the mammalian MHC. Annuol Reoieuof ltununology 21.,629-657 Kumar V & McNerney M.E (2005) A new self: MHC class Iindependent natural-killer cell self-tolerence Nsturc Reaiezos Innnunology5,363-374 Lanier L L (2005)NK cell recognition Annual Reaiezu oflntmunology 23.225 274

85 1

neous but include complement components linked to the formation of C3 convertases, heat-shock proteins and tumor necrosis factors. r Several different types of MHC class I and class II molecules are expressed by all cells. MHC genes also display remarkable polymorphism. A given MHC gene cluster is referred to as a'haplotype' and is usually inhe ited en blocas a single Mendelian trait, although its constituent genes have been revealed by crossover recombination events. . Classical class I molecules are present on virtually all cells in the body and presentpeptides to CD8+ cytotoxic T-cells. . Class II molecules are particularly associated with B-cells, dendritic cells and macrophages but can be induced on capillary endothelial cells and epithelial cells by y-interferon. Class II molecules present peptides to CD4* T-helpers for Bcells and macrophages. . The two domains distal to the cell membrane form a peptide binding cavity bounded by two parallel cr-helices sitting on a floor of p-sheet strands; the walls and floor of the cavity and the upper surface of the helices are the sites of maximum polymorphic amino acid substitutions. . Silent class I genes may increase polymorphism by gene conversion mechanisms. . Nonclassical MHC molecules and MHClike molecules have a number of functions, and include CD1 which presents lipid and glycolipid antigens to T-cells, and HLA-E which presents signal sequencepeptides from classical class I molecules to the CD94/NKG2 receptor of NK cells

Mak T W. (1998)T-cell receptor, oB In Delves PJ. & Roitt I.M. (eds) Encrlclopediaof Immunology,2nd edn, pp. 2264-2268. Academic Press, London (See also article by Hayday A & Pao W. on the y6 T CR; ibid., pp. 2268-2278 ) Matsuda F. ct al (1998) The complete nucleotide sequence of the heavy chain variable region locus. human immunoglobulin I ournal of Expcrimcn t al Med ic ine 188, 21,51,-21,62 MHC Sequencing Consortium (1999)Complete sequence and gene map of a human major histocompatibility complex Nature 407, 927-923. Moody D B , Zajonc D M. & Wilson I.A. (2005) Anatomy of CD1-1ipid antigen complexes. Nature Reaiews lmmunology 5, 387-399. Nemazee D (2000) Receptor editing in B cells Adaances in lmmunology 74,89-726. Parham P (ed ) (1999)Cenomic organisation of the MHC: structure, origin and function Immunological Reaiews1-67. Porcelli S A & Modlin R.L (1999) The CD1 system: antigenpresenting molecules for T cell recognition of lipids and glycolipids. An nual Reaiewof lmmunology T7,297-329. Prugnolle F et al. (2005) Pathogen-driven selection and worldwide HLA classI div ersrty.Current Biology 15, 7022-7027 Raulet D H (2004) Interplay ofnatural killer cells and their receptors with the adaptive immune response Nature lmmunology 5, 996-7002. de Wildt R.M.T. ef al (1999) Somatic insertions and deletions shape the human antibody repertote lournal of Molecular Biology 294,707-770.

Theprimoryinteroclion withontigen

INIRODUCTION In adaptive immunity, antigens are recognized by two classes of molecules: (i) antibodies, present either as soluble proteins or membrane-bound on the surface of B-cells; (ii) T-cell receptors present on the surface of Tcells. Antibodies recognize antigens on the surface of pathogens or as soluble foreign material such as toxins. T-cell receptors recognize peptides or glycolipids in the context of MHC molecules on the surface of host cells. Antibodies can thus be thought of as scanning for foreign material directly. T-cells are scanning for cells that are infected with pathogens.

W H A TA N I I B O D I ESSE E Antibodies recognize molecular shapes (epitopes) on antigens. Generally, the better the fit of the epitope (in terms of geometry and chemical character) to the antibody combining site, the more favorable the interactions that will be formed between the antibody and antigen and the higher the affinity of the antibody for antigen. The affinity of the antibody for the antigen is one of the most important factors in determining how effective the antibody the efficacy of in uioo. Epitopes come in ahuge variety of different shapes,as do antibody combining sites. Protein surfaces are typically recognized by a complementary surface in the antibody combining site, as illustrated in figure 5.1 which shows how an antibody recognizes an epitope on the human epidermal growth factor receptor, HER-2. The extent of complementarity of the interacting surfaces is readily appreciated. The area of antigen that contacts antibody is referred to as a footprint and is typically between about 400 and 1000 42. Footprints are of different sizes and irregular,

but a rough appreciation can be gained by projecting a 25Ax 25Asquare onto theprotein (figure 5.2). Antibodies recognize a topographic surface of a protein antigen. Most usually, key residues in the epitope will arise from widely different positions in the linear amino acid sequence of the protein (figure 5.3). This follows because of the manner in which proteins are folded; the linear sequence typically snakes from one side of the protein to the other a number of times. Such epitopes are described as discontinuous. Occasionally, key residues arise from a linear amino acid sequence. In such cases, the antibody may bind with relatively high affinity to a peptide incorporating the appropriate linear sequence from the antigen. Furthermore, the peptide may inhibit the antigen binding to the antibody. The epitope in such casesis described as continuous. An example of a continuous epitope would be a loop on the surface of the protein for which an antibody recognized successive residues in the loop. It should be noted, however, that an antibody that recognizes a continuous epitope does not bind a random or disordered structure. Rather it recognizes a defined structure that is found in the complete protein but can readily be adopted by the shorter peptide. The structure of an antibody that recognizes a linear epitope in complex with a peptide that contains the epitope is shown in figure 5.4; note that the structure of the peptide is largely helical in this example. Ihe 0ntibodycomplemenlority delerminingregions(CDRs) conlocllhe epilope The antibody combining site can vary greatly in shape and character depending upon the length and characteristics of the CDRs. Generally most or all of the CDRs

Figure 5.1. Complementarity of the antibody combining site and the epitope recognized on the antigen. The structure ofthe complex of the Fab of the antibody pertuzumab and its antigen HER2 is shown. HER2, the human epidermal growth factor receptor, is overexpressed on some breast cancer cells and pertuzumab is an antibody, similar to Herceptin,withpotential as atherapeutic againstbreastcancer. Below, the two molecules are shown separately with the interaction footprint shown on each. (Courtesv of Robvn Stanfield )

Figure 5.2. Antibodyfootprints (red) on a range of antigens. These footprints are determined from crystal structures of the antigens with antibody bound. The footprints are irregularbut canbe very roughly represented as a square of dimensions 25 Ax 25 Aas shown (Courtesy of Robyn Stanfield.)

Figure 5.3. Residues contributing to epitopes on the folded peptide chain of myoglobin. Amino acid residues 34,53 and 113 (black) contribute to the binding of a mAb and residues 83,144 and 145 to the bindingof another mAb (red). Bothepitopes are clearly discontinuous. By contrast, a third mAb binds to residues 18-22 (green). The mAb binds to isolated peptides containing the sequence corresponding to residues 1,8-22.The epitope is described as continuous. Much of the myoglobin structure is in a-helical conformation (Based on Benjamin D C et al. (1986)AnnualReaiew of lmmunology2,67.)

lrt

CHAPTER s - T H E P R I M A R YI N T E R A C I I O N W I T HA N T I G E N

Figure 5.4. The structure of an antibody bound to a peptide corresponding to a linear epitope. The antibody 4E10 neutralizes HlVbybinding to a linear epitope on the glycoprotein gp41 on the surface of the virus The antibodybinds to peptides containing the amino acid sequence NWFDIT and peptides containing this sequence can inhibit the binding of 4810 to gp41 The structure of the Fab fragment of4E10 bound to a peptide (gold) containing the NWFDIT sequence shows the peptide adopts a helical conformation. It is likely that the antibody recognizes its epitope in a helical conformation on the virus. (Courtesy of Rosa Cardoso.)

Figure 5.5. Conformational change in an antibody combining site. (a)An antiprogesterone antibody has a very hydrophobic pocket that is filled by a tryptophan residue (colored red) in the free antibody. (b) To bind progesterone (darkblue), the tryptophan residue swings out ofthe pocket and the antigen gains access (Courtesy of Robyn Stanfield.)

contribute to antigen binding but their relative contributions vary. The heavy chain CDRs, and particularly CDR H3, tend to contribute disproportionately to antigen binding. The CDR H3 in human antibodies can be quite long and has a finger-like appearance that could be used tobind into cavities on the antigen. The combining site of antibodies against smaller molecules such as carbohydrates and organic groups (haptens) are often more obviously grooves or pockets rather than the extended surfaces typically found in anti-protein antibodies. Structural changes and conformational rearrangements can occur in antibodies or antigens on interaction. In other words, on some occasions, the relationship between antibody and antigen will be like a'lock and key' but on other occasionsthe lock or key or both can be deformed to make a good fit. For the antibody, possible conformational changes include side chain rearrangements, segmental movements of CDRs or of the main-

chain backbone, and rotation of the Vr-V., domain upon antigen binding. Large changes in the conformation of the CDR H3 have been documented in crystal structures of Fab complexes. As shown in figure 5.5, an antibody to progesterone has a very hydrophobic combining pocket/ which is normally filled with a tryptophan from the CDR H3. Antigen binding involves this residue moving out of the pocket, the antigen molecule moving in and the trytophan stabilizing the antigen binding. As more and more structures have been solved it has become clear that antibody-antigen interactions come in all shapesand sizes with few general rules. It is important to bear in mind that high-affinity antibodies evolve in each individual following rounds of mutation and selection. There are multiple ways in which highaffinity recognition of an antigen can be achieved, and indeed no two antibody-antigen interactions are exactly the same.

CHAPTER s - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N

Antigens vsimmunogens An epitope on an antigen may bind very tightly to a given antibody but it may elicit such antibodies infrequently when the antigen is used to immunize an animal. In other words, there may be a perfectly good site on a pathogen for antibody binding but the antibody response to that site is so poor it cannot contribute to antibody protection against the pathogen. We say that

Conjugoie

8el

the site has low immunogenicity and the consequences can clearly be great. An extreme example of the distinction between the ability to be recognizedby an antibody (which we will term antigenicity) and the ability to elicit antibodies when used to immunize an animal (which we will term immunogenicity) is provided by experiments using small molecules known as haptens such as rflaminobenzene sulfonate. Immunization with free haptenproduces no antibodies to thehapten (figure 5.6). However immunization with hapten groups linked to a protein carrier generates antibodies that react with high affinity to hapten alone or linked to a molecule other than the carrier. It is logical to refer to the hapten as the antigen and the hapten-protein complex as the 'antigen' immunogen, although strictly the word is d eriv ed fr orn' an t ib o dy g en er atrng' sub stance.

IDENTIFYIN BG . C E t tE P I T O P EOSNA P R O T E I N

Figure 5.5. Antigenicity and immunogenicity. Afree small molecule hapten will not induce antibodies if injected in to an animal However, high affinity antibodies specific for the free hapten can be obtained by iniecting the hapten conjugated to a protein carrier molecule such as ovalbumin

Figure 5.7. Three epitopes on the small protein lysozyme. The crystal structures of lysozyme bound to three antibodies (HyHEL-5, HyHEL-10and D1.3) havebeen determined In the figure, the Fv fragment of eachantibody is shown separatedfrom lysozyme to reveal the footprint of interaction in each case The three epitopes are nearly nonoverlapping with only a small overlap between HyHEL-10 and D1.3. (After Davies ef al (1990)Annual Reaiezoof Biochemistrv59,439.)

How many epitopes are there on a single protein? This depends upon how one definesan epitope. For the small protein lysozyme (molecular weight -14 300 daltons), the structures of three noncompeting monoclonal antibodies in complex with the protein antigen have been determined. They have minimally overlapping footprints that cover just under half of the surface of the protein (figure 5.7). One could extrapolate that a small protein such as this could have of the order of between three and six nonoverlapping epitopes recognized by noncompeting antibodies. The specificity of a given antibody could then be defined by its ability to compete with the three to six'prototype' antibodies. In practice,

CHAPTER s - T H E P R I M A R YI N T E R A C T I OW NIIHANTIGEN this is often done; an antibody is said to be directed against a given epitope if it competes with a prototype antibody of known specificity. This is, of course, a rather simplistic view since many antibodies will compete with more than one prototype antibody allowing a more sophisticated B-cell epitope map to be constructed. An even more sophisticated map can be constructed by scanning mutagenesis of the antigen. In the latter case, single positions in the antigen can be substituted by differing amino acids (usually alanine-hence the term 'alanine scanning mutagenesis') and the effects on antibodybindingmeasured (seefigure 5.10).At this greater level of precision, it is likely that no two antibodies will give exactly the same footprint, and therefore no two antibodies recognize exactly the same epitope. What determines the strength of the antibody response to a given epitope on a protein? There appear to be a number of factors involved. Perhaps the most important is the accessibility of the epitope on the protein surface. Loops that protrude from the surface of the folded protein tend to elicit particularly good antibody responses. The sites on the hemagglutinin (HA) protein from the surface of influenza virus that elicit important antibody responses are indicated in figure 5.8. Antibodies to these regions, perhaps elicited by vaccination, neutralize the virus and protect against infection. Mutations in these regions allow the virus to 'escape' from neutralizing antibodies and infect human hosts who were protected against the original form of the virus. Influenza epidemics thus directly reflect antibody targeting to certain preferred epitopes. Note that the epitopes are all located towards the part of HAthat is distant from the membrane of the influenza virus and more accessible to antibody. Note also that three of the epitopes involve prominent loops on the HA structure. The combination of the ability of loops to accommodate changes in amino acid sequencemore readily than more compact structural features and their favored status in eliciting antibody responses is responsible for the frequent associationof loops withneutralization escape. HIV is anothervirus that makes use of the tendency of the antibody system to respond to highly exposed variable loops on the viral surface protein. Following primary infection, it takes some time (weeks) for neutralizing antibodies to reach a level where they begin to inhibit virus replication. These antibodies are typically elicited to exposed loops on the virus. While these antibodies are being elicited, the virus has diversified, i.e. it has become a swarm of related viruses through the errors associated with RNA transcription of this RNA retrovirus. Amongst this swarm is a virus that has sequence changes in the epitopes targeted by the neutralizing antibody response that allow it to escapefrom

Figure 5.8. Epitopes on the surface of influenza virus hemagglutinin (HA). The structure of HAfrom the 1918flu virus responsible for a pandemic that killed up to 50m people is shown The molecule is a hom*otrimer with monomers shown in cyan, gray and brown. The hom*otrimers form'spikes' on the surface of the virus. The spikes are anchored to the virus membrane (bottom of the figure). Sialic acid residues on the targetcellsbind to sites at the top ofthe spikes resulting in attachment of virus to target cell in the first step of infection Five principal epitopes on HA are represented as colored ovals located towards the tip of the molecule Three of the epitopes (yellow, green and purple) involve prominent loop structures (Courtesy of fames Stevens.)

the response. This new virus becomes predominant. Eventually a response is mounted to this virus and a second new virus emerges and so on. The antibody response chases the virus over many years but never gains control.

CHAPTER s - T H E P R I M A R YI N T E R A C I I O N W I T HA N T I G E N One point worthy of note is that accessibleloops on protein structures tend to be flexible. Therefore epitope dominance has also been associated with flexible regions of a protein antigen.

I H E R M O D Y N A M IOCFS A N I I B O D Y - A N T INGI E NIERACTIONS The interaction of antibody and antigen is reversible and canbe describedbv the laws of thermodvnamics. In particular, the reaction Ab +Ag+df-Agcomplex canbe studied and theposition of the equilibrium established under varying conditions. In other words, the amount of antibody bound to antigen under different conditions can be estimated. This is crucial information. If antibody coats a virus then it is likely that the virus will be prevented from entering target cells and infection will be avoided. If antibody can become attached to a bacterial cell in a high enough density then complement may be triggered and the cell killed. The position of equilibrium is describedby the association or binding constant, Ku: K" = [Ab-Agcomplex]/[Ab]x

[Ag]

where square brackets indicate molar concentrations. The units of Kuare thus moles per liter (v-1),or 1./ w.If K^ is a large number then the equilibrium is far to the right and Ab-Ag complex formation is favored. Typically high-affinity antibodies have Ku values of the order of 168-1gtor-t. Someresearchersprefer to think of binding in terms of a dissociationconstant,K6,simply defined as 1/ Kuand having the units of v. High affinity antibodies then have Ko values of the order of 10 8-10-10rr,r. Since a Ka = 10 e v corresponds to 1nv, high affinity antibodies are sometimes referred to as 'nM binders'. Moderate affinity antibodies such as IgMs are often referred to as pvbinders(Ka=1pr'a). Another way to look at the binding equation is that if half the available antigen sites are occupied by antibody then [Ag] = [Ab-Ag complex] and Ka = 1/[Ab] or Kd = [Ab]. In other words, Ko is equal to the antibody concentration at which half of the antibody is bound. Thus for example a nv binding antibody will begin to complex antigen when its concentration is in the nanomolar range. The antibody will bind very little if it is only in the picomolar (10-12)range of concentrations but will bind very effectively in the trrvrange. Similarly a ;,tv antibody will be effective in the plr range of concentration but not the nv. For IgG, nv is roughly 0.15 trtg/ ml and pv is 150 pglml. The average concentration of IgG in serum is about 10 mglml. Clearly then, if we require

erl

that antibodies be present in serum at concentrations where they are going to be effective in binding antigen, many many more specificities can be covered by a set of nv binding (high affinity) antibodies than a set of prra binding antibodies. Indeed, this seemsto be largely how Nature operates outside of extreme immunization protocols in animal models. Thus we are mostly protected, at least against re-infection following a primary infection or vaccination, by high affinity antibodies at relatively moderate concentrations. In the above discussion, we implicitly assume that antibody-antigen interactions are monovalent, involving just one Fab arm of the antibody molecule. In fact they may well be multivalent which complicates the issues somewhat, but the major points remain intact. We return to multivalency below. Binding constants for antibody-antigen interactions are often estimated from ELISA measurements but now can be determined with some precision by techniques such as surface plasmon resonance and isothermal calorimetry (see p. 369). For binding of antibodies to antigens on the cell surface, flow cytometry can give a good estimate of binding affinities. The binding constant for a reaction is directly related to the energy accompanying the reaction by the equation: AG=-RTlnKu where AG is called the free energy of the reaction, R is the gas constant and T is the temperature in oK. ln is natural log = 2.363x logrs.AG is then another way of describing how far a reaction will be driven to the left or right at equilibrium under certain conditions. If Ku = 10ervt-1, LG - -12 kcal/mole; if K" = 106n-1,AG - -8 kcal/mole. The advantageof considering the AG is thatit canhelp in beginning to understand the molecular forces that lead to antibody-antigen interaction. Thus the free energy of a reaction (AG) is the net effect of contributions from enthalpy (A.FI)and entropy (AS): AG=AH-TAS The enthalpy is the heat of the reaction; the more heat is given outby the reaction (negative AH) the more it will befavored (negativeAG).If heathas tobe supplied tothe reaction it is disfavored. The more entropy (or disorder) results from the reaction (positive AS), the more it is favored. For example, an antibody-antigen interaction would be favored by the formation of a H bond between the two molecules to the tune of approximately 13 kcal/mole. A salt bridge would provide a similar or slightly greater amount of energy. The reaction would also be favored by hydrophobic surfaces on the antibody and the antigen coming together because then

N I T HA N T I G E N CHAPTER s - T H E P R I M A R YI N T E R A C T I OW

water that was ordered around the hydrophobic faces would be released to increase entropy. It is estimated that burying 100 A2 of hydrophobic surface generates about 2.5 kcal/mole of binding energy. Some of the forces driving protein-protein interactions are summarized in figure 5.9. An epitope is often thought of in terms of the region of the antigen contacted by antibody, a picture provided from crystal structure studies of antibody-antigen complexes. However, it should be borne in mind that looking at contacts between antibody and antigen in a crystal structure does not tell us the contributions of individual interactions to the overall binding energy. This can be done by measuring the effects of scanning mutagenesis (see above) on antibody binding measured. The available data then suggest that only a few productive interactions ('hot spots') dominate the energetics of binding; many interactions are neutral or detrimental to binding even in a high affinity antibody-antigen pairing. In the interaction of an antibody with lysozyme, only about a third of the antibody contact residues actually contribute significantly to net binding (figure 5.10). A substitution in only one residue of antigen or antibody can be decisive in net binding of antibody to antigen. This can be readily appreciated intuitively. If a bulky residue replacesa small one in the epitope recognized, then the whole antibody-antigen interface may be disrupted. Pathogens typically evade antibodies by mutations in a small number of critical residues. Mullivolency in 0ntib0dy-0ntigen inleroclions The binding of a monovalent Fab fragment to a monovalent antigen canbe analyzed in a straightforward wayas described above. This should also be true for the corresponding divalent IgG molecule interacting with the monovalentantigen. However, oncewe consider a divalent IgG (or multivalent antibody of any class) interacting with a multivalent antigen, the analysis of binding becomesmore complex. Consider IgG binding to an antigen that is expressed as multiple copies on a cell surface. If the antigen molecules are appropriately spaced and in an appropriate orientation, IgG may be able to bind divalently (figure 5.11).This will lead to a higher affinity (often referred to as the avidity or functional affinity) of the IgG for the cell surface than the corresponding Fab. The 'bonus effect' of divalent binding can be understood intuitively in terms of the tendency of the divalent IgG to stick better to the cell surface than the corresponding Fab. For the Fab to 'fall off ' the cell, a series of interactions between a single antibody combining site and the antigen mustbe

broken. For the IgG to fall off , the interactions in two antibody combining sites must be broken simultaneously; a lower probability event. The bonus effect can be thought of in terms of AG. Divalent binding will produce a greatly increasedAHbecauseof the use of two antibody combining sites. However, an entropy price willbe paid in constraining the Fab arms of the IgG molecule. The net effect in AG is usually quite modest,producing an enhanced affinity of the order of 1-100-fold as the bonus effect. It should also be borne in mind that IgG maybind monovalently even to a multivalent antigen if the antigen molecules are inappropriately spaced or oriented. IgM is decavalent for antigen, which in theory could produce a huge bonus effect in functional affinity. In practice IgMs tend to be rather moderate affinity binders, suggesting limited use of multivalency and/or a high entropy price paid for multivalentbinding. One of the most dramatic effects of multivalent antibody interaction can be seen in the neutralization of toxins. Botulinum neurotoxins cause the paralytic human disease botulism and are considered a major potential bioterrorist threat. Monoclonal antibodies have been generated from phage libraries against the toxin. No single mAb protected mice against lethal challenge with toxin. However, a combination of three mAbs protected mice against a huge challenge with toxin. The difference could be attributed in part to a multivalency bonus effect (cooperative binding of the antibodies with more than one molecule of the toxin) that increased the functional affinities of the antibodies in to the pM range from the nM range in the individual mAbs. The origins of this effect are illustrated for a two-mAb combination in figure 5.12.

N DC R O S S . R E A C T I V I I Y SPECIFICIA TY O FA N I I B O D I E S Specificity is a commonly discussed concept in the context of antibodies. It can have different meanings. Sometimes, it is used simply to indicate that the antibody has high affinity for antigen. Generally this means that the antibody has a combining site that fits very well to an epitope on the antigen and is much less likely to fit other shapes very well. Therefore it is specific for the antigen. However, there maybe other shapesthat canbe accommodated, especially if they are related to the antigenic epitope in composition or character. Most likely is that other molecules will be recognized with lower affinity. It is important to remember from the discussion above that antibodies will be functional at concentrations around their Ko values. So if an antibody has a ntvl affinity for a given antigen and is present at nnaconcentrations in aizto, cross-reactivities with other antigens

N I T HA N T I G E N CHAPTER s - T H E P R I M A R YI N T E R A C T I OW

PEPTIDE

/ sncxaorur co

\+-,/

C H- C H " C H " C H " C H . N H .O C O - C H , C H , - C H

,/'\ \/ 0

cLUTAI/ATE

LYSTNE

COHCO

,/l\

-CHr-,CH

CH-CH,-0-H-0

\,/

IYROSINE

SERINE NH

CH" ; I CH,

6s--{

v,,3

c0 LEUCINE

ISOLEUCINE

Effecliveluter lrmrl0l IT

Figure 5.9. Protein-protein interactions. (a) Coulombic attraction between oppositely charged ionic groups on the two protein sidechains as illustrated by an ionized amino group (NHr*) on a lysine of one protein and an ionized carboxyl group (- COO-) of glutamate on the other. The force of attraction is inversely proportional to the square of the distance between the charges. Thus, as the charges come closer together, the attractive force increasesconsiderably: ifwehalve the distance apart, we quadruple the attraction Furthermore, since the dielectric constant of water is extremely high, the exclusion of water molecules through the proximity of the interacting residues would greatly increase the force of attraction Dipoles on antigen and antibody can also attract each other In addition, electrostatic forces may be generated by charge transfer reactions between antibody and antigen; for example, an electron-donating protein residue such as tryptophan could part with an electron to a group such as dinitrophenyl (DNP) which is electron accepting, thereby creating an effective +1 charge on the antibody and -1 on the antigen (b) Hydrogen bonding between two proteins involving the formation of reversible hydrogen bridges between hydrophilic groups, such as OH, NHt and COOH, depends very much upon the close approach of the two molecules carrying these groups. Although H bonds are relatively weak, because they are essentially electrostatic innafure, exclusion of waterbetween the reacting side-chains would greatly enhance thebinding energy through the gross reduction in dielectric constant. (c) Nonpolar, hydrophobic groups such as the side-chains of valine, leucine and isoleucine tend to

associate in an aqueous environment. The driving force for this hvdrophobic interaction derives from the fact that water in contact with hydrophobic molecules with which it cannot H bond will associate with other water molecules, but the number of configurations which allow H bonds to form will not be as great as that occurring when they are surrounded completely by other water molecules, i.e. the entropy is lower. The greater the area of contactbetween water and hydrophobic surfaces, the lower the entropy and the higher the energy state Thus, if hydrophobic grouPs on two Proteins come together so as to exclude water molecules, between them the net surface in contact with water is reduced and the proteins take up a lower energy state than when they are separated (in other words, there is a force of attraction between them). (d) Van der Waals force: the interaction between the electrons in the extemal orbitals of two different macromolecules may be envisaged (for simplicity!) as the attraction between induced oscillating dipoles in the two electron clouds. The nature of this interaction is difficult to describe irr nonmathematical terms, but ithasbeen likened to a temporary perturbation of electrons in one molecule effectively forming a dipole which induces a dipolar perturbation in the other molecule, the two dipoles then having a force of attraction between them; as the displaced electrons swingback through the equilibrium position and beyond, the dipoles oscillate. The force of attraction is inversely proPortional to the seventhpower ofthe distance and, as a result, this rises very rapidly as the interacting molecules come

in the sub-pM range are unlikely to be functionally significant. 'specificity' is A second meaning that is attached to the ability to discriminate between molecules. This clearly overlaps with the discussion above but could

also be applied to lower affinity antibodies. Thus in genomics there has been a demand for antibodies that can distinguish target proteins from many other proteins and identify the target Proteins in a variety of assays. This has not necessarily required high affinity

closer together.

F V D I3

^^G(kcol/mol): IIIE Figure 5.10. Energetic map of an antibody-antigen interface. The antibody Dl 3 (sFv shown here) binds with high affinity to hen eggwhite lysozyme (HEL) and the crystal structure of the complex has been solved (figure 5 7) The energetic contribution of contact residues for both antibody and antigen can be estimated by substituting the residue with the relatively 'neutral' residue aianine The effect can be expressed in terms of the loss of free energy of binding for the interaction on alanine substitution (MG). A large positive value for AAG

24 A shows that the alanine substitution has had a strong detrimental effect on binding and implies that the residue substituted forms a crucial contact in the interface between antibody and antigen Clearly, most contact residues, particularly on the antibody, contribute little to the overallbindingenergy. Thereareclear'hotspots'onbothantibodyand antigen and the hot spot residues on the antibody side of the interaction correspond to those on the antigen side. (After Sundberg and Mariuzza(2002) AdaancesinProtein Chemistry 61,119 )

Moderole

Moderote

Very strong

ANTIGEN ANTIGEN

Cellsurfoce

Singleonlibody '$tr bridge

seporolion

r,roseporolion

ANTIGEN

Figure 5.11. Divalent antibody binding to a cell surface. The affinity of an antibody that can bind divalently to a multivalent antigen, such as may be found on a cell surface, is enhanced relative to an antibodv that can onlybind monovalently.

Figure 5.12. The bonus effect of multivalent binding in antibody neutralization of a soluble molecule such as a toxin. (a-c) It is found that two antibodies binding to nonoverlapping epitopes on a soluble antigen may show considerably enhanced affinity as compared to the antibodies used separately (cooperative binding). (d,e) If only one antibody is present, dissociation of the complex requires only that the interaction between one antibody combining site and its epitope be disrupted. (f) Iftwo antibodies arebound to antigen and one antibodycombining site-epitope interaction is lost as shown, the complex stays together and the released Fab arm is in a position to re-complex. In effect, the koo values for the two antibodies are decreased in the complex and, if a Fab arm is dislodged, the ko. value is increased relative to a single bound antibody situation

N I T HA N T I G E N CHAPTER 5 - T H E P R I M A R YI N T E R A C T I OW buthas required good discrimination. Moderate affinity antibodies selected from phage libraries have been used successfully in this arena.

W H A IT H EI . C E T TS E E S We have on several occasions alluded to the fact that the T-cell receptor seesantigen on the surface of cells associated with an MHC class I or II molecule. Now is the time for us to go into the nuts and bolts of this relationship. Hoplotyperestriclionreveolsthe needfor MHCporticipotion It has been established in 'tablets of stone' that T-cells bearing crBreceptors,with some exceptions (cf. p. 105), only respond when the antigen-presenting cells express the same MHC haplotype as the host from which the T-cells were derived (Milestone 5.1). This haplotype restriction on T-cell recognition tells us unequivocally that MHC molecules are intimately and necessarily involved in the interaction of the antigen-bearing cell with its corresponding antigen-specific T-lymphocyte. We also learn that, generally, cytotoxic T-cells recognize antigen in the context of class I MHC, and helper T-cells interact when the antigen is associated with class II molecules. Accepting then the participation of MHC in T-cell recognition, what of the antigen? For some time it was perplexing that, in so many systems, antibodies raised to the native antigen failed to block cytotoxicity (cf. figure M5.1.1b), despite consistent success with anti-MHC classl sera.Wenow knowwhv. fromlhe onligen T-cellsrecognize o lineorpeplidesequence In Milestone 5.1,we commented on experiments involving influenza nucleoprotein-specific T-cells which could kill cells infected with influenza virus. Killing occurs after the cytotoxic T-cell adheres strongly to its target through recognition of specific surface molecules. It is curious then that the nucleoprotein, which lacks a signal sequenceor transmembrane region and so cannot be expressed on the cell surface, can nonetheless function as a target for cytotoxic T-cells, particularly since we have already noted that antibodies to native nucleoprotein have no influence on the killing reaction (figure M5.1.1b).Furthermore, uninfected cells do not become targets for the cytotoxic T-cells when whole nucleoprotein is added to the culture system. However, if, instead, we add a series of short peptides with sequences derived from the primary structure of the nucleoprotein, the uninfected cells now become susceptible to cytolytic T-cell attack (figure 5.13).

esl

Thus was the secret revealed! The startling reality is that T-cells recognize linear peptides derived from the antigen, and that is why antibodies raised against nucleoprotein in its native three-dimensional conformation do not inhibit killing. Note that only certain nucleoprotein peptides were recognizedby the polyclonal T-cells in the donor population and these are to be regarded as T-cell epitopes. \Alhen clones of identical specificity are derived from these T-cells, each clone reacts with only one of the peptides; in other words, like B-cell clones, each clone is specific for one corresPondingepitope. Entirely analogous results are obtained when T: helper clones are stimulatedby antigen-presenting cells to which certain peptides derived from the original antigen have been added. Again, by synthesizing a seriesof suchpeptides, the T-cell epitope canbe mapped with some precision. The conclusion is that the T-cell recognizes both MHC and peptide and we now know that the peptide which acts as a T-cell epitope lies along the groove formed by the o-helices and the p-sheet floor of the class I and class II outermost domains (seefigure 4.13)' ]ust how does it get there?

AA NR IIGEN OG FI N T R A C E T T U T PROCESSIN I MHC CTASS F O RP R E S E N T A T IBOYN theproWithin the cytosollurk proteolyticstructures, teasomes,involved in the routine turnover and cellular degradation of proteins (figure 5.14).Cytosolic proteins destined for antigen presentation, including viral proteins, are degraded to peptides via a pathway involving these structures, and further trimmed by cytosolic proteases including leucine- and aspartyl-aminopeptidases and finally by the endoplasmic reticulum (ER) resident aminopeptidase-1(ERAP-1). In addition to proteins that are already present in the cytosol, a proportion of membrane-bound and secretory proteins are transported from the ER back into the cytosol. This process may be mediated by the Sec61multi-molecular channel, or by the cytosolic p9TATPase (VCP; zralosin-containing protein) interacting with a membrane protein complex containing Derlin-1 and VIMP ( VCP-lnteracting membrane protein). Proteins that have undergone retrotranslocation from the ER into the cytosol can then be processed for class I presentation, as can proteins derived from mitochondria. Prior to processing, proteins are covalently linked to several ubiquitin molecules in an ATP-dependent process. The polyubiquitination targets the polypeptides to the proteasome (figure5.14). Only about 10% of the peptides produced by protea-

Iru

C H A P I E Rs - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N

LCM infeclion

K i l lz

E Figure M5.1.1. T-cell killing is restrictedby the MHC haplotype of the virus-infect€dtargetcells.(a)Haplotype-restrictedkilling of lymphocyticchoriomeningitis(LCM) virus-infectedtargetcellsby cytotoxic T-cells Killer cells from H-2d hosts only killed H-2dinfected targets,not those of H-2k haplotype and vice versa. (b) The realization that the MHC, which had figured for so long as a dominant controlling element in tissue graft rejection, should come to occupy the center stage in T-cell reactions has been a source of fascination and great pleasure to immunologists-almost as though a great universal plan had been slowly unfolding.

Killing of influenza-infected target cells by influenza nucleoprotein (NP)-specific T-cells from an HLA-A2 donor (cf p. 371 for human MHC nomenclature). Killing was restricted to HLA-A2 targets and onlyinhibitedby antibodies to.A2, nottoAl, nor to the class II HLADR framework or native NP antiqen.

ovolbumin-soecific T-cellclonedeilved trom

One of the seminal observations which helped to elevate the MHC to this lordly position was the dramatic Nobel prizewinning revelation by Doherty and Zinkernagel that cytotoxic T-cells taken from an individual recovering from a viral infection will only kill virally infected cells which share an MHC haplotype with the host. They found that cytotoxic Tcells from mice of the H-2d haplotype infected with lymphocytic choriomeningitis virus could kill virally infected cells derived from any H-2d strainbutnot cells of H-2k or other H-2 haplotypes. The reciprocal experiment with H-2k mice shows that this is not just a special property associated with H-2d (figure M5.1.1a). Studies with recombinant strains (cf. table 4.3) pin-pointed class I MHC as the restricting element and this was confirmed by showing that antibodies to classI MHC block the killing reaction. The samephenomenon has been repeatedly observed in the human. HLA-A2 individuals recovering from influenza have cytotoxic T-cells which kill HLA-A2 target cells infected with influenza virus, but not cells of a different HLA-A tissue-type specificity (figure M5.1.1b). Note how cytotoxicity could be inhibited by antiserum specific for the donor HLA-A type, but not by antisera to the allelic form HLA-AI or the HLA-DR class II framework. Of striking significance is the inability of antibodies to the nucleoprotein to block T-cell recognition even though the T-cell specificity in these studies was known to be directed towards this antigen. Since the antibodies react with nucleoprotein in its native form, the conformation of the antigen as presented to the T-cell must be quite different.

all

(ontigen) Anligen-presenling cellspulsed wilhovolbumin Figure M5.1.2. The Trcell clone only respondsby proliferation in oitro when the antigen-presentingcells (e.g. macrophages) pulsed with ovalbumin expressthe sameclassII MHC.

In parallel, an entirely comparable series of experiments has established the role of MHC class II molecules in antigen presentation to helper T-cells. Initially, it was shown by Shevach and Rosenthal that lymphocyte proliferation to antigen in aitro couldbe blocked by antisera raised between two strains of guinea-pig which would have included antibodies to the MHC of the responding lymphocytes. More stringent evidence comes from the type of experiment in which a T-cell clone proliferating in response to ovalbumin on antigen-presenting cells with the H-2Ab phenotype fails to respond if antigen is presented in the context of H-2Ak. However, if the H-2Ak antigen-presenting cells are transfected with the genes encoding H-2Ab, they now communicate effectively with the T-cells (figure M5.1.2).

CHAPTER s - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N

o/olysis Cytotoxicity: 05r015

< Influenzo infected cells < Conlrol cells,no oddedpeplide

325 1

?EN

--l

375

@ c o

l l

400 o

425

s7 l

butnot inthe mouse, calnexinis then replacedby calreticulin. The ER-resident protein, Erp57, which has thiol reductase, cysteine protease and chaperone functions, becomes associated with the complex of calreticulin-calnexin and classI heavy chain which now folds together with Br-microglobulin. The empty class I molecule bound to these chaperones becomes linked to TAPT/2 by tapasin. Upon peptide loading, the class I molecule can dissociate from the various accessory molecules, and the now stable peptide-class I heavy chain-pr-microglobulin complex traverses the Golgi stack and reaches the surface where it is a sitting target for the cytotoxic T-cell.

450

415 498

Figure 5.13. Cytotoxic T-cells, from a human donor, kill uninfected target cells in the presence of short influenza nucleoprotein peptides. The peptides indicated were added to 5lCrlabeled syngeneic (i e same as T-cell donor) mitogen-activated lymphoblasts and cytotoxicity was assessedby slCr releasewith a killer to target ratio of50 : 1. The three peptides indicated in red induced good killing Blasts infected with influenza virus of a different strain served as a posrtive control (Reproduced from Townsend A.R M ef ol (7986) Cell 44, 959-968,with permission Copyright 01986 by Cell Press)

somesare the optimal length (octamersor nonamers) to fitinto the MHC classI groove; about 70%are likely tobe too small to function in antigen presentation, and the remaining 20% would require further trimming by, for example/ cytosolic aminopeptidases.The cytokine IFNy increasesthe production of three specialized catalytic proteosomal subunits, the polymorphic LMP2 (Bri) and LMPT (Bri) and the nonpolymorphic MECL1 (Bri).These molecules replace the hom*ologous catalytic subunits (Br, Buand 02,respectively) in the housekeeping proteasome to produce what has been termed the immunoproteasome/ a process which modifies the cleavage specificity in order to tailor peptide production for class Ibinding. Both proteasome- and immunoproteasomegenerated peptides are translocated into the ER by the /ransporters associatedwith antigen processing (TAP1 and TAP2) (figure 5.15), a process which might also involve heat-shock protein family members. The newly synthesized class I heavy chain is retained in the ER by the molecular chaperone calnexin, which is thought to assist in folding, disulfide bond formation and promotion of assembly with Br-microglobulin. In the human,

F O RC T A S SI I OG FA N I I G E N PROCESSIN M H CP R E S E N T A T IFOONt I O W S PT ATHWAY A DIFFEREN Class II MHC complexes with antigenic peptide are generated by a fundamentally different intracellular mechanism, since the antigen-presenting cells which interact withT-helpercells need to sample the antigen fromboth the extracellular and intraceTlular compartments. In essence,a trans-Golgi vesicle containing class II has to intersect with a late endosome containing exogenous protein antigen taken into the cell by an endocytic mechanism. Regarding the classII molecules themselves, these are assembled from cr and B chains in the ER in association with the transmembrane invariant chain (Ii) (figure 5.16) which trimerizes to recruit three MHC class II molecules into a nonameric complex. Ii has several functions. First, it acts as a dedicated chaperone to ensure correct folding of the nascent class II molecule. Second, an internal sequence of the luminal portion of Ii sits in the MHC groove to inhibit the precocious binding of peptides in the ERbefore the classII molecule reachesthe endocytic compartment containing antigen. Additionally, combination of Ii with the cp class II heterodimer inactivates a retention signal and allows transport to the Golgi. Finally, targeting motifs in the N-terminal cytoplasmic region of Ii ensure delivery of the class Il-containing vesicle to the endocytic pathway. Meanwhile, exogenous protein is taken up by endocytosis and, as the early endosome undergoes progressive acidification, is processed into peptides by endopeptidases, exopeptidases and GILT (interferon-yinduced lysosomal thiol reductase).Particularly implicated are the cysteine proteases cathepsin S, L, B, F, H and (in humans) V, and asparagine endopeptidase (AEP). Most of these enzymes have broad specificity. The late endosomes characteristically acquire

lrt

CHAPTER 5 - T H E P R I M A R YI N T E R A C T I OW NIIHANTIGEN

a1-aJ

pl -p7 pl -p7 s1-a7

Figure 5.14. Cleavage of cytosolic proteins by the proteasome. Cytosolic proteins become polyubiquitinated in an ATP-dependent reaction in which the enzyme E1 forms a thiolester with the C-terminus of ubiquitin and then transfers the ubiquitin to one of a number of E2 ubiquitin-carrierproteins. The C-terminus of the ubiquitin is then conjugated by one of several E3 ubiquitin-protein ligase enzymes to a lysine residue on the polypeptide. There is specificity in these processes in that the individual E2 and E3 enzymes have preferences for different proteins. The ubiquitinated cytosolic protein binds to the ATPase-containing 19S regulator where ATP drives the unfolded protein chain into the cylindrical structure of the 20S core proteasome which is made up of28 subunits arranged in four stacked rings The two outer rings comprise seven different cr-subunits (or-ar) whilst the two rings of the central hydrolytic chamber are each made up of seven different B-subunits (Fr-Fz) Within the hydrolytic chamber the protein

is exposed to proteolytic activity (red shading) Anovelcatalytic mechanism is involved in which the nucleophilic residue that attacks the peptide bonds is the hydroxyl Sroup on the N-terminal threonine residue of the p-subunits. Three distinct peptidase activities havebeen 'ch)'rnotrypsin{ike' associated with specific p-subunits One is in that it hydrolyses peptides after large hydrophobic residues, one is 'trypsin-like' and cleaves afterbasic residues, and onehydrolyses after acidic residues. The low molecular weight proteins LMP2 (81t, LMPT (pui), and MECL1 (multicatalytic endopeptidase complex like 1, (Bri)) immunoproteasome-associated molecules show similar specificities but have enhanced chymotrypsin and trypsin activity and reduced postacidic cleavage compared to thefu counterparts in the housekeeping proteasome (Based on Peters J -M. et nl (7993)| ournal of Molecular Biology 234, 932-937 and Rubin D M. & Finley D (7995) Cur rent Biology 5,854t-858.)

lysosomal-associated mernbrane proteins (LAMPs), which are implicated in enzyme targeting, autophagy (the enveloping of cellular organelles into an autophagosome for subsequent degradation) and lysosomal biogenesis. These late endosomes fuse with the vacuole containing the class II-Ii complex. Under the acidic conditions within these MHC class l/-enriched compartments (MIICs), AEP and cathepsins S and L degrade Ii except for the part sitting in the MHC groove which, for the time being, remains there as a peptide referred to as CLIP (class Il-associated lnvariant chain peptide). An MHC-related dimeric molecule, DM, then

catalyzes the removal of CLIP and keeps the groove open so that peptides generated in the endosome can be inserted (figure 5.17). Initial peptide binding is determined by the concentration of the peptide and its onrate, but DM may subsequently assist in the removal of lower affinity peptides to allow their replacement by high affinity peptides, i.e. act as a peptide editor permitting the incorporation of peptides with the most stable binding characteristics, namely those with a slow offrate. Particularly in B-cells an additional MHC-related molecule, DO, associateswith DM bound to class II and modifies its function in a pH-dependent fashion. Its

C H A P I E Rs - T H E P R I M A R Y I N T E R A C T I OWNI T HA N T I G E N

esl

CLASIVPEPNDE MMPLEX]

Tronsporl intoERond ClossI/peptide complex formotion

CLASS II MHC+ Ii

Figure 5.15. Processing and presentation of endogenous antigen by class I MHC. Cytosolic proteins are degraded by the protcasomc complex into peptitles which are transported into the endoplasmic reticulum (ER) TAP1 and TAP2 arc members of the ABC family of ATP-dependent transport proteins and, under the influence of these transporters,the peptrdesare loaded into the groove ofthe membranebound classI MHC. The pcptide MHC cornplex is then releasedfrom all its associated transporters and chaperones, traverses the Golgi system,and appearson the cell surfaceready for presentationto the Tcell receptor Mutant cells deficient in TAPl /2 do not deliver peptrdes to class I and cannot function as cvtotoxic T-ccll targets

effectmay be to favor the presentation of antigens internalized via the BCR over those taken up by fluid phase endocytosis. The tetraspanin family member CD82 is also present in the MIIC, though its role is unclear at present. The classIl-peptide complexes are eventually transported to the membrane for presentation to Thelper cells.

Figure 5.16. Processing and presentation of exogenous antigen by classII MHC. ClassII rnoleculeswith Ii are assembledin the endoplasmic reticulum (ER) and transported through the Colgi to the transGolgi reticulum (actually as a nonamer consisting of three invariant, three o and three B chains -not shown) There it is sorted to a late endosomal vesicle with lysosomal characteristi.csknown as MIIC (meaning MHC class ll-enriched compartment) containing partially degraded protern derived from the endocytic uptake of exogenous antigen Degradation of the invariant chain leavesthe CLIP (classII-associated lnr.ariant chain peptide) lying in the groove but, under the influence of the DM molecule, this is replaced by other peptides in the vesicle including those derived from exogenous antigen, and the complexes are transported to the cell surface for presentation to T-helper cells This version of events is supported by the finding of high concentrations of invari.ant chain CLIP associated with class II in the MIIC vacuoles of DM-deficient mutant mice which are poor presenters of antigen to T-cells

I roo

C H A P T E sR- T H E P R I M A R Y I N T E R A C T I OWNI T HA N T I G E N

MHClllinvoriont ch0in(li)heterononomer lronsporl Degrodotion of li in MIIC

peplide replocement of CLIPwilh0ntigenic DM-cololysed

Figure 5.17. MHC class II transport and peptide loading illustrated by Tulp's gently vulgar cartoon (Reproduced from Benham A. et al (1995) Inmtunology Today16,359-362,with permission of the authors and Elsevier ScienceLtd )

C R O S S . P R E S E N I A TFIO ORN A C T I V A I I OONF N A T VC E D 8 *T - C E t t s We have just seenhow, in general,MHC classI presents endogenous antigen whilst MHC class II presents exogenous antigen. However, given that naive T-cells require dendritic cells for their activation, how can cytotoxic T-cells specific for peptides presented by MHC class I become aroused to destroy virus-infected cells?After all, most viruses are not tropic for dendritic cells and therefore not naturally present in the cytosol of the professional APCs. The answer to this conundrum lies in the phenomenon of 'cross-presentation'. Phagocytosed or endocytosed antigens can sneak out of the vacuole into which they have been engulfed and gain entry to the cytosol (figure 5.18).The escape route may involve the Sec61multimolecular channel. Once they enter the cytosol they are fair game for ubiquitination and subsequentdegradation by the proteasome,TAP-mediated transfer into the ER, and presentation by MHC class I. It is also possible that some endocytosed antigens can be loaded directly into recycling MHC class I molecules within the endosome without the need to be first processedin the cytosol. In addition to dendritic cells,macrophagesalso seemto be able to play the cross-presentation game, albeit less efficiently. Conversely, proteins and peptides within the ER are also potential clients for the class II groove and could make the journey to the MIIC. This may occur by chaperone-mediated autophagy involving various heat-

Figure 5.18. Cross-presentation of exogenous antigens. Engulfed exogenous antigens are able to accessthe classI processingpathwayby entering the cytosol from the MHC class II compartments (MIIC), perhaps through Sec61channels Other routes for the presentation of peptides derived from exogenous antigens on MHC class I may include peptide exchange with MHC class I molecules recycling from 'other way the cell membrane. Cross-presentation can also work the round'with endogenous antigens gaining entry into the class II processing and presentation Pathway.

shock proteins which bind to the protein to be processed.The protein complex is then recognized by LAMP-2a and dragged into the lumen of the lysosome for subsequent processing.

CHAPTER s - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N

I H E N A T U RO EFT H E ' G R O O V Y ' P E P T I D E

MHCCloss I - boundpeplide

MHCClossII - boundpeptide

ror I

The MHC grooves impose some well-defined restrictions on the nature and length of the peptides they accommodate and the pattern varies with different MHC alleles.Otherwise, at the majority of positions in the peptide ligand, a surprising degreeof redundancy is permitted and this relatesin part to residues interacting with the T-cell receptor rather than the MHC. Bindinglo MHc clossI X-ray analysis reveals the peptides to be tightly mounted along the length of the groove in an extended configuration with no breathing space for cr-helical structures (figure 5.19).The N- and C-termini are tightly H bonded to conserved residues at each end of the groove/ independently of the MHC allele. The naturally occurring peptides can be extracted from purified MHC class I and sequenced. They are predominantly eight or nine residueslong; longer peptides bulge upwards out of the cleft. Analysis of the peptide pool sequencesusually gives strong amino acid signals at certain key positions (table 5.1). These are called anchor positions and represent the preferred amino acid side-chainswhich fit into allele-specificpockets in the MHC groove (figure 5.2}a).There are usually two, sometimesthree,such major anchor positions for classIbinding peptides, one at the C-terminal end (peptide position 8 or 9) and the other frequently at peptide position2 (P2),but they may also occur at P3, P5 or P7. For example, the highly prevalent HLA-A*0201 has pockets for leucine or methionine at peptide position P2 and for valine or leucine at P9. Sometimes, a major anchor pocket may be replaced by two or three more weakly binding secondarypockets.Even with the constraintsof two or three anchor motifs, each MHC classI allele can accommodate a considerable number of different peptides. Except in the caseof viral infection, the natural class I Iigands will be self peptides derived from proteins endogenously synthesized by the cell, histones, heatshockproteins,enzymes,leadersignal sequences,and so on. It turns out that 75"h or so of these peptides originate in the cytosol (figure 5.21)and most of them will be in low abundance, say 100-400 copies per cell. Thus proteins expressed with unusual abundance, such as oncofetal proteins in tumors and viral antigens in infected cells, should be readily detectedby resting T-cells. Bindingto MHCclossil Unlike class I, where the allele-independent H bonding

Figure 5.19. Binding of peptides to the MHC cleft. T-cell receptor 'view'looking down on the s-helices lining the cleft (cf figure 4 13b) represented in space-filling models (a) Peptide 309-317 from HIV-1 reverse transcriptase bound tightly within the class I HLA-A2 cleft. In general, one to four of the peptide side-chains point towards the TCR, giving a solvent accessibllity of 77-27'/.. (b) Influenza hemagglutinin 306-318 lying in the class II HLA-DRI cleft. In contrast with class I, the peptide extends out of both ends of the binding groove and from four to six out of an average of 13 srde-chains point towards the TCR, increasing solvent accessibility to 35% (Based on Vignali D A A. & Strominger J L. (1994) The lmmunologist 2,172, wtthpermission of the authors and publisher )

to the peptide is focused at the N- and C-termini, the classII groove residuesHbond along the entire length of the peptide with links to the atoms forming the main chain. With respect to class II allele-specific binding pocketsfor peptide side-chains,motifsbased on three or four major anchor residues seem to be the order of the day (figure 5.20b).Secondarybinding pockets with less strict preference for individual side-chains can still modify the affinity of the peptide-MHC complex, while 'nonpockets' may also infl uence preferences for particular peptide sequences,especially if steric hindrance

I roz

N I T HA N I I G E N CHAPTER s - T H E P R I M A R YI N T E R A C T I OW

Table 5.1. Natural MHC class I peptide ligands contain two allelespecific anchor residues. (Based on Rammensee H.G, Friede T. & Stevanovic S (1995) lmmunogenetics 47, 178.) Letters represent the Dayhoff code for amino acids; where more than one residue predominates at a given position, the alternative(s) is given; . = any residue

I Closs ollelet2Z4b6789

AminoocidPosilion

.YI/L L/M

Y/F N

L/M/I

L/MA

.LN/|/M

.R/K/L/F

pockels Peptide binding lo clossI onchor

I cvtorotl. ffi Membrone-ossociofed

]

Exogenous

MHc-reloled

Figure 5.21. The origins of class I- and class II-bound peptides' The majority of class I peptides are derived from endogenous self proteins and, during infection, cell-resident pathogens such as viruses' Processing in the endosomal comPartments ensures that proteins of endogenous origin and those derived from membranes constitute over 90% of the peptides bound to the class II grooves (Reproduced from Vignali D A A. & Strominger J.L (1994) The Immunologist 2, 1L2, wlth permission ofthe authors and publisher )

to ClossII onchorpockets Peplide binding

grooves Figure 5.20. Allele-specific pockets in the MHC-binding bind the major anchor residue motifs of the peptide ligands. Crosssection through the longitudinal axis of the MHC groove. The two crhelices forming the lateral walls of the groove lie horizontally above andbelow the plane of the paper. (a) The class I groove is closed atboth ends The anchor at the carboxy terminus is invariant but the second anchor very often at P2 may also be at P3, P5 or P7 depending on the MHC al1e1e(cf. table 5 1) @) By contrast, the class II groove is open at both ends and does not constrain the length of the peptide There are usually three major anchor pockets at P1, P4, P6, P7 or P9 wiih Pl being the most important

becomes a factor. Unfortunately, we cannot establish these preferences for the individual residues within a given peptide because the open nature of the class II groove places no constraint on the length of the peptide, which can dangle nonchalantly from each end of the

groove, quite unlike the strait-jacket of the class I ligand site (figures 5.19 and 5.20).Thus, as noted earlier, each class II molecule binds a collection of peptides with a spectrum of lengths ranging fromeight to 30 amino acid residues, and analysis of such a naturally occurring pool isolated from the MHC would not establish which amino acid side-chains were binding preferentially to the nine available sites within the groove. One modern approach to get around this problem is to study the binding of soluble class II molecules to very large libraries of random-sequence nonapeptides expressed on the surface of bacteriophages (cf. the combinatorial phage libraries, p. 115).The idea is emerging that each amino acid in a peptide contributes independently of the others to the total binding strength, and it should be possible to compute each contribution quantitatively from this random binding data, so that ultimately we could predict which sequences in a given protein antigen would bind to a given class II allele. Becauseof the accessiblenature of the groove, as the native molecule is unfolded and reduced, but before any degradation need occur, the high affinity epitopes could immediately bury themselves in the class Il-binding groove where they are protected from proteolysis. At least for the HLA-DR1 molecule it has been shown that peptide binding leads to a transition from a more oPen conformation to one with a more comPact structure extending throughout the peptide-binding groove. Trimming can take place after peptide binding, leaving peptides from eight to 30 amino acids long. Several

CHAPTER s - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N factors will influence the relative concentration of peptide-MHC complex formed: the affinity for the groove as determined by the fit of the anchors, enhancement or hindrance by internal residues (sequences outside the binding residues have little or no effect on peptide-binding specificity), sensitivity to proteases and disulfide reduction, and downstream competition from determinants of higher affinity. The range of concentration of the different peptide complexes which result will engender a hierarchy of epitopes with respect to their ability to interact with Tcells; the most effective will be dominant, the less so subdominant. Dominant, and presumably subdominant, self epitopes will generally induce tolerance during T-cell ontogeny in the thymus (seep. 238).Complexes with some self peptides which are of relatively Iow abundance will not tolerize their T-cell counterparts and these autoreactive T-cells constantly pose an underlying threat of potential autoimmunity. Sercarz has labeled these cryptic epitopes, and we will discuss their possible relationship to autoimmune disease in Chapter 18.

T H Ea p T - C E t tR E C E P I 0 F R0 R M S A T E R N A RCYO M P T EW X I T HM H CA N D ANIIGENIP CE P I I D E The forces involved in peptide binding to MHC and in TCRbinding to peptide-MHC are similar to those seen between antibody and antigen, i.e. noncovalent. When soluble TCR preparations produced using recombinant DNA technology are immobilized on a sensor chip, they can bind MHC-peptide complex specifically with rather low affinities (K") in the 104 to 107 lr-1 range. This low affinity and the relatively small number of atomic contacts formed between the TCRs and their

Figure 5.22. T-cell receptor antigen combining site. Although the surface is relatively flat, there is a clearlyvisible cleftbetween the CDR3crand CDR3p which can accommodate a central upfacing sidechain of the peptide bound into the groove of an MHC molecule. The surface and loop traces of the VaCDRI andCDR2arecolorem d a g e n t aV, p C D R 1 and CDR2blue, VcrCDR3 and Vp CDR3 yellow, and the Vp fourth hypervariable region, which makes contact with some superantigens, orange. (Reproduced from GarciaK C et al (7996) Science274, 209-279, w ith permission )

IO3 I

MHC-peptide ligands when T-cellscontact their target cell make the contribution of TCR recognition to the binding energy of this cellular interaction fairly trivial. The brunt of the attraction rests on the antigenindependent major adhesion molecules,such as ICAM7/2,LFA-7/2 andCD2, but any subsequent triggering of the T-cell by MHC-peptide antigen must involve signaling through the T-cell receptor. Topologyof lhe lernorycomplex Of the three complementarity determining regions present in each TCR chain, CDR1 and CDR2 are much less variable than CDR3 which, Iike its immunoglobulin counterpart, has (D)J sequences which result from a multiplicity of combinatorial and nucleotide insertion mechanisms (cf. p. 68). Since the MHC elements in a given individual are fixed, but great variability is expectedin the antigenic peptide, a logical model would have CDR1 and CDR2 of each TCR chain contacting the cx-helices of the MHC, and the CDR3 concerned in binding to the peptide. In accord with this view, several studies have shown that T-cells which recognize small variations in a peptide in the context of a given MHC restriction element differ only in their CDR3 hypervariable regions. The combining sites of the TCRs which have been crystallized to date are relatively flat (figure 5.22),which would be expectedgiven the need for complementarity to the gently undulating surface of the peptide-MHC combination (figure 5.23a).In nearly all the structures so far solved, recognition involves the TCR lying either diagonally or orthogonally (figure 5.23b,c) across the peptide-MHC with the TCR Vs domain overlying the MHC class I ar-helix, or class II pr-helix, and the VB domain overlying the ar-helix. As suggestedabove, the

I ror

I N T E R A C T I OWNI I H A N T I G E N C H A P T E 5R- T H EP R I M A R Y

Figure 5.23. Complementarity between MHC-peptide and T-cell receptor. (a) Backbone structure of a TCR (designated 2C) recognizing a peptide (dEV8) presented by the MHC class I molecule H-2Kb. The TCR is in the top half of the picture, with the cr chain in pink and its CDR1 colored magenta, CDR2 purple and CDR3 yellow The p chain is colored light blue with its CDR1 cyan, CDR2 navy blue, CDR3 green and the fourth hypervariable loop orange Below the TCR is the MHC crchain rn green and pr-microglobulin in dark green The peptide with its side-chains is colored yellow (Reproduced from Garcia K C ef al (7998) Science279,1166-7172,with permission.) (b) The same complex

looking down onto a rnolecular surface representation of the H-2Kb in yellow, with the diagonal docking mode of the TCR in a backbone worm representationcolored pink. The dEV8 peptide is drawnin aball and stick format. (c) By contrast, here we see the orthogonal docking mode of a TCR recognizing a peptide presented by MHC classII The TCR (scD10) backbone worm representation shows the Vu in green and Vp in blue, and the I-Ak class II molecular surface representation has the a chain in light green and the p chain in orange, holding its conalbumin-derived peptide (Reproduced from Reinherz E'L et al'

CDR1 and CDR2 regions of both the TCR chains do indeed mainly bind to the u-helices of the MHC whilst the more variable CDR3 regions make contact with the peptide, particularly focusing in on the middle residues (P4 to P6). There is evidence to suggestthat the TCR initiallybinds to the MHC in a fairlypeptide-independent

fashion, followed by significant conformational changes in the CDR loops of the TCR and in the peptide-MHC to permit further contacts with the peptide. Activation through the TCR can operate if these adjustments Permit more stable and multimeric binding.

( I g 9 g ) 5 . 1l (' ? c { ' 2 8I6q,I 3 I q 2 l , w i t h p e r m i s s i o n . )

CHAPTER 5 - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N

r05 |

I . C E T T S W I I H A D I F F E R E N IO U T T O O K )H

Nonclossicol clossI molecules conolsopresentontigen MHC cl ass l-like mole cul es In addition to the highly polymorphic classical MHC classI molecules (HLA-A, -B and -C in the human and H-2K,D and L in the mouse), there are other loci encoding MHC molecules containing Br-microglobulin with relatively nonpolymorphic heavy chains. These are H2J|ld,Q and T in mice, and HLA-E, F and G in hom*o saprcns. The H-2M3 molecule encoded by the H-2M locus is unusual in its ability to present bacterial N-formyl methionine peptides to T-cells.Expression of H-2M3 is limited by the availability of these peptides so that high levels are only seen during prokaryotic infections. The demonstration of H-2M3-restricted CD8* aB T-cellsspecific for Listeria monocytogenes encourages the view that this class I-like molecule could underwrite a physiological function in infection. Discussion of the role of HLAG expression in the human syncytiotrophoblast will arise in Chapter 16 (seep. 381). Thef amily of CD1 non-MHC but cl ass l-like molecules presents lipid antigens After MHC classI and classII, the CD1 family (seep. 83) represents a third lineage of antigen-presenting molecules recognizedby TJymphocytes. The CD1 polypeptide chain associateswith Br-microglobulin as becomes an honest class I-like moiety, and the overall structure is similar to that of classicalclassI molecules, although the topology of the binding groove is altered (see figure4.23). CDl, molecules can present a broad range of lipid, glycolipid and lipopeptide antigens, and even certain small organic molecules, to clonally diverse aB and y6 T-cells. A common structural motif facilitates CD1-mediated antigen presentation and comprises a hydrophobic region of a branched or dual acyl chain and a hydrophilic portion formed by the polar or charged groups of the lipid and/or its associatedcarbohydrate or peptide. The hydrophobic regions are buried in the deep binding Broove of CD1, whilst the hydrophilic regions, such as the carbohydrate structures, are recognized by the TCR (figure 5.24). One major group of ligands for CDlb are glycophosphatidylinositols, such as the mycobacterial cell wall component Iipoarabinomannan. Both endogenous and exogenous lipids can be presented by CD1 and, like MHC class I, the CD1 heavy chain complexes with calnexin and calreticulin in the endoplasmic reticulum. Erp57 is then recruited into the

CDI b-Pldins

-{^

0" l u

Phosphotidyl-inosilol Figure 5.24. Antigen presentation by CD1. In this example the binding of phosphatidylinositol (Ptdins) to CDlb is shown with the binding pocket represented from a top view, looking directly into the groove Aliphatic backbones are in green, phosphor atom in blue and oxygen atoms in red (Reproduced with permission from Hava D L. et al (2005)Current Opinion in lmmunology 17, 88-94).

complex. Subsequent dissociation of the complex permits the binding of pr-microglobulin and, in a step that may involve the rzicrosomal friglyceride-transfer protein (MTP), the insertion of endogenous lipid antigens into the CD1 antigen-binding region. ]ust like their proteinaceous colleagues, exogenously derived lipid and glycolipid antigens are delivered to the acidic endosomal compartment. Localization of CD1b, c and d molecules to the endocytic pathway is mediated by a tyrosine-based cytosolic targeting sequence in the cytoplasmic tail. CD1a, which lacks this targeting motif, traffics through early recyling endosomes. Both humans and mice deficient in prosaposin, a precursor molecule of the sphingolipid activator proteins (SAPs) saposin A-D, are defective in the presentation of lipid antigens to T-cells. Various lines of enquiry indicate that these molecules are probably involved in the transfer of lipid antigens to CD1 in the endosomes.

l-cellshoveNKmorkers Some NKT-cells possessthe NK1.1 marker, characteristic of NK cells, together with a T-cell receptor. However, the TCR bears an invariant cr chain (Vs14Jcr18 in mice, Yu24lu78 in humans) with no N-region modifications and an extremely limited B chain repertoire based upon Vp11 in the human and VBS in the mouse. They recognize lipid antigens such as u-galactosylceramide

I roo

CHAPTER 5 - T H E P R I M A R YI N T E R A C T I OW N I T HA N T I G E N

and isoglobosides presented by CDld and constitute a major component of the T-cell compartment, accounting for 20-30% of T-cells in bone marrow and liver in mice, and up to 1% of spleen cells.The ability of NKT-cells to rapidly secrete IL-4 and IFNy following stimulation suggests they may have important regulatory functions.

y6 TGRshovesomefeoturesof ontibody Unlike uB T-cells, y6 T-cells recognize antigens directly without a requirement for antigen processing. In the mouse, y6 T-cells have been isolated which directly recognize the MHC class I molecule I-Ek and the nonclassical MHC molecules T10 and T22. Neither the polymorphic residues associatedwith peptide binding to classical MHC molecules nor the peptide itself are involved. T10 and T22 are expressedby aB T-cells following their activation, and ithas been suggestedthaty6 T-cells specific for these nonclassical MHC molecules may exert a regulatory function. Stressedor damaged cells appear to be powerful activators of y6 cells, and there is evidence for molecules such as heat-shock proteins as stimulators of yb T-cells. Low molecular weight phosphate-containing nonproteinaceous antigens, such as isopentenyl pyrophosphate and ethyl phosphate, which occur in a range of microbial and mammalian cells, have also been identified as potent stimulators. Evidence for direct recognition of antigenbyy6 T-cells came from experiments such as those involving a y6 Tcell clone specific for the herpes simplex virus glycoprotein-1. This clone could be stimulated by the native protein bound to plastic, suggesting that the cells are triggered by cross-linking of their receptors by antigen which they recognize in the intact native state just as antibodies do. There are structural arguments to give weight to this view. The CDR3 loops, which are critical for foreign antigen recognition by T-cells and antibodies, are comparable in length and relatively constrained with respect to size in the cxand p chains of the op TCR, reflecting a relative constancy in the size of the MHC-peptide complexes to which they bind. CDR3 regions in the immunoglobulin light chains are short and similarly constrained in length, but in the heavy chains they are longer on average and more variable in Iength, related perhaps to their need to recognize a wide range of epitopes. Quite strikingly, the yDTCRs resemble antibodies in that the ychain CDR3 loops are short with a narrow length distribution, while in the 6 chain they are long with a broad length distribution. Therefore, in this respect, the y6 TCR resembles antibody more than the cxpTCR. The X-ray crystallographic structure of a yD

TCR bound to its ligand, the nonclassical MHC moleculeT22mentioned above, has recentlybeen solved. In this example the extended CDR3 loop of the 6 chain, particularly the D62 segment encoded by a nonmutated (germline) sequence, mediates most of the binding with a minor contribution also made by the CDR3 of the y chain. Whilst structural determinations of additional y6TCR-antigen complexes will reveal whether this type of interaction is representative of other y6 TCRs, the broad length distribution of the different CDR3 V6 loops suggests that y6 TCRs will generally have topographically more adventurous binding sites than the TCRs of oB T-cells, thereby facilitating the ability of y6 T-cells to interact with intact rather than processed antigen. A particular subsetof y6 cells,which possessa diverse range of TCRs utilizing different D and/gene segments but always using the same Vgene segments,Vy9 (Vy2.in an alternative nomenclature system) and V 62,exp and in aiao to comprise a large proportion (8-60%) of all peripheral blood T-cells during a diverse range of infections. These V^pV62 T-cells recognize alkylamines and organophosphates. Indeed, individual V"BV62 T-cells can recognize both positively charged alkylamines and negatively charged molecules such as ethyl phosphate. However, this should be fairly straightforward for the receptor given the small hapten-like size of these antigens.Anumber of alkylamine antigens are producedby human pathogens, including Sqlmonellatyphimurium, Listeria monocytogenes, Yersinia enterocolitica and Escherichia coli. The above characteristics provide the y6 cells with a distinctive role complementary to that of the up population and enable them to function in the direct recognition of microbial pathogens and of damaged or stressed host cells. Recent evidence also suggeststhat they can express MHC class II and may be able to act as professional antigen-presenting cells for the activation of ap T-cells.

S U P E R A N T I G ESNI ISM U T A I E WHOTE F A M I T I EO SFT Y M P H O C YRTEEC E P T O R S Bocteri0l loxinsrepresenl 0nemoiorgroup of T-cell superontigens Whereas an individual peptide complexed to MHC will react with antigen-specific T-cells which represent a relatively small percentage of the T-cell pool because of the requirement for specific binding to particular CDR3 regions, a special class of molecule has been identified which stimulates the 5-20% of the total T-cell population expressing the same TCR VB family structure.

W I T HA N T I G E N CHAPTER s - T H E P R I M A R YI N T E R A C I I O N

r07 |

pyogenic toxin superantigen family include staphylococcal exfoliative toxins (ETs), Mycoplasma arthritidis and Yersinia pseudotuberculosis mitogen (MAM) mitogen. Superantigens are not processed by the antigenpresenting cell, but cross-link the class II and VB independently of direct interaction between MHC and TCR molecules (fi gure 5.25).

(MMTV) oct viruses mommory lumor mouse Endogenous ossuperonligens

Figure 5.25. Interaction of superantigen with MHC and TCR. In this composite model, the interaction with the superantigen staphylococcal enterotoxin B (SEB) involves SEB wedging itself between the TCR Vp chain and the MHC, effectively preventing interaction between the TCR and the peptide in the groove, and between the TCR p chain and theMHC. Thus directcontactbetweentheTCRandtheMHC islimited to Vo amino acid residues (Reproduced from Li H et al (7999)Annual Rettiewof lmmunology 17,435466, with permission.) Other superantigens disrupt direct TCR interactions with peptide-MHC to varying extents.

These molecules do this irrespective of the antigen specificity of the receptor. They have been described as superantigens by Kappler and Marrack. The pyogenic toxin superantigen family can cause food poisoning, vomiting and diarrhea and includes Staphylococcusaureus enterotoxins (SEA, SEB and several others), staphylococcal toxic shock syndrome toxin-1 (TSST-1),streptococcalsuperantigen (SSA)and several streptococcal pyogenic exotoxins (SPEs). Although these molecules all have a similar structure, they stimulate T-cells bearing different VB sequences. They are strongly mitogenic for these T-cells in the presenceof MHC classII accessorycells.SEA must be one of the most potent T-cell mitogens known, causing marked proliferation in the concentration range 10-13to 10 16u. Like the other superantigens it can cause the release of copious amounts of cytokines, including IL-2 and lymphotoxin, and of mast cell leukotrienes, which probably form the basis for its ability to produce toxic shock syndrome. Other superantigenswhich do notbelong to the

Very many years ago, Festenstein made the curious observation that B-cells from certain mouse strains could produce powerful proliferative resPonses in roughly 20% of unprimed T-cells from another strain of identical MHC. The so-calledMls gene product responsible for inciting proliferation turned out to be encoded by the open reading frame (ORF) located in the 3'long terminal repeat of MMTV. These are retroviruses which are transmitted as infectious agents in milk and are specific for B-cells. They associatewith class II MHC in the B-cell membrane and act as superantigens throughtheir affinity for certain TCR Vp families in a similar fashion to the bacterial toxins. Other proposed viral superantigens capable of polyclonally activating T-cells include the nucleocapsid protein of rabies virus, and antigens associated with cytomegalovirus and with Epstein-Barr virus. Microbesconolso provideB-cellsupel0ntigens Staphylococcal protein A reacts not only with the Fcy region of IgG but also with 15-50% of polyclonal IgM, IgA and IgG F(ab')t, all of which belong to the Vtt3 family. This superantigen is mitogenic for B-cells through its recognition by a discontinuous binding sequencecomposed of amino acid residues from FR1, CDR2 and FR3 of the V' domain. The human immunodeficiency virus (HIV) glycoprotein 9p120 also reacts which utilize V"3 family with immunoglobulins members. The binding site partially overlaPs with that for proteinAand utilizes amino acid residues from FR1, CDR1, CDR2 and FR3.

FO OF T RMS I H E R E C O G N I I I OONFD I F F E R E N B YB . A N DI . C E t t S I S ANIIGEN A D V A N I A G E OIUOSI H E H O S I It is our conviction that this section deals with a subject of the utmost importance, which is at the epicenter of immunology. Antibodies combat microbes and their products in the

I roa

N I T HA N T I G E N CHAPTER s - T H E P R I M A R YI N T E R A C T I OW

extracellular body fluids where they exist essentially in their native form (figure 5.26a).Clearly it is to the host's advantage for the B-cell receptor to recognize epitopes on the native molecules.

