Foot and Mouth Disease: Current Perspectives

Citation preview

Foot and Mouth Disease Current Perspectives

Copyright © 2004 By Horizon Bioscience

Foot and Mouth Disease Current Perspectives

Edited by

Francisco Sobrino Esteban Domingo Centro de BioJogia Molecular "Severo Ochoa" (CSIC-UAM), Cantoblano 28049, Madrid, Spain and Centro de Investigaci6n en Sanidad Animal (CISA-INIA), Valdeolmos 28130, Madrid, Spain

Copyright © 2004 By Horizon Bioscience

Copyright © 2004 Horizon Bioscience 32 Hewitts Lane Wymondham Norfolk NR18 OlA England

www.horizonbioscience.com

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 1-904933-00-9 Description or mention of instrumentation, software, or other products in this book does not imply endorsement by the author or publisher. The author and publisher do not assume responsibility for the validity of any products or procedures mentioned or described in this book or for the consequences of their use.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission ofthe publisher. No claim to original U.S. Government works. Printed and bound in Great Britain

Copyright © 2004 By Horizon Bioscience

This volume is dedicated to the memory of Dr Fred Brown, a pioneer of FMDV

research, who died on February 22,2004.

iii

Copyright © 2004 By Horizon Bioscience

Contents List of Contributors

vi

Preface

ix

Chapter 1 Stepping Stones in Foot-and-Mouth Research: A Personal View F. Brown Chapter 2 Genome Organisation, Translation and Replication of Foot-and-Mouth Disease Virus RNA Graham J. Belsham and Encarnacion Martinez-Salas

19

Chapter 3 Foot-and-Mouth Disease Virus Proteinases Martin D. Ryan, Michelle L.L. Donnelly, Mike Flint, Vanessa M. Cowton, Garry Luke, Lorraine E. Hughes, Caroline Knox and Pablo de Felipe

53

Chapter 4 Structure of Foot-and-Mouth Disease Virus Particles Cristina Ferrer-Dria and Ignacio Fita

77

Chapter 5 Clinical Signs of Foot-and-Mouth Disease Alex Donaldson

93

Chapter 6 Persistence of Foot-and-Mouth Disease Virus Jeremy Salt

103

Chapter 7 Molecular Aspects of Foot-and-Mouth Disease Virus Virulence and Host Range: Role of Host Cell Receptors and Viral Factors Barry Baxt and Elizabeth Rieder

145

Chapter 8 Immunology of Foot-and-Mouth Disease Kenneth C. McCullough and Francisco Sobrino

173

Chapter 9 Functional and Structural Aspects of the Interaction of Foot-and-Mouth Disease Virus with Antibodies Mauricio G. Mateu and Nuria Verdaguer

223

Chapter 10 Quasispecies Dynamics and Evolution of Foot-and-Mouth Disease Virus Esteban Domingo, Carmen M. Ruiz-Jarabo, Armando Arias, Juan Garcia-Arriaza and Cristina Escarmis

261

iv

Copyright © 2004 By Horizon Bioscience

Chapter 11 Modern Inactivated Foot-and-Mouth Disease (FMD) Vaccines: Historical Background and Key Elements in Production and Use S.J. Barteling

305

Chapter 12 Foot-and-Mouth Disease Virus Peptide Vaccines David J. Rowlands

335

Chapter 13 Mathematical Models of the Epidemiology and Control of Foot-and-Mouth Disease Mark E.J. Woolhouse

355

Chapter 14 Natural Habitats in Which Foot-and-Mouth Disease Viruse is Maintained Wilna Vosloo and Gavin R. Thomson

383

Chapter 15 Diagnosis and Control of Foot-and-Mouth Disease R.P.Kitching

411

Chapter 16 Control of Foot-and-Mouth Disease: Role of International Organizations J. Blancou, Y: Leforban, and J. E. Pearson

425

Chapter 17 Overview of Foot-and-Mouth Disease and its Impact as aRe-emergent Viral Infection Brian WJ. Mahy

437

Colour Plates

A.1

v Copyright © 2004 By Horizon Bioscience

Books of Related Interest Pathogenic Fungi: Host Interaction and Emerging Strategies for Control (2004) Edited by: GIOCONDA SAN-BLAS AND RICHARD A. CALDERONE Pathogenic Fungi: Structural Biology and Taxonomy (2004) Edited by: GIOCONDA SAN-BLAS AND RICHARD A. CALDERONE Malaria Parasites: Genomes and Molecular Biology (2004) Edited by: ANDREW P. WATERS AND CHRIS J. JANSE Peptide Nucleic Acids: Protocols and Applications, 2nd Edition (2004) Edited by: PETER E. NIELSEN Ebola and Marburg Viruses: Molecular and Cellular Biology (2004) Edited by: HANS-DIETER KLENK AND HEINZ FELDMANN Metabolic Engineering in the Post-genomic Era (2004) Edited by: BORIS N. KHOLODENKO AND HANS V. WESTERHOFF MRSA: Current Perspectives (2003) Edited by: AD C. FLUIT AND FRANZ-JOSEF SCHMITZ Tuberculosis: The Microbe Host Interface (2003) Edited by: LARRY S. SCHLESINGER AND Lucy DESJARDIN Genome Mapping and Sequencing (2003) Edited by: IAN DUNHAM Regulatory Networks in Prokaryotes (2003) Edited by: PETER DORRE AND BARBEL FRIEDRICH Bioremediation: A Critical Review (2003) Edited by: I.M. HEAD, I. SINGLETON AND M. MILNER Bioinformatics and Genomes: Current Perspectives (2003) Edited by: MIGUEL A. ANDRADE Frontiers in Computional Genomics (2003) Edited by MICHAEL Y. GALPERIN AND EUGENE V. KOONIN : Multiple Drug Resistant Bacteria: Emerging Strategies (2003) Edited by: CARLOS F. AMABILE-CUEVAS Vaccine Delivery Strategies (2003) Edited by: GUIDO DIETRICH AND W. GOEBEL Probiotics and Prebiotics: Where Are We Going? (2002) Edited by: GERALD W. TANNOCK Genomic Technologies: Present and Future (2002) Edited by: DAVID J. GALAS AND STEPHEN J. MCCORMACK Genomics of GC-Rich Gram Positive Bacteria (2002) Edited by: ANTOINE DANCHIN

Full details of these and all our books at:

www.horizonpress.com Copyright © 2004 By Horizon Bioscience

Contributors Armando Arias Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM) Universidad Aut6noma de Madrid Cantoblanco 28049 Madrid Spain

Pablo de Felipe Centre for Biomolecular Sciences School of Biology Biomolecular Sciences Building University of S1. Andrews North Haugh S1. Andrews KY16 9ST U.K

*S.J. Barteling Nieuwe Keizersgracht 438 1018 VG Amsterdam The Netherlands Email: [email protected]

*Esteban Domingo Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM) Universidad Aut6noma de Madrid Cantoblanco 28049 Madrid Spain Email: [email protected]

*Barry Baxt USDA,ARS Plum Island Animal Disease Center PO Box 848 Greenport NY 11944-0848 USA Email: [email protected]

and Centro de investigaci6n en Sanidad Animal (CISA-INIA) Valdeolmos 28130 Madrid Spain Email: [email protected]

*Graham J. Belsham BBSRC Institute for Animal Health Pirbright Woking Surrey GU240NF UK Email [email protected]

*Alex Donaldson 290 London Road Guildford Surrey GU4 7LB UK Email: [email protected]

*Jean Blancou 11, rue Descombes 75017 Paris France Email: jean.blancou@wanadooJr

Michelle L.L. Donnelly Marie Curie Research Institute The Chart Oxted Surrey RH8 OTL UK

*F. Brown Plum Island Animal Disease Research Center P.O. Box 848 Greenport

Cristina Escarmis Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM) Universidad Aut6noma de Madrid Cantoblanco 28049 Madrid Spain

11944 NY USA

Vanessa M. Cowton Dept. of Molecular and Cellular Pathology University of Dundee Ninewells Hospital and Medical School Dundee DD1 9SY U.K.