ExtrocellulorInfecliousogenl

W'T lr-\,- ta

Antibody reoctswilhnoliveepilope Figure 5.26. (a) Antibodies are formed against the native, not denatured, form of infectious agents which are attacked in the extracellular fluids (b) Effector T-ceils recognize infected cells by two surface markers: the MHC is a signal for the cell, and the foreign peptide is present in the MHC groove since it is derived from the proteins of an intracellular infectious agent Further microbial cell surface signals can be provided by undegraded antigens and low molecular weight phosphate-containing antigens (seen by yD T-cells), and lipids and glycolipids presented by CD1 molecules

Anllbody recognilion o Antibodies recognize molecular shapes (epitopes) on antigens. . Most protein epitopes are discontinuous involving key residues from different parts of the linear sequence of the protein, although some are continuous and can be mimicked by linear peptides. o The antibody combining site forms a complementary surfacetotheepitopeontheantigenandlargelyinvolvesthe CDRs of the antibody. . Antibody combining sites come in many shapes and sizes; anti-protein antibodies tend to have more extended recognition surfaces than antibodies to carbohydrates or peptides that are more likely to involve grooves or pockets. o Both antibody and antigen can sometimes undergo local changes in conformation to permit interaction. Elicilingonlibodies r Antigenicity (the ability of an antigen to be recognized by antibodies) can be distinguished from immunogenicity (the

oB T-cellshavequite a differentiob.Inthecase of cytotoxic T-cells, and the T-cells which secretecytokines that activate infected macrophages, they have to seek out and bind to the infected cells and carry out their effector function face to face with the target. First, with respect to proteins producedby intracellular infectious agents, the MHC molecules tell the effector T-lymphocyte that it is encountering a cell. Second, the T-cell does not want to attack an uninfected cell on whose surface a native microbial molecule is sitting adventitiously nor would it wish to have its antigenic target on the appropriate cell surface blocked by an excess of circulating antibody. Thus it is of benefit for the infected cell to express the microbial antigen on its surface in a form distinct from that of the native molecule. As will now be more than abundantly clear, the evolutionary solution was to make the T-cell recognize a processedpeptide derived from the intracellular antigen and to hold it as a complex with the surface MHC molecules. The single T-cell receptor thenrecognizesboth the MHC cell markerand the peptide infection marker in one operation (figure 5.26b). A comparable situation arises when CD1 molecules substitute for MHC in antigen presentation to T-cells, in this case associating with processed microbial lipids and glycolipids. The physiologicalrole of theyDcellshas yet tobe fully unraveled.

ability of an antigen (immunogen) to elicit antibodies when usedtoimmunizeananimal). . Small molecule haptens only elicit antibodies when linkedtoaproteincarriermolecule. o Certain epitopes, usually those with the greatest accessibility on the surface of the protein, e.g. loops, elicit far stronger antibody responses than others. o Many viruses, such as influenza and HIV, use the tendency of the antibody response to focus on immunodomi'escape' antibody control. nant epitopes to lhermodynomics of onlibody-onligen inleroclion o Antibody-antigeninteractionisreversibleandsubjectto thelawsof thermody'namics. o The tendency of antibody and antigen to interact is reflected in a binding constant (K") and a free energy for the interaction (AG). . Physiologically active antibodies mostly have binding constantsof the order of 10elv ('nvbinders'). o The energetics of antibody-antigen interaction is domi'hot spots'. nated by a few (Continuedp109)

C H A P I E Rs - T H E P R I M A R Y I N I E R A C T I OW N I T HA N T I G E N

o Multivalency can greatly enhance functional antibody affinity with significant physiological consequences, e.g in toxin inactivation. . High affinity physiologically active antibodies generally have much lower affinities for antigens other than the target antigen, i.e. they have low cross-reactivity. T-cell recognilion . oF T-cells seeantigen in association with MHC molecules. o The T-cells are restricted to the haplotype of the cell to whichtheywereinitiallyprimed. o Protein antigens are processed by antigen-presenting cells to form small linear peptides which associate with the MHC rnolecules, binding to the central groove formed by the cn-helicesand the B-sheetfloor. Processingof ontigenfor presenlofionby closs I MHC ' Endogenous cytosolic antigens such as viral proteins are cleaved by immunoproteasomes and the peptides so formed are transported to the ERby the TAP1 /2 system. o The peptide then dissociatesfrom TAP1/2 and forms a stable heterotrimer with newly synthesized class I MHC heavychainandBr-microglobulin. o This peptide-MHC complex is then transported to the surface for presentation to cytotoxic T-cells. Processingol ontigenfor ptesentofionby closs ll MHC ' The o and p chains of the class II molecule are synthesized in the ER and complex with membrane-bound invariantchain(Ii). . This facilitates transport of the vesicles containing class II across the Golgi and directs them to an acidified late endosome containing exogenous protein taken into the cell by endocytosisor phagocytosis. o Proteolyticdegradationofliintheclassllenrichedcompartments(MIIC)leavesapeptidereferredtoasCl-lPwhich protects the MHC groove ' Processingby endosomal proteasesdegrades the antigen to peptides which replace the CLIP . The class Il-peptide complex now appears on the cell surfacefor presentation to T-helper cells. Cross-presenl0lion o Naive CD8* cells are activated by dendritic cells which take up viruses by endocytosis and then transfer viral antigens into the cytosol through channels such as that created by the Sec61multimolecular complex o Proteasomal processing generates virus-derived peptides for presentation on the MHC class I molecules of the

o Theyareusuallyeightornineresiduesinlengthandhave two or three key anchors, relatively invariant residues which bind to allele-specific pockets in the MHC. . Class II peptides are between eight and 30 residues long, extend beyond the groove and usually have three or four anchor residues. . The other amino acid residues in the peptide are greatly variable and are recognized by the T-cell receptor. ComplexbelweenTCR,MHCond peplide o The first and second hypervariable regions (CDR1 and CDR2)of eachTCRchainmostlycontacttheMHCcr,-helices, while the CDR3s, having the greatest variability, interact with the antigenic peptide. SomeT-cellsore independenlof clossic0lMHCmolecules . MHC classI-like molecules, such as H-2M, are relatively nonpolymorphic and can present antigens such as bacterial N-formyl methionine peptides. . The CD1 family of non-MHC class I-like molecules can presentantigenssuchaslipidandglycolipidmycobacterial antigens. . y6 T-cells resemble antibodies in recognizing whole unprocessed molecules such as low molecular weight phosphate-containing nonproteinaceous molecules. Superontigens . These are potent mitogens which stimulate whole lymphocyte subpopulations sharing the same TCR VB or immunoglobulin V' family independently of antigen specificity. . Staphylococcusaureus enterotoxins are powerful human superantigens which cause food poisoning and toxic shock sy.ndrome. . T-cell superantigens are not processed but cross-link MHC class II and TCR VB independently of their direct interaction. . Mouse mammary tumor viruses are B-cell retroviruses which are superantigens in the mouse. Recognilionof differenlformsof onligenby B- ond T-cellsis 0n 0dvonloge o B-cells recognize epitopes on the native antigen; this is important because antibodies react with native antigen in the extracellular fluid. o T-cellsmustcontactinfectedcellsand,toavoidconfusion between the two systems, the infected cell signals itself to the T-cell by the combination of MHC and degraded antigen.

dendritic cells. o Autophagy can transfer ER resident proteins to the MIIC for subsequentprocessingand classII presentation.

Ihe nolureoflhepeptide . Class I peptides are held in extendedconformatron within theMHC groove.

rosI

(www.roifl,com) Seetheoccomponying website formultiole choice ouestions.

I I IO

C H A P T E sR- T H E P R I M A R Y I N I E R A C T T OWNI T HA N T I G E N

F U R T H ERRE A D I N G Burton D.R.,StanlieldR.L. & Wilson I.A. (2005)Antibody vs HIV in a clash of evolutionary titans. Proceedings of theNational Academyof SciencesUSA 102,1.494T1.4948. ChapmanH.A. (2006)Endosomalproteasesin antigenpresentation. CurrentOpinionin lmmunology18,78-a4. Davies D.R. & Padlan E.A. (1990)Antibody-antigen complexes. Annual Reoiews of Biochemistry 59,439473. Davis S.J.,IkemizuS.,EvansE.J.etal.(2003)Thenatureof molecular recognition by T-celIs.N atureImmunology 4, 217-224. van den Elmde B.J.& Morel S.(2001)Differential processingof class I-restricted epitopes by the standard proteasome and the -153. immunoproteasome.Currmt Opinionin Immunology 13,1.47 Heath W.R.,Belz G.T.,BehrensG.M. etaI.(2004)Cross-presentation, dendritic cell subsets,and the generation of immunity to cellular ar:irgens.lmmunologicalReaiews199,9+6. MacCallum R.M., Martin A.C.R. & Thomton J.M. (1996) Anti-

body-antigen interactions: contact analysis and binding site topogr aphy.I ournalofMol ecular Bialogy 262,732-745. Moody D.8., Zajonc D.M. & Wilson I.A. (2005) Anatomy of CDl-lipid antigen complexes. Nature Reviewslmmunology 5, 387-399. Nowakowski A., Wang C., PowersD.B. et al. (2002)Potent neutralization of bohrlinum neurotoxinby recombinant oligoclonal antibody. Proceedingsof the National Academy of SciencesUSA 99, 11346-11350. Padlan E.A. (1994)Anatomy of the antibody molecule. Molecular Immunology, 31,169-217. Rudd P.M.,Elliott T.,CresswelIP.et al. (2001)Glycosylationand the immune system.Science 291,2370-2376. SundbergE.J.& Mariuzza R.A. (2002)Molecular recognitionin antibody-antigen complexes. Adoancesin Protein Chemistry 61, 119-1,60. TrombettaE.S.& Mellman I. (2005)Cell biology of antigen processing invitro and invivo. Annual Reoiauof Immunology23,975-1028.

lmmunologicol methods ondqpplicotions

INTRODUCTION In addition to being quite handy for protecting our bodies from harmful infectious agents, antibodies are also incredibly useful and exquisitely-specific reagents for detecting and quantitating other proteins, as well as many other substances.Antibodies have, quite literally, numerous practical applications; ranging from the purification of proteins using antibody-based affinity columns, the detection of circulatinghormones inblood or urine samples for clinical diagnosis, to the exploration of expression and subcellular localization of proteins. A variety of experimental approaches are used to explore the composition and function of the immune system. These approaches range from the purely biochemical, to systems where genetic engineering techniques are used to create null mutations ('knockouts') in genes to explore their role in immunity.

MAKING A N I I B O D I EIS O ORDER Generolion ofpolyclonol 0ntibodies Although antibodies can be raised against practically any organic substance, some molecules elicit antibody responses much more readily than others. Proteins usually make excellent immunogens (i.e. substances that can elicit an immune response), although the immune response will typicallybe concentrated against small regions within the protein (called epitopes or antigenic determinants) spanning approximately 5-8 amino acids.As we discussedearlier (seeChapter 5), an epitope represents the minimal structure required for recognitionby antibody and a relatively large molecule, such as a protein, will usually contain multiple epitopes.

Thus, injection of the average antigen into an animal will almost always elicit the production of a mixture of antibodies that are directed against different epitopes within the antigen. It is also quite possible that some of the antibodies within this mixture may be directed towards epitopes that are also found in other antigens. Such antibodies are said to be cross-reactive against the other antigen to which they also bind. Small organic molecules are typically poor immunogens when injected on their own; the immune system appears unable to recognize these structures efficiently. Notwithstanding this, immunologists have found that such molecules can be made visible to the immune system by covalent coupling to a carrier protein (such as albumin) which is itself immunogenic. Such molecules are called haptens (seefigure 5.6). To generate an antibody against a protein of interest, the standard approach is to inject small samples of the protein (in the microgram range) into an animal such as a rabbit. However, administration of antigen alone is rarely sufficient to provoke a robust immune response, even if the antigen is composed of a high proportion of non-self determinants; co-administration of an adiuvant is required (figure 6.1).While it is not entirely clear exactly how adjuvants work, one important role they peform is to activate DCs and APCs at the site of antigen delivery (seep. 308). Activation of APCs dramatically enhances their ability to provide the costimulatory signals that are required for efficient T- and B-cell activation upon encounter with antigen (seeChapter 8). Potent adjuvants are usually crude preparations of bacterial extracts that contain mixtures of Toll-like receptor (TLR) ligands such as LPS or peptidoglycan. It has been proposed by ]aneway that DCs are incapable of providing costimulatory signals unless activated through their TLRs. Therefore, unless an antigen has intrinsic TLR-

I rrz

CHAPTER 6 _ I M M U N O L O G I C A TM E T H O D S AND APPTICATIONS

Boosler +4

within a cell, to quantify it within a mixture of other antigens, to neutralize its biological activity, and many other applications (these are elaborated upon later in this chapter). It is important to remind ourselves here that antisera generated in this way will also contain considerable amounts of other antibodies (directed against a variety of determinants) that the animal happens to have made in the recent past. These antibodies will usually be of a significantly lower titer than those directed against the antigen we have repeatedly used for immunization, but they can cause problems and may need to be removed from our antiserum for several applications. Fortunately, this canbe achieved by affinity purification (see figure 6.6). Because many antigens contain several distinct epitopes, antisera generated by injection of antigen will typically contain a mixture of antibodies directed against different antigenic determinants on the molecule. Some of these antibodies will bind to the antigen with high avidity, some will not, some will only recognize the native formof the antigen, while others will still recognize the antigen following denaturation to eliminate tertiary structure. Such antisera are said to be polyclonal as they contain a mixture of antibodies that are predominantly, although not exclusively, directed against the immunogen to which they were raised.

Boosler +4

Themonoclonol 0ntibodyrevolution

ANTIGEN ALONE '|05

ln4 E o o

l, nv J

? E l

t' vn 2

l0' 100

r++t+ Inifiol

Boosler c h o l l e n g e+ l wilhAg

Boosler +2

Boosler +3

r0 r1 t2

Weeks ANTIGEN PLUS ADJUVANT 105 rn4 o E E o

I, Nv J

? raz

o (n

1nl

100

0I2345678Iroilt2

ftf++

hiliol Boosler cholleng+ el wilhAg

Boosler +2

Boosler +3

Weeks Figure 6.1. Production of polyclonal antibodies. Repeated immunization with antigen plus adjuvant is required to generate efficient antibody responses as immunization with antigen alone is usually ineffective Polyclonal antisera are generated by immunizing, with a combination of antigen plus adjuvant, several times over a l2-week period Serum antibody titre (i e the highest dilution giving a positive test) frequently increases after each successiveboost with antiocn

binding activity, itwill fail to activate DCs and therefore fail to elicit a potent immune response on its own. As we outlined in Chapter 2, becausea single dose of antigen usually elicits a relatively modest response(see figure 2.72), the antigen is therefore injected several times over a period of 12 weeks or so. During this time, the concentration of antibodies (what is usually referred to as antibody titer) directed against the immunogen will increase (figure 6.1). All going well, we will now have an antiserum that is enriched with antibodies against our protein of interest and this can be used as a probe in many different contexts; to localize an antigen

First inrodents A fantastic technological breakthrough was achieved by Kohler and Milstein who devised a technique for the production of immortal' clones of cells making single antibody specificities by fusing normal antibodyforming cells with an appropriate B-cell tumor line. Theseso-called'hybridomas'are selectedout in a tissue culture medium which fails to support growth of the parental cell types and, by successivedilutions or by plating out, single clones can be established(frgwe 6.2). These clones can be grown up in the ascitic form in mice when quite prodigious titers of monoclonal antibody can be attained, but bearing in mind the imperative to avoid using animals wherever feasible,propagation in large-scale culture is to be preferred. Remember that, even in a good antiserum, over 90"/"of the Ig molecules have little or no avidity for the antigen, and the 'specific antibodies'themselves represent a whole spectrum of molecules with different avidities directed against different determinants on the antigen. What a contrast is providedbymonoclonal antibodies,where all the moleculesproducedby a givenhybridoma are identical: they have the same Ig class and allotype, the same variable

AND APPLICAIIONS CHAPTER 6 - I M M U N O T O G I C A LM E T H O D S

Figure 6.2. Production of rnonoclonal antibodies. Mice immunized with an antigen bearing (shall we say) two epitopes, a and b, develop spleen cells making anti-a and anti-b which appear as antibodies in the serum. The spleen is removed and the individual cells are fused in polyethylene glycol with constantly dividing (i.e 'immortal') B-tumor cells selected for a purine enzyme deficiency and usually for their inability to secrete Ig The resulting cells are distributed into microwell plates in HAT (hypoxanthine, aminopterin, thymidine) medium which kills off the fusion partners. They are seeded at such a dilution that on average each well will contain iess than one hybridoma cell Eachhybridoma -the fusionproduct of a single antibody-formingcell and a tumor cell -will have the ability of the former to secrete a single species of antibody and the immortality of the latter enabling it to proliferate continuously. Thus, clonal progeny can provide an unending supply of antibody with a single specificity-the monoclonal antibody. In this example, we considered the production of hybridomas with specificity for just two epitopes, but the same technique enables monoclonal antibodies to be raised against complex mixtures of multiepitopic antigens. Fusions using rat cells instead of mouse may have

il3|

ANTI-O SERUM

certain advantages in giving a higher proportion of stable hybridomas, and monoclonals which are better at fixing human complement, a useful attribute in the context of therapeutic applications to humans involving cell depletion. Naturally, for use in the human, the ideal solution is the production of purely human monoclonals. Human myeloma fusionpartners have not found wide acceptance since they tend to have low fusion efficiencies, poor growth and secretion of the myeloma Ig which dilutes the desired monoclonal. A nonsecreting heterohybridoma obtained by fusing a mouse myeloma with human B-cells can be used as a productive fusion partner for antibody-producing human B-celis. Other groups have turned to the well-characterized murine fusion Partners, and the heterohybridomas so formed grow well, clone easily and are productive There is some instability from chromosome loss and it appears that antibody production is maintained by translocation of human Ig genes to mouse chromosomes. Fusion frequency is even better if Epstein-Barr virus (EBV)-transformed lines are used instead of B-cells.

I rr+

CHAPTER 6 - I M M U N O T O G I C A LM E T H O D S AND APPLICAIIONS

region, structure, idiotype, affinity and specificity for a givenepitope. The large amount of non-specific, relative to antigenspecific, Ig in a polyclonal antiserum means that background binding to antigen in any given immunological test may be uncomfortably high. This problem is greatly reduced with a monoclonal antibody preparation, since all the antibody is antigen-specific, thus giving a much superior 'signal:noise' ratio. By being directed towards single epitopes on the antigen, monoclonal antibodies frequently show high specificity in terms of their low cross-reactivity with other antigens. An outstanding advantage of the monoclonal antibody as a reagent is that it provides a single standard material for all laboratories throughout the world to use in an unending supply if the immortality and purity of the cell line are nurtured; antisera raised in different animals, on the other hand, may be as different from each other as chalk and cheese. The monoclonal approach again shows a clean pair of heels relative to conventional strategies in the production of antibodies specific for individual components in a complex mixture of antigens. The uses of monoclonal antibodies are truly legion and include: immunoassay, diagnosis of malignancies, tissue typing, serotyping of microorganisms, the separation of individual cell types with specific surfacemarkers (e.g.lymphocyte subpopulations), therapeutic neutralization of inflammatory cytokines and 'magic bullet' therapy with cytotoxic agents coupled to antitumor-specific antibody-these and many other areashavebeen transformed by hybridoma technology. Catalytic antibodies An especially interesting development with tremendous potential is the recognition that a monoclonal antibody to a stable analog of the transition state of a given reaction can act as an enzyme ('abzyrne') in catalyzing that reaction. The possibility of generating enzymes to order promises a very attractive future, and some exceedingly adroit chemical maneuvers have already extended the range of reactions which can be catalyzed in this way. Arecent demonstration of sequence-specific peptide cleavage with an antibody which incorporates a metal complex cofactor has raised the pulse rate of the cognoscenti,since this is an energetically difficult reaction which has an enormous range of applications. Another innovative approach is to immunize with an antigen which is so highly reactive that a chemical reaction occurs in the antibody combining site. This recruits antibodies which are not only complementary to the active chemical, but are also likely to have some enzymic power over the immunogen-substrate

complex. Thus, using this strategy, an antibody with exceptionally broad substrate specificity for efficient catalysis of aldol and retro-aldol reactions was obtained. A key feature of this antibody is a reactive lysine buried within a hydrophobic pocket in the binding site. The antibody remains catalytically active for several weeks following i.v. injection into mice and has therapeutic potential for a version of antibody-directed enzyme prodrug therapy (seep.406), here with the enzyme componentbeing a catalytic antibody. Large combinatorial antibody libraries created by random associationbetween pools of heavy and light chains and expressedon bacteriophages(seebelow) can be screened for catalytic antibodies by using the substrate in a solid-phase state. Cleavage by the catalytic antibody leaves a solid-phase product which can now be identified by a double antibody system using antibodies specific for the product as distinct from the substrate. An area of great interest is the presence of catalytic autoantibodies in certain groups of patients, with hydrolytic antibodies against vasoactive intestinal peptide, DNA and thyroglobulin having been described. Catalytic antibodies capable of factor VIII hydrolysis have also recently been discovered in hemophiliacs given this clotting factor, the antibodies preventing the coagulation function of the factor VIII. Human mono cI on als cqn b e ffi 6 ile While scientists were quick to realize that monoclonal antibodies would make powerful and highly-specific therapeutic agents, particularly for the treatment of cancer, this proved to be rather more difficult than originally anticipated. Mouse monoclonals injected into human subjects for therapeutic purposes are frightfully immunogenic and the human anti-mouse antibodies (HAMAin the trade) so formed are a wretchednuisance, accelerating clearance of the monoclonal from the blood and possibly causing hypersensitivity reactions; they also prevent the mouse antibody from reaching its target and, in some cases,block its binding to antigen. In some circ*mstances it is conceivable that a mouse monoclonal taken upby a tumor cell couldbe processed and become the MHC-linked target of cytotoxic T-cells or help to boost the responseto a weakly immunogenic antigen on the tumor cell surface. In general, however, logic points to removal of the xenogeneic (foreign) portions of the monoclonal antibody and their replacement by human Ig structures using recombinant DNA technology. Chimeric constructs, in which the V' and V, mouse domains are spliced onto human C, and C, genes(figure 6.3a),are far lessimmunogenic inhumans. A more refined approach is to graft the six comple-

AND APPLICATIONS CHAPTER 6 - I M M U N O T O G I C A LM E T H O D S

cHn/ERrc ANTTBoDY

GMFTED CDRs RATCDRS

Figure 5.3. Genetically engineering rodent antibody specificities into the human. (a) Chimeric antibody with mouse variable regions fused to human ig constant regions. (b) 'Humanized' rat monoclonal in which gene segments coding for all six CDRs are grafted onto a human Ig framework.

mentarity determining regions (CDRs) of a high affinity rodent monoclonal onto a completely human Ig framework without loss of specific reactivity (figure 6.3b). This is not a trivial exercise, however, and the objective of fusing human B-cells to make hybridomas is still appealing, taking into accountnot only the gross reduction in immunogenicity, but also the fact that, within a species,antibodies can be made to subtle differences such as major histocompatibility complex (MHC) polymorphic molecules and tumor-associated antigens on other individuals. In contrast, xenogeneicresponsesare more directed to immunodominant structures common to most subjects, making the production of variantspecific antibodies more difficult. Notwithstanding the difficulties in finding good fusion partners, large numbers of human monoclonals have been established. A further restriction arisesbecausethe peripheral blood B-cells,which arethe only B-cellsreadily available in the human, are not normally regarded as a good source of antibody-forming cells. Immortalized Epstein-Barr virus-transf ormed B-cell lines have also been used as a source of human monoclonal antibodies. Although these often produce relatively low affinity IgM antibodies, some useful higher affinity IgG antibodies can occasionally be obtained. The cell lines frequently lose their ability to secreteantibody if cultured for long periods of time, although they can sometimesbe rescuedby fusion with a myeloma cell line to produce hybridomas, or the genescanbe isolated and used to produce a recombinant antibody. A radically different approach involves the production of transgenic xenomouse strains in which megabase-sizedunrearranged human Ig H and rlight chain loci have been introduced into mice whose endogenous murine 1g genes have been inactivated.

Immunization of these mice yields high affinity (10-10-1011u) human antibodies which can then be isolated using hybridoma or recombinant approaches. Potent anti-inflammatory (anti-Il--8) and anti-tumor (anti-epidermal growth factor receptor) therapeutic agentshave alreadybeen obtained using suchmice. There is still a snag in that even human antibodies can provoke anti-idiotype responses;these may have to be circumvented by using engineered antibodies bearing different idiotypes for subsequentinjections.Even more desirablewould be if the prospective recipients could be first made tolerant to the idiotype, perhaps by coadministering the therapeutic antibody together with a nondepleting anti-CD4. Despite the difficulties involved, a battery of humanized monoclonals have now been approved for therapeutic use. These include: anti-Il-2 (kidney transplant rejection), anti-VEGF (colorectal cancer), anti-TNFcr (rheumatoid arthritis), anti-CD11cr (psoriasis), antiCD52 (B-cell chronic lymphocytic leukaemia), antiCD33 (acute myelogenous leukaemia), anti-HER-2 (a subset of metastatic breast cancers)and several others (cf. tabIe77.3).

onlibodies Engineering There are other ways around the problems associated with the production of human monoclonals which exploit the wiles of modern molecular biology. Refer'humanizing' of ence has already been made to the rodent antibodies (figure 6.3), but an important new strategy based upon bacteriophage expression and selection has achieved a prominent position. In essence, mRNA from primed human B-cells is converted to cDNAand the antibody genes, or fragments therefrom, expanded by the polymerase chain reaction (PCR). Single constructs are then made in which the light and heavy chain genes are allowed to combine randomly in tandem with the gene encoding bacteriophage coat protein III (pIII) (figure 6.4).This combinatorial library containing most random pairings of heavy and light chain genesencodesa huge repertoire of antibodies (or their fragments) expressed as fusion proteins with pIII on the bacteriophage surface. The extremely high number of phagesproduced by E. collinfection can now be panned on solid-phase antigen to select those bearing the highest affinity antibodies attached to their surface (figure 6.4). Becausethe genes which encode these highest affinity antibodies are already present within the selected phage, they can readily be cloned and the antibody expressedin bulk. It should be recognized that this selection procedure has an enormous advantage over techniques which employ screening

I rro B-CELLS FROIV IIVIVUNIZED DONOR

C H A P T E 6R- I M M U N O L O G I C A L M E T H O DA SN D A P P T I C A T I O N S

Pool VHCHI

PCR

& l i g h lc h o i n

Construct combinotoriol librory

Express in bocteriophoge in ondexpond Ecoli

gene& Clone produce Ab

Figure 6.4. Selection of antibody genes from a combinatorial library. B-cells from an immunized donor (in one important experiment, human memory peripheral blood cells were boosted with tetanus toxoid antigen after transfer to SCID mice; Duchosal M A cl n1 (1992)Nafurc 355,258)are used for the extraction ofIgC mRNA and the light chain (V.C,) and VrrCrrl genes (encoding Fab) randomly com-

bined in constructsfused to the bacteriophagepIII coat protein gene as shown These were incorporated into phagemids such as pHENI and cxpanded in E coll After infection with helper phage, the recombinant phagesbearrng the highest affinity were selectedby rounds ofpanning on solid-phase antigen so that the genes encoding the Fabs could be

becausethenumber of phageswhichcanbe examined is severallogs higher. Combinatorial Iibraries have also been established using mRNA from unimmunized human donors. Vn, V" and V^ genes are expanded by PCR and randomly recombined to form single-chain Fv (scFv) constructs (figure 6.5a) fused to phage pIII. Soluble fragments binding to a variety of antigens have been obtained. Of special interest are those which are autoantibodies to molecules with therapeutic potential such as CD4 and tumor necrosis factor-cr(TNFcr);lymphocytes expressing such autoantibodies could not be obtained by normal immunization since they would probably be tolerized,but the random recombinatiorrof VHandV, can produce entirely new specificitiesunder conditions in aitro where tolerancemechanismsdo not operate. 'test-tube'operation, Although a this approach to the generation of specific antibodies does resemble the affinity maturation of the immune responsein aiuo (see pp.20a-206) in the sensethat antigen is the determining factor in selectingout the highest affinity responders. In order to increase the affinities of antibodies produced by thesetechniques,antigen canbe used to select higher affinity mutants produced by random mutagenesis or even more effectively by site-directed replacements at mutational hotspots (figure 6.5b), again mimicking the natural immune responsewhich involves random mutation and antigen selection (see pp. 200-202). Afflnity has also b een improved by gene'shuf fling'in which a 7r, gene encoding a reasonableaffinity antibody is randomly combined with a pool of 7. genes and subjected to antigen selection. The process can be

further extended by mixing the V. from this combination with a pool of 7,, genes.It has also proved possible to shuffle individual CDRs between variable regions of moderate affinity antibodies obtained by panning on antigen, thereby creating antibodies of high affinity from relatively small libraries. The isolation of high affinity llama heavy chain antibody Vrr' fragments from immunized animals represents yet another approach. Other novel antibodies have been created.In one construct, two scFvfragments associateto form an antibody with two different specificities (figure 6.5c). Another consistsof a single heavy chain variable region domain (DAB) whose affinity can be surprisingly high-of the 'stickorder of 20 nrra.If it were possible to overcome the iness' of these miniantibodies, their small size could be exploited for tissuepenetration. The design of potential 'magic bullets' for immunotherapy can be based on fusion of a toxin (e.g. ricin) to an antibody Fab (figure

cloned. L, bacterial leader sequence

6.5d). Fields of antibodies Not only can the genes for a monoclonal antibody be expressedin bulk in the milk of lactating animals but plants can also be exploited for this purpose. So-called 'plantibodies' have been expressed in bananas, potatoes and tobacco plants. One can imagine a high-tech farmer drawing the attention of a bemused visitor to one field growing anti-tetanus toxoid, another antimeningococcal polysaccharide, and so on. Multifunctional plants might be quite profitable with, say,the root beingharvested asa food crop and the leavesexpressing

AND APPLICATIONS CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S

'DIABODY BISPECIFIC

SINGLE CHAIN FvFRAGMENT

AFFINITY MUTANT WITHINCREASED

IIVI\4UNOTOXIN

ALANINE ARGININE tl

F(ob')2

Figure 6.5. Other engineered antibodies. (a) A single gene encoding Vn and Vrjoined by a sequence ofsuitable lengthgives rise to a singlechain Fv (scFv) antigen-binding fragment (b) By site-specific mutagenesis of residues in or adjacent to the complementarity determining region (CDR), it is possible to increase the affinity of the antibody. (c) Two scFv constructs expressed simultaneously will associateto form a

'diabody' with two specificities. These bispecific antibodies have a number of uses. Note that such a bispecific antibody directed to two different epitopes on the same antigen will have a much higher affinity 'bonus effect' of cooperation between the two binding sites due to the (cf p. 9a). (d) Potential'magic bullets' canbe constructedby fusing the gene for a toxin (e.g ricin) to the Fab.

groups to cyanogenbromide-activated Sepharoseparticles or some other solid support. Insolubilized antibody, for example, can be used to extract the corresponding antigen out of solution, in which it is present as one Drugs canbebaseil onthe CDRs of minibodies component of a complex mixture, by absorption to its Millions of minibodies composed of a segment of the surface. The uninteresting garbage is washed away and V' region containing three p-strands and the H1 and H2 the required ligand released from the affinity absorbent hypervariable loops were Beneratedby randomization by disruption of the antigen-antibodybonds by changof the CDRs and expressed on the bacteriophage pIII ing the pH or adding chaotropic ions such as thiocoat protein. By panning the library on functionally important ligand-binding sites,such as hormone recepcyanate (figure 6.6). This technique can be used to identify the antigen to which an antibody binds where tors, useful lead candidates for drug design programs this is not known; in the case of an autoantibody for can be identified and their affinity improved by loop example. A very similar aPProach can also be used to optimization, loop shuffling and further selection. identify binding partners for an antigen; such molecules will usually stay attached to the antigen if the P U R I F I C A I I OONFA N I I G E N A SN D immunopurification procedure is carried out under A N T I B O D I EBSYA F F I N I I Y CHROMATOGRAPHY gentle conditions. Many of the proteins that participate in T-cell receptor signal transduction, for instance, were The principle is simple and uery widely applied. initially identified by using antibodies directed against Antigen or antibody is bound through its free amino some desirable gene product. At this rate there may not be much left for sciencefiction authors to write about!

I rra Conjugole ontibody ondpockintocolumn

CHAPTER AND APPTICATIONS 6 - I M M U N O T O G I C A LM E T H O D S

Runonligen mixlure through offinity column

Wosh

Elule

WOSh unD0un0 0nTrgen from thecolumn

Purificofion of onfigen

Anlibody

Activoted sephorose

Wosh unDoun0 onlibody from c0rumn

EC o"+

Elute specific ontibody

Purificotion of onfibody

8 p -E8 Antigen

Conjugole onligen ondpockinlocolumn

Runonliserum lhrough offinily column

Wosh

Figure 6.6. Affinity purification of antigen and antibody. Antibody can be immobilized on activated sepharose and used to affinity-purify antigen Depending on the conditions used to carry out the assay, antigen-associated proteins may also be captured by this procedure. For the purification of specific antibody from a polyclonal antiserum,

antigen is immobilized on Sepharose beads and nonspecific ulbound antibodies fail to be captured and can be washed away. After capture, specific antibody can be eluted by transiently lowering the pH or rncreasing the salt concentration of the buffer

known TCR signalling components to pull out these components from complex protein mixtures, along with their binding-partners. Isolated llama heavy chain (Vgs) fragments are proving to be valuable for repeated cyclesof antigen purification becauseof their resistance to denaturation by repeated cycles of exposure to low pH. In a similar manner, an antigen immunosorbent can be used to absorb out an antibody from a mixture whence it can be purified by elution (figure 6.6).This is especially useful where an antiserum displays high levels of nonspecific reactivity against other antigens rendering it unusable. Affinity-purification of such an antiserum, by means of the antigen that was used to generate it, can often dramatically improve its specificity.

inhibitory effect on the metabolism of the organisms in culture.

MODUTATIO T IIVITY ON FB I O T O G I C A C B YA N I I B O D I E S Todetectantibody A number of biological reactions can be inhibited by addition of specific antibody. Thus the agglutination of red cells by interaction of influenza virus with receptors on the erythrocyte surface can be blocked by antiviral antibodies and this forms the basis for their serological detection. A test for antibodies to SalmonellaH antigen present on the flagella depends upon their ability to inhibit the motility of the bacteria in aitro. Likewise, Mycoplasmaantibodies can be demonstrated by their

Using antibody as aninhibitor The successful treatment of cases of drug overdose with the Fab fragment of specific antibodies has been described and may become a practical proposition if a range of hybridomas can be assembled.Conjugates of cocaine with keyhole limpet hemocyanin (the latter is used as a carrier to elicit efficient Ab production to cocaine) can provoke neutralizing antibodies. Antibodies to hormones such as insulin and thyroid-stimulating hormone (TSH), orto cytokines, canbeused toprobe the specificity of biological reactions in aitro. For example, the specificity of the insulin-like activity of a serum sample on rat epididymal fat pad can be checkedby the neutralizing effect of an antiserum. Such antibodies can be effective in aiao, and anti-TNF treatment of patients with rheumatoid arthritis has confirmed the role of this cytokine in the diseaseprocess.Likewise, as part of the worldwide effort to prevent disastrousoverpopulation, attempts are in progress to immunize against chorionic gonadotropin using fragments of the B chain coupled to appropriate carriers, since this hormone is needed to sustain the implanted ovum. In a totally different context, antibodies raised against myelin-associated neurite growth inhibitory proteins revealed their importance in preventing nerve repair/ in that treatment with these antibodies permitted the

AND APPTICATIONS CHAPTER 6-IMMUN.OLOGICAM L ETHODS regeneration of corticospinal axons after a spinal cord lesion had been induced in adult rats.This quite remarkable finding significantly advances our understanding of the processes involved in regeneration and gives ground for cautious optimism concerning the development of treatment for spinal cord damage, although for various reasons this mav not ultimatelv be based on antibody therapy. Using antibody as an actiuator Antibodies can alsobe used to substitute for naturalbiological ligands, either because the ligand is unknown, is difficult to purify, or would require a small mortgage to be able to afford it! For example, antibodies can be used instead of ligand to stimulate cell-surface receptors that propagate signals into the cell upon crosslinking. Normally, the natural ligand for the receptor would promote receptor crosslinking but antibodies can be used to mimick this very efficiently. Such an approach hasbeenused to greateffectto study intracellular events that take place upon stimulation of T- or B-cell receptor complexes by antibodies directed against these receptors or associatedproteins (such as the CD3 complex). In a similar vein, antibodies directed against the Fas (CD95) cell surface receptor can substitute for the natural ligand (FasL/CD9SL)in order to stimulate the receptor and study the consequencesof this. In the latter case, stimulation of Fas by anti-Fas antibodies induces rapid programmed cell death (apoptosis) in cells bearing this receptor (figure 6.7).Another good example is the induction of histamine release from mast cells by divalent F(ab')2 anti-FceRl but notby the univalent fragment. Antibody-induced activation can be used to study the signal transduction cascadedownstream of receptor engagement by ligand, even where the ligand has not yet been identified. Unlreoled

Figure 6.7. Antibody-induced receptor activation. Transformed Jurkat T-cells were either left untreated, or were treated with anti-Fas IgM antibody for 4 hours. Crosslinking of the Fas (CD95) receptorwith antibody activates the receptor and results in a signal transduction cascade that culminates in activation of a series of cysteine proteases, called caspases,that provoke apoptosis in the stimulated cell. Apop-

I re I

A N T I G EINN C E t t S I M M U N O D E I E C T IOOFN A N DI I S S U E S microscopy Iuorescence Immunof Antibodies can be used as highly sensitive probes to explore the subcellular localization of a protein (or other antigenic determinant) within a cell or a tissue. Because fluorescent dyes such as fluorescein and rhodamine can be coupled to antibodies without destroying their specificity, the conjugates can combine with antigen present in a tissue section and be visualized using a microscope equipped with an appropriate light source (typically UV light). Looked at another way, the method can also be used for the detection of antibodies directed against antigens already known tobe present in a given tissue section or cell preparation. Before aPPlying the antibody to the cell or tissue preparation, samples require fixation and permeabilization in order to preserve cellular structures and to permit free passage of antibody across the plasma membrane. There are two general ways in which the test is carried out. Direct test zuith labeled antibody The antibody to the tissue antigenis directlyconjugated withthe fluorochrome and applied tothe sample (figure 6.8a).Binding of the antibody to the antigen is betrayed by that part of the cell becoming fluorescent when illuminated using UV light. For example, suPPose we wished to show the distribution of a thyroid autoantigen reacting with the autoantibodies present in the serum of a patient with Hashimoto's disease,a type of thyroid autoimmunity. We would isolate IgG from the patient's serum, conjugate itwith fluorescein, and apply it to a section of human thyroid on a slide. \Alhenviewed onti-Fos

totic cells exhibit plasma membrane blebbing and collapse of the cell 'apoptotic bodies' Similar into small fragments or vesicles termed effects are also seen when the natural ligand, FasL, is used instead of anti-Fas antibody. (Kindly provided by Dr. Colin Adrain, Dept. of Genetics, Trinity College, Dublirl Ireland.)

I rzo

CHAPTER 6 - I M M U N O T O G I C A TM E T H O D S AND APPIICATIONS

TISSUE SECTION UNLABELEDFLUORESCEIN FLUORESCEIN -LABELED ANTIBODY -LABELED ANTIBODY ANTI-IMMUNOGLOBULIN

Exciting lighl

FLUORESCEIN

Direcllesl

Unlreoted

Indirecl lest

Sondwich lest

Figure 6.8. The basis of fluorescence antibody tests for identification of tissue antigens ortheirantibodies. O =fluorescein labeled.

LPSJreoled

Figure 6.9. Immunolocalization of a transcription factor upon receptor stimulation Transformed human monocytes (THP-I cells) were either left untreated (a), or were stimulated with bacterial lipopolysaccharide (LPS) for 2 hours (b). Cells were then fixed and immunostained with an antiNFrcBantibody. Note that in unstimulated cells NFrB is abundantly present in the cell cytoplasm but is excluded from the nucleus, whereas the reverse is true in LPSstimulated cells. (Courtesy of Dr Lisa Bouchier-Hayes,St. Jude's Hospital, Memphis,USA )

in the fluorescencemicroscope we would see that the cytoplasm of the follicular epithelial cells was brightly stained (cf. figure 18.1a). Let's consider another example to illustrate the versatility of this technique. We have just generated a monoclonal antibody to a transcription factor (NFrB for example) that is known to be important for LPS-induced macrophage activation and IL-1B production. We could compare resting versus LPS-treated macrophages to determine whether the transcription factor does anything 'interesting' upon exposure of macrophagesto LPS.Inthis casewewould observethat, whereas resting macrophages contain lots of NFrB, it all appears to be in the cytoplasm. However, we would also certainly note that within minutes of exposure to LPS, practically all of the NFrB had moved to the nucleus (figure 6.9). By using two (or even three) antisera conjugated to dyes which emit fluorescence at different wavelengths (figure 6.10),several different antigens canbe identified

simultaneously in the same preparation. In figure 2.6e, direct staining of fixed plasma cells with a mixture of rhodamine-labeled anti-IgG and fluoresceinconjugated anti-IgM craftily demonstrates that these two classesof antibody are produced by different cells. The technique of coupling biotin to the antiserum and then finally staining with fluorescent avidin is often employed. lndirect test with labeled secondary antibody In this double-layer technique, which is the most commonly adopted approach, the unlabeled antibody (the primary antibody) is applied directly to the tissue and visualized by treatment with a fluorochromeconjugated anti-immunoglobulin serum (the secondary antibody; figure 6.8b). Anti-immunoglobulin antisera are widely available conjugated to different fluorochromes. This technique has several advantages. In the first place the fluorescence is brighter than with the direct

121 |

AND APPTICATIONS CHAPTER 6 - I M M U N O T O G I C A LM E T H O D S test since several fluorescent anti-immunoglobulins bind on to each of the antibody molecules present in the first layer (figure 6.8b). Second, even when many sera have to be screened for specific antibodies it is only necessary to prepare (or, more usually the case,purchase) a single secondary antibody. Furthermore, the method

(nm) Moxemission

480

500

520

540

560 580 600 (nm) Excilolion

620

640

660

680

Figure 6.10, Fluorescent labels used in immunofluorescence microscopy and flow cytometry. The fluorescein longer wave emission overlaps with that of Texas Red and is corrected for in the software The phycobiliproteins of red algae and cyanobacteria effect energy transfer of blue light to chlorophyll for photosynthesis; each molecule has many fluorescent groups giving a broad excitation range, but fluorescence is emitted within a narrow wavelength band with such high quantum efficiency as to obviate the need for a second amplifying antibody.

has great flexibility. For example, by using a mixture of primary antibodies directed against different target antigens, it is possible to compare the relative positions arrd/ or expression of two different antigens within the same cell. Note, however, that in the latter scenario the primary antibodies must not have been generated in the same speciesor the secondaryreagentwill notbe able to discriminate between them. For example, to simultaneously label cytochrome c and tubulin in the same cell, one would need to use anti-tubulin antibody that has been raised in the mouse, in combination with anticytochrome c antibody that has been raised in rabbit, or visa versa. By using species-specificsecondary detection reagents (i.e. anti-mouse and anti-rabbit Ig) that are Iabeled with different flourochromes, it is a simple matter to detect both proteins within the same cell (figure 6.11). Further applications of the indirect test maybe seenin Chapter 18. Confocolmicroscopy Fluorescent images at high magnification are usually difficult to resolve because of the flare from slightly out of focus planes above and below that of the object. The resulting blurred images are usually of little help

Phoseconlrosl

Figure 6.11. Confocal immunofluorescence microscopy. Human HeLa cell immunostained with mouse anti-p-tubulin antibody detected with FITClabeled anti-mouse Ig (green) and rabbit anti-cytochrome c antibody detected with Texasred-labeled antirabbit Ig (red) Cells were also stained with the DNA-binding dye, DAPI (blue) Aphase contrast image of the same cell is aiso shown for comparison Images were acquired on an Olympus Flouroview 1000 confocal microscope (Courtesy of Petrina Delivani.)

c cylochrome

Tubulin

Composils

in exploring the finer points of cellular architecture. All that is now a thing of the past, with the advent of commercially available scanning confocal microscopes which focus the laser light source on a fine plane within the cell and collect the fluorescence emission in a photomultiplier tube (PMT) with a confocal aperture. Fluorescence from planes above and below the object plane fails to reach the PMT and so the sharpness of the image is dramatically enhanced over

conventional immunofluorescence microscopy (figure 6.11). An X-Y scanning unit enables the whole of the specimen plane to be interrogated quantitatioely and, with suitable optics, three or four different fluorochromes can be used simultaneously. The instrument software can compute three-dimensional fluorescent images from an automatic series of such X-Y scansaccumulated in the Z axis (hgure 6.12) and rotate them at the whim of the operator. Such Z-stacks can be used to

Figure 6.12. Construction of a three-dimensional fluorescent image with the confocal microscope. A sphericai thyroid follicle in a thick razor-blade section of rat thyroid fixed in formalin was stained with a rhodamine-phalloidin conjugate which binds F-actin (similar results obtained with antibody conjugates) Although the sample was very thick, the microscope was focused on successive planes at 1-pm intervals from the top of the follicle (image no. 1) to halfway tfuough (image no. 8), the total of the images representing a hemisphere Note how the fluorescence in one Dlane does not interfere with that in another and

that the composite photograph (image no 9) of images 1-8 shows all the fluorescent staining in focus throughout the depth of the hemisphere. Clearly the antibody is staining hexagonal structures close to the apical (inner) surface of the follicular epithelial cells. Erythrocytes are visible near the top of the follicle. (Negatives kindly suppliedby Dr Anna Smallcombe were takenby Bio-Rad staff on a Bio-Rad MRC-500 confocal imaging system using material provided by Professor V Herzog and Fr. Brix of Bonn University.)

123 |

CHAPTER AND APPLICATIONS 6 - I M M U N O L O G I C A LM E T H O D S reconstruct a three-dimensional view of a cell, tissue or organelle, and offer unparalleled insights into cell and molecular structure. Timelapseexperiments can also be carried out using the confocal microscope and this often transforms our understanding of events previously only viewed as snapshotsin time. Often, seeingreally is believing!

THYMOCYTES .l04

'l03 .:1.,': -..:j

: 1 1 ' r 1 i : . ..1. , j

,-:

Flowcylomelry o

When a cell population is immunostained for a particuIar marker (CD4 for example) a subsetof the population may express this marker at high levels, a different subset may express the same marker at low levels and the remainder of the population may be negative. To add further complexity, one may wish to examine simultaneously the expression of a different marker (CD8 for example) to determine whether expressionof theseproteins is mutually exclusive. Assessmentof the percentage of cells in a population expressing either CD4 or CDS, or both, would be quite a chore using fluorescence microscopy or confocal microscopy, as this would involve manual counting of several hundred cells to obtain reliable figures. Quite apart from the labor involved, such analyseswould also be quite subjective and results may vary depending on the skill of the operator. Fortunately, the flow cytometer makes such determinations rather trivial as this instrument can analyze the fluorescence levels associated with thousands of cells per minute in a highly reproducible and quantitative way (figure 6.13). In its most basic form, the flow cytometer is an instrument equipped with a fluid-handling system capableof moving thousands of cells in single file through a narrow chamber illuminated by a laser.The passageof an immunolabeled cell through the chamber (called a flow cell) results in excitation of the fluorochrome attached to the cell by the laser. The resulting emission from the fluorochrome is detected by a sensitive photomultiplier-based detectorwhich permits precisequantitation of the fluorescenceassociatedwith the cell. Thus it is possible to rapidly discriminate between cells that are negative, slightly positive or highly positive for a given marker or antigen. Most modern flow cytometers are equipped with three or four lasers of different wavelengths (along with associated detectors) and each laser-detector combination can gather signals from different fluorochromes (figure 6.14).As a result, it is possible to immunostain a cell population for four different markers (with a different fluorochromelabeled antibody used for each one) and to gather data relating to the expressionof all four markers as the cell passesthrough the flow cytometer.

.l02

.,_:1

:., ::+ . , ; ..,,.: :;

r..,?

: ],

i'.t".,1,.,....

;' i;i"i'*:...'t '"

l0l

,;+.;l'l

: .-'.a!t''

i.r- :.l'.'].' .: . :,r."?. . .. ,. . : r i l ' :{i:-

, . ..::.

100 100

l0r

102

l03

'l04

cD4 SPLENOCWES

l02

Figure 6.13. Flow cytometric analysis of CD4 and CD8 expression in thymocytes and splenocytes, Mouse thymocytes and splenocytes were stained using FlTC-conjugated anti-CD4 and Rhodamineconjugated anti-CD8 antibodies Note that the majority of thymocytes are positive for both CD4 and CD8 and are therefore present in the upper right quadrant; thymocytes single-positive for CD4 (bottom right) or CD8 (top left) are also detected, as are double negative cells (bottom left). In the spleen, few double positive cells are found, with the majoriiy of cells (most likely B-cells) negative for CD4 or CD8 along with cells that are single-posiiive for either marker. (Data kindly provided by Professor Thomas Brunner and Daniela Kassahn.)

ln+

C H A P T E 6R- I M M U N O L O G I C A L M E T H O DA SN D A P P L I C A T I O N S

620/20 bondposs

640nm reflector

The good news doesn't end there; the flow cytometer is also capable of providing information relating to cell size and granularity (organelle density) due to the way in which the laser light is scattered or reflected as it passes through the cell. The latter information (called forward and side scatter)is also very useful as this alone is often enough to permit discrimination between distinct cell types (figure 6.15a). Thus, the flow cytometer records quantitative data relating to the antigen content and physical nature of each individual cell, with multiple parameters being assessedper cell to give a phenotypic analysis on a single cell rather than a population average. With the impressive number of monoclonal antibodies and of fluorochromes to hand, highly detailed analyses are now feasible, with a notable contribution to the diagnosis of leukemia (cf. p. 394). We can also probe the cell interior in several ways. Permeabilization to allow penetration by fluorescent antibodies (preferably with small Fab or even singlechain Fv fragments) gives a readout of cytokines and other intracellular proteins. Cell cycle analysis can be achieved with DNA-binding dyes such as propidium iodide to measure DNAcontent (figure 6.15b)and antibody detection of BrdU incorporation to visualize DNA synthesis. In addition, fluorescent probes for intracellu-

Figure 6.14. Six-parameter flow cytometry optical system for multicolor immunofluorescence analysis. Ce11fluorescence excited by the blue laser is divided into green (fluorescein) and orange (phycoerythrin) signals, while fluorescence excited by the orange laser is reflected by a mirror and divided into near red (TexasRed) and far red (allophycocyanin) signals. Blue light scattered at small forward angles and at 90ois also measured in this system, providing information on cell size and internal granularity respectively. PMI photomultiplier tube. (Based closely on H a r d y R R ( 1 9 9 8 )I n D e l v e s P J & R o i t t I M (eds) Encyclopediaoflmmunology, 2nd edn, p 946 Academic Press,London ) The recent use of three lasers and nine different fluorochromes pushes the system even further, providing 11parameters!

lar pH, thiol concentration, Caz*,ll4g2* and Na+ have been developed.

olherlobeledonlibodymefhods A problem with fluorescent conjugates is that the signals emitted by these probes fade within a relatively short time; photobleaching of the fluorescent label upon exposure to an excitation source (such as UV light) can also occur. In practice, this is not a problem so long as the labeled sample is analyzed in a timely fashion. However, enzymes such as alkaline phosphatase (cf. figure 17.9)or horseradish peroxidase can be coupled to antibodies and then visualized by conventional histochemical methods under the light microscope (figure 6.16).Such stains are relatively stable and are particularly useful for staining tissue sectionsas opposed to cell susPenslons. Colloidal gold bound to antibody is being widely used as an electron-dense immunolabel by electron microscopists.At least three different antibodies can be applied to the same section by labeling them with gold particles of different size (cf. figure 8.13).A new ultrasmall probe consisting of Fab' fragments linked to undecagold clusters allows more accuratespatial localization of antigens and its small size enablesit to mark

CHAPTER 6 - I M M U N O L O G I C A tM E T H O D S AND APPLICATIONS

125 |

sites which are inaccessible to the larger immunolabels. However, clear visualization requires a high-resolution scanning transmission electron mrcroscope.

ONFA N I I G E N DEIECIIOA N N DS U A N T I I A T I O B YA N T I B O D Y lmmunoossoy of onligenby EIISA

o o

r00 80 C

:z.6 0 o

GO/GI

400

600

800

I 000

DNA conlent Figure 6.15. Analysis of cell size, granularity and cell cycle position by flow cytometry. (a) Staining of peripheral blood with anti-CD4, anti-CD8, and anti-CD14 antibodies followed by analysis of cells by forward-scatter (FSC) and side-scatter (SSC) characteristics (top left) Cells with low forward- and side-scatter characteristics (lymphocytes, R1) were analyzed for CD4 or CD8 expression, as indicated Cells wiih high forward- and side-scatter characteristics (monocytes, R2) were anaiyzed for CD14 expression. (Data kindly provided by Professor Thomas Brunner.) (b) Cel1 cycle analysis of transformed furkat T-cells by propidium iodide staining The fluorescent DNA-binding dye, propidium iodide, stains cells in proportion to their DNA content; cells with normal diploid (2N) DNA content appear in the G0lG1 phase of the cell cycle, cells actively synthesizing DNA have greater than 2N DNAcontent and are therefore assigned to S-phase,whereas cells with 4N (diploid) DNAcontentareinG2 ormitosis (C2lM) (Courtesyof Dr ColinAdrain.)

The ability to establish the concentration of an analyte (i.e. a substance to be measured) through fractional occupancy of its specific binding reagent is a feature of any ligand-binding system (Milestone 6.1), but because antibodies can be raised to virtually any structure, its application is mostversatile in immunoassay. Large analytes, such as protein hormones, are usually estimated by a noncompetitive two-site assay in which the original ligand binder and the labeled detection reagent are both antibodies (figure M6.1.1). By using monoclonal antibodies directed to two different epitopes on the same analyte, the system has greater power to discriminate between two related analytes; if the fractional cross-reactivity of the first antibody for a related analyte is 0.1and of the secondalso 0.1,the final readout for cross-reactivity will be as low as 0.1 x 0.'l',i.e. 1o/". Using chemiluminescent and time-resolved fluorescent probes, highly sensitive assays are available for an astonishing range of analytes. For small molecules like drugs or steroid hormones, where two-site binding is impractical, competitive assays (figure M6.1.1) are appropriate. The ELISA (Enzyme-Linked Immunosorbent Assay) is one of the most commonly used techniques for measuring antigens, such as cytokines, from serum or cell culture fluid. The technique is quite straightforward and involves immobilizing antibody to the protein of interest within the plastic wells of a microtiter plate. Unbound protein-binding siteswithinthe plate are then blocked by incubation with an irrelevant protein such as albumin. Samples containing the antigen of interest are then added to the antibody-coated wells and incubated for a couple of hours to allow captureof the antigen by antibody. Following washing to remove nonbinding material, the bound antigen is then detectedby adding a second antibody which is directed against a different binding-site on the antigen to the one recognized by the capture antibody. The antigen is now sandwiched between the two antibodies giving rise to the terms 'sandwich ELIS,{ or antigen-capture assay.The detection antibody is conjugated to an enzyme such as horseradish peroxidase or alkaline phosphatase which, upon addition of the enzyme substrate, produces a colored or chemiluminescent reaction product. Comparison

H&E

Anli-cD3

Anli-CD68

Anti-CD2l

Anli-cD20

Anti-cD3(brown) Anti-CD2o(ed)

Figure 6.16. Immunohistochemical analysis of human tonsil follicle centers. Human tonsil preParations were stained either with the histochemical stain hematoxylin and eosin (H&E), or were immunostained with antibodies against CD21 (complement receptor 2, expressed on follicular dendritic cells and B-cells), CD68 (expressed on macrophages), CD3 (T-ce1ls),CD20 (B-cells), or a combination of anti-CD3 and anti-CD20, as shown. (Images kindly provided by Dr Andreas Kappeler )

between a range of standards of known concentration enables the concentration of antigen in the test samples tobe calculated.

antigen excess/it is important to ensure that the value for antigen was obtained in antibody excessby running a further control in which additional antigen is included.

Thenephelomelric0ssoyfor 0nligen If antigen is added to a solution of excessantibody, the amount of complex which can be assessedby forward light scatter in a nephelometer (cf. p. 133) is linearly related to the concentration of antigen. With the ready availability of a wide range of monoclonal antibodies which facilitate the standardization of the method, nephelometry is frequently used for the estimation of immunoglobulins, C3, C4, haptoglobin, ceruloplasmin and C-reactive protein in those favored laboratories which can sport the appropriate equipment. Very small samplesdownintherange 1-10 plcanbe analyzed. Turbidity of the sample can be a problem; blanks lacking antibody can be deducted but a more satisfactory solution is to follow the rate of formation of complexes whichis proportional to antigenconcentrationsincethis obviates the need for a separate blank (figure 6.77). Because soluble complexes begin to be formed in

lmmunoblotling(weslernblofiing) This widely adopted technique can be used to determine the relative molecular mass of a protein and to explore its behavior within a complex mixture of other proteins. Issues relating to whether the protein of interest is upregulated, downregulated, cleaved, phosphorylated, glycosylated or ubiquitinated in response to a particular stimulus can be addressed by immunoblot analysis. This involves first rururing a mixture of proteins through a gel matrix that is formed by polymerization of acrylamide and bisacrylamide between a pair of glassplates.Polyacrylamide gel electrophoresis (PAGE) of proteins is typically carried out using protein mixtures that have been denatured by heating in the presence of a detergent, sodium dodecyl sulfate (SDS). SDS is a negatively charged molecule that becomes covalently coupled to proteins along their length upon

127 |

CHAPTER AND APPLICATIONS 5-IMMUNOIOGICAt METHODS

The appreciation that a ligand could be measured by the fractional occupancy (F) of its specific binding agent heralded a new order of sensitive wide-ranging assays.Ligand-binding assayswere first introduced for the measurement of thyroid hormone by thyroxine-binding protein (Ekins) and for the estimation of hormones by antibody (Berson & Yalow). These

reaction. F can be measured by noncompetitive or competitive assays (figure M6.1.1) and related to a calibration curve constructed with standard amounts of analyte. For competitive assays,the maximum theoretical sensitivity is givenby the term e/Kwhere eis the experimental error (coefficient of variation). Suppose the error is 1% and K is 1011M-1, the maximum sensitivity will be 0.01 x 10-11v =

findings spawned the technology of radioimmunoassay, so called because the antigen had to be trace-labeled 1n some way and the most convenient candidates for this were radioisotopes. The relationship between fractional occupancy and analyte

10-13rra or 6 x 107molecules/ml. For noncompetitive assays, labels of very high specific activity could give sensitivities down to 102-103molecules/ml under ideal conditions. In practice, however, since the sensitivity represents the lowest analyte concentration which can be measured against a background containing zero analyte, the error of the measurement

concentration [An] is givenby the equation: F= 1- (1/1+I(Anl)

of background poses an ultimate constraint on sensitivity.

where K is the association constant of the ligand-binding

OCCUPIED

UNOCCUPIED

ANALYIE L I G A NB DI N D E R

Figure M5.1.L. The principle of ligand-binding assays.The ligandbinding agent maybe in the soluble phase orbound to a solid support as shownhere, the advantage of the latter being the easeof separation of bound from free analyte. After exposure to analyte, the fractional occupancy of the ligand-binding sites can be determined by competitive or noncompetitive assays using labeled reagents (in orange)

Meosure unoccupied siteswilhlobeled onolyte

;

\72 ASSAY COMPETITIVE

NONCOI\4PETITIVE ASSAY

asshown.

A Light

450-550nm

o a

;

o a

o e

Anligen-{nlibody oggregote

Time>

Figure 6.17. Rate nephelometry. (a) On addition of antiserum, small antigen-antibody aggregates form (cf figure 6.24) which scatter incident light filtered to give a wavelength band of450-550 nm. For nephelometry, the light scattered at a forward angle of 70" or so is measured. (b)After additionof thesample (1) and thenthe antibody (2), the rateat which the aggregates form (3) is determined from the scatter signal.

> Anligen concentrofion

(c)The software in the instrument then computes the maximumrate of light scatter which is related to the antigen concentration as shown in reaction (d) (Copied from the operating manual for the'Array'rate automated immunonephelometer with permission from Beckman CoulterLtd )

I rze

SN D A P P L I C A I I O N S C H A P I E R6 _ I M M U N O L O G I C AM L E T H O DA

exposure to heat; apart from denaturing the protein, this also imparts a negative charge in proportion to its length. Upon introduction of the protein sample to the gel and the application of a vertical electric field from the top of the gel to the bottom, proteins are repelled from the negative pole (the cathode) and migrate towards the positive pole (the anode). Due to the molecular sieving effect of the gel matrix, proteins within the mixture become resolved into discrete zones (bands) with the smallest proteins moving furthest through the gel (figure6.18). In order to probe the electrophoretically separated protein mixture with antibody to identify the protein of

interest, it is necessaryto allow the antibody accessto the proteins within the gel. Becauseantibodies are relatively large proteins they cannot readily penetrate the gel matrix; the solution to this problem is to'blot'the gel onto a positively charged membrane that traps the charged proteins and immobilizes them on the surface of the membrane (figure 6.18).This is achieved by again applying an electric field to the gel to drive the proteins horizontally out of the gel onto the blotting membrane; polyvinylidene difluoride (PVDF) and nitrocellulosebased membranes are typically used for this purpose' The blot can then be probed with either polyclonal or monoclonal antibodies directed against the protein of

lo blol of oroteins Electrolronsler

Applyprolein lo gel somples Protein somples

Blotting Gelmembrone

Polyocrylomide gelbelween glossplotes

wilhonlibody Bloiis probed forAgof inlerest specific wilh bydelection followed conjugote dlg-HRP

175 82 63 41

325

analysis. (a) Denatured Figure 5.18. Principle of immunoblot protein mixtures can be separated on the basis of their relative mobilities through a ge1matrix that is formed by polymerizing acrylamide and bis-acrylamide, to form polyacrylamide, between closely spaced (1 0-1 5 mm) glass plates. Prior to loading on the gel, proteins are first denatured by heating in a sample buffer containing SDS followed by introductionto the samplewellsof the geland applicationof anelectric current for 2-3 hours Separated proteins are then electrotransfered

onto blotting membranes composed of PVDF or nitrocellulose which canthenbeprobedwithantibody, followedbydetectionof bound antibody using anti-Ig coniugated to horseradish peroxidase or similar (b) SDS-PAGE gel of various cell iysates (ianes 1-5), purified proteins (lanes 6-12) and molecular weight markers (lane 13), stained with Coomassie Blue to reveal all proteins ran on the gel (Data kindly provided by SeanCullen.)

AND APPTICATIONS CHAPTER 6 - I M M U N O T O G I C A TM E T H O D S

<- Fulllength rR,

.ii*ri

tFilp*,ffi

<- p20 <- pll

Figure 5.19. Immunoblot analysis. Analysis of caspase-3processing by the CTL/NK protease, granzyme B (GzmB). Protei.n extracts, derived from Jurkat T-cells, were incubated rn the presence of decreasing concenhations of granzyme B, a serine protease that is delivered into target cells upon attackby CTLs or NK cells Cell lysates were then separated on an SDS-PAGE gel followed by transfer to nitrocellulose membrane and immunoblotted with antibodies against caspase-3. Note how caspase-3becomes processed (cleaved) by the higher concentrations of granzyme B; such processing activates the caspase-3 precursor within the iarget cells and promotes apoptosis (Data kindly provided by Dr ColinAdrain )

interest. Binding of antibody is detected using horseradish peroxidase-conjugated anti-Ig secondary antibodies, followed by application of a suitable enzyme substrate(figure 6.19). Obviously, such a procedure will not work with antigens which are irreversibly denatured by this detergent, and it is best to use polyclonal antisera for blotting to increase the chance of including antibodies to whichever epitopes do survive the denaturation procedure; a surprising number do.

lmmunoprecipitoli0n 0f 0nfigencomplexes Antibodies immobilized on a solid support/ such as agarose beads, can be used to purify an antigen from a complex mixture of other antigens to explore the nature of the antigen and the proteins to which itbinds (figure 6.20).To illustrate this approach, let us imagine that we have generated a monoclonal antibody againsi a cellsurface receptor, such as TLR4, that is known to play an important role in the recognition of pathogen components. We would like to immunoprecipitate (IP) the receptor in the (possibly vain!) hope that there will be a proteinhanging onto the cytoplasmic tail of the receptor that may shed some light uponhow the receptor signals deep into the bowels of the cell. To do this, we would immunoprecipitate the receptor using our lovinglyprepared anti-TLR4 monoclonal antibody immobilized on agarose beads. We would then wash away unbound material by centrifugation of the beads a couple of times in a suitable wash buffer, followed by applying the immunoprecipitated material onto an SDS-PAGE gel to see what we have bagged. In the event that only the receptor has been immunoprecipitated, we would, to our obvious dismay, see only a single band on the gel

of membrane antigen. Analysis Figure 6.20. Immunoprecipitation of membrane-bound class I MHC antigens (cf p 75). The membranes from human cells pulsed with 3sS-methionine were solubilized in a detergent, mixed with a monoclonal antibody to HLA-A and B molecules and immunoprecipitated with staphylococci An autoradiograph (A) of the precipitate run in SDIPAGE shows the HLA-A and B chains as a 43 000 molecular weight doublet (the position of a 45 000 marker is arrowed) If membrane vesicles are first digested with proteinase K before solubilization, a labeled band of molecular weight 39 000 can be detected (B) consistent with a transmembrane orientation of the HLA chain: the 4000 Da hydrophilic C-terminal fragment extends into the cytoplasm and the major portion, recognized by the monoclonal antibody and by tissue typing reagents,is Presenton the cell surface (cf. figure 4.13). (From data and autoradiographs kindly supplied by Dr M J. Owen )

along with bands corresponding to the antibody that we have used to perform the IP. Any unexPected bands are candidate receptor-interacting proteins which we can identifybypicking a sample of the proteinspotfromthe gel and subjecting this to mass spectrometry or Protein sequencing analysis. The cynics among you will guess that this is often rather more simple in theory than in practice. However, given that the sensitivity of protein identification techniques has increased in leaps and bounds in recent years, such approaches have become increasingly fruitful. Immunoprecipitation can alsobe used to testwhether protein A binds to protein B by coexpressing these proteins within the same cell, followed by immunoPrecipitation of protein A (or protein B) using a suitable antibody and running on an SDS-PAGE gel. Following transfer of the gel to a membrane support by Westem blotting, the membrane can now be probed with antibodies directed against protein B to see whether it has co-immunoprecipitated with protein A. The latter technique is undoubtedly the most widely used form of the IP method and has been employed to great effect in the study of protein-protein interactions.

I rso

CHAPTER 6 - I M M U N O L O G I C A TM E T H O D S AND APPLICAIIONS

Proleinondqntibodymicroorroys With the ready availability of cDNA copies of essentially all human genes, it is now possible to express virtually any human protein 'off-the-shelf' and to purify it to hom*ogeneity using simple molecular tricks. This is also true for protein-coding genes from several species of yeast, many bacteria, the fruitfly and other organisms. Using large-scaleprotein expressionapproaches,arrays of proteins have been produced that contain thousands of independent proteins, or protein fragments, arranged on glass slides as discrete microdots. On such arrays, each microdot contains a single protein and the identity of thisprotein is knownfromits positionwithin the array. Such arrays canbe probed with an antiserum, or monoclonal antibody, to determine the spectrum of proteins to which the antibody reacts.Thus, it may be possible in the near future to determine the full spectrum of autoantibodies (and their relative titer) present in a patient sample in a single step. One has only to cross-reference the spots 'lighting up' after incubation with antibodywith a listthat specifiesthe identity of the protein that has been placed within those particular spots (figure 6.21).Such arrays are likely to be useful for the identification of novel autoantigens, for the rapid diagnosis and classification of autoimmune conditions, and may also be useful for monitoring disease progression. Additional applications include the rapid determination of proteins that cross-react against monoclonal or polyclonal antibodies. In practice, smaller arrays focused on particular disease states are used for diagnostic purposes. Similar to protein anays, antibodies can also be arrayed as discrete spots on glass slides or other solid supports. Such arrays can be used to capture multiple antigens from the same sample simultaneously. Thus, for example, an anti-cytokine antibody array can be used to detect the presence of multiple cytokines within the same sample.

E P I T O PM EA P P I N G T-cellepilopes Where the primary sequence of the whole protein is known, the identification of T-cell epitopes is comparatively straightforward. Since these epitopes are linear in nature, multipin solid-phasesynthesiscanbe employed to generate a series of overlapping peptides, 8-9-mers for cytotoxic T-cells and usually 10-14-mers for Thelpers (figure 6.22), and their ability to react with antigen-specific T-cell lines or clones can be deciphered to characterize the active epitopes.

Proteins of knownidentiiy orroyed onloglosssupporl

+ Incubote wilhpotient onliserum

Incuboie wilhcxlg-conjugote lo detect binding of primory 0nltserum

+ o o o o o o o oo o o o o o oo o o o o o o o o o o o o o o o o o o oo o o o o o o o o o o o o o o o o o o o ooo oo o o o o o o o o oo o o o o o oa o o o o o o o o o o o oo o o o o o o o o o o o o o o o o o o o o o o o o o o o o oo o o o o o o o o o o o o o o o o oo oo o o o o o o o o o o o o o o o o oo oo o o o o o o o o o o oo o o o o o ooo o oo o o o o o o o oo o o o o o o oo o oo o o o o a o o oo o o o o o ooo o oo o o o o o o o o o o o o o o ooo o oo o o o o o o o o o o o a o o o oo oooooooooooooooooooo o o o o o o o o o o o ooo o o o o o o o o o o o o o o o o o ooo o o o o o o o o o o o o o o o o o o oo o o o o o o o o o oo oo o o o o o o o o o o o o o o o o o o o oo o o o o o oo o o o o o o o o o o o o o o o o o o oo o o o o o o o o o o o o o o o o o o ooo o o o o

Addsubslrote Proteins thotbind onlibody orereveoled

Figure 6.21. Serum profiling by protein microarray analysis. Protein arrays, consisting of thousands of proteins of known identity arrayed in a specific order, can be probed with a sample of a patient's serum to determine the range of proteins to which there are antibodies present Bound antibodies can be detected with appropriate anti-Ig secondary antibody which leads directly to the identity of the proteins within the positive spots

Dissecting out T-cell epitopes where the antigen has not been characterized is a more daunting task. Randomized peptide libraries can be produced but strategies need to be devised in order to keep these within manageable numbers. Information from the accumulated data deposited invarious databankscanbe used to identify key anchor residues and libraries constructed that maintain the relevant amino acids at these positions. Thus, a positional scanning approach employs a peptide library in which one amino acid at a particular

AND APPTICATIONS CHAPTER 6 . I M M U N O T O G I C A LM E T H O D S

PINS

MICROTITER PLATE WELLS AMINO ACID ADDED

coo

Merrifield solid-phose synthesis

Second synlhesis

AflerI 2lh synthesis cleovepeptides offpin

likely to be brought together in space to form the epitope. To the extent that small linear sequences may contribute to a discontinuous ePitoPe, the overlapping peptide strategy may provide some clues. Apotentially promising approach to this problem of mimicking the residues which constitute such epitopes (termed mimotopes by Geysen) is through the production of libraries of bacteriophages bearing all possible random hexapeptides. These are produced by ligating degenerate oligonucleotide inserts (coding for hexapeptides) to a bacteriophage coat protein in a suitable vector; appropriate expression inE. coli can provide up to 10edifferent clones. The beauty of the system is that a bacteriophage expressing a given hexapeptide on its external coat protein also bears the sequence encoding the hexapeptide in its genome (cf. p. 116).Accordingly, sequential rounds of selection, inwhich the phages react with a biotinylated monoclonal antibody and are then panned on a streptavidin plate, should isolate those bearing the peptides which mimic the epitope recoSnized by the monoclonal; nucleotide sequencing will then give the peptide struchtre. Even nonproteinaceous antigens can occasionally be mimicked using peptide libraries, one example being the use of a o-amino acidhexapePtide library to identify a mimotope forN-acetylglucosamine. Others have used a single-chain Fv (scFv) library to isolate an idiotypic mimic of a meningococcal carbohydrate.

O FA N I I B O D Y ESIIMAIION

nll(D

r3r I

Figure 6.22. Synthesis of overlapping peptide sequences for (PEPSCAN) epitope analysis.Aseries of pinswhich sitindividually in the wells of a 96-well microtiter plate each provide a site for solidphase synthesis of peptide. A sequence of such syntheses as shown in the figure provides the required nests of peptides Incorporation of a readily cleavable linkage allows the soluble peptide to be released as the synthesis is terminated.

position is kept constant and all the different amino acids are used at the other Dositions. B-cellepilopes If they are linear protein epitopes formed directly from the primary amino acid sequence, then binding of antibody to individual overlapping peptides synthesized as described above will identify them. Unfortunately, most epitopes on globular proteins recognized by antibody are discontinuous and this makes the lob rather demanding, since one cannot predict which residues are

Since antigens and antibodies are defined by their mutual interactions, they can each be used to quantify each other. Before we get down to details, it is worth 'What does serum "antibody posing the question content" mean?' If we have a solution of a monoclonal antibody, we can define its affinity and specificity with considerable confidence and, if pure and in its native conformation, we will know that the concentration of antibody is the same as that of the measurable immunoglobulin in nglml or whatever. When it comes to measuring the antibody content of an antiserum, the problem is of a different order because the immunoglobulin fraction is composed of an enormous array of molecules of varying abundance and affinity ( figwe 6.23a). An average Ku for the whole IgG can be obtained by analyzing the overall interaction with antigen as a mass action equation. But how can the antibody content of the IgG be defined in a meaningful way? The answer is of course that one would usually wish to describe antibody in practical functional terms: does a serum protect against a given infectious dose of virus, does it promote

I rsz

CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S AND APPTICATIONS

A o F6^ 6

| > (Ko)lV1 Affinilyof individuol molecules

Amounlol Abbinding required for: o

*runt /

c o

./-

| I

o 6 c

= E

SeTUm 2 E

lndividuol (4) > Aboffinily

Figure 6.23. Distribution of affinity and abundance of IgG molecules in an individual serum. (a) Distribution of affinities of IgG molecules for a given antigen in the serum of a hypothetical individual There is a great deal of low affinity antibody which wouldbe incapable of binding to antigen effectively, and much lower amounts of high affinity antibody whose skewed distribution is assumed to arise from exposure to infection (b) Relationship of affinity distribution to positivity in tests for antigen binding. Rearranging the mass action equation, for all molecules of the same affinity K, and concentration of unbound antibody [Ab,] : theamountof complexformed [AgAb] * K,tAbxl forfixed [Ag]. Starting with the lowest affinity molecules in the serum, we have charted the cumulative total of antibody bound for each antibody species up to and including the one being plotted As might be expected, the very low affinity antibodies make no contribution to the tests. Serum 2 has more low affinity antibody and virtually no high affinity, butitcan producejust enough complex to react in the sensitive agglutination test although, unlike serum 1, it forms insufficient to give a positive precipitin. Becauseof its relatively high'content' ofantibody, serum 1 canbe diluted to a much greater extent than serum 2 and yet still give positive agglutination, i.e. it has a higher titer. The precipitin testis less sensitive, requiringmore complex formation, and serum 1 cannotbe diluted muchbefore this testbecomes negative, i.e. the precipitin titer will be far lower than the agglutination titer for the same serum

effective phagocytosis of bacteria, does it permit complement-mediated bacteriolysis, does it neutralize toxins, and so on? For such purposes, very low affinity molecules would be useless because they form such inadequate amounts of complex with the antigen. At the practical level in a diagnostic laboratory, the functional tests are labor intensive and therefore expensive, and a compromise is usually sought by using immunochemical assayswhich measure a composite of medium to high affinity antibodies and their abundance. The majority of such tests usually measure the total amount of antibody binding to a given amount of antigen; this could be a modest amount of high affinity antibody or much more antibody of lower affinity, or all combinations inbetween. Seraare compared for high or 'antibody low content' either by seeing how much antibody binds to antigen at a fixed serum dilution, or testing a series of serum dilutions to seeat which level a standard amount of antibody just sufficient to give a positive result is bound. This is the so-called antibody titer. To take an example, a serum might be diluted, say, 10000 times and still just give a positive agglutination test (cf. hgwe 6.29).This titer of 1:10000 enables comparison to be made with another much'weaker'serum which has a titer of only, say, 1 : 100.Note that the titer of a given serum will vary with the sensitivity of the test, since much smaller amounts of antibody are needed to bind to antigen for a highly sensitive test, such as agglutination, than for a test of low sensitivity, such as precipitation, which requires high concentrations of antibody-antigen product (figure 6.23b). To summarize: the 'effective antibody contents' of different sera can be compared by seeing how much antibody binds to the fixed amount of test antigen, or the titer can be determined, i.e. how far the serum can be diluted before the test becomes negative. This is a compromise between abundance and affinity and for practical purposes is used as an approximate indicator of biological effectiveness.

Anligen-{nlibodyinler0clionsin solulion The clqs sicsl precipitin reaction \A/hen an antigen solution is added progressively to a potent antiserum, antigen-antibody precipitates are f ormed (fi gu r e 6.24a,b).The cross-linking of antigen and antibody gives rise to three-dimensional lattice structures, as suggested by ]ohn Marrack, which coalesce, largely through Fc-Fc interaction, to form large precipitating aggregates.As more and more antigen is added, an optimum is reached (frgure 6.24b) after which consistently less precipitate is formed. At this stage the supernatant can be shown to contain soluble complexes of

CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S AND APPTICATIONS

Precipitoting complexes

Antibody excess

Equivolence

Soluble complexes

-r? Y

Antigen excess

Monovolenf hoplen

Figwe 6,24.Diagrammatic representation complexes formed between a hypothetical tetravalent antigen ( ) and bivalent antibody ( H ) mixed in different proportions. In practice, the antigen valencies are unlikely to lie in the same plane or tobe formedby identical determinants as suggested in the figure. (a) In extreme antibody excess,the antigen valencies are saturated and the molar ratio Ab:Ag approximates to the valency of the antigen. (b) At equivalence, most of the antigen and antibody combines to form large lattices which aggregate to produce typical immune precipitates (c) In extreme antigen excess, where the two valencies of each antibody molecule become rapidly saturated, the complex AgrAb tends to predominate. (d) A monovalent hapten binds but is unable to cross-link antibody molecules

antigen (Ag) and antibody (Ab), many of composition AgnAbr, AgrAb, and AgrAb (figure 6.24c).In extreme antigen excess (figure 6.24c), ultracentrifugal analysis reveals the complexes to be mainly of the form Ag.Ab, a result directly attributable to the two combining sites (divalence) of the IgG antibody molecule (cf. electron microscopestudy, figure 3.10). Serums frequently contain up to 10% of nonprecipi-

Figure 5.25. Binding capacity of an antiserum for labeled antigen (*Ag) by precipitation of soluble complexes either: (i) by changing the solubility so that the complexes are precipitated while the uncombined Ag and Ab remain in solution, or (ii) by adding a precipitating anti-immunoglobulin antibody or staphylococcal organisms which bind immunoglobulin Fc to theproteinAon their surface; the complex can thenbe spun down The level of label (e.g. radioactivity) in the precipitate willbe a measure of antigenbinding capacity.

r33 |

tating antibodies which are effectively monovalent because of the asymmetric presence of oligosaccharide on one antigen-binding arm of the antibody molecule which stereochemically blocks the combining site. Also, frank precipitates are only observed when antigen, and particularly antibody, are present in fairly hefty concentrations. Thus, when complexes are formed which do not precipitate spontaneously, more devious methods must be applied to detect and estimate the antibody level. N onpre cipit ating antib o dies can b e detected bynephelometry The small aggregates formed when dilute solutions of antigen and antibody are mixed create a cloudiness or turbidity which can be measured by forward angle scattering of an incident light source (nephelometry). Greater sensitivity canbe obtained by using monochromatic light from a laser and by adding polyethylene glycol to the solution so that aggregatesize is increased. In practice, nephelometry is applied more to the detection of antigen than antibody (cf. figure 6.17). Compl exesf ormed by nonprecipit ating antib o dies c6nbe precipitated The relative antigen-binding capacity of an antiserum which forms soluble complexes can be estimated using radiolabeled antigen. The complex can be brought out of solution either by changing its solubility or by adding an anti-immunoglobulin reagent as in figwe6.25. Messurement of antibody ffinity As discussed in earlier chapters (cf. p.97), the binding strength of antibody for antigen is measured in terms of the association constant (Ku)or its reciprocal, the dissociation constant (K6), governing the reversible interaction between them and defined by the mass action equation atequilibrium: ,, = lAgAb complex] n' J6r""4r16r"sAbl *Ag+ Ab t

-AgAbS0LUBLE C0MPLEX

*AgAbPRECIPITATE

I r34

AND APPTICATIONS C H A P I E R6 - I M M U N O L O G I C A TM E T H O D S number of antigen-binding sites oPerative at the highest levels of antigen used. Various types of ELISA have been developed which provide a measure of antibody affinity. Inone system the antibody is allowed to firstbind to its antigen, and then a chaotropic agent such as thiocyanate is added in increasing concentration in order to disrupt the antibodybinding; the higher the affinity of the antibody, the more agent that is required to reduce the binding. Another type of ELISA for measuring affinity is the indirect competitive system devised by Friguet and associates(figure 6.26b).A constant amount of antibody is incubated with a series of antigen concentrations and the free antibody at equilibrium is assessed by

With small haptens, equilibrium dialysis can be employed to measure K., but usually one is dealing with larger antigens and other techniquesmustbe used. One approach is to add increasing amounts of radiolabeled antigen to a fixed amount of antibody, and then separate the free from bound antibody by precipitating the soluble complex as described above (e.9. by an antiimmunoglobulin). The reciprocal of the bound, i.e.complexed, antibody concentration can be plotted against the reciprocal of the free antigen concentration, so allowing the affinity constant to be calculated (figure 6.26a). For an antiserum this will give an affinity constant representing an average of the heterogeneous antibody components and a measure of the effective

plol Longmuir modified Melhod of Steword-Petly: Solution *i^'AhngTAu

Figure 6.26. Determination of affinity with large antigens. The equilibria between Ab and Ag at different concentrations are determined as follows: (a) For a polyclonal antiserum one can use the Steward-Petty modification of the Langmuir equation:

*n^Ah ^V^u

Rooid ---+pprn

o c

= c

1 / b = l / ( A b . c K " )+ 1 / A b r where Ab, = 161a1 Ab combining sites, b = bound Ab concentration, c = free Ag concentration and Ko = average affinity constant At infinite Ag concentration, all Ab sitesare bound andl /b =1/ Abt Whenhalf the Ab sites are bound, 1/c = K" (b) The method of Fdguet ef al for monoclonal antibodies. First, a calibration curve for free antibody is established by estimating the proportion binding to solid-phase antigen, bound antibody being measured by enzymeJabeled anti-Ig (ELISA: see text) Using the calibration curve, the amount of free Ab in equilibrium with Ag in solution is determined by seeing how much of the Ab binds to solid-phase Ag (the amount of solid-phase antigen is insufficient to affect the solution equilibrium materially) Combination of the Klotz and Scatchard equations gives:

plol Method of Friguet et o/,: Klolzlscotchord

Solulion AbAg=-

Ag+Ab

I Sotioptrose

nolng l

Ao/Ao-A=7+ Ko/a,. where Ao = ELISAoptical density (OD) for Ab in the absence of Ag, A = OD in the presence ofa*g concentration aowhere aois approximately 10 x concentration of Ab. The slope of the plot gives Ko. (Labeled molecules are marked with an asterisk )

curveforfreeAb Colibrolion

Slope:Ko

o o C o

1/oo

FreeAb odded >

AND APPTICATIONS CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S

l e

-l

zuv

o o E tnn

AsO

Flow Figwe 6.27. Surface plasmon resonance. (a) The principle: as antigen binds to the antibody-coated sensor chip italters the angle ofreflection (b) This signals the rates of association during the antigen pulse and dissociation. In this example, the same antigen was injected over three immobilized monoclonal antibodies The arrows point to the beginning and end of the antigen injection,which is followedbybuffer flow. Note the differences between the antibodies in the association and dis-

secondary binding to solid-phase antigen. In this way, values for Kuare not affectedby any distortion of antigen by labeling. This again stressesthe superiority of determining affinity by studying the primary reaction with antigen in the soluble state rather than conformationally altered throughbinding to a solid phase. Increasingly, affinity measurements are obtained using surface plasmon resonance. A sensor chip consisting of a monoclonal antibody coupled to dextran overlying a gold film on a glass prism will totally internally reflect light at a given angle (figure 6.27a) . Antigen present in a pulse of fluid will bind to the sensor chip and, by increasing its size, alter the angle of reflection. The system provides data on the kinetics of association and dissociation (and hence K) (figure 6.27b) and permits comparisons between monoclonal antibodies and also assessmentof subtle effectsof mutations. p0rticles Agglulin0li0nof 0nligen-cooled Whereas the cross-linking of multivalent protein antigensby antibody leads to precipitation, cross-linking of cells or large particles by antibody directed against surfaceantigens leads to agglutination. Sincemost cells are electrically charged, a reasonable number of antibody links between two cells are required before the mutual repulsion is overcome. Thus agglutination of

sociation rates (Data kindly provided by Dr R. Karlsson, Biacore AB, and reproduced from Panayotou G. (1998)Surface plasmon resonance. In Delves P.J.& Roitt I.M. (eds) Encyclopediaof lmmunology,2nd edn Academic Press, with permission.) The system can be used with antigen immobilized on the sensor chip and antibody in the fluid phase, or can be applied to any other single ligand-binding assay.

Figure 6.28. Mechanism of agglutination of antigen-coated particles by antibody crossJinking to form large macroscopic aggregates If red cells are used, several cross-links are needed to overcome the electrical charge at the cell surface. IgM is superior to IgG as an agglutinator because of its multivalent binding and because the charged cells are further apart.

cells bearing only a small number of determinants may be difficult to achieve unless special methods such as further treatment with an antiglobulin reagent are used. Similarly, the higher avidity of multivalent IgM antibody relative to IgG makes the former more effective as an agglutinating agent, molecule for molecule (figure 6.28). Agglutination reactions are used to identify bacteria and to type red cells; they have been observed with leukocytes and platelets, and even with spermatozoain certain cases of male infertility due to sperm agglu-

I rso

C H A P I E R6 - I M M U N O L O G I C A TM E T H O D S AND

tinins. Because of its sensitivity and convenience, the testhasbeen extended to the identification of antibodies to soluble antigens which have been artificially coated on to erythrocytes, latex or gelatin particles. Agglutination of IgG-coated latex is used to detect rheumatoid factors. Similar tests using antigen-coatedparticles can be carried out in U-bottom microtiter plates where the settling pattern on the bottom of the well may be observed (figure 6.29); this provides a more sensitive indicator than macroscopic clumping. Quantification of more subtle degrees of agglutination can be achieved by nephelometry or Coulter counting.

lmmunoossoy for onlibodyusingsolid-phose ontigen The principle The antibody content of a serum can be assessedby the ability to bind to antigen which has been immobilized by physical adsorption to a plastic tube or microtiter plate with multiple wells; the bound immunoglobulin may then be estimated by addition of a labeled anti-Ig raised in another species (figure 6.30). Consider, for example, the determination of DNA autoantibodies in SLE (cf. p. a1{). When a patient's serum is added to a microwell coated with antigen (in this case DNA), the autoantibodies will bind to the antigen and the remaining serum proteins can be readily washed away. Bound antibody can now be estimated by addition of 125Ilabeledpurified rabbitanti-human IgG; after rinsing out excess unbound reagent, the radioactivity of the tube will clearly be a measure of the autoantibody content of the patient's serum. The distribution of antibody in different classescan be determined by using specific antisera. Take the radioallergosorbent test (RAST) for IgE antibodies in allergic patients. The allergen (e.g.pollen extract) is covalently coupled to an immunoabsorbent,

Figure 6.30. Solid-phase immunoassay for antibody. To reduce nonspecific binding of IgG to the solid phase after adsorption of the first reagent, it is usual to add an irrelevant protein, such as dried skimmed milk powder or bovine serum albumin, to block any free sites on the plastic Note that the conformation of a protein often alters on binding

'f'

r'f

f " t..#n

."n '"r-)

tt.f]

j'*

{frlr

iltlr

ll,.r

P{r.t

llr

rj P" .1. rn lr

} O !' ...J prtrt'!J

'rpO 46

test for thyroglobulin Figure 6.29. Red cell hemagglutination autoantibodies. Thyroglobulin-coated cells were added to dilutions of patients'serums. Uncoated cells were added to a 1:10dilution of serum as a control In a positive reaction, the cells settle as a carpet over the 'VlshaPed cross-section of these bottom of the cup. Because of the cups, in negative reactions the cells fall into the base of the'V', forming a small, easiiy recognizable button The reciprocal of the highest serum dilution giving an unequivocally positive reaction is termed the titer. The titers reading from left to right are: 640,20, >5120, neg, 40, 320, neg, >5120 The control for serum no. 46 was slightly Positive and this serum should be tested again after absorption with uncoated cells.

to plastic, e g. a monoclonal antibodywhich distinguishesbetween the apo and holo forms of cytochrome c in solution combines equally well with both proteins on the solid phase Covalent coupling to carboxyderivatized plastic or capture of the antigen substrate by solid-phase antibody can sometimes lessen this effect.

AND APPTICATIONS CHAPTER 6-IMMUNOtOGICAt METHODS in this case a paper disk, which is then treated with patient's serum. The amount of specificIgEbound to the paper can now be estimated by the addition of labeled anti-IgE. Awide aariety of labels are aoailable Whilst providing extremely good sensitivity, radiolabels have a number of disadvantages,including loss of sensitivity during storage due to radioactive decay, the deterioration of the labeled reagent through radiation damage, and the precautions needed to minimize human exposure to radioactivity. Therefore, other types of label are often employed in immunoassays. ELISA (enzyme-Iinkedimmunosorbentassay).Enzymes which give a colored soluble reaction product are currently the most commonly used labels, with horseradish peroxidase (HRP) and calf intestine alkaline phosphatase (AP) being by far the most popular. Aspergillus niger glucose oxidase, soy bean urease and Escherichia coli B-galactosidase provide further alternatives. One clever ploy for amplifying the phosphatase reaction is to use nicotinamide adenine dinucleotide phosphate (NADP) as a substrate to generate NAD which now acts as a coenzyme for a second enzyme system. Otherlnbels.Enzyme-labeled streptococcalprotein G or staphylococcalprotein A will bind to IgG. Conjugation with the vitamin biotin is frequently used since this can readily be detected by its reaction with enzyme-linked avidin or streptavidin (the latter gives lower background binding), both of which bind with ferocious specificity and affinity (K = tOts*-t). Chemiluminescent systems based on the HRP-catalyzedenhanced luminol reaction, where light from the oxidized luminol substrate is intensified and the signal duration increased by the use of an enhancing reagent, provide increased sensitivity and dynamic range. Specialmention should be made of time-resolved fluorescenceassaysbased upon chelatesof rare earths such as europium 3*, although these have a more important role in antigen assays.

D E T E C I I OONFI M M U N E C O M P T EFXO R M A T I O N Many techniques for detecting circulating complexes have been described and because of variations in the size, complement-fixing ability and Ig class of different complexes, it is useful to apply more than one method. TWo fairly robust methods for general use are: 1 precipitation of complexed IgC from serum at concentrations of polyethylene glycol which do not bring down significant amounts of IgG monomer, followed

r37I

by estimation of IgG in the precipitate by single radial immunodiffusion (SRID) or laser nephelometry, and 2 binding of C3b-containing complexes to beads coated with bovine conglutinin (cf . p. 77) and estimation of the bound Ig with enzyme-labeled anti-Ig. Other techniques include: (i) estimation of the binding of 12sI-C1qto complexes by coprecipitation with polyethylene glycol, (ii) inhibition by complexes of rheumatoid factor-induced aggregation of IgG-coated particles, and (iii) detection with radiolabeled anti-Ig of serum complexes capable of binding to the C3b (and to a lesser extent the Fc) receptors on the Raji cell line. Sera from patients with immune complex diseaseoften form a cryoprecipitate when allowed to stand at 4'C. Measurement of serum C3 and its conversion product C3c is sometimes useful. Tissue-bound complexes are usually visualized by the immunofluorescent staining of biopsies with conjugated anti-immunoglobulins and anti-C3 (cf. figure 15.18).

TU EB P O P U T A I I O N S I S O T A T I OONFT E U K O C YS Because of the complexity of the interactions between cells of the immune system, it is often well-nigh impossible to sort out who is doing what to whom unless one adopts a reductionist approachbypurifying specific cell populations to study in isolation. Clearly, this approach also has its pitfalls as purified cell populations often behave differently in aitro to the way they do in aiao. However, the combination of in aitro and in aiao approaches has been very powerful and each has its place in the immunologist's armory. A number of techniques are routinely employed to enrich immune cell populations to varying degrees of purity. Most of these rely upon unique characteristics of particular cell populations ranging from their size, abillty to adhere to plastic, or expression of a particular cell surface antigen. Antibodies to particular CD markers are especially useful for isolating specific populations of leukocytes when used in conjunction with a range of clever panning methods as we shall seebelow.

Bulktechniques Separ ation b ased on phy sical p ar amet ers Separation of cells on the basis of their differential sedimentation rate, which roughly correlates with cell size, can be carried out by centrifugation through a density gradient. Cells can be increased in mass by selectively binding particles such as red cells to their surface, the most notable example being the rosettes formed when

I rse

CHAPTER AND APPLICATIONS 6 - I M M U N O L O G I C A LM E T H O D S

t:l Centrifugolion

cYToToxtcmf Plolelels Peripherol) blood | ,- mononucteor I f^lfy *" cells | (PBMCS) )

MAGilETIC BEADS ANTI-lg COATED^ I\4AGNETIC ( BEADS

\ \

ANTI-lg-RlClN CONJUGATE or COIVPLE[/ENT

Erythrocytes ondgronulocytes I

Figure 6.31. Separation of leukocytes by density gradient centrifugation. Whole blood is carefully layered onto Ficoll-Hypaque or similar medium of known density, followed by centrifugation at 800.q for 30 min This results in the sedimentation of erythrocytes and granulocytes to the bottom of the centnfuge tube A peripheral blood mononuclear cell 'buffy coat' consisting mainly of T- and Blymphocytes, NK cells and monocytes is found at the interface between the two layers.

sheep erythrocytes bind to the CD2 marker present on humanT-cells. Buoyant density is another useful parameter. Centrifugation of whole blood over isotonic FicollHypaque (sodium metrizoate) of density 1.077g/rnl leavesthe mononuclear cells (lymphocytes, monocytes and natural killer (NK) cells) floating in a band at the interface, while the erythrocytes and polymorphonuclear leukocytes,being denser,travel right down to the base of the tube (figure 6.31).Adherence to plastic surfaces largely removes phagocytic cells, while passage down nylon-wool columns greatly enriches lymphocyte populations for T-cellsat the expenseof B-cells. Separ ation exploiting biolo gical p ar ameters Actively phagocytic cells which take up small iron particles can be manipulated by a magnet deployed externally. Lymphocytes which divide in response to a polyclonal activator (see p. 778), or specific antigen, can be eliminated by allowing them to incorporate 5bromodeoxyuridine (BrdU); this renders them susceptible to the lethal effect of UV irradiation. Selectionby antibody Several methods are available for the selection of cells specifically coated with antibody, some of which are illustrated in figure 6.32. Addition of complement or anti-Ig toxin conjugates will eliminate such populations. Magnetic beads coated with anti-Ig form clusters with antibody-coated cells which can be readily separated from uncoated cells.Another useful bulk selection technique is to pan antibody-coated cells on anti-Ig adsorbed to a surface.One variation on this theme used to isolate bone marrow stem cells with anti-CD34 is to coat the cells withbiotinylated antibody and selectwith an avidin column or avidin magnetic beads.co*cktails of

ANTIBOD /-COATED n t LL

FLUORESCENT ANTI-lg J ACTIVATED FTUORESCEl{CE CELtSORIIl{G

DISHES ANTI-lgCoATED PAt{t{t1{G

Figure 6.32. Major methods for separating cells coated with a specific antibody.

antibodies coated onto beads are used in cell separation columns for the depletion of specific populations Ieading to, for example, enriched CD4+ CD45RA- or CD4+CD45RO lymphocytes. by FACS Cellselection Cells coatedwith fluorescentantibody canbe separated by fluorescence-activated cell sorting (FACS) as described in Milestone 6.2 and figure 6.33 (seemore in'Flow cytometry' , p. 123).The depth discussion under technique is relatively simple but the technology required to achieve it is highly sophisticated. Cells are typically stained with antibodies against particular cell surfacemarkers (such as CD4 or CD19) and cellsthat are positive or negative for this marker are sorted into different collection tubes by the instrument. p0pul0ti0ns Enrichment 0f 0nligen-specific Selective expansion of antigen-specific T-cells by repeated stimulation with antigen and presenting cells in culture, usually alternated with interleukin-2 (IL-2) treatment, leads to an enrichment of heterogeneous Tcells specific for different epitopes on the antigen. Such T-cell lines can be distributed in microtiter wells at a high enough dilution such that on average there is less than one cell per well; pushing the cells to proliferate

CHAPTER AND APPTICATIONS 6 - I M M U N O L O G I C A LM E T H O D S

The FACS was developed by the Herzenbergs and their colleagues to quantify the surface molecules on individual white cells by their reaction with fluorochrome-labeled monoclonal antibodies and to use the signals so generated to separatecells of defined phenotype frorn a heterogeneousmixture. In this elegant but complex machine, the fluorescent cells are made to flow obediently in a single stream past a laser beam. Quantitative rneasurementof the fluorescent signal in a suitably placed photomultiplier tube relays a signal to the cell as it emerges in a single droplet; the cell becomescharged and

r3eI

can be separated in an electric field (figure M6.2.1). Extra sophistication can be introduced by using additional lasers and fluorochromes, and both 90" and forward light scatter. This is elaborated upon in the section on flow cytofluorimetry describing how this technique can be used for quantitative multiparameter analysis of single-cell populations (cf. figure 6.14).Suffice to statethat these latest FACS machines permit the isolation of cells with a complex phenotype from a heterogeneous population with a high degree of discrimination.

SHEATH STAINED FLUID CELL SUSPENSION V v

ffi

CHARGESIGNAL_-+ l

FLUORESCENCE DETECTORS LIGHT SCATTER DETECTORS

Figure M5.2.1. The principle of the FACS for flow cytofluorimetry of the fluorescence on stained cells (green rimmed circles) and physical separation from unstained cells. The charge signal can be activated to separate cells of high from low fluorescence and, using light scatter, of large from small size and dead from living.

with antigen or anti-CD3 produces single T-cell clones which can be maintained with much obsessionalcare and attention, but my goodness they can be a pain! Potentially immortal T-cell hybridomas, similar in principle to B-cell hybridomas, can be established by fusing cell lines with a T-tumor line and cloning. Animals populated essentially by a single T-cell specificity can be produced by introducing the T-cell receptor aand Bgenesfrom a T-cell clone, as a transgene (seebelow); sincethe genesare already rearranged,their presence in every developing T-cell will switch off any other VB gene recombinations. No one has succeeded in cloning primary B-cells as they die rapidly upon introduction to cell culture. It is possible however to culture immortalized B-cell hybridomas or Epstein-Barr virus-transformed cell lines, and, as with T-cells, transgenic animals expressing the same antibody in all of their B-cells havebeen generated.

ANATYSIS G E N EE X P R E S S I O The analysis of gene expression patterns can tell us a lot about what a cell or cell population is doing, or about to do, at a particular moment in time. To analyze the cohort of genes that are expressed by a cell population, either at a steady-state level or in resPonse to a particular stimulus, messenger RNA (mRNA) is extracted and is analyzed by a method that enables genes of interest to be detected. mRNA can b e analyzed by Northern blot, where a single gene Probe is hybridized to the mRNA sample, or by reverse transcriptase (RT)-primed PCR where genes of interest can be amplified by initiallymaking a cDNAcopy using RT followed by gene amplification by means of specific primers that are complementary to the sequence of interest. While Northern blotting and RT-PCR can give information concerning more than one transcript, this

I r+o

C H A P I E R6 - I M M U N O T O G I C A M L E T H O DA SN D A P P T I C A T I O N S

UNSORTED

POSITIVE SORTED

NEGATIVE SORTED

& GATE

6 o z.

,i*

6 o z.

E o z

tr ]00

W' &_ l0r

102

!03

104 100

Anti-CD45R0 Relolive f luorescence

l0r

102

103

'|04

Relotive f luorescence Anti-CD45R0

* loo

lor

lo2

lo3

lo4

fluorescence Anli-CD45R0 Relolive

60 000 50 000 c

E 40 000

30 000 20 000

r0000 memory cells octivoled CD45R0

CD4sRA noivecells Figure 6.33. Separation of activated peripheral blood memory Tcells (CD4SRO positive) from naive T-cells (CD45RO negative; but positive for the CD4SRAisoform) in the FACS after staining the surface of the living cells in the cold with a fluorescent monoclonal antibody to the CD45RO (see p 207) The unsorted cells showed two peaks (a); cells with fluorescence intensity lower than the arbitrary gate were separated from those with higher intensity giving (b) negative

(CD45RA) and (c) positive (CD45RO) populations, which were each tested for their proliferative resPonse to a mixture of two anti-CD2 monoclonals (OKT11 and GT2) in the presence of 10% antigen-presenting cells (d). 3H-Thymidine was added after 3 days and the cells counted after 15h Clearly the memory cell population proliferated, whereas the naive population did not (Data kindly provided by D

usually requires significant amounts of mRNA and is relatively slow. The development of microarray technologies now permits the simultaneous measurementof expressionof thousands of genes in a single experiment. Oligonucleotides or CDNA fragments are robotically spotted onto a gene chip and cDNA generated from, for example/ T-cell mRNA is labeled and hybridized to the genes on the microarray. This provides a quantitative comparison of expression for every gene present on the chip. By accumulating such data it is possible to build up a complete picture of which genes are expressed in which cells (figure 6.34). One area in which this technology is being rapidly deployed is in the analysis of differences in gene expression between a tumor cell and its normal counterpart, thereby illuminating possible targets for therapeutic intervention.

All that glitters is not gold however and it is certainly true to say that DNA microarrays are not a solution to all our problems. Background is a troublesome feature of this type of approach and often threatens to drown out interesting data in a cacophany of experimental noise. Well-controlled experimental set-ups are a must for large-scale microarray approaches/ otherwise any gene expression differences observed could well be due to slamming the tissue culture incubator door rather than 'garbage in, garbage the intended stimulus. The term out' comes to mind in these situations.

Wallace and R. Hicks )

ASSESSMEO N IFF U N C I I O N AATC I I V I T Y cells ofphogocylic Theoclivity The major tests employed to assessneutrophil function are summarized in table 6.1.

CHAPTER 6 . I M M U N O t O G I C A t M E I H O D SA N D A P P T I C A T I O N S

r4r I

Table 6.1. Evaluation of neutrophil function.

such0s Meosure theuplokeof porlicles or by bycounting lofexor boclerio chemiluminescence lelrozolium reduction of nitroblue Meosure

fillersupconcentrolion through Movement grodient ogenlsuch0s of chemotoclic Phe formvl.Mel.Leu GerminalCenter B Cell ActivatedB Cell

B/TCell

Figure 6.34. Gene expression during lymphocyte development and activation. The data were generated from over 3 8 million measurements of gene expression made on 13 637 genes using 243 microarrays. Each experiment represents a different cell population. For example, experiment 1 utilized polyclonally activated fetal CD4+ thymic cells, whereas experiment 2 shows the same population prior to stimulation. Overexpressed or induced genes are colored red, underexpressed or repressed genes green Certain gene expression signafures become apparent in the different cell populations, indicated on the right For example, the T-cell gene expression signature includes CD2, TCR, TCR signaling molecules and many cytokines (Reproduced with permission of the authors and the publishers from Alizadeh A.A. & Staudt L M. (2000)Current Opinion in lmmunology72,219.)

lymphocyte ]esponsiveness When lymphocytes are stimulated by antigen or polyclonal activators in zsitrothey usually undergo cell division (cf. figure 2.6b) and releasecytokines. Cell division is normally assessedby the incorporation of radiolabeled 3H-thymidine or 125I-UdR(5-iododeoxyuridine) into the DNA of the dividing cells. Cell division can also be measured by incorporation of fluorescent lipophilic dyes, such as CFSE, into the plasma membrane of lymphocytes or other cells.Upon division of cells labeled in this way, the fluorescent dye is equally partitioned to

each of the daughter cells such that each daughter has onlyhalf the dye content of the parent (figure 6.35a).The decrease in membrane dye content can be measured accurately using a flow cytometer and this gives information concerning the number of cell divisions a cell has undergone since it was labeled (figure 6.35b). This method is especially useful when using mixed cell PoPulations where it is important to know which cell type is dividing; by membrane labeling of purified cells, folIowedby adding thesecellsinto a mixed cellpopulation or even injecting these into an animal, it is possible to trackthe number of cell divisions the labeled cells subsequently undergoby measuring their dye content. Cytokines released into the culture medium can be measured by immunoassay orby abioassay using a cell line dependent on a particular cytokine for its growth and survival. Individual cells synthesizing cytokines can be enumerated in the flow cytometer by permeabilizing and staining intracellularly with labeled antibody; alternatively the ELISPOT technique (seebelow) can be applied. As usual, molecular biology has a valuable, if more sophisticated, input since T-cells transfected with anIL-2enhancer-lacz construct will switch onIncZ B-galactosidaseexpression on activation of the IL-2 cytokine response(cf.p.175) and this canbe readily revealed with a fluorescent or chromogenic enzyme substrate. The ability of cytotoxic T-cells to kill their cell targets extracellularly is usually evaluated by a_chromium 5tCr and the release assay.Target cells are labeled with releaseof radioactive protein into the medium over and above that seen in the controls is the index of cytotoxicity. The test is repeated at different ratios of effector to target cells.A similar technique is used to measure extracellular killing of antibody-coated or uncoated targets by NK cells. Now a word of caution regarding the interpretation of In aitro assays.Since one can manipulate the

l'oz

CHAPTER 6 - I M M U N O L O G I C A TM E T H O D S AND APPTICAIIONS

Figure 6.35. Analysis of cell proliferation by CFSE-labeling. Lymphocytes, or other celis with proliferative potential, canbe labeled with the fluorescent lipophilic dye, CFSE, and subsequentiy analyzed for partitioning of fluorescent dye into daughter cells. (a) Schematic depiction of a CFSE-antilabeling experiment and corresponding flow cytometry plots. (b) Human peripheral Tcells were labeled with CFSE and stimulated withplate-coated anti-CD3 monoclonal antibody for 4 days. Left panel: no stimulation; right panel: anti-CD3 stimulation. Numbers and bars on the top ofeach histogram refer to respective division peaks with the peak of undivided ce1lsto the extreme right in each histogram (Courtesy of DrAntioneAttinger )

culture conditions within wide limits, it is possible to achieve a result that might not be attainable in aiao.Let us illustrate this point by reference to cytotoxicity for murine cells infected with lymphocytic choriomeningitis virus (LCMV) or vesicular stomatitis virus (VSV). The most sensitive in aitro technique proved to be chromium release from target cells after secondary stimulation of the lymphocytes. However, this needs 5 days, during which time a relatively small number of memory CDS cytotoxic T-cell precursors can replicate and surpass the threshold required to produce a measurable assay.Nonetheless, a weak cytotoxicity assay under these conditions was not reflected by any of the lr oioo assessments of antiviral function implying that they had no biological relevance. Apoplosis Programed cell death occurs frequently in the immune system and is particularly important for the resolution of immune responses.Antigen-driven clonal expansion of T- and B-cells is typically followed by death of many

of these cells within a relatively short period, with the remaining cells making up the memory cell population; interference with this cell elimination process can result in accumulation of lymphocytes that may break tolerance and result in autoimmunity. The Fas (CD95) receptor plays an important role in peripheral tolerance and homeostatic control of lymphocyte cell populations; inactivation of this membrane receptor protein, or its ligand, in the mouse results in severe enlargement of the spleen and lymph nodes due to accumulation of lymphocytes that would normally have been eliminated through Fas-dependentapoptosis (figure 6.36).Engagement of the Fas receptor on activated lymphocytes normally results in rapid induction of apoptosis in these cells (figure 6.7). Cytotoxic T-cells also eliminate target cells by inducing apoptosis through a variety of strategies.Apoptosis is also important in shaping the T- and Bcell repertoires; negative selection of both lymphocyte populations involves triggering apoptosis. A variety of approachescan be used to measure apoptosis, ranging from morphological assessment (figure 6.7) or by exploiting biochemical alterations to the cell

r43 |

AND APPLICATIONS CHAPTER 6 - I M M U N O T O G I C A TM E T H O D S that occurs during this process. One of the most widely used assays for apoptosis takes advantage of the fact that phosphatidylserine (PS), a phospholipid that is normally confined to the inner leaflet of the plasma membrane, becomes exposed on the outer leaflet during apoptosis. This can be readily detected using fluorescently labeled annexin V, a PS-binding protein; apoptotic cells display markedly enhanced binding of arnrexin V relative to healthy cells (figure 6.37). Other assays take advantage of the fact that extensive DNA fragmentation is also a common feature of apoptosis and this can be assessedby agarose gel electrophoresis of DNAextracted from apoptotic cells or the TUNEL (TdT-mediated dUTP (deoxyuridine triphosphate) nick end /abeling) assay; the latter assay utilizes the enzyme terminal deoxynucleotidyl transferase (TdT) to add biotinylated nucleotides to the 3' ends of DNA fragLympn nodes

Spleen

Wild-type

Fosknockout

Figure 6.35. Gross enlargement of spleen and lymph nodes from Fas 'knockout'mice. Lymph nodes and spleen from wild type versus Fas knockout mice are compared Both organs are increased approximately 20-fold in size in the knockout due to accumulation of excessTand B-cells due to a failure of peripheral deletion in these animals (Kindly provided by Professor Shigekazu Nagata and adapted from Adachi et aI, 7995N ature Genetics11, 294, with permission.)

ments and this can then be detected using fluorescently labeled streptavidin. Several members of the caspase family of cysteine proteases become activated during apoptosis and this can be assessed by immunoblot analysis (figure 6.19) orby using labeled syrthetic substrate peptides that can be cleaved by active caspases.

frequency Precursor The magnitude of lymphocyte responses in culture is closely related to the number of antigen-specific lymphocytes capable of responding. Because of the clonality of the responses, it is possible to estimate the frequency of these antigen-specific precursors by limiting dilution analysis. In essence,the method depends upon the fact that, if one takes several replicate aliquots of a given cell suspension which would be expected to contain on aaerageone precursor per aliquot, then Poisson distribution analysis shows that 37"h of the aliquots will contain r?oprecursor cells (through the randomness of the sampling). Thus, if aliquots are made from a series of dilutions of a cell suspension and incubated under conditions which allow the precursors to mature and be recognized through some amplification scheme, the dilution at which 37% of the aliquots give negative responseswillbe known to contain an average of one precursor cell per aliquot, and one can therefore calculate the precursor frequency in the original cell suspension. An example is shown in some detail in figure 6.38. It has been argued that limiting dilution analysis often underestimates the true precursor frequency. This is likelybecause cells generally do not survive verywell when cultured in isolation (i.e. as a single cell per well) because most cells, with few exceptions, require signals from other cells to survive. Martin Raff showed that in the absenceof such signals cells typicallyundergo apoptosis. An accurate measure of the percentage of lyrnphocytes bearing a specific antigen receptor can be obtained by flow cytometry of cells stained with labeled antigen.

AnnexinV FITC Figure 6.37. Analysis of apoptosis by Annexin V-labeling. Phosphatidylserine (PS)is externalized on the outer leaflet of the plasma membrane during apoptosis and this canbe readily detected using the PSbinding protein, Annexin V (a) Untreated human T-lymphoblastoid cells and (b) apoptotic TJymphoblastoid cells were stained with FlTC-conjugated annexin V. (Data kindly provided by Dr Gabriela Brumatti )

Annexin V FITC

100

l0r

102 F L ]- H

]03

lo4

100

lot

lo2 F L ]- H

lo3

104

C H A P I ER 6 - | M M UN O t O G l C A t M E T H O D SA N D A P P U , C A I I O H S ,

70 60 50 40 30 20 t0 0

No of responder cellsperwellxl04020105

No of responder cellsperwellx I 1234

No of responder cellsperwellx I 0' 125 0,625 0 3125

60 2 . 5 50 40 30 20 t0 0 t0

:-L

Figure 6.38. Limiting dilution analysis of cytotoxic T-cell precursor frequency in spleen cells from a BALB/c mouse stimulated with irradtated C57BL/ 6 spleen cells as antigen BALB/c splenic responder cells were set up in 24 replicates at each concentration tested together with antigen and an excessof T-helper factors The generation of cytotoxicity in each well was looked for by addrng 5lCr-labeled tumor cells (EL-4) of the C57BL/ 6 haplotype; cytotoxicity was then revealed by measuring the release of soluble slcr-labeled intracellular material into the medium. (a) The points show the percentage of specific lysis of individual wells The dashed line indicates three standard deviations

In the case of B-cells this is fairly straightforward given that their antigen receptors recognize native antigen. However, it is only recently that technical finesse, in the form of peptide-MHC tetramers, has brought this technique to T-cells (figure 6.39). This approach overcomes the problem of the relatively weak intrinsic affinity of TCR for peptide-MHC by presenting a tagged peptideMHC as a multivalent tetramer, thereby exploiting the bonus effect of multivalency (cf. p. 94). Peptide-MHc complexes are produced by permitting recombinant MHC molecules to refold with the appropriate synthetic peptide. The recombinant MHC molecules are biotinylated on a special carboxy-terminal extension, which ensures that the biotin is incorporated at a distance from the site to which the TCRbinds, and mixed with fluorescently labeled streptavidin, which not only binds biotin with a very high affinity but also has a valency of four with respect to the biotin-hence the formation of tetramers. Numerous adaptations of this technology are appearing. For example, incubationof tetramersbound to their cognateTCRleads to internalization at 37"C;by tagging them with a toxin individual TJymphocytes of a single specificity canbe eliminated. Another approach is to use the FACS to directly sort stained cells into an ELISPOT microtiter plate in which cytokine secretion is measured, providing a functional analysis of the cells.

above the medium release control, and each point above that line is counted as positive for cytotoxicity. @) The data replotted in terms of the percentage of negative wells at each concentration of responder cells over the range in which the data titrated (5 x 10 3/well to 0.625 x 10 3/well). The dashed line is drawn at3T"knegative wells and this intersects the regression line to give a precursor (T.o) frequency of 1 rn 2327 responder cells. The regression line has an rz value of 1.00 in this experiment. (Reproduced with permission from Simpson E & Chandler P. (1986) In Weir D.M. (ed,) Handbook of Experimental lmmunology, figure 68.2 Blackwell Scientific Publications, Oxford )

Highovidity of peplide-NilHC Telr0mer Figure 6.39. Peptide-MHC tetramer. A single fluorochromelabeled peptide--MHC complex (top right) has only a low affinity for the TCR and therefore provides a very insensitive probe for its cognate receptor. However, by biotinylatint 1.; the MHC molecules and then mixing them with streptavidin, which has a valency of four with respect to biotin binding, a tetrameric complex is formed which has a much higher functional affinity (avidity) when used as a probe for the specific TCRs on the T-cell surface.

CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S AND APPLICATIONS

clear of red cells will be seen around each antibodyforming cell (figure 6.40).Direct plaques obtained in this way largely reveal IgM producers since this antibody has ahighhemolytic efficiency. To demonstrate IgG synthesizing cells itis necessaryto increase the complement binding of the erythrocyte-IgG antibody complex by 'indirect plaques' adding a rabbit anti-IgG serum; the thus developed can be used to enumerate cells making antibodies in different immunoglobulin subclasses, provided that the appropriate rabbit antisera are available. The method can be extended by coating an antigen such as Pneumococcuspolysaccharide on to the red cell, or by coupling hapten groups to the erythrocyte surface. In the ELISPOT modification, the antibody-forming cell suspension is incubated in microtiter wells containing filters coated with antigen. The secreted antibody is captured locally and is visualized, after removal of the cells, by treatment with enzymelabeled anti-Ig and development of the color reaction with the substrate. The macroscopic spots can be readily enumerated (figure 6.41).

Enumerolion 0f 0nlibody-forming cells The immun oflu orescencesan dwich test This is a double-layer procedure designed to visualize specific intracellular antibody. If, for example, we wished to see how many cells in a preparation of lymphoid tissue were synthesizing antibody to Pneumococcus polysaccharide, we would first fix the cells with ethanol to prevent the antibody being washed away during the test, and then treat with a solution of the polysaccharide antigen. After washing, a fluoresceinlabeled antibody to the polysaccharide would then be added to locate those cellswhichhad specificallybound the antigen. The name of the test derives from the fact that antigen is sandwiched between the antibody present in the cell substrate and that added as the second layer (figure 6.8c). Plaque techniques Antibody-secreting cells can be counted by diluting them in an environment in which the antibody formed by each individual cell produces a readily observable effect. In one technique, developed from the original method of jerne and Nordin, the cells from an animal immunized with sheep erythrocytes are suspended together with an excessof sheep red cells and complement within a shallow chamber formed between two microscope slides. On incubation, the antibodyforming cells release their immunoglobulin which coats the surrounding erythrocytes. The complement will then cause lysis of the coated cells and a plaque

Radiation chimerqs The entire populations of lymphocytes and polymorphs can be inactivated by appropriate doses of X-irradiation. Animals ablated in such a way maybe reconstituted by injection of bone marrow hematopoietic stem cells which provide the precursors of all the formed elements of the blood (cf. figure 11.1).Thesechimeras of hostplus

t *r],

rt t -'/ - r ( ) r - - \ O

C a

Y t

vl*

\--/

(--j

()

r'-r (-')

&'--/ & Secreled onlibody cootssurrounding redcells

Anolysisof funcfionoloclivityby cellul0rreconslilulion

COMPLE[/ENT

r45 |

('-) &

r--r

lo CooledeMhrocytes orelysedbycomplemenl formoloouewiihontibodvjorminq cellol center

Figure 6.40. jeme plaque technique for enumerating antibodyforming cells (Cunningham rnodification). (a) The direct technique for cells synthesizing IgM hemolysin is shown. The lndlrecf technique forvisualizingcells producinglgGhemolysinsrequiresthe additionof anti-IgG to the system. The differencebetween theplaques obtained by direct and indirect methods gives the number of 'IgG' plaques. The reaerseplaque assay enumerates total Ig-producing ce1lsby capturing

secreted Ig on red cells coated with anti-Ig Multiple plaque assayscan be carried out by a modification using microtiter plates. (b) Photograph of plaques which show as circular dark areas (some of which are arrowed) under dark-ground illumination They vary in size dependingupon the antibodyaffinityand the rate of secretionbythe antibodyforming ce1l.(Courtesy of C. Shapland, P Hutchings and Professor D Male.)

I r+o

AND APPTICATIONS CHAPTER 6 - I M M U N O L O G I C A TM E T H O D S

Secretion of ontiboov

Secreted onlibodybindsto dish

Locolize secreled ontibody with phosphotose-onti-lg \l/

A A^ V

Addilion0f subslrote formslorge 'spof reoction of Droduct \t/

\l/

Figure 6.41. ELISPOT (frorn ELISA spot) system for enumerating antibody-forming cells. The picture shows spots formed by hybridoma cells making autoantibodies to thyroglobulin revealed by alkaline phosphatase-linked anti-Ig (courtesy of P Hutchings) Increasing numbers of hybridoma cells were added to the top two and bottom left-hand wells which show corresponding increases in the 'ELISPOTs' number of The bottom right-hand well is a control using a hybridoma of irrelevant specificity.

hematopoietic grafted cellscan be manipulated in many ways to analyze cellular function, such as the role of the thymus in the maturation of T-lymphocytes from bone marrow stem cells (figure 6.42).

Figure 6.42. Maturation of bone matrow stem cells under the influence of the thymus to become immunocompetent lyrnphocytes capable of cell-mediated immune reactions. X-irradiation (X) destroys the ability of host lymphocytes to mount a cellular immune response, but the stem cells in inlected bone marrow can become immunocompetent and restore the response (1) unless the thymus is removed (2), in which caseonly already immunocompetent lymphocytes are effective (3) Incidentally, the bone marrow stem cells also restore the levels of other formed elements of the blood (red cells, platelets, neutrophils, monocytes) which otherwise fall dramatically after X-irradiation, and such therapy is crucial in cases where accidental or therapeutic exposure to X-rays or other antimitotic agents seriously damages the hematopoietic cells.

Mice with seaerecombined immunodeficiency (SCID) Mice with defects in the genes encoding the IL-2 recePtor y chain, the nucleotide salvage Pathway enzymes adenosine deaminase or purine nucleoside phosPhoryIase,or the RAG enzymes, develop SCID due to a failure of B- and T-cells to differentiate. These special animals can be reconstituted with various human lymphoid tissues and their functions and responses analyzed. Coimplantation of contiguous fragments of human fetal liver (hematopoietic stem cells) and thymus allows Tlymphopoiesis, production of B-cells and maintenance of colony-forming units of myeloid and erythroid lineages for 6-72 months. Adult peripheral blood cells injected into the peritoneal cavity of SCID mice treated with growth hormone can sustain the production of human B-cellsand antibodies and canbe used to generate human hybridomas making defined monoclonal antibodies. Immunotherapeutic antihlmor resPonses can also be played within theseanimals. Cellul ar inter acti o ns in a itr o It is obvious that the methods outlined earlier for depletion, enrichment and isolation of individual cell populations enable the investigator to study cellular interactions through judicious recombinations. These interactions are usually more effective when the cells are operating within some sort of stromal network resembling the set-up of the tissues where their function is optimally expressed. For example, colonization of murine fetal thymus rudiments in culture with T-cell

AND APPLICATIONS CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S precursorsenablesone to follow the pattern ofproliferation, maturation, TCR rearrangement and positive and negative selection normally seenin aiuo (cf . pp. 233-234). An even more refined system involves the addition of selected lymphoid populations to disaggregated stromal cells derived from fetal thymic lobes depleted of endogenous lymphoid cells with deoxyguanosine. The cells can be spun into a pellet and cocultured in hanging drops; on transfer to normal organ culture conditions after a few hours, reaggregation to intact lobes takes place quite magically and the various differentiation and maturation processesthen unfold.

E N G I N E E R IO NG GENEIIC FC E t t S Insenion ondmodificolion ofgenes inmommolion cells Becausegene transfer into primary (i.e. untransformed) mammalian cells is inefficient, it is customary to use immortal cell lines for such transfections and to include a selectable marker such as neomycin resistance. Genes can be introduced into cells using bacterial plasmid vectors; however, because cells do not readily take up free DNA, methods to improve the rate of uptake have been developed. Increased uptake can be achieved through precipitating plasmid DNA using calcium phosphate or by electroporation where an electric current is used to open transient pores in the plasma membrane. Another approach is to incorporate the plasmid into liposomes which fuse with the cell membrane. Direct microinjection of DNA is also effective but is labor intensive and requires specialized equipment. Integration of the gene into the genome of avirus such as vaccinia provides an easy ride into the cell, although more stable long-term transfections are obtained with modified retroviral vectors. One of the latest fads is transfection by biolistics , the buzz word for biological ballistics. DNA coated on to gold microparticles is literally fired from a high-pressure helium gun and penetrates the cells; even plant cells with their cellulose coats are easy meat for this technology. Skin and surgically exposed tissuescan also be penetrated with ease. Studying the effect of adding a gene, then, does not offer too many technological problems. How does one assessthe impact of remoaing a gene? One versatile strategy to delete endogenous gene function is to target the gene's mRNA as distinct from the gene itself. Nucleotide sequences complementary to the mRNA of the target gene are introduced into the cell, usually in a form which allows them to replicate. The antisense molecules so produced base pair with the target mRNA and block translation into protein. Although antisense RNA approaches showed early promise, this approach has

147 |

been largely superseded by a recent innovation called RNA interf erence (RNAi). 'knock down' expression of RNAi can be used to particular target genes within a cell by introducing a double-stranded (ds) RNAmolecule hom*ologous to the target gene. This method takes advantage of a natural antiviral system that selectively targets mRNA when it is detected in double-stranded form in the cell; normally dsRNA spells trouble, as this form of RNA is rarely present in cells unless they are infected by a virus. The cellular machinery that naturally responds to dsRNA selectively degrades only mRNAs that are hom*ologous to the dsRNA molecule that initiated the response. In theory, this can be mimicked by synthesizing a dsRNA copy of the gene to be silenced and introducing this into the cell; in practice there are problems with this approach when using mammalian cells and so an alternative strategy is widely employed (figure 6.43). Short-interfering RNA (siRNA) molecules of 27-25 nucleotides, hom*ologous to the gene of interest, can be syrrthesized and these overcome some of the nonspecific effects seen with large dsRNA molecules. Because of the simplicity of the siRNA approach, genome-wide cell-based screens are underway io knockdown essen-

shortinlerfering Synlhelic duplex RNA(SiRNA)

Formolion of RNA-induced silencing complex

(Rrsc)

Recycling of RISC complex

ol Degrodotion lorgelmRNA

Figure 6.43. Gene silencing via siRNA. Synthetic short-interfering double-stranded RNAmolecules (siRNAs), complementary to a gene of interest, are introduced into cells by transfection and, in complex with proteins within the transfected cell, Iead to the formation of an RNA-induced silencing complex (RISC) which binds to mRNA molecules complementary to the introduced siRNA This results in degradation of the target mRNA and recycling of the RISC to target additiona I mRNA molecules.

I r+e

CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S AND APPLICATIONS

tially every gene in the genome and explore the consequences of this. It is important to note however that gene knockdown approaches are rarely, if ever, 100% effective and there is always the uncertainty that any observed effects could also be due to unintentional silencing of other genes along with the gene of interest. Inlroducing newgenesinlo0nimols Est abli shing' d esign er mi ce' b earing n eu)genes Female mice are induced to superovulate and are then mated. The fertilized eggs are microinjected with the gene and surgically implanted in females. Between 57o and 40'/" of the implanted oocytes develop to term and, of these, 70-25"hhave copies of the injected gene, stably integrated into their chromosomes, detectable by PCR. These'founder' transgenicanimals are mated withnontransgenic mice and pure transgenic lines are eventually established (figwr e 6.44). Expressionof the transgenecanbe directed to particular tissues if the relevant promoter is included in the construct, for example the thyroglobulin promoter will

FERTILIZED EGG

ffi

-llFigure 5.44. Production of pure strain transgenic mice by microinjectionof fertilized egg, implantation into a foster mother and subsequent inbreeding.

confine expression to the thyroid. A different approach is to switch a gene on and off at will by incorporating an inducible promoter. Thus, the metallothionine promoter will enable expression of its linked gene only if zinc is added to the drinking water given to the mice. One needs to confirm that only the desired expression is obtained as, in some situations, promoters may misbe'leaky' have leading to expression of the associated gene. Transg enes intr o du ced int o embry o nic st em cell s Embryonic stem (ES)cells can be obtained by culturing the inner cell mass of mouse blastocysts. After transfection with the appropriate gene, the transfected cells can be selected and reimplanted after injection into a new blastocyst.The resulting mice are chimeric, in that some cells carry the transgene and others do not. The same will be true of germ cells and, by breeding for germ-line transmission of the transgene, pure strains can be derived (figure 6.45). The advantage over microinjection is that the cells can be selected after transfection, and this is especially important if hom*ologous recombination is required in 'knockout order to generate mice' lacking the gene which has been targeted. In this case, a DNA sequence whichwill disrupt the reading frame of the endogenous gene is inserted into the ES cells. Becausehom*ologous recombination is a rare event compared to random integration, selectable markers are incorporated into the constructin order to transfer onlythose EScellsinwhich the endogenous gene hasbeen deleted (figure 6.46).This is a truly powerful technology and the whole biological community has been suffused with boxing fever, knocking out genes right, left and center. Just a few examples of knockout mice of interest to immunologists are listed in table 6.2. It is not a particularly rare finding to observe that knocking out a gene leads to unexpected developmental defects. Whilst this in itself can provide important information concerning the role of the gene in developmental processes,it can frustrate the original aim of the experiment. Indeed, a number of knockouts are nonviable due to embryonic lethality. Never fear, ingenuity once again triumphs, in this caseby the harnessing of viral or yeast recombinase systems. Instead of using a nonfunctional gene to create the knockout mouse, the targeting construct contains the normal form of the gene but flanked with recognition sequences (loxP sites) for a recombinase enzyme called Cre. These mice are mated with transgenic mice containing the bacteriophage P1derived Cre transgene linked to an inducible or tissuespecific promoter. The endogenous gene of interest will be deleted only when and where Cre is expressed

CHAPTER 6 - I M M U N O T O G I C A LM E T H O D S AND APPLICATIONS

r4eI

thereby creating a tissue-specific or conditional knockout (figure 6.47).TheCre/ lorP system can alsobe organized in such a way as to turn on expression of a gene by incorporating a stop sequenceflanked by loxP sites. Mice in which an endogenous gene is purposefully replaced by a functional gene, be it a modified version of the original gene or an entirely different gene, are referred to as'knocked in mice'. Hence, in the example above, knocking in a loxP flanked gene leads eventually to a knocked out gene in a selected cell type. EMBRYONIC STEMCELL CULTURE

Table 6.2. Some gene'knockouts'and their effects.

T-cells Absence of cvlofoxic Defective signoling in lhymocyles T-cells bulnolperioherol

Nobonelosswhenovoriectomized (implicotions forosieoporosis?) molor;shift t0 Leishmonio Suscepfible (decreosed fromThl 10Th2response IFNT 0ndincreosed lL-4produclion)

^L-^5Figure 5.45. Introduction of a transgene through transfection of embryonic stem cells. The transfected cells can be selected, e g for hom*ologous recombinant'knockouts', before reimplantatton

lmooired CTLondNKcellfunction

shock; Resislonl lo endoloxic lo Lisleflo susceDlible (1995)Current BiologY 5,625. fromBrondon Modified

Torgeling sequence contoining nontuncfionol gene

Plosmid

Endogenous normolgene in EScells

Chromosome

Knocked out genein EScells

Chromosome

Figure 5.46. Gene disruption by hom*ologous recombination with plasmid DNAcontaining a copyof the gene of interest (in thisexample R 4G-1) into which a sequence specifying neomycin resistance (reoR.) has been inserted in such a way as to destroy the R 4G-1 reading frame between the 5'and 3'ends of the gene. Embryonic stem (ES) cells in which the targeting sequence has been incorporated into the chromosomal DNA by hom*ologous recombination will be resistant to the

neomycin analog G418. Stem cells in which nonhom*ologous recombination into chromosomal DNAhas occurred would additionally incorporatethe thymidinekinase (tk) gene which canbe used to destroy such cells by culturing them in the presence of ganciclovir, leaving only ES cells in which hom*ologous recombination has been achieved. These are then used to create a knockout mouse.

I rso

CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S AND APPTICATIONS

Torgeting sequence

Chromosome

Deleted

Genelheropyin humons We seemtobe catching up with sciencefiction and are in the early stages of being able to correct genetic misfortune by the introduction of 'good' genes. For example, one form of severe combined immunodeficiency (SCID) is due to a mutation in the yc gene which encodes a subunit of the cytokine receptors for IL-2, -4, -7, -9, -15 and -21.Correction of this defectin children has been achieved by in oitro transfer of the normal gene into CD34+ bone marrow stem cells using a vector derived from a Moloneyretrovirus, a convincingproof of principle for human gene therapy. Major problems yet to be overcome concern both the efficiencyof delivery of replacement genes as well as fargetingof tl:.egene-deliveryvector to the desired cell population. Where it is possible to remove the target cell population and treat ex aiao the risk of mis-targeting to other tissues is diminished but not entirely eliminated. In situations where the target tissue cannot be removed for treatment, the efficiencyof gene deliverycanbe poor.

Figwe 6.47. Conditional knockout. The endogenous gene that is under study (here B7.2) is hom*ologously replaced in ES cel1s with an identical gene, as in figure 6.46, but here flanked by loxP sequences (brown boxes) and with the neoRgene incorporated in a nondisruptive manner purely for selection purposes. Nonhom*ologous recombinants will contain the fk gene and are eliminated using ganciclovir Transgenic animals are then generated from ES cells which are resistant to G418. If hom*ozygous 87.2 loxP transgenics are mated with mice which contain a transgene for the Cre recombinase under the control of specific regulatory elements, only those cells in which the promoter is active will produce the Cre enzyme necessary to delete the sequence flankedby loxP The example given would represent an experiment aimed at investigating the effect of specifically knocking out B7 2 in B-cells whilst maintaining its expression in, for example, dendritic cells.

Other risks include insertion of the replacement gene at random chromosomal sites; insertion into a tumor suppressor gene for example would be highly undesirable and may lead to tumor development. Some gene delivery vectors such as adeno-associated virus (AAV) insert atpredictable chromosomal locations and seemthe way forward in this regard. This still leaves the problem of efficient gene delivery in aiao. Viruses represent the most efficient gene delivery vehicles, being perfectly adapted to the task of invading human tissues and inserting their genomes. Thus it is not surprising that the most promising gene delivery vectors are currently assembled around modified forms of adenovirus, AAV and lentiviruses such as HIV. Ironically, the immune system turns out to be one of the biggest obstacles to efficient gene delivery due to robust immune responses agains these viral vectors. However, some viruses (such as AAV) provoke only modest or ineffective immune responses which can be exploited, in this instance at least.to ourbenefit.

CHAPTER 6-IMMUNOIOGICAI METHODS AND APPTICATIONS

I5I

Mokingonlibodieslo order . Polyclonal antisera can be generated by repeated immunization with antigen. . Polyclonal antibodies recognize a mixture of determi-

M0dul0llon0f blologicoloctlylly o Antibodies can be detected by inhibition of biological functions such as viral infectivity or bacterial growth. . Inhibition of biological function by known antibodies

nants on the antigen. . Adjuvants are required for efficient immune responses to antlgen. . Immortal hybridoma cell lines making monoclonal antibodies provide powerful immunological reagents and insights into the immune response. Applications include

helps to define the role of the antigen, be it a hormone or cytokine for example, in complex responses in aiao and in aitro. o Activation of biological ftrnction by receptor-stimulating or receptor-crosslinking antibodies can substitute for natural ligand and can be used to explore biological function inaitro or inaiao.

enumeration of lymphocyte subpopulations, cell depletion, immunoassay, cancer diagnosis and imaging, purification of antigen from complex mixtures, and recently the use of monoclonals as artificial enzymes (catalytic antibodies). . Genetically engineered human antibody fragments can be derived by expanding the V, and V. genes from unimmunized, but preferably immunized, donors and expressing them as completely randomized combinatorial libraries on the surface of bacteriophage. Phages bearing the highest affinity antibodies are selected by panning on antigens and genes can then be cloned from the isolated :l:u::jibody . Single-chain Fv (scFv) fragments encoded by linked 7" and V. genes and even single heavy chain domains can be created. o The human anti-mouse antibody (HAMA) response ls a significant obstacle to use of mouse monoclonal antibodies for therapeutic purposes. o The HAMA response against mouse monoclonal antibodies can be reduced by producing chimeric antibodies with mouse variable regions and human constant regions or, better still, using humanized antibodies in which all the mouse sequences except for the CDRs are replaced by humansequences. . Humanized antibodies are now in clinical use for the treatment of a variety of conditions such as rheumatoid arthritis and B-cell lymphoma. o tansgenic mice bearing human Ig genes can be immunized. The mice produce high affinity fully human antibodies. o Recombinant antibodies can be expressed on a large scale inplants. o Combinatorial libraries of diabodies containing the Hl and H2 V*, CDR may be used to develop new drugs. Purificolionol ontigen0nd ontibodyby offinitychromotogr0phy o Insoluble immunoabsorbents prepared by coupling antibody to Sepharose can be used to affinity-purify antigens from complex mixtures and reciprocalty to purify antibodies. o Affinity chromatography can also be used to co-purify proteins that serve as binding partners of antigens

lmmunodelecti0n0f 0nllgenin cells 0ndlissues o Antibodies can be used as highly specific probes to detect the presence of antigen in a tissue and to explore the subcellular localization of antigen. Antigens can be localized if stained by fluorescent antibodies and viewed in a fluorescence mrcroscoPe. . Fixation and permeabilization of cells permits entry of antibodies and allows intracellular antigens to be detected. o Confocal microscopy scans a very thin plane at high magnification and provides quantitative data on extremely sharp images of the antigen-containing structures which can alsobe examined in three dimensions. . Antibodies can either be labeled directly or visualized by a secondary antibody, a labeled anti-Ig. o Different fluorescent labels can be conjugated to secondary antibodies enabling simultaneous detection of several different antigens in the same cell. . Flow cytometry is a highly quantitative means of detecting fluorescence associated with immunolabeled or dyelabeled cells and thousands of cells per minute can be analyzed by such instruments. o In a flow cytometer single cells in individual droplets are interrogated by one or more lasers and quantitative data using different fluorescent labels can be logged, giving a complex phenotypic analysis of each cell in a heterogeneous mixture. In addition, forward scatter of the laser light defines cell size and 90oscatter,cell granularity. o Fluorescent antibodies or their fragments can also be used for staining intracellular antigens in permeabilized cells. Intracellular probes for pH, Ca2+,Mg2*, Na*, thiols and DNA content are also available. o Antibodies can be enzyme-labeled for histochemical definition of antigens at the light microscope level, or coupled with different-sized colloidal gold particles for ultrastructural visualization in the electron microscope. Deteclionond quonliloli0nof 0nligenby 0ntibody . Exceedingly low concentrations of antigens can be measured by immunoassay techniques which depend upon the (Continuedp 152)

I rsz

AND APPTICATIONS CHAPTER 6 - I M M U N O T O G I C A LM E T H O D S

relationship between Ag concentration and fractional occupancy of the binding antibody. Occupied sites are measured with a high speci{ic activity second antibody directed to a different epitope; altematively, unoccupied sites can be estimated by labeled Ag. o Antigens can be separated on the basis of molecular mass upon electrophoresis through polyacrylamide gels. Antigens separated in this way can be blotted onto PVDF or nitrocellulose membranes and their presence detected by probing with suitable antibodies. r Antigens and antigen-associated molecules can be immunoprecipitated using antibodies that recognize the antigen in its native form. . Higher concentrations of antigens are frequently estimated by nephelometry. r Protein microarrays, containing thousands of proteins immobilized on a solid support, can be probed with antibody for the simultaneous screening of many antigens. Similarly, antibody microarrays can be used to screen for the presence of multiple antigens in a single sample. Epitopemopping . Overlapping nests of peptides derived from the linear sequence of a protein can map T-cell epitopes and the linear elements of B-cell epitopes. Bacteriophages encoding all possible hexapeptides on their surface have provided some limited success in identifying discontinuous B-cell

mediated cytotoxicity or anti-Ig-ricin conjugates; they can be isolated by panning on solid-phase anti-Ig or by cluster formation with magnetic beads bearing anti-Ig on their surface. . Smaller numbers of cells can be fractionated by coating with a fluorescent monoclonal antibody and separating them fromnonfluorescent cells in the FACS. . Antigen-specific T-cells can be enriched as lines or clones by driving them with antigen; fusion to appropriate T-cell tumor lines yields immortal antigen-specific T-cell hybridomas.

onolysis Gene expression o mRNA expression can be analyzed by Northem blotting orRT-PCR. o A complete picture of cellular gene expression is now attainable by hybridization to microarray chips. Assessmenlof funclion0l 0clivily . Lymphocyte resPonses to antigen are monitored by proliferation and/or cytokine release' Proliferation can be measuredby uptake of 3H-thymidine orby CFSE-labeling' r Individual cells secreting cytokines can be identified by the ELISPOT technique inwhich the secretedproduct is captured by a solid-phase antibody and then stained with a

determinants.

second labeled antibody. . Extracellular killing by cytotoxic T-cells, and NK cells, slCr from can be measured by the release of radioactive

Eslimolionof onlibody . The antibody content of a polyclonal antiserum is defined entirely in operational terms by the nature of the assay

prelabeled target cells. . APoPtosis can be measured by assessment of annexin V-binding which detects the extemalization of phosphatidylserine on the outer leaflet of the plasma membrane

employed. . Nonprecipitating antibodies can be measured by laser nephelometry or by salt or anti-Ig coprecipitation with radioactive antigen. . Affinity is measured by a variety of methods including surface plasmon resonance which gives a measure of both the on- and off-rates. . Antibodies can also be detected by macroscopic agglutination of antigen-coated particles, and by one of the most important methods, ELISA, a two-stage procedure inwhich antibody bound to solid-phase antigen is detected by an enzyme-linked anti-Ig. lsol0fion of leukocylesubpopul0li0ns . Cells canbe separated on thebasis of physical characteristics such as size, buoyant density and adhesiveness. . Phagocytic cells can be separated by a magnet after taking up iron particles, and cells which divide in response to a specific stimulus, e.g. antigen, can be eliminated by ultraviolet li ght af ter incorporation of S-bromodeoxyuridine. . Antibody-coated cells can be eliminated by complement-

of dyingcells. . The precursor frequency of effector T-cells can be measured by staining the cells with peptide-MHC tetramers or by limiting dilution analysis. . Antibody-forming cells can be enumerated, either by an immunofluorescence sandwich test or by plaque techniques in which the antibody secreted by the cells causes complement-mediated lysis of adjacent red cells, or is captured by solid-phase antigen in an ELISPOT assay. . Functional activity canbe assessedby cellular reconstifution experiments in which leukocyte sets and selected l)'rnphoid tissue can be transplanted into unresponsive hosts such as X-irradiated recipients or SCID mice. Defined cell populations can also be separated and selectively recombinedinr.titro. . Antibodies can be used to probe cellular ftmction by cross-linking cell surface comPonents or by selective destruction of particular intracellular sites by laser irradiation of chromophore-conjugated specific antibodies which localize to the target area by penetrating permeabilized cells. (Continuedp 753)

AND APPLICATIONS CHAPTER 6 - I M M U N O L O G I C A TM E T H O D S

Genelic engineering olcells . Genes can be inserted into mammalian cells by transfection using calcium phosphate precipitates, electroporation, Iiposomes and microinjection. o Genes can also be taken into a cell after incorporation into vaccinia or retroviruses. o Endogenous gene function can be inhibited by antisense RNA, RNA interference, short-interfering RNA or by hom*ologous recombination with a disrupted gene. o Tiansgenic mice bearing an entirely new gene introduced into the fertilized egg by microinjection of DNA can be established as inbred lines. . Genes canbe introduced into embryonic stem cells; these modified stem cells are injected back into a blastocyst and can develop into founder mice from which pure transgenic animals can be bred. One very important application of this

F U R I H ER E A D I N G Alkan S.S. (2004) Monoclonal antibodies: the story of a discovery that revolutionized science and medicine Nature Reaiews Immunology 4,753-156. Ausubel FM, Brent R., Kingston R E., Moore D D, Seidman I G, Smith J.A. & Struhl K. (eds) Current Protocols in Molecular Biology John Wiley, New York. Bonifacino J.S., Dasso M, Harford J B , Lippincott-Schwartz J. & Yamada K M (eds) Current Protocols in Cell Biology. John Wiley, NewYork Brandtzaeg P (1998) The increasing power of immunohistochemistry and immunocytochemistry. lournal of lmmunological Methods 276,49-47 Brannigan f A. & Wilkinson A.f . (2002) Protein engineering 20 years on. Nature ReaiewsMolecular and CelI Biology 3,964-970. Carter L.L. & Swain S.L. (1997) Single cytokine analysis of cytokine production. Cu rrent Opinion inlmmunology 9,177-182 Cavazzana-Calvo S el al (2000) Cene therapy of human severe (SCID)-XI drsease Science 288, combined immunodeficiency 669-672 Chatenoud L (2003) CD3-specific antibody-induced active tolerance: from bench to bedside. Na ture Rez.tiews lmmunology 3,723-732. Chowdhury P.S.& Pastan I. (7999) Improving antibody affinity by mimicking somatic hypermutation in aituo. Nature Biotechnology 17,568-572. Coligan J.E., Bierer B E , Margulies D.H., Shevach E.M. & Strober W (eds) Current Protocolsin Immunology. John Wiley, New York. Cotter T.G. & Martin S.J.(eds) (1996)Techniques in Apoptosis:AUser's G uide. P ortland Press, London. Delves P.f. (7997) Antibody Production.J.Wlley & Sons, Chichester. Fishwild D.M. ef al. (7996) High-avidity human IgGr monoclonal antibodies from a novel strain of minilocus transgenic mice. N at ure B iotechnology 14, 845-851. Friguet B., Chafotte A.F., Djavadi-Ohaniance L. & Goldberg M E.J (1985) Measurements of the true affinity constant in solution of antigen-antibody complexes by enzymeJinked imrnunosorbent assay.lournal of lmmunological Methods 77, 305-379. George A.f , Lee L. & Pitzalis C. (2003) Isolating ligands specific for human vasculature using in vivo phage selection. Tiends in B io t echnology 5, 799--203

r53 |

technique involves the disruption of a targeted gene in the embryonic stem cell by hom*ologous recombinatioq producing'knockout' mice lacking a specific gene. Conditional knockouts employ recombinase systems such as Cre/loxP in order to control the deletion either temporally or in a tissue-specific manner. o 'Knock in' mice have a specified endogenous gene hom*ologously replaced with either a variant of that gene or an entirely different gene. . Human gene therapy promises an exciting future but has to overcome major obstacles conceming safe and effective delivery of therapeutic genes. Delivery of genes by vectors based on retroviruses or adeno-related virus is under intensive investigation. o Robust immune responses to many viral vectors reduces their utility as gene delivery vehicles.

Green L.L (1999) Antibody engineering via genetic engineering of the mouse: xenomouse strains are a vehicle for the facile generation of therapeutic human monoclonal antibodies lournal of I mmuno logical Metho ds 231, 11,-23 Huppi K., Martin S.E. & Caplen N.J. (2005) Defining and assaying RNAi in mammalian cells. Mole cular Cell 17. 1-10. Lacroix-Desmazes S et al. (7999) Catalytic activity of antibodies against factor VIII in patients with hemophilia A Nature Medicine 5,7044-1,047 Lefkovits I & Waldmann H (7999) Limiting Dilution Analysis of Cells of the lmmune System,2nd.edn Oxford University Press, Oxford Liu M. ef aI. (2004) Gene-based vaccines and immunotherapeutics. Proceedingsof the National Academy of SciencesUSA 101, Suppl. 2, 74567-74571. Malik V.S. & Lillehoj E.P. (1994) Antibody Techniques.Academic Press, London. (Laboratory rnanual of pertinent techniques for production and use of monoclonal antibodies for the nonimmunologist.) McGuinness B.T.et al. (7996)Phage diabody repertoires for selection of large numbers of bispecific antibody fragments.NatureBiotechnology 74,7749-11,54 Monroe R.l. et al. (1999) RAG2:GFP knockin mice reveal novel aspects of RAG2 expression in primary and peripheral lymphoid trssttes. I mmunit y 17, 201-272 the mouse. Seminars in Mosier D E. (ed.) (1996) Humanizing lmmunology 8,185-268. tetrameric comOgg G.S. & McMichael A J (1998) HlA-peptide plexes Current Opinion in lmmunology 10,393-396 Pinilla C et al. (L999) Exploring immunological specificity using synthetic peptide cornbinatorial libraries Current Opinion in lmmunology\1,193-202 Robinson J.P.& Babco*ck G.F. (eds) (7998) PhagocyteFunction: A Guide for Researchand Clinical Eaaluatron.Wiley-Liss, New York. SambrookJ & RussellDW (2001) Molecular Cloning: ALaboratory Manual,3rd edn. Cold Spring Harbor Laboratory Press, New York Shabat D., Rader C., List B, Lerner R.A. & Barbas C.F, III (1999) Multiple event activation of a generic prodrug triggerby antibody catalysis Proceedingsof the National Academy of SciencesUSA 96, 69254930. Staudt L.M. & Brown PO. (2000) Genomic views of the immune system. Annual Reztiewsof lmmunology 18,829-859

CHAPTER 6 - I M M U N O L O G I C A LM E T H O D S AND APPTICATIONS Storch W.B (2000) Immunofluorescence in Clinicol Immunology: A Primer and Atlas Birkhiiuser Verlag AG, Basel. Vaughan T.J et al. (1996) Human antibodies with subnanomolar affinities isolated from a large nonimmunized phage display libr ary. N nture Biotechnology 14, 309-31,4

Weir D.M. ef a/. (eds) (7996)Handbookof Experimentallmmunology,Sth edn. Blackwell Scientific Publications, Oxford. Zola H (1999) Monoclonal Antibodies Bios Scientific Publishers, Oxford.

Theonotomy of theimmune response

INTRODUCIION Immunologists from the far corners of the world who haveproduced monoclonalantibodies directed to surface molecules on B- and T-cells, macrophages, neutrophils andnaturalkiller (NK) cells, and so on, gettogether every so often to compare the specificities of their reagents in international workshops whose spirit of cooperation should be a lesson to most politicians. \4/here a cluster of monoclonals are found to react with the same polypeptide, they clearly represent a seriesof reagents defining a given marker and are labeled with a CD (cluster of differentiation) number. Currently, there are nearly 340 CD numbers assigned, with some of them having subdivisions, but those in table 7.7 are most relevant to our discussions.It is important to appreciate that the expression level of cell surfacemolecules oftenchanges ascells differ'subpopulations' entiate or become activated and that of cells exist which differentially express particular mole'presence' cules. When expressed at a low level the or 'absence' of a given CD antigen may be rather subjective, but be aware that low level expression does not necessarily imply biological irrelevance.

I H E N E E DF O RO R G A N I Z E D T Y M P H O ITDI S S U E For an effective immune response,an intricate seriesof cellular events must occur. Antigen must bind and if necessary be processed by antigen-presenting cells, which must then make contact with and activate T- and B-cells;T-helpersmust assistB-cellsand cytotoxic T-cell precursors, and there have to be mechanisms which amplify the numbers of potential effector cells by proliferation and then bring about differentiation to generate the mediators of humoral and cellular immunity. In

addition, memory cellsfor secondaryresponsesmustbe formed and the whole response controlled so that it is adequate but not excessive and is appropriate to the type of infection being dealt with. By working hard, we can isolate component cells of the immune system and persuade them to carry out a number of responsesto antigen in the test-tube, but compared with the efficacy of the overall development of immunityinthebody, our efforts still leave much to be desired. In aiao the integration of the complex cellular interactions which form the basis of the immune response takes place within the organized architecture of peripheral, or secondary, lymphoid tissue which includes the lymph nodes, spleen and unencapsulated tissue lining the respiratory, gastrointestinaland genitourinary tracts. These tissues become populated by cells of reticular origin and by macrophages and lymphocytes derived from bone marrow hematopoietic stem cells, the Tcells first differentiating into immunocompetent cells by a high-pressure training period in the thymus, the B-cells undergoing their education in the bone marrow itself (figure 7.1).In essence,the lymph nodes receive antigen either draining directly from the tissues or carriedbyMHC classII* dendritic cells,the spleenmonitors the blood and the unencapsulated lymphoid tissue is strategically integrated into mucosal surfaces of the body as a forward defensive system based on IgA secretion. The anatomical disposition of these lymphoid tissues is illustrated in figure 7.2. The lymphatics and associated lymph nodes form an impressive network, draining the viscera and the more superficial body structures before returning to the blood by way of the thoracic duct (figre7.3). Communicationbetween these tissuesand the rest of the body is maintainedby a pool of recirculating lym-

RESPONSE CHAPTE7 R- T H EA N A T O M YO F T H ET M M U N E

| 156

Table 7.1. Some of the major clusters o{ differentiation markers on human cells.

lDC,B subset T,NK T

(CD)

glycolipid Presents ondolhernonpeptide onligens lo T-cells Receplor forCD58(LFA-3) costimulolor Tronsducing elemenls ot T-cellreceplor

MHCclossll reslricted I Mo,MS,IDC

Receptor for MHCclossll

T,B subsel

Involved in ontigen receplor signoling

MHCclossI restricled T

Receplor for MHCclossI

G,MO,MO

for LPS/LBP Receplor complex

tissue. Figure 7.1. The functional organization of lymphoid Hematopoietic stem cells (SC) arising in thebone marrow differentiate into immunocompetent T- and B-cells in the primary lymphoid organs and then colonize the secondary lymphoid tissues where immune responses are organized. The mucosa-associated lymphoid tissue (MALT) together with diffuse collections of cells in the lamina propria and the lungs produces antibodies for mucosal secretions.

(medium G, NK,B, M0,rDC FcyRlll oftinilylgc receplor) B,FDC

Port0f B-cellontigen receptor complex

B B, FDC

Provides signols for B celloctivolion ondproliferotion forC3dondEpsleinCR2.Receplor Borrvirus.Portof B-cellontigen receplor complex

B, Mo,FDC

(lowoffinitylgEreceplor) FceRll

*T,*8, *Mo, (xchoin "M0 lL-2receplor T,"B Mo,M0, lDC, FDC,G,NK,B Progenilors B, M0,tDC,FDC

(87.1ond forCD80/CD86 Receptor 87.2)costimulolors (lowoffinitylgc receplor) FeyRll Adhesion molecule. Stemcellmorker Receplor forCDI54 (CD40L) costimulolor Phosphotose, celloclivolion

Resting/Noive l cel|s,B,G,Mo,NK Etfector T-cell Phosphotose, celloclivotion Mo,MQ,IDC Mo,MQ,DC FeEl(highoffiniiylgc receptor) B

lg([/lgBlronsduclng elemenls ot B-cellrecepior

*8, *l Mq,DC

87.l receptor forCD28costimulotor ondforCTLA4 inhibilory signol forCD28costimulotor 87 2 receptor ondforCTLA4 inhibitory signol Fosreceplor for FosL(CDl78). Tronsmits opoplotic signols

B,IDC,MO Wdespreod

*,octivoted; B,B-lymphocyles; FDC, folliculor dendritic cells; G,gronulocyles; lDC, interdigitoting dendritic cells; Most, most cells; M0,mocrophoges; Mo, monocytes; NK,nolurol killer cells, I T-lymphocytes. phocytes which pass from the blood into the lymph nodes, spleen and other tissues and back to the blood by the major lymphatic channels such as the thoracic duct (figures 7.4and7.12).

PRIMARYLYMPHOID ORGANS Figure7.2. The distribution throughout the body.

MUCOSA-ASSOCIATED (MALT) TISSUE LYMPHOID of major lymphoid

organs and tissues

BETWEEN ATFIC T Y M P H O C Y TI R ES ISSUES TYMPHOID This traffic of lymphocytes between the tissues, the bloodstream and the lymph nodes enables antigensensitive cells to seek the antigen and to be recruited to sites at which a response is occurring, while the dissemination of memory cells and their progeny enables a more widespread response to be organized throughout the lymphoid system. Thus, antigen-reactive cells are

CHAPTER 7 _ T H EA N A T O M YO F T H EI M M U N ER E S P O N S E

r57 |

LEFT SUBCLAVIAN VEIN THORACIC DUCT

LYIVPHATIC VESSELS

Figure 7.3. The network of lyrnph nodes and lymphatics. Lymph nodes occur at junctions of the draining lymphatics. The lymph finally collects in the thoracic duct and thence returns to the bloodstream via the left subclavian vein

depleted from the circulating pool of lymphocytes within 24 hours of antigen first localizing in the lymph nodes or spleen;several days later, after proliferation at the site of antigen localization, a peak of activated cells appears in the thoracic duct. When antigen reaches a lymph node in a primed animal, there is a dramatic fall in the output of cells in the efferent lymphatics, a phenomenon described variously as 'cell shutdown' or 'lymphocyte trapping' and which is thought to result from the antigen-induced releaseof soluble factors from T-cells (cf. the cytokines, p. 185);this is followed by an output of activated blast cells which peaks at around 80 hours.