* indicates corresponding author

vii

Copyright © 2004 By Horizon Bioscience

Cristina Ferrer-Orta Institut de Biologia Molecular de Barcelona (CSIC) Jordi-Girona 18-26 08034 Barcelona Spain

Caroline Knox Centre for Biomolecular Sciences School of Biology Biomolecular Sciences Building University of St. Andrews North Haugh St. Andrews KY16 9ST U.K

*Ignacio Fita Institut de Biologia Molecular de Barcelona (CSIC) Jordi-Girona 18-26 08034 Barcelona Spain Email: [email protected]

Yves Leforban Conseil General Veterinaire 251 rue de Vaugirard 75 732 Paris Cedex 15 France Email: [email protected]

Mike Flint Center for the Study of Hepatitis C Laboratory of Virology and Infectious Disease The Rockefeller University New York New York 10021 USA Juan Garcia-Arriaza Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM) Universidad Autonoma de Madrid Cantoblanco 28049 Madrid Spain Lorraine E. Hughes Centre for Biomolecular Sciences School of Biology Biomolecular Sciences Building University of St. Andrews North Haugh St. Andrews KY16 9ST U.K *R.P. Kitching National Centre for Foreign Animal Disease Canadian Food Inspection Agency 1015 Arlington Street Winnipeg MB R3E 3M4 Canada Email: [email protected]

Garry Luke Centre for Biomolecular Sciences School of Biology Biomolecular Sciences Building University of St. Andrews North Haugh St. Andrews KY16 9ST U.K *Brian W. J. Mahy National Centre For Infectious Disease 1600 Clifton Road N E Mailstop C12 Atlanta GA30333 USA Email: bxm1 @cdc.gov *Mauricio G. Mateu Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM) Universidad Autonoma de Madrid Cantoblanco 28049 Madrid Spain Email: [email protected] Encarnacion Martinez-Salas Centro de Biologia Molecular "Severo Ochoa" CSIC-UAM Cantoblanco 28049 Madrid Spain Email: [email protected]

viii Copyright © 2004 By Horizon Bioscience

Kenneth C. McCullough Immunology Department Institute of Virology and Immunoprophylaxis Mittelhausern Switzerland James E. Pearson 4016 Phoenix St.Ames Iowa 50014 USA Email: [email protected] Elizabeth Rieder United States Department of Agriculture Agricultural Research Service Plum Island Animal Disease Center Greenport NY 11944-0848 USA Email: [email protected]

*Jeremy Salt Pfizer Animal Health Ramsgate Road Sandwich Kent CT3 1DY UK Email: [email protected] *Francisco Sobrino Centro de Biologia Molecular "Severo Ochoa" (CSIC-UAM) Cantoblano 28049 Madrid Spain Email: [email protected] and Centro de investigaci6n en Sanidad Animal (CISA-INIA) Valdeolmos 28130 Madrid Spain

*Martin D. Ryan Centre for Biomolecular Sciences School of Biology Biomolecular Sciences Building University of St. Andrews North Haugh St. Andrews KY16 9ST U.K Email: [email protected] *David J. Rowlands Division of Microbiology School of Biochemistry and Molecular Biology University of Leeds Leeds LS2 9JT UK Email: [email protected] Carmen M. Ruiz-Jarabo Department of Microbiology and Immunology University of California San Francisco 513 Parnassus Avenue San Francisco California 94143-0414 USA

Gavin R. Thomson AU - Inter-African Bureau for Animal Resources POBox 30786 Nairobi Kenya Nuria Verdaguer Institut de Biologia Molecular de Barcelona (CSIC) Jordi Girona 18-26 08034 Barcelona Spain Email: [email protected] *Wilna Vosloo ARC -Onderstepoort Veterinary Institute Private Bag X5 Onderstepoort 0110 South Africa Email: [email protected] *Mark E.J. Woolhouse Centre for Infectious Diseases College of Med and Veterinary Med University of Edinburgh Easter Bush Midlothian EH25 9RG U.K. Email: [email protected] ix

Copyright © 2004 By Horizon Bioscience

Preface A book on foot-and-mouth disease, no matter how flooded the literature with review articles, requires no justification. If anybody still had doubts, the shock produced by the 2001 UK epizootic associated with an Asian virus, should dispel them. This first XXI st century outbreak in Europe will rank with the famous 1946-1954 Mexican or 1965-1966 European epizootics. With an important difference: the public perception of the disease today is very different from the perception decades ago. The world is becoming increasingly "global" and it is easy nowadays to establish a pathetic contrast between the TV images of massive slaughtering of farm animals with the mounting hunger in many deprived areas of this planet. And this should motivate experts to free of FMD countries in need of economic development, to eliminate trade restrictions that at present represent an unsurmountable burden. And there is only one reliable action to this goal: scientific progress. This book is aimed at exposing the reader to the major historic and current developments in FMD research. This includes basic research on molecular biology and epidemiology, disease manifestations, diagnostic procedures, economic and environmental impact for agriculture and wildlife, and alternative control measures, with the role of international organizations. We are deeply indebted to all authors. Unfortunately, for well justified reasons, not all experts could accept our invitation to contribute. The observant reader will soon identify salient absences and, as a consequence, some rather lengthy chapters. We apologize for uneven emphases and disequilibria in the distribution of contents. We made a point of letting each author to decide on coauthors and on chapter length. We have not worried about possible redundancies since, if they do not exceed a certain limit, they can have pedagogic value by reflecting different points of view on a topic. We have limited the participation of each author to a single chapter, to extend the input to the book to as many experts as possible. Obviously, research on FMD and FMDV has involved many more people than represented in the book, and most contributions, we believe, are recognized in the extensive bibliography which amounts to more than a thousand references. We are all indebted to the many experts who have contributed to the understanding of FMD, and are paving the way to its effective control. We would like the potential reader (an expert or a student considering a career in infectious disease) to be aware that current FMD research represents a window through which major challenges in infectious diseases of humans, animals and plants are viewed. Such challenges include disease emergence and reemergence, the recognition of the complexity of microbial populations and genetic heterogeneity (quasispecies dynamics) as an underlying difficulty for disease control, the need to understand basic multiplication mechanisms (RNA replication, viral protein synthesis, assembly of particles, etc.), the architecture and interactions that affect the stability of virus particles, the antiviral immune response of a variety of host species, viral pathogenesis, and the design of preventive (vaccination or nonvaccination policies) and control measures. These constitute interlinked sets of

Copyright © 2004 By Horizon Bioscience

relevant questions that apply to many disease agents. Hopefully, the book will offer a double perspective of what is interesting for FMD and for infectious diseases generally. It is the integration of knowledge form seemingly unrelated sources that can help in the development of more effective, rational and socially acceptable FMD control measures. The preparation of the book has been possible thanks to the continuous interaction with Annette Griffin of Horizon Press, and the editorial help of Ms. Lucia Horrillo at Centro de Biologia Molecular "Severo Ochoa". We wish to express our gratitude to Instituto Nacional de Investigaci6n Agraria y Alimentaria (INIA, Spain) for financial support that has permitted publication of the color plates that illustrate several chapters. We hope that a broad readership will find valuable elements in this book. While this book was in press we received the sad news that Fred Brown, a pioneer of FMDV research, died on February 22, 2004. Fred devoted his entire life to FMD, was extremely influential, and all the scientific community and the authors of different chapters of this book owe a lot to him. Fred contributed the first chapter of this volume, but he could neither correct the galleys nor see the finished book. The editors respectfully have kept Fred's chapter exactly as he wrote it, and dedicate this volume to his memory. Francisco Sobrino, and Esteban Domingo Madrid, April, 2004

xi

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrina and Esteban Domingo

Chapter 1 Stepping Stones in Foot-and-Mouth Research: A Personal View

F. Brown Abstract The history of foot-and-mouth disease falls into several distinct areas. 1. Loeffler and Frosch's landmark description in 1897 that the disease is caused by a filterable agent, the first observation that an animal disease could be caused by a virus. 2. The search for experimental laboratory animals, culminating in the demonstration by Waldmann and Pape of the susceptibility of the guinea pig in 1920 and the suckling mouse by Skinner in 1951. 3. The discovery of three distinct serotypes 0, A and C in the 1920s by Vallee and Carre in France and by Waldmann in Germany, and the subsequent recognition in the 1940s and 1950s by the Pirbright group of the three Southern African Territory Types SAT 1-3, and Asia 1.4. The development of in vitro techniques for the growth of the virus which have been crucial for the largescale production of vaccines and for the accurate assay of virus infectivity. Work by Hecke and the Maitlands in the early 1930s, followed by the crucial demonstration by Frenkel in 1947 that large amounts of the virus could be produced in surviving tongue epithelium, formed the basis for the vaccination programmes initiated in Europe in the 1950s. The subsequent development of cell lines has brought a remarkable degree of sophistication to the study of virus growth. 5. The impact of molecular studies on the structure of the virus and its mode of replication which have led to practical applications such as an in vitro test for vaccine potency, rapid diagnosis methods, and international epidemiological surveys. In addition, they have provided the means to design molecular vaccines. 1. Introduction The massive outbreaks of foot-and-mouth disease (FMD) in Taiwan in 1997 and more recently in the United Kingdom in 2001 provided stark reminders that, despite all the knowledge we have accumulated during the 100 or so years since Loeffler and Frosch (1897) discovered that the causal agent is a virus, even well organized veterinary and scientific services are hard pressed or even helpless when the virus is on the march. In this chapter I will outline the advances made in our knowledge of the disease and its virus since 1897 but initially I will describe why FMD is important, not only to the immediate farming community but also to world trade and even industries seemingly unconnected with an outbreak. As long ago as the 16th century the importance of the disease was recognised, if indeed the description by Fracastorius (1546) was that of FMD. Most certainly, Copyright © 2004 By Horizon Bioscience

12

Stepping Stones in FMD Research

towards the end of the 19th century its devastating effects in Europe, particularly in Germany, led the Government there to invite scientists to study the disease and its cause. Germany was not the only country in mainland Europe with the problem and the disease was also present in the United Kingdom in 1839. In addition, there are reports of the disease in Asia in 1842, South America (1871) and Africa (1892). Clearly there can be no certainty that these occurrences were FMD but the descriptions are sufficiently similar to what is seen today that there is little doubt about their identity. It is generally regarded that FMD is the most contagious of all diseases of farm animals. Cattle, pigs, sheep and goats are all susceptible. Typically, when an animal becomes infected, most members of the herd or flock will become infected. Clinical signs vary considerably in different species. In domestic cattle, following a short incubation period of two to eight days, the disease is characterized by an initial period of pyrexia, depression and anorexia. The vesicles which develop on the dorsal surface of the tongue, dental pad, lips, buccal mucosa and muzzle vary in size but in severe cases may involve the greater part of the tongue. The lesions normally rupture within 24h, releasing vesicular fluid containing up to 108 infectious units per ml. The epithelium is shed, salivation is profuse, and saliva frequently hangs from the muzzle. Subsequent to or concurrent with the mouth lesions, vesicles appear on the feet, particularly in the clefts or on the coronary bands, causing pain and lameness. Vesicles may also occur on the teats and udder, sometimes leading to bacterial mastitis. Loss of productivity is usually estimated to be about 25%. Although the disease is not usually fatal for adult animals, young animals often die, as found with pigs in Taiwan in 1997. Ironically, with most strains of the virus it is unusual for sheep to show clearly overt signs of the disease, but they proved to be effective transmitters of the disease in the United Kingdom in 2001. The need to stop movement of animals during an outbreak leads to considerable financial losses in the farming community and the size and widespread nature of the U.K. 2001 outbreak led to much greater financial losses to the tourist industry. The disease also occurs in many wildlife species which makes its control in several countries problematic. Of particular importance in this connection is the susceptibility of the African buffalo in which infection normally occurs in the absence of observable clinical signs. Hedger (1981) is of the opinion that, although many hosts are susceptible under experimental conditions (e.g. by inoculation of the virus or close contact with infected animals), under natural conditions these hosts would not represent a danger to domestic stock. Even the suggestion by Capel-Edwards (1971) that brown rats (Ratus norvegicus), because of their close association with farm animals and their ability to migrate over considerable distances, could playa significant role in the spread of the disease, was discounted by Hugh-Jones in 1970 on the basis of a study of the extensive 1967-69 epizootic in England. Transmission of the disease is primarily from the infected animal itself, especially during the early febrile stage when virus is present in the organs, tissues and body fluids. Affected animals shed virus in vesicular epithelium and fluid, saliva, milk, faeces, urine, semen and vaginal secretions. Moreover, virus is excreted before the development of clinical signs.

Copyright © 2004 By Horizon Bioscience

Brown

31

2. Landmarks in the History of the Disease The landmark papers on the disease were published by Loeffler and Frosch in 1897 when they showed that it was caused by a filterable agent and that serum from convalescent animals "neutralized" it. Clearly influenced by the brilliant pioneering studies of Pasteur and Koch and their colleagues, the primary objective of the work on FMD following the discovery of its aetiology was to develop a vaccine. It is a sobering thought that, despite this recognition, it was not until 1952 that comprehensive vaccination against the disease was first undertaken. The initiation of vaccination of cattle in Holland, France and Germany was an immediate success and it is somewhat bewildering to the author that the policy was stopped in 1991, in his opinion for spurious reasons. I enlarge on these reasons below. The findings which I regard as the subsequent landmarks in the history of the disease are arranged chronologically in Table 1. Although most are self explanatory, a small section of text summarizing their importance is provided for each landmark. To avoid unnecessary repetition of the more extensive descriptions provided by other authors which appear elsewhere in the book, I have kept these remarks brief. 2.1. Identification of the Causal Agent 1897

This observation was not only important in FMD history but also because it was the first example of an animal disease caused by a virus. All the early work on the disease stemmed from the observation that sera from convalescent animals "neutralized" the agent by protecting them from infection. 2.2. Laboratory Models for the Disease 1920 and 1951

The demonstration that the guinea pig could be infected experimentally was particularly important because the lesions obtained resembled those found in the naturally susceptible hosts, Le. blisters on the feet and tongue, and loss of condition. Moreover, they provided the means to test experimental vaccines and to produce hyperimmune sera for serological tests. The subsequent discovery by Skinner in 1951 that the suckling mouse could be infected by inoculation provided a cheap laboratory animal which was widely used before the introduction of tissue culture methods, because it is generally as sensitive as cattle for titration of the virus. Even no\v it is used in some laboratories for vaccine potency assays. It was also used extensively by Skinner in attempts to produce an attenuated vaccine. 2.3. Demonstration of Antigenic Variation 1922-1954

This finding is of crucial importance in studies on the epidemiology of the disease and its control by vaccination. The initial discovery of three serotypes 0, A and C in the 1920s was followed by the observation that there was significant variation within these serotypes which led to problems in the major outbreak in Mexico in 1946-1954. The excellent experimental type A vaccine produced from a virus isolate in a 1932 outbreak of the disease in the U.K. was ineffective against the serotype A virus causing the outbreak in Mexico (Galloway et al., 1948). Similarly, the serotype 0 vaccines being used on a large scale in Western Europe in the 1950s proved ineffective in the face of a serotype 0 virus which was introduced from

Copyright © 2004 By Horizon Bioscience

\4

Stepping Stones in FMD Research

Table 1. Landmarks in the history of foot-and-mouth disease Year

Achievement

1546

Description of the disease (Fracastorius).

1897

Identification of the causal agent (Loeffler and Frosch); first example of an animal disease caused by a virus.

1920

Demonstration that guinea pigs are susceptible to the disease; the first small experimental animal to be used for FMD work (Waldmann and Pape).

1922-6

Recognition of three distinct serotypes 0, A and C, of the virus (Vallee and Carre; Waldmann and Trautwein).

1925

First vaccination against the disease (Vallee et al.)

1927

Recognition of antigenic variation within serotypes (Bedson; Waldmann and Trautwein).

1930

Growth of virus outside animal body (Heeke; Maitland and Maitland).

1931

First estimates of size of virus (Galloway and Elford).

1947

Growth of virus on large scale in tongue tissue fragments, making mass vaccination possible (Frenkel).

1948

Recognition of three serotypes from Southern Africa distinct from 0, A and C (Brooksby).

1948

Vaccination of 2 x 106 animals in Argentina (Rosenbusch et al.).

1951

Use of suckling mouse as experimental animal (Skinner).

1952

Mass vaccination using in vitro grown virus started in Holland.

1954

Recognition of 7 th serotype, Asia 1 (Brooksby).

1958

Observation of virus in electron microscope (Bachrach and Breese; Bradish et al.).

1958

Role of 146 S particle in immunization (Brown and Crick). Isolation of infectious ribonucleic acid from the virus (Brown et al.).

1959

Isolation of virus from carrier animals (van Bekkum et al.).

1962-4

Growth of virus in a cell line (Mowat and Chapman; Capstick et al.; de Castro).

1963

Purification of the virus particle (Brown and Cartwright; Bachrach et al.).

1965

Demonstration of recombination between virus strains (Pringle).

1966

Viral polymerase recognised by serological methods (Cowan and Graves).

1969

Protein composition of virus determined (Wild et al.; Burroughs et al.).

1969

Importance of single protein in immune response recognised (Wild et al.).

1973

Demonstration that a single isolated protein will evoke neutralizing antibody (Laporte et al.).

1977

Biochemical mapping of the virus genome (Sangar et al.).

1978

Viral RNA polymerase within virus particle (Denoya et al.).

1981

Expression in E. coli cells of the immunogenic protein (Kleid et al.)

1982

Chemical synthesis of a peptide which evokes protective response (Bittle et al.; Pfaff et al.).

Copyright © 2004 By Horizon Bioscience

Brown

51

Table 1, Continued

1984

First crystals of the virus particle (Fox et al.)

1986

First complete nucleotide sequence of the virus genome (Clarke and Carroll).

1989

Structure of virus at 2.9A resolution (Acharya et al.).

1990

Infectious cDNA (Zibert et al.).

19962003

Identification of receptors on susceptible cells (Jackson et al.; Duque and Baxt).

South America. These observations were forerunners of the quasispecies theory which has been studied extensively by Domingo et ale (2003) (Chapter 10). The three South African Territories serotypes SAT 1-3 described by the Pirbright laboratory in 1948 and the ASIA 1 isolate unique to that continent emphasised the considerable diversity of the virus and poses interesting questions regarding its evolution. With the application of nucleic acid sequencing it has become possible to follow precisely the movement of individual virus strains between different countries. This is illustrated in the movement of the serotype 0 virus which caused the outbreak in the United Kingdom in 2001. Such information should help considerably in international control of the disease. 2.4. Development of a Vaccine Against FMD 1925-1948

Once it had been established that FMD was caused by a virus, the major research objective was to follow the lead provided by Pasteur and Koch in their brilliant pioneering vaccination studies with anthrax, rabies, tetanus and diphtheria. The absence of a laboratory model meant that progress was slow and the earliest report on the development of a vaccine was not published until 1925, when Vallee, Carre and Rinjard showed that epithelial tissue from calves infected with the virus and treated with formaldehyde provided protection against challenge with the virus. This approach was adopted by Waldmann's group in Germany, but incorporated aluminium hydroxide in the inactivation procedure to function as an adjuvant. This method was applied successfully by Rosenbusch et ale (1948) to the vaccination of two million cattle in Argentina against the three serotypes 0, A and C and about 100,000 sheep but did not give good results in pigs. The history of the development of the classical vaccines has been reviewed recently (Sutmoller and Barteling, 2003). In view of the problems which arose later regarding incomplete inactivation of the virus by formaldehyde, it is interesting to note that the preparations were made in glycocoll buffer, which is alkaline. This issue is discussed below but it has been shown that the virus itself is inactivated more quickly in alkaline conditions than at neutral pH because of the enhanced activity of the ribonuclease present in the particle. Despite the promise of the vaccines prepared by this method it was the demonstration by Frenkel in 1947 that the virus could be grown in surviving tongue epithelial cells on the scale required for large-scale vaccination programmes to be mounted which enabled Holland, France and Germany to undertake such programmes. The effect was dramatic (Figure 1) and deserving as