Noivelymphocytes homelo lymphnodes Naive lymphocytes enter a lymph node through the afferent lymphatics and by guided passage across the specializedhigh-walled endothelium of the postcapillary venules (HEVs) (figure 7.5). Their destination is determined by a series of homing receptors which include members of the integrin superfamily (table 7.2), chemokine receptorsand selectins.Integrins canbind to extracellular matrix, plasma proteins and to other cell surface molecules, and they are widely involved in

Figure 7.4. Traffic and recirculation of lymphocytes through encapsulated lymphoid tissue and sites o{ inflammation. Blood-bornelymphocytes enter the tissues and lymph nodes passing through the high-walled endothelium of the postcapillary venules (HEV) and leave via the draining lymphatics The efferent lymphatics join to form the thoracic duct which returns the lymphocytes to the bloodstream In the spleen, which lacks HEVs, lymphocytes enter the lymphoid area (white pulp) from the arterioles, pass to the sinusoids of the erythroid area (red pulp) and leave by the splenic vein. Tiaffic through the mucosal irnmune system is elaborated in figure 7.12

embryogenesis, cell growth, differentiation, adhesion, motility, programed cell death and tissue maintenance. Within the immune system their complementary ligands include cell surfacevascular addressins such as highly glycosylated and sulfated sialomucins present only on the HEVs of the appropriate blood vessels, in this case peripheral lymph nodes (figurc 7.6). Chemokines presented by vascular endothelium play a key role in triggering lymphocyte arrest, the chemokine receptors on the lymphocyte being involved both in binding to their ligand and in the functional activation of integrins. Thus, naive lymphocytes, and also dendritic cells, express CCRT and are therefore directed into peripheral lymph nodes by virtue of the fact that the HEVs in the nodes have CCL19 and CCL21 (cf. table 9.3) on their luminal surface.Whilst CCL2I is produced by the endothelial cells themselves, CCL19 is secretedby local stromal cells and subsequently transferred to the HEV. The plt/plt mouse, which lacks expression of both of these chemokines, not unsurprisingly exhibits defective T-cell migration into peripheral lymph nodes. Chemokine activation of integrins occurs as a result of the chemokine signals facilitating their lateral mobility in the cell membrane and also by inducing structural changes in the integrins which results in a state of increased affinity.

occursin threesloges Tronsmigrolion St ep 1: Tethering and r olling In order for the lvmphocvte to become attached to the

I rsa

(a)

7 - I H E A N A T O M YO F T H EI M M U N ER E S P O N S E CHAPTER

(b)

Figure 7.5. Lymphocyte association with postcapillary venules. (a) High-walled endothelial venules (HEV) in rat cervical lymph nodes showing intimate association with lymphocytes (Ly) (b) Flattened capillary endothelial cell (EC) for comparison. (c) Lymphocytes adhering to HEV (scanning electron micrograph) ((a) and (b) Kindly provided by Dr Ann Ager and (c) by Dr W van Ewijk )

Fostflow (-4000pm/s)

t I

Sheorforce

I Slowllow (rolling ol -40pm/s)

C C L9 I , CCL2I

Figure 7.5. Homing and transmigration of lymphocytes into peripheral lymph nodes. Fast-moving lymphocytes are tethered (Step 1) to the vessel walls of the tissue they are being guided to enter through an interaction between specific homing receptors, such as L-selectin (.) located on the microvilli of the lymphocyte, and its peripheral node addressin (PNAd) ligands on the HEV of the vesselwall. PNAd comprises several molecules, including CD34 and CIyCAM-1, which possessfucosylated, sulfated and sialylated Lewisx structures Various chemokine receptors (.) are also present on these T- and B-cells After rolling along the surface of the endothelial cells (Step 2), activation of the lymphocyte LFA-1 integrin (.) (cf table 72) occurs (Step 3) in

I C A M - I I, C A M - 2J,A M - I

response to stimulation by chemokines For T-celis this step is mainly regulated by CCL19 and CCL21 binding to CCRT as shown, whereas for B-celis CXCL13 binding to CXCR5 provides additional signals Note that, because LFA-1 is absent from the microvilli, firm binding occurs by the body of the lymphocyte to its ligands, ICAM-1/2, on the endothelium. This process results in cell arrest and flattening (Step 4) followed by migration of the lymphocyte between adjacent endothelial cells, a process referred to as diapedesis which involves LFA-1 binding not only to ICAM-1/2but additionally to the junctional adhesion molecule-1 (JAM-1) which is presentbetween the endothelial cells (Step s)

CHAPTER 7 - T H E A N A T O M YO F T H EI M M U N ER E S P O N S E Table 7.2. The integrin superfamily. In general, the integrins are concerned with intercellular adhesion and adhesion to extracellular matrix components. Many of them are also involved in cell signal transduction They are op heterodimers selected from 18 c chains and 24 p chains which pair to form 24 different combinations. The VLA subfamily took its name from VLA-I and -2 which appeared as very late antigens (VLA) on T-cells, 2-4 weeks after in aitro activation

'very However, VLA-3, -4 and -5 belong to the same family but are not monocytes, lymphocytes, found to different extents on late' and are platelets and hematopoietic progenitors. A structure called the I (inserted) domain is present in many integrin subunits and contains themetal ion-dependentadhesion site (MIDAS) which, in thepresence of Mg2*, is involved inbinding the Arg Gly.Asp. (RGD) motif on many of the ligands essential for cell adhesion.

c,P, (VLA-1)

CD49o/CD29

Widespreod

LM,CO

crp, (VLA-2)

cD49b/CD29

Widespreod

LM,CO,CHAD,MMP-I

arp, $iLA-3)

CD49c/CD29

Widespreod

FN,LM,CO,EN

cop, (VLA-4)

cD49d/CD29

Widespreod

FN,VCAM-I,OP

c6P' 0LA-5)

CD49e/CD29

Widespreod

oup,0LA-6)

cD497CD29

Widespreod

0u0r

-tcD29

Widespreod

ce0r

-/cD29

Widespreod

VN,FN,TN,NN,OP

Widespreod

ogPr

tcD29

crroFr

-tcD29

Widespreod

0r rFr

-lcD29

Musculoskeletol

du0r

cD5l/cD29

MostIeukocytes

FN,VN

o,p, (LFA-',1)

C D Il o / C D l 8

Mostleukocytes

tcAM-1.-2,-3

dMP2(CR3[Moc-I ])

C D II b / C D I 8

N,Mo,M0

ICAM-I,C3bi,FG,FX

crp, (pl 50,95)

CDIlc/CDl8

lDC,lEL,NK Mo,M0

C3bi,LPS

do0z

cDl I d/cDtB

ICAM-3

(GPllb/lllo) cq,oB,

cD4t/cD6i

plotelefs Megokoryocytes,

THR FN,VN,FG,VWE,

%Fs

cD5t/cD6t

Widespreod

VN,FN,FG,VWF,THR,TN,OP

0o0r

cD4gf/CDl04

Epithelium, endolhelium, LM T-cells Schwonn cells,

%9s

cDsl/,

Widespreod

%Fe

cD51/-

Epithelium

FN,TN

croB,(LPAM1)

cD49d/-

B-cells I-cells.

VCAM-I,FN MAdCAM-I,

IEL

E-codherin

Neurons

FN,LM,CO

0eFr

%0s

cDSr/-

CHAD, chondroodherin; CO,coll0gen; CR3,complemenl receptor 3j EN,entoctin; glycoproleins FG,fibrinogen; FN,fibroneclinr FX,foctor X;GPllb/lllo, integrin llb ondlllo;ICA[/,intercellulor odhesion molecule; lDC,inlerdigitoting dendritic cell; lEL,introepitheliol lymphoc)4e; LFA, leukocyte funclion-ossocioted molecule; LM, potchodhesion lominin; LPAM, lymphocyte Peyer's molecule, IilO,mocroph0gel MAdCA[/,mucosoloddressincell odhesionmolecule;[/MP, motrix

r5eI

NK,noturolkillercell;NN, N, neutrophil; Mo,monocyle; metolloproteinose-; VCAM, TN,lenoscin; THR,thrombospondin; 0P, osleoponlin; nephronectin; (olthough theyorenot0ll molecule; VLA,veryloleonligen vosculor cellodhesion foctor+CDmorkers ore vonWillebrond VWF, lote!);VN,vitronectin; expressed -, yef0ssigned exploined onp I 55 noCDdesignotion

160

CHAPTE7 R- T H EA N A T O M YO F T H EI M M U N ER E S P O N S E

endothelial cell, it has to overcome the shear forces created by the blood flow. This is effected by a force of attraction between the homing receptors and their ligands on the vessel wall which operates through microvilli on the leukocyte surface (figure 7.6).After this tethering process, the lymphocyte rolls along the endothelial cell, with L-selectin and other adhesion molecules onthe lymphocytebinding to their ligands on the endothelium. The selectins generally terminate in a lectin domain (hence 'selectin'), as might be expected given the oligosaccharide nature of the ligands.

lymphocyfehomingt0 otherlissues Homing of activated and memory lymphocytes to other tissues involves a similar process but with different receptors and ligands involved. The codes for skin and gut homing are fairly well established, whilst those for lung and liver are onlypartially defined (figure 7.7). It appears that dendritic cells from the appropriate tissue play an important role in selectively imprinting the correct address code during their activation of naive T-cells. Cells concerned in mucosal immunity are imprinted to enter Peyer's patches by binding to HEVs in this location. In other casesinvolving migration into normal and inflamed tissues, the lymphocytes bind to and cross nonspecialized flatter endothelia.

Step 2: LFA-I actiaation resulting infirm adhesion This process leads to activation and recruitment of LFA-1 to the nonvillous surface of the lymphocyte. This integrin binds very strongly to ICAM-1 and -2 on the endothelial cell, the intimate contact causing the lymphocyte rolling to be arrested and a flattening of thelymphocyte.

t Y MP H N O D E S The encapsulated tissue of the lymph node contains a meshwork of reticular cells and their fibers organized into sinuses. These act as a filter for lymph draining the body tissues, and possibly bearing foreign antigens, which enters the subcapsular sinus by the afferent vessels and diffuses past the lymphocytes in the cortex to reach the macrophages of the medullary sinuses

Step 3:Diapedesis The flattened lymphocyte now uses the LFA-1 to bind to the ICAMs and junctional adhesion molecule-1 (JAM-1) on the endothelial cells to elbow its way between the endothelial cells and into the tissue in response to chemotacticsignals.

SKIN

LUNG

-1 1t1 tt--J---J--J l, llcctin

{

Itegrin.

n 0/

e l e c t i nl O , AM-l/2, M --t] ,c nctl 1l 77 , lI , 2 ? '7 M

PSGL-I, VCAI/.I tcAv-l/2

GUT

L-seleolin. (r4lirinlegrin, ECRg

r'/

,/ #-.f3, integrin, I CXCR3 V 6 CTJRS PNAd,ICA[/-I /2, CCL]9,2I, VCA[/-I

MAdEAM.I

0er25

iuAtul-r, vcA['i1. F.$elflrl in,

cxctg.16.0cr.5

Figlure7.7. Access to tissues require the correct address code. T-cells (and dendritic cells) destined for various locations carry a combination code of cell surface molecules which recognize their respective ligands on the vascular endothelium at their destination Some ligand-ligand pairs are the same irrespective of the destination tissue, such as LFA-1 binding to ICAM-1 and -2, and onp' (VLA-4) integrin binding to VCAM-I Other interactions utilize adhesion molecules that bind a number of different ligands, each expressed at different locations Thus, Lselectin recognizes PNAd (peripheral lymph node addressin) on peripheral lymph node endothelium but MAdCAM-1 (mucosal vascular addressin cell adhesion molecule-1), which is also recognized by the cr,rp,integrin, on gut endothelium. Both L- and P-selectin bind PSGL-1 (P-selectin glycoprotein ligand-1) on lung endothelium. The recognition of E-selectinby CLA (cutaneous lymphocyte antigen) directs skinbound lymphocytes to the correct locatron Chemokine receptors (cf table 9 3) recognize tissue-specific chemokines.

r6r I

C H A P T E 7R- T H EA N A T O M YO F T H EI M M U N ER E S P O N S E (figure 7.8a,b) and thence the efferent lymphatics (figures 7.4 and 7.8b). What is so striking about the organization of the lymphnode is that the T- and B-lymphocytes are very largely separated into different anatomical compartments, a processdirected to a large extent by chemokines. Lymph node stromal cells and

IDCs secreteCCL19 and CCL21 in the T-cell zone which attractsCCRT-bearingT-cells,whilst CXCL13 produced in the B-cell areas attract CXCR5-positive B-cells. T-B interaction occurs when antigen stimulation upregulates CCRT on B-cells thereby directing them into the Tcell zone.

CONNECTIVE SUBCAPSULAR AFFERENT CAPSULE SINUSTISSUE VALVEMARGINAL FLOWLYMPHATICS LYMPH SECONDARY FOLLICLE WITH MANTLE OF SMALL B-LYMPHOCYTES ANDGERMINAL CENTER

PARACORTEX AREA) O-CELL

MEDULLARY CORDS

CORTEX OUTER (B.CELL AREA)

MEDULLARY SINUSES

L, NODE L. NODE HILUM EFFERENT LYMPHATICS VEIN ARTERY B BLASTS PRIIVARY IVANTLE ZONE

Figure 7.8. Lyrnph node. (a) Human lymph node, low-powerview. (b) Diagrammatic representation of section through a whole node (c) Secondary lymphoid follicle showing germinal center surrounded by a mantle of small B-lymphocytes stained by anti-human IgD labeled with horseradish peroxidase (brown color) There are few IgD-positive cells in the center butboth areas contain IgM-positive BJymphocytes. (d) Diagram showing differentiation of B-cells during passage through different regions of an active germinal center FDC, follicular dendritic ce1l;MQ, macrophage; x, apoptotic B-cell ((a)Photographed by ProfessorPM Lydyard; (c) by Dr K.A. Maclennan.)

DARK ZONE

LIGHT ZONE BASAL

BLASTS

@ APICALLIGHTZONE

B-CELLS PLASI\4A CELLS MEIVIORY

CHAPTE7 R- T H EA N A I O M Y O F T H EI M M U N ER E S P O N S E B-cell areas The follicular aggregations of B-lymphocytes are a prominent feature of the outer cortex. In the unstimulated node they are present as spherical collections of cells termed primary follicles, but after antigenic challenge they form secondary follicles which consist of a corona or mantle of concentrically packed/ resting, small B-lymphocytes possessing both IgM and IgD on their surface surrounding a pale-staining germinal center (figure 7.8cd). This contains large, usually proliferating, B-blasts, a minority of T-cells, scattered conventional reticular macrophages containing'tingible bodies' of phagocytosed lymphocytes, and a tight network of specialized follicular dendritic cells (FDCs).The FDCs are of mesenchymal origin, are nonphagocytic and lack lysosomesbut have very elongated processes which make intimate contact with the lymphocytes. The B-cell activating factor BAFF, a TNF family member, is produced by FDCs and promotes Bcell survival in the germinal centerby inhibiting apoptosis of proliferating B-cells.Germinal centersare greatly enlarged in secondary antibody responses during which they constitute sites of B-cell maturation and the generation of B-cell memory. In the absence of antigen drive, the primary follicles are composed of a mesh of FDCs whose spacesare filled with recirculating, but resting, small B-lymphocytes. On primingwith a single dose of a T-dependent antigen (i.e. antigen for which the B-cells require cooperation from T-helper cells; cf. p.778), the FDC network can be colonized by as few as three primary B-blasts which undergo exponential growth, producing around 104 so-called centroblasts and displacing the original resting B-cells which now form the follicular mantle. These highly mitotic centroblasts, with no surface IgD (sIgD) and very little sIgM, then differentiate into light zone centrocytes which are noncycling and begin to upregulate their expression of slg. At this stage there is very extensive apoptotic cell death, giving rise to DNA fragments which are visible as 'tingible bodies' within the macrophages, the final resting place of the dead cells. The survivors undergo their final training in the apical light zone. A proportion of those which are shunted down the memory cell pathway take up residence in the mantle zone population, the remainder joining the recirculating B-cell pool. Other cells differentiate into plasmablasts with a well-defined endoplasmic reticulum, prominent Colgi apparatus and cytoplasmic Ig; these migrate to become plasma cells in the medullary cords which project between the medullary sinuses (figure 7.8b). This maturation of antibody-forming cells at a site distant from that at which antigen triggering has occurred is also seen in the spleen, where plasma cells

are found predominantly in the marginal zone. One's guess is that this movement of cells acts to prevent the generation of high local concentrations of antibody within the germinal center, so avoiding neutralization of the antigen and premature shutting off of the immune response. The remainder of the outer cortex is also essentiallya B-cell areawith scatteredT-cells. T-cell areas T-cells are mainly confined to a region referred to as the paracortex, or thymus-dependent area (figure 7.8a,b). In nodes taken from children with selective '1,4.6), T-cell deficiency (figure or from neonatally thymectomized mice, the paracortical region is seen to be virtually devoid of lymphocytes. Techniques such as intravital two photon scanning laser microscopy allow observation of lymphocyte behavior within lymphoid tissue. T-cells are seen to move rapidly and randomly within the paracortex, desperately trying to find an IDC bearing'their' antigen. Should the TCR on the T-cell recognize the cognate MHC-peptide, a stable binding occurs which is largely cemented by LFA-1 on the T-cell binding to ICAM-I on the IDC. An immunological synapse is generated and contact maintained for 3648 hours in order to fully activate the T-cel1.

SPTEEN On a fresh section of spleen, the lymphoid tissue forming the white pulp is seen as circular or elongated gray areas (figure 7.9a) within the erythrocyte-filled red pulp which consists of splenic cords lined with macrophages and venous sinusoids. As in the lymph node, T- and B-cell areas are segregated (hglre7.9b). The spleen is a very effective blood filter removing effete red and white cells and responding actively to blood-borne antigens, the more so if they are particulate. Plasmablasts and mature plasma cells are present in the marginal zone extending into the red pulp (figure 7.9c).

I H E S K I NI M M U N E SYSTEM Pathogens will first be encountered at body surfaces, either the skin or the mucosae (seebelow). The surfaces of thebody are endowed with a variety of externalbarriers against infection (cf. figure 7.2), andonly if these are breached will the cells of the immune system come into play. In a normal, noninflammed, state the epidermis is provided with resident Langerhans cells and T-cells whilstthe underlying dermis contains dendritic cells, T-

CHAPTER 7 - T H E A N A T O M YO F T H EI M M U N ER E S P O N S E

163 |

an inflammatory reaction in the skin then other cells of the immune system will fairly rapidly appear on the scene, including neutrophils, monocytes, eosinophils and plasma cells. In diseasessuch as atopic eczemathe number of leukocytesinthe skin substantially increases. Cutaneous inflammation is directed by several adhesion molecules amongst which LFA-1, arB, and crnB, integrins and cutaneous leukocyte antigen (CLA) have key roles in the recruitment of appropriate cells. The CCR4 chemokine receptor is expressedby most CLA* Tcells,with its ligand CCL|7 (TARC, fhymus and activation regulated chemokine) being presented on blood vessel walls in the skin. Another chemokine, CCL27 (CTACK, cutaneous T-cell-attracting chemokine) is expressedby keratinocytesand its receptor,CCR10,on a subpopulation of CLA*T-cells. Someof theCLA* T-cells present in the skin are CD4+CD25nFoxp3+ regulatory cells.

M U C O S AI IM . MUNIIY

Figure 7.9. Spleen. (a) Low-power view showing lymphoid white pulp (WP) and red pulp (RP). (b) Diagrammatic representation of an area of white puip surrounded by red pulp (c) High-power view o{ germinal center (GC) and lymphocyte mantle (M) surrounded by marginal zone (MZ) and red pulp (RP) Adjacent to the follicle, an arteriole (A) is surrounded by the periarteriolar lymphoid sheath (PALS) predominantly consisting of T-cells Note that the marginal zone is only present above the secondary follicle ((a) Photographed by Professor PM Lydyard; (c) by Professor I C M. Maclennan )

cells,macrophagesand mast cells.There is a continuous migration of leukocytes into the skin from the blood vessels,with thesecellslooking out for signs of infection and then returning to the circulation via the lymphatic system and lymph nodes. Should a pathogen provoke

Many pathogens will first be encountered at mucosal surfaces, for example if ingested, inhaled or sexually transmitted. The gastrointestinal, respiratory and genitourinary tracts are guarded immunologicallyby subepithelial accumulations of cells and of lymphoid tissues which are not constrained by a connective tissue capsule (figure 7.10).These may occur as diffuse collections of lymphocytes, plasma cells and phagocytes throughout the lung and the lamina propria of the intestinal wall (figure 7.10c),or as organized tissue (mucosa-associated lymphoid tissue, MALI) with well-formed follicles. In humans, the latter includes the lingual, palatine and pharyngeal tonsils, the Peyer's patches of the small intestine (figure 7.10a) and the appendix. Gut lymphoid tissue is separated from the lumen by columnar epithelium with tight junctions and a mucous layer. This epithelium is interspersed with microfold (M)-cells (figures 7.10b and 7.11);specialized antigen-transporting cells with short, irregular microvillae, and strong nonspecific esterase activity. They overlay intraepithelial lymphocytes and macrophages(figure 7.11b,c). Collectively the cells and tissues involved in mucosal immunity form an interconnected secretory system within which B-cells committed to IgA (and IgE) synthesis may circulate (figure 7.72).Itis however noteworthy that, unlike other mucosal tissues, in both the female and male reproductive tract the dominant isotype is

Igc. Peyer'spatchesform the sitefor induction of immune in thegut responses Foreignmaterial,including bacteria,is taken up by M-

Figure 7.10. Gut-associated immunity. (a) Immunofluorescence staining indicating the B-cells (with anti-CD20, green), T-cells (with anti-CD3, red) and the follicle-associated epitheliurn (FAE) (with anticytokeratrn,blue) inPeyer'spatchof humansmallintestine (b) Details from the antigen-samplingmicrofold-cell (M-cell) area. (c) Staining for IgA (green) and IgG (red) in a section of human large bowel mucosa Crypt epithelium shows selective transport of IgA Only a few scat-

tered IgG-producing cells are seenin the lamina propria, togetherwith numerous IgAplasma cells. (d) Staining for CD4 (red) and CD8 (green) T-cells in human duodenal mucosa The epithelium of the villi is blue (cytokeratin) The weak CD4 expression seen in the background is either macrophages or dendritic cells. (Reproduced from Brandtzaeg P & Pabst R. (2004) Trendsin Immunology 25,570-577 with permission frorn the publishers.)

ANTIGEN

Figure 7.11. M-cell within Peyer's patch epithelium. (a) Scanning electron micrograph of the surface of the Peyer's patch epithelium The antigen-sampling M-cell in the center is surrounded by absorptive enterocytes covered by closely packed, regular microvilli. Note the irregular and short microfolds of the M-cell (Reproduced with permission of the authors and publishers from Kato T & Owen R.L (1999) In Ogra R ef al (eds) Mucosallmmunology,2nd edn. Academic Press, San Diego ) (b) After uptake and transcellular transport by the M-cell (M), antigenis processedbymacrophages and thenceby dendritic cells which present antigen to T-cells in Peyer's patches and mesenteric lymph nodes E, enterocyte; IDC, interdigitating dendritic cell; L, lym-

phocyte; MQ, macrophage. (c) Electron photomicrograph of an M-cell (Minnucleus) withadjacentlymphocyte (Linnucleus). Notetheflanking epithelial cells are both absorptive enterocytes with a typicalbrush border In some cases,proteases on the surface of the M-cells modify the pathogen so that it can adhere and be taken up. Pathogenic Salnonella can invade and destroy M-celis, making a hole through which other bacteria can invade the underlying tissue (Lead citrate and uranyl acetate, x1600.) ((b) Based on Sminia T & Kraal G (1998) In Delves PJ. & Roitt I.M (eds) Encyclopediaof Immunology, 2nd edn, p 188.AcademicPress,London )

O F T H EI M M U N ER E S P O N S E cells and passed on to the underlying Peyer's patch antigen-presenting cells which then activate the appropriate lymphocytes. Thus, the Peyer's patches constitute the inductiae site for immune responses in the gut. After their activation is induced the lymphocytes travel via the lymph to the mesenteric lymph nodes where additional activation and proliferation may occur. A feature of dendritic cells from Peyer's pathches and mesenteric lymph nodes is that, unlike dendritic cells from peripheral lymph nodes or spleen, they express enzymes which convert vitamin D to retinoic acid. \A/hy is this relevant? -well because it turns out that retinoic acid induces T-cells to upregulate guthoming receptors. These lymphocytes thenmove via the thoracic duct into the bloodstream and finally on to the lamina propria (figure 7.72).In this responsive site they assist IgAforming B-cells which, because they are now broadly distributed, protect a wide area of the bowel with protective antibody. T- and B-cells also appear in the lymphoid tissue of the lung and in other mucosal sites guided by the interactions of specific homing receptors with appropriate HEV addressins as discussed earlier. Similarly, intranasal immunization is particularly effective at generating antibody production in the genitourinary tract.

F igtre 7.72. Circulation of lymphocytes within the mucosa-associated lymphoid system. Antigen-stimulated cells move from Peyer's patches to colonize the lamina propria and the other mucosai surfaces ( rar'a\.^ ), f orming what has been described as a common mucosal immune system.

lnt estin al ly mpho cy te s The intestinal lamina propria is home to a predominantly activated T-cell population rich in the cxnpt (LPAM-1) integrin (table7.2), the ligand for MAdCAM1 on the lamina propria postcapillary venules (figure 7.13).These T-cells bear a phenotype roughly comparable to that of peripheral blood lympho cytes:viz. >95"/oTcell receptor (TCR) up and a CD4 : CD8 ratio of 7 :3. Unwarranted immune responses may be dampened down following the secretion of IL-10 and transforming growth factor-B (TGFp) by inducible regulatory T-cells. Within the lamina propria there is also a generous sprinkling of activated B-blasts and plasma cells secreting IgA for transport by the poly-Ig receptor to the intestinal lumen (cf.p.51). Intestinal intraepithelial lymphocytes (IELs) are 'kettle of fish'. They are also mostly Tquite a different cells, in humans about 10% of which have a T6 TCR although in other speciesyDT-cells may rePresent up to 40% of the IEL T-cells. Of those bearing an op TCR, most are CD8* positive and can be divided into two populations. One-third of them possessthe conventional form of CD8, which is a heterodimer composed of a CD8 u chain and a CDS p chain. However, two-thirds of them instead express a CD8 acr hom*odimer which is almost

I roo

C H A P T E 7R- T H EA N A T O M YO F T H EI M M U N ER E S P O N S E

(a)

(b)

exclusively found only on IELs Whilst the CD8 oB TCR crBIELs are conventional T-cells restricted by classical MHC class I molecules for the recognition of foreign peptides, CD8 cxcrTCR uB IELs are efficiently generated in class I knockout mice Whether or not they are restricted by nonclassical MHC molecules (cf. p. 83) such as TL and Qa1 remains somewhat unclear.Intraepithelial lymphocytes and intraepithelial dendritic cells expresshigh levels of the cruB,integrin which binds Ecadherin on intestinal epithelial cells. There is a relatively high proportion of TCR y6 intestinal T-cellsin many species,and most of thesecells also express the CDS cxcxcharacteristic of IELs. It has been postulated that they act as a relatively primitive first line of defenseat the outer surfacesof the body.Reflectfor a moment on the fact that roughiy 1014bacteria reside in the intestinal lumen of the normal adult human. That is a pretty impressive number of 'noughts'to swallow. Yet combined with the barrier of mucins produced by goblet cells and the protective zone of secretedIgAantibodies, these collections of intestinal lymphocytes represent a crucial line of defense. Indeed, the number of IEL in the small intestine of the mouse accounts for nearly 50"1,of the total number of T-cellsin all lymphoid organs It is not only the intestine that has a localized immune system composed mostly of resident T-lymphocytes; we have already mentioned the resident T-cellsin skin and a similar set-up appearsto apply to the liver which in the human contains 101(llymphocytes, mostly long lived memory T-cellsbut also a relatively high proportion of NK and NKT cells.NonclassicalMHC antigens seem to play an important role in these specializedlocales,with the MHC class I chain-related (MIC) family members MICA and MICB (cf. p. 84) involved in the activation of

(c)

Figure 7.13. Selective expression of the mucosal vascular addressin MAdCAM-1 on endothelium involved in lymphocyte homing to gastrointestinal sites. Immunoh i s t t r l o g i tr t . r i n i n gr e v e a l st h c p r c s c n c co I MAdCAM-1 (a) onpostcapillaryvenules in the small intestinal lamina propria and (b) on HEV in Pever's patches,but its absence from (c) HEV in peripheral lymph nodes (Reproclucedwith permission from Butcher F.C ct ol (1999)Adunncesin lnttunology 72, 209 ) At least some component of intestinal trafficking appearsto operate as a subcomponcnt of a common mucosal immune system (cf figure 7 12),MAdCAM-1 being largely absent from the genitourinary tract, lung, salivary and lacrimal gland, although it is present on vascular endothelium in the mammarv gland

Figure 7.14. Plasma cells in human bone marrow. Cytospin prcparrtion stained with rhodamine (orange)for IgAheavy chain and fluorescein (green) for lambda light chain One cell is IgA 7', another lgA non-)" .rnd the third rs non-IgA I positive (Photograph kindly : u p p l i e d b y D r s B c n n e rH, i i m a n sa n d H a a i i m a n)

human y6 TCR IELs, and CDld in the presentation of glycolipids to liver NKT cells(cf.p. 105).

C A NB EA M A J O RS I I E O F B O N EM A R R O W A N T I B O DSYY N I H E S I S A few days after a secondary response, activated memory B-cellsmigrate to the bone marrow where they mature into plasma cells (figure 7.14).The bone marrow is a major source of serum Ig, contributing up to 80% of the total lg-secreting cells in the 10O-week-oldmouse. The peripheral lymphoid tissue responds rapidly to antigen, but only for a relatively short time, whereas bone marrow starts slowly and gives a long-lasting massive production of antibody to antigens which repeatedly challenge the host.

CHAPTER 7 - T H E A N A T O M YO F T H EI M M U N ER E S P O N S E

IHEENJOYMEN OIFP R I V I T E G E SD ITES Certain locationsin thebody, for examplebrain, anterior chamber of the eye and testis,are referred to asimmunologically privileged sites because antigens located within them do not provoke reactions against themselves.It has long been known, for example, that foreign corneal grafts can take up long-term residence, and a number of viruses have been expanded by repeated passagethrough animal brain. Generally, privileged sites are protected by rather strong blood-tissue barriers and low permeability to hydrophilic compounds and carrier-mediated transport systems. Functionally insignificant levels of complement reduce the threat of acute inflammatory reactions and unusually high concentrations of immunomodulators, such as IL-10 and TGFB (cf. p. 187),endow macrophageswith an immunosuppressive capacity.Immune privilege may also be maintained by Fas (CD95)-induced apoptosis of autoaggressivecells. Lesley Brent put it rather well: 'It may be supposed that it is beneficial to the organism not to turn the anterior chamber or the cornea of the eye, or the brain, into an inflammatory battle-field, for the immunological responseis sometimesmore damaging than the antigen insult that provoked it.'

I H E H A N D T I NO GFA N I I G E N Where does antigen go when it enters the body? If it penetrates the tissues, it will tend to finish up in the draining lymph nodes. Antigens which are encountered in the upper respiratory tract or intestine are trapped by local MALT, whereas antigens in the blood provoke a reaction in the spleen. Macrophages in the liver will filter bloodborne antigens and degrade them without producing an immune response since they are not strategically placed with respectto lymphoid tissue. Mocrophoges oregenerolonligen-presenting cells 'Classically',

it has always been recognized that antigens draining into lymphoid tissue are taken up by macrophages. The antigens are then partially, if not completely, broken down in the phagolysosomes; some may escapefrom the cell in a soluble form tobe takenup by other antigen-presenting cells and a fraction may reappear at the surface,either as a large fragment or as a processedpeptide associatedwith class II major histocompatibility molecules. Although resting, resident macrophagesdo not expressMHC classII, antigens are usually encountered in the context of a microbial infectious agentwhich caninduce the expressionof classIIby

167 |

its adjuvant-like properties involving molecules such as bacterial lipopolysaccharide (LPS) which activates the macrophage through TLR4. dendriticcellspresenl0nligen Interdigil0ting lo T-lymphocyles Notwithstanding the impressive ability of the mighty macrophage to present antigen, there is one function where it is deficient, namely the priming of naive Iymphocytes. Animals which have been depleted of macrophages, by selective uptake of liposomes containing the drug dichloromethylene diphosphonate, are as good as their controls with intact macrophages in responding to T-dependent antigens. We must conclude that cells other than macrophages prime T-helper cells, and it is now generally accepted that these are the interdigitating dendritic cells (IDCs). These cells, which are of bone marrow origin, have the awesome capacity to process four times their own volume of extracellular fluid in t hour, thereby facilitating antigen capture and processing in their abundant intracellular MHC classIIrich compartments (MIIC; cf. p. 97). The IDCs are the crDmede la crDmeof the antigenpresenting cells and, if pulsed with antigen before injection into animals, usually produce stunning immune responses. In this connection, it is relevant to note that large numbers of these dendritic cells can be generated from peripheral blood by cultivation with granulocyte-macrophage colony-stimulating factor (CM-CSF) (cf. p. 187) to promote proliferation and IL-4 to suppress macrophage overgrowth. Their use in immunotherapy is beginning to be explored, e.g. by pulsing autologous dendritic cells with the patient's tumor antigens and then reinjecting them to evoke an rmmune response. Precursordendritic cellsin theblood that are destined to become skin Langerhans' cells express cutaneous leukocyte antigen (CLA), directing their homing to skin via interaction with E-selectin on the relevant vascular endothelial cells just as occurs for cutaneous T-cells. The Langerhans'cells, and dendritic cells in other tissues, act as antigen sampling agents. They are only moderately phagocytic but display extremely active endocytosis. Receptors involved in antigen capture, including the mannose receptor and Fc receptors forboth IgG and IgE, are present on dendritic cells. The expression of cell surface MHC class II, and of adhesion and costimulatory molecules, is low at this early stage of the dendritic cells' life. However, as they differentiate into fully fledged antigen-presenting cells, they decrease their phagocytic and endocytic activity, show reduced levels of molecules involved in antigen capture, but dramati-

I toa

c H A p T E 7R- T H EA N A T o M vo F T H ET M M U NREE s p o N s E

cally increase their MHC class II. Costimulatory molecules such as CD40, CD80 (B7.1) and CD86 (87.2) arc also upregulated at this stage,asis the ICAM-1 adhesion molecule which is thought to contribute to both the migratory and antigen-presenting properties of these cells. Their expression of CD4 and the chemokine receptors CCRS and CXCR4 (cf. table 9.3) means that they are attracted to and migrate into T-cell areas and incidently become susceptible to infection by HIV (seep. 326). There is evidence for different populations of IDCs, although this is still a somewhat shaky area. TWo separate developmental pathways have been described, the myeloid pathway which generates CD11c+ interstitial dendritic cells and skin Langerhans'cells, and the lymphoid pathway which produces CD11c plasmacytoid dendritic cells. There appear to be a number of subpopulations within these divisions. Furthermore, in the absence of activation, the dendritic cells remain in an immature state and do not upregulate the expression of costimulatory molecules such as CD80 and CD86. Antigen presented by these'tolerogenic' dendritic cells will causeT-cell anergy or deletion, or induce regulatory T-cells to secreteimmunosuppressive cytokines such as IL-10 and TGFp. In some circ*mstances semi-mature dendritic cells can also exhibit a regulatory phenotype by secretingindoleamine 2,3-dioxygenase(IDO) which catalyzes the depletion of trytophan, in the absence of which T-cellsundergo apoptosis. It is worth noting that, unlike macrophages, which in a senseare'brutal microbe crunchers'.the dendritic cells

are not strongly phagocytic and they do need help from the macrophages in the preprocessing of particulate antigens, since these responses are completely abolished in macrophage-depleted animals. To summarize, the scenario for T-cell priming appears to be as follows. Peripheral immature dendritic cells such as the Langerhans' cells (cf. figure2.6),whichbind to skin keratinocytes through surface expression of Ecadherin, can pick up and process antigen. As maturation proceeds, they lose their E-cadherin and produce collagenase, presumably to facilitate their crossing of 'veiled' the basem*nt membrane. They then travel as cells in the lymph (figure 7.15b) before settling down as IDCs in the paracortical T-cell zone of the draining Iymph node (figure 7.1,5a).There, maturation is completed (figure 7.76), the IDC delivers the antigen with costimulatory signals for potent stimulation of naive and subsequently of activated, specific T-cells, which take advantage of the large surface area to bind to the MHC-peptide complex on the IDC membrane. We will meet IDCs again in Chapter 11 when we discuss their central role within the thymus where they present self-peptides to developing autoreactive T-cells and trigger their apoptotic execution (known more 'clonal deletion'; cf. p.2aQ. gently as

Figure 7.15. Dendritic antigen-presenting cell. (a) Interdigitating dendritic cell (IDC) in the thymus-dependent area of the rat lymph node. These antigen-presenting cells are derived from the Langerhans' cell inthe skin and interstitial dendritic cells in other tissues, and travel to the node in the afferent lymph as'veiled'cells bearing antigen on their profuse surface processes Intimate contacts are made with the surface membranes (arrows) of the surrounding TJymphocytes (TL)

(x2000). (b) Scanning electron micrograPh of a veiled cell. ln contrast with these dendritic cells whichPresent antiSen to T-cells, the follicular dendritic cells in germinal centers stimulate B-cells. ((a) Reproduced with permission of the authors and publishers from Kamperdijk E W.A., HoefsmitE.Ch.H., Drexhage H.A. & Balfour B.H. (1980)InVan Furth R. (ed.) Mononuclear Phagocytes,3rdedn Rijhoff Publishers, The

0nd Folliculor dendriliccellsbindimmunecomplexes slimuloleB-cells The FcyRII, FceRII and CR1 (CD35) and CR2 (CD21)

Hague (b) Courtesy of Dr G G. MacPherson.)

r6eI

CHAPTE7 R- T H EA N A T O M YO F T H EI M M U N ER E S P O N S E

Nonlymphoid tissue Figure 7.15. Migration and maturation of interdigitating dendritic cells. The precursors of the IDCs are derived from bone marrow stem cells They travelvia theblood to nonlymphoid tissues These immature IDCs, e g Langerhans' cells in skin, are specializedfor antigen uptake Subsequently they travel via the afferent lyrnphatics as veiled cells (cf figure 7 15b)to take up residencewithin secondary lymphoid tissues(cf. figure 7 15a) where they cxpress high ler.els of MHC classII and costimulatory moleculessuch as B7 Thesecells are highly specializedfor the actir.ation of naive T-cel1s The activated T-cell may carry out its function rn the lymph node or, af ter i m p r i n t i n gw i t h r e l e v a nht o m i n g molecules,recirculateto the appropriate tissue

complement receptors (cf. p. 265) on the surface of the nonphagocytic MHC classII- follicular dendritic cells (FDC) enables these cells to trap complexed antigen very efficiently and hold it in its native form on their surface for extended periods. Memory B-cells can then be stimulatedby recognition of the retained antigen and costimulated through the B-cell CD27 (cf. p. 181)recognizing complement fragments held on the surfaceof the FDC. However, to what extent such complexes form an essentialdepository of antigen for the stimulation of Bcells is a matter of some controversy. Classically,a secondary responsewould be initiated

Ihesurloce mo*ersofcellsinfhelmmune syslem . Individual surface molecules are assigned a cluster of differentiation (CD) number defined by a cluster of monoclonal antibodies reacting with that molecule. 0rgonizedlymphoidlissue . The complexity of immune responses is catered for by a sophisticatedstructure. o Lymph nodes filter and screen lymph flowing from the body tissueswhile spleenfilters the blood. o B- and T-cell areasare separated,at least partly under the direction of chemokines

il\ rci'.i&l&iif

.\#N

Afferenl lymphotics

Lymph node

Antlgenpresentotion

Efferenl

at the T-helper level by antigen, alone or as a complex, being takenupbylDCs and macrophages.Howevet the capture of immune complexes on the surface of FDCs opens up an alternative pathway. One to three days after secondary challenge, the filamentous denddtes on the follicular cells, to which the immune complexes are bound, form into beads which break off as structures called'iccosomes' (immune complex-coatedbodies). These bind to germinal center B-cells which then endocytose and process the antigen for presentation by the B-cell MHC class II, and subsequent stimulation of T-helper cellsto kick off the secondaryresponse.

o B-cell structures appear in the lymph node cortex as primary follicles which become secondary follicles with germinal centers after antigen stimulation. . Germinal centers with their meshwork of follicular dendritic cells expand B-cell blasts produced by secondary antigen challenge and direct their differentiation into memory cells and antibody-forming plasma cells.

immunity Mucosol . Specialized antigen-transporting M-cells provide the gateway for antigens to the mucosal lymphoid tissue. . Lymphoid tissue guarding the gastrointestinal tract is ( C o n t i n u e d p1 7 0 )

I rzo

CHAPTER 7 - T H E A N A T O M YO F T H EI M M U N ER E S P O N S E

unencapsulated and somewhat structured (tonsils, Peyer's patches, appendix). There are also diffuse cellular collections in the lamina propria. Intraepithelial lymphocytes are mostly T-cells and include some novel subsets, e.g. CD8 aobearing cells which use nonclassical MHC molecules as restriction elements for antigenpresentation. . Together with the subepithelial accumulations of cells Iining the mucosal surfaces of the respiratory and genitourinary tracts, these cells and lymphoid tissues form the 'secretory immune system' which bathes the surface with protective antibodies.

ofhersiles . T-cells in the skin characteristically bear cutaneous lymphocyte antigen (CLA) and the chemokine receptor CCR4. . Bone marrow is a major site of antibody production. . The respiratory tract and the liver contain substantial numbers of lymphocytes and phagocytic cells. o The brain, anterior chamber of the eye and testis are privileged sites in which antigens can be safely sequestered. Lymphocylelroftic . Lymphocyte recirculation between the blood and tissues is guided by specializedhoming receptors on the surfaceof the high-walled endothelium of the postcapillary venules.

F U R I H ER E A D I N G Barclay A N ef a/ (7997) Thc LeucoctlteAntigen Factsbook,2ndedn Academic Press,London Bos J.D. (ed) (2004)Skfulmnune System(SIS):utaneous innnunologtl ottdclinical imnurunodermatology,3rdedn. Boca Raton CRC Press. Brandtzaeg P.& Pabst R. (2004)Let's go mucosal: communication on slippery gro und Trends in lmm unol ogy 25, 570-577 Caligarls-Cappio F. (1998) Germinal centers In Delves PJ & Roitt I M (eds) Ercyclop ediaof lm mtmology, 2nd edn, p p 992-995 Academic Press,London Cheroutre H (2004)Starting at the beginning: new perspectives on thebiology of mucosai T cells.Annual Reaiewsof lmmunology 22, 217-246 Iwata M., Hirakiyama A., Eshima Y. et al. (2004) Retinoic acid imprints gut-homing specificity on T cells lmmwity 21, 527 -538

. Lymphocytes are tethered and then roll along the surface of the selected endothelial cells through interactions between selectins, integrins and chemokine receptors, and their respective ligands. Arrestof the lymphocyte following LFA-1 activation results in cell flattening and subsequent transmigration acrossthe endothelial cell. Ihe hondlingof onligen . Macrophages are general antigen-presenting cells for primed lymphocytesbut cannot stimulate naive T-cells. o This is effected by dendritic cells of hematopoietic origin which process antigen, migrate to the draining lymph node and settle down as interdigitating dendritic cells. They can present antigen-derived peptides to naive T-cells, thereby powerfully initiating primary T-cell responses. . Follicular dendritic cells in germinal centers bind immune complexes to their surface through Ig and C3b receptors. The complexes are long-lived and can provide a sustained source of antigenic stimulation for B-cells.

(www.roilt.com) website Seetheoccomp0nying questions, choice formultiple

Kraehenbuhl J P. & Neutra M R. (2000) Epithelial M cells: differentiation and function Annual Reaiezusof Cell and Deaelopmental Biologt176,307132 Lefrangois L. & Puddington L (2006) Intestinal and pulmonary mucosal T-cells: Local heroes fight to maintain the status quo of Immunology 24,681-704 Annuol Reztiezu McHeyzer-Williams LJ., Driver DJ. & McHeyzer-Williams M'G. (2001)Germinal center reaction Current Opinion in Hematology8' 52-59 PribilaJ.T, QualeA.C , MuelierKL& ShimizuY. (2004)Integrins and T cell-mediated immunity. Annual Reztiewsof lmmunology 22, 157-180 Wardlaw, A J , Guillen C & Morgan A. (2005) Mechanisms of T cell migration to thelung. Clinical and Experimental Allergy 35,4-7 -

Lymphocyte oclivqtion

INIRODUCTION The adaptive immune response begins as a result of an encounterbetween a B- or T-lymphocyte and its specific antigen which typically results in 'activation' of the lymphocyte and a radical shift in cell behavior-from a quiescent nondividing state, to a more active proliferative one. This simultaneously achieves two goals: the number of cells that are capable of responding to a particular antigen are multiplied (clonal expansion), and these new recruits are equipped with the ability to produce large quantities of cytokines or antibodies to help repel the intruder. Becauseof the potential dangers associated with inappropriate lymphocyte activation (to 'self ' or innocuous substances),signals that promote T- or B-cell activation usually require costimulation by other cells of the immune system. The requirement for costimulation raises the threshold for lymphocyte activation and provides a safeguard against autoimmunity (seeChapter 18). In previous chapters we learned that B- or T-cellsuse related, but nonetheless distinct, antigen receptors to 'see' antigen. Stimulation of T- or B-cells through their respective antigen receptors initiates a cascadeof signal transduction events that rely heavily upon protein kinases; proteins that can add phosphate groups to other proteins. While there are differences in the nature of the specific kinases that relay signals from the B- and T-cell receptors, there are also many similarities. Inboth cases,these signal transduction events result in the activation of many of the same transcription factors, entry into the cell division cycle, and the expression of an array of new proteins by the activated lymphocyte.

CTUSTERIN OG FM E M B R A NREE C E P T O R S ACTIVAIION TEADS T OT H E I R All cells use plasma membrane-borne receptors to extract information from their environment. This information is propagated within the cell by signaling molecules and enables the cell to make the appropriate response; whether this is reorganization of the cell cytoskeleton (to facilitate movement), expression of new gene products, increased cellular adhesiveness, or all of the above. In many instances, occupation of the receptor with its specific ligand (whether this is a growth factor, a hormone or an antigen) results in aggregation of the receptor and creates the conditions under which the receptors,or proteins associatedwith the receptor cytoplasmic tails, can mutually interact. Becausemany plasma membrane receptors are protein kinases,or can recruit protein kinasesuponengagement with their specific ligands, aggregation of the receptors typically results in phosphorylation of the receptor, or of associatedproteins. In the caseofthe B- and T-cell receptors, the receptors themselves do not have any intrinsic enzymatic activity but are associated with invariant accessory molecules (the CD3 16e and ( chains in the caseof the T-cell receptor, and the Ig-cr/ p complex in the caseof the B-cell receptor) that can attract the attentions of a particular class of kinases. Central to this attraction is the presence of special motifs called ITAMs (f mmunoreceptor fyrosine-based activation motifsl within the cytoplasmic tails of these accessory molecules (see also p. 63, Chapter 4). Phosphorylation of ITAMs at tyrosine residues-in response to TCR or BCR stimulation-enables these motifs to interact with adaptor proteins that have an affinity for phosphorylated tyrosine motifs, thereby initiating signal transduction. We will deal, in turn, with the signaling events that

take place upon encounter of a T-cell or a B-cell with antigen.

T . T Y M P H O C Y IAENSDA N I I G E N . P R E S E N T I N G C E t t s I N I E R A CIIH R O U G S H E V E R APTA I R S O FA C C E S S O R MYO T E C U T E S Before we delve into the nuts and bolts of TCR-driven signaling events, it is important to recall that T-cells can only recognize antigen when presented within the peptide-binding groove of major histocompatibility complex (MHC) molecules. Furthermore, while the TCRis the primary means bywhich T-cellsinteractwith the MHC-peptide complex, T-cells also express coreceptors for MHC (either CD4 or CD8) that define functional T-cell subsets.Recall that CD4 molecules act as coreceptors for MHC class II and are found on T-helper cell populations which provide'help' for activation and maturation of B-cells and cytotoxic T-cells (figure 8.1). CD8 molecules act as coreceptors for MHC class I molecules and are a feature of cytotoxic T-cells that can kill virally infected or precancerous cells (figure 8.1). Note, however, that the affinity of an individual TCR for its specific MHC-antigen peptide complex is relatively low (figure 8.2).Thus, a sufficiently stable association with an antigen-presenting cell (APC) can only be achieved by the interaction of several complementary pairs of accessorymolecules such as LFA-1/ICAM-1, CD2/LFA-3 and so on (figure 8.3). However, these

(4) M-l AFFTNTTY Figure 8.2. The relative affinities of molecular pairs involved in interactions between T-lymphocytes and cells presenting antigen. The ranges of affinities for growth factors and their receptors, and of antibodies, are shown for comparison. (Based on Davies M M & Chien Y.-H (1993)Current Opinion in lmmunology 5, 45 )

vcAM-t-> tcAM-t->

CD4+T-cell

Figure 8.1. Helper and cytotoxic T-cell subsets are restricted by MHC class. CD4 on helper T-cells acts as a coreceptor for MHC class II and helps to stabilize the interaction between the TCR and peptide-MHC complex; CD8 on cytotoxic T-cells performs a similar functionby associating with MHC class I

signol@+ @

Signol@

CD4+T-cell

Figure 8.3. Activation of resting T-cells. Interaction of costimulatory molecules leads to activation of resting T-lymphocyte by antigenpresenting cell (APC) on engagement of the T-cell receptor (TCR) with its antigen-MHC complex Engagement of the TCR signal 1 without accompanying costimulatory signal2leads to anergy. Note, a cytotoxic rather than a helperT-cell would, of course, involve coupling of CD8 to MHC class I. Signal 2 is delivered to a resting T-cell primarily through engagement of CD28 on the T-cellby 87 .7 or 87 .2 on the APC CTLA-4 competes withCD2SforBTligands and has a muchhigher affinity than CD28 for these molecules Engagement of CTLA-4 with 87 downregulates signal 1 ICAM-L/2, lntercellular adhesion moleculLe-1'/2; LF A-1/ 2, /ymphocyte Tfunction-associated molecule-7 / 2; VCAM-I, zrascularcelladhesion rrlolecule-1;YLA-4, aery late antigen-4.

CHAPTER 8 - L Y M P H O C Y TA EC T I V A T I O N

HIGHAFFINIry

Figure 8.4. Integrin activation. Integrins such as LFA-1 can assume different conformations that are associated with different affinities The bent head-piece conformation has a 1ow affinity for ligand but can be rapidly transformed into the extended high-affinity conformation by activation signals that act on the cytoplasmic tails of the integrin o and p subunits; a processknown as'inside-out'signaling.

molecular couplings are not necessarily concerned with intercellular adhesion alone; some of these interactions also provide the necessarycostimulation that is essential for proper lymphocyte activation. Unstimulated lymphocytes are typically nonadherent but rapidly adhere to extracellular matrix components or other cells (such as APCs) within seconds of encountering chemokines or antigen. Integrins such as LFA-1 and VLA-4 appear to be particularly important for lymphocyte adhesion.The easewith which lymphocytes can alter their adhesivenessseemsto be related to the ability of integrins to change conformation; from a closed, low-affinity state, to a more open, high-affinity, one (figure 8.4). Thus, upon encounter of a T-cell with an APC displaying an appropriate MHC-peptide complex, the affinity of LFA-I for ICAM-1 is rapidly increased and this helps to stabilize the interaction between the T-cell and theAPC. This complex has come to be known, in fashionable terms, as the immunological synapse. Activation of the small GTPase Rapl by TCR stimulation appears to contribute to the rapid change in integrin adhesiveness.How Rapl achieves this remains somewhat hazy,blrt it is likely that modification of the integrin cytoplasmic tail serves to trigger a conformational change in the integrin extracellular domains; a process that has been termed 'inside-out' signaling.

I H EA C I I V A I I O N O FT - C E t t SR E S U I R E S I W OS I G N A T S Stimulation of the TCR by MHC-peptide (which can be

I 7 3' l I

mimicked by antibodies directed against the TCR or CD3 complex) is not sufficient to fully activate resting helper T-cells on their own. Upon addition of interleukin-1 (IL-1), however, RNA and protein synthesis is induced, the cell enlarges to a blastlike appearance, interleukin-2 (IL-2) synthesis begins and the cell moves from G0 into the G1 phase of the cell division cycle. Thus, two signals are required for the activation of a resting helper T-cell (figure 8.3). Antigen in association with MHC class II on the surface of APCs is clearly capable of fulfilling these requirements. Complex formation between the TCR and MHC-peptide provides signal 1, through the receptor-CD3 complex, and this is greatly enhanced by coupling of CD4 with the MHC. The T-cell is now exposed to a costimulatory signal (signal 2) from the APC. Although this could be IL-7,it would appear that the most potent costimulator is 87 on the APC which interacts with CD28 on the T-cell. Thus activation of resting T-cells can be blocked by anti-B7; surprisingly, this renders the T-cell anergic, i.e. unresponsive to any further stimulation by antigen. As we shall see in later chapters, the principle that two signals activate, but one may induce anergy in, an antigen-specific cell provides a potential for targeted immunosuppressive therapy. However, unlike resting T-lymphocytes, activated Tcells proliferate in response to a slngle signal. Adhesion molecules such as ICAM-1, VCAM-1 and LFA-3 are not intrinsically costimulatory but augment the effect of other signals (figure 8.3); an important distinction. Early signaling events also involve the aggregation of lipid rafts composed of membrane subdomains enriched in cholesterol and glycosphingolipids. The cell membrane molecules involved in activation become concentrated within these structures.

P R O T E IIN Y R O S I NP EH O S P H O R Y T A TI S ION A N E A R TE Y V E N I N I - C E t TS I G N A T I N G As we shall seeshortly, the signaling cascadesthat result from TCR stimulation canbecomequite complex (figure 8.5);but take it one step at a time and a senseof order can be extracted from the apparent chaos. Interaction between the TCR and MHC-peptide complex is greatly enhanced by recruitment of either coreceptor for MHC, CD4 or CD8, into the complex. Furthermore, because the cytoplasmic tails of CD4 and CD8 are constitutively associated with Lck, a protein tyrosine kinase (PTK) that can phosphorylate the three tandemly arranged ITAMs within the TCR ( chains, recruitment of CD4 or CD8 to the complex results in stable associationbetween Lck and its ( chain substrate(figure 8.6).

CHAPTER 8 - T Y M P H O C Y TA EC T I V A T I O N

174

Help,nypoater istnosnall \.._

)ie 7t'lYl

tuq

t\ l t4 6e*.t

Figure 8.5. Signaling pathways can become quite complex. (Reproducedwith permissionfromZolnierowicz S & Bollen M (2000)EMBO Iournal19,483 )

Phosphorylation of ( chain by Lck creates binding sites for the recruitment of another PTK, ZAP-70 (zela chain associated protein oI 70 kDa), into the TCR signaling complex. Recruitment of ZAP-70 into the receptor complex results in activation of this PTK by Lck-mediated phosphorylation. ZAP-70, in turn, phosphorylates two adaptor proteins, LAT (linker for activation of T-cells) and SLP76 (SH2-domain containing leukocyte protein of 76 kDa) that can instigate divergent signaling cascadesdownstream (figure 8.6). LAT plays an especially significant role in subsequent events by serving as a platform for the recruitment of several additional players to the TCR complex. LAT contains many tyrosine residues that, uponphosphorylation by ZAP-70, can bind to other adaptor proteins through motifs (called SH2 domains) that bind phosphotyrosine residues. Thus, phosphorylation of LAT results in recruitment of GADS (GRB2-related adapter protein) which is constitutively associatedwith SLP76. SLP76 has been implicated in cytoskeletal rearrangements due to its ability to associatewith Vavl and NCK (figure 8.6). Thus, TCR-stimulation induced cell shape changes are most likely due to recruitment of SLP76 into the TCR signaling complex. Phosphorylated LAT also attracts the attentions of two additional phosphotyrosine-binding proteins; the 11isoform of phospholipase C (PLCfl), and the adaptor protein GRB2 (growth factor receptor-binding protein 2). From this point on, at least two distinct signaling cascadescan ensue; the Ras/MAPK pathway and the phosphatidylinositol pathway.

Figure 8.6. T-cell signaling leads to activation. Signals through the MHC-antigen complex (signal 1) and costimulator B7 (signal 2) initiate a cascadeofprotein kinase activation events and a rise in intracellular calcium, thereby activating transcription factors which control entry in the cell cycle from G0 and regulate the expression of IL-2 and many other cytokines. Stable recruitment of CD4 or CD8 to the TCR complex initiates the signal transduction cascadethrough phosphorylation of the tandemly arranged ITAM motifs within the CD3 ( chains Subsequent events which creates binding sites for theZAP-Tlkinase are marshaled through ZAP-7}-rnediated phosphorylation of LAI; recruitment of several signaling complexes to LAT result in triggering of theRas/MAPKandPLCyl signalingpathways The latterpathways culminate in activation of a range of transcription factors including NFrB, NFAT and Fos/Jun heterodimers Note that other moiecules can also contribute to this pathway but have been omitted for clarity. Seemain body of text for further details. Abbreviations: DAG, diacylglycerol; ERK, extracellular signal regulated kinase; IPr, inositol triphosphate; LA! linker for activated T-cells; NFrB, nuclear/actor rB; NFAI, ruclear /actor of activated T-cells; OCT-1, octamer-binding f actor; P ak7, p21-activated kinase; PIP., phosphatidylinositol diphosphate; PKC, protein kinase C; PLC, phospholipase C; SH2, Src-ftomology domain 2; SLAP, SLP-76-nssociated phosphoprolein; SLP-76, SH2-domain containing leukocyte-specilic 76 kDa phosphoprotein; Positive signal ZAP-70, ( chain-associated protein kinase *, transduction

FOttOWING D O W N S I R E AEMV E N I S I C RS I G N A T I N G polhwoy TheRos/MAPK Ras is a small G-protein that is constitutively associated with the plasma membrane and is frequently activated

CHAPTER 8-LYMPHOCYTA ECTIVATION

Ligond

(Grb2) Adoptor

GTP

Cytosol

GDP

I 7 5I I

then sets in motion a series of further kinase activation events culminating in phosphorylation of the transcription factor Elk1. Elkl phosphorylation permits translocation of this protein to the nucleus and results in the expression of Fos, yet another transcription factor. The appearance of Fos results in the formation of heterodimers with Jun to form the AP-1 complex which has binding sites on the IL-2 promoter as well as on many other genes (figure 8.6).Deletion of AP-1 binding sites from the IL-2 promoter abrogates 90% of IL-2 enhancer activity. polhwoy Thephosph0lidylin0sit0l

5 ( Y

MAPklnosekinosekinose(RoD

v MAPkinosekinose(Mek)

v MAPkinose(ERK) Figure 8.7. Regulation of Ras activity controls kinase amplification cascades. A number of cel1 surface receptors signal through Rasregulated pathways. Ras cycles between inactive Ras-CDp and active Ras-GTP, regulated by guanine nucleotide exchange factors (GEFs) which promote the conversion of Ras-GDP to Ras-GTp, and by GTPase-activating proteins (GAPs) which increase the intrinsic GTPase activity of Ras Upon ligand binding to receptor, receptor tyrosine kinases recruit adaptor proteins, e.g Grb2, and GEF proteins, such as Sos ('son of sevenless'), to the plasma membrane. These events generate Ras-GTP which activates Raf. (Modified with permission from Olson M F & Marais R (2000\ Semi nar s i n lmm u nol ogy12,63 )

in response to diverse stimuli that promote cell division (figure 8.7). Ras can exist in two states, GTP-bound (active) and GDP-bound (inactive). Thus, exchange of CDP for CTP stimulates Ras activation and enables this protein to recruit one of its downstream effectors, Raf. Sohow does TCR stimulation result in activation of Ras? One of the ways inwhich Rasactivationcanbe achieved is through the activity of GEFs (guanine-nucleotide exchange factors) that promote exchange of GDP for CTP on Ras. One such GEF, SOS (son of sevenless),is recruited to phosphorylated LAT via the phosphotyrosine-binding protein GRB2 (figure 8.6).Thus, phosphorylation of LAT by ZAP-70 leads directly to the recruitment of the GRB2/SOS complex to the plasma membrane where it can stimulate activation of Ras through promoting exchange of GDP for GTP. In its CTP-bound state, Ras can recruit a kinase, Raf (also called MAPKKK mitogen-associated protein kinase kinase kinase!), to the plasma membrane which

Phosphorylation of LAT by ZAP-70 not only promotes docking of the GRB2/SOS complex on LAI but also stimulates recruitment of the y1 isoform of phospholipase C (PLCy1;figure 8.6).PLCyI plays a crucial role in propagating the cascade further. Phosphorylation of PLCyI activates this lipase which enables it to hydrolyze the membrane phospholipid, phosphatidylinositol biphosphate (PIP,), into diacylglycerol (DAG) and inositol triphosphate (IPr) (figure 8.6). Interaction of IP, with specific receptors in the endoplasmic reticulum triggers the release of Caz+into the cytosol which also triggers an influx of extracellular calcium. The raised Ca2*concentration within the T-cell has at least two consequences. First, it synergizes with DAG to activate protein kinase C (PKC); second, it acts together with calmodulin to increase the activity of calcineurin, a protein phosphatase that can promote activation of an important transcription factor (NFAT) required forIL-2 production. The Ca2*-dependent activation of PKC by DAG is instrumental in the activation of yet another transcription factoq, NFrB. NFrB is actually a family of related transcription factors that are involved in the regulation of transcription of many genes, including cytokines (such as IL-2), as well as genes that can promote cell survival byblocking signals that promote apoptosis. Conlrol0f lL-2 genelr0nscriplion Transcription of IL-2 is one of the key elements in preventing the signaled T-cell from lapsing into anergy and is controlled by multiple binding sites for transcriptional factors in the promoter region (figure 8.6). Under the influence of calcineurin, the cytoplasmic component of the ruclear factor of activated T-cells (NFAT")becomes dephosphorylated and this permits its translocation to the nucleus where it forms a binary complex with NFAT., its partnerwhich is constitutively expressed in the nucleus. The NFAT complex binds to

CHAPIER 8.IYM PHOCYTEACTIVATION] two different IL-2 regulatory sites (figure 8.6).Note here that the calcineurin effect is blocked by the anti-T-cell drugs cyclosporine and tracolimus (see Chapter 16). PKC- and calcineurin-dependent pathways synergize in activating the multisubunit kB kinase (IKK), which phosphorylates the inhibitor kB thereby targeting it for ubiquitination and subsequent degradation by the proteasome. Loss of kB from the kB-NFrB complex exposes the nuclear localization signal on the NFxB transcription factor which then swiftly enters the nucleus. In addition, the ubiquitous transcription factor Oct-L interacts with specific octamer-binding sequence motifs. We have concentrated on IL-2 transcription as an early and central consequence of T-cell activation, but more than 70 genes are newly expressed within 4 hours of T-cell activation, leading to proliferation and the synthesis of several cytokines and their receptors (see Chapter 9). CD28costimuloli0n 0mplifiesTCRsign0ls As discussed earlier, naive T-cells typically require two signals for proper activation; one derived from TCR ligation (signal 1), and the other most likely provided by simultaneous engagement of CD28 on the T-cell (signal 2) by 87.7 or B7.2 on the APC (figure 8.3). Indeed, Tcells derived from CD28-deficient mice, or cells treated with anti-CD28 blocking antibodies, display severely reduced capacity to proliferate in response to TCR stimulation in aitro and in aiao. Moreover, CD28 deficiency also impairs T-cell differentiation and the production of cytokines required for B-cell help. Similar effects are also seen when B7.1 or B7.2 expression is interfered with. Sowhat does tickling the CD28 receptor do thatis so special? \A/hile early studies suggested that CD28 stimulation may result in qualitatiaely different signals to those that are generated through the TCR, recent studies suggest that this might not be the case. Instead, these studies suggest that while CD28 engagement undoubtedly activates pathways within the T-cell that TCR stimulation alone does not (such as the phosphatidylinositol 3kinase [PI3K] pathway), the primary purpose of costimulation through CD28 may be to quantitatiaely amplify signals through the TCRby converging on similar transcription factors such as NFrB and NFAT. In support of this view, microarray analyses of genes upregulated in response to TCR ligation alone, versus TCR ligation in the presence of CD28 costimulation, found, rather surprisingly, thatessentially the same cohorts of geneswere expressed in both cases. While signals through CD28 enhanced the expression of many of the genes switched

on in response to TCR ligation, no new genes were expressed. This indicates that CD28 costimulation may be required in order to cross signaling thresholds that are not achievable via TCR ligation alone. One is reminded here of the choke that earlier generations of cars were supplied with to provide a slightly more fuelrichmixture tohelp start a coldengine! CD28 costimulation of naive T-cells may serve a similar purpose, with 'choke' no longer needed when these cells the CD28 have warmed up as a result of previous stimulation. Furlherlhoughls0n T-celllliggering A seri al T CR engagem ent m o d el f o r T-cell actia ati o n We have already commented that the major docking forces which conjugate the APC and its T-lymphocyte counterpart must come from the complementary accessorymolecules such as ICAM-1/LFA-1 and LFA3/CD2, rather than through the relatively low affinity TCR-MHC plus peptide links (figure 8.3). Nonetheless, cognate antigen recognition by the TCR remains a sine qua non for T-cell activation. Fine, but how can as few as 100 MHC-peptide complexes on anAPC, through their low affinity complexing with TCRs, effect the Herculean task of sustaining a raised intracellular calcium flux for the 60 minutes required for full cell activation? Any fall in calcium flux, as may be occasioned by adding an antibody to the MHC, and NFAT. dutifully returns from the nucleus to its cytoplasmic location, so aborting the activation process. Surprisingly, Valitutti and Lanzavecchia have shown that as few as 100 MHC-peptide complexes on an APC can downregulate 18 000 TCRs on its cognate Tlymphocyte partner. They suggest that each MHCpeptide complex can serially engage up to 200 TCRs. In theirmodel, conjugationof anMHC-peptide dimerwith two TCRs activates signal transduction, phosphorylation of the CD3-associated ( chains with subsequent downstream events, and then downregulation of those TCRs. Intermediate affinity binding favors dissociation of the MHC-peptide, freeing it to engage and trigger another TCR, so sustaining the required intracellular activation events. The model for agonist action would also explain why peptides giving interactions of lower or higher affinity than the optimum could behave as antagonists (figure 8.8). The irnportant phenomenon of modified peptides behaving as partial agonists, with differential effects on the outcome of T-cell activation, is addressed in the legend to figure 8.8. The immunolo gical synap se Experiments using peptide-MHC and ICAM-1 molecules labeled with different fluorochromes and inserted

C H A P T E R8 - L Y M P H O C Y T EA C T I V A T I O N

177 |

No ol Durolion, frequency ond successful quolityof complex formolioncomplexes formedin giventime

Figure 8.8. Serial triggering model of TCR activation (Valitutti S. & Lanzavecchia A (1995) The Immunologist 3,122) Intermediate affinity complexes between MHC-peptide and TCR survive long enough for a successful activation signal to be transduced by the TCR, and the MHC-peptide dissociates and fruitfully engages another TCR. A sustained high rate of formation of successful complexes is required for full T-cell activation Low affinity complexes have a short halflife which either has no effect on the TCR or produces inactivation, perhaps through partial phosphorylation of ( chains (O successful TCR activation; O, TCR inactivation; -, no effect: the length of the horizontal bar indicates the lifetime of that complex) Being of low affinity, they recycle rapidly and engage and inactivate a large number of TCRs High affinity complexes have such a long lifetime before dissociation that insufficient numbers of successful triggering events occur Thus modified peptide ligands of either low or high affinity can act as antagonists by denying the agonist accessto adequate numbers of vacant TCRs. Sorne modified peptides act as partial agonists in that they produce differential effects on the outcomes of T-cell activation For example,a singleresiduechangeinahemoglobinpeptidereduced IL-4 secretion 1O-fold but completely knocked out T-cell proliferation The mechanism presumably involves incomplete or inadequately transduced phosphorylation events occurring through a truncated half-life of TCR engagement, allosteric effects on the MHC-TCR partners, or orientational misalignment of the peptide within the complex (Reproduced with permission of Hogrefe & Huber Publishers )

into a planar lipid bilayer on a glass support have provided evidence for the idea that T-cell activation occurs in the context of an immunological synapse. A clustered area of integrins acts as an anchor to permit optimal interaction between the opposing cell surfaces. Initially unstable TCR-MHC interactions occur in a broad outer ring surrounding the integrins. The peptide-MHC molecules then move towards the center of the synapse,changingplaceswith the adhesionmolecules which now form the outer ring (figure 8.9). It has been suggested that the generation of the immunological s).napse only occurs after a certain initial threshold level of TCR triggering hasbeen achieved, its formation being dependent upon cytoskeletal reorganization and leading to potentiation of the signal. Dompingl-cell enlhusiosm We have frequently reiterated the premise that no selfrespecting organism would permit the operation of an expanding enterprise such as a proliferatingT-cellpopulation without some sensible controlling mechanisms.

Figure 8.9. The immunological synapse. (a) The formation of the irnmunological synapse. T-cells were brought into contact with planar lipid bilayers and the positions of engaged MHC-peptide (green) and engaged ICAM-1 (red) at the indicated times after initial contact are shown (Reproduced with permission from Grakoui A., Dustin M.L et al ('1.999)Science285,221. @American Association for the Advancement of Science ). (b) Diagrammatic representation of the resolved synapse in which the adhesion molecule pairs CD2/LFA-3 and LFAl/ICAM-1, which were originally in the center, have moved to the outside and now encircle the antigen recognition and signaling interactionbetween TCR and MHC-peptide and the costimulatoryinteraction between CD28 and 87 The CD43 molecule has been reported to bind to ICAM-1 and E-selectin, and upon ligation is able to induce IL-2 mRNA, CD59 and CD154 (CD40L) expression and activate the DNAbinding activity of the AP-1, NFrB and NFAT transcription factors

Whereas CD28 is constitutively expressed on T-cells, CTLA-4 is not found on the resting cell but is rapidly upregulated following activation. It has a 10- to 20-fold higher affinity for both 87.1 andB7.2 and, in contrast to costimulatory signals generated through CD28, 87 engagement of CTLA-4 downregulates T-cell activation. The mechanism by which CTLA-4 suppresses T-cell activation remains enigmatic, as this recePtor

I rza

CHAPTER 8-LYMPHOCYTA ECTIVATION

appears to recruit a similar repertoire of proteins (such as PI3K) to its intracellular tail as CD28 does.It has been proposed, however, that CTLA-4 may antagonize the recruitment of the TCR complex to lipid rafts, which is where many of the signaling proteins that propagate TCR signals reside. A number of adaptor molecules have been identified which may be involved in reigning in T-cell activation. Theseinclude SLAR SIT and members of the Cbl family. Cbl-b appears to influence the CD28 dependenceof IL-2 production during T-cell activation, perhaps via an effect on the guanine nucleotide exchange factor Vav. Another member of the Cbl fam1ly, Cbl-c, is a negative regulator of Syk andZAP-7}, and may thereby alter the triggering threshold of the antigen receptors on both T- and B-cells. Tempting though it might be, phosphatases should not automatically be equated with downregulation of a phosphorylation cascade.The observation that T-cell mutants lacking CD45 do not possess signal transduction capacity was at first sight deemed to be strangebecauseCD45 has phosphataseactivity and was thought thereby to downregulate signaling. However, the Lck kinase in the CD45-deficient T-cellsis phosphorylated on tyrosine-SO5which is a negative regulatory site for kinase activity; hence dephosphorylation by CD45 activates the Lck enzyme and the paradox is resolved.

B . C E t t SR E S P O NI D O THREE DIFFERENT I Y P E SO FA N I I G E N I Type I fhymus-independenl onligens Certain antigens,such as bacterial lipopolysaccharides, at a high enough concentration, have the ability to activate a substantialproportion of the B-cellpool polyclonally, i.e. without reference to the antigen specificity of

TypeI lhymus-independent 0ntigens 0repolyclonol octivolors

the surfacereceptor hypervariable regions.They do this through binding to a surface molecule which bypasses the early part of the biochemical pathway mediated by the specific antigen receptor. At concentrations which are too low to cause polyclonal activation through unaided binding to these mitogenic bypass molecules, the B-cell population with Ig receptorsspecificfor these antigens will selectively and passively focus them on their surface,where the resulting high local concentration will suffice to drive the activation process (figure 8.10a).

2 lype 2lhymus-independent onligens Certain linear antigens which are not readily degraded in the body and which have an appropriately spaced, highly repeating determinant - Pneumococcus polysaccharide, ficoll, o-amino acid polymers and polyvinylpyrrolidone, for example - are also thymus-independent in their ability to stimulate B-cells directly without the need for T-cell involvement. They persist for long periods on the surface of specialized macrophages located at the subcapsularsinus of the lymph nodes and the splenic marginal zone, and can bind to antigenspecificB-cellswith great avidity through their multivalent attachment to the complementary Ig receptors which they cross-link (figure 8.10b). In general, the thymus-independent antigens give rise to predominantly low affinity IgM responses, some IgG3 in the mouse, and relatively poor, if any, memory. Neonatal B-cellsdo not respond well to type2 antigens and this has important consequences for the efficacy of carbohydrate vaccinesin young children.

ontigens 3 Thymus-dependenl Th e n eed f o r co ll ab o r ati on utith T-helper cell s Many antigens are thymus-dependent in that they

Type 2 thymus-independenl ontigen with repeoling delerminonls cross-link lg receplors

Figure8.10. B-cellrecognitionof (a) typel and (b) type 2 thymus-independent antigens. The complex gives a sustained signal to the B-cellbecause of the long halflife of this type of molecule. - ^), activation - - -, signai; J,ssfacelgreceptor ; crossJinking of receptors

r7eI

ACTIVATION CHAPTER 8 - L Y MP H O C Y T E provoke little or no antibody responsein animals which have been thymectomized at birth and have few T-cells (Milestone 8.1). Such antigens cannot fulfil the molecular requirements for direct stimulation; they may be univalent with respect to the specificity of each determinant; they may be readily degraded by phagocytic cells; and they may lack mitogenicity. If they bind to B-cell receptors,they will sit on the surfacejust like a hapten and do nothing to trigger the B-cell (figure 8.11). Cast your mind back to the definition of a hapten-a small molecule like dinitrophenyl (DNP) which binds to preformed antibody (e.g.the surface receptor of a spe-

In the 1960s,as the mysteries of the thymus were slowly unraveled, our erstwhile colleaguespushing back the frontiers of knowledge discovered that neonatal thymectomy in the mouse abrogated not only the cellular rejection of skin grafts, but also the antibody responseto some but not all antigens (figure M8 1.1) Subsequentinvestigations showed that both thymocytes and bone marrow cellswere neededfor optimal antibody responsesto such thymus-dependent antigens (figure M8 1.2) By carrying out thesetransfers with cells from animals bearing a recognizable chromosome marker (T6), it became evident that the antibody-forming ceils were derived from the bone marrow inoculum, hence the nomenclature'I' for lhymus-derived lymphocytes and'B' for antibody-forming ce11precursors originating in the Bone marrow. This convenient nomenclature has stuck even though bone marrow contains embryonic T-cell precursors since the immunocompetent T- and B-cells differentiate in the thymus and bone marrow respectively(seeChapter 11).

cific B-cell) but fails to stimulate antibody production (i.e. stimulate the B-cell). Remember also that haptens become immunogenic when coupled to an appropriate carrier protein (seep. 89). Building on the knowledge that both T- and B-cells are necessary for antibody responses to thymus-dependent antigens (Milestone 8.1),wenow know thatthe carrier functions to stimulate T-helper cells which cooperate with B-cells to enable them to respond to the hapten by providing accessory signals (figure 8.11). It should also be evident from figure 8.11 that, while one determinant on a typical protein antigen is behaving as a hapten inbinding to the

Ag INJECTED

IHYIVECTOIVIY

RESPONSE ANTIBODY

+++

Shom

Neonolol

+++

Neonolol

Figure M8.1.1. The antibody resPonse to some antigens is The thymus-dependent and, to others, thymus-independent. response to tetanus toxoid in neonatally thymectomized animals could be restored bv the iniection ofthlrnocvtes

Cellsinjected

N0ne

(B) (T) Bonemorrow Thymocytes

Produclion of Ab Figure M8.1.2. The antibody response to a thymus-dependent antigen requires two different T-cell populations. Different populations of cells from a normal mouse histocompatible with the recipient (i.e. of the same H-2 haplotype) were injected into recipients which had been X-irradiated to destroy their own lymphocyte responses. They were then primed with a thymus-dependent antigen such as sheep red blood cells (i.e. an antigen which fails to

Thymocyles & Bonemorrow

+++

give a response in neonatally thymectomized mice; figure M8.1 1) and examined for the production of antibody after 2 weeks. The small amount of antibody syrthesized by animals receiving bone marrow alone is due to the presence of thymocyte precursors in the cell inoculum which differentiate in the intact th1'rnus gland of the recipient.