Copyright © 2004 By Horizon Bioscience

16

Stepping Stones in FMD Research

much recognition as was directed towards the results of the introduction of the Salk inactivated poliovaccine in the mid-1950s. The advantages of either monolayers or suspensions derived from a cell line have resulted in the replacement of tongue epithelium by pig kidney or baby hamster kidney cells (Mowat and Chapman, 1962; Capstick et ai., 1962, de Castro, 1964). This technology is now capable of producing virus in quantities sufficient to furnish hundreds of millions of doses of vaccine each year. At the personal level, my own introduction to FMD vaccination came in the mid-1950s and the unease regarding the failure of formaldehyde to inactivate the virus completely. This uncertainty had also been found with the Salk poliovaccine in the much-publicized Cutter incident. The suspicion, first expressed by Moosbrugger in 1948, that treatment of FMDV with formaldehyde did not give an innocuous product, was vindicated by Wesslen and Dinter (1957), Graves (1963) and Brown (1963). Despite this evidence, formaldehyde continued to be used by some manufacturers until an outbreak of the disease in France in 1981 was shown to be related to an improperly inactivated vaccine (King et ai. 1981). More extensive molecular analysis of virus isolates from other outbreaks in Western Europe during the 1970s and 1980s provided more critical evidence of the unreliability of formaldehyde as an inactivant for FMDV (Beck and Strohmaier 1987). The suggestion by E. Weston Hurst of Imperial Chemical Industries that Nacetyl ethyleneimine might prove superior was correct and the reagent was adopted by the Wellcome Laboratories at Pirbright for their vaccine production in the 1960s. This reagent (Brown and Crick, 1958) and the parent imine (Bahnemann, 1975) have been used exclusively for FMD vaccine production since the 1980s. The product is innocuous and is highly immunogenic. Surprisingly, the chemical reaction involved in the inactivation of viruses by imines was not studied until recently. However, work by Broo et ale (2001) provides good evidence that the reaction with several RNA viruses is overwhelmingly between the imine and the RNA with only slight modification of the virus proteins. The topic was reviewed recently by the author (Brown, 2002). In the section on molecular studies, the components present in virus harvests are described in detail. Early fractionation studies by Brown and Crick showed that the infectious particle, suitably inactivated, elicited good levels of protective neutralizing antibody whereas the fraction which was not deposited in the ultracentrifuge did not. Subsequent work showed that the so-called empty particles also elicited protective levels of neutralizing antibody. However, antibodies against the viral RNA polymerase were not protective. An observation which led to the identification of a major immunogenic site on the virus particle was the effect of trypsin on the infectivity and immunogenicity of a virus of serotype O. Both the infectivity and immunogenicity were reduced to about 1% of the untreated particle. This provided important clues about the proteins evoking the immune response because it was found that only one of the four structural proteins, VP1, had been cleaved by the enzyme.. This observation opened the way to the possibility of making a vaccine which consisted of only part of the virus particle and did not involve the use of infectious material. But such an approach had to await the characterisation of the virus particle and the other components of virus harvests (see Section 4). Copyright © 2004 By Horizon Bioscience

Brown

71

105 1()4

10 3 102

10

0 VI

ra

105

..c .....

104

0

103

...llil!

~

::J

~

0

OJ

102

E ;:,

10

.c

z

0

105

Federal Republic of Germany

104 103 102 10

0

1950

1980

Figure 1. Effect of mass vaccination on the number of outbreaks of foot-and-mouth disease in three countries in western Europe. Arrows indicate the start of mass vaccination.

2.5. Attenuated Vaccines

The development of the highly successful attenuated vaccines for yellow fever in the 1920s and poliomyelitis in the 1950s led to many attempts to produce a similar product for FMD. In general, the strains were obtained empirically by passing the virus serially in foreign hosts such as one-day-old chicks, embryonated chicken eggs, baby rabbits or mice. The adapted viruses were then tested in cattle at regular passage levels. When the pathogenicity for cattle had been reduced to the point at which macroscopic lesions did not occur and the animals were resistant to challenge, they were tested on a larger scale. Unfortunately, at this point it became clear that when larger numbers of animals were used, more were seen to have characteristics typical of the disease. Moreover, species other than cattle became infected and showed typical FMD. Consequently, such vaccines were regarded as too risky and were never used on an extensive scale (Sutmoller and Barteling, 2003; Sutmoller et al. 2003). It is interesting to speculate whether our knowlegde of the virus genome will eventually allow us to identify those factors responsible for pathogenicity and

Copyright © 2004 By Horizon Bioscience

\8

Stepping Stones in FMD Research

to manipulate the genome accordingly. So far, there is little evidence to suggest that we have the first clues on this subject. The empirical approach which was so successful with yellow fever and poliomyelitis can be attributed to sheer luck. Nevertheless, the way in which FMDV replicates has been studied extensively and has been reviewed in great detail recently (Mason et al. 2003) so it is not proposed to describe it here. Such studies may eventually provide us with the knowledge we require to produce attenuated strains in a logical manner.

3. Molecular Biology 1931 - Present 3.1. Early Studies 1931-1935

If we accept that the term molecular biology includes the structure of the virus, it can be fairly stated that the first molecular report on FMDV was that by Elford and Galloway (1931) on the size of the infective particle. Using Gradocol membrane filtration, these authors estimated the size of the unit to be in the range of 8-12 nm. That this was two to three times lower than the actual size was due to the application of the incorrect factor for the relation between particle size and pore diameter. Subsequent work by the same group using ultracentrifugation gave a value of 20nm (Elford and Galloway, 1937; Schlesinger and Galloway, 1937). Shortly afterwards Traub and Pyl (1943) made an observation which was to be of considerable significance in understanding the structural relationships between the specific components in virus harvests when they showed that prolonged centrifugation which deposited 99% of the infectivity left more than 50% of the complement fixing activity in the supernatant. Subsequent fractionation studies using the more sophisticated sucrose gradient centrifugation method have shown that virus harvests contain at least four components (Table 2). 3.2. Characteristics of Virus Components 1955 - Present

It was in the mid 1950s that more attention was given to the molecular biology of FMDV. The purification and crystallisation of poliovirus (Schaffer and Schwerdt, 1955) and the demonstration that tobacco mosaic virus contained infectious RNA (Gierer and Schramm, 1956) were crucial observations that were made as I started work at the Pirbright laboratory, and the Plum Island Animal Disease Center in the U.S.A. was opened for work on FMDV. Electron microscopy studies by Bachrach in the U.S.A. and Bradish in the U.K., and their colleagues, showed that the virus was spherical and of the same size as poliovirus and, soon after the Wistar Institute group had shown that poliovirus contained infectious RNA, Brown and his colleagues at Pirbright and Strohmaier at the Tubingen laboratory in Germany made similar observations with FMDV. It is interesting and amusing to recall that some of the classical virologists at Pirbright were happy to tell me that the infectivity of the RNA was due to residual virus, despite the vulnerability of the extracted RNA to ribonuclease, whereas the virus is unaffected by the enzyme. At this time (March 1956) the Ciba Foundation organised a symposium on The Nature of Viruses, with most of the leading figures in virology present and they pointed the way forward, both orally and in the classic publication which came from the meeting. The development of cell cultures for growing the virus allowed it to be grown in defined media so that it became possible to obtain the different components labelled in their nucleic acid and protein moieties and made their analysis feasible. Copyright © 2004 By Horizon Bioscience

Brown

91

Table 2. Antigenic components of foot-and-mouth disease virus Component

Sedimentation Constant

Composition

Particle Virus

146S

I molecule ssRNA (M r =2.6 x 106). 60 copies of each ofVPI-3 (Mrc. 24 x 103) and VP4 (M r c. 8 x 103)

Empty particle

75S

60 copies each ofVPO a, VPI and VP3

Protein subunit

12S

Pentamer ofVPI-3

Virus infectionassociated antigen

3.8S

RNA polymerase (M r c. 56 x 103)

A VPO comprises VP4 and VP2 covalently linked

Purification of the virus particle was achieved in 1963 (Brown and Cartwright, 1963) and its characterisation and relationship with the other components present in virus harvests. The particle was found to consist of one molecule of ssRNA sedimenting at 35S, 60 copies of three structural proteins of mol. wt. 24 x 103 (VP1, VP2 and VP3) and 60 copies of VP4 mol. wt. 8 x 103 and, significantly, one copy of a protein of mol. wt. 53 x 103 (Sangar et al., 1976, Denoya et al., 1978; Table 2). Subsequent X-ray studies of the virus particle provided an explanation for many of its properties, especially when these were considered in conjunction with the amino acid sequence of the constituent proteins. Interestingly, the RNA of FMDV contains a polycytidylic acid sequence located near the 5' end. A similar sequence occurs in encephalomyocarditis virus but, to my knowledge, in no other RNA. Its role, if any, has not been described. It had been known for several years that the infectivity of the virus and the virus particle itself were lost rapidly by heating at 56°C or by reducing the pH below 7 (Randrup, 1954). The availability of the radioactively labelled purified virus particles allowed the transformation to be analysed in detail. This analysis showed that the four structural proteins comprising the virus capsid were transformed into a 12S particle, indistinguishable from that occurring naturally in virus harvests and consisting of a pentamer of VP1, VP2 and VP3. An aggregate of VP4 accounted for the fourth structural protein and the RNA was present as the infectious entity. Hidden among this mixture was one copy of the protein which turned out to be the viral RNA polymerase attached loosely to the RNA (Newman and Brown, 1997). This component was discovered by Cowan and Graves (1966) by agar gel analysis of virus harvests produced in BHK 21 cells. The component was distinct from the two precipitates produced by the 146S virus particle and 12S particle in that it was not serotype-specific and the authors suggested, correctly, that it might be an enzyme involved in virus replication. The copy that is incorporated in virus particles produces low levels of antibody in animals given repeated doses of vaccine; thus the detection of the antibody does not, as the authors had hoped, provide categorical evidence of previous infection. The remarkable lability of the virus particle just below pH? was an intriguing problem for many years but the information emerging from the X-ray analysis

Copyright © 2004 By Horizon Bioscience

\10

Stepping Stones in FMD Research

appears to offer an acceptable explanation. Lowering the pH disrupts contacts between neighbouring pentamers. Acharya et ale (1989), in their paper on the structure of the virus, drew attention to the high density of histidine residues on the interface between VP2 and VP3. Since the particle dissociates at pH6.8, the pKa of histidine, this led them to propose that the dissociation is triggered by electrostatic repulsion between the protonated imidazole side chains. Twomey et ale (1995) identified two of these histidines, VP3-141 and VP3-144, as potential destabilisers because they are the only two in a location such that the number of histidine, lysine or arginine residues greatly exceeds the number of negatively charged aspartic and glutamic acid residues within lOA on the opposite side of the pentamer-pentamer interface. Moreover, they are conserved among all the serotypes of the virus. 3.3. Application to Field Problems The statement somewhat reluctantly accepted initially, that the chemistry of nucleic acids and proteins would have much to offer to practical problems in the field, has been amply justified during the past few years by its application to several issues. In 1976 it was shown that T1 fingerprinting of the viral RNAs allowed the identity of a virus to be established much more precisely than by serology (Frisby et al., 1976). This method pinpointed the source of the outbreak in France in 1981 (King et al., 1981) and nucleic acid sequencing established without doubt that incompletely inactivated vaccines were the source of some other outbreaks in Europe (Beck and Strohmaier, 1987). With the introduction of nucleic acid sequencing, movement of a virus throughout the world can be followed precisely, clearly demonstrated by the recent outbreaks in Japan, S. Korea, S. Africa and Western Europe. Information about the genetic make-up of the virus, which clearly showed that the 3D protein is highly conserved among all the serotypes, has allowed a virus-specific rapid diagnostic test (real time RT-PCR) to be developed. Such a test, which can be applied at the farm gate (Callahan et al., 2002), would have considerable advantages because it would avoid the need to transport samples from animals suspected of having the disease to a laboratory. Indeed, it is an advantage to work in an environment which is FMDV-free, to avoid adventitious contamination, which is all too easily encountered. The need to identify animals that have been infected with FMDV, and hence become potential carriers of the virus, from those that have been vaccinated but not infected is one of the problems in international trade. Compared with vaccinated animals, the sera of convalescent animals contain antibodies to some non-structural proteins. These proteins have been identified and several tests based on these observations have been published and await validation. In addition, chemically synthesised peptides corresponding to these proteins have been used successfully for the same purpose.

4. The Rocky Road Towards New Vaccines The observation that treatment of a serotype 0 virus with trypsin reduced its immunogenic activity to about I % of that of the untreated virus (Wild et ale 1969) and that only VP1 of the four capsid proteins was hydrolysed opened the way to the possibility of making a vaccine which did not involve the use of the infectious particle. This evidence was buttressed by the demonstration by Laporte et ale

Copyright © 2004 By Horizon Bioscience

Brown

11\

(1973) that the separated VPl was the only protein which elicited the production of neutralizing antibodies. The sequencing of the genome and the determination of the sequence coding for VP1 enabled the protein to be expressed, first in E. coli cells and subsequently in other expression systems. In no instance, however, was a good immunogen produced. In fact, this result was not unexpected because the 12S particle, obtained under the mildest of conditions, was also a poor immunogen although it possessed intact VP1. Clearly configuration of the protein is an important issue. A much better neutralizing antibody response was obtained when the entire capsid protein region \vas expressed. Empty viral capsids are present in virus harvests and they contain all the capsid proteins. They induce an immune response that is as good as that obtained with inactivated virus particles. Several expression systems have been used including E.coli, vaccinia virus and baculovirus. The most recent is the replication defective human adenovirus type 5 (Grubman et al., 2003). Since adenoviruses target the upper respiratory and gastrointestinal tracts, thereby inducing local mucosal immune responses, they could have considerable advantages over other expression systems. In an extension of this work, the same group have inserted the gene coding for interferon type 1. Swine inoculated with Ad5-IFN were protected from challenge with virus one day after the inoculation. Thus by combining the two inocula it should be possible to obtain the early protection before the specific antibody response is produced. 4.1. The Peptide Approach The evidence that VPl of FMDV possesses significant immunogenic activity has led to considerable work on the active sites of this protein. With the emphasis on protection and consequently on neutralization tests, most of the work has focused on those sites which are concerned with the stimulation of neutralizing antibody. Several approaches have been used to identify the antigenic sites on the protein. In pioneering work, Strohmaier and his colleagues (1982) analysed fragments of the protein obtained by proteolytic enzyme or cyanogen bromide cleavage and concluded that the immunogenic activity would be confined to two regions encompassing amino acids 146-154 and 200-213. In a highly variable virus such as FMDV, it would be expected that antigenic differences would correlate with amino acid sequence differences. The occurrence of seven serotypes of the virus and the multiplicity of antigenic variants within these serotypes meant that ample material ,vas available to test the correctness of this postulated correlation. Studies of the growing number of sequences of VPl which have been available have shown that there is about 98% conservation of the amino acid sequences of VP4, 90% of VP3 and VP2 and 80% of VP1. More particularly, sequences of variants within a serotype are more conserved. Of particular importance was the observation that sequence variability occurs mainly at three sites in VPl, namely in the regions 42-61, 131-160 and 193-204. Taken together with the evidence that proteolytic enzymes cleave intact virus particles at the latter two regions, it seemed likely that the major immunogenic site would be located in those two regions and this supposition has been amply confirmed (Bittle et aI., 1982).

Copyright © 2004 By Horizon Bioscience

\12

Stepping Stones in FMD Research

In another approach the 'Pepscan' method described by Geysen et ale (1984) identified amino acid sequence 146-152 as a potential immunogenic site. In this method hexapeptides corresponding to overlapping sequences of the entire amino acid sequence ofVP1 were allowed to react with the type-specific antibody and the reactive regions of the protein detected with anti-species antiserum. Taken together with the other two approaches, amino acid sequence 141-160 appeared to contain the most immunogenic site and in general work has concentrated on this region of VP1 (Chapter 9). In a study of four antigenic variants which occurred in a virus of serotype A which was isolated in the U.K. in 1932 and had been used as a laboratory model at Pirbright for many years, if was found that their capsid protein amino acid sequences differed only at positions 148 and 153 of VP1. This confirmed the antigenic importance of this region and reinforced the view that the sequence 141160 was important in eliciting neutralizing antibodies. A CD spectroscopy study confirmed that the amino acid substitutions at these positions were also responsible for the conformational differences of the peptides. Antigenic sites have also been identified on VP2 and VP3 by growing the virus in the presence of neutralizing monoclonal antibodies. However, it is likely that the major immunogenic site is contained in VP1 because peptides corresponding to the sequence absorb greatly in excess of 50% of the neutralizing activity of of polyclonal sera from convalescent and vaccinated animals. Considerable success has been achieved by vaccination of pigs by United Biomedical Inc, a small company located on Long Island, New York (Wang et al., 2002) but the results are less promising in cattle (Taboga et al., 1988). In preliminary studies it has been shown that a retro-inverse peptide assembled from D-amino acids in the reverse order with respect to the original sequence, i.e. the direction of the peptide bonds is reversed while the side-chain orientation of the amino acids is retained) elicited high levels of neutralizing antibodies in guinea pigs which persist longer than those induced by the corresponding L-peptide and conferred protection against the cognate virus (Briand et al., 1997). Promising results were also obtained in pigs. This approach, which provides a more stable product, would have distinct advantages in the field. 4.2. Hybrid Viruses

The demonstration by Racaniello and Baltimore (1981) that the DNA complementary to the genomic RNA of poliovirus was infectious provided the opportunity for constructing hybrid viruses by appropriate manipulation of the eDNA. The presence of the poly (C) tract in the RNA of FMDV proved to be a stumbling block for many years but cDNA of a serotype 0 virus was eventually obtained by Zibert et ale (1990). This work was confirmed by Rieder et ale (1993) using a virus of serotype A and hs been used in experiments to study the basis of attenuation.