I rao

CHAPTER 8 - T Y M P H O C Y TA EC T I V A T I O N

B-cell, the other determinants subserve a carrier function in recruiting T-helper cells. Antigen p ro cessing by B - cells The need for physical linkage of hapten and carrier strongly suggests that T-helpers must recognize the carrier determinants on the responding B-cell in order to provide the relevant accessorystimulatory signals. However, since T-cells only recognize processed membrane-bound antigen in association with MHC molecules, the T-helpers cannot recognize native antigen bound simply to the Ig receptors of the B-cell as naively depicted in figure 8.11. All is not lost, however, since primed B-cells can present antigen to T-helper cells-in fact, they work at much lower antigen concentrations than conventional presenting cells because they can focus antigen through their

surface receptors.Antigen bound to surface Ig is internalized in endosomes which then fuse with vesicles containing MHC classII molecules with their invariant chain. Processingof the protein antigen then occurs as describedinChapter 5 (seefigure 5.16)and the resulting antigenic peptide is recycled to the surface in association with the classII molecules where it is available for recognition by specificT-helpers (figures 8.12and 8.13). The need for the physical union of hapten and carrier is now revealed; the hapten leads the carrier to be processed into the cell which is programed to make antihapten antibody and, following stimulus by the Thelper-recognizing processedcarrier,it will carry out its program and ultimately produce antibodies which react with the hapten (is there no end to the wiliness of nature?!).

T H EN A I U R E O FB - C E TA t CIIVAIION

'^^-*^^^^r*,r'-

NOB-CELL STIMULATION T-HELPER-INDUCED B-CELL STIMULATION Figure 8.11. T-helper cells coopetate through protein carrier determinants to help B-cells respond to hapten or equivalent determinants on antigens by providing accessorysignals. (For simplicity we are ignoring the MHC component and epitope processing in T-cell recognition, but we won't forget it )

Similar to T-cells,naive or resting B-cells are nondividing and activation through the B-cell receptor (BCR) drives thesecells into the cell cycle.As is the casefor the T-cell receptor,the B-cell receptor (surface Ig) does not possessany intrinsic enzymatic activity. Once again, it is the accessory molecules associated with the antigen receptor that propagate activation signals into the Bcell. It was noted in Chapter 4 that the BCR complex is composed of membrane-anchored immunoglobulin which is associatedwith a disulfide-linked Ig-a and Ig-B heterodimer, the cytoplasmic tails of whicheach contain a single ITAMmotif (figure 4.3).As we willnow discuss in more detail, antigen-driven cross-linking of the BCR results in the initiation of a PTK-driven signaling cascade,seeded by the Ig-a/B ITAMs, that awaken a

B-CELL invorionl choin Y

coplure of ontigen

Endosome formotion & lusion ACTIVATED T-HELPER :

t^ |

-"j

I EH{"l l . - - - l .": I I :t::

l"-l

RFSTING B.CELL

processing Anligen

peptide-|\/Hc Corrier onligenic II interocts onsurfoce withT-cellreceDlor

Figure 8.12. B-cell handling of a thymusdependent antigen. Antigen captured by the surface Ig receptor is internalized within an endosome, processed and expressed on t h e s u r f a c ew i t h M H C c l a s sl l ( c f f i g u r e 5 16) Costimulatory signals through the CD40-CD40L (CD154) interaction are required for the activation of the resting cell by the T-helper slg cross-linking by antigen on the surface of an antigen-presenting cell is likely during secondary responses within the germinal centers when complementcontaining complexes on follicular dendritic cells interact with BJymphoblasts. ----> , activation signal; , cross-linking of receprors

CHAPTER E - T Y M P H O C Y TAEC T I V A T I O N

(

r8r I

tt' 'o ao.

Figure 8.13. Demonstration that endocytosed B-cell surface Ig receptors enter cytoplasmic vesicles geared for antigen processing. Surface IgG was cross-linked with goat anti-human Ig and rabbit antigoat Ig conjugated to 15-nm gold beads (large, dark arrow). After 2 minutes, the cell sections were prepared and stained with anti-HLADR invariant chain (2 nm gold; arrowheads) and an antibody to a cathepsin protease (5 nm gold; open arrows). Thus the internalized IgG is exposed to proteolysis in a vesicle containing class II molecules. The presence ofinvariantchain shows thatthe classII molecules derive from the endoplasmic reticulum and Golgi, not from the cell surface Note the clever use of different-sized gold particles to distinguish the antibodies used for localizing the various intravesicular proteins, etc (Photograph reproduced with permission from the authors and the publishers from Guagliardi L.E et aI (799O)Nature 343, 133.Copyright O 1990Macmillan Magazines Ltd )

panoply of transcription factors from their cellular slumber.

B-cellsrequirecostimul0tion lor etficientoctivotion B-cells also require costimulation to mount efficient effector responses.While CD28 is the primary costimulatory molecule for T-cells, a complex of costimulatory molecules perform essentially the same function in Bcells (figure 8.14).The mature B,cell coreceptorcomplex appears to be composed of four components: CD19, CD21 (complement receptor type 2, CR2),CD81 (TAPA1) and LEU13 (interferon-induced transmembrane protein 1). While the details of how the B-cell coreceptor complex augments B-cell receptor signaling are still under investigation, it is clear that CD19 plays an especially important role in this process. CD19 is a B-cell-specifictransmembrane protein that is expressed from the pro-B-cell to the plasma-cell stage and possessesa relatively long cytoplasmic tail. Upon Bcell receptor stimulation, the cytoplasmic tail of CD19

Figure 8.14. The BCR coreceptor complex. The B-cell coreceptor complex provides costimulatory signals for B-cell activation through recruitment of a number of signaling molecules, inciuding phosphatidylinositol 3-kinase and Vav, which can amplify signals initiated through the BCR. On mature B-cells, CD19 forms a tetrameric complex with three other proteins: CD21 (complement receptor type 2), CD81 (TAPA-1) and LEU13 (interferon-induced transmembrane protein 1).

undergoes phosphorylation at multiple tyrosine residues (by kinases associated with the BCR) which creates binding sites on CD19 for several proteins, including the tyrosine kinase Lyn and phosphatidylinositol 3-kinase (PI3K). Similar to the role that CD28 plays on T-cells, the B-cell coreceptor amplifies signals transmitted through the BCR approximately 100-fold. BecauseCD19 and CR2 (CD21) molecules enjoy mutual association, this could be brought about by bridging the Ig and CR2 receptors on the B-cell surface by antigen-C3d complexes bound to the surface of APCs. Thus, antigen-induced clustering of the B-cell coreceptor complex with the BCR lowers the threshold for B-cell activation by bringing kinases that are associated with the BCR into close proximity with the coreceptor complex. The action of these kinases on the coreceptor complex engages signaling pathways that reinforce signals originating from the BCR.

surfocelg B-cellsoreslimuloledby cross-linking B-cell activation begins with interaction between antigen and surface immunoglobulin. Recruitment of the BCR to lipid rafts is thought to play an important role in B-cell activation as surface Ig is normally excluded from lipid rafts but becomes rapidly recruited to rafts within minutes of Ig cross-linking (figure 8.15); this event probably serves to bring the PTK Lyn into close proximity with the ITAMs within the cytoplasmic tails of the BCR-associatedIg-crl F heterodimer as Lyn is constitutively associated with lipid rafts. Upon recruit-

lrcz

CHAPTER 8 - T Y M P H O C Y TA EC T I V A T I O N

CDI9

lg cr" ig F

ITAM

Figure 8.15. Receptor cross-linking recruits the BCR to lipid rafts. Antrgen-induced receptor crossJinking recruits the BCR, which is normally excluded from membrane cholesterol-rich lipid raft domains, to membrane rafts where signaling proteins such as the protein tyrosine kinase Lyn reside Stable recruitment of the BCR to rafts facilitates Lyn-medrated phosphorylation of ITAMs within the cytoplasmic tails of the Ig-u and Ig-B accessory molecules that initiate the BCR-driven signaling cascade

ment, Lyn then adds phosphate groups to the tyrosine residues within the ITAMs of the cytoplasmic tails of the lg-u/F complex. This is rapidly followed by binding of the PTK Syk to the ITAMs along with another kinase Btk (Bruton's tyrosine kinase). Active Lyn also phosphorylatesresidueswithin the CD19 coreceptormolecule that initiate signals which reinforce those initiated by the BCR (figure8.16). Syk fulfills a critical role within the B-cell activation process; disruption of the gene encoding Syk in the mouse has profound effects on downstream events in Bcell signaling and results in defective B-cell development. In this respect,Syk servesa similar role in B-cells to that servedby ZAP-7}in T-cells.Active Syk phosphorylates and recruits BLNK (B-cell linker; also called SLP65, BASH and BCA) to the BCR complex. Upon phosphorylation by Syk, BLNK provides binding sites for phospholipase Cyz (PLC^/2),Btk and Vav. Recruitment of Btk in close proximity to PLC/2 enablesBtk to phosphorylate the latter and increase its activity. Just as in the T-cell signaling pathway, activated PLC{2 initiatesa pathway that involves hydrolysis of PIP'to generate diacylglycerol and inositol triphosphate and results in increasesin intracellular calcium and PKC activation (figure 8.16).PKC activation, in turn, results in activation of the NFrB and JNK transcription factors and increased intracellular calcium results in NFAT activation, just as it does in T-cells. The Vav family of guanine nucleotide exchange factors consistsof at least three isoforms (Vav-1, -2 and -3) and is known to play a crucial role in B-cell signaling through activation of Racl and regulating cytoskeletal changes after BCR crosslinking; Vav-1 deficient B-cells

Figure 8.16. Signaling cascadedownstream of antigen-driven B-cell receptor ligation. Upon interaction with antigen, the BCR is recruited to lipid rafts where ITAMs within the Ig-u/p heterodimer become phosphorylatedby Lyn. This is followed by recruitment and activation of the Syk and Btk kinases. Phosphorylaiion of the B-cell adaptor protein, BLNK, createsbinding sites for several other proteins, including PLC/2. which promotes PIP, hydrolysis and instigates a chain of signaling events culminating in activation of the NFAT and NFrB transcription factors The CD19 coreceptor molecule is also phosphorylated by Lyn and can suppress the inhibitory effects of GSK3 on NFAT through the PI3K/Akt pathway. BCR stimulation also results in rearrangement of the cell cytoskeleton via activation of Vav which acts as a guanine nucleotide exchange factor for small G-proteins such as Rac and Rho

are defective in proliferation associatedwith crosslinking of the BCR. Vav is also recruited to the CD19 coreceptor molecule upon phosPhorylation by Lyn and, along with PI3K which is also recruited to CD19 as a result of Lyn-mediated phosphorylation, plays a role in the activation of the serine/threonine kinase AkU the latter may also enhance NFAT activation through neutralizing the inhibitory effects of GSK3 (glycogen synthase kinase 3) on NFAT. BecauseGSK3 can also phosphorylate and destabilize Myc and Cyclin D, which are essential for cell cycle entry, Akt activation also has positive effects on proliferation of activated B-cell (figure 8.16). The BCR cross-linking model seems appropriate for an understanding of stimulation by type 2 thymusindependent antigens, since their repeating determi-

C H A P T E R8 - L Y M P H O C Y T EA C T I V A T I O N nants ensure strong binding to, and cross-linking of, multiple Ig receptorson the B-cell surfaceto form aggregates which persist owing to the long halflife of the antigen and sustain the high intracellular calcium needed for activation. On the other hand, type 1 Tindependent antigens, like the T-cell polyclonal activators, probably bypass the specific receptor and act directly on downstream molecules such as diacylglycerol and protein kinase C since Ig-u and Ig-B are not phosphorylated.

DompingdownB-celloctivotion Several cell surface receptors, including FcyRIIB, CD22 and PIRB (paired immunoglobulinlike receptor B), have been implicated in antagonizing B-cell activation through recruitment of the protein tyrosine phosphatase SHP-I to ITIMs (immunoreceptor tyrosinebased lnhibitory motifs) in their cytoplasmic tails. SHP-1 impairs BCR signaling by antagonizing the effects of the Lyn kinase on Syk and Btk; by dephosphorylating both of these proteins SHP-1 blocks recruitment of PLC"(2 to the BCR complex. Coligation of the BCR with any of thesereceptorsis therefore likely to block Bcell activation.

T-helper cellsoclivoterestingB-cells T-dependent antigens pose a different problem, since

lmmunocompelenlI- ond B-cells differ in mony respects o The antigen-specific receptors, TCR/CD3 on T-cells and surface Ig on B-cells, provide a clear distinction between these two cell types. . T- and B-cells differ in their receptors for C3d, IgG and certain viruses. . There are distinct polyclonal activators of T-cells (PHA, anti-CD3) and of B-cells (anti-Ig, Epstein-Barr virus).

r83 |

they are usually univalent with respect to B-cell receptors, i.e. each epitope appears once on a monomeric protein, and as a resultthey cannot cross-link the surface Ig. Howevet we have discussed in some detail how antigen captured by a B-cellreceptor canbe internalized and processed for surface presentation as a peptide complexed with classII MHC (cf. figure 8.12).This can now be recognized by the TCR of a carrier-specific Thelper cell and, with the assistance of costimulatory signals arising from the interaction of CD40 with its ligand CD40L (cf. figure 8.12), B-cell activation is ensured. In effect, the B-lymphocyte is acting as an antigenpresenting cell and, as mentioned above, it is very efficient because of its ability to concentrate the antigen by focusing onto its surface Ig. Nonetheless, although a preactivated T-helper can mutually interact with and stimulate a resting B- cell, a resting T-cell can only be triggered by a B-cell that has acquired the 87 costimulator and this is only present on activated, not resting, B-cells. Presumably the immune complexes on follicular dendritic cells in germinal centers of secondary follicles can be taken up by the B-cells for presentation to T-helpers, but, additionally, the complexes could cross-link the slg of the B-cell blasts and drive their proliferation in a Tindependent manner. This would be enhanced by the presenceof C3 in the complexessincethe B-cell complement receptor (CR2)is comitogenic.

recruitment of the TCR to lipid rafts where many membrane-associated signaling proteins reside. Aclivolion of T.cells requirestwo signols . TWo signals activate T-cells, but one alone produces unresponsiveness (anergy). . One signal is provided by the low affinity cognate TCR-MHC plus peptide interaction. . The second costimulatory signal is mediated through ligation of CD28 byBT andgreatly amplifies signals generated

T-fymphocyles ond onligen-presenling cells inleroclthrough poirs of occessorymolecules . The docking of T-cells and APCs depends upon strong mutual interactions between complementary molecular pairs ontheir surfaces: MHC II{D4, MHC I{D8, VCAM1-vLA-4, ICAM-1-LFA-1, LFA-3{D2, B7-CD28 (and CTLA-4). c B7-CTLA-4 interactions are inhibitory whereas B7-CD28

Proteintyrosinephosphorylotionis0neorlyevenlin T-ceilsignoling . The TCR does not possessany intrinsic enzymatic activity

interactions are stimulatory. CTLA-4 may antagonize the

butisassociatedwithaccessoryproteinsthatcanrecruitPTKs.

through TCR-MHC interactions. . Previously stimulated T-cells only require one signal, throughtheirTCRs,forefficientactivation.

(Continuedp 184)

I ra+

ACTIVATION C H A P I ER 8 - L Y M P H O C Y T E

. The TCR signal is transduced and amplified through a protein tyrosine kinase (PTK) enzymic cascade. . Recruitment of CD4 or CD8 to the TCR complex leads to phosphorylation of ITAM sequences on CD3-associated ( chainsby the CD4-associated LckPTK. Thephosphorylated ITAMs bind and then activatetheZAP-70 kinase. DownslreomevenlsfollowingTCRsignqling . Nonenzymic adaptor proteins form multimeric

complexes with kinases and guanine nucleotide exchange factors (GEFs). . Hydrolysis of phosphatidylinositol diphosphate by phospholipase Cy1 or C{2 produces inositol triphosphate (IP.) and diacylglycerol (DAG). . IPs mobilizes intracellular calcium. . DAG and increased calcium activate protein kinase C. r The raised calcium together with calmodulin also stimulates calcineurin activity. o Activation of Ras by the guanine nucleotide exchange factor Sos sets off a kinase cascade operating through Raf, the MAP kinase kinase MEK and the MAP kinase ERK CD28 through PI3 kinase can also influence MAP kinase. . The transcription factors Fos and Jun, NFAT and NFrB are activated by MAP kinase, calcineurin and PKC, respectively, and bind to regulatory sites in the IL-2 promoter reglon. . A small number of MHC-peptide complexes can serially trigger a much larger number of TCRs thereby providing the sustained signal required for activation. . Initial binding of integrins facilitates the formation of an immunological synapse, the core of which exchanges integrins for TCR interacting with MHC-peptide.

F U R T H ERRE A D I N G Abraham R.T. & Weiss A (2004) Jurkat T-cells and development of the T-cell receptor signaling paradigm. Nature Reztiewslmmunology 4,301-308. Acuto O & Cantrell D. (2000) T-cell activation and the cytoskeleton Annual Reaiezu of lmmunology 18,165-184 Acuto O & Michel F (2003) CD28-mediated costimulation: a quantitative support for TCR signaling Noture Reztiewslmmunology 3, 939-957 Bromley S.K , Burack W R , Johnson K G et al (2007) The rmmunolo gical synapse . An nual Reaiew of I mmu nology 19, 37 5-39 6. Collins M ef al (2005)The B7 family of immune-regulatory ligands. CenomeBiology6,223. Grakoui A et al (1999) The immunologicai synapse: a molecular machine controlling T-cell activ atton Science285, 227-227

. Cbl family adaptor molecules are involved in negative signaling pathways. . The phosphatase domains on CD45 are required to remove phosphates at inhibitory sites on kinases. B-cellsrespondto lhree diffetenflypesof 0nligen . T)?e 1 thymus-independent antigens are polyclonal activators focused onto the specific B-cells by slg receptors. . Type 2 thymus-independent antigens are Polymeric molecules which cross-link many slg receptors and, because of their long half-lives, provide a persistent signal to the B-cell. . Thymus-dependent antigens require the cooperation of helper T-cells to stimulate antibody productionby B-cells. o Antigen captured by specific slg receptors is taken into the B-cell, processed and expressed on the surface as a peptide in association with MHC II. . This complex is recognized by the T-helper cell which activates the resting B-cell. . The ability of protein carriers to enable the antibody response to haptens is explained by T-B collaboration, with T-cells recognizing the carrier and B-cells the hapten. Ihe nolureof B-celloclivolion o Cross-linking of surface Ig receptors (e.g. by type 2 thymus-independent antigens) activates B-cells. . T-helper cells activate resting B-cells through TCR recognition of MHC Il-carrier peptide complexes and costimulation through CD40L-CD40 interactions (analogous to the B7-CD28 second signal for T-cell activation). . B-cell costimulation is also provided by the B-cell coreceptor complex consisting of CDl9,CD21,,CD81 and LEU13'

Jenkins M.K., Khoruts A ,Ingu'lliE. et al (2001) In vivo activation of antigen-specific CD4 T-cells. Annual Reaiew of Immunology 19, 2345 Kinashi T. (2005) Intracellular signaling controlling integrin activation in lymphoc ytes . N ature ReaiewsImmunology 5, 546-559 . Kurosaki T (2002) Regulation of B-cell signal transduction by adaptor proteins Nature Reztiewslmmunology 2'354-363. An Atlas of Biochemistry Michel G. (ed.) (1999)BiochemicalPathzuays: ond MolecularBlology.John Wiley & Sons,New York. Niiro H. & Clark E A (2002) Regulation of B-cell fate by antigenreceptor signal s N ature Rettiewslmmunology 2, 9 45-956 Olson M.F. & Marais R (2000) Ras protein signaling. Seminars in lmmunology12,63-73 Schraven B. et al (1'999)Integration of receptor-mediated signals in T-ce1lsby transmembrane adaptor proteins lmmunology Todny20, 431434

Theproduction ofeffeclors

INIRODUCTION In the previous chapter we explored the requirements for successful T- and B-cell activation through their respective membrane receptors for antigen. Having crossed the threshold required for activation, a stimulated T-cell enters the cell division cycle and undergoes clonal proliferation and differentiation to effectors. A succession of genes are upregulated upon T-cell activation. Within the first half hour, nuclear transcription factors such as Fos,/]un and NFAI, which regulate interleukin-2 (IL-2) expression and the cellular protooncogenec-myc, areexpressed,but the next few hours seethe s1'nthesis of a range of soluble cytokines and their specific receptors.Much later we seemolecules like the transferrin receptor related to cell division and very late antigens such as the adhesion molecule VLA-1 which enables activated T-cells to bind to vascular endothelium at sites of infection. The effector functions that Tcells acquire include the abilityto activate macrophages, provision of cytokine-mediated help for antibody productionby B-cells, and the ability to eliminate virally infected targets by inducing apoptosis in such cells.As we shall see, activated T-cells can differentiate along somewhat different paths to produce distinct subsets of effector T-cells that secretedifferent cohorts of cytokines broadly tailored towards the nature of the pathogen (whether intracellular or extracellular) that initiated the response. Activated B-cells also enter the cell cycle and undergo clonal proliferation to swell their ranks. Some of the activated B-cells eventually differentiate into plasma cells that migrate to the bone marrow where they produce and secrete large amounts of antibody for relatively long periods. Here we will consider the main issues surrounding the acquisition of effector function by T- and

B- lymphocytes and the roles that effector lymphocytes play in the immune response.

AS CIAS CYTOKINE I N I E R C E T T U TMAERS S E N G E R S Whereas the initial activation of T-cells and Tdependent B-cells involves intimate contact with dendritic cells (DCs), subsequent proliferation and maturation of the response are orchestrated by secreted proteins, termed cytokines (figure 9.1). In essence/ cytokines are messenger molecules that can communicate signals from one cell type to another and, amongst other things, can instruct the cell receiving the signal to proliferate, differentiate, secrete additional cytokines, migrate or die. To date, many different cytokines have been described and no doubt some remain to be discovered (table 9.1). One of the most important cytokine groupings, to the immunologist's way of thinking, is the interleukin family as this contains cytokines that act as communicators between leukocytes. Members of the interleukin family are somewhat diverse, belonging to different structural classes,because the primary qualification for membership of this family is biological with activity on leukocytes rather than sequenceor structural hom*ology. Indeed, while additional hom*ologs of the interleukin family are known, their status as interleukins awaits evidence that these proteins exert functional effects upon leukocytes. Approximately 30 interleukins have been described to date (IL-1 to IL-33) with the status of IL-14 as an interleukin in doubt. Other cytokine families have been established on the basis of their ability to support proliferation of hematopoietic precursors (colony stimulating factors), or cytotoxic activity towards transformed cell types (tumor necrosis factors), or the ability to interfere with

I rao

OF EFFECTORS 9-THEPRODUCTION CHAPTER

B-SHEET CYTOKINES

o-HELICAL CYTOKINES (co,l 5 oo) Shorl(' helices

(co 25 oo) LongcI'helices Exomples: TNF LT

Figure 9.1. Cytokine structures. Cytokines can be divided into a number of different structural groups Illustrated here are three of the main types of structure and some named examplesof eachtype: (a) four short (-15 amino acids) o-helices,(b) four long (-25 amino acids) cr-helicesand (c) a p-sheet JohnWiley& structure (ReproducedwithpermissionfromMichalG (ed)(1999) BlocheticnlPathzuays:AnAtlasofBiochemistryandMolecularBiology Sons,New York )

viral replication (interferons). It is important to note, however, that cytokines often do much more than their somewhat descriptive (and often misleading) names would suggest. Indeed, the response that these molecules elicit depends, to a large extent, on the context in which the cytokine signal is delivered. Thus, factors such as the differentiation stage of the cell, its position within the cell cycle (whether quiescentor proliferating) and the presence of other cytokines, can all influence the responsemade to a particular cytokine.

Cylokineoclionis tronsientondusuollyshorlr0nge Cytokines are typically low molecular weight (1525 kDa) secreted proteins that mediate cell division, inflammation, immunity, differentiation, migration and repair. Becausethey regulate the amplitude and duration of the immune-inflammatory responses,cytokines mustbe produced in a transient manner tightly coupled to the presenceof foreign material. Cytokine production can also occur in responseto the releaseof endogenous 'danger signals' that betray the presenceof cells dying by necrosis,a mode of cell death that is typically seenin pathological situations and is typically provoked by infectious agents or tissue injury. It is relevant that the AU-rich sequencesin the 3'untranslated regions of the mRNA of many cytokines prime thesemRNAs for rapid degradation thereby ensuring that cytokine production rapidly declines in the absenceof appropriate stimulation. Unlike endocrine hormones, the majority of cvtokines normallv act locallv in a paracrine or even

autocrine fashion. Thus cytokines derived from lymphocytes rarely persist in the circulation, but nonlymphoid cells can be triggered by bacterial products to releasecytokines which may be detected in the bloodstream, often to the detriment of the host. Septic shock, for example, is a life-threatening condition that largely results from massive overproduction of cytokines such as TNF and IL-1 in response to bacterial infection and highlights the necessity to keep a tight rein on cytokine production. Certain cytokines, including IL-1 and tumor necrosis factor (TNF), also exist as membraneanchored forms and can exert their stimulatory effects without becoming soluble. Cylokinesocl throughcell surtoceleceplors Cytokines are highly potent, often acting at femtomolar (10-1sv) concentrations,combining with smallnumbers of high affinity cell surfacereceptorsto produce changes in the pattern of RNA and protein synthesis in the cells they act upon. Cytokine receptors typically possess specific protein-protein interaction domains or phosphorylation motifs within their cytoplasmic tails to facilitate recruitment of appropriate adaptor proteins upon receptor stimulation. A recurring theme in cytokine receptor activation pathways is the ligandinduced dimer- or trimerization of receptor subunits; this facilitates signal propagation into the cell through the interplay of the transiently associatedreceptor cytoplasmic tails. There are six major cytokine receptor structural families (figure 9.2).

Table 9.1. Cytokines: their origin and function.

CYTOKINE

lo, lL-1p

SOURCE

EFFECTOR FUNCTION

rL-33

produclion including lL-2ondits T octivolion of cytokines Coslimulotes byenhoncing NKcytoloxicity; induces lL-1,-6,-8, TNF, receplor; enhonces B proliferolion ondmolurolion; byinducing chemokines ondICAM-lond GM-CSF ondPGE,byMQ;proinflommolory fever,APBboneresorplion byosteoclosls VCAM-lonendothelium; induces proliferolion Thl T-ondB-cells; enhonces NKcylotoxicily ondkilling of lumor Induces of octivoled ondMO cellsondbocferio bymonocyles precursors; MCgrowlh Growth onddifferentiotion of hemotopoielic I NK,MC proliferolion Ih2,Ic2,NK,NKIy6I MC B,I MC;upregulotes MHCclossII on lnduces Th2cells;sfimuloles of oclivoted inhibils Thl lL-'l2production ondlhereby B ondMO,ondCD23onB;downregulotes jnduces swilcht0 lgcl ondlgE increoses M0phogocylosis; differenliotion; proliferotion Th2,[/C B; induces switchlo lgA Induces ol eosinoondoclivoted Th2,I/ono,M0,DC,B[/ stromo APP;enhonces T Differentiolion slemcellsondof B intoolosmocells;induces of mveloid proliferolion BMondthymicstromo molureT T ondB;oclivoles Induces differenliolion of lymphoid slemcellsintoprogenitor Mono,M0,Endo of neutrophils Mediotes chemoloxis ondoclivolion proliferotion Th MCgrowth;synergizes wilhlL-4in swilchl0 Induces ol thymocytes; enhonces lgcl ondlgE Th(Th2in mouse), MHCclossII Tc,B,Mono,MQ Inhibits Thl cells;downregulofes lFNysecretion bymouse, ondlL-2byhumon, (including inhibiling Thl lL-12)production bymono,M0ondDC,fhereby ondcylokine B differentiolion inhibits T proliferolion; enh0nces ditferentiolion; BMslromo APP induces Promoles differentiolion of pro-Bondmegokoryocyles; proliferolion produclion Mono,MQ,DC,B induces ondIFNY byThl, Crilicol cytokine forThl ditferenfiotion; NKondCD8*T cyloloxicily CD8+ondy6T ondNK;enhonces Th2,MC B proliferolion; upregulotes Inhibils ondcylokine secrelion byM0;co-oclivoies octivolion VCAM-l swilchlo lgcl ondlgE;induces IVIHC clossII ondCD23onB ondmono;induces onendo production proliferolion ondcytotoxicily in NK B ondcytokine Induces of T-,NKondoctivoled I NK,Mono,MO,DC,B growtho1inleslinol epilhelium forT sfimulotes ondCD8+I chemotoclic Th,TC induces MHCclossII forCD4I monoondeosino; Chemooflroctonl production T including TNF,IL-l G-CSF of cylokines Proinflommolory; slimuloles B,-6,-8, MO,DC NKcytoloxicity Induces lFNyproduclion byI enhonces Mono Modulotion of Thl octivily Mono,Kerolinocyles to skin Regulotion responses of inflommotory Th T costimul0ti0n NKdifferenliotion; B oclivotion; Regulotion of hemolopoiesis; T Inhibils lL-4production byTh2 proliferolion proliferotion DC of memory cells byThl; induces Induces ondlFNyproduclion Th2,I/ono,M0 Induclion octivily of TNF,lL-1,lL-6,onti-tumor polhologies Thl,MO,I/osl Induclion ot lL-4,lL-5,lL-13ondTh2-ossocioted production Enhonced of lL-8ondlL-,l0byepilhelium INK produclion DC,Mono IFN-y Induclion of THI responses, enhonced Mono,DC inhibilion of virolreplicolion TypeI IFN-like oclivify, MOno, DC inhibilion of virolreplicolion TypeI IFN-like oclivity, T responses in skin Promoles inflommotory NK,T Promotes inllommotion, Rolein octivolion-induced T-celloooolosis DC,MO lnduclion of lL-4.lL-5.lL-l3

GM-CSF G-CSF M-CSF SLF

Th,M0,Fibro, MC,Endo Fibro, Endo Fibro, Endo, Epith BMsfromo

growthof progenitors M0 eosinoondboso;oclivotes ol mono,neutro, Slimuloles growthol neutroprogenitors Slimuloles growlhof m0n0progenilors Slimuloles (c-kltligond) stemcelldivision Slimuloles

TNFflNFc)

Th,Mono,M0,DC,MC,NK,B

Lymphofoxin ONFp)

ThI, TC

(weighlloss);induces induces E-seleclin 0n cylokine secretion; Tumorcyfotoxicily; cochexio endo;octivoles M0;onlivirol phogocylosis in lymphoid 0rg0n by neutro ondMO;involved Tumor enhonces cylotoxicity; develo0menl onlivirol

lFNa lFNy

Leukocyles Fibroblosls T h l ,T c l ,N K

MHCclossI Inhibits virolreDlicolion; enhonces enhonces MHCclossI hhibitsvirolreolicofion; M0;induces swilchl0 lgc20 MHCclossI ondII; ocfivoles Inhibils virolreolicolion; enhonces of Th2 lL-4oclions;inhibilsproliferolion onlogonizes sever0l

TGFp

Th3,B,M0,[/C

LIF Eto-I oncostolin M

Thymic epith,BMstromo T IMO

by, of monoondMQbutolsoonli-inflommolory Proinflommolory by,e,g.chemoottroction proliferolion; tissuerepoir induces swilchlo lgA;promofes lymphocyte e,g,inhibiting Induces APP lL-10produclion byMQ ondinhibils Stimuloles lL-12produclion lnduces APP

tL-2 tL-3 tL-4 tL-5 tL-6 rL-7 tL-8 tL-9

rL-l0 rL-lI tL-12 tL-l3 tL-l5 tL-t6 tL-l7 tL-18 rL-lI tL-20 tL-21 tL-22 tL-23 tG24 tL-25 rL-26 tL-27 tL-28 tL-29 tL-31 tL-32

rFNp

Mono,M0,DC,NK,B, Endo

APP, acute phase proteins; B, B-cell; baso,basophil; BM, bone marrow; Endo, endothelium; eosino, eosinophil; Epith, epithelium; Fibro, fibroblas! GM-CSE granulocyte-macrophage colony-stimulating factor; IL, interleukin; LIF, leukemia inhibitory factor; MQ, macrophage; MC, mast cell; Mono, monocyte; neutro, neutrophil; NI! natural killer; SLF,steel

locus factor; T, T-cell; TGFp, transforming growth factor-B.Note that there is not an interleukin-14. This designationwas given to an activity that, upon further investigation, could notbe unambiguously assigned to a single cytokine. IL-30 also awaits assignment IL-8 is a member of the chemokine family. These cytokines are listed separately in table 9.3

OF EFFECTORS 9-THEPRODUCTION CHAPTER

Hemolopoielin receptor fomily IL-2R

IFNreceptor fomily

TNFreceplor fomily TNFR

IFNYR

rss.F

receprors

I L - IR

Chemokine receplor fomily

TGF receplor fomily

IL-8R

eieeEEe ' 7 B

a< >>

OAU

Cyiosol Figure 9.2. Cytokine receptor families. One example is shown for each family. (a) The hematopoietin receptors operate through a common subunit (yc, Bc or gp130, depending on the subfamily)which transduces the signal to the interior of the cell In essence,binding of the cytokine to its receptor must initiate the signaling process by mediating hetero- or hom*odimer formation involving the common subunit In some casesthe cytokine is active when bound to the receptor either in soluble or membrane-bound form (e g IL-6) The IL-2 receptor is interesting with respect to its ligand binding The crchain (CD25, reacling with the Tac monoclonal) of the receptor possessestwo complement control protein structural domains and binds IL-2 with a low affinity; the pchain (CD122)hasamembraneproximalfibronectintype III structural domain and a membrane distal cytokine receptor structural domain, and associateswith the common ychain (CD132) which has a sirnilar structural organization The p chain binds IL-2 with intermediate affinity. IL-2 binds to and dissociates from the o chain very rapidly but the same processes involving the p chain occur at two or three orders of magnitude more slowly. When the a, p and y chains form a single receptor, the a chain binds the IL-2 rapidly and

facilitates its binding to a separate site on the p chain from which it can only dissociate slowly. Since the final affinity (K6) is based on the ratio of dissociation to association rate constants, then Ko = 16 + r-t 71Q7r,r-1s 1 = 10-11M, which is a very high affinity. The y chain does not itself bind IL-2 but contributes towards signal transduction (b)The interferon receptor family consists of heterodimeric molecules each of which bears two fibronectin type III domains. (c) The receptors for TNF and related molecules consist of a single polypeptide with four TNFR domains The receptor trimerizes upon ligand binding and, in common with a number of other recePtors, is also found in a soluble form which, when released from a cell following activation, can act as an antagonist (d) Another group of receptors contains varying numbers of Ig superfamily domains, whereas (e) chemokine receptors are members of the G-protein-coupled receptor superfamily and have seven hydrophobic transmembrane domains. (f) The final family illustrated are the TGF receptors which require association between two molecules, referred to as TGFR type I and TGFR type II,

Hemat op oietin receptors These are the largest family, sometimes referred to simply as the cytokine receptor superfamily, and are named after the first member of this family to be defined-the hematopoietin receptor. These receptors generally consist of one or two polypeptide chains responsible for cytokine binding and an additional shared (common or'c') chain involved in signal transduction. The yc (CD132)chain is used by the IL-2 receptor (figure 9.2a) andIL-4,IL-7,IL-9, IL-15 and IL-21 receptors/a Bc (CDw131) chainby IL-3,IL-s and granulocyte-macrophage colony-stimulating factor (GMCSF)receptors/and 9p130 (CD130) shared chainby the IL-6, IL-17, IL-12, oncostatin M, ciliary neurotrophic factor and leukemia inhibitory factor (LIF) receptors.

Interferonreceptors These also consist of two polypeptide chains and, in addition to the IFNcr, IFNB and IFNI receptors (figure 9.2b),this family includes the IL-10 receptor.

for signaling to occur.

TNF receptors Members of the TNF recePtor suPerfamily possesscysteine-rich extracellular domains and most likely exist as preformed trimers that undergo a conformational change in their intracellular domains uPon ligand binding. They include the tumor necrosis factor (TNF) receptor (figure 9.2c)and the related Fas(CD95IAPO-1) and TRAIL receptors. This family also contains the lymphotoxin (LT) and nerve growth factor (NGF) recePtors, as well as the CD40 recePtor, which plays an important

OF EFFECTORS 9-THEPRODUCTION CHAPTER

r8eI

role in costimulation of B-cells and dendritic cells by activated T-cells. lgSF cytokine receptors Immunoglobulin superfamily members are broadly lized in many aspects of cell biology (cf. p.252) include the IL-1 receptor (figure 9.2d), and macrophage colony-stimulating factor (M-CSF) stem cell factor (SCF/c-kit) receptors.

utiand the and

Chemokine receptors Chemokines share a common functional property of promoting chemotaxis and their receptors comprise a family of approximately 20 different G-protein-coupled, seven transmembrane segment polypeptides (figure 9.2e). Each receptor subtype is capable of binding multiple chemokines within the same family. For example, CXC receptor 2 (CXCR2)is capableof binding seven different ligands within the CXC ligand (CXCL) family. TGF receptors Receptors for transforming growth factors such as the TCFB receptor (figure 9.2f)possesscytoplasmic signaling domains with serine/threonine kinase activity. Sign0llr0nsduclionlhroughcylokine receptors The ligand-induced hom*o- or heterodimerization of cytokine receptor subunits represents a common theme for signaling by cytokines. The two major routes that are utilized are the Janus kinase (JAK)-STAT and the Ras-MAP kinase pathways. Members of the cytokine receptor superfamily (hematopoietin receptors) lack catalytic domains but are constitutively associatedwith one or more jAKs (figure 9.3).There are four members of the mammalian JA K f amily : jAK 1,JAK 2, J AK3 andTyk2 (tyrosine kinase 2) and all phosphorylate their downstream substrates at tyrosine residues. Genetic knockout studies have shown that the various JAKS have highly specific functions and produce lethal or severe phenotypes relating to defects in lymphoid development, failure of erythropoiesis and hypersensitivity to pathogens. Upon cytokine-induced receptor dimerization, JAKs reciprocally phosphorylate, and thereby activate, each other. Active JAKs then phosphorylate specific tyrosine residues on the receptor cytoplasmic tails to create docking sites for members of the STAT (signal fransducers and activators of franscription) family of SH2 domain-containing transcription factors. STAIs reside in the cytoplasm in an inactive state but, upon recruitment to cytokine receptors (via their SH2 domains),

Figure 9.3. Cytokine receptor-mediated pathways for gene transcription. Cytokine-induced receptor oligomerization activates JAK kinases that are constitutively associated with the receptor cytoplasmic tails Upon activation, JAK kinases phosphorylate tyrosine residues within the receptor tails, thereby creating binding sites for STAT transcription factors which thenbecome recruited to the recePtor complex and are, in turn, phosphorylatedbyJAKs. Phosphorylation of STAIs triggers their dissociation from the receptor and promotes the formation of STAT dimers which translocate to the nucleus to direct transcription of genes that have the appropriate binding motifs within therr promoter regions Members of the SOCS family of inhibitors can suppress cytokine signaling at several points, either through inhibition of JAK kinase activity directly or by promoting polyubiquitination and proteasome-mediated degradation of JAKs The PIAS family of STAT inhibitors can form complexes with STAT proteins that either result in decreased STAT-binding to DNA or recruitment of transcriptional corepressors that can block STAT-mediated transcription. Cytokine receptors can also recruit additional adaptor proteins such as Shc, Grb2 and Sos,that can activate the MAPkinase (seefigure 8.7) and PI3 kinase signaling cascades,but these havebeen omitted for clariry

become phosphorylated by |AKs and undergo dimerization and dissociation from the receptor. The dimerized STATs then translocate to the nucleus where they play an important role in pushing the cell through the mitotic cycle by activating transcription of various genes (figure 9.3). Seven mammalian STATs have been described and each plays a relatively nonredundant role in distinct cytokine signaling pathways. Individual cytokines usually employ more than one type of STAI to exert their biological effects; this is because the hematopoietin receptors are composed of two different receptor chains that are capable of recruiting distinct

CHAPTER 9-THEPRODUCTION OF EFFECTORS STAT proteins. Further complexity is achieved due to the ability of STATs to form heterodimers with each other, with the result that a single cytokine may exert its transcriptional effects via a battery of STAT combinations. ]AKs may also act through src family kinases to generate other transcription factors via the Ras-MAP kinase route (see figure 8.7). Some cytokines also activate phosphatidylinositol 3-kinase (PI3K) and phospholipaseC (PLCy). Downregulation of JAK-STAT signaling is achieved by proteins that belong to the SOCS (suppressor of cytokine signaling) and PIAS (protein lnhibitor of actrvated STAT) families (figure 9.3). SOCS proteins are induced in a STAl-dependent manner and therefore represent a classical feedback inhibition mechanism where cytokine signals induce expression of proteins that dampen down their own signaling cascades.The SOCS family contains eight members (namely CIS and SOCSI-SOCS7),and theseproteins utilize two distinct mechanisms to downregulate cytokine signals. On the one hand, SOCSproteins can interact with JAKs, as well as other signaling proteins such as Vav, and target these proteins for degradation by the ubiquitin-proteasome pathway (cf. p. 98).Alternatively, SOCS family proteins can interact with SH2-domain binding sites found within the activation loop of the JAK kinase domains, therebyblocking accessof jAKs to their downstream substrates (figure 9.3). Some SOCS family members, such as CIS (cytokine-lnducible src hom*ology domain 2 [SH2]-containing), can also directly interact withthe STAl-bindingSH2 domains found on cytokine receptors and by doing so can block recruitment of STAT molecules to the receptor complex. Targeteddeletion of SOCSgenesin the mouse has revealed the importance of these proteins for normal cytokine signaling. SOCS-I.deficient mice display marked growth retardation and lymphocytopenia and die from inflammationassociated multi-organ failure within 3 weeks of birth. Consistent with the role of SOCS proteins as negative regulators of cytokine signaling, lymphocytes derived from SOCS-l-deficient mice undergo spontaneous activation even in pathogen-free conditions. SOCS-Ideficient mice generated on a RAG2-deficient background do not display any of the phenotypes observed on a normal genetic background, confirming that SOCS1 exerts its effects primarily within the lymphocyte compartment. The PIAS family consists of four members (PIAS1, PIAS3, PIASX and PIASY) and can act to repress STATinduced transcriptional activity by interacting with these proteins to either restrict their ability to interact with the DNApromoter elements they associatewith, or alternatively, by recruiting transcriptional corepressor

proteins such as histone deacetylase to the STAI transcriptional complexes (figure 9.3). JAK-STAT pathways can also be regulated by other mechanisms such as protein tyrosine phosphatasemediated antagonism of JAK activity, for example. Cylokinesoflenhovemullipleetfects In general, cytokines are pleiotropic, i.e. exhibit multiple effectson a variety of cell types (table 9.1),and there is considerable overlap and redundancy between them with respect to individual functions, partially accounted for by the sharing of receptor components and the utilization of common transcription factors. For example, many of the biological activities of lL-4 overlap with those of IL-13. However, it should be pointed out that virtually all cytokines have at least some unique properties. Their roles in the generation of T- and B-cell effectors, and in the regulation of chronic inflammatory reactions (figure 9.4a,b),will be discussed at length later in this chapter. We should note here the important role of cytokines in the control of hematopoiesis (figure 9.4c). The differentiation of stem cells to become the formed elements of blood within the environment of the bone marrow is carefullynurtured throughthe production of cytokines by the stromal cells. These include GM-CSF, C-CSF (granulocyte colony-stimulating factor), M-CSF, IL-6 and -7 and LIF (table 9.1), and many of them are also derived from T-cells and macrophages. It is not surprising therefore that, during a period of chronic inflammation, the cytokines that are produced recruit new precursors into the hematopoietic differentiation pathway - a useful exercisein the circ*mstances. One of the cytokines,IL-3, should be highlighted for its exceptional ability to support the early cells in this pathway, particularly in synergywith IL-6 and C-CSF.

Networkinleroclions The complex and integrated relationships between the different cytokines are mediated through cellular events. The genes forIL-3, -4 and -5 and GM-CSF are all tightly linked on chromosome 5 in a region containing genes for M-CSF and its receptor and several other growth factors and receptors. Interaction may occur through a cascade in which one cytokine induces the production of another, through transmodulation of the receptor for another cytokine and through synergism or antagonism of two cytokines acting on the same cell (figure 9.5). Because of the number of combinations that are possible and the almost yearly discovery of new cytokines, the meansby which target cells integrate

OF EFFECTORS CHAPTER 9-THEPRODUCTION

growth Conlrol of lymphocyte

rer I

and interpret the complex patterns of stimuli induced by these multiple soluble factors is only slowly unfolding.

C A NM A K E I.C DIFFEREN I Ett SUBSEIS C IY T O K I NPEA T T E R N S DIFFEREN ThebipolorThl/Ih2 concept Aclivolion of innoleimmune mechonisms

HemoloDoiesis in bonemorrow STRO[/AL CELL

FIBROBLAST ORENDOTHELIAL CELL Figure 9.4. Cytokine action. A general but not entirely comprehensive guide to indicate the scope of cytokine interactions (e g for reasons of simplicity we have omitted the inhibitory effects of IL-10 on monocytes and the activation of NK cells by IL-12) EOSINQ, eosinophil; LAK, lymphokine-actir.ated killer; MQ, rnacrophage; NK, natural killer cell; PMN, polymorphonuclear neutrophil

Helper T-cell clones can be divided into two main phenotypes, Th1 and Th2, with each displaying distinct cytokine secretion profiles (table 9.2). This makes biological sensein that Th1 cells producing cytokines such as IFNT would be especially effective against intracellular infections with viruses and organisms which grow in macrophages,whereas Th2 cells are very good helpers for B-cells and would seem to be adapted for defense against parasites and other extracellular pathogens which are vulnerable to Il-4-switched IgE, Il-S-induced eosinophilia and IL-3 / 4-stimulated mast cell proliferation. Thus, studies on the infection of mice with the pathogenic proto zoan Leishmaniamaior demon' strated that intravenous or intraperitoneal injection of killed promastigotes leads to protection against challenge with live parasites associatedwith high expression of IFNy mRNA and low levels of IL-4 mRNA; the reciprocal finding of low IFNyand high IL-4 expression was made after subcutaneous immunization which failed to provide protection. Furthermore, nonvaccinated mice infected with live organisms could be saved by injection of IFNyand anti-Il-4. These results are consistent with the preferential expansion of a population of protective IFNy-secretingTh1 cellsby intraperitoneal or intravenous immunization, and of nonprotectiveTh2 cells producing IL-4 in the subcutaneously injected animals. The ability of IFNy, the characteristic Th1 cytokine, to inhibit the proliferation of Th2 clones, and of Th2-derivedlL-4 and -10 to block both proliferation and cytokine releaseby Th1 cells,would seemto put the issuebeyond reasonabledoubt (figure 9.6). The original Mosmann-Coffman classification into Th1 and Th2 subsetswas predicated on data obtained with clones which had been maintained in culture for long periods and might have been artifactsof conditions in aitro. The use of cytokine-specific monoclonal antibodies for intracellular fluorescent staining, and of ELISPOT assays (cf. p. 145) for the detection of the secretedmolecules,has demonstrated that the Th1/Th2 dichotomy is also apparent in freshly sampled cells and thus also applies in aiao.Nonetheless,it is perhaps best not to be too rigidly constrained in one's thinking by the Tlnl/Th2 paradigm, but rather to look upon activated T-cells as potentially producing a whole sPectrum of

I rsz

C H A P T E9 R- T H EP R O D U C T I OO NF E F F E C T O R S

Coscode

Receptor tronsmodulotion

,r-,oraroroo I

rL-2RECEPToR I

RECEPTOR

TNF..>

Synergism

Antogonism

GENES

Table9.2. Cytokine patterns of helper T-cell clones. Interleukin-10 is not listed in the table. Although classed as a Th2 cytokine in the mouse, it is produced byboth Th1 and Th2 cells in the human.

Th'l

Ih2

lFNl

tL-2 (TNFB) Lympholoxin TNFCINFa) GM-CSF I L-J

tL-4 tL-5 tL-6

tL-t3

Thl/2,T-helperl/2.

cytokine profiles (Th0, figure 9.6), with possible skewing of the responses towards the extreme Th1 and Th2 patterns depending on the nature of the antigen stimulus. Thus, other subsets may also exist, in particular the transforming growth factor-B GGFB) and IL-10producing Th3/TrI (T-regulatory 1)cells, which are of interest because these cvtokines can mediate immuno-

Figure 9.5. Network interactions of cytokines. (a) Cascade: in this example TNF induces secretion ofIL-1 and ofitself (autocrine) in the macrophage. (Note that all diagrams in this figure are simplified in that the effects on the nucleus are due to messengers resulting from the combination of cytokine with its surface receptor.) (b) Receptor transmodulation showing upregulation of each chain forming the high affinitylL-2 receptor in an activated T-cell by individual cytokines and downregulationby TGFp (c) Synergy of TNF and IFNyin upregulation of surface MHC class II molecules on cultured pancreatic insulin-secreting cells. (d) Antagonism of IL-4 and IFN^yon transcription of silent ('sterile') mRNArelating to isotype switch (cf. figure 9 17)

suppressive effects and may be involved in the induction of mucosally induced tolerance(cf. p. a51).

lnteroctionswith cellsof lhe innoleimmunesyslembioses rheThl/Ih2 lesponse The cytokine milieu thatbecomes establishedby cells of the innate immune system during the early stages of infection has a major influence on the adaptive immune response. Typically, innate immune responses dominate initially as T-lymphocytes require priming by DCs or other APCs to initiate clonal expansion and maturation to effectors. Upon migration of antigen-specific Tcells to lymph nodes where they come in contact with mature DCs fresh from their encounters with microbial pathogens, the pathogen products encountered by the DC will have polarized the latter in favor of secreting particular cytokines. This in turn can polarize T-cell responses in favor of differentiating towards a Th1 or Th2 phenotype. The reader will be well aware by now that efficient T-cell activation requires two signals from theAPC; signal 1 is providedby stimulation through the TCR and signal 2 is provided by costimulation through CD28. Polarizationof T-cellstowards aTh1,Th2 or other fate is achieved via signal 3 and the nature of this signal is strongly influenced by the conditions under which the APC is primed (figure 9.6). IL-12 and its recentlv discovered relatives. IL-23 and

9-THEPRODUCTION OF EFFECTORS CHAPTER

IhI -promoting pothogen producls

\ T L - ^\

lL-12 lL'27 cD80 ,/coeo \ \ r ccD86 D 8CD28 6cD28J I

)-'

Th2-promoling \ pdinoge." \ producls

/ /-./

\ tL-6 tL-t0 tL-l3

Humorol immunily Figure 9.6. The generation of Thl and Th2 CD4 subsets. Following initial stimulation of T-cells, a range of cells producing a spectrum of cytokine patterns emerges. Depending on the nature of the pathogen and the response of cells of the innate immune system during the initial stages of infection, the resulting T-helper cel1population can be biased towards two extremes Th1-promoting pathogen products (such as LPS) engage Tolllike receptors (TLRs) on dendritic cells (DC) or macrophages and induce the secretion of Thl-polarizing cytokines such as IL-12 and IL-27. The latter cytokines promote the development of Th1 cells which produce the cytokines characteristic of cell-mediated immunity IL-4, possibly produced by interaction of microorganisms with the lectinlike NK1 1* receptor on NKT-cells or through interaction of Th2-promoting pathogen products with TLRs on DCs, skews the development to the production of Th2 cells whose cytokines assist the progression of B-cells to antibody secretion and the provision of humoralimmunify. Cytokines produced by polarized Th1 and Th2 subpopulations are mubually inhibitory. LI, lymphotoxin (TNFF); Th0, early helper cell producing a spectrum of cytokines; other abbreviations as in table 9 1

IL-27, are instrumental in polarizing towards a Th1 cell phenotype while IL-4 is pivotal for the production of a Th2 cell phenotype. Polarization of T-cells towards an inducibleregulatoryphenotype (Th3/Tr1) can also occur during this phase,with IL-10 and TCFBbeing influential in this situation. Invasion of phagocytic cellsby intracelIular pathogens induces copious secretion of IL-12, which in turn stimulates IFNyproduction by NK cells. Engagement of many of the known Tolllike receptors (TLRs) on DCs by microbial products (such as LPS, dsRNA and bacterial DNA) triggers DC maturation and induces IL-12 production, thereby favoring Th1 responses.Bacterial priming also induces CD40 receptor expression on DCs and induces responsivenessto CD40L, expressedby activated T-cells,for optimal IL-12 slmthesis.IL-12 is also particularly effective at inducing IFN1 by activated T-cells and secretion of the latter by the T-cell further enhances IL-12 production and secretion by DCs; this acts as a classical positive feedback loop for enhancement of IL-12 production and further skews the responsetowards Th1.

r e3 |

While IL-12 and IFN^ypromote a Th1 response, these cytokines also inhibit Th2 responses (figure 9.6). However, IL-4 effectsappear to be dominant over IL-12 and therefore the amounts of IL-4 relative to the amounts of IL-1,2 and IFNy will be of paramount importance in determining the differentiation of ThO cells into Th1 or Th2. IL-4 downregulates the expression of the IL-12R B, subunit necessaryfor responsiveness to IL-72, further polarizing the Th2 dominance. It is still unclear whether signals from the innate immune system drive T-cells in the direction of a Th2 response or whether this is a default differentiation pathway for Th cells unless suppressedby Th1-polarizing signals such as IL-12 or IFNy. Aspecial cell population, the NKT-cells bearing the NK1.1* marker, rapidly releasesan IL-4dominated pattern of cytokines on stimulation. These cells have many unusual features.They may be CD4 8 or CD4+8 and expresslow levels of T-cell crBreceptors with an invariant cr chain and very restricted B, many of these receptors recognizing the nonclassical MHClike CD1 molecule. Their morphology and granule content are intermediate between T-cells and NK cells. Although they express TCRaB, there is an inclination to 'innate' immune classify them on the fringe of the system with regard to their primitive characteristics and possessionof the lectin-like NK1.1 receptor which may be involved in the recognition of microbial carbohydrates. Whilst there is a certain amount of evidence indicating the existence of subpopulations of dendritic cells specialized for the stimulation of either Th1 or Th2 populations, it seemsthat DCs are relatively plastic and can adopt a Th1- or Tln2-polarizing phenotype depending on the priming signals they encounter from microbial and tissue-derived sources. However, it should be obvious from the above discussion that the cytokines produced in the immediate vicinity of the T-cell will be important. To give one recent example, Cantor and colleagues hom*ologously deleted the gene for the cytokine Eta-1 (osteopontin) in mice and found that these animals had severely impaired immunity to infection with herpes simplex virus and to the intracellular bacterium Listeria monocytogenes.Thisw as due to a deficient Th1 immunity caused by reduced IL-72 and IFNy and enhanced IL-10 production. It appears that Eta-1 productionby activated T-cellsstimulates IL-12 production by macrophage lineage cells and downregulates IL-10 production. Interestingly, both serine phosphorylated and nonphosphorylated forms of Eta-1 are secreted by T-cells, and the IL-12 effect is phosphorylationdependent whereas the IL-10 effect is not, indicating that phosphorylation can regulate the activity of secretedproteins.

Regulotory T-cells dompen immune responses 0ndprotect ogoinsl ouloimmunity Another subset of T-cells, the naturally-occurring regulatory T-cells (Tregs), has been the subject of much attention in recent years. These cells are a population of CD25+ CD4+ T-cells that can suppress immune responses of autoreactive T-cells by mechanisms that are still poorly understood but do not seem to involve cytokines. These Tregs are a nntural, as opposed to an inducible,population of regulatory T-cells distinct from TGFp-secretingTh3 cellsand IL1O-secretingTi1 cells (cf. figure 10.10).The current view is that these self-antigenreactive T-cells develop in the thymus and are released as functionally mature cells that can act to suppress the activation of other self-reactive T-cells that escapenegative selection in the thymus, possibly through competition for self-antigens presented by APCs or through CTlA4-mediated signals from the Treg to theAPC. IL-2 is crucial for the maintenance of natural Tregs as these Tcells are incapable of making their own IL-2, unlike activated T-cells, and rely fully on paracrine IL-2 for their survival. Consequently, the number of suchcells is drastically reduced in IL-2 and IL-2R knockout mice, with the result that these mice develop lymphoproliferation followed by lethal autoimmunity. The source of IL-2for the maintenance of Tregs is unresolved but could come from autoreactive or antigen-activated T-cells that are interacting on the same DC as the Tieg. In contrast to natural regulatory T-cells, Th3/Tr1 cells are generated from naive T-cells in the periphery after encounter with antigen presented by DCs. The conditions under which DCs polarize towards Th3/Tr1cells remain uncertain but there is some evidence that such DCs have a semi-mature phenotype with low levels of CD40 and ICAM1. Pathogen molecules that induce IL-10 and suppress IL-12 production by DCs may be instrumental in evoking a Th3/Tr1 response.

ing activation is critically dependent upon IL-2 (figure 9.7). This cytokine is a single peptide of molecular weight 15.5kDa which acts only on cells which express high affinity IL-2 receptors (figure 9.2). These receptors are not present on resting cells, but are synthesized within a few hours after activation. Separation of an activated T-cell population into those with high and low affinity IL-2 receptors showed clearly that an adequate number of high affinity receptors were mandatory for the mitogenic action of IL-2. The numbers of these receptors on the cell increase under the action of antigen and of IL-2 and, as antigen is cleared, so the receptor numbers decline and, with that, the responsiveness to IL-2. It should be appreciated that, although IL-2 is an immunologically nonspecific T-cell growth factor, it only functions appropriately in specific responsesbecauseunstimulated T-cells do not express IL-2 receptors. The T-cell blasts also produce an impressive array of

lL-2 Synfhesis ondlL-2receplor IL-2RECEPTOR

lL-2driven proliferofion

CyloloxicT-cellsconolso be subdividedintoTcl/tc2 Clones of human cytotoxic T-cells obtained by limiting dilution also characteristically secrete particular cytokines. Thus, Tc1 cells secrete IFNy but not IL-4, whilst Tc2 cells secreteIL-4 but not IFNy. These clones show no differences in their cytolytic functionbut, when cocultured with CD4+ T-cells, Tc1 clones induce Th1 cells,whilst Tc2clonesinduce Th2 cells.

ACIIVAIED T - C E t t SP R O I . I F E R AI N TE R E S P O N SI O E CYIOKINES In so far as T-cells are concerned, amplification follow-

Figure 9.7. Activated T-blasts expressing surface receptors for IL-2 proliferate in response to IL-2 produced by itself or by another T-cell subset Expansion is controlled through downregulation of the IL-2 receptor by IL-2 itself The expanded population secretes a wide variety of biologically active cytokines of which IL-4 also enhances T-cell proliferation

CHAPTER 9-THEPRODUCTION OF EFFECTORS other cytokines, and the proliferative effect of IL-2 is reinforced by the action of IL-4 and, to some extent, IL-6, which react with corresponding receptors on the dividing T-cells. We must not lose sight of the importance of control mechanisms, and obvious candidates to subsume this role are TGFp, which blocks Il-2-induced proliferation (figure 9.5b) and the production of TNF (TNFa) and lymphotoxin (TNFp), and the cytokines IFNy, IL-4 and IL-12, which mediate the mutual antagonism ofThl and Th2 subsets.

I - C E t t E F F E C I O RI N S C E t t . M E D I A I EIDM M U N I T Y Cylokines mediole chtonic intlommotory responses In addition to their role in the adaptive response, T-cell cytokines are responsible for generating antigenspecific chronic inflammatory reactions which deal with intracellular parasites (figures 9.4b and 9.8), although there is a different emphasis on the pattern of factors involved (cf.p.282).

re5 |

Early eoents The initiating event is probably a local inflammatory response to tissue injury caused by the infectious agent which provokes the synthesis of adhesion molecules such as VCAM-1 (vascular cell adhesion molecule) and ICAM-1 on adjacent vascular endothelial cells. These adhesion molecules permit entry of memory T-cells to the infected site throughtheir VLA-4 and LFA-I homing receptors (cf. p. 159). Contact with processed antigen derived from the intracellular parasite will activate the specific T-cell and induce the release of secreted cytokines. TNF will further enhance the expression of endothelial accessory molecules and increase the chancesof other memory cells in the circulationhoming in to meet the antigen provoking inflammation. Chemotaxis The recruitment of T-cells and macrophages to the inflammatory site (figure 9.8) is greatly enhancedby the action of chemotactic cytokines termed chemokines (chemoattractantcytokine). These can be produced by a variety of cell types and are divided into four families

ANTIGEN

I L - I , 6 T N FI F N O G-CSFG[/-CSF

Figure 9.8. Cytokines controlling the antibody and T-cell-mediated inflammatory responses. Abbreviations as in table 9 1.

crorrrnp-rxepnooucrrox Os:Si

I rge

based on the disposition of the first (N-terminal) two of the four canonical cysteine residues (table 9.3). CXC chemokines have one amino acid and CX3C have three amino acids between the two cysteines. CC chemokines have adjacent cysteines at this location, whereas C chemokines lack cysteines 1 and 3 found in other chemokines. Chemokines bind to G-protein-coupled

seven transmembrane receptors (figure 9.2).Despite the fact that a single chemokine can sometimes bind to more than one receptor, and a single receptor can bind several chemokines, many chemokines exhibit a strong tissue and receptor specificity. They play important roles in inflammation, lymphoid organ development, cell trafficking, cellular compartmentalization within lym-

Thble 9.3. Chemokines and their receptors. The chemokines are grouped according to the arrangement of their cysteines (see text). The letter L designates ligand (i.e the individual chemokine), whereas the letter R designates receptors. Names in parentheses refer to the murine hom*ologs of the human chernokine where the names of these differ, or the murine chemokine alone if no human equivalenthasbeen described.

CXCLI CXCL2 CXCL3 CXCL4 CXCLS CXCL6 CXCLT CXCLS CXCL9 CXCLI O cxclt I

cxcLt2 cxcll3 cxcll4 CXCLI 5 cxclr6

CXCR2>CXCRI CXCR2 CXCR2 CXCR3-B CXCR2 CXCRI, CXCR2 CXCR2 CXCRI, CXCR2 CXCR3-A, CXCR3-B CXCR3-A, CXCR3-B CXCR3-B CXCR3-A, CXCR4 CXCRS DC,MOnO 2 CXCR6

GROc/MGSAcr GR0p/rvGSAp GRO"y/MGSAT PF4 ENA-78 GCP-2l(CKa-3) NAP-2 tL-8 Mig

rP-t 0 I-TAC SDF-l crlp BLC/BCA-I BRAC/Bolekine Lungkine None Lymphotoclin/SCM1a/ATA C

scM-r B

CCLI

r-309/(TCA-3/P500)

ccL2 ccL3

MCP-I/MCAF MIP-la/LD78a

CCL4 ccL5 (ccL6) ccLT ccLS (cclg/r0) cclt I (cclr2) cclt 3 cclr4 cclr5 cclr6 7 CCLI cclr8 ccLlI ccL20 ccl2r ccL22 CCL23 ccL24 ccL25 ccL26 ccL2l ccL28

RANTES (cr 0/MRP-r ) t\4cP-3 MCP-2 (MRP-2/CCFI 8/MlP-ly) EoloxinI (MCP-5) MCP-4 HCC-l /HCC-3 HCC-2lleukoloctinI /MlP-l6 HCC-4/LEC/(LCCl) TARC DCCKI /PARC/AMAC-I MIP-3p/ELC/Exodus-3 MIP-3a/LARC/ExodusI 6Ckine/SLC/Exodus-2/OCA-4) MDC/STCP-I /ABCD-I MPIF-I MPIF-2/Eotoxin-2 TECK SCYA26/Eoloxin-3 CTACK/ALP/ESKine IVIEC

p MrP-r

Mono T,NK,DC,Mono,Boso I, NK,DC,Mono,Eosino I NK,DC,Mono Boso Eosino, I NK,DC,lvlono, Mono,MQ,I Eosino Boso I NK,DC,Mono,Eosino, I NK,DC,Mono,Boso I Mono Boso I DC,Eosino, I, NK,DC,Mono,Boso Boso T,NK,DC,Mono,Eosino, I Mono,Eosino T T I DC,Mono lDc IB,DC DC IDC T,DC,I/ono T Boso I DC,Eosino, I DC,Mono T T T,B, Eosino

CCR8

ccR2 ccRr, ccRS ccR5 ccRr,ccR3, ccRs ccRt ccR3 ccRr,ccR2, ccR3 CCRI

ccR3 ccR2 ccR2, ccR3 ccRr ccRl,ccR3 ccRl ccR4 2

ccRT ccR6 ccRT ccR4 ccRl ccR3 ccR9 ccR3 0 ccRr 0 ccR3/ccRr

killelT,T-cell neulrophil; NK,noturol Neutro, Mono, monocyfe; Boso, chemokine; B,B-cell; bosophil; DC,dendritic cell;Eosino, eosinophil; MEC, mucosol epilheliol

CHAPTER 9 _ T H EP R O D U C T I O N OF EFFECTORS phoid tissues,Th1/TM development, angiogenesisand woundhealing. Macrophage actiaation Macrophages with intracellular organisms are activated by agents such as IFNy, GM-CSF,IL-z and TNF and should become endowed with microbicidal powers. During this process, some macrophages may die (probably helped along by cytotoxic T-cells) and release living parasites, but these will be dealt with by fresh macrophages brought to the site by chemotaxis and newly activated by local cytokines so that they have passed the stage of differentiation at which the intracellular parasites can subvert their killing mechanisms ( c f. p . 2 6 e ) . Co mb ating a ir al inf ecti on Virally infected cells require a different strategy and one strand of that strategy exploits the innate interferon mechanism to deny the virus accessto the cell's replicative machinery. IFNy, TNF and lymphotoxin all induce 2'-5' (A) synthetase, a protein which is involved in viral protection. TNF has another string to its bow in terms of its ability to kill certain cells, since death of an infected cell before viral replication has occurred is obviously beneficial to the host. The cytotoxic potential of TNF was first recognized using tumor cells as targets (hence the name), and IFNy and lymphotoxin can act synergistically, with IFNy setting up the cell for destruction by inducing the formation of TNF receptors. It is worth pointing out, however, that TNF kills by inducing apoptosis rather than necrosis in the majority of cases. KillerT-cells The generation of cytotoxic T-cells Cytotoxic T-cells (Tc),also referred to as cytotoxic T-lymphocytes (CTLs), represent the other major arm of the cell-mediated immune response and are of strategic importance in the killing of virally infected cells and possibly in contributing to the postulated surveillance mechanisms against cancercells (seep. 388). CTL precursors recognize antigen on the surface of cells in association with class I major histocompatibility complex (MHC) molecules and, like B-cells, they usually require help from T-cells. The mechanism by which help is proffered may, however, be quite different. As explained earlier (seep. 180),effective T-B collaboration is usually 'cognate' in that the collaborating cells recognize two epitopes that are physically linked (usually on the same molecule). If we may remind the readerwithout causing offense, the reasonfor this is that the surface Ig receptors on the B-cell capture native

r e7 |

antigen, process it internally and present it to the Th as a peptide in association with MHC class II. Although it has been shown that linked epitopes on the antigen are also necessaryfor cooperation between Th and the cytotoxic T-cell precursor (Tcp), the nature of T-cell recognition prevents native antigen being focused onto the Tcp by its receptor for subsequent processing, even if that cell were to express MHC II, which in its resting state it does not. It seemsmost likely that Th and Tcp bind to the same APC, for example a dendritic cell, which has processed viral antigen and displays processed viral peptides in associationwithboth classII (for the Th cell) and classI (for the Tcp) on its surface; one cannot exclude the possibility that the APC could be the virally infected cell itself. Cytokines from the triggered Th will be released in close proximity to the Tcp which is engaging the antigen-MHC signal and will be stimulated to proliferate and differentiate into a Tc under the influence of IL-2 and -6 (figure 9.9a). However, interaction of the APC with the Th and the Tc cell canbe temporally separated and, in this case, it appears that the helper T-cell 'licenses' the dendritic cell for future interaction with the cytotoxic T-cell. It does this by activating the dendritic cell through CD40, thereby upregulating costimulatory molecules and cytokine production, in particular lL-I2,by the dendritic cell (figure 9.9b). An entirely Th-independent mechanism of Tc activation is also thought to occur. This has been demonstrated in, for example, the response to protein antigens given with potent adjuvants such as immunostimulatory DNA sequences (ISSs), in this case possibly involving adjuvant-induced production of proinflammatory cytokines and cell surface costimulatory molecules. The lethalprocess Cytotoxic T-cells (Tc) are generally of the CD8 subset, and their binding to the target cell through TCRmediated recognition of peptide presented on class I MHC is assisted by interactions between CD8, the coreceptor for class I, and by other accessorymolecules such as LFA-1 and CD2 which increase the affinity of the interaction between the CTL and the target cell (seefigure 8.3). Tc are unusual secretory cells which contain modified lysosomes equipped with a battery of cytotoxic proteins. Following activation of the Tc, the cytotoxic granules are driven at a rare old speed (up to 7.2 ytrn/sec) along the microtubule system and delivered to the point of contact between the Tc and its target (figure 9.10). This ensures the specificity of killing dictated by TCR recognition of the target and limits collateral damage to surrounding cells, as well as to the killer cell itself. As with NK cells, which have comparable

OF EFFECTORS CHAPTER 9_IHE PRODUCTION

Denddliccell

ICR

cD40f--l

MHCII

Figure 9.9. T-helper cell activation of cytotoxic T-cells. Activation of the CD4* helper T-cells (Th) by the dendritic cell involves a CD40-CD40 ligand (CD154) costimulatory signal and recognition of an MHC class II peptide presented by the T-cell receptor (a) If both ihe Th and the cytotoxic T-lymphocyte (Tc) are present at the same time, the release of cytokines from the activated Th cells stimulates the differentiation of the CD8+ precursor into an activated, MHC class Irestricted Tc However, as shown in (b), the Th and the Tc donotneed to 'licenses' interact with the APC at the same time In this case,the Th cell the dendritic celi for future interaction with a Tc cell Thus the Th cell, by engaging CD40, drives the dendritic cell from a resting state into an activated state with upregulation of costimulatory molecules such as B7.1 and 87 2 (CD80 and CD85, respectively) and increased cytokine production, particularly of IL-12

granules (cf. p. 18),exocytosisof the cytotoxic granules delivers a range of cytotoxic proteins into the target cell cytosol that cooperate to promote apoptosis of the target. Videomicroscopy shows that Tc are serial killers. 'kiss of death', the T-cell can disengage and After the seek a further victim, there being raPid synthesis of new granules. Cytotoxic T-cell granules contain perforin, a poreformingprotein similar to the C9 component of complement, and a battery of cathepsin-like proteases that are collectively referred to as granzymes. Perforin facilitates the entry of the other granule constituents into the target cell in a manner that is still much debated, but all

Figure 9.10. Conjugation of a cytotoxic T-cell (on left) to its target, here a mouse mastocytoma, showing polarization of the granules towards the target at the point of contact The cytoskeletons of both cells are revealed by immunofluorescent staining with an antibody to tubulin (green) and the lytic granules with an antibody to granzyme A (red) Twenty minutes after conjugation the target cell cytoskeleton maystillbeintact (above),butthis rapidlybecomes disrupted (below). (Photographs kindly provided by Dr Gillian Griffiths )

are agreed that perforin Plays an essential role in the killing process; mice deficient in perforin are severely impaired in clearing several viral pathogens. It is not clear how all of the granzymes contribute to target cell death upon delivery into the cell cytoplasm but granzyme A and B are known to play particularly significant roles in this process. Granzyme A can Promote the activation of a nuclease through proteolysis of its inhibitor and this results in the formation of numerous single-stranded DNA breaks within the target cell (figure 9 .17). Granzyme B can directly process and activate several members of the caspasefamily of cysteine proteases that can rapidly initiate apoptosis through restricted proteolysis of hundreds of proteins within the target cell. Granzyme B can also Promote caspaseactivation indirectly, through activation of Bid, a protein that promotes permeabilization of mitochondria and release of mitochondrial cytochrome c into the cytosol; the latter

g-THEPRODUCTION CHAPTER OF EFFECTORS

gronules Cytoloxic

oo \A

,Perforin chonnels

.i:

,.\

e,liilirE=

GronzYmetc - 'it'o

----r

/.- - \

eqts' nonl'er

I +l {Apopfosis <-

cvmn../c =' poptosome

--/

Figure 9.11. Cytotoxic granule-dependent killing of target cells by cytotoxic T-cells and NK cells. In response to an appropriate stimulus, Tc and NK cells deliver the contents of their cytotoxrc granule onto the surfaceof targetcells Thecytotoxic granuleproteinperforinisthought to polymerize within the target cell membrane forming pores that perrnit passage of other granule constituents, which includes several serine proteases (granzymes), into the target ceil. Upon entry into the target, granzyme B orchestrates apoptosis by cleaving and activating BID, which translocates to mitochondria and triggers the opening of a pore or channel within the mitochondrial outer membrane composed ofBax and/or Bak; the latter channel permits the releaseof cytochrome c from the mitochondriai intermembrane space into the cytoplasm where it acts as a co-factor for the assembly of a caspase-9-activating complex (the apoptosome) The apoptosome promotes activation of downstream caspases,such as caspase-3and caspase-7,and the latter proteases coordinate apoptosis through restricted proteolysis of hundreds of substrate proteins Granzyme B can also proteolytically process and activate caspases-3 and -7direct1y, providinga more direct route to caspaseactivation Another granule protein, granzyme A, can cleave a protern within the SET complex (an endoplasmic reticulumassociated protein complex) This permits the translocation of a nuclease (NM23-H1) to the nuclear compartment that can catalyze single-strand DNA breaks Cytotoxic granules also contain other granzymes that contribute to target cell killing but substrates for these proteases have yet to be identified.

event arms a caspase-activatingcomplex that has been termed 'the apoptosome' and this complex promotes the activation of several downstream caspases (figure 9.11). Several additional granzymes have also been found within cytotoxic granules but their precise func-

reeI

tional role in Tc killing remains the subject of ongoing investigation. Collectively, entry of the full spectrum of granzymes into the target cells results in very swift cell killing (within 60 minutes or so) and several parallel pathways to apoptosis are most likely engaged during this process.Tc also expressproteaseinhibitors, such as PI-9, that may protect them from the lethal effects of their own granule contents. The induction of apoptosis,as opposed to necrosis,by the Tc is likely to have several benefits. Apoptotic cells, by virtue of specific alterations to their plasma membranes, are swiftly recognized by macrophages and other phagocytic cells and undergo phagocytosis before their intracellular contents can leak; this has the desirable effect of minimizing collateral damage to neighboring cells and may also prevent escape of viral particles from an infected cell. Moreover, nucleasesand caspase proteases that become activated within the target cell during apoptosis are also likely to degrade viral nucleic acids and structural proteins and may also contribute to ensuring that infectious viral particle releaseis kept to a minimum. Tc are also endowed with a second killing mechanism involving Fas and its ligand (cf. p. 19).In this situation, engagement of the trimeric Fas receptor by membraneborne Fas ligand on the Tc initiates a signaling pathway within the target cell that results in the recruitment and activation of caspase-8at the receptor complex. Upon activation, caspase-8 can further propagate the death signal through restricted proteolysis of Bid, similar to the granzyme B pathway discussed above, or can directly processand activate downstream caspases such as caspase-3. However, the inability of perforin knockout mice to clear viruses effectively suggests that the secretory granules provide the dominant means of killing virally infected cells. One should also not lose sight of the fact that CD8 cells slmthesize other cytokines such as IFNywhich also have potent antiviral effects. muslbe reguloted Inflommolion Once the inflammatory process has cleared the inciting agent,thebodyneeds to switch it off. IL-10has profound anti-inflammatory and immunoregulatory effects,acting on macrophages and Th1 cells to inhibit release of factors such as IL-1 and TNF. It induces the release of soluble TNF receptors which are endogenous inhibitors of TNF, and downregulates surface TNF receptor. Soluble IL-1 receptors released during inflammation 'decoy' IL-1 itself. IL-4 not only acts to concan act to strain Th1 cells but also upregulates production of the natural inhibitor of IL-1, the IL-1 receptor antagonist

C H A P T E R9 - T H E P R O D U C T I O NO F E F F E C T O R S (IL-1Ra).The role of TGFp is more difficult to teaseout because it has some pro- and other anti-inflammatory effects, although it undoubtedly promotes tissue repair after resolution of the inflammation.

P R O T I F E R A I IAONNDM A T U R A I I OONFB . C E t t RESPONSE AS R EM E D I A I E D B YC Y I O K I N E S The activation of B-cells by Th cells, through the TCR recognition of MHClinked antigenic peptide plus the costimulatory CD40L-CD40 interaction, leads to upregulation of the surface receptor for IL-4. Copious local release of this cytokine from the Th then drives powerful clonal proliferation and expansion of the activated B-cellpopulation.IL-2 and IL-13 also contribute to this process(figure 9.12). Under the influence of IL-4 and IL-13, the expanded clones can differentiate and mature into IgE symthesizing cells. TGFp and IL-5 encourage cells to switch their Ig class to IgA. IgM plasma cells emerge under the tutelage of IL-4 plus -5, and IgG producers result from the

tL-2,4,13

TGFB, tL-S'

lL-4.5

rL-4.5.6.1 3

Figure 9.12. B-cell response to thyrnus-dependent (TD) antigen: clonal expansion and mafuration of activated B-cells under the influence of T-cell-derived soluble factors. Costimulation through the CD40L{D40 interaction is essential for primary and secondary imnune responses to TD antigens and for the formation of germinal centers and memory. c-myc expression, which is maximal 2hours after antigen or anti-p stimulation, parallels sensitivity to growth factors; transfection with c-rayc substitutes for anti-p.

combined influence of IL-4, -5, -6, -73 and IFNy (figure 9.72). Type 2 thymus-independent antigens can activate B-cells directly (cf. p. 178) but nonetheless still need cytokines for efficient proliferation and Ig production. These may come from accessory cells such as NK and NKT-cells which bear lectin-like receptors.

O NI N W H A II S G O I N G T H EG E R M I N ACLE N I E R ? The secondary follicle with its corona or mantle of small lymphocytes surrounding the pale germinal center is a striking and unique cellular structure. First, let us recall the overall events described in Chapter 8. Secondary challenge with antigen or immune complexes induces enlargement of germinal centers, formation of new ones, appearance of memory B-cells and development of Ig-producing cells of higher affinity. Bcells entering the germinal center become centroblasts which divide with a very short cycle time of 6 hours, and then become nondividing centrocytes in the basal light zone, many of which die from apoptosis (figure 9.13). As the surviving centrocytes mature, they differentiate either into immunoblast plasma cell precursors, which secrete Ig in the absence of antigen, or into memory B-cells. \Mhat then is the underlying scenario? Following secondary antigen challenge, primed B-cells may be activated by paracortical Th cells in association with interdigitating dendritic cells or macrophages, and migrate to the germinal center. There they divide in response to powerful stimuli from complexes on follicular dendritic cells (cf. p. 169) and from cytokines released by T-cells in response to antigen-presenting Bcells. During this particularly frenetic bout of cell division, somatic hypermutation of B-cell Ig genes occurs. The cells also undergo Ig class switching. Thereafter, as they transform to centrocytes, they are vulnerable and 'tingible die readily, whence they are taken up as the bodies'by macrophages,unless rescuedby association with antigen on a follicular dendritic cell. This could result from cross-linking of surface Ig receptors and is accompanied by expression of Bcl-x and Bcl-2 which protect against apoptosis. Interactions between BAFF (B-cell-activating /actor of the tumor necrosis factor farnlly; also called BLyS) on the T-helper cell and TACI (fransmembrane 4ctivator and calcium modulator and cyclophilin ligand [CAML] lnteractor), its receptor on the B-cell, may also be important for the maintenance of germinal center B-cells. Signaling through CD40 and TACI, during presentation of antigen to Th cells, would also prolong the life of the centrocyte. In either case,the

201 |

CHAPTER e - r H E P R O D U C T T OONF E F F E C T O R S

PROLIFEMTION FOLLICULAR DENDRITIC CELL

lg SWITCH SOMATIC MUTATION

CENTROBLASTS

CENTROCYTES FOLLICULAR DENDRITIC CELL

APOPTOSIS

T-HELPER

SELECTIVE SURVIVAL

MACROPHAGE WITHPHAGOCYTOSED LYMPHOCYTE

IMMUNE COMPLEX (Ag-Ab-c3d) MATUMTION OF SURVIVING CELLS

Figure 9.13. The events occurring in lymphoid germinal centers. Germinal center B-cells can be enriched through their affinity for the peanut agglutinin lectin They show numerous mutations in the antibody genes Expression of LFA-1 and ICAM-1 on B-cells and foliicular dendritic cells (FDCs) in the germinal center makes them'sticky' Centroblasts at the base of the follicle are strongly CD77 positive. The Th cells bear the unusual CD57 marker The FDCs all express CD21 and CD54; those in the apical lightzone are stronglyCD23 positive, those in the basal light zone express little CD23. Through their surface recep-

tors, FDCs bind immune complexes containing antigen and C3 which, in turn, are very effective B-cell stimulators since coligation of the surface receptors for antigen and C3 (CR2) lowers their threshold for activation The costimulatory molecules CD40 and 87 play pivotal roles Antibodies to CD40 prevent formation of germinaltenters and anti-CD40L can disrupt established germinal centers within l2hours Antl-B7 2, given early in the immune response, Prevents germinal center formation and, when given at the onset of hypermutation,

interactions will only occur if the mutated surface Ig receptor still binds antigen and, as the concentration of antigen gradually falls, only if the receptor is of high affinity. In other words, the system can deliver high affinity antibody by a Darwinian process of high frequency mutation of the Ig genes and selection by antigen of the cells bearing the antibody which binds

most strongly (figure 9.14). This increase of affinity as the antibody level falls late in the response is of obvious benefit, since a small amount of high affinity antibody can do the job of a large amount of low affinity (as in 'goodun' will generally be a match for a boxing, a small mediocre'bigun'). Further differentiation now occurs. The cells either

suppresses thatprocess

NF E F F E C T O R S C H A P T E 9R- T H EP R O D U C T I OO

| 202

migrate to the sites of plasma cell activity (e.9. lymph node medulla) or go to expand the memory B-cell pool depending upon the cytokine and other signals received.CD40 engagementby CD40ligand on a T-cell guides the B-cell into the memory compartment.

T H ES Y N I H E S IO SFA N I I B O D Y The sequential processesby which secreted Ig arises

are illustrated in figure 9.15. In the normal antibodyforming cell there is a rapid turnover of light chains which are present in slight excess. Defective control occurs in many myeloma cells and one may see excessive production of light chains or complete suppression of heavychain synthesis. The variable and constantregions are spliced together in the mRNA before leaving the nucleus. Differential splicing mechanisms also provide a rational explanation for the coexpression of surface IgM and IgD with identical V regions on a single cell, and for the switch from production of membrane-bound IgM receptor to secretory IgM in the antibody-forming cell (cf. figures 4.7and4.2).

SWIICHING I M M U N O G T O B UC TT I NA S S t -cEtts 0 c c u R s l N l N D l v l D U AB

Figure 9.14. Darwinian selection by antigen of B-cells with antibody mutants of high affinity protects against cell death in the germinal centet either through cross-linking of slg by antigen on follicular dendritic cells, or through Th cell recognition ofprocessed antigen and signaling through CD40 ln both cases,capture of antigen, particularly as the concentration falls, willbe crltically affected by the affinity ofthe surface receptor.

The synthesis of antibodies belonging to the various immunoglobulin classes proceeds at different rates. Usually there is an early IgM response which tends to fall off rapidly. IgG antibody synthesis builds up to its maximum over a longer time period. On secondary challenge with antigen, the time-course of the IgM response resembles that seenin the primary. By contrast, the synthesis of IgG antibodies rapidly accelerates to a much higher titer and there is a relatively slow fall-off in serum antibody levels (figure 9.16). The same probably

SRP

GOLGI

U SECREIIOII SSRECEPTOR

ENDOPLASMIC RETICULUM

)sRP RECEPToR

Figure 9.15. Synthesis of immunoglobulin. As nRNA is translated on the ribosome, the N-terminal signal sequence (SS) is bound by a signal recognition particie (SRP) which docks onto a receptor on the outer membrane of the endoplasmic reticulum (ER) and facilitates entry of the nascent Ig chain into the ER lumen The SSassociateswith a specific membrane receptor and is cleaved; the remainder of the chain, as it elongates, complexes with the molecular chaperone BiP

(heavychain-bindingprotein) whichbinds to theheavychainCrrl and V. domains to control protein fotding. The unassembled chains oxidize and dissociate as the full HrLrlg molecule. The assembled HrL, molecules can now leave the ER for terminal glycosylation in the Golgi and final secretion. Surface receptor Ig would be inserted by its hydrophobic sequencesinto the membrane of the endoplasmic reticulum as itwas synthesized

OF EFFECTORS CHAPTER 9-THEPRODUCTION

I sl injeclion of ontigen

2ndinjection of ontigen

lg(

'/, \ l0M

Weeks4 Figure 9.16. Synthesis of IgM and IgG antibody classes in the primary and secondary responsesto antigen.

holds for IgA, and in a senseboth these immunoglobulin classesprovide the main immediate defenseagainst future penetration by foreign antigens. Individual cells can switch over from IgM to IgC production. For example, antigen challenge of irradiated recipients receiving relatively small numbers of lymphoid cells produced splenic foci of cells, each synthesizing antibodies of differentheavy chain classbearinga single idiotype; the common idiotype suggeststhat each focus is derived from a single precursor cell whose progeny can forrn antibodies of different class. Antibody synthesis in most classesshows considerable dependence upon T cooperation in that the responses in T-deprived animals are strikingly deficient; such is true of mouse lgGl,IgG2a, IgA, IgE and part of the IgM antibody responses.T-independent antigens such as the polyclonal activator, lipopolysaccharide (LPS) endotoxin, induce synthesis of IgM with some IgC2b and IgG3. Immunopotentiation by complete Freund's adjuvant, a water-in-oil emulsion containing antigen in the aqueous phase and a suspension of killed tubercle bacilli in the oil phase (see p. 308), seems to occur, at least in part through the activation of Th cells which stimulate antibody production in Tdependent classes.The prediction from this, that the responseto T-independent antigens (e.g.Pneumoclccus polysaccharide, p. 178) should not be potentiated by Freund's adjuvant, is borne out in practice;furthermore, as mentioned previously, these antigens evoke primarily IgM antibodies and poor immunological memory, as do T-dependent antigens injected into T-cell-deficient, neonatally thymectomized hosts. Thus, in rodents at least, the switch from IgM to IgG and other classes appears to be largely under T-cell control critically mediated by CD40 and by cytokines (seep. 200).Let us take another look at the stimulation of small, surfaceIgM-positive, B-cellsby LPS.As we noted,

203 |

onits own, thenonspecific mitogenevokes the synthesis of IgM, IgG3 and some IgG2b. Following addition of IL4 to the system, there is class switching from IgM to IgE and IgGl production, whereas IFNy stimulates class switchingfrom IgMto IgG2a and TGFBinduces switching from IgM to IgA or IgG2b. These cytokines induce the formation of germ-line sterile transcripts which start at the I (initiation) exon 5' of the switch region for the antibody classto which switching will occur and terminate at the polyadenylation site 3'of the relevant Cngene (figure 9.17). The transcripts are not translated but instead remain associated with the template DNA, forming RNA-DNAhybrids within the S regions of the DNAwhich might act as targets for enzymes involved in the recombination process. Under the influence of the recombinase, a given VDI gene segment is transferred from p6 to the new constant region gene (figure 9.17),so yielding antibodies of the same specificity but of different class. B-cells0resubiecllo high mulolionroles Closs-switched otterlhe iniliolresponse The reader will no doubt recollect that this idea was raised in Chapter 4 when discussing the generation of diversity, and that the germinal center has been identified as the site of intense mutagenesis. The normal Vregion mutation rate is of the order of 10 s/base pairlcell division, but this rises to 10-3lbase pair / gen' eration in B-cells as a result of antigenic stimulation. This process is illustrated in figure 9.18which charts the accumulation of somatic mutations in the immunodominant Vrr/V* antibody structure during the immune response to phenyloxazolone. With time and successive boosting, the mutation rate is seen to rise dramatically and, in the context of the present discussion, it is clear that the strategically targeted hypermutations occurring within or adjacent to the complementarity determining hypervariable loops (figure 9.19) can give rise to cells which secrete antibodies having a different combining affinity to that of the original parent cell. Randomly, some mutated daughter cells will have higher affinity for antigen, some the same or lower and others perhaps none at all (cf. figure 9.14).Similarly, mutations 'silent' or, if they in the framework regions may be disrupt the folding of the protein, give rise to nonfunctional molecules. Pertinently, the proportion of germinal center B-cellswith'silent'mutations is high early in the immune response but falls dramatically with time, suggesting that early diversification is followed by preferential expansion of clones expressing mutations which improve their chances of reacting with and being stimulated by antigen.

I zoa

CHAPTER 9-THEPRODUCTION OF EFFECTORS

LoopIormolion r---------l

(sTERtLE OtOr*ttC.'Ot[

)

CLASS SWITCH I

I'T'FII t

FT'FE

EM-_--]

Figure 9.17. Class switching to produce antibodies of identical specificity but different immunoglobulin isotype (in this example from IgM to IgGl) is achieved by a recombination process which utilizes the specialized switch sequences (f ) and leads to a loss of the intervening DNAloop (p, 6 and y3) Each switch sequence is 1-10 kilobases in length and comprises guanosine-rich repeats of 20-100 base pairs Becausethe switch sequence associated with each C' gene has a unique nucleotide sequence, recombination cannot occur hom*ologously and therefore probably depends upon nonhom*ologous end joining. DNA repair proteins including Ku70, Ku80 and the catalytic subunit of the DNAdependent protein kinase (DNA-PK.') are involved in this process

F A C I O RA SF F E C T I NAG N T I B O DAYF F I N I TIY N T H EI M M U N E RESPONSE @

*25

Theefleclofonligen dose

820

Other things being equal, the binding strength of an antigen for the surface receptor of a B-cell will be determined by the affinity constant of the reaction:

U a co

= (J rn =rv a OC z

Figure 9.18. Increasing somatic mutations in the immunodominant germ-line antibody observed in hybridomas isolated following repeated immunization with phenyloxazolone (Data from Berek C & Apel M (1989)In Melchers F. ef a1.(eds) Progressin lmmunology7, 99. Springer-Verla g, Berlin )

Ag+(surface)Ab =AgAb and the reactants will behave according to the laws of thermodynamics (cf. p. 91). It may be supposed that, when a sufficient number of antigen molecules are bound to the receptors on the cell surfaceand processedfor presentation to T-cells,the lymphocyte will be stimulated to develop into an antibody-producing clone. When only small amounts of antigen are present/ only those lymphocytes with high affinity receptors will be able to bind sufficient antigen for stimulation to occur and their daughter cells will, of course, also produce high affinity antibody. Consideration of the antigen-antibody equilibrium equation will show that, as the concentration of antigen is increased, even antibodies with relativelv low affinitv will bind

205 |

CHAPTER 9-THEPRODUCTION OF EFFECTORS

Figure 9.19. An'antigen's eye view'of sequencediversity in human antibodies. The sequence diversity has been plotted on a scale of blue (more conserved) to red (more diverse). The V" domain is on the right and the V* domain on the left in both pictures (a) GermJine diversity prior to somatic hypermutation is focused at the center of the antigenbinding site (b) Somatic hypermutation spreads diversity to regions at the periphery ofthe binding site that are highly conserved in the germ-

line V gene repertoire Somatic hypermutation is therefore complementary to germJine diversity. The Vtr CDR3, which lies at the center of the antigen-binding site, was not included inthis analysis and therefore is shown in gray as a loop structure. The end of the V* CDR3 (also excluded) lies at the center of the binding site and is not visible in this representation (Reproduced with kind permission from Tomlinson IM. et al (1996)lournal of Molecular Biology 256, 81'3)

more antigen; therefore, at high doses of antigen, the Iymphocytes with lower affinity receptors will also be stimulated and, as may be seen from figure 9.20, these are more abundant than those with receptors of high affinity. Furthermore, there is a strong possibility that cells with the highest affinity will bind so much antigen as to become tolerized (cf. p. 2a\. Thus, in summary, low amounts of antigen produce high affinity antibodies, whereas high antigen concentrations give rise to an antiserum with low to moderate affinitv. Molurolionof offinity In addition to being brisker and fatter, secondary responses tend to be of higher affinity. There are probably two main reasons for this maturation of affinity after primary stimulation. First, once the primary response gets under way and the antigen concentration declines to low levels, only successively higher affinity cells will bind sufficient antigen to maintain proliferation. Second, at this stage the cells are mutating madly in the germinal centers, and any mutants with an adventitiously higher affinitywillbind well to antigen on follicular dendritic cells and be positively selectedfor by its persistent clonal expansion. Modification of antibody specificity by somatic point mutations allows gradual

lAgl', V

AFFINIilANTIBODIES MODERATE

Froc'tion of memory cells for Ag

Afriniiy(K)

Figure 9.20. Relationship of antigen concentration to affinity of antibodies produced. Low concentrations of antigen ([Ag].o) bind to and permit stimulation of a range of high affinity memory cells and the resulting antibodies are of high affinity. High doses of antigen ([Ag]*rr) are able to blnd sufficiently to the low affinity cells and thereby allow their stimulation, whilst the highest affinity cells maybind an excessof antigen and be tolerized (dashed line); the resulting antiserum will have a population of low to moderate affinity antibodies

9-THEPRODUCTION CHAPTER OF EFFECTORS diversification on which positive selection for affinity can act during clonalexpansion. It is worth noting that responses to thymusindependent antigens, which have poorly developed memory with very rare mutations, do not show this phenomenon of affinity maturation. Overall, the ability of Th to facilitate responses to nonpolymeric, nonpolyclonally activating antigens, to induce expansive clonal proliferation, to effect class switching and, lastly, to fine-tune responses to higher affinity has provided us with bigger, better and more flexible immune responses.

M E M O RC YE T T S As the immune response subsides, the majority of recently expanded effector cells are culled by large-scale induction of apoptosis in this population. However, a subpopulation of cells escape the culling process and these form the memory compartment thatlive to mount a more rapid and efficient secondary immune response upon re-exposure to the same antigen. It is possible that the memory cell population represents a subpopulation of cells thatbypass the effector cell stage entirely, but this concept remains controversial. The process of memory cell generation is central to the concept of vaccination and memory cells have been the subjects of much investigation as a consequence. Antibodies encoded by unmutated germ-line genes represent a form of evolutionary memory, in the sense that they tend to include specificities for commonly encountered pathogens and are found in the so-called 'natural antibody'fraction of serum. Memory acquired during the adaptive immune response requires contact with antigen and expansion of antigen-specific memory cells, as seen for example in the 20-fold increase in cytotoxic T-cell precursors after immunization of females with the male H-Y antigen. Memory of early infections such as measles is longlived and the question arises as to whether the memory cells are long-lived or are subject to repeated antigen stimulation from persisting antigen or subclinical reinfection. Fanum in 1847described a measlesepidemic on the Faroe Islands in the previous year in which almost the entire population suffered from infection except for a few old people who had been infected 65 years earlier. While this evidence favors the long half-life hypothesis, memory function of B-cells transferred to an irradiated syngeneic recipient is lost within a month unless antigen is given or the donor is transgenic for the bcl-2 gene (remember that signals in the germinal center which prevent apoptosis of centrocytic B-cells also upregulate bcl-2 expression). It is envisaged that B-cell

memory is a dynamic state in which survival of the memory cells is maintained by recurrent signals from follicular dendritic cells in the germinal centers, the only long-term repository of antigen. Evidence from mouse models strongly suggests that memory T-cells can, at least in principle, persist in the absence of antigen. T-cells isolated from mice several months after they were immunized with lymphocytic choriomeningitis virus (LCMV) were transferred into two groups of genetically modified mice which lacked endogenous T-cells, one of the groups additionally lacking MHC class I expression. T-cells were parked in these mice for 10 months and then analysed in aitro. Functional virus-specific CD8+ CTLs were still present in both groups of mice, and in similar numbers, even though those from the class I- mice could not have had antigen presented to their TCR. Indeed, these memory T-cells undergo antigen- and MHC-independent proliferation in uioo, their numbers controlled, at least in part, by a balance between proliferation-inducing signals from IL-15 and cell death-inducing signals from IL-2 released in the local environment, both cytokines binding to the IL-2R p chain (cf. figure 9.2). Other recent findings indicate that helper T-cell memory also does not require the continued presence of antigen or MHC and, at least in some cases,Th memory is maintained in the absenceof cell division. However, we should not lose sight of the fact that, while these experiments in transgenic and knockout animals clearly demonstrate that immunological memory cnn be maintained in the absence of antigen, usually antigen persists as complexes on follicular dendritic cells. Therefore, there is the potential for antigenpresenting cells within the germinal center to capture and process this complexed antigen and then present it to memory T-cells. Some evidence/ again recent, suggests that it is a type of dendritic cell, and not the germinal center B-cells, that may subserve this function. To add complexity, there is also accumulating evidence that the mechanisms used to maintain memory T-cells in the mouse, a relatively short-lived animal, may differ significantly from those employed by the human immune system. Specific antigen may play a much more important role in maintaining T-lymphocyte memory in man, not least because ongoing entry of new memory cells specific for diverse antigens to the memory compartment will generate competition between memory cells.Becausethe naive and memory cell pools are maintained at a relatively constant size, il is likely that memory cells which receive periodic re-stimulation with antigen are likely to persist for longer than those that fail to re-encounter antigen. Competition may be absent or diminished in mouse models where animals

OF EFFECTORS CHAPTER 9-THEPRODUCTION are typically maintained in artificially clean environments; such cosseting is likely to reduce the rate of entry of new T-cell specificities to the memory compartment and therefore reduce competition between memory cell populations. In support of this view, while there is evidence that T-cell memory in humans can persist for decadesafter exposureto particular antigens,immunity does indeed decline over time and estimates of the half-life of T-cell responseshave put this between 8 and 15 years. In addition, becausethe lifespan of the laboratory mouse is far shorter than the average human, the problems associatedwith retention of memory cells in the human are likely to be greater than those faced by laboratory mice. Ongoing attrition of memory T-cells, in the absence of antigenic re-stimulation, may contribute to the increased rate and severity of infectious diseases in the elderly and may also explain why latent viruses, such as varicella zoster (human herpesvirus 3), may reactivate many years after initial infection. Thememoryp0pulolion is notsimplyon exponsion of noivecells corlesponding In general, memory cells are more readily stimulated by a given dose of antigen becausethey have a higher affinity. In the caseof B-cells,we are satisfiedby the evidence that links mutation and antigen-driven selection, occurring within the germinal centers of secondary lymph node follicles, to the creation of high affinity memory cells. The receptors for antigen on memory T-cells also have higher affinity but, since they do not undergo significant somatic mutation during the priming response, it would seem that cells with pre-existing receptors of relatively higher affinity in the population of naive cells proliferate selectively through preferential binding to the antigen. Intuitively one would not expect to improve on affinity to the same extent that somatic hypermutation can achieve for the B-cells, but nonetheless memory T-cells augment their binding avidity for the antigenpresenting cell through increased expression of accessory adhesion molecules, CD2, LFA-1, LFA-3 and ICAM-1. Since several of these molecules also function to enhance signal transduction, the memory T-cell is more readily triggered than its naive counterpart. Indeed, memory cells enter cell division and secrete cytokines more rapidly than naive cells, and there is some evidence that they may secretea broader range of cytokines than do naive cells. A phenotypic change in the isoform of the leukocyte common antigen CD45R, derived by differential splicing, allows some distinction to be made between naive

207 |

and memory cells.Expressionof CD4SRAhasbeen used as a marker of naive T-cells and of CD4SRO as a marker of memory cells capable of responding to recall antigens. However, most of the features associated with the CD45RO subset are in fact manifestations of activated cells and CD45RO cells can revert to the CD45RAPhenotype. Memory cells, perhaps in the absence of antigenic stimulation, may therefore lose their activated status and join a resting pool. Another marker used for differentiating naive from memory cells takes one step back on the CD ladder and utilizes differences in the relative expression of the adhesion molecule CD44; naive T-cells seem to express low levels of CD44whilst memory T-cells express high levels. Lanzavecchia and colleagues have proposed that the CCRT chemokine receptor allows a distinction to be 'central memory' T-cells, which made between CCRT+ 'effector differentiate from naive T-cells, and CCRTmemory' T-cells, which subsequently arise from the central memory T-cells (figure 9.21).Both populations are long-lived. The central memory cells provide a clonally expanded pool of antigen-primed cells which can travel to secondary lymphoid organs under the influence of the CCL21 (SLC) chemokine (cf. table 9.3) and, following re-encounter with antigen, can stimulate dendritic cells, help B-cells and generate effector cells. In contrast, effector memory T-cells possessCCR1, CCR3 and CCR5 receptors for proinflammatory chemokines and constitute tissue-homing cells which mediate inflammatory reactions or cytotoxicity. Recently, IL-7 has emerged as a key regulator of peripheral T-cell survival and homeostatic turnover. Unlike most other cytokines that use receptors containing the common y-(CD732),IL-7 is produced constitutively at low levels, is detectable in human serum, and may contribute to the antigen-independent maintenance of CD4 and CD8 memory T-cells by stimulating homeostatic division of these cells. \Atrhereasstudies using MHC-deficient mice have shown that peptideMHC interactions are not essential for the persistence of memory T-cells, CD4 T-cells decline rapidly in the absence of IL-7. The expression of IL-7R is highest on resting cells, ensuring that these cells comPete more effectively for available IL-7 than activated effector Tcells.Indeed, stimulationvia the TCRinduces downregulation of the receptor for IL-7 as effector T-cells come under the influence of cytokines produced during immune responses (su ch as IL-2, IL- 4, IL-7, IL-1'5 and lL21).As the responsesubsides,T-cellsbecomedependent on IL-7for their continued survivaloncemore. Thus, the current view is that lL-7 contributes to the antigenindependent maintenance of T-cells by permitting homeostatic division of these cells in the absence of

CHAPTER 9 _ T H EP R O D U C T I O N OF EFTECTORS

Anligen

Antigen

Anligen

l Effector T-cells

Thymic T-cells

CD45RA+ * CCRT

CD45RAccRT

Figure 9.21. Central and effector memory T-cells. Naive T-cel1sbear the CD45RA splice variant of the CD45 molecule and are attracted from the thymus into secondary lymphoid tissue under the influence of CCRT-binding chemokines such as CCL19 (MIP-38) and CCL21 (6Ckine/SLC) Upon encounter with antigen, some of these cells become effectors of the primary immune response, whilst others differentiate into central memory T-cells which retain the CCRT chemokine receptor but lose expression of CD45RA Subsequent reencounter with antigen will push these cells into the effector memory

NAIVE IL-7R IL-2R IL-4R rL-t5R

MEMORY lL-7R1 lL-2Rr lL-4R{ r rL-l5R

EFFECTOR l L - 7 RI t L - 2 R1 lL-4R1

t L - t 5 rR

tL-t5 . Removol tL-7 of onligen

Figwe9.22. Cytokine receptor expression and cytokine availability control T-cell proliferation and survival. Naive CD4 and CD8 T-ce1ls express high levels of IL-7R and low levels of receptors for other cytokines, such as IL-2, IL-4 and IL-15, which can influence T-cell proliferation and survival. Antigenic stimulation induces downregulation of IL-7R and upregulation of receptors for IL-2, IL-4 and IL-15 as these cytokines sustain T-cell clonal expansion and survival during the effector phase of the immune response During resolution of the immune response, massive apoptosis occurs within the effector cell compartment leaving only the 'fittest' cells to become memory cells The memory cell compartment appears to rely upon IL-7 for long-term survival with IL-15 also thought to be required, particularly for the maintenance of memory CD8 T-celis

antigenic stimulation (figur e 9 .22).IL- 15 also appears to be more important for the maintenance of CD8 memory T-cellsas mice deficient in either IL-15 or IL-15Rcrchain displayreduced CD8 T-cell memorywhichcanbe rescued

CD45RA CCRT c c R l ,3 . 5 r

compartment with replacement of CCRT by other chemokine recePtors such as CCR1, CCR3 and CCR5 This changes the homing characteristics of these cells which can now relocate as cytokine-secreting or cytoioxic T-cells to inflammatory sites under the influence of a number of chemokines including CCL3 (MIP-1cr), CCL4 (MIP-18) and CCL5 (RANTES) (see table 9 3) Note that whilst the activation and subsequent differentiation of these cells is dependent on antigen, both central memory and effector memory T-cells are thought to be longhved in the absenceof antigen

by transfer of thesecellsto normal mice. Thus,IL-7 andIL15 appear to act in concert to maintain the memory T-cell pool, the latter being particularly important for the maintenanceof CD8 memoryT-cells (figure 9.22). The persistence of memory cells may also be influencedby physical factors,such as the length of chromosomal telomeres, that impose limits on the number of divisions that most mammalian cells can undergo; the so-calledHayflick limit. The progressive erosion of chromosomal telomeres during each cell division can result in cells entering a state of senescencefrom which they cannot exit. In this situation, cells are unable to divide further and are likely to be functionally compromised and therefore of little further use to the immune system. For many cell types, the Hayflick limit is typically reached within 40-50 cell divisions, but lymphocytes may be permitted somewhat more cell divisions than this due to the upregulation of the telomerelengthening enzyr;:.e, telomerase, within activated lymphocytes. Ithasbeen reported that CD8 T-cellsfail to upregulate telomerase after four restimulations with antigen while CD4 T-cells may retain this ability for Ionger. Virgin B-cells lose their surface IgM and IgD and switch receptor isotype onbecomingmemory cells,and the differential expression of these surface markers has greatly facilitated the separation of B- and T-cells into naive and memory populations for further study. The costimulatory molecules B7.1 (CD80) and 87.2 (CD86) are rapidly upregulated on memory B-cells, and the

OF EFFECTORS CHAPTER 9 _ T H EP R O D U C T I O N potent ability of these cells to present antigen to T-cells could well account for the brisk and robust nature of secondary responses.A schemesimilar to that outlined in figure 9.27 for T-cells may also exist for the B-

Asuccessi0n 0lgenes oreupreguloled byT-oell 0clivotion . Within L5-30 minutes, genes for transcription factors concerned in the progression G0 to G1 and in the conkol of IL-2

areexpressed. . Up to 14 hours, cytokines and their receptors are expressed. o Later, a variety of genes related to cell division and adhesionareupregulated.

2oeI

lymphocyte compartment, with an initial population of memory cells possessingthe 8220 marker developing into 8220- memory B-cells which then go on to generate antibody-secretingeffector cells.

. Interaction of antigen with macrophages or dendritic cells, via their Toll-like receptors (TLRs) and other pattem recognition receptors, Ieads to production of lL-12 andIL-27 which skews T-cell responses to the Th1 fype, or IL-4 which will skew the responses to the Th2 pole. r Other subsets may exist, including natural tegs and inducible TGFp-secreting Th3 (Tr1) regulatory cells. . lc1 (IFNT) and Tc2 (IL-4) populations can also be distinguished.

Cyfokines oclosinlercellulor messengeF . Cytokines act transiently and usually at short range, although circulating IL-1 and IL-6 can mediate release of acute phase proteins from the liver. . Cytokines are mostly small proteins that act through surface receptors belonging to six structural families. . Cytokine-induced dimerization of individual subr.rnitsof the main (hematopoietin) receptor family activates protein

Acfivoled T-cellsprollferote in responselo cyloldnes . IL-2 acts as an autocrine growth factor for Th1 and paracrine for TM cells which have upregulated their IL-2

tyrosine kinases, including JAKs, and leads to phosphorylation and activation of STAT transcription factors. . Cytokine signaling can be downregulated by members of the SOCS and PIAS family of inhibitors that act to suppress JAK activity or STAT-dependent transcription, respectively. . Cytokines are pleiotropic, i.e. have multiple effects in the general areas of: (i) control of lymphocyte growth, (ii) activation of innate immune mechanisms (including inflammation), and (iii) control of bone marrow hematopoiesis (cf.

T-cellefleclorcin cell-medioledimmunity . Cytokines mediate chronic inflammatory responses and induce the expression of MHC class II on endothelial cells, a variety of epithelial cells and many tumor cell lines, so facilitating interactions between T-cells and nonlymphoid cells' . Differential expression of chemokine receptors permits selective recruitment of neutrophils, macrophages, dendritic cells and T- and B-cells. o TNF synergizes with IFNyin killing cells. o Cytotoxic T-cells are generated against cells (e.g. virally infected) which have intracellularly derived peptide associated with surface MHC class L They kill usinglytic granules containing perforin, granzl'rnes and TNF. . The cytotoxic granule-dependent pathway to apoptosis is orchestratedby granzyme B, a serine protease that can

figure 9.4). . Cytokines may act sequentially, through one cytokine inducing production of another or by transmodulation of the receptor for another cytokine; they can also act synergistically or antagonistically. . The roles of cytokines in ztiao can be assessedby gene 'knockout', transfection or inhibition by specific antibodies. Differenll-cell subselscon mokedifferenlcylokines o The cytokine milieu that is established within the initial stages of infection has a significant influence on the pattern of cytokines secretedby Th cell populations. o As immunization proceeds, Th tend to develop into two subsets: Th1 cells concerned in inflammatory processes, macrophage activation and delayed sensitivity make IL-2 and -3,IFNy, TNF,lymphotoxin and GM-CSF; Th2 cells help B-cells to synthesize antibody and secreteIL-3, -4, -5, -6 and -13,TNF and GM-CSF.IL-10 is secretedby Th2 cells in mice but by both Th1 and Th2 subsets in humans.

recePtors. . Cytokines act on cells which express the appropriate cytokine receptor.

process and activate the mitochondrial-permeabilizing protein, Bid, as well as members of the caspasefamily of cell death proteases. Granzyme A also plays an important role in granule-dependent killing. . T-cell-mediated inflammation is strongly downregulated by IL-4 and IL-10.

bYcylokines ismedioled ofB-cellresponses Prolilerotion . Early proliferation is mediatedby IL-4 which also aids IgE synthesis. . IgAproducersaredrivenby TGFBandIL-5. . IL-4 plus IL-5 promote IgM and lL-4, -5,-6 and -13plus IFNystimulateIgC synthesis. (Continuedp 210)

I zro

C H A P T E 9R_ T H EP R O D U C I I O N OFEFFECTORS

Evenfsin lhe germinolcenler . There is clonal expansion, isotype switch and mutation in the dark zone centroblasts. o The B-cell centroblasts die through apoptosis unless

ated their production, the vast rnajority of effector lymphocytes are eliminated via apoptosis. A fraction of antigenresponsive cells are retained, possibly those with the highest affinity for antigen, and these form the memory

rescued by certain signals which upregulate bcl-2. These include cross-linking of surface Ig by complexes on follicular dendritic cells and engagement of the CD40 receptor which drives the cell to the memory compartment o The selection of mutants by antigen guides the development of high affinity B-cells.

compartment. . Murine memory T-cells canbe maintained in the absence of antigen but human T-cell memory may require periodic restimulation with antigen. . Immune complexes on the surface of follicular dendritic cells in the germinal centers provide a long-term source of

Thesynlhesisof onlibody o RNAfor variable and constant regions is spliced together before leaving the nucleus o Differential splicing allows coexpression of IgM and IgD with identical V regions on a single cell and the switch from

antigen. o Memory cells have higher affinity than naive cells, in the case of B-cells through somatic mutation, and in the case of T-cells through selective proliferation of cells with higher affinity receptors and through upregulated expression of associated molecules such as CD2 and LFA-1, which increase the avidity (functional affinity) for the antigen-

membrane-bound to secretedIgM. lg clossswilchingoccursin individuolB-cells . IgM produced early in the responseswitches to IgG, particularly with thymus-dependent antigens. The switch is Iargely under T-cell control . IgG, but not IgM, responses improve on secondary challenge. Anlibodyotfinilyduringlhe immuneresponse . Low doses of antigen tend to selecthigh affinity B-cells and hence antibodies since onlv these can be rescued in the germinal center o For the same reasons, affinity matures as antigen concentration falls during an immune response.

presenting cell. . Activated memory and naive T-cells are distinguished by the expression of CD45 isoforms, the former having the CD45RO phenotype, the latter CD45RA. It seems likely that a proportion of the CD4SRO population reverts to a CD45RA pool of resting memory cells CD45RA- memory cells can be divided into CCRT+central memory and CCRT effector memory cells. . High levels of CD44 expression are also characteristicof memory T-cells, low level expression being associated with naive T-cells. . IL-7 appears to be critical for the long-term survival of CD4 T-cell populations and is preferentially bound by resting T-cells. Memory CD8 T-cells require IL-15 for their long-term survival.

Memorycells . Upon disappearance of the source of antigen that initi-

F U R T H ERRE A D I N G Ber.erlv PC L. (2004) Kinetics and clonality of immunological memory in hum ans Settin nrs i n I m rnu nology 16,315-321 Bradley L M, Haynes L & Swain S L (2005) IL-7: maintaining T-ce11memory and achieving homeostasis Trendsin Innrunology 26,772-176 Camacho S A, Kosco-Vilbois MH &BerekC (1998)The dynamic structure of the germinal center.lnmruttologyTodny19,511,-51,4 Fujimoto M & Naka T. (2003) Regulation of cytokine signaling by SOCS familv molecules. Trettdsinlnrmunology 24,659-666 Kapsenberg M L (2003) Dendritic cell control of pathogen-drir.en T-cell polarization Nnf lre Rezrieuslmmunology 3,984-993 Kinosh*ta K & Honjo T. (2000)Unique and unprecedented recombination mechanisms in class switching, Current Oytinion in Intn u noIogy 12, 195-198 Lanzavecchia A. & Sallusto F (2002)Progressive differentiation and selection of the fittest ln the immune response Nnture Reaiews Inunmology 2,982-987 Mills K H C (2004)Regulatory T-cells: friend or foe in immunity to infection? Nafu re Reztiezus lntnrunology4,841-855

Moser B & Loetscher P (2001) Lymphocyte traffic controi by chemokines Nature lm munology 2, 123-128 Sallusto F, Mackay CR. & Lanzavecchia A (2000) The role of chemokine receptors in primary, effector, and memory immune responses Annunl Reaiewof lmmunology 18,593-620 Schluns K.S & Lefrangois L (2003) Cytokine control of memory Tcell development and survrval Nnture Re'oiews lmmunology 3, 269-279 Shuai K & Liu B (2003) Regulation of JAK-STAT signaling in the immune system . Ntlture ReriezosImmunology 3,900-977 Sprent J & Surh C D (2001)Ceneration and maintenance of memory T-cells. Current Opinion in lnmtunology 13,248-254 Tough D F, Sun S., Zhang X. & Sprent J (1999)Stimulation of naive and memory T-cel1sby cytokines. lmmunological Reaiews 170, 3947 Trinchieri G (2003) Interleukin-12 and the regulation of innate resistanceand adaptive immunity. Nature Reaiewslnrmunology 3, 133-146. Zlotnik A & Yoshie O (2000) Chemokines: a new classification system and their role in lmmunity. hununity 12,127-127

mechonisms Control

INIRODUCIION The acquired immune responseevolved so thatitwould come into play upon contact with an infectious agent. The appropriate antigen-specific cells expand, often to form a sizableproportion of the lymphocytes in the local lymphoid tissues, the effectors eliminate the antigen and then the responsequietens down and leaves room for reaction to other infections. Feedback mechanisms must operate to limit the response;otherwise, after antigenic stimulation, we would become overwhelmed by the responding clones of T-cells and antibody-forming cells and their products-obviously an unwelcome state of affairs, as may be clearly seen in multiple myeloma, where control over lymphocyte proliferation is lost. It makes sensefor antigen to be a major regulatory factor and for lymphocyte responsesto be driven by the presence of antigen, falling off in intensity as the antigen concentration drops (figure 10.1).There is abundant evidence to support this view Antigens can stimulate the proliferation of specific lymphocytes in aitro. Clearanceof antigen in aiao by injection of excessantibody during the course of an immune responseleads to a dramatic drop in antibody synthesis and the number of antibody-secretingcells.

A N T I G E NCSA NI N I E R F E RWEI I H E A C HO T H E R The presence of one antigen in a mixture of antigens can drastically diminish the immune responseto the others. This is true even for epitopes within a given molecule; for example, the response to epitopes on the Fab fragment of IgG is far greater when the Fab rather than whole IgC is used for immunization due to the inhibitory nature of the Fc region. Factors which determine immunodominance include the precursor

frequency of the B-cells bearing antigen receptors for different epitopes on the antigen, the relative affinity of these antigen receptors for their respective epitopes, the degree to which the surface membrane antibody protects the epitope from proteolysis following internalization of the antibody-antigen complex, and the level of competition of processed antigenic peptides for the najor ftistocompatibility complex (MHC) groove. There is a clear hierarchy of epitopes with respect to this competitive binding based on differential accessibility to proteasesasthe molecule unfolds, and thepresenceor absenceof particular amino acid sequenceswhich facilitate breakdown to yield peptides in high abundance and with relatively high affinity for the MHC (figure 10.2). Thus, Sercarz envisages dominant epitopes, which bag the lion's share of the available MHC grooves, subdominant epitopes, which are less successful,and cryptic epitopes, which generatemiserably low concentrationsof peptide-MHC that are ignoredby potentially reactivenaive T-cells. Clearly, the possibility that certain antigens in a mixture, or particular epitopes in a given antigen, may block a desired protective immune response has obvious implications for vaccine design. Contrariwise, the identification of inhibitory peptides with a predatory affinity for the MHC groove(s) should provide therapeutic agents to quash unwanted hypersensitivity reactions.

C O M P T E M EA NN I DA N T I B O DAYL S O P t A YA R O T E Innate immune mechanisms are usually first on the scene and activation of the alternative pathway of complement will lead to C3d deposition on the microbe. When C3d-coatedantigens are recognizedby the B-cell,

212

CHAPTER I O - C O N T R O LM E C H A N I S M S

cross-linking of the BCR and the CD21 complement receptor, with its associated signal-transducing molecule CD19, enhancesB-cell activation (figure 10.3a).By contrast, cross-linking of the BCR with FcyRIIBI (cf. p.46) delivers a negative signalby suppressing tyrosine phosphorylation of CD19 (figure 10.3b).Thus, removal of circulating antibody by plasmapheresis during an ongoing response leads to an increase in synthesis, whereas injection of preformed IgG antibody markedly

EARLY

Antigen cofobolism & removol 0ytmmune response

hastens the fall in the number of antibody-forming cells (figure 10.4)consistent with feedback control on overall synthesis. In complete contrast/ injection of IgM antibodies enhances the response (figure 10.4), presumably by cross-linking antigen bound to the sIgM receptors without activating the Fcy inhibitory receptol, and

+I LATE

ACCESSIBILITY

Denoturotion Proleolysis ANTIGEN CONCN.

Proleolysrs PREFERENTIAL PROTEOLYSIS

++ gffi

>-ifrF+I

PREFERENTIAL BINDING TO MHC TIMEAFTER INFECTION

oovtnnnrrptropr

SuBDOI/TNANT EprTopE

Figure 10.1. Antigen drives the irnmune response. As antigen concentration falls due to catabolism and elimination by antibody, the intensity of the immune response declines, but is maintained for some time at a lower level by antigen trapped on germinal center follicular dendritic cells

Figure 10.2. Mechanisms of epitope dominance at the MHC level. The other factor which can influence dominance is the availability o{ reactive T-cells; if these havebeen eliminated, e.g through tolerization by cross-reacting self-antigens, a peptide which may have dominated the MHC groove would be unable to provoke an immune response.

Figure 10.3. Cross-linking of surface IgM antigen receptor to the CD21 complement receptor stimulates, and to the Fc1 receptor FCyRIIBI inhibits, B-cells. (a) Following activation of complement, C3d becomes covalently bound to the microbial surface. The CD21 complement receptor binds C3d and signals through its associated CD19 molecule The CD21 and CD81 (TAPA-1) Leu13 molecules form the B-cell coreceptor (cf p 181) and cross-linking of this complex to the

surface IgM of the BCR leads to tyrosine phosphorylation ofCD19 and subsequent binding of phosphatidylinositol 3-kinase (PI 3-K), leading to B-cell activation (b) The FcyRIIBI molecule possessesa cytoplasmic immunoreceptor tyrosine-based inhibitory motif (ITIM) and, upon cross-linking to membrane Ig, becomes phosphorylated and binds the inositol polyphosphate 5lphosphatase SHIP This suppresses phosphorylation of CD19 and thus inhibits B-cell activation.

CHAPTER I O - C O N T R O LM E C H A N I S M S

U a

{, o o o o 9?

rn4

lgM

E o o q

CONTROL

lgG

l0r!

Figure 10.4. Time-course of enhancement of antibody response to sheep redblood cells (SRBC) due to injection of preformed IgM, and of suppression by preformed IgG antibodies. Mice received monoclonal IgM anti-SRBC,IgG anti-SRBC or medium alone intravenously 2 hours prior to immunization with 105SRBC (Data provided by J. Reiter, P Hutchings, P Lydyard and A Cooke )

perhaps also via the generation of C3d by the classical pathway of complement activation. Since antibodies of this isotype are either already present amongst the broadly reactive natural antibodies, or if not will certainly appear at an early stage after antigen challenge, they would be useful in boosting the initial response.

213 |

ligand (TRAIL/Apo2L) retains the ability to signal through the receptor TRAIL-R1. Such soluble ligands can potentially mediate either paracrine or autocrine cell death ln aiao, and show promise as tumor therapeutics. The death receptors are members of the tumor necrosis factor receptor (TNF-R) family and include TNFRI, CD95 (Fas), TRAMP (TNF receptor apoptosismediating protein), the aforementioned TRAIL-RI (death /eceptor DR4), TRAIL-R2 (DR5), DR3 and DR6. Apoptosis induction through these receptors initially involves cleavage of the inactive cysteine Protease Procaspase8 to yield active caspase8. Ultimately, this activation pathway converges with the apoptosis pathway induced by cellular stress,both leading to the activation of downstream effector caspases(figure 10.5).There are also a numb er of decoy recePtors, including the membrane bound TRAIL-R3 (DcR1) and TRAIL-Ra (DcR2) and the soluble TRAIL-RS (osteoprotegerin), which bind the potentially apoptosis-inducing ligands but do not signal. When initially activated by peptide-MHC, T-cells are resistant to apoptosis but become progressively sensitive. A number of molecules are known to be protective against apoptosis; for example, bcl-2 and bcl-Xt, which appear to act as watchdogs preventing the release of pro-apoptotic proteins from the mitochondria' Of patticular relevance to death receptor-mediated AICD, however, is the molecule FLIP (FLICE lnhibitory protein, FLICE being an older name for caspase8). FLIP bears structural similarity to casPase 8, and therefore by competitive inhibition prevents recruitment of this caspase into the death-lnducing signaling complex (DISC) (figure 10.5).Thus, FLIP levels can determine the fate of the cell when the death receptor is engaged by its ligand but does not affect apoptosis induced by the stress-activatedmitochondrial pathway (figure 10.6)'

A C I I V A I I O N - I N D U CCEEDT TD E A T H Although clearance of antigen from the body by the immune system will clearly lead to a downregulation of lymphocyte proliferation due to the absenceof a signal through the antigen receptor, even in the presence of antigen the signals provided do not lead to the continuous proliferation of cells, but rather set off a train of events that, unless the cells are protected in some way, leads to activation-lnduced cell death (AICD) by apoptosis. Subsequent to activation, T-cells upregulate death receptorsand their ligands. If the ligands remain associated with the cell surface they can activate apoptosis in adjacent cells. However, they are often released from the cell surface by proteases, producing soluble forms which in some cases retain activity, for example the soluble version of the TNF-related npoptosis-lnducing

I . C E T tR E G U T A I I O N T-cel I speciolizolion Helper There is abundant evidence to suggest that different populations of Th cells are specialized for different helper functions. With respect to help for antibody production, T-cell lines derived from Peyer's patches are much better at helping IgA-producing B-cells from Peyer's patch precursors than are splenic T-cell lines. The help is for IgA-precommitted B-cells rather than for induction of the classswitch to IgA, since Peyer's patch T-cells do not markedly enhance IgA production by splenic B-cells. It should also be evident from the previous discussion of AICD that Th cells will not be around to expand B-cell and Tc clone sizes indefinitely.

214

CHAPTER I O - C O N T R O LM E C H A N I S M S

Deothreceplor induced opoptosis

Stress induced opoplosis

l8[

(CDr 78) FASL

(CD95) FAS

I DISC

l l Cospose

tt -.'t

Effecl cosp(

rnnnSLomlolr ro Errroor,rLLcLense Figure 10.5. Activation-induced cell death. Receptor-basedinduction of apoptosis involves the trirnerization of TNF-R family members (e g. Fas)by trimerized ligands (e.g FasJigand). This brings together cytoplasmic death domains (DD) which can recruit a number of death effector domain (DED)-containing adaptor molecules to form the death-inducing signaling complex (DISC) The different receptors use different combinations of DED-containing adaptors; Fas uses FADD (Fas-associated protein with death domain). The DISC, which also includes caspase10, induces the cleavage of inactive procaspase 8 lnto active caspase 8 with subsequent activation of downstream effector caspases This process eventually leads to the release of the endonuclease known as caspase-activated DNase (CAD) from a restraining protein (inhibitor of CAD; ICAD) in the cyioplasm, with subsequent translocation of the endonuclease to the nucleus A second pathwav of

I-cell suppression It is perhaps inevitable that nature, having evolved a functional set of T-cells which promote immune responses,should also develop a regulatory set whose job would be to modulate the helpers. T-cell mediated suppressionwas first brought to the serious attention of the immunological fraternity by a phenomenon colorfully named by its discoverer, Dick Gershon, as 'infectious tolerance'.Quite surprisingly it was shown that, if mice were made unresponsive by injection of a high dose of sheepred blood cells (SRBC),their T-cellswould suppress specific antibody formation in normal recipients to which they had been transferred (figure 10.7).It may not be apparent to the reader why this result was at all surprising, but at that time antigen-induced tolerance was regarded essentially as a negative phenome-

apoptosis induction, often triggered by cellular stress, involves a number of mitochondria-associated proteins including cytochrome c, Smac/DIABLO and thebcl-2 family memberbax. Caspase9 activation is the key event in this pathway and requires association of the caspase with a number of other proteins including the cofactor Apaf-1; the complex formed incorporates cytochrome c and is referred to as the apoptosome The activated caspase 9 then cleaves procaspase 3 Although the death receptor and mitochondrial pathways are shown as initially separate in the figure, there is cross-talk between them. Thus, caspase 8 can cleave the bc1-2 family member bid, a process which promotes cytochrome c release from mitochondria. Other members of the bcl-2 family, such as bcl-2 itself and bci-Xl, inhibit apoptosis, perhaps by preventing the release of pro-apoptotic molecules from the mitochondrra M, mitochondrion.

non involving the depletion or silencing of clones rather than a state of active suppression. Over the years, T-cell mediated suppression has been shown to modulate a variety of humoral and cellular responses,the latter including delayedtype hypersensitivity, cytotoxic Tcells and antigen-specific T-cell proliferation. However, the existence of dedicated professional T-suppressor cells is a question which has generated a great deal of heat. Suppressorandhelper epitopes cenbe discrete Detailed analysis of murine responses to antigens such as hen egg-white lysozyme tells us that certain determinants can evoke very strong suppression rather than help depending on the mouse strain, and also that T-cell mediated suppression directed to one determinant can switch off helper and antibody responses to other deter-

215

I O - C O N T R O tM E C H A N I S M S CHAPTER

8 see-sow TheFLIP/Cospose

Cospose I

Figure 10.6. Life and death decisions. (a) The relative amounts of anti-apoptotic FLIP and pro-apoptotic caspase8 can determine the fate of the cell. (b) Experiments involving overexpression of FLIP in transgenic mice indicate that this protein protects T-cells from AICD stimulated through the death receptor pathway by Fasligand, but not from cell death triggered via the mitochondrial pathway using the drug staurosporine. (Based on data obtained byf Tschopp and colleagues.)

Figure 10.7. Demonstration of T-suppressor cells. Amouse of an appropriate strain immunized with animmunogenic dose of sheep erythrocytes makes a strong antibody response. However, if spleen cells from a donor of the same strain previously injected with a high dose of antigen are first transferred to the syngeneic animal, they depress the antibody response to a normally immunogenic dose of the antigen. The effect is lost if the spleen cells are first treated with a T-cell-specific antiserum (anti-Thy-1 ) pius complement, showing that the suppressors are T-cells (After Gershon R K. & Kondo K (1,971,)lmmunology 21, 903-914.)

polhwoy Milochondriol

pothwoy Deothreceptor

Wild-type

Wild-lype

FLIP Tronsgenic d

d o

FLIP Ironsgenic

(pu) -----> Siourosporine

Fos-L(pg/ml)---->

ffi'---\

/ _-.Xd nolron \-

minants on the same molecule. Thus m ice of H-2bhaplotype respond poorly to lysozyme because they develop dominant suppression; howevet if the three N-terminal amino acids are removed from the antigen, these mice now make a splendid response, showing that the Tregulation directed against the determinant associated with the N-terminal region has switched off the response to the remaining determinants on the antigen. Similar results have been obtained in several other systems. This must imply that the antigen itself acts as a form of bridge to allow communication between regulatory T-cells and cells reacting to the other determinants,

with lreoled Cells onti'Thy-l C

tr1t--\ /

--H

as might occur through these cells binding to an antigenpresenting cell expressing several different Processed determinants of the same antigen on its surface (figure 10.8). Ch ar acteristics of suppre ssion Originally, suppressor T-cells in mice were found to possessLy2 (now called CDScr)and Ly3 (CD8p) on their surface.As researchersbegantocharacterizetheseCD8* T-suppressor cells they were described as expressing a molecule called I-] encoded within the MHC region and able to produce soluble suppressor factors that were

216

CHAPTER I O - C O N T R O LM E C H A N I S M S

\,r' \./

,,S'JfrilX,. CELL

SUPPRESSOR EPITOP

ELPER EPITOPE ANTIGEN Figure 10.8. Possible mechanism to explain the need for a physical linkage between suppressor and helper epitopes. The helper and suppressor cells can interactbybinding close together on the surface of an antrgen-presenting cell, which processes the antigen and displays the different epitopes on separate MHC molecules on its surface

frequently antigen-specific. These suppressor factors proved impossible to define biochemically, and when the entire MHC was cloned it was found that I-| didn't exist! There then, perhaps not unsurprisingly under the circ*mstances,followed a period of extreme skepticism regarding the very existence of suppressor T-cells. However, during the last decadethey have made a dramatic comeback,although it is now appreciatedthat the majority of these cells belong to the CD4 rather than the CD8 T-cell lineage, and the current vogue is to refer to them as regulatory T-cells (Tregs).The characterization of these cells has itself, however. not been without its problems and there seemto be severaldifferent types of Tregs.Whilst some require cell-cell contact in order to be suppressive, others depend upon soluble cytokines to mediate their effect. Let us first look at CD8+ T suppressors.One experimental example that relates to the 810.,4 (2R) mouse strain which has a low immune response to lactate dehydrogenase B GDHB) associatedwith the possessionof the H-2EB gene of k rather than b haplotype. Lymphoid cells taken from these animals after immunization with LDHB proliferate poorly in oitro in the presence of antigen, but if CD8+ cells are depleted, the remaining CD4* cells give a much higher response.Adding back the CD8+ cells reimposes the active suppression. Human suppressor T-cells can also belong to the CD8

Figure 10.9. CD8* suppressor T-cells can inhibit T-cell activation by B-cells. Upon stimulation by peptide-MHC through the T-cell receptor, the CD40 ligand (CD40L, CD154) is upregulated on the partially activated Th2 cell. Interaction with CD40 on B-cel1sleads to NFrBmediated upregulation of 87 1 and 87 2 (CD80 and CD86, respectively) on the B-cell leading to mutual stimulation of the helper T-cell and the B-cell through B7-CD28 interactions In the presence of CD8*CD28- suppressor T-cells there is a failure to upregulate 87 and the lack of costimulation results in the Th2 cel1sbecoming anergic. Unlike T-cells in various stages of activation, resting anergic T-cells do not expressCD40L.

subset. Thus, CD8+ CD28- cells can prevent antigenpresenting B-cells from upregulating costimulatory 87 molecules in response to CD4O-mediated signals from the CD40 ligand on Th cells (figure L0.9).Following interaction with the CD8 T-cells, the APCs are then capable of inducing anergy in Th cells. This effect on B7 is mediated by inhibition of NFrB activation in theAPC, an event necessary for transcription of both the B7.1 (CD80) andB7.2 (CD86) genes. Although it is clear from such experiments that CD8* T-cellscan mediate suppression,the current view is that regulatory CD4* cells are perhaps the major effectorsof suppression. If anti-CD2s and complement are used to deplete the CD25* cells from the lymph nodes or spleen of BALB/c mice and then the remaining CD25 cells transferred into athymic (nude) BALB/cmice,the recipients develop multiple autoimmune diseases.However, if CD4*CD25* cells are subsequently given shortly after the CD25- cells the mice do not develop autoimmune disease, suggesting that the CD4+CD25* population

I O - C O N I R O LM E C H A N I S M S CHAPTER

217I

mechanism they use to suppress immune responsesto either the initiating antigen or to other antigens is still being established, but usually requires cell-cell contact between the regulator and the regulated. Several other types of Tregs have also been described, many of which do not require cell-cell contact (figure 10.10).Human CD4 cells stimulated with antigen in the presenceof IL10 can develop into Ti1 cells which themselves secrete IL-10, a cytokine that can mediate immunosuppressive functions. Although these cells constitutively exPress Foxp3, they only express CD25 upon activation. Th3 cells are defined by the fact that they secrete TGFB, another cytokine with the capacity to be immunosuppressive. Immunoregulatory y6 T-cells may recognize activation-induced nonclassical MHC molecules, as appearsto be the casewith the recognition of the class Ib molecules T22 and T10 in the mouse, although the precise role of such cells is still being elucidated. It is, however, known that some Y6 T-cells secrete IL-10 and TGFB and such cells would clearly have the potential for immunosuppression. Ongoing researchwill hopefully help clarify the roles of these different types of regulatorv cell.

contains Tiegs. Many similar experiments have established that CD4+CD25* T-cells do indeed include a population of Tregs able to mediate suppression of autoimmunity, allograft rejection and allergic responses.However, becauseCD25 (the cx-chainof the IL-2 receptor) is a general marker of cell activation, it is notpossible to use this as a defining molecule for the regulatory subset. These naturally occurring Tiegs also express CTLA-4, OX40, GITR @lucocorticoid-lnduced TNF receptor family related molecule), cell surface TGFB and the forkhead transcription factor Foxp3. Whilst the presenceof eachof thesemolecules is helpful in defining the regulatory population, it is their expression of Foxp3 which is currently thought to uniquely define these cells as regulatory T-cells. Indeed, if the Foxp3 gene is introduced into naive CD4*CD25 T-cells they are converted into cells capableof suppressing the development of autoimmune disease in a number of animalmodels. The activation of CD4*CD25+Foxp3*naturally occurring Tregs is usually antigen-specificbut they can subsequently suppress the responsesto other antigens, a situation referred to as linked suppression.The precise

(b)

tL-l0

rGFp

rL-4,rr-r0, TGFp, rFNy

Figure 10.10. T-cell mediated control of immune responses.Th1 and Th2 cells tend to show a mutual antagomsm based upon the fact that some of the cytokines they produce downregulate the opposing population Superimposed on this are a number of different types of suppressor/regulatory T-cell populations. These may include: (a) IL-10 secreting Tr1 cells which acquires CD25 expression upon activation; (b) the naturally occurring Tregs whrch arise in the thymus and constitutively expressCD25 and Foxp3 Their mode of suppression usually requires cell-cell contact, perhaps utilizing a membrane-bound

version of TGFB; (c) Th3 cells which functionvia their ability to secrete TGFp; (d) CD8 cells which may suPPress using cytotoxicity or cytokrnes; (e) immunosuppressive T-cells bearing a y6 TCR; and (f) NKT cells for which cytokine and cytotoxicity mediated modes of operation have been proposed Whilst it is thought that these varrous tvpes of T-cell can act directly on effector T-ce1ls,they (and perhaps other) populations of suppressor/regulatory T-cells may aiso function r.ia effects on dendritic cells

I zra

CHAPTER I O - C O N T R O LM E C H A N I S M S

Suppression is a regulated phenomenon We have already entertained the idea that antigenlinked T-T interactions can occur on the surface of an antigen-presenting cell (figure 10.8)and the concept of Tc1 and Tc2 CD8 subsets paralleling the Th1/Th2 dichotomy. Furthermore, there is mutual antagonism (suppression) between Th1 and Th2 cells One could postulate downregulation of Th1 cells by type 2IL-4producing CD8 cells, and suppression of Th2 cells by type 1 IFNy-producing CD8 cells, interacting on the surface of an antigen-presenting cell (figure 10.11).In this model, when the immune responsehas locked onto a particular mode, e.g.Th1-mediated cellular immunity, other types of response, such as T-B collaboration, are restricted through a cytokine inhibitory effect. Although these cells mediate T-suppression, they would notbe called dedicated professional suppressors since, in a sense, their suppressive powers are a byproduct of their main defensive function. Perhaps we need these cytokine-secreting Tc cells to prevent Th cells getting out of hand by excessiveproliferation, just as IgG holds back the B-cellsby feedbackcontrol.

ANTIGEN.PRESENTING CELL

Figure 10.11. Mutual antagonisms between T-cell subsets hnked indirectly by processed antigen on an antigen-presenting cell lead to functionaily distinct modes of suppression (Leaning heavily on Bloom B R., Salgame P & Diamond B (1992) Immunology Todny L3, 131-136 ) Yet another mechanism may pror.e to be important Unhke the mouse, many other mammalian species can express MHC class II on a proportion of their activated T-cells; presentation of processed peptide by thesecells can induce CD4+ cytotoxic cellswith suppressor potential We also need to know more about the circ*mstances leading to the production of TGFp by suppressors since this cytokine inhibits T-cell proliferation

The activities of Tregs are to some extent controlled by dendritic cells, with resting dendritic cells favoring the development of a regulatory phenotype. The activation of dendritic cells which occurs through microbial engagementof pattern recognition receptorswill lead to the production of IL-6 and other soluble mediators which can curb the Tregs when an anti-pathogen responseis required.

IDIOIYPE NETWORKS f erne'snetworkhypothesis In 1974the Nobel laureateNeils Jernepublished a paper 'Towards entitled a network theory of the immune system' in which he proposed that structures formed by the variable regions of antibodies (i.e.the antibody idiotype) could recognizeother antibodyvariable regions in such a way that they would form a network based upon mutual idiotype-anti-idiotype interactions. BecauseBcellsuse the antibody molecule as their antigen receptor this would provide a connectivity between different clones of B-cells, and therefore the potential for regulation of the individual clones that are members of the network (figure 10.12).This concept was later extended to include the idiotypes presenton the T-cellreceptorsof both CD4* and CD8+ T-cells.There is no doubt that the elements which can form an idiotypic network are present in the body, and autoanti-idiotypes occur during the course of antigen-induced responses.For example, certain strains of mice injected with pneumococcalvaccinesmake an antibody responseto the phosphorylcholine groups in which the germ-line-encoded idiotype T15 dominates. Waves of T15+and of anti-Tl5 (i.e. autoanti-idiotype) cells are demonstrable. Antiidiotypic reactivity has alsobeen demonstrated in T-cell populations using various experimental systems. Indeed, anti-idiotypic T-cells recognizing peptide derived from the CDR2, CDR3 and framework regions of other TCRs appear to form a part of the normal human T-cell repertoire. Anetzuork is eaident in early life If the spleens of fetal mice which are just beginning to secreteimmunoglobulin are used to produce hybridomas, an unusually high proportion are interrelated as idiotype-anti-idiotype pairs. This high level of idiotype connectivity is not seen in later life and suggests that theseearly cells,largely the CD5+B-1 subset (cf.p.244), are programed to synthesize germ-line gene specificities which have network relationships. P ria at e and p ublic idioty p es Whilst certain idiotypes (private idiotypes) are present

CHAPTER I O - C O N T R O LM E C H A N I S M S

Figure 10.12. Elernents in an idiotypic network in which the antigen receptors on one lymphocyte reciprocally recognize an idiotype on the receptors of another. T-helper, T-suppressor and B-lymphocytes interact through idiotype-anti-idiotype reactions producing either stimulation or suppression T-T interactions could occur through direct recognition of one T-cell receptor (TCR) by the other, or more usually by recognition of a processed TCR peptide associated with MHC. One of the anti-idiotype sets,Abrp, may bear an idiotype of similar shape to (i e provides an intemal image of ) the antigen. The same idiotype (?)

maybe shared by receptors of different specificity, the nonspecific parallel set (since the several hypervariable regions provide a number of potential idiotypic determinants and a given idiotype does not always form part of the epitope-binding site, i.e. the paratope), so thatthe anti(anti-Idr) does not necessarily bind the original antigen (The following abbreviations are often ernployed: cras aprefix = anti; Id = idiotyPe; Ab, = Id; {!161= crld not involving the paratope; Abrp = internal image old involving the paratope; Ab, = a(old) )

only on antibodies of a defined single specificity, crossreacting idiotypes (public idiotypes) are present on a variety of antibodies (and thus B-cell receptors) of different specificities. Such frequently occurring and usually germ-line-encoded public idiotypes seem to be provoked fairly readily with anti-Id and are therefore candidates for regulatory Id which can be under some degree of control by a limited idiotypic network. The phenomenon of 'original antigenic sin' occurs when the immune responsebecomes'locked in' to particular epitopes originally encountered on a microorganism, such that it largely ignores even normally immunodominant epitopes during a subsequent encounter with an antigenically related but nonidentical microorganism. Although competition for antigenby the expanded population which forms the memory B-cells plays a major role, idiotype-specific memory Th cells could also contribute to this phenomenon. Idiotype networks may also, by a mutual low-level stimulation of lymphocytes within a network, allow the immune response to 'tick over' for extended periods and maintain the memory cell population.

M anipul ation of the immune response throughidiotypes Quite low doses of anti-idiotype, of the order of nanograms, can greatly enhance the expression of the idiotype in the response to a Siven antigen, whereas doses in the microgram range lead to a suPPression (figure 10.13). Thus the idiotypic network provides interesting opportunities to manipulate the immune response, particularly in hypersensitivity states such as autoimmune disease, allergy and graft rejection. However, the B-cell response is normally so diverse, suppression by anti-Id is likely to prove difficult; even when the response is dominated by a public Id and that Id is suppressed, compensatory exPansion of clones bearing other idiotypes ensures that the fall in the total antibody titer is relatively undramatic (cf. figure 10.13). Conceivably, Th cells may exPressa narrower sPectrum of idiotypes, thereby being more susceptible to suppressionby Id autoimmunization. Reports that'vaccination' with irradiated lines of Th cells specific for brain or thyroid antigens prevents the induction of experimental autoimmunity against the relevant organ are encourag-

CHAPTER I O - C O N T R O IM E C H A N I S M S ing. A totally different approach would be to use monoclonal anti-Id of the 'antigen internal image' set (figure 10.12) to stimulate antigen-specific regulatory T-cells capable of turning off B-cells directed to other epitopes on the antigen through bridging by the antigen itself (cf . figure 10.8).

Since we know that under suitable conditions anti-Id can also stimulate antibodyproduction, itmightbe pos'internal sible to use image' monoclonal anti-Ids as 'surrogate' antigens for immunization in cases where the antigen is difficult to obtain in bulk-for example, antigens from parasites such as filaria or the weak embryonic antigens associated with some cancers.Another example is where protein antigens obtained by chemical synthesis or gene cloning fail to fold into the configuration of the native molecule; this is not a problem with the anti-Id whichby definition has been selectedto have the shape of the antigenic epitope. The main factors currently thought to modulate the immune responseare summarized in figure 10.14.

WEEKS 067 lnjecl onli-idiotype

hjecl 0nTrgen

> ANTIBODY LEVEL

LOWDOSE

rd- lrdl IHEINFTUENC OEFG E N E T IFCA C I O R S HIGHDOSE

genes Some otfeclgenerol responsiveness

T----l --T---l

IRRELEVANT old

Mice can be selectively bred for high or low antibody responses through several Benerations to yield two lines, one of which consistently produces high-titer antibodies to a variety of antigens, and the other, antibodies of relatively low titer (figure 70.75; Biozzi and colleagues). Out of the ten or so different genetic loci involved, some give rise to a higher rate of B-cell proliferation and differentiation, while one or more affect macrophagebehavior.

Figure 10.13. Modulation of a major idiotype in the antibody response to antigen by anti-idiotype. In the example chosen, the idiotype is present in a substantial proportion of the antibodies produced in controls injected with rrrelevant anti-Id plus antigen (i e this is a public or cross-reacting Id; see p. 52) Pretreatment with 10 ng of a monoclonal anti-Id greatly expands the Id+ antibody population, whereas prior injection of 10 pg of anti-Id almost completely suppresses expression of the idiotype without having any substantial effect on total antibody production due to a compensatory increase in Id- antibodv clones

Anligenreceplorgenesore linkedlo lhe immuneresponse Clearly, the Ig and TCR y, D and / genes encoding the specific recognition sites of the lymphocyte antigen

Slimulotion ANTIGENPRESENTING CELL

I ANTI-IDIOryPE I (NETWORK) ^l

t-, I

t--_______ ldiolype recognilion

Figure 10.14. Regulation of the immune response. T-help for cell-mediated immunity will be subject to similar regulation. Some of these mechanisms maybe interdependent; for example, one could envisage regulatory T-cells with specificity for the idiotlpe on Th or B-cell.s To avoid too many confusing arrows, we have omitted the recruitment of B-cells by anti-idiotypic Th cells. Th, T-helper cell; Treg, regulatory T-cell

22r I

CHAPTER I O-CONIROI MECHANISMS

o o ^ o

@ @ @ @

@ o e

o o E a4

Figure 10.15. Selective breeding of high and low antibody responders (after Biozzi and colleagues) A foundation population of wild mice (with crazy mixed-up genes and great variability in antibody response) is immunized with sheep red blood cells (SRBC), a multideterminant antigen The antibody titer of each individual mouse is shown by a circle. The male and female giving the highest titer antibodies () were bred and their litter challenged with antigen. Again, the best responders were bred together and so on for 20 generations when all mice were high responders to SRBC and a variety of other

receptors are of fundamental importance to the acquired immune response. However, since the mechanisms for generating receptor diversity from the available genes are so powerful (cf. p. 68), immunodeficiency is unlikely to occur as a consequence of a poor Ig or TCR variable region gene repertoire. Nevertheless, just occasionally, we see holes in the repertoire due to the absence of a gene; failure to respond to the sugar polymer a7-6 dextran is a feature of animals without a particular immunoglobulin V gene, and mice lacking the Vu, TCR gene cannot mount a cytotoxic Tcell response to the male H-Y antigen. lmmuneresponse conbe influenced bytheMHC There was much excitement when it was first discovered that the antibody responses to a number of thymus-dependent antigenically simple substances are determined by genes mapping to the MHC. For example, mice of the H-2bhaplotype respond well to the sl,nthetic branched polypeptide (!G)-A-L, whereas H-2kmice respond poorly (table 10.1). It was said that mice of the H-2bhaplotype (i.e. a particular set of H-2 genes) are high responders to (T,G)A-L because they possess the appropriate immune

@ @ @

&@ o a

o o o o O a

Low RESPOl{DERS

antigens. The same was done for the poorest responders (f), yielding a strain of low responder animals The two lines are comparable in their ability to clear carbon particles or sheep erythrocytes from the blood by phagocytosis, but macrophages from the high responders present antigen more efficiently (cf p 167).On the other hand, the low responders survive infection by Salmonella typhimuriurn better and their macrophages support much slower replication of Listeria (cf p.268), indicative of an inherently more aggressive microbicidal ability.

Table 10.1. H-2 haplotype linked to high, low and intermediate immune responses to synthetic peptides. (lG)-A-L, polylysine with polyalanine side-chains randomly tipped with tyrosine and glutamine; (H,G)-A-L, the same with histidine in place of tyrosrne.

response (1r) gene. With another synthetic antigen, (H,G)-A-L, having histidine in place of tyrosine, the 'poor (T,G)-A-L responders' position is reversed, the 'good now giving a good antibody resPonse and the GG)-A-L responders' a weak one/ showing that the capacity of a particular strain to give a high or low responsevaries with the individual antigen (table 10.1). These relationships are only apparent when antigens of highly restricted structure are studied because the response to each single determinant is controlled by an 1rgene and itis less likely that the different determinants on a complex antigen will all be associated with consistently high or consistently low responder 1r genes; however, although one would expect an average of

I zzz

CHAPTER I O - C O N T R O LM E C H A N I S M S

randomly high and low responder genes, since the various determinants on most thymus-dependent complex antigens are structurally unrelated, the outcome willbebiasedby the dominance of one or more epitopes (cf . p. 277). Thus H-2linked immune responses have been observed not only with relatively simple polypeptides, but also with transplantation antigens from another strain and autoantigens where merely one or two determinants are recognized as foreign by the host. With complex antigens, in most but not all cases,H-2 linkage is usually only seenwhen the dose administered is so low that iust one immunodominant determinant is recognizedby the immune system.In this way, reactions controlled by 1r genes are distinct from the overall responsivenessto a variety of complex antigenswhich is a feature of theBiozzimice (above). The lr genesmap to the H-21 region qnd control T-B cooperation Table 10.2gives some idea of the type of analysisused to map the 1r genes.The three high responder strains have individual H-2 genes derived from prototypic pure strains which have been interbred to produce recombinations within the H-2 region. The only genes they have in common areAk and Db; since the 8.10 strain bearing the D'gene is a low responder, high responsemust be linked in this caseto possessionof Ak. The I region molecules must represent the 1r gene product since a point mutation in the H-2A subregion in one strain led to a change in the class II molecule at a site affecting its polymorphic specificity and changed the mice from high to low responder status with respect to their thymusdependent antibody response to antigen in aiao. The mutation also greatly reduced the proliferation of Tcells from immunized animals when challenged inaitro with antigen plus appropriate presenting cells, and there is a good correlation between antigen-specific Tcell proliferation and the responder status of the host. The implication that responder status may be linked to the generation of Th cells is amply borne out by adop-

Table 10.2. Mapping of the Ir gene for (H,G)-A-L responses by analysis of different recombinant strains.