5. The Carrier Issue The carrier state is the largest and most controversial issue confronting international trading following an outbreak of FMD in any country. It is well established that animals which have been infected with the virus, whether vaccinated or not, can Copyright © 2004 By Horizon Bioscience

Brown

131

carry it for years (Salt, 1993). The important issue is whether these animals can transmit the virus to naive animals. There seems to be no agreed view. Sutmoller et ale (2003) have reviewed the reports in which outbreaks of the disease were described following the introduction of healthy convalescent animals into areas free of the disease. If these reports are authentic, then no country that has had the disease can ever be regarded as FMD-free because it is inferred by Hedger (1968) that carrier animals may transmit the virus to their offspring. But to say that the evidence is confusing is an understatement. Nevertheless, the focus regarding trading has been on those countries which vaccinate against the disease, on the pretext that vaccinated animals could become infected without showing signs of the disease. But naive animals can also become infected without showing signs, as witnessed by the sheep in the U.K. 2001 epidemic. Consequently the rules about disease-free status are not based on scientific knowledge and should be revised accordingly. The only way to ensure that animals are "safe" to be exported is to test them for the presence of virus by the most sensitive method available. At present, this is by RT-PCR. It should not be done in an FMD laboratory because of the risk of contamination. If RNA is detected, this means that virus is present or has been present and is no longer detectable. This analysis could be done on pooled samples because of the great sensitivity of the method, following the strategy of the blood transfusion services. As a back-up, the sera of suspect animals could be tested for the presence of virus-specific antibody but this method is less sensitive than RTPCR. 6. The Future The major outbreaks of FMD in Taiwan in 1997 and, more particularly in the U.K. in 2001, stimulated international debate about how the disease should be handled in the future. Most of the research effort since it was found that the disease is caused by a virus has been aimed at producing an effective vaccine which was cheap and could be supplied in the quantities necessary for comprehensive vaccination programmes. The examples of what had been achieved in Western Europe and in some countries in South America showed the effectiveness of comprehensive vaccination programmes (Figure 1). Consequently the abandonment of the programmes in Europe in 1991, based on misguided international trading rules, appear ludicrous to the author. What is even more perplexing is that several laboratories are pursuing better vaccines, although other considerations mean that they will not be used. Another issue which emerged in the U.K. outbreak was the need to be able to

avoid the slaughter of uninfected animals by rapid diagnosis. It is not appreciated that it is unnecessary to isolate the infectious virus - or fail to do so - to make a diagnosis. The RNA of the virus can be detected by RT-PCR within two hours of receiving the tissue from the suspected case. Moreover, with the advanced instrumentation now available, there is no need to send the sample to a specialist laboratory. What is urgently required is the need to validate this new technology and then be prepared to accept the results. We have gained an enormous amount of knowledge over the past few years on virus-cell interaction and the determinants which define tissue tropism (Chapter 7). Copyright © 2004 By Horizon Bioscience

\14

Stepping Stones in FMD Research

The question to be asked is whether we know enough to make a start on breeding animals which are not susceptible to the disease. Success in that approach would obviate the need to slaughter or even have to consider the blunderbuss approach of vaccination. References Acharya, R., Fry, E., Stuart, D., Fox, G., Rowlands, D.J. and Brown, F. 1989. The threedimensional structure of foot-and-mouth disease virus at 2.9 A resolution. Nature 337: 709-716. Bachrach, H.L., and Breese, S.S. 1958. Purification and electron microscopy of foot-andmouth disease virus. Proc. Soc. Exp. BioI. Med. 97: 659-665. Bahnemann, H.G. 1975. Binary ethyleneimine as an inactivant for foot-and-mouth disease virus and its application for vaccine production. Arch. Virol. 47: 47-56. Beck, E., and Strohmaier, K. 1987. Subtyping of European foot-and-mouth disease virus strains by nucleotide sequence determination. J. Virol. 47: 47-56. Bittle, J.L., Houghten, R.A., Alexander, H., Shinnick, T.M., Sutcliffe, J.G., Lerner, R.A., Rowlands, DJ., and Brown, F. 1982. Protection against foot-and-mouth disease by immunization with a chemically synthesized peptide predicted from the viral nucleotide sequence. Nature 298: 30-33. Bradish, C.J., Henderson, W.M., and Kirkham, J.B. 1960. Concentration and electron microscopy of the characteristic particle of foot-and-mouth disease. J. Gen. Microbial. 22: 379-391. Briand, J.-P., Benkirane, N., Guichard, G., Newman, J.F.E., Van Regenmortel, M.H.V., Brown, F. and Muller, S. 1997. Aretro-inverso peptide corresponding to the GH loop of foot-and-mouth disease virus elicits high levels of long-lasting protective neutralising antibodies. Proc. Natl. Acad. Sci. USA. 94: 12,445-12,550. Broo, K., Wei, J., Marshall, D., Brown, F., Smith, TJ., Johnson, J.E. et al. 2001. Viral capsid mobility: a dynamic conduit for inactivation. Proc. Natl. Acad. Sci. USA. 98: 2274-2277. Brooksby, J.B. 1958. The virus of foot-and-mouth disease. Adv. Virus Res. 5: 1-37. Brown, F. Inactivation of viruses by aziridines. 2002. Vaccine 20: 322-327. Brown, F., and Cartwright, B. 1963. Purification of radio-active foot-and-mouth disease virus. Nature 31: 1168-1170. Brown, F., Hyslop, N.St.G., Crick, J., and Morrow,A.W. 1963. The use ofacetylethyleneimine in the production of inactivated foot-and-mouth vaccines. J. Hyg. 61: 337-344. Brown, F., and Crick, 1. 1959. Application of agar gel diffusion analysis to a study of the antigenic structure of inactivated vaccines prepared from the virus of foot-and-mouth disease. 1. Immuno!. 82: 444-447. Brown, F., Sellers, R.F., and Stewart, D.L. 1958. Infectivity of ribonucleic acid from mice and tissue culture infected with the virus of foot-and-mouth disease. Nature 182: 535536. Callahan, J.D., Brown, F., Osorio, F.A., Sur, J.H., Kramer, E., Long, G.W., Lubroth, J., Ellis, S.J., Shoulars, K.S., Gaffney, K.L., Rock, D.L., and Nelson, W.M. 2002. Use of a portable real-time reverse transcriptase-polymerase chain reaction assay for rapid detection offoot-and-mouth disease virus. JAVMA 220: 1636-1642. Capel-Edwards, M. 1971. The susceptibility of small mammals to FMDV. Vet. Bull. 41: 815-823. Capstick, P.B., Telling, R.C., Chapman, W.G., and Stewart, D.L. 1962. Growth of a cloned strain of hamster kidney cells in suspended cultures and their susceptibility to the virus offoot-and-mouth disease. Nature 195: 1163-1164. Cowan, K.M., and Graves, J.H. 1966. A third antigenic component associated with foot-andmouth disease infection. Virology 30: 528-540. Copyright © 2004 By Horizon Bioscience

Brown

151

De Castro, M.P. 1964. Behaviour of the foot-and-mouth disease virus in cell cultures: susceptibility of IB-RS-2 line. Archivos do Instituto Biologico Sao Paulo 31: 63-78. Denoya, C.D. Scodeller, E.A., Vasquez, C., and La Torre, J. 1978 . Foot-and-mouth disease virus. Endoribonuclease activity within purified virions. Virology 89: 67-74. Domingo, E., Escarmis, C., Baranowski, E., Ruiz-Jareabo, C.M., Carrillo, E., Nunez, J.L, Sobrino, F. 2003. Evolution of foot-and-mouth disease virus. Virus Research 91: 4763. Duque, H., and Baxt, B. 2003. Foot-and-mouth disease virus receptors: comparison of bovine alpha(V) integrin utilization by type A and 0 viruses. J. Virol. 77: 2500-2511. Elford, WJ., and Galloway, LA. 1937. Centrifugation studies, III - the viruses of foot-andmouth disease and vesicular stomatitis. Br. 1. Exp. Path. 18: 155. Fracastorius, H. 1546. In: De Contagione et Contagiosis Morbis et Curation, Libri iii. Frenkel, H.S. 1947. La culture de la virus de la fievre aphteuse sur I'epithelium de la langue des bovides. Bull. Off. Int. des Epizoot. 28: 155-162. Frisby, D.P., Newton, C., Carey, N.H. , Fellner, P., Newman, J.F.E., Harris, T.J.R., and Brown, F. 1976. Oligonucleotide mapping of picornavirus RNAs by two-dimensional electrophoresis. Virology 71: 379-388. Galloway, LA., and Elford, W.J. 1931. Filtration of the virus of foot-and-mouth disease through a new series of graded collodion membranes. Br. J. Exp. Path. 12: 407. Galloway, LA., Henderson, W.M., and Brooksby, J.B. 1948. Strains of the virus offoot-andmouth disease recovered from outbreaks in Mexico. Proc. Soc. Exp. BioI. Med. 69: 57-63. Geysen, H.M., Meloen, R.H. and Barteling, SJ. 1984. Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natn. Acad. Sci. USA. 81: 3998-4002. Gierer, A., Schramm, G. 1956. Infectivity of ribonucleic acid from tobacco mosaic virus. Nature, Lond. 177: 702-703. Graves, J.H. 1963. Formaldehyde inactivation offoot-and-mouth disease virus as applied to vaccine preparation. Am. J. Vet. Res. 24: 1131-1135. Grubman, M.J., Moraes, M.P., Chinsangaram, J., Mayr, G.A., and Mason, P. 2003. New approaches to control foot-and-mouth disease. In: Foot-and-Mouth Disease: Control Strategies. Dodet, B. and Vicari, M. eds. Elsevier Amsterdam and New York. p. 337343. Hecke, F. 1930. Zeuchtungsversuche des Maul-und-Klauenseuchevirus in Gewebekulturen. ZentBI. Bakt. Parasitkde I, 116: 386-414. Hedger, R.S. 1968. The isolation and characterization of foot-and-mouth disease from clinically normal herds of cattle in Botswana. J. Hyg. Camb. 68: 53-60. Hedger, R.S. 1981. Foot-and-mouth disease. In: Infectious Diseases of Wild Mammals. David, J.W., Karstad, L.H. and Trainer, D.O. eds. Hugh-Jones, M.E. 1970. Epidemiological studies on the 1967-68 foot-and-mouth disease epidemic: the possible spread of infection by farm rats (Rattus norvegicus). Brit. Vet. J. 126: 368-371. Jackson, T. King, A. M., Stuart, D. I., and Fry, E. 2003. Structure and receptor binding. Virus Res. 91: 33-46. King, A.M.Q., Underwood, B.O. McCahon, D., Newman, J.W.I., and Brown, F. 1981. Biochemical identification of viruses causing the 1961 outbreaks of foot-and-mouth disease in the U.K. Nature 293: 479-480. Laporte, J., Grosclaude, J., Wantyghem, J., Bernard, S., Rouze, P. 1973. Neutralisation en culture cellulaire du pouvoir infectieux du virus de la fievre aphteuse par des serums provenant de porcs immunises aI' aide d 'une proteine viral purifiee. C.R. Acad. Sci. 276: 3399-3401.