(H,G)-A-L Response

H-2 region A E

A,TL

flir

B r0A (4R)

Br0

ASW

Slroin

tive transfer studies showing that irradiated (H-2bxH2k)F1 mice make good antibody responsesto (T,G)-A-L when reconstituted with antigen-primed B-cells from another F1 plusT-cellsfrom aprimed H-2b (highresponder); T-cellsfrom the low responder H-2k mice only gave poor help for antibody responses.This also explains why these H-2 gene effects are seen with thymusdependent but not T-independent antigens. Three mechanisms can account for class Illinked high and low responsiveness. I Defectiaepresentation.Ina high responder,processing of antigen and its recognition by a corresponding T-cell lead to lymphocyte triggering and clonal expansion (figure 10.16a).Although there is (and has to be) considerable degeneracy in the specificity of the class II groove for peptide binding, variation in certain key residues can alter the strength of binding to a particular peptide (cf.p. 101)and convert a high to a low responder because the MHC fails to present antigen to the reactive T-cell (figure 10.16b).Sometimesthe natural processing of an antigen in a given individual does not produce a peptide which fits well into their MHC molecules. One study showed that a cytotoxic T-cell clone restricted to HLA-A2, which recognized residues 58-68 of influenza A virus matrix protein, could cross-react with cells from an HLA-A69 subject pulsed with the same peptide; nonetheless,the clone failed to recognize HLA-A69 cells infectedwith influenza A virus. Interestingly, individuals with the HLA-A69 class I MHC develop immunity to a different epitope on the same protein. 2 DefectiaeT-ceIIrepertoire.T-cells with moderate to high affinity for self-MHC molecules and their complexes with processed self-antigens are tolerized (cf. p. 237), 'hole' in the T-cell repertoire. If there is a so creating a cross-reaction,i.e. similarity in shape at the T-cell recognition level between a foreign antigen and a selfmolecule which has already induced unresponsiveness, the host will lack T-cells specific for the foreign antigen and thereforebe a low responder (figure 10.16c).To take a concrete example, mice of DBA / 2 str ain respond well to the synthetic peptide polyglutamyl, polytyrosine (GT), whereas BALB/c mice do not, although both have identical class II genes.BALB/c B-cell blasts express a structure which mimics CT and the presumptionwould be that self-tolerance makes these mice unresponsive to GT. This was confirmed by showing that DBA/2 mice made tolerant by a small number of BALB/c hematopoietic cells were changed from high to low responder status. To round off the story in a very satisfying way, DBA/2 mice injected with BALB/c B-blasts, induced by the polyclonal activator lipopolysaccharide, were found tobe primed for GT.

CHAPTER I O _ C O N T R OM L ECHANISMS

HGHREsPoNDER

LoWRESPoNDER

slimulrtlon

T-CELL

2 2 g II

Agprocessing Microbicidol porency

T,CELL RECEPTOR ANTIGENIC PEPTIDE

MHCN

MHCII

Anfigen bindsMHC T-cellseesonligen

Anligen doesnotbindMHC Noontigen forT-cell lo see

LowRESPoNDER +NOT-CELL Figure 10.17. Genetic control of the immune response,

MHCII

Selfpeptide + MHC lolerizes immolure T-cell

MHCX peplide+ MHC X-reocting foreign NoT-celllo seeontigen

result of regulatory cell activity (figure 10.16d). Low responsecanbe dominantin classIlheterozygotes, indicating that suppression can act against Th restricted to any other class II molecule. In this it differs from models 1 and 2 above where high response is dominant in a heterozygote because the factors associated with the low responder gene cannot influence the activity of the high responder. Factors influencing the genetic control of the immune response are summari zed in figure 70.17.

R E G U T A I O IRMYM U N O N E U R O E N D O C R I N E NE T W O R K S

T-cell seesontigen buflriggering suppressed byTregcell Figure 10.16. Different mechanisms can account for low T-cell response to antigen in association with MHC classIL

3 T-suppression. We would like to refer again to the MHC-restrictedlow responsiveness which canoccurto relativelycomplexantigens(seep. 214),sinceit illustratesthe notion that low responderstatuscanariseasa

There is a danger, as one focuses more and more on the antics of the immune system, of looking at the body as a collection of myeloid and lymphoid cells roaming around in a big sack and of having no regard to the integrated physiology of the organism. Within the wider physiological context, attention has been drawn increasingly to interactions between immunological and neuroendocrine systems. Immunological cells have the receptors which enable them to receive signals from a whole range of hormones: corticosteroids, insulin, growth hormone, estradiol, testosterone, prolactin, p-adrenergic agents, acetylcholine, endorphins and enkephalins. By and large, glucocorticoids and androgens depress immune responses, whereas estrogens, growth hormone, thyroxine and insulin do the opposite.

Aneuroendocrine feedb0ck immune responses l00poffecling The secretion of glucocorticoids is a major responseto stresses induced by a wide range of stimuli, such as extreme changes of temperature, fear, hunger and physical injury. They are also released as a consequence of immune responses and limit those responses in a neuroendocrine feedbackloop. Thus,IL-1 (figure 10.18),IL6 and TNF are capable of stimulating glucocorticoid synthesis and do so through the hypothalamic-pituitary-adrenal axis. This, in turn, leads to the downregulation of Th1 and macrophage activity, so completing the negative feedback circuit (figure 10.19). However, the glucocorticoid dexamethasone can prevent activation-lnduced cell death (AICD) in T-cells by inducing expression of GILZ (glucocorticoid-lnduced leucine zipper). The situation is therefore somewhat complex becauseglucocorticoids can themselvestrigger apoptosis in T-cells, yet counteract apoptosis activated by peptide-MHC interaction with the TCR. In the absence of glucocorticoid, the activation of T-cells by peptide-MHC leads to a progressive loss of GILZ and eventual cell death by apoptosis. By contrast, if activation through the TCR occurs in the presence of glucocorticoids, then expression of GILZ is increased and this directly inhibits activation and nuclear translocation of NFrB, thereby protecting the cells fromAICDIt has been shown that adrenalectomy prevents

spontaneous recovery from experimental allergic encephalomyelitis (EAE). This demyelinating disease is associated with progressive paralysis and is produced by immunization with myelinbasic protein in complete Freund's adjuvant. Induction of the disease can be blocked by implants of corticosterone. Spontaneous recovery fromEAE in intact animals is associated with a dominance of Th2 autoantigen-specific clones, indicative of the view that glucocorticoids suPPress Th1 and may augment Th2 cells. Individuals with a genetic predisposition to high levels of stress-induced glucocorticoids might therefore be expected to have increased susceptibility to infections with intracellular pathogens

Hypolholomus

piluilory Anlerior

2 I

0.25 0.50 (t"rg/onimot) rtL-1

Figure 10.1E. Enhancement of ACTH and corticosterone blood levels in C3H/HeJ mice 2 hours after injection of recombinant IL-1 (values are means + SEM for groups of seven or eight mice) The significance of the mouse strain used is that it lacks receptors for bacterial lipopolysaccharide (LPS), and so theeffects cannotbe attributed to LPS contamination of the IL-1 preparation. (Reprinted from Besedovsky H., del Rey A , Sorkin E & Dinarello C A (1986) Science233,652454, with permission Copyright @ 1986by the AAAS.)

Figure 10.19. Glucocorticoid negative feedback on cytokine production. Additional regulatory circuits based on neuroendocrine interactions with the immune system are almost certain to exist given that lymphoid and myeloid cells inboth primary and secondary lymphoid organs can produce hormones and neuropeptides, and classical endocrine glands as well as neurons and glial cells can synthesize cytokines and appropriate receptors Production of prolactin and its receptors by peripheral lymphoid cells and thyrnocytes is worthy of attention. Lymphocyte expression of the prolactin receptor is upregulated following activation and, in autoimmune disease, witness the beneficial effects of bromocriptine, an inhibitor of prolactin synthesis, in the NZB x W model of murine systemic lupus erythematosus (cf. p. 82). CRH, corticotropin-releasing hormone; ACTH, adrenocorticotropic hormone

CHAPTER I O - C O N T R O LM E C H A N I S M S such asMyco bacterium Iepraewhichrequire effective Th1 cell-mediated immunity for their eradication. Neonatal exposure to bacterial endotoxin (LPS) not only exerts a long-term influence on endocrine and central nervous system development, but substantially affects predisposition to inflammatory disease and therefore appears to program or'reset' the functional development of both the endocrine and immune systems. Thus, in adult life, rats which had been exposed to endotoxin during the first week of life had higher basal levels of corticosterone compared with control animals, and showed a greater increasein corticosteronelevels in responseto noise stressand a more rapid rise in corticosterone levels following challenge withLPS. Sexhormones comeinlolhe piclure Females are far more susceptible to autoimmune disease,an issue that will be discussedin greater depth in Chapter lS,buthere let us note that estrogenreceptors are present on various cell types in the immune system, including lymphocytes and macrophages. Although investigations into the role of estrogen in immune responses have often led to apparently contradictory data, some of its more clearly establishedeffectson different populations of lymphocytes are highlighted in figure 10.20.It has often been found to enhance T-cell proliferation, B-cell survival, and humoral responses. Conversely, androgen deprivation induced by castration of postpubertal male mice increasesthe levels of Tcells in secondary lymphoid tissuesand enhancesT-cell proliferation. Estrogen also has effects on regulatory cells, expanding CD4*CD25*T-cells and upregulating Foxp3 expression. Another type of regulatory cell is the NKT cell that

produclion onlibody Upreguloted

Figure 10.20. Some effects of estrogen on lymphocyte function.

bears an invariant TCR, responds to the synthetic glycolipid a-galactosylceramidepresented by CDld (cf. p. 83), and is able to produce both the Th1 cytokine interferon-y (IFNy) and the Th2 cytokine IL-4. Substantially enhanced levels of IFNy are produced by these cells when stimulated with antigen in the presence of physiological concentrations of estrogen,providing a possible explanation for the observation that female mice produce higher levels of this cytokine in response to antigen challenge.

'Psychoimmunology' The thymus, spleen and lymph nodes are richly innervated by the sympathetic nervous system. The enzyme dopamine B-hydroxylase catalyses the conversion of dopamine to the catecholamine neurotransmitter norepinephrine which is released from sympathetic neurons in these tissues.Mice in which the gene for this enzyme has been deleted by hom*ologous recombination exhibited enhanced susceptibility to infection with the intracellular pathogen Mycobacteriumtuberculosis and impaired production of the Th1 cytokines IFNy and TNF in response to the infection. Although these animals showed no obvious developmental defects in their immune system,impaired Th1 responseswere also found following immunization of these mice with the hapten TNP coupled to KLH. These observations suggest that norepinephrine can play a role in determining the potency of the immune response. Denervated skin shows greatly reduced leukocyte infiltration in response to local damage, implicating cutaneous neurons in the recruitment of leukocytes. Sympathetic nerves which innervate lymphatic vessels and lymph nodes are involved in regulating the flow of lymph and may participate in controlling the migration of B-adrenergic receptor-bearing dendritic cells from inflammatory sites to the local lymph nodes. Mast cells and nerves often have an intimate anatomical relationship and nerve growth factor causesmast cell degranuIation. The gastrointestinal tract also has extensive innervation and a high number of immune effector cells. In this context, the ability of substance P to stimulate, and of somatostatin to inhibit, proliferation of Peyer's patch lymphocytes may prove to have more than a trivial significance.The pituitary hormone prolactin has also been brought to our attention by the experimental observation that inhibition of prolactin secretion by bromocriptine suppresses Th activity. There seems to be an interaction between inflammation and nerve growth in regions of wound healing and repair. Mast cells are often abundant, IL-6 induces neurite growth and IL-1 enhances production of nerve

I zza

CHAPTER I O - C O N T R O LM E C H A N I S M S

growth factor in sciatic nerve explants. IL-1 also increases slow-wave sleep when introduced into the lateral ventricle of the brain, and both IL-1 and interferon produce pyrogenic effects through their action on the temperature-controlling center. Although it is not clear just how these diverse neuroendocrine effects fit into the regulation of immune responses,at a more physiological level, stressand circadian rhythms modify the functioning of the immune system. Factors such as restraint, noise and exam anxiety have been observed to influence a number of immune functions including phagocytosis, lymphocyte proliferation, NK activity and IgAsecretion. Amazingly, it has been reported that the delayed-type hypersensitivity Mantoux reaction in the skin can be modified by hypnosis. An elegant demonstration of nervous system control is provided by studies showing suppression of conventional immune responses and enhancement of NK cell activity by Pavlovian conditioning. In the classicPavlovianparadigm, a stimulus such asfood that unconditionally elicits a particular response,in this case salivation, is repeatedly paired with a neutral stimulus that does not elicit the same response. Eventually, the neutral stimulus becomes a conditional stimulus and will elicit salivation in the absenceof food. Rats were given cyclophosphamide as a unconditional and saccharin as a conditional stimulus repeatedly; subsequently, there was a depressed antibody response when the animals were challenged with antigen togetherwith just the conditional stimulus, saccharin.As more and more data accumulate, it is becoming clearer how immunoneuroendocrine networks could play a role in allergy and in autoimmune diseasessuch as rheumatoid arthritis, type I diabetes and multiple sclerosis.

E F F E C TOSFD I E TE , X E R C I SIER, A U M AA N D A G EO NI M M U N I I Y M0lnutrilion diminishes theeffectiveness of lheimmune response The greatly increased susceptibility of undernourished individuals to infection can be attributed to many factors: poor sanitation and personal hygiene, overcrowding and inadequate health education. But, in addition, there are gross effects of protein-calorie malnutrition on immunocompetence. The widespread atrophy of lymphoid tissues and the 50% reduction in circulating CD4 T-cellsunderlie serious impairment of cell-mediated immunity. Antibody responsesmay be intact but they are of lower affinity; phagocytosis of bacteria is relatively normal but the subsequent intracelluIar destructionis defective.

Deficiencies in pyridoxine, folic acid and vitamins A, C and E resultingenerallyimpaired immune responses. Vitamin D is an important regulator. It is produced not only by the UV-irradiated dermis, but also by activated macrophages, the hypercalcemia associatedwith sarcoidosis being attributable to production of the vitamin by macrophages in the active granulomas. The vitamin is a potent inhibitor of T-cell proliferation and of Th1 cytokine production. This generates a neat feedback loop at sites of inflammation where macrophages activated by IFN^yproduce vitamin D which suppresses the T-cells making the interferon. It also downregulates antigen presentation by macrophages and promotes multinucleated giant cell formation in chronic granulomatous lesions. Nonetheless/ as a further emphasis of the potential duality of the CD4 helper subsets, it promotes Th2 activity, especially at mucosal surfaces: quite a busy little vitamin. Zinc deficiency is rather interesting; this greatly affects the biological activity of thymus hormones and has a major effect on cell-mediated immunity, perhaps as a result. Iron deficiency impairs the oxidative burst in neutrophils since the flavocytochrome NADP oxidase is an iron-containing enzyme. Of course there is another side to all this in that moderate restriction of total calorie intake and/or marked reduction in fat intake ameliorates age-related diseasessuch as autoimmunity. Oils with an n-3 double bond, such as fish oils, are also protective, perhaps due to increased synthesis of immunosuppressive prostaglandins. Given the overdue sensitivity to the importance of environmental contamination, it is important to monitor the nature and levels of pollution that may influence immunity. Here is just one example: polyhalogenated organic compounds (such as polychlorinated biphenyls) steadily pervade the environment and, being stable and lipophilic, accumulate readily in the aquatic food chain where they largely resist metabolic breakdown. It was shown that Baltic herrings with relatively high levels of these pollutants, as compared with uncontaminated Atlantic herrings, were immunotoxic when fed to captive harbor seals,suggesting one reason why sealsalong the coastsof northwestern Europe succumbed so alarmingly to infection with the otherwise nonvirulent phocine distemper virus in 1988. 0lher f0cfors Exercise, particularly severe exercise, induces stress and raises plasma levels of cortisol, catecholamines, IFNcr, IL-1, p-endorphin and metenkephalin. It can lead to reduced IgAlevels, immune deficiency and increased

I O - C O N T R O LM E C H A N I S M S CHAPTER susceptibility to infection. Maniacal joggers and other such like masoch*sts-you havebeen warned! Multiple traumatic injury, surgery and major burns are also immunosuppressive and so contribute to the

60 Agein yeors

r20l I

120

Agein yeors

Figure 10.21. Age trends in some immunological parameters. (Based on Franceschi C., Monti D., Sansoni P.& CossarizzaA. (7995)Immunology Today76,72-76)

Conlrol byontlgen . Immune responses are largely antigen driven. As the Ievel of antigen falls, so does the intensity of the response. o Antigens can compete with each other: a result of competition between processed peptides for the available MHC grooves. Feedb0ckGlnlrol by complement0nd0nlibody . Early IgM antibodies and C3d boost antibody responses/ whereas IgG inhibits responsesvia the Fcyreceptor on B-cells.

I-cellregulolion o Activated T-cells express members of the TNF receptor family, including Fas, which act as death receptors and restrain unlimited clonal expansion by a process referred to as activation-induced cell death (AICD). . Regulatory T-cells (Tregs) can suppress the activity of helper T-cells, presumably as feedback control of excessive Thexpansion. . Suppressor and helper epitopes on the same molecule can be discrete. o Effectors of suppression include naturally occurring CD4'CD25'Foxp3* cell-contact dependent Tregs, IL-10 secreting Tr1 cells, TGFp-secreting Th3 cells, immunoregulatory y6 T-cells and CD8* suppressors. o Resting dendritic cells can preferentially promote the development of regulatory cells. . Suppression can occur due to T-T interaction on the

227 |

increased risk of sepsis. Corticosteroids produced by stressful conditions, the immunosuppressive prostaglandin E, released from damaged tissues and bacterial endotoxin derived from the disturbance of gut flora are all factors which influence the outcome after trauma. Accepting that the problem of understanding the mechanisms of aging is a tough nut to crack, it is a trifle disappointing that the easier task of establishing the influence of age on immunological phenomena is still not satisfactorily accomplished. Perhaps the elderly population is skewed towards individuals with effective immune systems which give a survival advantage. Be that as it may, IL-2 production by peripheral blood lymphocytes (figure 10.21) and T-cell-mediated functions such as delayed-type hypersensitivity reactions to common skin test antigens decline with age and so, it is thought, does T-cell mediated suppression, although this is a notoriously elusive function to measure.

surfaceof antigen-presentingcells.|ust asThl and Th2 cells mutually inhibit each other through production of their respectivecytokinesIFNyand IL-4/10, sotheremaybe two types of CD8 cellswith supPressoractivity: one of Tc2fype making IL-4 and suppressingTh1 cells,and the other Tc1 cellsmaking IFNycapableof suppressingTh2 cells. Irliotypenelworks . Antigen-specificreceptorson lymphocytes can interact with the idiotypes on the receptorsof other lymphocytesto form a network (Jerne). . Anti-idiotypes canbe inducedby autologousidiotypes. o An idiotype network involving mostly CD5 B-1 cells is evidentin early life. . T-cellidiotpic interactionscanalsobe demonstrated. . Idiotypes which occur frequently and are shared by a multiplicity of antibodies(public or cross-reactingId) are targetsfor regulationby anti-idiotlpes in the network, thus providing a further mechanismfor control of the immune resPonse. o The network offers the potential for therapeutic intervention to manipulateimmunitY. theimmuneresponse foclorsinfluence Genefic . Multiple genescontrol the overall antibody resPonseto complexantigens:someaffectmacrophageantigenprocessing and microbicidalactivity and somethe rate of proliferation of differentiatingB-cells. (Continuedp 228)

I zza

CHAPTER I O - C O N T R O LM E C H A N I S M S

. ImmunoglobuLin and TCR genes are very adaptable because they rearrange to create the antigen receptors, but 'holes'in the repertoire can occur. . Immune response genes are located in the MHC class II locus and control the interactions required for T-B collaboration. o Class ll-linked high and low responsiveness may be due to defective presentation by MHC, a defective T-cell repertoire caused by tolerance to MHC+self-peptides and T-suppression.

. Estrogens can enhance both T- and B-cell responses, but can also promote the activity of regulatory cells. Efleclsof diet ond olherf0clorson immunily . Protein-calorie malnutrition grossly impairs cellmediated immunity and phagocyte microbicidal potency. o Exercise,trauma, age and environmental pollution can all act to impair immune mechanisms. The pattern of cytokines produced by peripheral blood cells changes with age, IL-2 decreasing and TNR IL-1 and IL-6 increasing; the latter is associated with a lowered DHEA level.

lmmunoneuroendocrine nelworks . Immunological, neurological and endocrinological systems interact, forming regulatory circuits. . Feedbackby cytokines augments the production of corticosteroids and is important because this shuts down Th1 and macrophage activity.

C o r t i -s o l

o These have figured with some prominence in this chapter and a summary of some of the major influences on the balance between Th1 and Th2 responses is presented in figure10.22.

il -----l

DHEAsulfote

lhebiosbelween ThI ondTh2subsets Foclors influencing

VitominD 3 A I

F U R I H ER E A D I N G Bevan M J (2004)Helping the CD8+ T-cell response Noture Rexiews lnn tu nology 4, 595-602. Chandra R K (1998) Nutrition and the immune system In Delves PJ & Roitt I M (eds) Encyclopediaof ImtLunology, 2nd edn, pp 1869-1871 Academic Press,London. (Seealso other relevant articles in the Encqclopedrn: 'Aging and the immune system', pp 59-61;'Behavioral regulation of rmmunity', pp. 336-340;'Neuroendocrine regulation of immunity', p 1824; 'Sex hormones and immunity', pp 2775-2778;'Stress and the immune system', pp 2220-2228;'Yitamin D', pp 2494-2499 ) Cohen I R. (2000) Tending Adam's Gnrden Academic Press, London Crowley MP. et tl. (2000) A population of murine yD T cells that recognize an inducible MHC class Ib molecule Science 287, 374-376 Fearon DT & Carroll MC (2000) Regulation of B lymphocyte responses to foreign and self-antigens by the CD19ICD21 complex An n ual Reui ews of Immunolo gy 18, 39 3-422

Figure 10.22. Summaryof major f actors affecting Th1/Th2 balance. Preferential stimulation of mucosal antibody synthesis by vitamin D involves the promotion of dendritic cell migration to the Peyer's patches By downregulating macrophage activity, Th1 effectiveness is decreased. Cortisol and dehydroepiandrosterone (DHEA) are products of the adrenal and have opposing effects on the Th1 subset A relative deficiency of DHEA wiil lead to poor Th1 perforrnance NKT cells bear an crBTCR, the natural killer cell marker NK1 L, and secrete cytokrnes including IL-4 which stimulate Th2 cells

Jiang H & Chess L (2006) Regulation of immune responses by T cells Tfte Nezr Engl and I our nnl of Me dic ine 354, 11,66-177 6 Krammer PH (2000) CD95's deadly mission in the immune system. Noture 407,789-795 O'Garra A & Vieira P (2004) Regulatory T-cells and mechanisms of immune system control Nafure Medicine 1O,801--805 Randolph D.A. & Garrison Fathman C. (2006) CD4* CD25* regulatory T-ce1ls and their therapeutic potential Annual Reaiezu of Medicine 57,38L402 Reiche E.M V, Nunes S.O V & Morrmoto H K (2004)Stress,depression, the immune system, and cancer Lancet Oncology 5,617-625 Shanks N et al (2000) Early life exposure to endotoxin alters function and predisposition to hypothalamic-pituitary-adrenal inflammation Proceedingsof the National Academy of Sciencesof the United Statesof America9T,5645-5650. Steinman L (2004) Elaborate interactions between the immune and nervous systems Nature Imtnunology 5,575-587

Ontogeny ondphylogeny

INIRODUCIION Hematopoiesis originates in the early yolk sac but, as embryogenesis proceeds, this function is taken over by the fetal liver and finally by the bone marrow where it continues throughout life. The hematopoietic stem cell (HSC) which gives rise to the formed elements of the blood (figure 11.1)can be shown to be multipotent, to seed other organs and, under the influence of the cytokine leukemia inhibitory factor (LIF), to have a relatively unlimited capacity to renew itself through the creation of further stem cells. Thus an animal can be completely protected against the lethal effects of high doses of irradiation by injection of bone marrow cells which will repopulate its lymphoid and myeloid systems. The capacity for self-renewal is not absolute and declines with age in parallel with a shortening of the telomeres and a reduction in telomerase, the enzyme which repairs the shortening of the ends of chromosomes which would otherwise occur at everv round of cell division.

H E M A I O P O I E ISI C T E MC E t t S The bone marrow contains at least two types of stem cell, the HSC mentioned above and the mesenchlrmal stem cell (MSC) which constitute the bone marrow stroma and under appropriate signals can differentiate into adipocytes, osteocytes,chondrocytes and myocytes. The -,gca-7+ HSC in the mouse is CD34to-/ ,Thy-1*/lo-, CD38*, c-kit (CD117)*and lin-, whereas the surface phenotype of the equivalent cell in human is CD34+,CD59+, Thy-1*, 6p33low/-,6-(l;/low and lin-. Impressively, less than 100 HSCs can prevent death in a lethally irradiated animal. The hematopoietic stem cells differentiate within the microenvironment of the sessile stromal cells which

produce various growth factors including IL-3, -4, -6 and -7, G-CSF,GM-CSF, stem cell factor (SCF),flt-3 (flk2 ligand), erythropoietin (EPO), thrombopoietin (TPO) and so on. SCFremains associatedwith the extracellular matrix and acts on primitive stem cells through the tyrosine kinase membrane receptor c-kit. The importance of this interactionbetween undifferentiated stem cells and the microenvironment which guides their differentiation is clearly shown by studies on mice hom*ozygous for mutations at the zuor the sl loci which, amongst other defects, have severe macrocytic anemia. sl/sl mutants have normal stem cells but defective stromal production of SCF which can be corrected by transplantation of a normal spleen fragment; zu/ w rnutantrnyeloid progenitors lack the c-kit surface receptor for SCF, and so can be restored by injection of normal bone marrow cells (figure 11.2). Hematopoiesis needs to be kept under tight control, for example by transforming growth factor B GGFB) which exerts a cytostatic effect on HSCs via induction of the cyclin-dependent kinase inhibitor pS7KIP2. Mice with severe combined lmmunodeficiency (SCID) provide a happy environment for fragments of human fetal liver and thymus which, if implanted contiguously, will produce formed elements of the blood for 6-12months.

SEENVIRONMENI T H EI H Y M U SP R O V I D EI H t t D I F F E R E N I IATION F O RT . C E The thymus is organized into a series of lobules based upon meshworks of epithelial cells derived embryologically from an outpushing of the gut endoderm of the third pharyngeal pouch and which form well-defined cortical and medullary zones (figure 11.3).This framework of epithelial cells provides the microenvironment

MULTIPOTENTIAL STEM CELL

COMMON PROGENITORS

DIFFERENTIATED PROGENY

SPECIALIZED PROGENITORS

[-r

r(Eilt/uLUUI r c

ERYTHROCYIE

l--

PLATELET

FEGAMRyocyTEI-

FRoMoNocYfl--

lL-I tL-3 tL-6

SELF RENEWAL

tL-3 G[/-CSF

tGRANULgcruli9-

tL-3 tL-6

tL-2 tL-3 tL-7 tL-tI

tL-2 tL-4 tL-5 tL-6

tL-I tL-l TNF

tL-]

IL-4IL-9GM-CSF

tL-7

tl-s tL-21

Figure 11.1. The multipotential hematopoietic stem cell and its progeny which differentiate under the influence of a series of soluble growth factors within the microenvironment of the bone marrow The expression ofvarious nuclear transcription factors directs the differentiation process. For example, the Ikaros gene encodes a zinc-fingered transcription factor critical for driving the development of a cornmon myeloid /lymphoid precursor into a lymphoid-restricted progenitor giving rise to T-, B- and NK cells SCF, stem cell factor; LIF, leukemia inhibitory factor; IL-3, interleukin-3, often termed the multi-CSF

because it stimuiates progenitors of platelets, erythrocytes, all the types of myeloid cells, and also the progenitors of B-, but not T-, cells; GM-CSF, granulocyte-macrophage colony-stimulating factor, socalledbecause itpromotes the formation of mixedcolonies of these two cell types from bone marrow progenitors either in tissue culture or on transfer to an irradiated recipient where they appear in the spleen; G-CSF, granulocyte colony-stimulating factor; M-CSF, monocyte colony-stimulating factor; EPO, erythropoietin; TPO, thrombopoietin; TNF, tumor necrosis factor; TGFB, transforming growth factor B.

for T-cell differentiation. In both neonatal and adult mice c-kit+ CD44+ T-cell progenitors arrive from the bone marrow in waves of immigration that appear to be regulated by the accessibility of putative niches in the thymus. There are subtle interactions between the extracellular matrix proteins and a variety of adhesion/ homing molecules which, in addition toCD44, include the au integrin. Several chemokines also play an essential role, with CXCL12 (stromal-derived factor-1,SDF-1) being a particularly potent chemoattractant for the CXCR4* progenitor cells in man. Inaddition, the epithelial cells produce a series of peptide hormones which mostly seem capable of promoting the appearance of Tcell differentiation markers and a variety of T-cell functions on culture with bone marrow cells lzl aitro. T]lre circulating levels of these hormones ir? aiuo begin to decline from puberty onwards, reaching vanishingly

small amounts by the age of 60 years.Severalhave been well characterizedand sequenced, including thymulin, thymosin u' fhymic humoral factor (THF) and thymopoietin (and its active PentapePtide thymopontin, TP-s). Of these, only thymulin is of exclusively thymic origin. This zinc-dependent nonapeptide tends to normalize the balance of immune resPonses:it restores antibody avidity and antibodyproduction in aged mice and yet stimulates suppressor activity in animals with autoimmune hemolytic anemia induced by crossreactive rat red cells (cf. p. a2$. Thymulin may be looked upon as a true hormone, secreted by the thymus in a regulated fashion and acting at a distance from the thymus as a fine physiological immunoregulator contributing to the maintenance of T-cell subset homeostasis. Specialized large epithelial cells in the outer cortex/ 'nurse' cells, are associated with large known as

C H A P T E RI I - O N T O G E N Y A N D P H Y I O G E N Y

MYELOID STEMCELLS

Reslorofion in mulontmiceby normolgrofts of hemqlopoiesis

MtcR0-

SPLEENFRAGMENT

ffie

NORMAL DONOR GMFT

ANEMIC MUTANT RECIPIENT HEMATOPOIESIS

*, ! *r*

)i'roeL

DEFECTIVE a

w/w

iie&

tl/sl

Figure 11.2. Hematopoiesis requires normal bone marrow stem cells differentiating in a normal microenvironment, The zu locus codes for c-kit, a stem cell tyrosine kinase membrane receptor for the stem cell factor (SCF) encoded by the sl 1ocus.Mice which are hom*ozygous for mutant alleles at these loci develop severe macrocytic anemia

NORMAL

which canbe corrected by transplantation ofappropriate normal cells. The experiments show that the w f u m:utant lacks normal stem cells and the sl/sl mutant lacks the environmental factor needed for their development.

Figure 11.3. Cellular features of a thyrnus lobule. Seetext for description. (Adapted from Hood L E, Weissmanl L, Wood W.B. & Wilson J.H. (1984)Immunology, 2nd edn, p. 261 Benjamin Cummings, California )

numbers of lymphocytes which lie within pockets produced by the long membrane extensions of these epithelial cells. The epithelial cells of the deep cortex have branched processes, rich in class II MHC, and connect through desmosome cell junctions to form a network through which cortical lymphocytes must pass on their way to the medulla (figure 11.3).The cortical lymphocytes are densely packed compared with those in the medulla, many are in division and large numbers of them are undergoing apoptosis. On their way to the

'sentinel' medulla, the lymphocytes pass a cordon of macrophages at the corticomedullary junction' A number of bone marrow-derived dendritic cells are present in the medulla and the epithelial cells have broader processes than their cortical counterparts and express high levels of both class I and class II MHC. \A/horled keratinized epithelial cells in the medulla form the highly characteristic Hassall's corpuscles beloved of histopathology examiners. These structures serve as a disposal system for dying thymocytes, perhaps follow-

CHAPTER I I _ O N T O G E N YA N D P H Y T O G E N Y

Ludwig Gross had found that a form of mouse leukemia could be induced in low-leukemia strains by inoculating filtered leukemic tissue fromhigh-leukemia strainsprovided that this was done in the immediate neonatal period. Sincethe thymus was known to be involved in the leukemic process, Jacques Miller decided to test the hypothesis that the Gross virus

results in 1961 (Miller I.F.A.P., Lancet ii, 748479), MiIler 'during embryogenesis the thymus would opined that produce the originators of immunologically competent cells, manyof whichwouldhavemigrated to othersitesataboutthe time of birth' All in all a superb example of the scientific method and its applicationby a top-flight scientist.

could only multiply in the neonatal thymus by infecting neonatally thymectomized mice of lowleukemia strains. The results were consistent with this hypothesis but, strangely, animals of one strain died of a wasting diseasewhich Miller deduced could have been due to susceptibility to infection, since fewer mice died when they were moved from the converted horse stables which served as an animal house to 'cleaner' quarters. Autopsy showed the animals to have atrophied lymphoid tissue and low blood lymphocyte levels, and Miller therefore decided to test their immunocompetence before the onset of wasting disease.To his astonishment, skin grafts, even from rats (figure M11.1.1)as well as frorn other mouse strains, were fully accepted. These phenomena were not induced by thymectomy later in life and, in writing up his preliminary

ing their phagocytosis by dendritic cells, and are the only location where apoptotic cells are found in the medulla. A fairly complex relationship with the nervous system awaits discovery; the thymus is richly innervated with both adrenergic and cholinergic fibers. Both thymocytes and thymic stromal cells expressreceptors for a number of neurotransmitters and neuropeptides. Somatostatin is expressed by both cortical and medullary thymic epithelial cells and is able to induce the migration of thymocytes which express the SSTR2 receptor for this neuropeptide. Theglial cell line-derived neurotrophic/actor (GDNF) and the GFRo(1component of the GDNF receptor are expressed in CD4-CDS('double negalive' , DN) thymocytes. Furthermore, fetal thymocyte precursors fail to grow in a serum-free medium culture system unless recombinant GDNF is added. Thus, GDNF may be involved in the survival of the DN immature thymocytes. Glucocorticoid hormones, such as hydrocortisone (cortisol) and cortisone, are classicallyassociatedwith the adrenal gland but are also produced by thymic epithelial cells. During stress the output of these hormones increasesand directly provokes thymic involution and apoptosis of thymocytes. The sex hormones testosterone,estrogen and progesterone can also promote thymic involution. In the human, thymic involution naturally commences within the first 12 months of life, reducing by

Figure M11.1.1. Acceptance of a rat skin graftby a mouse which had been neonatallv thvmectomized

around 3o/oa year to middle age and by 7% thereafter. The size of the organ gives no clue to these changes because there is replacement by adipose tissue. In a sense,the thymus is progressively disposable because, aswe shall see,it establishesa long-lasting peripheral Tcell pool which enables the host to withstand loss of the gland without catastrophic failure of immunological function-witness the minimal effects of thymectomy in the adult compared with the dramatic influence in the neonate (Milestone 11.1).Nevertheless, the adult thymus retains a residue of corticomedullary tissue containing a normal range of thymocyte subsets with a broad spectrum of TCR gene rearrangements. Adult patients receiving either T-cell-depleted bone marrow or peripheral blood hematopoietic stem cells following ablative therapy are able to generate new naive T-cells at a rate that is inversely related to the age of the individual. Theseobservationsestablishthat new T-cellscan be generated in adult life, either in the thymus or in the still 'extrathymic' sites that have been rather mysterious proposed as additional locations for T-cell differentiation and which might include the intestinal cryptopatchesof the gut-associatedlymphoid tissues. l-cells Bonemorrowstemcellsbecomeimmunocompelent in thelhymus The evidence for this comes from experiments on the

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y reconstitution of irradiated hosts.An irradiated animal is restored by bone marrow grafts through the immediate restitution of granulocyte precursors; in the longer term, also through reconstitution of the T- and B-cells destroyed by irradiation. However, if the animal is thymectomized before irradiation, bone marrow cells will not reconstitute the T-lymphocyte population (cf.figwre6.42). By day 77-72in the mouse embryo, lymphoblastoid stem cells from the bone marrow begin to colonize the periphery of the epithelial thymus rudiment. If the thymus is removed at this stage and incubated in organ culture, a whole variety of mature T-lymphocytes will be generated. This is not seen if 10-day thymuses are cultured and shows that the lymphoblastoid colonizers give rise to the immunocompetent small lymphocyte ProSeny.

T . C E t tO N I O G E N Y Differenli0li0n isoccomponied bychonges insurfoce mofters T-cell progenitors arriving from the bone marrow enter the thymus through venules at the cortico-medullary junction. Theseearly thymocytes lackboth the CD4 and CD8 coreceptorsand are therefore referred to as double negative (DN) cells. They do, however, express the chemokine receptor CCRT and, under the influence of chemokines such as CCL19 and CCL21, they migrate through the thymic cortex towards the outer subcapsular zone (figure 11.4).The earliest of these progenitors, the DN1 cells, retain pluripotentiality, still express the stem cell marker CD34, and also express high levels of the adhesion molecule CD44 and the stem cell factor (SCF) receptor (c-kit, CD117) (p.229). As they mature into DN2 cells they lose CD34 expression and begin to expressthe IL-2 receptor crchain (CD25).Thesecells are restricted to producing T-cells and lymphoid dendritic cells. T-cell development is severely impaired in Notch-l / knockout mice, Notch-1 signaling being necessary for T-cell lineage commitment of the DN1 and DN2 cells. Indeed, the Notch-l ligands with the rather exotic names of ]agged-L, Jagged-2 and 6like-1 are expressed on thymic epithelial cells in a highly regulated way. Further differentiation into DN3 cells and a downregulation of CCRT expressionaccompaniestheir arrival in the subcapsular zone. Transient expressionof the recombinase-activating genes RAG-1 and RAG-2, together with an increase in chromatin accessibility around one allele of the TCR B-chain genes, results in pchain rearrangement and commitment to the T-cell lineage. Expression of CD3, the invariant signal trans-

ducing complex of the TCR, occurs at this stage, whilst CD44 and c-kit are lost. The later loss of CD25 signifies passage into the DN4 population, which subsequently differentiate into the CD4+ (the marker for MHC class II recognition) and CD8* (class I recognition) double positive (DP) thymocytes. TCR cr-chain gene rearrangement occurs when RAG-1 and RAG-2 are again transiently expressed in either the DN4 cells or immediately following expression of CD4 and CD8. The DP thymocytes then begin to re-express high amounts of CCRT causing them to migrate back through the cortex, eventually crossing the cortico-medullary junction into the medulla. The CD4 and CDS markers segregatein parallel with differentiation into separate immunocompetent populations of single positive (SP) CD4. (mostly Thelpers) and CD8- (mostly cytotoxic) T-cell Precursors. In addition to the factors mentioned above, thymocyte development is critically dependent on IL-7 which is necessary for the transition to the DN3 stage. This cytokine is produced by the thymic epithelial cells and is thought to be retained locally by binding to glycoaminoglycans in the extracellular matrix. Signaling through the IL-7 receptor and c-kit also help drive the early extensive proliferation that occurs in thymocytes prior to the rearrangement of the TCR genes, with thymic stromal lymphopoietin (TSLP) acting as an additional ligand for the IL-7 receptor cr-chain. SCF, IL-7 and TSLP are aided and abetted in this task by activation through the Wnt/F-catenin/nuclear T-cell factor (TCF) pathway, which also causes the upregulation of adhesion molecules required for the ordered migration of the chemokine-responsive thymocytes through the thymus. Stimulation of this pathway occurs through binding of Wnt family members to cell surfaceFrizzled (Fz) receptors on the thymocytes, Wnt-4, -7a, -7b, -10a and -10b being expressed by thymic epithelium.The yD cells remain double negative, i.e. CD4-8-, except for a small subsetwhich expressCDS. The factors which determine whether the doublepositive cells become single-positive CD4* class IIrestricted cells or CD8+ class I-restricted cells in the thymus are still not fully established. TWo major scenarios have been put forward. The stochastic/selection hypothesis suggests that expression of either the CD4 or CD8 coreceptor is randomly switched off and then cells that have a TCR-coreceptor combination capable of recognizing an appropriate peptide-MHC are selected for survival. By contrast, the instructive hypothesis declares that interaction of the TCR with MHC-peptide results in signals which instruct the T-cells to switch off 'useless' coreceptor incapable of recexpression of the ognizing that particular classof MHC. In order to reconcile the fact that there is supporting data for both

SUB. CAPSUI,AR ZONE

THYMIC CORTEX

MHC resfuiclion ol cB T-cells through

FPosrTrvE I I SELECT0N ieexponsion of self-MHC reocling cells c0RTrc9MEDUtLA[{Y JUNCTION

THYMIC MEDULI-A

Self loleronce of ap T-cells through

ieeliminotion of ontiself cells reocling with introlhymic Ag + MHC Immunocompelenl l-cells Figure 11.4. Differentiation of T-cells within the thymus. The transition between the different populations of CD4-CD8- (double negative, DN) T-cell precursors in the thymus is marked by the differential expression of CD44, CD25, c-kit and CD24. DN3 cells are unable to progress to the DN4 stage unless they successfully rearrange one of their two TCR p-chain gene loci. Successful rearrangement of the TCR cr-chain to form the mature receptor is obligatory for differentiation beyond the early CD4+CD8+ (double positive, DP) stage. The double positive T-cells which bear an ap TCR are subjected to positive and negative selection events, the self-reactive negatively selected cells are indicated in gray. Autoreactive cells with specificity for self-antigens

not expressed in the thy"rnus may be tolerized by extrathymic peripheral contact with antigen. y6 T-cells, which develop from the DN2 or DN3 T-cell precursors, mainly appear to recognize antigen directly, in a manner analogous to the antibody molecule on B-cells, although some may be restricted by nonclassical MHC-like molecules. Details and location of positive and negative selection events for y6 T-cells are not well characterized. NKT cells, which are not shown in the diagram for the sake of clarity, arise from the double positive T-cells to become CD4+CD8-, CD4-CD8+ or CD4-CD8-cellsbearing aninvariantaB TCR and the NKl.l marker (cf. p. 105).They are usually restrictedby CD1d.

hypotheses, a signal strength hypothesis has been put forward in which the strength of the p55l"k signal (seep. 174)correlateswithlineage choice and is determinedby the duration of TCR engagement during the double positive stage of thymocyte differentiation. A stronger cumulative signal is thought to favor the generation of CD4 cells (figure 11.5).Because this process is likely to generate some cells with an incorrect TCR-coreceptor combination, the cells instructed in this way subsequently undergo a selection process in which they are checked for the correct lineage choice so that cells

wrongly instructedto becomeCD4 or CD8 SPareeliminated. NKT cells constitute a distinct subsetof aBT-cells Thesecells,which appear in the thymus rather late in ontogeny,expressmarkersassociatedwith both T-cells and NKcells, suchasa TCRand the C-lectin-typereceptor NK1.1,respectively.The TCR of NKT cellsis mostly composedof an invariant TCR o, chain (Vo14Ja18in mice,Ya24la18in human) togetherwith a VB8(mouse) or Vp11 (human) p chain. They recognizeglycolipids

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y

235I

Thymic epilheliol cetl Figure 11.5. CD4versus CDS lineage commitment. In this model, the duratron of signaling through the TCR and coreceptors on the double positive (CD4+8+)thymocyte d e t e r m i n e st h e i n t e n s i t yo f t h e p 5 b l ' L s i g n a l , which in turn instructs the thymocyte to become either a CD4 T-cell if there are sustained p56l'k signals or a CD8 T-cell if the signals are less intense

such as the endogenous lysosomal isoglobotrihexosylceramide (icb3) and the microbial antigen phosphatidylinositolmannoside from mycobacteria. These antigens are presented to the invariant TCR by the nonpolymorphic MHC classIlike CDld molecule, and when activated the NKT cells secretelarge amounts of cytokines including IFN1 and IL-4, i.e. both Th1 and Th2-type cytokines. It has been proposed that NKT cells may function primarily as regulatory cells.

Receplor reorrongemenl Rearrangementof V-,D- and/-region genesare required to generatethe TCR (seep. 66).By day 15 of fetal development cells with the yD TCR can be detected in the mouse thymus followed soon by the appearance of a 'pre-TCR' version of the sF TCR. Notch-1 signals (again!) appear to play a role in crBversus y6 lineage commitment, although the details are stillbeingworked out. The deuelopment of aBreceptors The TCR B-chain gene is usually rearranged at the DN3 stageand associateswith an invariant pre-crchain, pTc, to form a 'pre-TCR', functional rearrangement of the B-chain being required for transition to the DN4 stage (figure 11.4).Signaling through the pre-TCR occurs in a ligand-independent manner whereby the pre-TCR molecules constitutively target to lipid rafts. Activation of signaling cascadesinvolving Ras/MAPK and phospholipase Cy1 pathways recruit Ets-1 and other transcription factors, stimulating the proliferation and differentiation of DN3 cells into DN4 and subsequentlv

into DP cells, as well as mediating feedback inhibition on further TCR VB gene rearrangement. Subsequent development of pre-T-cells requires rearrangement of thLeVu gene segments so allowing formation of the mature crBTCR. Rearrangement of the V B genes on the sister chromatid is suppressed following the expression of the preTCR (remember each cell contains two chromosomes for each u and B cluster). Thus each cell only exPressesa single TCR B chain and the process by which the hom*ologous genes on the sister chromatid are suppressed is called allelic exclusion. This exclusion is at least partially due to the methylation of histones maintaining a closed chromatin structure which Prevents access of the recombinases to the TCR gene segments on the excluded allele. The a chains are not always allelically excluded, so that many immature T-cells in the thymus have two antigen-specific receptors, each with their own o chain but sharing a common B chain. However, exPression of one of the crchains is usually lost during T-cell maturation, leaving the cell with a single specificity aB TCR. It has, however, been proposed that some dual-TCR Tcells do enter the periphery and thereby extend the TCR repertoire to include specificities for foreign antigens thatwould not otherwise be selected in the thymus. The deoelopment of ySreceptors Unlike the cB TCR, theybTCRinmanycases seemstobe able to bind directly to antigen without the necessity for antigen presentation by MHC or MHClike molecules, i.e. it recognizes antigen directly in a manner similar lpreto antibody. The y6 lineage does not produce a

I zso

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y

receptor'and mice expressingrearranged yand 6 transgenesdo not rearrangeany further y or 6 gene segments, indicating allelic exclusion of sister chromatid genes. y6 T-cells in the mouse, unlike the human, predominate in association with epithelial cells. A curious feature of the cells leaving the fetal thymus is the restriction in V gene utilization. Virtually all of the first wave of fetal y6 cells express VyS and colonize the skin; the second wave predominantly utilize Vy6 and seed the uterus in the female.In adult life, there is far more receptor diversity due to a high degree of junctional variation (cf. p.69), although the intraepithelial cells in the intestine preferentially use Vy4 and those in encapsulated lymphoid tissue tend to expressYy4,Yy1.7 and Vy2. It should be noted that, just to confuse everyone, other nomenclatures are also currently in circulation regarding the numbering of the individual murine Vy genes. The Vy set in the skin readily proliferates and secretes IL-2 on exposure to heat-shockedkeratinocytes,implying a role in the surveillance of trauma signals. The yD T-cellsin peripheral lymphoid tissuerespond well to the tuberculosis antigen PPD ('purified protein derivative') and to conserved epitopes from mycobacterial and selfheat-shock protein hsp65. However, evidence from y6 TCRknockoutmice suggeststhatoverall, in the adult,y6 T-cells may make a minor contribution to pathogenspecific protection. It has therefore been proposed that their primary role may be in the regulation of uB T-cells, with most y6 T-cells biased towards a Th1 cytokine secretionpattern. TWo major y6 subsets predominate in the human, V\9,V62 and Vy1,V62. The V19 set rises from 25% of the total y6 cells in cord blood to around 70% in adult blood; at the same time, the proportion of Vy1 falls from 50% to less than 30%. The majority of the Vp set have the acti-

Thymectomize , xf mice

Groflwith lrrodiole ond Tnymus reconslilule wilh of hoplotype , x I(bonemorrow

vated memory phenotype CD4SRO,probably as a result of stimulationby common ligands for the V19,V62TCR, such as non-proteinacious phosphate-bearing antigenic components of mycobacteria, Plasmodiumfalciparum and the superantigen staphylococcal enterotoxin A. Gellsqre positivelyselectedf0r self-MHCreslriclion in lhe lhymus The ability of T-cells to recognize antigenic peptides in association with self-MHC is developed in the thvmus. If an (H-2k x H-2b) F1 animal is sensitized to an antigen, the primed T-cells can recognize that antigen on presenting cellsof eitherH-2kor H-2bhap\otype,i.e.theycan use either parental haplotype as a recognition restriction element. However, if bone marrow cells from the (H2k x H-2b) FL are used to reconstitute an irradiated F1 which had earlier been thymectomized and given an H2kthymus, the subsequently primed T-cells can only recognize antigens in the context of H-2k, not of H-2b (figure 11.6). Thus it is the phenotype of the thymus that imprints H-2 restriction on the differentiating T-cells. It will also be seenin figure 11.6that incubation of the thymus graft with deoxyguanosine, which destroys the cells of macrophage and dendritic cell lineage, has no effect on imprinting, suggesting that this function is carried out by epithelial cells. Confirmation of this comes from a study showing that lethally irradiated H2k mice, reconstituted with (b x k) F1 bone marrow and then injected intrathymically with an H-2b thymic epithelial cell line, developed T-cells restricted by the b haplotype. The epithelial cells are rich in surface MHC molecules and the current view is that double-positive (CD4+8+)T-cellsbearing receptorswhich recognize selfMHC on the epithelial cells are positively selected for

Proliferolive response of primedT-cellsto KLH Prime with onontigen-presenting KLH Figure 11.6. Imprinting of H-2 T-helper cellsot hoplolype H-2x

t xk

dGuotreoted ,

J.1

dGuolreoledk

restriction by the haplotype of the thymus. Host mice were F1 crossesbetween strains of haplotypeH-2b andH-2k Theywere thymectomized and grafted with 14-day fetal thymuses, irradiated and reconstituted with F1 bone marrow. After priming with the antrgen keyhole limpet hemocyanin (KLH), the proliferative response of lymph node Tcells to KLH on antigen-presenting cells of each parental haplotype was assessed In some experiments, the thymus lobes were cultured in deoxyguanosine (dGuo), which destroys intrathymic cells of macrophage/ dendritic cell lineage, but this had no effect on positive selection (From Lo D. & Sprentf (1986)N at ur e 379,672-675.)

C H A P I E RI I - O N T O G E N YA N D P H Y L O G E N Y

differentiation to CD4+8 or CD4-8+ single-positive cells.The evidencefor this comeslargely from studies in transgenic mice. Since this is a very active area, we would like to cite some experimental examples;nonprofessionalsmay need to hang on to their haplotypes, put on their ice-packsand concentrate. One highly sophisticatedstudy startswith a cytotoxic T-cell clone raised in H-2b females against male cells of the same strain. The clone recognizesthe male antigen, H-Y and this is seenin associationwith the H-2Db selfMHC molecules, i.e. it reactswith the H-2b/Y complex. The o and B chains for the T-cell receptor of this clone are now introduced as transgenes into SCID mice which lack the ability to rearrange their own germ-line variable region receptor genes; thus the only TCR which could possiblybe expressedis that encodedbythe transgenes,provided of coursethat we are looking at females rather than males, in whom the clone would be eliminated by self-reactivity.If the transgenic SCID females bear the original H-2b haplotype (e.g. F1 hybrids between b x dhaplotypes), then the anti-H-2b/Y receptor is amply expressedon CDS+cytotoxic precursor cells (table 11.1a),whereas H-2d transgenics lacking H-2b produce only double CD4+8+thymocytes with no single CD4*8- or CD4-8* cells. Thus, as CD4+8+cells express their TCR transgene, they only differentiate into CD8* immunocompetent cells if they come into contact with

2s7 |

thymic epithelial cells of the MHC haplotype recognized by their receptor. We say that such self-recognizing thymocytes are being positively selected. Positive intracellular events accompany the positive selection process since the protein tyrosine kinases fyn and lck are activated in double-positive CD4+8+thymocytes maturing to single-positive CD8+cellsin the b haplotypebackground, but are low in cells which fail to differentiate into mature cells in the nonselectived haplotype. In another example, genes encoding an o(p recePtor from a T-helper clone (2B4), which responds to moth cytochrome c in association with the class II molecule H-2Eck,Bb(remember H-2E has an c, and B chain), are transfected into H-2k and H-2b mice. For irrelevant reasons, H-2k mice express the H-2E molecule on the surface of their antigen-presenting cells, but H-2b do not. In the event, the frequency of circulating CD4+ Tcellsbearing the 284 receptor was 10 times greater in the H-2k relative to H-2b strains, again speaking for positive selection of double-positive thymocytes which recognize their own thymic MHC. In a further twist to the story, positive selectiononly occurred in mice manipuIated to expressH-2E on their cortical rather than their medullary epithelial cells, showing that this differentiation step is effected before the developing thymocytes reach the medulla. ('Read it again Sam' as Humphrey Bogart might have said!)

I-CEtT TOTERANCE Table 11.1. Positive and negative selection in SCID transgenic mice bearing the op receptorsof an H-2Db T-cellclone cytotoxic for the male antigen H-! i e the clone is of H-2' haplotype and is female anti-male (a) The only T-cells are thosebearing the already rearranged transgenic TCR, since SCID mice cannot rearrangetheir own Vgenes. The clones are only expanded beyond the CD4+8+stagewhen positively selected by contact with the MHC haplotype {H-2r) recognized by the original clone from which the transgenewas derived Also, since the TCR recognized class I, only CDS* cells were selected (b) When the anti-male transgenic clone is expressed on intrathymic T-cells in a male environment, the strong engagement of the TCR with male antigen-bearing cells eliminates them (Based on data from von Boehmer H et al (1989) In Melchers F ef a/ (eds) Progressin Immunology7, p. 297 SpringerVerlag,Berlin.)

,/ Posiiiveselecfion Phenotype o{thymocytes

Hoplotype of tronsgenic femoles H_2D/0

CD4.8- TCRC D 4 + 8 +T C R t

cD4-8' TCRr*

++ +

H_2 a/0

++ +

Negoliueselecfion Tronsgenic H-2bmice Moles

+++

Femoles +

+++ I

cD4*8- TCR** +, crudemeosure of therelolive numbers in lhethymushovingthe of T-cells nh6n^h/^a

inrli.nlod

l0 lolelonceis necesso]y Theinduclionof immunologicol ovoidself-reocfivity In essence,lymphocytes use receptors that recognize foreign antigensthrough complementarity of shape(see p. 86).To a large extent the building blocks used to form microbial and host molecules are the same, and so it is the assembled shapes of self and nonself molecules which must be discriminated by the immune system if potentially disastrous autoreactivity is to be avoided. The restriction of each lymphocyte to a single specificity makes the job of establishing self-tolerance that much easier, simply because it just requires a mechanism which functionally deletes self-reacting cells and leaves the remainder of the repertoire unscathed. The most radical difference between self and nonself molecules lies in the fact that, in early life, the developing lymphocytes are surrounded by self and normally only meet nonself antigens at a later stage and then within the context of the adjuventicity and cytokine releaseusually associatedwith infection. Evolution has exploited these differencesto establishthe mechanisms of immunological tolerance to host constituents (Milestone 11.2).

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y

Over 60 years ago, Owen made the intriguing observatron that nonidentical (dizygotic) twin cattle, which shared the same placental circulation and whose circulations were thereby linked, grew up with appreciable numbers of red cells from the other twin in their blood; if they had not shared the same circulation at birth, red cells from the twin iniected in adult life would have been rapidly eliminated by an immunoIogical response From this finding, Burnet and Fenner conceived the notion that potential antigens which reach the lymphoid cells during their developing immunologically immature phase can in some way specifically suppress any future response to that antigen when the animal reaches immunological maturity. This, they considered, would provide a means whereby unresponsiveness to the body's own constituents could be established and thereby enable the lymphoid cells to make the important distinction between 'self ' 'nonself '. and On this basis, any foreign cells introduced into the body during immunological development should trick the animal into treating them as 'self'-components in later life, and the studies of Medawar and his colleagueshave shown that immunological tolerance, or unresponsiveness, can be artificially induced in this way. Thus neonatal injection of CBAmouse cells into newbornAstrain animals suppresses their ability to reject a CBAgraftimmunologically in adultlife (figure M11.2.1). Tolerance can also be induced with soluble antigens; for example, rabbits injected with bovine serum albumin without adjuvant at birth fail to make antibodies on later challenge with this protein. Persistenceof antigen is required to maintain tolerance. In

removal of the thymus, it is of interest to note that the tolerant statepersists for much longer in thymectomized animals. The vital importance of the experiments by Medawar and his team was their demonstration that a state of immunological tolerance can result from exposure to an antigen. As willbe discussed in the text, there is a window of susceptibility to clonal deletion of self-reacting T-lymphocytes at an immature phase in their ontogenic development within the thymus (and in the caseof B-cells within the bone marrow). The concept of a neonatal window for tolerance induction is more aPparent than real and stems from the relatively lownumber of peripheralized immunocompetent T-cells, which do not differ in behavior from resting T-cells in the adult in their tolerizability or capacity for an immune resPonse.Note that resting T-cells are generally more readily tolerizable than memory cells.

Medawar's experiments, the tolerant state was long lived becausethe injected CBA cells survived and the animals continued to be chimeric (i.e. they possessedboth A and CBA cells). With nonliving antigens, such as soluble bovine serum albumirU tolerance is gradually los! the most likely explanation is that, in the absence of antigen, newly recruited immunocompetent cells which are being generated thloughout life are not being rendered tolerant. Since recruifment of newly competent T-lymphocytes is drastically curtailed by

Self-toleronce conbe inducedin fhethymus Since developing T-cells are to be found in the thymus, one mightexpectthis tobe the milieu inwhich exposure to self-antigens on the surrounding cells would induce tolerance. The expectation is reasonable.If stem cells in bone marrow of H-2k haplotype are cultured with fetal thymus of H-2d origin, the maturing cells become tolerant toH-2d, as shown by their inability to give a mixed lymphocyte proliferative response when cultured with stimulators of H-2d phenotype; third-

Figure M11.2.1. Induction of tolerance to foreign CBA skin graft in A strain mice by neonatal injection of antigen. The effect is antigen specific since the tolerant mice can reiect third-party grafts normally. (After Billingham R, Brent L. & Medawar P.B. (1953) NnttLreTT2.603406.)

party responsiveness is not affected. Further experiments with deoxyguanosine-treated thymuses showed that the cells responsible for tolerance induction were deoxyguanosine-sensitive, bone marrowderived macrophages or dendritic cells which are abundant at the corticomedullary junction (table 11.2). Introthymic clon0ldelelionleodsto self-loleronce There seems little doubt that strongly self-reactive Tcells canbe physically deleted within the thymus.If we

look at the experiment in table 11,.1b,we can see that SCID males bearing the rearranged transgenes coding for the c,p receptor reacting with the male H-Y antigen do not possess any immunocompetent thymic cells expressing this receptor, whereas the females which lack H-Y do. Thus, when the developing T-cells react with self-antigen in the thymus, they are deleted. In other words, self-reactive cells undergo a negative

Thble 11.2. Induction of tolerance in bone marrow stem cells by incubation with deoxyguanosine (dGuo)-sensitive macrophages or dendritic cells in the thymus. Clearly, the bone marrow cells induce tolerance to their own haplotype. Thus the thymic tolerance-inducing cells can be replaced by progenitors in the bone marrow inoculum fenkinson E.J.,]hittay P, Kingston R. & Owen J.]. (1985) kansplantation 39,337) or by adult dendritic cells from spleen, showing that it is the stage of differentiation of the imrnature T-cell rather than any special nature of the thymic antigen-presenting cell which leads to tolerance (Matzinger P.& Guerder S. (1989)Nature 338,74)

Figure 11.7. AIRE directs the ectopic expression of organ-specific self antigens in the thymus. Double transgenic mice were generated by crossing transgenic mice expressing hen egg lysozyme (HEL) under the control of the rat insulin promoter (RIP) with mice expressing the transgenic 3A9 crB TCR specificfor the amino acid 46-61 peptide from HEL presented by the I-Ak MHC class II molecule. These mice normally tolerize the transgenic T-cells in the th)'rnus (a), but this did not occur if the mice were backcrossed to mice in which theAIRE gene had been knocked out (b). The incidence of type I diabetes was dramatically increased in the absenceofAIRE expression. (Based on data from Liston A. et aI. (2004)lournal of ExperimentalMedicine 20O, 1015-1,026\

selection process in the thymus, a process which constitutes central tolerance of the T-cells. Expression of the AIRE (autoimmune regulator) gene in medullary thymic epithelial cells acts as a master switch directing the transcriptional activation of the genes for a number of organ-specific self antigens. The ectopic exPression of these antigens provokes the elimination of the corresponding self-reactive thymocytes. Confirmation of the importance of AIRE expression for such clonal deletion has come from experiments using a double transgenic model developed by Goodnow and colleagues. In these mice a membrane-bound version of hen egg lysozyme ' (HEL) is transgenically expressed as a'neo-self antigen (becauseit is always present itbecomes essentially a self antigen), and high numbers of thymocytes specific for this antigenare also generatedbyintroductionof the relevant TCR as the other transgene. When the HEL transgene is linked to the tissue-specific ralinsulinpromoter 'self ' (RIP), expression of the antigen occurs in both the p-cells in the islets of Langerhans of the pancreas and in the thymus. In the absence of AIRE the RlP-driven expression of HEL fails to occur in thymic epithelium, but still occurs in the pancreatic islets. The developing transgenic T-cells which are normally deleted in the thymus escape deletion in the AIRE deficient mice (figure11.7).

CHAPTER I I - O N T O G E N YA N D P H Y T O G E N Y Deletion of thymocytes also occurswhen thymic cells bear certain self-components which act as superantigens (cf. p.707), in this casebecausethe antigen reacts with a whole family of VB receptorsthrough recognition of nonvariable structures on a VB segment.An example is the H-2E molecule which reacts with receptors belonging to the VB17a family; strains which cannot expressH-2E becauseof a defect in the Ea gene possess mature T-cells utilizing Yp17a, whereas strains which express H-2E normally delete their VBl7a-positive Tcells. Likewise, mice of the MIso genotype delete VB6bearing cells, the Mls being a locus encoding a B-cell superantigen which induces strong proliferation in VB6 T-cells from a strain bearing a different Mls allele (cf. p. 107).Even exogenous superantigens,such as staphylococcalenterotoxin B which activatesthe VB3 and VB8 Tcell families in the adult, can induce apoptosis in early immature thymocytes utilizing these receptor families. F s ct o r s aff ectin g p o sitia e o r n eg atia e seIecti on inthethymus T-cells that either fail to express a TCR at all, or which express a TCR of very low affinity, do not receive survival signals and die from neglect. For the remaining cells the engagement of the TCR by MHC-peptide underlies both positive and negative selection.But how can the same MHC-peptide signal have two totally different outcomes? Well, positive and negative selection may occur at low and high degrees of TCR ligation, respectively.For example, high concentrations of antibody to the TCR-associatedCD3 induce apoptosis in thymocytes (figure 11.8),whereas low concentrations do not. Furthermore, many examples have been published showing that the same peptide will induce positive selection at low concentration and negative selection at high concentration. This has led to the avidity model, which postulates that a functionally low avidity interaction between T-cell and peptide-MHC will positively select double positive CD4+8+thymocytes,while a high avidity interaction will lead to clonal deletion. Thus, engagement of the TCR with self-MHC on cortical epithelial cells leads to expansion and positive selection for clones which recognize self-MHC, perhaps with a whole range of affinities, but that engagement of the TCRwith high affinity for self-MHC (+ self-peptide) on medullary epithelial cells and dendritic cells will lead to elimination and hence negative selection. Although still not fully worked out, there are also obvious differences in the biochemical pathways used for positive and negative signaling. Positive selection is cyclosporine-sensitive and dependent on the Ras-MEK-ERK pathway (cf. p.775), whereas negative selection is cyclosporine-resistantand independent of

Figure 11.8. Electron micrograph of cells induced to undergo apoptosis in intact fetal thymus lobes after short-term exposure to anti-CD3. A and N indicate representative apoptotic and normal lymphocytes, respectively. Note the highly condensed state of the nuclei of the apoptotic lymphocytes. (Photograph kindly donated by Professor J J T Owen, from Smith et al (1989) Nature 337, 181-184 Reproduced by permission from Macmillan Journals Ltd, London )

this pathway. Different intensities of signaling from the TCR, Notch and other receptors may influence which pathway is utilized. The Ikaros transcripton factor seems to be important in regulating negative selection, and also appears to help determine whether DP thymocytesbecomeCD4 or CD8 SPcells during positive selection. Let us finish on a cautionary note: the avidity model may be substantially correct but it could be an oversimplification. For instance, certain superantigens, which can cause clonal deletion of certain Vp families, fail to expand them even at very low concentrations when the model would have indicated positive selection. This has spawned other models involving conformational changes at the TCR-peptide-MHC binding interface, and given the complex interactions of peptides behaving as agonists,partial agonists and antagonists (cf. p.177); the last word has not yet been spoken (not that it ever is in science!). conolsobeduel0 clonolonelgy T-celltoleronce We have already entertained the idea that engagement of the TCR plus a costimulatory signal from an antigenpresenting cell are both required for T-cell stimulation, but, when the costimulatory signal is lacking, the T-cell becomes tolerized or anergic, or, if you prefer, paralyzed. T-cell tolerance occurring outside the thymus is referred to as peripheral tolerance. Thus, anergy canbe induced in extrathymic T-cells by peripheral antigens ln aiao when presented by cells lacking costimulatory molecules.If a transgene construct of H-2Eb attached to an insulin promoter is introduced into a mouse which normally fails to express H-2E, the H-2Eb transgene

CHAPTER AND PHYLOGENY I I _ONTOGENY product appears on the B-cells of the pancreas and induces toleranceto itself . Whereasthe expressionof H2E onbone marrow-derived cellsin the thymic medulla deletes T-cells bearing Y$77a receptors, these cells are not lost in the toleranttransgenicmouseexpressingpancreaticH-2E, i.e. there is a stateof clonal anergy,not deletion. The altered immunological status of these cells is revealed by their inability to proliferate when their receptorsare cross-linkedby an antibody to VB17a. It is unlikely that these results are due to low level expression of antigen in the thymus. Similar experiments showed that mice expressing influenza hemagglutinin on the pancreatic B-isletsalso became tolerant irrespective of whether the transgenic thymus was replaced by a normal gland or not. Nonetheless,anergic cellscan alsobe generatedwithinthe thymic population as seenin mice transgenicfor both an anti-Kb TCR and a K'gene controlled by a truncated fragment of a keratin IV promoter which allowed expression on thymic medullary cells. Peripheral T-cell anergy can occur at different levels depending upon the circ*mstancesof antigen exposure. If the above double transgenic experiment is repeated with a full keratin IV promoter, the Kb antigen is expressedon keratinocytes and induces full tolerance, even though the same high frequency of cytotoxic T-cell precursors with the transgene TCR is seen as in single transgenic animals lacking K'. If Kb is expressedon cells of neuroectodermal origin or hepatocytes, again the double transgenicmice are tolerant but there is dramatic downregulation of TCR and CD8 molecules; in the

former but not the latter case, downregulation of TCR could be reversed by exposure to antigen in aitro. In some experimental models, tolerance can be abrogated byIL-2. To recapitulate, autoreactive T-cells leaving the thymus can be rendered anergic in the periphery and can display different degrees of potentially reversible unresponslveness. lnfectious anergy If a clone of T-helpers is subject to a limiting dilution experiment (p. 743),the minimal unit of proliferation in response to peptide on an APC is usually several cells not just one. This implies that triggering only occurs in small groups or clusters of cells and suggests that paracrine or multicellular interactions between potential responders bound to a single APC are needed to drive the cells into division (figure 71.9a).It will be appreciated that, if a newly arising extrathymic naive T-cell binds to its antigen, even on a professional APC, it will not be stimulated if its neighbors in the cluster have already been made anergic. Indeed, instead of being triggered, it will itself become anergic, so perpetuating the infectious anergic process (figure 11.9b). CD4* CD25+ Foxp3* Tieg cells, which often naturally exhibit a state of profound ar.ergy,rnaybe the critical cell here. It is thought that such cells can downregulate MHC class II and the costimulatory CD80 (B7.1) and CD86 (F7.2)molecules on the APC, generating an infectious anergy which does not require the simultaneous presenceof the Tregs and the cells whichwill themselves be made anergic (figure 11.9c).Anergic regulatory T-

T-cells Muluol helpbynormol

T-Tinleroction onergy: inodequole Infectious

oo Figure 11.9. Infectious anergy. (a) Clusters of normal T-cells (green) around newly immunocompetent cells (gray) reacting with the sameAPC mutually support activatron and proliferation (b) Newly immunocompetent cells surrounded by anergic T-cells(red) receiveno stimulatory signals from their nerghbors and are themselvesrendered anergic (c) The anergic T-cells can act as regulatory T-cel1s whereby they downregulate expression of MHC classII, CD80 (B7 1) and CD86 (87 2) molecules on the APC This effect requires cell-cell contactbetween the anergic T-cel1s and the APC, is not blocked by neutralizing antibodies to the cytokines IL-4, IL-10 or TGFp, and would exhibit linked suppression to other epitopes on the antigen presented by the APC Subsequent encounter of newly immunocompetent c e l l sw i t h t h i s A P C w i l l l e a d t o r n e r " v

241 I

of APCfunction Infectious onergy; downregulolion

C H A P T E RI I - O N T O G E N Y A N D P H Y I . O G E N Y cells might also be capable of inducing apoptosis in dendritic cells via CD95-CD95L interactions, or cause the dendritic cells to induce apoptosis in responder Tcells, again by the CD95 pathway. We shall see later in Chapter 16 that the induction of transplantation immunosuppression with a nondepleting anti-CD4 can be long-lasting because the production of anergic regulatory T-cells prevents the priming of newly immunocompetent T-lymphocytes by the transplantation antigen(s).

Lock ofcommunicolion concouse untesponsiveness It takes two to tango: if the self-molecule cannot engage the TCR, there can be no response. The anatomical isolation of molecules, like the lens protein of the eye and myelin basic protein in the brain, virtually precludes them from contact with lymphocytes, except perhaps for minute amounts of breakdown metabolic products which leak out and may be taken up by antigen-presenting cells,but at concentrationsway below that required to trigger the corresponding naive T-cell. Even when a tissue is exposed to circulating lymphocytes, the concentration of processed peptide on the cell surface may be insufficient to attract attention from a

potentially autoreactive cell in the absenceof costimulatory B7. This was demonstrated rather elegantly in animals bearing two transgenes: one for the TCR of a CD8 cytotoxic T-cell specific for LCM virus glycoprotein, and the other for the glycoprotein itself expressed on pancreatic B-cells through the insulin promoter. The result? A deafening silence: the T-cells were not deleted or tolerized, nor were the p-cells attacked. If these mice were then infected with LCM virus, the naive transgenic T-cells were presented with adequate concentrations of the processed glycoprotein within the adjuvant context of a true infection and were now stimulated. Their primedprogeny/ having an increasedavidity (cf.p.424) and thereby being able to recognize the low concentrations of processed glycoprotein on the B-cells, attacked their targets even in the absenceofBT and caused diabetes(figure 11.10).This may sound a trifle tortuous,but the principle could have important implications for the induction of autoimmunity by cross-reacting T-cell epitopes. Molecules that are specifically restricted to particular organs which do not normally express MHC class II represent another special case, since they would not have the opportunity to interact with organ-specific CD4 T-heh:er cells.

PROFESSIONAL

ANTIGENPRESENTING CELL

LcN,4 Infection

PRIMED

FORLCMg p

Nostimulolion

I Kitl I I

PROCESSED g p PEPTIDE

MHCI TRANSGENE LCMgp

PANCREATIC P CELL

PANCREATIC OCELL

Figure 11.10. Mutual unawareness of a naive cytotoxic precursor T-cell and its B7negative cellular target bearing epitopes present at low concentrations. Priming of the naive cell by a natural infection and subsequent attack by the higher avidity primed cells on the target tissue. LCM, lym phocytic choriomeningi tis virus; 8.p., glycoprotein (From Ohashi P.S.et aI (7997\ Cell6s.3O5117.\

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y Immunological silence would also result if an individual has no genes coding for lymphocyte receptors directed against particular self-determinants; analysis of the experimentally induced autoantibody response to cytochrome c suggests that only those parts of the molecule which show species variation are autoantigenic, whereas the highly conserved regions where the genes have not altered for a much longer time appear to be silent, supposedly becausethe autoreactivespecificitieshave had time to disappear.

243 |

IMMATURE MATURE LYMPHOID B-CELL STEMCELL PRO-B-CELLPRE-B-CELL B-CELL +slgD slglV

TdT RAG-I/2 c-kit

B . C E t t SD I F F E R E N T I A I NI E I H E F E T ATTI V E R A N DI H E NI N B O N EM A R R O W

cDr27 (L-7R) ontigen) CD24(heot-sloble

TheB-lymphocyte precursors, pro-B-cells, arepresent among the islands of hematopoietic cells in fetal liver by 8-9 weeks of gestation in humans and 14 days in the mouse. Production of B-cells by the liver wanes and is mostly taken over by the bone marrow for the remainder of life. Stromal reticular cells, which express adhesion molecules and secrete IL-7, extend long dendritic processesmaking intimate contact withIL-7 receptorpositive B-cell progenitors. Although early B-cellscomprise only a minor subpopulation of the cells in those cultures, it is possible to analyze the different stages in their development by in aitro rescue with the Abelson murine leukemia virus (A-MuLV), a replicationdefective retrovirus capable of transforming pre-B-cells at various points in their development into clones. A series of differentiation markers associated with B-cell maturation have now been established(figure 11.11). P0x5is0 m0i0rdetermining f0ctorin B-celldifferentiotion Development of hematopoietic cells along the B-cell lineage requires expression of E2A and of early B-cell factor (EBF);the absenceof either of theseprevents proB-cellsprogressing to the pre-B-cell stage (figure 17.\2). Also required is expression of the PaxS gene which encodesthe BSAP (B-cell-specificactivator protein) transcription factor. Thus, in Pax| / knockout mice, early pre-B-cells (containing partially rearranged immunoglobulin heavy chain genes) fail to differentiate into mature, surface Ig*, B-cells (figure 77.12). However, if the pre-B-cells from Par5 i knockout mice are provided with the appropriate cytokines in aitro, they can be driven to produce T-cells,NK cells, macrophages,dendritic cells, granulocytes and even osteoclasts!These unexpected findings clearly show that the early pre-Bcell has the potential to be diverted from its chosenpath and instead provide a source of cells for many other hematopoietic lineages.However, these pre-B-cellsare not pluripotent as, unlike bone marrow hematopoietic

MHCCLASSII

cDl0 (CALLA)

CD43(Leukos

(lg-c/lg-F) CD79o/CD79b (8220isoformof leukocyle commononligen) I cD45R

I l

cD22 cDlI cD40

CD25(lL-2Rr'___)

cD20

cD2r(cR2) CD32(FcyRlI) CD23(FceRII)

Figure 11.11. Some of the differentiation markers of developing B-cells. The time of appearance of enzymes involved in Ig gene rearrangement and diversification (blueboxes) and of surface markers defined by monoclonal antibodies (orange boxes, seetable 7 lfor list of CD members) is shown.

stem cells, they are unable to rescue lethally irradiated mice. It is clear lhat Pax\ actsas a critical master gene by directing B-cell development along the correct pathway, and does this by repressing expression of genes such as those encoding Notch-L, myeloperoxidase and monocyte/macrophage colony-stimulating factor receptor which are associated with other lineages, whilst activating B-cell specific genes including Ig-cr, CD19 and the adaptor protein BLNK.

I Figure 11.12. Par5 is required for B-cell differentiation. Hematopoietic stem cells (SC) under the influence of stem cell factor (SCF), IL-3 and the Ikaros and PU.1 transcription factors can differentiate into proB-cells Further differentiation into pre-B-cells requires the E2A transcription factor together with early B-cell factor (EBF) and IL-7 hom*ozygous E2A mutant mice lack pre-B-cells, there being a block to D rl ,rearrangement in the Ig heavy chain locus plus severe reduction in RAG-1, Ig-cr, CD19 and 1,,transcripts If atthe earlypre-B stagePaxs is not expressed, then differentiation along the B-cell lineage pathway

B . I A N DB . 2 C E t t S R E P R E S E INWTO D I S T I N CPI O P U T A T I O N S We have previously drawn attention to the subpopulation of B-cells which, in addition to surface IgM, express CD5 (cf. p. 218). The progenitors of this subset move from the fetal liver to the peritoneal cavity fairly early in life, at which stage they are the most abundant B-cell type and predominate in their contribution to the idiotype network and to the production of low affinity, multispecific IgM autoantibodies and the so'natural' called antibodies to bacterial carbohydrates, which seemingly arise slightly later in the neonatal period without obvious exposure to conventional antigens. The B-1 phenotype, viz. high surface IgM, low surface IgD, CD43+and CD23-, is shared by a minority subpopulation which is, however, CD5-; these two populations are referred to as B-1aand B-1brespectively (figure 11.13).The phenotype of conventional B-2 cells (table 11.3),low surface IgM, high surface IgD, CD5-, CD43- and CD23+,reflects the fact that they represent a separate developmental lineage (figure 11.13). Some general comments may be in order. Although B-1 cells can shift to a B-2 phenotype, and possibly vice versa, there is minimal conversion between the two lineages under normal circ*mstances. B-1 cells are particularly prevalent in the peritoneal and pleural cavities, maintain their numbers by self-replenishment and limit their

comes to an abrupt halt. These early pre-B-cells have rearranged Ig Dt to J" indicating their intention to become B-cells. However, even at this late stage, they can make other lineage choices as evidenced by the fact that, in the absenceofPax5 expression, theycan give rise to a number of other ce1l types if they are provided with appropriate cytokines. Indeed, Par5 / clones are able to develop into T-cells if transferred to immunodeficient mice, in which case they exPress rearranged TCR genes in addition to their initial Ig heavy chain gene rearrangement

de noaoproduction from progenitors by feedback regulation. They can express both CD5 and its ligandCDT2 on their surface, which should encourage mutual interaction, but a major factor influencing self-renewal could be the constitutive production of IL-10, since treatment of mice with anti-Il-10 from birth virtually wipes out the 8-L subset. The predisposition for self-renewal may underlie their undue susceptibility tobecome leukemic, with the malignant cells in chronic lymphocytic leukemia being almost invariably CD5+. B-1 cells tend to use particular germ-line y genes,and the autoantibody response to bromelain-treated erythrocytes is restricted to this subset which utilizes the rather diminutiveV n11andV ,12 families. Clonal expansion seems to be driven by reaction with self-antigens (seelegend to figure 11.13).They tend to respond to type 2 thymus-independent antigens (ct. p.178) and, unlike the B-2 population, they do not enter into liaisons with thymus-dependent antigens, do not enter germinal centers and hence do not undergo somatic mutation or form high affinity antibodies. This may be just as well if the harmless low affinity autoantibodies which are produced by many B-1 cells are not automatically driven to high affinity pathogenic autoantibodies. In a 'good' and'bad' weak moment one sometimes hears of 'good guys'Possibly having autoantibodies, with the the job of sweeping up broken down self-comPonents, as envisaged by Pierre Grabar many years ago when he thought of themas globulinestransporteurs.

C H A P I E RI I - O N T O G E N YA N D P H Y L O G E N Y

FETAL LIVER

ADULT BONEMARROW

245 |

Table 11.3. Comparison of two mouse B-cell subsets. (Developed from Herzenberg L A, StallA , Melchers F.efal.(eds) (7989)Progressin lnmunologyT,p 409 Springer-Verlag, Berlin )

t setect I selfor tP*t-"

loD+

cos-toa3* J+

--t

cD23-

lgM* loD+++

Reploced bylgMprecursors inbone m0rrow

I co5-co+scD23+

w NATUML ANTIBODY

Self-renewing produclion Conslilutive tL-10

w HIGHAFFINITY ANTIBODY

ADULT B-CELL POOL

PRODUCTION ANTIBODY Serum lgM,lgc3 lgGl lgG2o, lgG2b lgMouloontibody lgMonli-ld Ab lgMonli-bocteriol Anti-hoplen/prolein T-dependence Affinilymolurotion

+++ + + f o+ + +++ +++ +++

+ +++ + + t o+ + + ,)

Figure 11.13. The development of separateB-cell subpopulations. It is presumed that B-1 cells of su{ficiently high avidity for, say, selfsurface antigens are eliminated, leaving positively selected lower affinity specificities for soluble self-antigens and the spectrum of 'natural antibody'-producing B-1 cells By contrast,B-2 cells undergo negative, rather than positive, selection by self-antigens and the surviving B-2 cells give rise to the higher affinity IgG antibody produced by helper T-cell-dependent class-switched B-cells It is thought that, although these subsets might be able to give rise to each other under some circ*mstances, generally they are maintained as separate lineages Direct evidence that self-antigens positively selectB-1 cells is provided by mice made transgenic for the V"3609 heavy chain gene. The transgene-encoded heavy chain pairs with endogenous Vr21C light chain to produce an anti-thymocyte autoantibody associated with CD5+ B-cells and which recognizes a Thy-1-associated carbohydrate epitope High levels of the transgenic B-cells and of the serum autoantibody were found only in the presence of the autoantigen, being absent in Thy-1 knockout mice SC, hematopoietic stem cel1; CD23, FceRII;CD43. leukosialin.

gene reofiongemenls Thesequence ofimmunoglobulin

Other functions of B-1 cells may be the generation of an idiotype network concerned in self-tolerance, the responseto conservedmicrobial antigens, and possibly the idiotypic regulation of B-2 responses.They are certainly the sourceof 'natural antibodies' which provide a pre-existing first line of IgM defense against common microbes. Up to 50% of the IgA-producing cells in the lamina propria are derived from peritoneal cavity B-1 cells. These cells are therefore an important source of the mucosal IgA which coats the normal microflora of the gut.

There is an orderly sequence of Ig gene reaffangements during early B-cell differentiation. Stage 1. Initially, the D-/ segments on both heavy chain coding regions (one from each parent) rearrange (figure 77.14). Stage 2. A V-DI recombinational event now occurs on one heavy chain. If this proves to be a nonproductioe rearrangement (i.e. adjacent segments are joined in an incorrect reading frame or in such a way as to generate a termination codon downstream from the splice point), then a second V-Dl reanangement will occur on the sister heavy chain region. If a productive rearrangement is not achieved, we can wave the pre-B-cell a fond farewell.

+ f o+ + + +++ ++ ++

rCBp/N gene(Xid)producing micehoveonX-linked immunodeficiency o defect (btk)ossocioted in theBruton tyrosine kinose wilhpoorB-l cellmoturotion ond responses inodequoie lotypell T-lndependent ontigens 2Molheolen phospholose theprotein tyrosine micehovethemevmutotion offecting whichdr0motic0lly olters lhethreshold forontigen ondslrongly I C(PfP-/ C)gene Themicehovewidespreod development toword theB-l subpopulotion bioses outoimmunity ondmostoftheirB-cells oreB-I

D E V E T O P M EONFTB . C E t tS P E C I F I C I I Y

PROGENITOR (MATERNAL) GERMLINE lg GENES

(PATERNAL)

STAGE I DJ', REARMNGEMENT

STAGE 2 VDJH REARRANGEMENT

STAGE 3 SYNTHESIS OFSURROGATE .IgM, SURFACE RECEPTOR

'E 'E

4 STAGE

ALLELIC EXCLUSION H GENES OFSISTER <*EXTERNAL SIGNAL?

STAGE 5

VJLREARRANGEMENT slgMSYNTHESIS

STAGE 6

ALLELIC EXCLUSION OFREI/AINING LIGHT CHAIN GENES

ffi

= produclive : non-produclive reorrongemenl reorrongemenl; t\/::qJ_l

Figure 11.14. Sequence of B-cell gene rearrangements and postulated mechanism of allelic exclusion (seetext).

2 4 7 II

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y Stage3. If a productive rearrangementis made, the preB-cell now synthesizes p chains. At around the same time, two genes/ V,,,,." (CD179a)and /.. (CD179b), with hom*ology for the V. and C. segments of l, light chains respectively, are temporarily transcribed to form a 'pseudo-light chain'which associateswith the p chains to generatea surface surrogate'IgM'receptor, together with the lg-u(CD79a) and Ig-B (CD79b) chains conventionally required to form a functional B-cell receptor. Expression of this receptor is absolutely essential for further differentiation of the B-lymphocytes since disruption of the membrane exon of the p chain or of the 2. gene by hom*ologous recombination of embryonic stem cells (cf. p. 148) arrests development at the pre-B stage and the animal is devoid of mature B-cells.This surrogate receptor closely parallels the pre-Tcr/Breceptor on pre-T-cellprecursors of crBTCR-bearing cells. Stage 4. The surface receptor is signaled, although it is unclear whether this normally occurs in a cell autonomous manner in the absence of an external Iigand or whether ligands on stromal cells are engaged. Signaling by the pre B-cell receptor through tgu/F causesallelic exclusion whereby there is suppressionof any further rearrangement of heavy chain genes on a sister chromatid. Stage 5. Expression of the lnterferon regulatory/actors IRF-4 and IRF-8 downregulates the production of Vpre-B and ),5 and induces rearrangement of the conventional light-chain genes.This involves V-/recombinations on first one and then the other r allele until a productive V^.-l rearrangement is accomplished. If this fails, an attempt is made to achieve productive rearrangement of the l, alleles. Subsequent expression of a productively rearranged light chain is thought to involve binding of IRF-4 and -8 together with the PU.1 and Spi-B transcription factors to the r or l" enhancers. Synthesisof conventional sIgM now proceeds. Stage 6. The sIgM molecule prohibits any further gene shuffling by allelic exclusion of any unrearranged light chain genes.sIgD bearing an identical VDJ sequenceto the IgM heavy chain is produced by alternative splicing of the heavy chain RNA transcript and the naive IgM*IgD* B-cell is now ready for its encounter with antigen. Upon antigenic stimulation IgD is lost and, in the presence of appropriate T-cell help, classswitching from IgM to IgG, IgA or IgE antibody production can occur. At the terminal stagesin the life of a fully mature plasma cell, virtually all surface Ig is shed. Theimporlonce of qllelicexclusion Since each cell has chromosome complements derived

from each parent, the differentiating B-cell has four light and two heavy chain gene clusters to choose from. We have described how, once the VD/ DNA rearrangement has occurred within one heavy and Vl rearrangement in one light chain cluster, the V genes on the other four chromosomes are held in the embryonic state by an allelic exclusion mechanism so that the cell is able to express only one heavy and one light chain. This is essential for clonal selection to work since the cell is then only programed to make the one antibody it uses as its cell surface receptor for antigen. Furthermore, this gene exclusion mechanism prevents the formation of molecules containing two different light or two different heavy chains which would have nonidentical combining sites and therefore be functionally monovalent with respect to the majority of antigens; such antibodies would be nonagglutinating and would tend to have low avidity as the bonus effect of multivalency could not operate.

Differenlspecificresponses conoppeqrsequenti0lly The responses to given antigens in the neonatal period appear sequentially, suggesting a programed rearrangement of I/ genes in a definite order (figure 11.15).Early in ontogeny there is a bias favoring the rearrangement of the V" genes most proximal to the D/ segment.

E I H E I N D U C I I OO N FI O T E R A N CI N P H O C YEI S B .T Y M byclonol delelion ondclonol onergy Toleronce conbecoused just as for T-cells, so both mechanisms can operate on Bcells to prevent the reaction to self. Excellent evidence

SHEEP BRUCELLARBC Ol.'{}A

PNEUMOCOCCAL DONKEY POLYSACCHARIDE SIII RBC KLH

+++

0 "0

6 7 Age(doys)

1418222630

Figure 11.15. Sequential appearance ofresponsiveness to different antigens in the neonatal rat. RBC, red blood cell; KLH, keyhole limpet hemocvanin.

CHAPTER I I - O N T O G E N YA N D P H Y L O G E N Y for deletion comesfrom mice bearing transgenescoding forlgMwhichbinds to H-2Kmolecules of allH-2 haplotypes except d andf.Miceof H-2dhaplotype expressthe transgenic IgM abundantly in the serum, while 25-50% of total B-cells bear the transgenic antibody. (d x k) F7 crosses completely failed to express the transgene, either in the serum or on B-cells,i.e. B-cellsprogramed for anti-H-2Kk were expressedinH-2d mice but deleted in mice positive for H-2Kk which in thesecirc*mstances actsas an autoantigen. Tolerance through B-cell anergy was clearly demonstrated in another study in which double transgenic mice were made to expressboth soluble lysozyme and a high affinity antibody to lysozyme. The animals were completely tolerant and could not be immunized to make anti-lysozyme; nor did the transgenic antibody appear in the serum althoughitwas abundantlypresent on the surface of B-cells.These anergic cells could bind antigen to their surface receptorsbut could not be activated. Like the aged rou6, wistfully drinking in the visual attractions of some young belle, these tolerized lymphocytes can 'see' the antigen but lack the ability to do anything aboutit. Whether deletion or anergy is the outcome of the encounter with self probably depends upon the concentration and ability to cross-linklg receptors.In the first of the two B-cell tolerance models above, the H-2Kk autoantigen would be richly expressed on cells in contact with the developing B-lymphocytes and could effectively cause cross-linking. In the second case, the lysozyme, masquerading as a 'self'-molecule, is essentially univalent with respect to the receptors on an antilysozyme B-cell and would not readily bring about cross-linking. The hypothesis was tested by stitching a transmembrane hydrophobic segment onto the lysozyme transgene so that the antigen would be inserted into the cell membrane. Result? B-cells expressing the high affinity anti-lysozyme transgene were eliminated. Another self-censoring mechanism, receptor editing, can come into play. We have already discussedone type of receptor editing (.f. p. 69) in which secondary rearrangements substitute another V gene onto an already rearranged 7(D)/ segment. However, receptor editing can also occur by wholesale replacement of an entire light chain. This can best be explained by an example. If the heavy and light chain Ig genes encoding a high affinity anti-DNA autoantibody are introduced as transgenes into a mouse, a variety of light chains are produced by genetic reshuffling until a combination with the heavy chain is achieved which no longer has anti-DNA activity, i.e. the autoreactivity is edited out. This will often involve replacement of a r light chain

with a new rearrangement made on the l, light chain locus and is associated with re-expression of the R 4G1 / 2 genes. Most peripheral B-cellsin mice are ligand selectedas revealed by analysis of the V' repertoire at the cDNA level of bone marrow pre-B-cellscompared withmature spleen B-cells. Once peripheralized, the bulk of the Bcell pool is stable; lymph node B- (and T-) cells from unprimed mice survived comfortably for at least 20 months on transfer to H-2 identical SCID animals. moylesullfromhelplessB-cells Toleronce With soluble proteins at least, T-cells are more readily tolerized than B-cells (figure 11.16) and, depending upon the circulating protein concentration, a number of self-reacting B-cells may be present in the body which cannot be triggered by T-dependent self-components since the T-cells required to provide the necessary T-B help are already tolerant-you might describe the Bcells as helpless. If we think of the determinant on a selfcomponent which combines with the receptors on a self-reacting B-cell as a hapten and another determinant which has to be recognizedby a T-cell as a carrier (cf. figure 8.11), then tolerance in the T-cell to the carrier will prevent the provision of T-cell help and the B-cell will be unresponsive. Take C5 as an examPle; this is normally circulating at concentrations which tolerize T- but not B-cells. Some strains of mice are congenitally deficient in C5 and their T-cells can help C5-positive strains

BASIC THYROMYELIN PROTEIN GLOBULIN

SERUM ALBUMIN

YVV

o o

protein> 0ulologous of circuloling Molorily

Figure 11.16. Relative susceptibility of T- and B-cells to tolerance by circulating self-antigens. Those circulating at low concentration induce no tolerance; at intermediate concentration, e.g. thyroglobulin, T-cells are moderately tolerized; molecules such as albumin which circulate athrgh concentrations tolerize both B- and T-cells

I I - O N T O G E N YA N D P H Y L O G E N Y CHAPTER to make antibodies to C5, i.e. the C5-positive strains still have inducible B-cells but they are helpless and need nontolerized T-cells from the C5-negative strain (figure 77.77). It is worth noting the observation that injection of high doses of a soluble antigen without adjuvant, even when given several days after primary immunization with that antigen, prevented the emergence of high affinity mutated antibodies. Transfer experiments showed the T-cells to be tolerant. This tells us that, even when an immune response is well underway, T-helpers

DONORS

TRANSFER C5 POSITIVE IMMUNIZE T-HELPERS NORMAL WITHC5 ANTI-C5 RECIPIENTS

N A T U R AKTr t t E R( N K )C E t t 0 N T 0 G E N Y

o r'r'zro

in the germinal center are needed to permit the mutations which lead to affinity maturation of antibody and, as a further corollary, that soluble self-antigens in the extracellular fluids can act to switch off autoreactive Bcells arising in the germinal centers by hypermutation. Presumably, self-tolerance in both B- and T-cells involves all the mechanisms we have discussed to varying degrees and these are summarized in figure 11.18.Remember that, throughout the life of an animal, new stem cells are continually differentiating into immunocompetent lymphocytes and what is early in ontogeny for them can be late for the host; this means that self-tolerance mechanisms are still acting on early lymphocytes even in the adult, although it is always comforting to note that the threshold concentration for tolerance induction is very much lower for pre-B-cells relative to mature B-cells.

24sI

r.,on-to,e'on,

Figure 11.17. Circulating C5 tolerizes T- but not B-cells leaving them helpless. Animals withcongenital C5 deficiency donot tolerize their Thelpers and can be used to break tolerance in normal mice.

SILENCE

DELETION

The precise lineage of NK cells is still to be established. They differentiate in the bone marrow from precursors which express the shared B chain of the IL-2 and IL-15 receptor but lack other markers of mature NK cells such as NK1.1, CD16 (FcyRIII) and the NK inhibitory and stimulatory receptors. Although sharing a common early progenitor with T-cells, in contrast to NKT cells (cf . p. 105),classicalNK cells do not develop in the thymus.

ANERGY

HELPLESSNESS

CELLS OFACTIVATED SUPPRESSION

lnl

A I I I V

NOHELP

I I I V NosAgreceptors

I oreblue or suppressed Cells whichoredeod,unreoclive

Figure11.18. Mechanismsof self-tolerance(seetext) sAg, self-antigen;XsAg/Id, cross-reactingantigenoridiotype;APC, antigen-presentingcell; Th, T-helper; Ts, T-suppressor/T-regulatory cell; Tc, cytotoxic T-cell precursor

I zso

C H A P I E RI I - O N T O G E N YA N D P H Y L O G E N Y

T H EO V E R A tR t E S P O N SI N ET H EN E O N A T E Lymph node and spleen remain relatively underdeveloped in the human at the time of birth, except where there has been intrauterine exposure to antigens as in congenital infections with rubella or other organisms. The ability to reject grafts and to mount an antibody responseis reasonably well der.elopedby birth, but the immunoglobulin levels, with one exception, are low, particularly in the absenceof intrauterine infection. The exception is the IgG acquired by placental transfer from the mother using the neonatal Fc-receptor, FcRn (cf. figure 3.17). These antibodies are catabolized with a half-life of approximately 3-4 weeks and there is a fall in IgG concentration over the first 3 months accentuated by the increasein blood volume of the growing infant. Thereafter, the rate of IgG synthesis by the newborn's own B-cellsovertakesthe rate of breakdown of maternal IgG and the overall concentration increases steadily. The other immunoglobulins do not cross the placenta and the low but significant levels of IgM in cord blood are synthesizedby the baby (figure 77.79).IgM reaches adult levels by 9 months of age. Only trace levels of IgA, IgD and IgE are present in the circulation of the newborn.

BIRTH V

(

txwxtx

T H E E V O T U T I O NO F T H E I M M U N E R E S P O N S E ogoinslinfecfion Plonldefenses Plants are able to detect pathogens using a number of resistance proteins, encoded by R genes, which are capable of initiating an immediate hypersensitive response(HR). This leads to localized apoptosis thereby rapidly curtailing the growth of the infectious agent. 'immune state' of systemic acquired The HR induces an resistance (SAR) which persists for several weeks and extends to a broad range of bacterial, viral and fungal pathogens beyond the initiating infective agent. A series of SAR genesencodea wide variety of microbicidalproteins which can be induced through endogenous chemical mediators such as salicylic acid, jasmonic acid and methyl-2,6-dichloroisonicotinic acid. ]asmonic acid also contributes to resistance against herbivorous insects. Salicylic acid may also contribute to the acute defensive responseby inducing a catalase-dependentincreasein Hror' microbioldefensemechonisms lnvenebrote Mechanisms for the recognition and subsequent rejection of nonself can be identified in invertebrates as far down the evolutionary scale as marine sponges (figure 77.20). Phagocytosis is of importance throughout the animal kingdom (cf. Milestone 1.1). In many phyla, phagocytosis is augmented by coating with agglutinins and bactericidins capable of binding to pathogenassociated molecular patterns (PAMPs) on the microbial surface so providing the basis for the recognition of

NECROSIS

>s -50

0- J:

0612 ^ ^ Age(months) INFANT MATERNAL

Figure 11.19. Development of serum immunoglobulin levels in the human. (After Hobbs J R (1969)In Adinolfi M (ed ) ImmrLnology and p 118.Heinemann, London ) Deuelopntent,

Figure 11.20. Recognition and rejection of nonself. Parabiosed fingers of a marine sponge from the same colony are permanently united but members of different colonies reiect each otherby 7-9 days

I I - O N T O G E N YA N D P H Y L O G E N Y CHAPTER 'nonself'. It is notable that infection very rapidly induces the synthesis of an impressive battery of antimicrobial peptides in higher insectsfollowing activation of transcription factors which bind to promoter sequence motifs hom*ologous to regulatory elements involved in the mammalian acutephaseresponse.Thus, the toll molecule in drosophila is a receptor for PAMPs that activate NFrB in theseflies. Drosophila with a lossof-function mutation in toll are susceptible to fungal infections. Antimicrobial peptides produced by insects include disulfide-bridged cyclic peptides such as the 4 kDa anti-Gram-positive defensins and the 5 kDa antifungal peptide, drosomycin. Linear peptides inducible by infection include the cecropins and a series of antiGram-negative glycine- or proline-rich polypeptides. Cecropins, which have also been identified in mammals, are 4 kDa strongly cationic amphipathic o-helices causing lethal disintegration of bacterial membranes by creating ion channels. Elements of a primordial complement system also exist among the lower orders. A protease inhibitor, an or-macroglobulin structurally hom*ologous to C3 with internal thiolester,is present in the horseshoecrab. Conceivably this might represent an ancestral version of C3 which is activated by proteasesreleased at a site of infection, deposited onto the microbe and recognized there as a ligand for the phagocytic cells. The complement receptor CR3 is an integrin, and related integrins in insects may harbor common ancestors.Mention of the horseshoe crab may have stirred a neuronal network in readers with good memories, to recall its synthesis of limulin (cf. p. 17) which is hom*ologous with the mammalian acute phase C-reactive protein (CRP); presumably, it acts as a lectin to opsonize bacteria and is likely to be a product of the evolutionary line leading ultimately to C1q, mannose-binding lectin and lung surfactant. The other major strategy effectively deployed by invertebrates is to wall off an invading microorganism. This is achieved, for example, through proteolytic 'gelled' cascades which produce a coagulum of hemolymph around the offender.

Adoptiveimmuneresponsesoppeorwilh the verlebroles Lower aertebrates Lymphocytes and genuine adaptive T- and B-responses do not emerge in the phylogenetic tree until we reach the vertebrates, althoughneither canbe elicited in the lowliest vertebrate studied-the California hagfish. This unpleasant cyclostome (which preys upon moribund fish by entering their mouths and eating the flesh from

25r I

the inside) can respond to hemocyanin provided that it is maintained at temperatures approaching 20"C (in general, poikilotherms make antibodies better at higher temperatures), but true immunoglobulins are not involved. Further up the evolutionary scale in the cartilaginous fishes, well-defined 185 and 75 immunoglobulins with heavy and light chains are present, but the responsesare T-independent. T-cells appear The toad, Xenopus,is a pliable, if unlovely, speciesfor study since it is possible to make transgenics and cloned tadpoles fairly readlly, and it has a less complex lymphoid system than mammals, characterizedby a small number of lymphocytes and a restricted antibodyrepertoire not subject to somatic mutation. Furthermore, positive and negative thymic selection have been demonstrated in frogs. The emergence of a thymus in the teleosts (bony fishes), amphibians, reptiles, birds and mammals was of course associatedwith MHC molecules, cell-mediated immunity, cytotoxic T-cells and allograft rejection. It could be argued that we also seephylogenetically more ancient, T-independent B-1 (CD5-positive) cells joined by a new T-dependent B-2 population. However, T-dependent, high affinlty, heterogeneous, rapid secondary antibody responses are only seen with warmblooded vertebrates such as birds and mammals, and these correlate directly with the evolution of germinal centers. Generation of antibody dioercity Mechanisms for the generation of antibody diversity receive quite different emphasis as one goes from one species to another. We are already familiar with the mammalian system where recombinational events involve multiple V,D andJ gene segments. The horned shark also has many I/ genes, but the opportunities for combinatorial joining are tightly constrained by close linkage between individual V, D, I and Csegments and this may be a factor in the restricted antibody response of this species. In sharp contrast, there is only one operational V gene at the light chain locus in the chicken,but this undergoes extensive somatic diversification utilizing nonfunctional adjoining V pseudogenes in a somatic gene conversion-like process. Camel lovers should note that not only do they get by on little water but, like the llamas, a proportion of their functional antibodies lack light chains. The especially long CDR3 loop in the heavy chain variable region aPpears to compensate for the lack of a light chain in these antibodies.

Table 11.4. Effect of neonatal bursectomy and thymectomy on the development of immunologic competence in the chicken. (From Cooper M D, PetersonR D A., South M A & Good R A (1966)I ournal of ExperimentalMedicine l2S,T5,withpermission of the editors )

Peripherol AIIXDeloyed blood rgconc Antibody irrodioled offer skinreoclion Groft lymphocyte birth to luberculin rejeclion c0unl lntocl t4 800 ++ ++ Thymecfomized I 000 Bursectomized 1 32 0 0

T H EE V O T U T I O ONFD I S I I N C IB - A N DI . C E T T T I N E A G EWSA SA C C O M P A N I E BD Y I H E D E V E T O P M EONFIS E P A R A T E S I T E SF O RD I F F E R E N T I A I I O N The differential effects of neonatal bursectomy and thymectomy in the chicken on subsequent humoral and cellular responses paved the way for our eventual recognition of the separate lymphocyte lineages which subserve these functions. Like the thymus, the bursa of Fabricius develops as an embryonic outpushing of the gut endoderm, this time from hindgut as distinct from foregut, and provides the microenvironment to cradle incoming stem cells and direct their differentiation to immunocompetent B-lymphocytes. As may be seen from table 11.4,neonatal bursectomy had a profound effect on overall immunoglobulin levels and on specific antibody production following immunization, but did not unduly influence the cell-mediated delayed-type hypersensitivity (DTH) response to tuberculin or affect graft rejection. On the other hand, thymectomy grossly impaired cell-mediated reactions and inhibited antibody production to most protein antigens. The distinctive anatomical location of the B-cell differentiation site in a separate lymphoid organ in the chicken was immensely valuable to progress in this field because it allowed the above types of experiments to be carried out. However, many years wentby in a fruitless search for an equivalent bursa in mammals before it was realized thatthe primary site for B-cellgenerationwas in fact the bone marrow itself.

C E T T U T ARRE C O G N I T I O MN OtECUtES E X P T O ITTH EI M M U N O G T O B UG TE I NN E S UP ER F A M ItY When nature fortuitously chances upon a protein structure ('motif is thebuzz word) which successfully mediates some useful function, the selective forces of evolution make sure that it is widely exploited. Thus,

( r, r.) HEAW LIGHT

BOND DISULFIDE hom*oLOGY WITHIgV H0M0L0GY WITHlgc

g V

rrr DoMATN lTSl flsnoNrcrNTYPE MHCDOMAIN

{ GPIANCHOR

nvonoRroerc MEMBRANE SEGN4ENT

superfamily comprises a large Figure 11.21. The immunoglobulin number of surface molecules which all share a common structure, the immunoglobulin-type domain, suggesting evoiution from a single primordial ancestral gene. Just a few examples are shown. (a) Multigene families involved in antigen recognition (the single copy p2-microglobulin is included because of its association with class I) (b) Single copy genes Thy-1 is present on T-cells and neurons. The poly-Ig receptor transports IgAacrossmucosalmembranes. N-CAMis an adhesion molecuie binding neuronal cells together. It is also found on NK cells and a subpopulation of T-cells, but its function on these lymphoid cells is unknown (Reprinted by permission from Nature 323, 1.5. Copyright @ 1986, Macmillan Magazines Ltd with some updating )

CHAPTER I I . O N T O G E N YA N D P H Y L O G E N Y the molecules involved in antigen recognition which we have described in Chapters 3 and 4 are members of a gene superfamily related by sequenceand presumably a common ancestry. All polypeptide members of this family, which includes heavy and light Ig chains, T-cell receptor chains, MHC molecules and Br-microglobulin, are composed of one or more immunoglobulin hom*ology units. Each Ig-type domain is roughly 110 amino acidsin length and is characterizedby certain conserved residues around the two cysteines found in each domain and the alternating hydrophobic and hydrophilic amino acids which give rise to the familiar antiparallel p-pleated strands with interspersed short variable lengths having a marked propensity to form reversed 'immunoglobulin turns -the fold' (cf. p. a1). A very important feature of the Ig domain structure is mutual complementarity which allows strong interdomain noncovalent interactions, such as those between V,, and 7, and the two Crr3 regions. Gene duplication and diversification can create mutual families of interacting molecules. such as CD4 with MHC class IL CDS with

Mulllpotenf l0lhemolopoief icslemcellslromthebonemorrow giveriselo 0lllheformed offhebl00d elemenls . Expansion and differentiation are driven by soluble growth (colony-stimulating) factors and contact with reticular stromal cells. Ihe difierenfioti0n0f l-cells occurswilhin

lhemicroenvironmenl oflhelhymus . Precursor T-cells arising from stem cells in the bone marrow travel to the thymus under the influence of chemokines in order tobecome immunocompetent T-cells T-cellonlogeny o Differentiation

to immunocompetent T-cell subsets is accompanied by changes in the surface phenotype which canbe recognized with monoclonal antibodies. . TCR genes rearrange in the thymus cortex, producing a'y6 TCR or a pre-op TCR, consisting of an invariant pre-To associated with a conventional VB, before final rearrangement of the Vcr to generate the mature oB TCR. r Double-negative CD4-8- pre-T-cells are driven and expanded by Notch-mediated and other signals to become double positive CD4*8*. . The thymus epithelial cells positively select CD4t8* Tcells with avidity for their MHC haplotype so that singlepositive CD4+ or CD8* T-cells develop that are restricted to the recognition of antigen in the context of the epithelial cell haplotype.

253 |

MHC class I and IgA with the poly-Ig recePtor (figure 71.21). Likewise, the intercellular adhesion molecules ICAM-1 and N-CAM (figure 17.21) are richly endowed with these domains, and the long evolutionary history of N-CAM strongly suggeststhat these structures made an early appearance in phylogeny as mediators of intercellular recognition. In marine sponges, Ig superfamily structures are found both on the extracellular portion of the receptor tyrosine kinases (RTKs) and in the cell recognition molecules (CRMs), both thought to be involved in allograft rejection. Arecent trawl of the protein sequence databaserevealedhundreds of knownmembers of the Ig superfamily. Some family! The integrins, whose members include the leukocyte function-associated antigen-1 (LFA-1) and the very late antigens (VLAs), form another structural superfamily which contains a number of hematopoietic cell surface molecules concerned with adhesion to extracellular matrix proteins and to cell surface ligands; their function is to direct leukocytes to particular tissue sites (see discussion onp.757).

. NKT cells, which express both a TCR and NK cell markers such as NK1.t, have highly restricted TCR variable regions and recognize glycolipid antigens presented by the MHClike molecule CD1d. They seuete IL-4 and IFNy and rnay function as regulatory cells. I-cell toleronce . The induction of immunological tolerance is necessary to avoid self-reactivity. . High avidity T-cells which react with self-antigens presented by medullary epithelial cells and dendritic cells are eliminated by negative selection. The paradigm that low avidity binding to MHC-peptide produces positive selection and high avidity negative, is probably broadly true but may need some amendment. . The autoimmune regulator (AIRE) directs the ectopic expression of several organ-specific self antigens in the thymic medullary epithelial cells leading to deletion of the relevant T-cells. . Self-tolerance can also be achieved by anergy. . Anergic cells attached to a dendritic cell can downreguIate the antigen-presenting ability of that cell, resulting in infectious anergy. . Regulatory T-cells normally suppress the activities of self-reactive T-cells that escape the deletional or anergy Processes. . A state of what is effectively self-tolerance also arises when there is a failure to adequatelypresent a self-antigen to (Continuedp254)

I zs+

CHAPTER I I _ONTOGENY AND PHYLOGENY

lymphocytes, either because of compartmentalization, lack of class II on the antigen-presenting cell or low concentration of peptide-MHC (cryptic self). B-cellsrlifferentiolein lhe tetol liver ondthen in lhe bonemorrow . They become immunocompetent B-cells after passing through pro-B-, pre-B- and immature B-cell stages. . Pax5 expression is essential for progression from the preB- to immature B-cell stage B-l ond B-2 represenltwo dislinctsubpopulotions of B-cells o B-1 cells represent a minor population expressing high sIgM and low sIgD. B-1a cells are CD5*, B-1b are CDS-. The majority of conventional B-cells, the B-2 population, are slgMlo, slgDhi, CDs-. The 8-L population predominates in early life, shows a high level of idiotype-anti-idiotype connectivity, produces low affinity, IgM polyreactive antibodies, many of them autoantibodies, and is responsible for the T-independent'natural' IgM antibacterial antibodies which appear spontaneously. Developmenfot B-cellspecificity . The sequence of Ig variable gene rearrangements is D/ and then VDl o VD/ transcription produces p chains which associatewith Vp."e .Is chains to form a surrogate surface IgMJike recePtor. . This receptor signals allelic exclusion of unrearranged heavy chains and initiates rearrangement of V-l- and, if unproductive, V-,f^.

CIRCULATING ll\ilMUN0COMPETENT CELLS

THYMUS D2,3,5,1 | CD2,3,5, CDlt^D ,- \48 "f \4

6D)',-(O)

coRTtcAL I trlrourrnnv BONEMARROW

VT\

I MHCII

(r\)-!AL v iv PRE-B

Ihe induclionotloler0ncein B-lymphocyles o B-cell tolerance is induced by clonal deletion, clonal anergy, receptor editing and'helplessness' due to preferential tolerization of T-cells needed to cooperate in B-cell stimulation Nolurolkillet (NK) cell ontogeny . NK cells develop in the bone marrow and express inhibitory receptors for MHC class I and stimulating receptors recognizing a variety ofcell surface ligands. Theover0lltesponsein lhe neonole o Maternal IgG crosses the placenta and provides a high level of passive immunity at birth The antigen-independent differentiation within primary lymphoid organs and antigen-drivenmaturation in secondary lymphoid organs are summarized in figsre lL22.

response oflheimmune Theevolulion o Plants develop a systemic acquired resistance (SAR) to infection which shows broad specificity and lasts for several weeks. o Recognition of self is of fundamental importance for multicellular organisms, even lowly forms like marine sponges. ArfigerFdepe!{gId ' frnturotioo

Anflgon l rdeperdeffdlft rensotltrll STEI/CELL

r If the rearrangement at any stage is unproductive, i.e. does not lead to an acceptable gene reading frame, the allele on the sister chromosome is rearranged. o The mechanisms of allelic exclusion ensure that each lymphocyte is programed for only one antibody. Responses to different antigens appear sequentially with age.

TMMATURE

Exlensive I Generotion or proliferotion I subpopulotions Generotron I of diversity sull-tor.,on.,

l l

LYI\iIPHOID TISSUE SECONDARY

/r-\\

IVEIVORYCEI-LS

v v

EFF--CTOR CE

Figwe 11.22. Antigen-independent dif{erentiation and antigen-dependent maturation of T- and B-cells. Cortical thymocytes are positively selected to recognize self-MHC haplotype Tdl terminal deoxynucleotidyl transferase (Corttinuedp 255)

CHAPTER I I - O N T O G E N YA N D P H Y T O G E N Y

o Invertebrates have defense mechanisms based on phagocytosis, killing by a multiplicity of microbicidal peptides, and imprisonment of the invader by coagulation of the hemolymph. o B- and T-cell responses are well defined in the vertebrates and the evolution of these distinct lineages was accompaniedby the development of separate sites for differentiation. . The success of the immunoglobulin domain structure,

F U R T H ERRE A D I N G Baba Y, Pelayo R & Kincade P.W. (2004) Relationships between hematopoietic stem cells and lymphocyte progenltors. Trends in lmmuno Iogy 25, 645-649. BIom B & Spits H. (2006) Development of human lymphoid ce1ls Annual Reaieu of lmmunology 24,287-320. Bona C (2005)NeonatalImmunity. Humana Press,Totowa, NJ Busslinger M. (2004) Transcriptional control of early B ce1ldeveloprnent Annual Reaiezuof lmmunology 22,55-79. Chien Y, JoresR & Crowley M.P. (1996)Recognition by 1/6 T cells. Annual Rettiewof Immunology L4, 511-532. Germain R N (2002) T-cell development and the CD4{D8 lineage decision. Naf are Reaiewslmmunology 2,309-322 Horton J. & Ratcliffe N. (2002) Evolution of immunity. In Roitt I M., Brostoff J & Male D.K (eds) Immunology,6th edn, pp. 217-234 Mosby,London Kamradt T & Mitchison N.A. (2001) Tolerance and autoimmunity. NeruEnglandlournal of Medicine344,655464. Kang J & Der S.D (2004) Cytokine functions in the formative stages of a lymphocyte'shfe Current Opinion in Immunology L6, 180-190

255 |

possibly through its ability to give noncovalent mutual binding, has been exploited by evolution to produce the very large Ig superfamily of recognition molecules, including Ig, TCRs, MHC class I and II, Br-microglobulin, CD4, CD8, the poly-Ig receptor and Thy-1 Another superfamily, the integrins, which includes LFA-1 and the VLAmolecules, is concernedwith leukocyte binding to endothelial cells and extracellular matrix proterns.

Kronenberg M (2005) Towards an understanding of NKT cell biology: progress and paradoxes. Annual Rez,iewof Immunology 23, 877-900. Montecino-Rodriguez E., Leathers H & Dorshkind K. (2006) Identification of a B-1 B cell-specified progenitor Nature Immunology 7,

293-301. M. (1999) CommitNutt S.L.,Heavey8.,RolinkA.G.& Busslinger ment to the B lymphoid lineage depends on the transcription factor Pax5 Nature 4Ol, 556-562 Radtke F., Wilson A. & MacDonald H R. (2004) Notch signaling in T- and B-cell development. Current Opinion in lmmunology 16, 774-1,79. Rothenberg E.V & Taghon T. (2005) Molecular genetics of T ceil development Annual Reoiewof lmmunology 23, 607-649 . Staal F.J.T & Clevers H C. (2004) WNT signaling and haematopoiesis: a WNT-WNT situation Nature Reaiews lmmunology 5,

2t10. Starr T.K.,JamesonS.C & Hogquist K.A (2003)Positive and negative selectionolT cells AnnualReaiewof lmmunology21,139-176 YokoyamaW.M, Kim S & FrenchA.R. (2004)The dynamic life of natural killer k ells.Annual Reaiewoflmmunology22, 405429.

Adverso ies rioI strofeg duringinfection

INIRODUCTION We are engaged in constant warfare with the microbes which surround us and the processesof mutation and evolution have tended to selectmicroorganisms which have evolved means of evading our defense mechanisms. Pathogens continue to take a terrifying toll (figure 12.1),particularly in the developing world. Furthermore, the decline in mortality from infectious diseasewhich had been seen in countries such as the USA has gone into reverse,with a rise in deaths in that country from infectious disease increasing at a rate of 4.8% per year between 1981 and 1995. Among the current long list of problems are newly emerging infections including HSNI infl rter.za,E. coIi C-157:H7, p rions, Legionnellapneumophila,HIV and ebola virus. Furthermore, old adversarieshave re-emerged,such as dengue, West Nile virus, cholera, plague, rift valley fever and lyme disease.During the middle of the last century it had seemed that the introduction of antibiotics had finally beaten infectious disease, but now multidrug resistancehas become an extremely worrying development, as seen with tuberculosis, malaria, Streptococcus pneumoniae, Enterococcus aeruginosa faecalis,Pseudomonas and methicillin-resistant Staphylococcus aureus(MRSA). Resistance to sulfa drugs occurred in S. aureus in tt.,:re 1940s,to penicillin in the 1950s,to methicillin in the 1980sand to vancomycin in 2002. Infections arising after 48 hours of hospital admission may well have been acquired in the hospital and are then referred to as 'nosocomial' infections; MRSA and other multidrug-resistant organisms often lurk in such institutions. It is also becoming increasingly appreciated that infectious agents are related to many 'noninfectious' diseases,such as the association of Helicobacter pylori witln gastric ulcers and gastric cancer, and of

various viruses with other cancers.In this chapter, we look at the varied, often ingenious, adversarial strategies which we and our enemieshave developed.

INFTAMMATIR OE NV I S I I E D The acute inflammatory process involves a protective influx of leukocytes, complement, antibody and other plasma proteins into a site of infection or injury and was discussed in the introductory chapters. Let's now reexamine the mechanisms of inflammation in greater depth. The reader may find it helpful to have another look at the relevant sectionsin Chapters 1 and 2, particularly those relating to figures 1.15,7.76,1.\7 and2.78.

Mediolorsof i nflommolion A complex variety of mediators are involved in acute inflammatory responses(figure 12.2).Some act directly on the smooth muscle wall surrounding the arteriolesto alter blood flow. Others act on the venules to cause contraction of the endothelial cells with transient opening of the interendothelial junctions and consequent transudation of plasma. The migration of leukocytes from the bloodstream is facilitated by mediators which upregulate the expression of adherence molecules on both endothelial and white cells and others which lead the leukocytes to the inflamed site through chemotaxis. cellsthloughp0iled leukocylesbindfo endotheli0l odhesionmolecules Redirecting the leukocytes charging along the blood into the site of inflammation is somewhat like having to encouragebulls stampeding down the Pamplona main streetto move quietly into the side roads.The adherence

CHAPTER I 2 . A D V E R S A R I A LS T R A T E G I EDSU R I N GI N F E C T I O N

nisms to destroy the pathogen (cf. p. 6). Additionally, they releaseneutrophil extracellular fraps (NETs)which act like a spider's web to ensnare the prey and thereby prevent them from spreading (figure 12.4).The NETs contain a number of antimicrobial agents including elastase,cathepsin G, myeloperoxidase and lactoferrin, thereby also contributing directly to destruction of the

Figure 12.1. Deaths from infectious disease. These six diseases caused almost 90% of the 14 million deaths from infectious disease worldwide, as estimated by the World Health Organization for the int/topics/ year 2000 For more information, see http://wwwwho infectious diseases/en

of leukocytes to the endothelial vesselwall through the interaction of complementary binding of cell surface molecules is an absolutely crucial step.Severalclassesof molecule subserve this function, some acting as lectins to bind a carbohydrate ligand on the complementary partner.

lniti0li0nof lhe oculeinflommotory response A very early event is the upregulation of P-selectinand platelet activating factor (PAF) on the endothelial cells lining the venules by histamine or thrombin released by the original inflammatory stimulus. Recruitment of these molecules from intracellular storage vesicles ensures that they appear within minutes on the cell surface.Engagement of the lectin-like domain at the tip of the P-selectin molecule with sialyl Lewis* carbohydrate determinants borne by the P-selectinglycoprotein ligand-1 (PSCL-1)on the neutrophil surface causesthe cell to slow and then roll along the endothelial wall and helps PAF to dock onto its corresponding receptor.This, in turn, increases surface expression of the integrins leukocyte function-associatedmolecule-1 (LFA-1) and Mac-1 which bind the neutrophil very firmly to the endothelial surface (figure 12.3). Activation of the neutrophils also increases their responsivenessto chemotactic agents and, under the influence of CSa and leukotriene 84, they exit from the circulation by moving purposefully through the gap between endothelial cells, across the basem*nt membrane (diapedesis) and up the chemotactic gradient to the inflammation site. Here they phagocytose the microorganisms and use their various killing mecha-

microorganisms. Damage to vascular endothelium, which exposesthe basem*nt membrane, and bacterial toxins such as LPS, trigger the blood coagulation and fibrinolysis pathways. Activation of platelets, for example by contact with basem*nt membrane collagen or induced endothelial PAF, leads to the release of many inflammatory mediators including histamine and a number of chemokines that are stored in granules. Some newly synthesizedmediators such as IL-1B are translated from mRNA in the anucleate platelets. Aggregation of the activated platelets occurs and thrombus formation is initiated by adherence through platelet glycoprotein Ib to von Willebrand factor on the vascular surface' Such platelet plugs are adept at stemming the loss of blood from a damaged artery, but in the venous system the damaged site is sealed by a fibrin clot resulting from activation of the intrinsic clotting system via contact of Hageman factor (factor XII) with the exposed surfaceof the basem*nt membrane. Activated Hageman factor also triggers the kinin and plasmin systems and several of the resulting products influence the inflammatory process, including bradykinin and fibrinopeptides which, together with complement components C3a and C5a, increase vascular permeability and thrombin which contributes towards activation of endothelium. process Theongoinginflommofory One must not ignore the role of the tissue macrophage which, under the stimulus of local infection or injury, secretes an imposing array of mediators. In particular, the cytokines IL-1 and TNF act at a later time thanhistamine or thrombin to stimulate the endothelial cells and maintain the inflammatory process by upregulating Eselectin and sustaining P-selectin expression. Thus, expressionof E-selectinoccurs2-4 hours after the initiation of acute inflammation, being dependent upon activation of gene transcription. The E-selectin engages the glycoprotein E-selectin ligand-1 (ESL-1) on the neutrophil. Other later acting components are the chemokines (chemotacticcytokines) IL-8 (CXCL8) and epithelial-derived neutrophil attractant-78 (ENA-78, CXCLS)which are highly effective neutrophil chemoattractants. IL-1 and TNF also act on endothelial cells, fibroblasts and epithelial cells to stimulate secretion of

I zse

CHAPTER I 2 - A D V E R S A R I ASLT R A T E G I E DS U R I N GI N F E C T I O N

DILATATION HISTA[/INE

BRADYKININ

DCF /I

+++

VIP

+++

NITRIC OXIDE

+++

CONSTRICTION

INCREASE PERMEABILIry

ENDOTHELIUM

NEUTROPHIL

NEUTROPHIL CHEMOTMIS

++ ++ Polenliole othermediolors +++ +

LEUKOTRIENE-D4

LEUKOTRIENE-C4 C5o

LEUKOTRIENE-84

f f

f,Met,Leu Phe

PLATELET ACTIVATING FACTOR

ADHESION i/OL, UPREGULATE

+++

rL-8

+++

NAP-2(CXCL7)

tL-I

TNF

\ INCREASE/DECREASE BLOOD FLOW

NEUTROPHIL DIAPEDESIS

TRANSUDATION OFPLASIVA

HELIAL RETRACTION ARTERY

ARTERIOLE

CAPILLARIES

VENULE

VEIN

Figure12.2. Theprincipalmediatorsofacuteinflammation.Thereadershouldreferbacktofigurel.l5torecalltherangeofproductsgeneratedby the mast cel1.The later acting cytokines such as IL-1 are largely macrophage-derived and thesecells also secreteprostaglandin E, (PGE'), leukotriene B,,and the neutrophil activating chemokine NAP-2 (CXCL7) VIP, vasoactive intestinal peptide

another chemokine, MCP-1 (CCL2), which attracts mononuclear phagocytes to the inflammatory site to strengthen and maintain the defensive reaction to infection. Perhapsthis is a good time to remind ourselves of the important role of chemokines (table 9.3) in selectively attracting multiple types of leukocytes to inflammatory foci. Inflammatory chemokines are typically induced by microbial products such as lipopolysaccharide (LPS) and by proinflammatory cytokines including IL-1, TNF and IFNy. As a very broad generalization, chemokines of the CXC subfamily, such as IL-8, are specific for neutrophils and, to varying extents, lymphocytes, whereas chemokines with the CC motif are chemotactic for T-cells,monocytes, dendritic cells, and variably for natural killer (NK) cells, basophils and eosinophils.

Eotaxin (CCL11)is chemotacticfor eosinophils, and the presence of significant concentrations of this mediator together with RANTES (regulated upon activation normal T-cell expressedand secreted,CCL5) in mucosal surfaces contribute towards the enhanced population of eosinophils in those tissues. The different chemokines bind to particular heparin and heparan sulfate glycosaminoglycans so that, after secretion, the chemotactic gradient can be maintained by attachment to the extracellular matrix as a form of scaffolding. Clearly, this whole operation serves to focus the immune defenses around the invading microorganisms. These become coated with antibody, C3b and certain acute phase proteins and are ripe for phagocytosis by the neutrophils and macrophages; under the influence of the inflammatory mediators these have

U R I N GI N F E C T I O N I 2 _ A D V E R S A R I ASLT R A T E G I D ES CHAPTER

ROLLING

25sI

ACTIVATION

BAsem*nT IVIE[/BRANE

TNF,LPS Figure 12.3. Early events in inflammation affecting neutrophil margination and diapedesis. Induced upregulation of P-selectinon the vessel u.alls plays the major role in the initial leukocyte endoihelial interaction (rolling) by interaction with ligands on the neutrophil such as P-selectin glycoprotein ligand-1 (PSGL-1,CD162). Recognition of extracellular gradients of the chemotactic mediators by receptors on the polymorphonuclear neutrophil (PMN) surface triggers intraccllular signals rvhich generate motion Neutrophil migration along the extracellular matrix vitronectin is dependent upon very rapid cyclesof

Figure 12.4. Neutrophil extracellular traps. Releaseof granule proteins and chromatin from neutrophils leads to the formation of neutrophil extracellular traps (NETs)which prevent bacterial spreading and ensurethat microbicidal substancesreleasedfrom the neutrophils are kept in the immediate vicinity of the bacteria for optimal killing of the microbe and minimal collateral damage to host tissues Scanningelectron micrograph of NETs from IL-8 activated neutrophils trapping: (A) St oyl tyl ttcoccLL.nur cus; (B) S t y1ilr ntLrirLnt;(C) S /crne ll The bar indicates 500 nm (Rcproduced from Brinkman c/ rI (2004)Sci,ricc303,1532,with permission ftom the Publishers )

integrin-dependent adhesion and detachment regulated by calcineurin. The cytokine-induced expression of E-selectin, whrch is recognized by the glycoprotein E-selectin ligand-L (ESL-1) on the neutrophil, occurs as a later event Chemotactic factors such as IL-8, which is secretedby a number of cell types including the endothelium itself, are important mediators of the inflammatory Process (Compare events involved in homing and transmigration of lymphocytes, figure 76)

upregulated complement and Fc receptors, enhanced phagocytic responsesand hyped-up kilting powers, all adding up to bad news for the bugs. Of courseit is beneficial to recruit lymphocytes to sites of infection and we should remember that endothelial cells in these areas express VCAM-1 (cf. p. 160) which acts as a homing receptor for VlA-4-positive activated memory T-cells, while many chemokines (cf. table 9.3) are chemotactic for lymphocytes. Regulotion 0ndresolution of inflommolion With its customary prudence, evolution has established regulatory mechanisms to prevent inflammation from getting out of hand. At the humoral level we have a seriesof complement regulatory proteins: C1 inhibitor, C4b-binding protein, the C3 control proteins factors H and I, complement receptor CR1 (CD35), decay accd.er ating /actor (DAF, CD55), ruembrane cofactor protein (MCR CD46), immunoconglutinin and hom*ologous restriction factor 20 (HRF20, CD59) (cf. p. 314).Some of the acute phase proteins derived from the plasma transudate, including a-1 antichymotrypsinogen, o(-1antitrypsin, heparin cofactor-2 and plasminogen-activator inhibitor-1, are proteaseinhibitors. At the cellular level, PGE', transforming growth factor-B GGFB) and glucocorticoids are powerful regulators.PGE, is apotentinhibitor of lymphocyte proliferation and cytokine production by T-cells and macrophages. TGFB deactivates macrophages by inhibiting the production of reactive oxygen intermediates, inhibiting the class II MHC transactivator (CIITA) and thus downregulating class II expression, and quelling the cytotoxic enthusiasm of both macrophages and NK cells. Endogenous glucocorticoids produced via the hypothalamic-pituitary-adrenal axis exert their anti-inflammatory effects both through the repression of a number of genes for proinflammatory cytokines and adhesion molecules, and the induction of the inflammation inhibitors lipocortin-1, secretory leukocyte proteinase inhibitor (SLPI, an inhibitor of neutrophil elastase)and IL-1 receptor antagonist. IL-10 inhibits antigen presentation, cytokine production and nitric oxide killing by macrophages, the latter inhibition being greatly enhanced by synergistic action with iL-4 andTGFB. Once the agent that has provoked the inflammatory reaction has been cleared, these regulatory processes will normalize the site. When the inflammation traumatizes tissues throughits intensity and extent, TGFB plays a major role in the subsequent wound healing by stimulating fibroblast division and the laying down of new extracellular matrix elements.

Chronicinflommolion If an inflammatory agent persists, either because of its resistance to metabolic breakdown or through the inability of a deficient immune system to clear an infectious microbe, the character of the cellular response changes.The site becomes dominated by macrophages with varying morphology: many have an activated appearance, some form what are termed 'epithelioid' cells and others fuse to form giant cells. Lymphocytes in various guises are also often present. This characteristic granuloma walls off the persisting agent from the remainder of the body (see type IV hypersensitivity in Chapter 15,p. 356).

E X T R A C E T T UB TAC R I E R IS A U S CP ET I B T E T OK I T T I N G B YP H A G O C Y T O S I S A N DC O M P T E M E N T Bocleriol suruivol slrolegies As with virtually all infectious agents, if you can think of a possible avoidance strategy, some microbe will already have used it (table 12.1). Ea ading phagocytosis The cell walls of bacteria are multifarious (figure 12.5) and in some casesare inherently resistant to microbicidal agents;but many other strategiesare used to evade the immune response (figure \2.6). Acommon mechanismby whichvirulent forms escapephagocytosis isby synthesis of an outer capsule, which does not adhere readily to phagocytic cells and covers carbohydrate molecules on the bacterial surface which could otherwise be recognized by phagocyte receptors. For example, as few as 10 encapsulated pneumococci can kill a mouse but, if the capsule is removed by treatment with hyaluronidase, 10 000 bacteria are required for the job. Many pathogens evolve capsuleswhich physically prevent accessof phagocytes to C3b deposited on the bacterial cell wall. Other organisms have actively antiphagocytic cell surface molecules and some go so far as to secrete exotoxins, which actually poison the leukocytes. Yet another ruse is to gain entry into a nonphagocytic cell and thereby hide from the professional phagocyte. Presumably, some organisms try to avoid undue provocation of phagocytic cellsby adhering to and colonizingthe externalmucosalsurfacesof the intestine. Challenging the compl ement sy st em Poor actiaationof complemenf.Normal mammalian cells are protected from complement destruction by regula-

Table 12.1. Examples of mechanisms used by bacteria to avoid the host immune response. (Partly based on Merrell D.S. and Falkow S (2004)Nature 430,250 )

Phogocylosis Yersinio

lnhibition ofoctin skelefon in phogocytes byYopT cleovoge ofRhoA (seefigure 12,7)

Legionello

Dot/icm inlrocellulor mulliplicoiion phogolysosomefusion genesinhibit

Complemenl Slreplococcus M profeinbindingof C4b-binding pyogenos proleinreducssC3 convertose oclivity Apoptosis

Shigelloflexnel

lpoB-medioled ofCospose1 oclivolion (seefigureI 2,7) rnduces opoplosis

MycobocteriumIncreosed expression ofbcl2ondRb luberculosis inhibits opoplosis Cytokine production

Vibriocholeroe Cholero loxininhibition oflL-l2 secrelion Bordetello perlussis

Antibody

induclion oflL-l ond Pertussistoxin tL-4

StophylococcutlgGopsonizotion forphogocylosis oureus blocked byProtein Abinding the onlibody lhe'wrongwoy'round Neisseilo gonoffhoeoe

T-celloclivotion Helicobocter pyloti

Antigenic voriotion byrecombinotion withinp/Egene inhibits VocAvocuololing cyfotoxin pothwoys colcineurin signoling

tory proteins such as MCP and DAR which cause C3 convertase breakdown (see p. 314 for further discussion). Microorganisms lack these regulatory proteins so that, even in the absence of antibody, most of them would activate the alternative complement pathway by stabilization of the C3bBb convertase on their surfaces. However, bacterial capsules in general tend to be poor activators of complement and selective Pressures have favored the synthesis of capsules whose surface comPonents do not permit stable binding of the convertase. Accelerntion of complementbreakdown.Members of the regulators of complement activation (RCA) family which diminish C3 convertase activity include C4bbinding protein (C4BP), factor H and factor Hlike protein 1 (FHL-1). Certain bacterial surface molecules, notably those rich in sialic acid, bind factor H (figure 12.6), which then acts as a focus for the degradation of C3b by the serineproteasefactor I (cf.p. 10).This is seen, for example, with Neisseriagonorrhoeae.Similarly, the hypervariable regions of the M-proteins of certainStreptococcuspyogenes(group A streptococcus) strains are able to bind FHL-I, whilst other strains downregulate complement activation by interacting with C4BP, this time acting as a cofactor for factor I-mediated degradation of the C4b component of the classical pathway C3 convertase C4b2a. All group A streptococci, and SrouP B, C and G streptococci of human origin, produce a CSa peptidase which acts as a virulence factor by proteolytically cleaving and therebyinactivating C5a. Complement deaiation. Some species manage to avoid lysis by deviating the complement activation site either to a secreted decoy protein or to a position on the bacterial surface distant from the cell membrane.

GLYCOLIPIDS MYCOLIC ACIDS ARABINOGALACTAN

LIPOTEICHOIC

PEPTIDOGLYCAN (MUREIN) CELLMEMBRANE GRAM-NEGATIVE

Figure 12.5. The structure of bacterial cell walls. All types have an inner cell membrane and a peptidoglycan wall which can be cleaved by lysozyme and lysosomal enzymes The outer lipid bilayer of Gramnegative bacteria, which is susceptible to the action of complement or cationic proteins, sometimes contains lipopolysaccharide (LPS; also known as endotoxin; composed o{ a membrane-distal hydrophilic

polysaccharide (which forms the highly polymorphic O-specific antigens) attached to a basal core polysaccharide, itself linked to the hydrophobic membrane-anchoring lipid A' 179 O antigen variants of Escherichiacoli areknown) . The mycobacterial cell wall is highly resistant to breakdown. When present, outer caPsules may protect the bacteria from phagocytosis.

I zaz

CHAPTER I 2 - A D V E R S A R I AS L T R A T E G I EDSU R I N G

t--:'l

Irt /--Y--\

(3lj,

Cleovoge

c' devi0li0n

Poorstobilisolron

lEuyorrRbmel -I qreerl t7 a t -Atg ixIE- | i fdx@el or Anti-phogocytic

(s)

No inserlion

\c l

Poorodherence

Phogocyte

Figure 12.6. Avoidance strategies by extracellular bacteria. Clockwise from left: (a) microbe attaches to surface component to enter nonphagocytic cell; (b) accelerating breakdown of complement by action of microbial products; (c) complement effectors are deviated from the mrcrobial cell wall; (d) capsule provides nonstabilizing surface for

alternative pathway convertase; (e) cell wall impervious to complement membrane attack complex (MAC); (f) capsule gives poor phagocyte adherence; (g) exotoxinpoisons phagocyte; (h) surface inhibitor of phagocytosis.

Resistance to insertionof terminalcoffiplement clmponents. Gram-positive organisms (cf. figure 12.5)have evolved thick peptidoglycan layers which prevent the insertion of the lytic C5b-9 membrane attack complex into the bacterial cell membrane (figure 12.6).Many capsulesdo the same.

spirochete Borrelia burgdorferi, of enzymes involved in synthesizing surface structures in Campylobacterj ejuni, and of the pili tn Neisseriameningitidis. In addition, new strains can arise, as has occurred with the lifethreatening E. coli O757:H7 which can causehemolytic uremic syndrome and appears to have emerged about 50 years ago by incorporationof Shigellatoxin genesinto the E. coll 055genome.

lnterfering with internal eaents in the macrophage Enteric Gram-negative bacteria in the gut have developed a number of ways of influencing macrophage activity, including inducing apoptosis, enhancing the production of lL-I, preventing phagosome-lysosome fusion and affecting the actin cytoskeleton (figwre12.7). Antigenic oariation Individual antigens can be altered in the face of a determined host antibody response. Examples include variation of surface lipoproteins in the lyme disease

Thehostcounler-oltock Antibodies can defeat these devious attempts to avoid engulfment by neutralizing the antiphagocytic molecules and by binding to the surface of the organisms to 'opsonizing' focus the site for fixation of complement, so them for ingestion by neutrophils and macrophages or preparing them for the terminal membrane attack complex (Milestone 12.1).However, antibody produc-

CHAPTER I 2 - A D V E R S A R I AS T T R A T E G I EDSU R I N GI N F E C T I O N

Figure 12.7. Evasion of macrophage defenses by enteric bacteria. The IpaB (lnvasion plasmid nntigen B ) and SipB (Salmonella invasion protein B) molecules secreted by Shigella and Solmonello,respectively, can activate caspase1 and thereby set off a train of events that will lead to the death of the macrophage by apoptosis Activated caspase 1 also triggers a protease which cleaves pro-Il-1, thereby causrng the release of large amounts of this proinflammatory cytokine from the macrophage Paradoxically,this may be advantageous to the bacteria because the subsequent migration of neutrophils to the intestinal lumen results in a loosening of intercellular junctions between the enterocytes, permitting cellular invasion of the basolateral surface by organisms from the lumen The SpiC (5almonella pathogenicity lsland C) protein from Salmonellainhibits the trafficking of cellular vesicles, and thereforeis able to prevent lysosomesfusing with phagocytic vesicles Yersininproduces a number of Yop molecules (Yersinia outer proteins) able to interfere with the normal functioning of the phagocyte For example, YopJ inhibits TNF production and downregulates NFrB and MAP kinases, thereby facilitating apoptosis by inhibiting antiapoptotic pathways YopT prevents phagocytosis by modifying the CTPase RhoA involved in regulating the actin cytoskeleton (Based on Donnenberg M S. (2000)Na/rre 406,768)

The pioneering research which led to the recognition of the antibacterial protection afforded by antibody clustered in the Iast years of the 19th century. A good place to start the story is the discovery by Roux and Yersin in 1888,at the Pasteur Institute in Paris, that the exotoxin of diphtheria bacillus could be isolated from a bacterium-free filtrate of the medium used to culfure the organism. von Behring and Kitasato at Koch's Institute in Berlin in 1890then went on to show that animals could develop an immunity to such toxins which was due to the development of specific neutralizing antidotes referred to generally as antibodies. They further succeededin passively transferring immunityto another animal with serum containing the antitoxin. The dawning of an era of serotherapy came in 1894 with Roux's successful treatment of patients with diphtheriabyinjection of immune horse serum. Sir Almroth Wright in London in 1903 proposed that the main action of the increased antibody produced after infec-

263 |

tion by B-cells usually requires T-cell help, and the Tcellsneed to be activated by antigen-presentingcells. As already discussed in Chapter 1, but so important that it is worth repeating, pathogen-associated ruolecular patterns (PAMPs), such as the all-important lipopolysaccharide (LPS) endotoxin of Gram-negative bacteria, peptidoglycan, lipoteichoic acids, mannans/ bacterial DNA, double-stranded RNA and glucans, are molecules which are broadly expressed by microbial pathogens but not present on host tissues. Thus these molecules serve as an alerting service for the immune system, which detects their presence using pattern recognition receptors (PRRs)expressed on the surface of antigen-presenting cells. It will be recalled that such receptors include the mannose receptor (CD206) which facilitates phagocytosis of microorganisms by macrophages and the scavenger receptor (CD204) which mediates clearance of bacteria from the circulation. LPS-binding protein (LBP) transfers LPS to the CD14 PRR on monocytes, macrophages,dendritic cells and B-cells. This leads to the recruitment of the toll-like receptor 4 (TLR4) molecule which triggers expressionof proinflammatory genes, including those for IL-7,IL-6, IL-12 and TNF, and the upregulation of the CD80 (87.1) and CD86 (87.2) costimulatory molecules. Other TLRs, e.g. TLM, recognize Gram-positive bacterial cell wall components. Although each of the 11 or so toll-like receptors so far characterized recognize broadly expressed microbial structures, it has been suggested that collectively they are able to some extent to discriminate between differentpathogens by detection of partic-

tion was to reinforcekiliing by the phagocytes.He called the antibodiesopsonins (Gk opson,a dressingor relish),because and they preparedthe bacteriaasfood for the phagocyticce11s, amply verified his predictions by showing that antibodies dramatically increasedthe phagocytosisof bacteria in aitro, thereby cleverly linking in nateLondaptirteimmunity. George Bernard Shaw even referred to Aimroth Wright's proposal in his play The Doctor's Dilemma In the preface he 'the gave an evocativedescriptionof the function of opsonins: white corpuscles or phagocytes which attack and devour disease germs for us do their work only when we butter the disease germs appetizingly for them with a natural sauce which Sir Almroth named opsonins. . . .' (A more extended account of immunology at the turn of the 19th century may be found in Silverstein A M. (1989) A History of lmmunology. Academic Press,SanDiego.)

(Contintedp,264)

I za+

Figure M12.1.1. Emil Behring (1854-1917).

CHAPTER I 2 - A D V E R S A R I AS L T R A T E G I EDSU R I N G

Figure M12.1.3. Sir Almroth Wright. (S1idekindly supplied by The WellcomeCentre Med iLibrary, ca1 Photographic London )

von

Figure M12.1.2. von Behring extracting serum using a tap. Caricature by Lustigen Bliittern, 1894 (Legend:'Serum direct from the horse! Freshly drawn') (S1idekindly supplied by The Wellcome Centre Medicai Photographic Lrbrary, London.)

ular combinations of PAMPS in a 'bar-code'-tvpe apProach. Toxin neutralization Circulating antibodies can neutralize the soluble antiphagocytic molecules and other exotoxins (e.g. phospholipase C of Clostridiumwelchii)releasedby bacteria. Combination near thebiologically active site of the toxin would stereochemically block reaction with the substrate, whilst combination distant from the active site may also causeinhibition through allosteric conformational changes. In its complex with antibody, the toxin maybe unable to diffuse away rapidly and willbe susceptibleto phagocytosis.

and their rate of clearance from the bloodstream is strikingly enhanced (figure 12.8).The less effective removal of coated bacteria in complement-depleted animals emphasizesthe synergism between antibody and com-

Op soniz ation of b acteri a Independentlyof antibody. Differencesbetween the carbohydrate structures on bacteria and self are exploited by the collectins (cf. p. 17), a series of molecules with similar ultrastructure to C1q and which bear C-terminal Iectin domains. These include mannose-binding lectin (MBL) which, on binding to terminal mannose on the bacterial surface, initiates antibody-independent complement activation (cf. p. 17). Other collectins, lung surfactant proteins SP-A and SP-D and, in cattle, conglutinin, also recognize carbohydrate ligands and can all act as opsonins (see Milestone 12.1) mediating phagocytosis by virtue of their binding to the C1q recepror. Augmented by antibody. Encapsulated bacteria which resist phagocytosis become extremely attractive to neutrophils and macrophages when coated with antibody

o o

Figure 12.8. Effect of opsonizing antibody and complement on rate of clearanceof virulent bacteria from the blood. The uncoated bacterra are phagocytosed rather s\owly (innate immunity) but, on coating with antibody, adherence to phagocytes is increased many-fold (acquired immunity) The adherence is less effective in animals temporarily depleted of complement This is a hypothetical but realistic situation; the natural proliferation of the bacteria has been ignored.

CHAPTER I 2 . : A D V E R S A R I ASLT R A T E G I EDSU R I N GI N F E C T I O N

Fca/F

ECEPTOR

265 |

FcT nrniprnn. ''""-"JJrCeerOn

Figure 12.9. Immunoglobulin and complement greatly increase the adherence of bacteria (and other antigens) to macrophages and neutrophils. Uncoated bacteria adhere to pattern recognition receptors (PRRs) such as various toll-like receptors (TLR) and the mannosebinding receptor The Fca/pR on macrophages binds to IgM (rl)-coated bacteria High affinity receptors for IgG Fc (O) and for C3b (CR1) and iC3b (CR3) (I) on the macrophage and neutrophil

surface considerablyenhance the strength ofbinding. The augmenting effect of complement is due to the fact that two adiacent IgG molecules can fix many C3b molecules, thereby increasing the number of links to 'bonus effect of multivalency'; p. 92) Bacteria the macrophage (cf opsonized with IgA (V) can adhere to the phagocyte via the Fcct/ltR already mentioned, or via FccrRI (CD89) which is Present on the s u r f a c eo f b o t h m a c r o p h a g e sa n d n e u t r o p h i l s

plement for opsonization which is mediated through specific receptors for immunoglobulin Fc and complement on the phagocyte surface (figure 72.9).It is clearly advantageous that the IgC subclasses which bind strongly to the IgG Fc receptors (e.g. igGl and IgG3 in the human) also fix complement well, it being appreciated that C3b bound to IgC is a very efficient opsonin because it engages two receptors simultaneously. Complexes containingC3b and C4b may show immune adherenceto the CR1 complement receptorson erythrocytesto provide aggregateswhich are transported to the liver and spleen for phagocytosis. Some elaboration on complement receptors may be pertinent at this stage.The CR1 receptor (CD35) for C3b is also present on neutrophils, eosinophils, monocytes, B-cellsand lymph node follicular dendritic cells (FDC). Together with the CR3 receptor (CD11b/CD18), it has the main responsibility for clearance of complexes containing C3. The CRl gene is linked in a cluster with C4b-binding protein and factor H, all of which subserve a regulatory function by binding to C3b or C4b to disassemblethe C3/C5 convertases,and act ascofactors for the proteolytic inactivation of C3b and C4b by factor I. CR2 receptors (CD21) for iC3b, C3dg and C3d are present on B-cells and FDC, and transduce accessory signals for B-cell activation especially in the germinal centers (cf. p. 200). They act as the receptor for Epstein-Barr virus (EBV), binding to the 9p350 major viral envelope glycoprotein and thereby facilitate entry of the virus into B-cells, with MHC class II molecules acting as a coreceptorbinding to viralgp42. CR3 receptors (CD11b/CD18 on neutrophils, eosinophils, monocytes and NK cells) bind iC3b, C3dg and C3d. They are related to LFA-1 and CR4 (CD11c/CD18,binds iC3b and C3dg) inbeing members

of the B, integrin subfamily (cf. table 7.2).CRs is found on neutrophils and platelets andbinds C3d and C3dg. A number of other complement receptors have been described including some with specificity for C1q, for C3a and C4a, and the CD88 molecule with specificity for CSa. Somefurther effects of complement Some strains of Gram-negative bacteria which have a lipoprotein outer wall resembling mammalian surface membranes in structure are suscePtible to the bactericidal action of fresh serum containing antibody. The antibody initiates the development of a comPlementmediated lesion which is said to allow accessof serum lysozyme to the inner peptidoglycan wall of the bacterium to cause eventual cell death. Activation of complement through union of antibody and bacterium will also generate the C3a and C5a anaphylatoxins leading to extensive transudation of serum components, including more antibody, and to the chemotactic attraction of neutrophils to aid in phagocytosis, as described earlier under the acute inflammation umbrella (cf. figures 2.18 and 12.3). The secretory immune systemprotects the external mucosal surfeces We have earlier emphasized the critical nature of the mucosal barriers where there is a potentially hostile interface with the microbial hordes. With an area of around 400 rn2, give or take a tennis court or two, the epithelium of the adult mucosaerepresentsthe most frequent portal of entry for common infectious agents, allergens and carcinogens. The need for wellmarshaled, highly effective mucosal immunity is glaringly obvious. The gut mucosal surfaces are defended by both

I zoo

NFEcTroN DEUsR T NrG c n n p T E Rr 2 - A D V E R S A R Ts A TR L ATEGT

antigen-specific and nonantigen-specific mechanisms. Among the nonspecific mechanisms, antimicrobial peptides are produced not only by neutrophils and macrophages but also by mucosal epithelium. As described in Chapter 1, the group of antimicrobial peptides called defensins lyse bacteria via disruption of their surface membranes. Specific immunity is provided by secretoryIgA and IgM, with IgAl predominating in the upper areasand IgA2 in the large bowel. Most other mucosal surfaces are also protected predominantly by IgA with the exception of the reproductive tract tissues of both male and female, where the dominant antibody isotype is IgG. The size of the task is highlightedbythe fact that 80%of the Ig-producingB-cells in the body are present in the secretory mucosae and exocrine glands. IgA antibodies afford protection in the external body fluids, tears, saliva, nasal secretionsand those bathing the surfaces of the intestine and lung by coatingbacteria and viruses and preventing their adherence to the epithelial cells of the mucous membranes, which is essentialfor viral infection and bacterial colonization. SecretoryIgAmolecules themselveshave very little innate adhesivenessfor epithelial cells, but high affinity Fc receptors for this Ig class are present on macrophages and neutrophils and can mediate phagocytosis (figure 72.10a). If an infectious agent succeedsin penetrating the IgA barrier, it comes up against the next line of defense of the secretory system (seep. 163) which is manned by IgE. Indeed, most serum IgE arises from plasma cells in mucosal tissues and their local draining lymph nodes. Although present in low concentration, IgE is firmly bound to the Fc receptorsof the mast cell (seep. 48) and contact with antigen leads to the release of mediators

which effectively recruit agents of the immune response and generatea local acute inflammatory reaction. Thus histamine, by increasing vascular permeability, causes the transudation of IgG and complement into the area, and while chemotactic factors for neutrophils needed to dispose the cells attract effector eosinophils of the infectious organism coated with specific IgG and C3b (figure 72.70b). Engagement of the Fcy and C3b receptorson local macrophagesby such complexeswill lead to secretion of factors which further reinforce these vascular permeability and chemotacticevents.Broadly, one would say that immune exclusion in the gut is noninflammatory, but immune elimination of organisms which penetrate the mucosa is proinflammatory. Where the opsonized organism is too large to be en gulf ed, pha gocytes can employ antibo dy - dependent cellular cytotoxicity (ADCC, p. 32) and there is evidence for its involvement in parasitic infections (see

p.27e). The mucosal tissues contain various T-cell populations, but their role and that of the mucosal epithelial cells, other than in a helper function for local antibody production, is of less relevance for the defense against extracellular bacteria. Somespecific bocteri0linfections First let us see how these considerations apply to defenseagainstinfectionby common organisms such as streptococciand staphylococci. The p-hemolytic stteptococci were classified by Lancefield according to their carbohydrate antigen, the most important for human disease belonging to group A. Streptococcuspyogenes most commonly causes acute pharyngitis (strep sore

)t\I ffi

lgc + C' BLOOD VESSEL Releose of vosooctive foclors & chemoloctic

PLASMA CELL Figure 12.10. Defense of the mucosal surfaces.(a) IgA opsonizes organisms and prevents adherenceto the mucosa. (b) IgE recruits agents of the immune response by firing the release of mediators from mast ce1ls

CHAPTER I 2 - A D V E R S A R I ASLT R A T E G I D U R I N GI N F E C T I O N ES throat) and the skin condition impetigo, but is also responsible for scarletfever and has emerged as a cause of the much rarer but often fatal toxic shock syndrome and of the always alarming necrotizing faciitis (flesh-eatingdisease).Rheumatic fever and glomerular nephritis sometimes occur as serious postinfection sequelae. The most important virulence factor is the surface Mprotein (variants of which form the basis of the Griffith typing). This protein is an acceptor for factor H which facilitates C3b breakdown, and binds fibrinogen and its fragmentswhichcover sitesthatmay act ascomplement activators. It thereby inhibits opsonization and the protection afforded by antibodies to the M-component is attributable to the striking increase in phagocytosis which they induce. The ability of group A streptococci to elicit cross-reactive autoantibodies which bind to cardiac myosin is implicated in poststreptococcal autoimmune disease. High titer antibodies to the streptolysin O exotoxin (ASO), which damages membranes, indicate recent streptococcal infection. The streptococcal pyrogenic exotoxins SPE-A, -C and -H, and the streptococcalrzitogenic exotoxin SMEZ-2, are utperantigens associatedwith scarlet fever and toxic shock syndrome. The toxins are neutralized by antibody and the erythematous intradermal reaction to i

Roitt's Essential Immunology, 11th Edition - PDF Free Download (2024)