Copyright © 2004 By Horizon Bioscience

116

Stepping Stones in FMD Research

Loeffler, F., and Frosch, P. 1897. Summarischer Bericht ueber der Ergebnisse der Untersuchungen zur Erforschung der Maul- und Klauenseuche. ZentBl. Bakt Parasitkde I, 22: 257-259. Maitland, M.C., and Maitland, H.C. 1931. Cultivation ofFMDV. I Compo Path. Ther. 44: 106-113. Mason, P.W., Grubman, MJ., and Baxt, B. 2003. Molecular basis of pathogenesis of FMDV. Virus Research 91: 9-32. Moosbrugger, G.A. 1948. Recherches experimentales sur la fievre aphteuse. Schweiz. Arch. Tierheilk.90: 176-198. Mowat, G.N., and Chapman, .W.G. 1962. Growth of foot-and-mouth disease virus in a fibroblastic cell derived from hamster kidneys. Nature, Lond. 194: 253-255. Mussgay, M., and Strohmaier, K. 1958. Gewinnung eines infektioesen Prinzips von Ribonucleinsaure Character aus homogenaten mit dem maul-und-klauenseuche virus infizierten Jungmause. Zentr. Bakteriol. Parasitenk. Abb. I orig. 173: 163-174. Newman, J.F.E., and Brown, F. 1997. Foot-and-mouth disease virus and poliovirus particles contain proteins of the replication complex. J. Virol. 71: 7657-7662. Racaniello, V.R. and Baltimore, D. 1981. Cloned poliovirus complementary DNA is infectious in mammalian cells. Science 214: 916-919. Randrup, A. 1954. On the stability of foot-and-mouth disease virus dependent on pH. Investigations on the complement fixing and the immunizing antigen as well as on the infective agent. Acta Path. Et Microbiol. Scand. 35: 388-395. Rieder, E., Bunch, T., Brown, F., and Mason, P.W. 1993. Genetically engineered foot-andmouth disease viruses with polyC tracts of two nucleotides are virulent in mice. J. Virol. 67 (9): 5139-5145. Rosenbusch, C.T., Decamps, A., and Gelormini, N. 1948. Intradermal foot-and-mouth disease vaccine. Results obtained from the first million head of cattle vaccinated. J. Am. Vet. Med. Assoc. 112: 45-47. Salt, J.S. 1993. The carrier state in foot-and-mouth disease: an immunological review. Br. Vet. J. 149: 207-223. Sangar, D.V., Rowlands, D. J., Cavanagh, D., Brown, F. 1976. Characterization of the minor polypeptides in the foot-and-mouth disease particle. J. Gen. Virol. 31: 35-46. Schaffer, F.L., and Schwerdt, C.E. 1955. Crystallisation of purified MEF-l poliomyelitis virus particles. Proc. Nat. Acad. Sci. 41: 1020-1023. Schlesinger, N., and Galloway, LA. 1937. Sedimentation of the virus of foot-and-mouth disease in the Svedberg super centrifuge. J. Hyg., Camb. 37: 445. Skinner, H.H., 1951. Propagation of strains of foot-and-mouth disease virus in unweaned mice. Proc. R. Soc. Med. 44: 1041-1044. Strohmaier, K., Franze, R., Adam K.-H. 1982. Localisation and characterisation of the antigenic portion of the foot-and-mouth disease virus protein. J. Gen. Virol. 59: 295306.

Sutmoller, P. and Barteling, SJ. 2003. The history of foot-and-mouth disease vaccine development: a personal perspective. In: Foot-and-Mouth Disease: Control Strategies. Dodet, B. and Vicari, M. eds. Elsevier Amsterdam and New York. p. 337-343. Sutmoller, P., Barteling, S. S., Olascoaga, R. C., Sumption, K. J. 2003. Control and eradication offoot-and-mouth disease. Virus Res. 91: 101-144. Taboga, 0., Tami, C., Carrillo, E., Nunez, II., Rodriguez, A., Saiz, J.C., Blanco, E., Valero, M.-L., Roig, A., Camarero, J.A., Andreu, D., Mateu, M.G., Giralt, E., Domingo, E., Sobrino, F., and Palma, E.L. 1997. A large-scale evaluation of peptide vaccines against foot-and-mouth disease: lack of solid protection in cattle and isolation of escape mutants. J. Virol. 71: 2606-2614. Traub, E., and Pyl, G. 1943. Untersuchungen ueber das komplementbindende Antigen bei der Maul- und Klauenseuche. Z. Immun. Forsch. Exp. Ther. 104: 158.

Copyright © 2004 By Horizon Bioscience

Brown

171

Twomey, T., France, L.L., Hassard, S., Burrage, T.C., Newman, J.F.E., and Brown, F. 1995. Characterization of an acid-resistant mutant of foot-and-mouth disease virus. Virology 206: 69-75. Vallee, H., and Carre, H., 1922. Sur la pluralite du virus aphteux. C. r. hebd. Acad. Sci. Paris 174: 1498-1500. Vallee, H., Carre, H., and Rinjard, P. 1925. On immunisation against foot-and-mouth disease. Rech. Med. Vet. 101: 297-299. Waldmann, 0., and Pape, J. 1920. Die Kuenstliche Uebertragung der Maul-undKlauenseuche aufDas Meerschweinchen. Berl. Tierarztl. Wschr. 36: 519-520. Waldmann, 0., and Trautwein, K. 1926. Experimentelle Untersuchungen ueber die Pluralitet des Maul-und-Klauenseuche Virus. Berl. Tierarztl. Wschr. 42: 569-571. Wang, C.Y., Chang, T.Y. Walfield, A.M., Ye, J., Shen, M., Chen, S.~, Li, M.C., Lin, Y.L., Jong, M.H., Yang, P.C., Chyr, N., Kramer, E., and Brown, F. 2002. Effective synthetic peptide vaccine for foot-and-mouth disease in swine. Vaccine 20: 371-391. Wesslen, T., and Dinter, Z. 1957. The inactivation of foot-and-mouth disease virus by formalin. Arch. Ges. Virusforsch. 7: 394-402. Zibert, A., Maass, G., Strebel, K., Falk, M.M., and Beck, E. 1990. Infectious foot-and-mouth disease virus derived from a cloned full-length eDNA. J. Virol. 64: 2467-2473.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 2 Genome Organisation, Translation and Replication of Foot-and-Mouth Disease Virus RNA Graham J. Belsham and Encarnacion Martinez-Salas

Abstract Foot-and-mouth disease virus (FMDV) is a member of the picornavirus family. The virus has a positive-sense RNA genome that functions like a mRNA and encodes a viral polyprotein. This polyprotein is co-translationally processed, largely by virus encoded proteases, to produce about 15 mature proteins plus many different precursors. These proteins have functions in RNA replication, modification of the host cell and in the assembly of new virus particles. The RNA genome also has to act as the template for RNA replication. This process occurs in two main stages. Initially synthesis of negative strands occurs using the positive strand template and then the production of many positive-sense infectious RNAs is achieved from the negative strands. Some of these infectious RNAs are then packaged by the structural proteins to produce new virus particles. Particular emphasis in this review is placed on the mechanism of protein synthesis initiation on the viral RNA. Early on in FMDV-infected cells, the synthesis of host cell proteins is inhibited, as a result of modifications to the cellular translation machinery that are induced by virus-encoded proteases. However, viral protein synthesis is maintained under these conditions. Initiation of protein synthesis occurs at two different start sites and is directed by a large RNA element of about 450 nt termed the internal ribosome entry site (IRES). 1. Introduction Foot-and-mouth disease virus (FMDV) is the causative agent of one of the most economically important diseases of farm animals. The virus is classified within the aphthovirus genus of the Picornaviridae. All picornaviruses, including FMDV, have a single stranded positive-sense RNA genome. A single copy of the genome is contained within each virus particle. A protein shell comprised of 60 copies of the 4 different structural proteins lA (VP4, which is internal), IB (VP2), Ie (VP3) and ID (VPl) surrounds the RNA. The FMDV genome is over 8,300 nt in length and its structure has certain similarities to eukaryotic cellular mRNAs in that it contains a single, long, open reading frame (ORF) and has a poly(A) tail at its 3' terminus (see Figure 1). However, in contrast to cellular mRNAs, there is no cap structure (m7GpppN... ) at the 5' terminus of the genomic RNA but a virus-encoded peptide, termed VPg (or 3B), is covalently linked to the terminal nucleotide. Copyright © 2004 By Horizon Bioscience

S-fragment CC

_

PK W

I

ere

~~

IRES

I

Lb

.-_A_U_G

I'

_

I

L~

1------------:

f\ - - -.-..

I

·-· -· · · · · · A-· :~·~· ·-· · · · · ~:;::~

_.

~ 1300nt ~ ..' IAB'---3D--~ AAA(n) 3B (VPg)

VP3 VPl

VPO

myristoylation

~

Pl-2A

ILProl

~

lAB

+

~

lAB

encapsidation

I0

~

~~ VP4

VP2

+

P2

P3 3D

I

+

~

B~~ I ~~ Gill ~ ~ ~ EJ I 11 3A

I 2 3 3C

3D

3D

DJB

Jcpr

JDpol

I

I

Figure 1. Genome organization of FMDV. The important features of the FMDV genome are shown. The positions of RNA elements within the 5' UTR are indicated. The viral RNA encodes a polyprotein. Many of the processing intermediates and mature proteins that are generated by the action of virus-encoded proteases on the polyprotein precursor are depicted. Each of these features is discussed within the text. Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

211

The 5' untranslated region (UTR) of the FMDV RNA is about 1300 nt in length. This is much longer than the 5'UTRs of most cellular mRNAs (typically 50-100 nt). The large open reading frame within the RNA encodes a polyprotein. However, the complete polyprotein is never detected since protein processing, largely by virus-encoded proteases, commences during synthesis to generate the mature viral proteins required for RNA replication and virus assembly. A notable feature of the FMDV genome is that there are two different sites on the RNA at which the initiation of protein synthesis occurs. Their presence is responsible for generating two alternative forms of the N-terminal component of the viral polyprotein, the leader (L) protease (see Figure 1). A very important feature of picornavirus RNA is that it is infectious (see, e.g. Belsham and Bostock, 1988). No viral proteins are required by the viral RNA to initiate the infection cycle. Thus, the virus particle essentially serves to protect the genome while it is outside of cells and to bring about its delivery to the cytoplasm of a cell where a new cycle of infection can be initiated. The entire replication cycle of the virus occurs in the cytoplasm of cells and for certain picornaviruses it has been demonstrated that their replication is maintained in enucleated cells (Follett et aI., 1975). The viral RNA contains all the information required to allow the virus to take over the cellular machinery. Most picornaviruses induce within infected cells the shut down of nearly all of the host macromolecular synthesis and only the production of viral RNA and viral proteins is observed. FMDV RNA is replicated with high efficiency within susceptible cells and after a few hours of infection the amount of viral RNA (all genome length) within cells can reach a level similar to that of all the cellular cytoplasmic mRNAs (about 5% of total RNA). The viral RNA is synthesised by the virus encoded RNA-dependent RNA polymerase (3Dpol) (see Figure 1). This process also requires other viral proteins, e.g. VPg (3B) which acts as the primer for RNA synthesis, inside a membraneassociated replication complex that contains several other viral proteins. The infectivity of viral RNA demonstrates that the first phase of the infection cycle has to be translation of the viral RNA so that each of the viral proteins is generated within the cell. These viral proteins are required for the formation of the RNA replication complex and to generate the capsid proteins that are required to form new virus particles. At some point, it appears that there has to be a switch in the function of the input viral RNA so that translation is blocked and RNA replication can commence. This is necessary since the process of translation of picornavirus RNA (in which ribosomes move along the RNA in a 5'-3' direction) is apparently not compatible with the initial step in the replication of the viral RNA that is achieved by the movement of the RNA polymerase from the 3' end to the

5' terminus (see Gamarnik and Andino, 1998). It is clear that the distinct processes of protein synthesis, RNA replication and RNA packaging (to form virus particles) need to be regulated and probably differentially localized within the cell. Separating the various activities to different sites within the cell could help to overcome potential incompatibilities between these functions which each require positive-sense RNA and hence may compete. The targets of any signals required for regulation of function must be contained within the sequence of the genome.

Copyright © 2004 By Horizon Bioscience

122

FMDV RNA Structure and Function

This chapter will seek to examine the different features of the FMDV genome, which are responsible for directing the production of virus proteins and viral RNA within the infected cell. There is much that is currently not known about these processes and we will seek to identify these areas and point out where relevant information is available from other related viruses. However, it should always be borne in mind that the various members of the picornavirus family are distinct and what is true for one virus may not always apply to others. Several unique features of FMDV are already known and undoubtedly more remain to be identified. 2. The Structure and Function of the FMDV RNA 5' UTR As indicated above, the 5' UTR of FMDV RNA is about 1300 nt long. This is significantly larger than most other picornavirus 5'UTRs, e.g., the poliovirus (PV) RNA 5' UTR is about 740 nt in length while the encephalomyocarditis virus (EMCV) RNA 5' UTR is about 830 nt long. The 5' UTR of each picornavirus RNA includes sequences required for the initiation of protein synthesis on the viral RNA, for RNA replication and probably for other, currently unknown, functions. The FMDV RNA 5' UTR can be considered in several distinct regions, including the S-fragment, a poly(C) tract, several pseudoknots and elements termed the cis-acting replication element (ere) and the internal ribosome entry site (IRES). The IRES is about 450 nt in length and directs the initiation of protein synthesis on the viral RNA (see Belsham and Sonenberg, 1996; 2000; Belsham and Jackson, 2000; Martinez-Salas et ai., 2001), its properties will be discussed in detail below. The ere is a recently defined structure (about 55 nt in length) that is required for FMDV RNA replication (Mason et ai., 2002) and will also be discussed further below. 3. RNA Replication Elements 3.1. The S-Fragment At the 5' terminus of the FMDV genome is a region of about 360 nt, termed the Sfragment, which is predicted to form a large hairpin structure (Clarke et al., 1987; Escarmis et al., 1992; Witwer et al., 2001). The S-fragment is the name given to the smaller of the two RNA fragments that are generated when viral RNA is treated \vith oligo(dG) and RNAseH (Rowlands et al., 1978). This treatment cleaves the RNA within the poly(C) tract and generates the 5'-terminal S-fragment plus the major portion of the genome (the large or L-fragment) that includes the rest of the 5' UTR, the polyprotein coding region and the 3' UTR. It is assumed that the S-fragment is required for RNA replication but there are no data on the specific role of this sequence in FMDV replication or about the nature of any protein interactions with it. In contrast, there is now a considerable body of data indicating that both cellular and viral proteins interact with the 'cloverleaf' structure that is located at the 5' terminus of the PV genome (Gamarnik and Andino, 1997; Parsley et al., 1997). This element (only about 80 nt in length) binds to a cellular RNAbinding protein termed the poly(rC) binding protein 2 and to the PV 3CD protein (by interaction with sequences within 3C). The role of these interactions is being investigated intensively. It has been suggested that they may play a role in the switch from translation to replication (Gamarnik and Andino, 1998). Additionally, they may serve to circularise the RNA (since the poly(A) binding protein can Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

23\

bind to both the 3' terminal poly(A) tail and the poly(rC) binding protein) which may facilitate replication and/or translation (Herold and Andino, 2001). It is also possible that the presence of proteins at the RNA termini serves to protect the viral RNA from degradation (Murray et aI, 2001). 3.2. The poly(C) Tract and Pseudoknots On the 3' side of the S-fragment within FMDV RNA there is a long, almost homopolymeric, stretch of C residues termed the poly(C) tract. This is a common feature of the genomes of FMDV and most cardioviruses (e.g. EMCV, but not Theiler's murine encephalitis virus). There is considerable variation in the length of this poly(C) tract amongst different strains of FMDV, from about 80 nt to about 420 nt. In general, the shorter tracts are found in laboratory strains and typically field strains have a tract of about 200 nt (Brown et al., 1974; Harris and Brown, 1977). However, the largest poly(C) tract identified was found in a strain of FMDV (R 100) recovered from persistently-infected cells in culture (Escarmis et al., 1992) but the significance of the size of this homopolymeric sequence is not clear. The R 100 strain has been described as hypervirulent in BHK-21 cells (de la Torre et aI., 1988) but it is attenuated for growth in mice and cattle (Diez et al., 1990). However since the R 100 genome is different from the parental virus at about 80 positions (1 % of the genome), it is not clear which changes (including the long poly(C) tract) are responsible for the different phenotypes. There is a strong selection pressure for the presence of the poly(C) tract within the viral RNA. When virus was recovered from infectious RNA transcripts, derived from plasmids that initially contained as few as 6 C residues at this location, it was found that the length of the poly(C) tract in the replicated RNA had increased to about 80 nt or more (Rieder et aI., 1993). In contrast, RNA transcripts containing just 2 C residues at the location of the poly (C) tract did not regenerate a poly(C) tract but these rescued viruses grew fairly poorly in tissue culture cells. However, each of the rescued viruses retained virulence in mice irrespective of the length of the poly(C) tract (Rieder et al., 1993). The instability in the length of this tract in FMDV RNA is in marked contrast to observations made using mengovirus RNA (Duke et aI., 1989, 1990). Modified mengoviruses, with different size poly(C) tracts, were rescued in tissue culture and their virulence was determined in mice. The viruses with short poly(C) tracts were greatly attenuated but remarkably no selection for viruses with enlarged tracts occurred within these animals (Duke et al., 1990). Surprisingly, the role of the poly(C) tract in EMCV, another cardiovirus like mengovirus, appears quite different (Hahn and Palmenberg, 1995). Mutant EMC viruses, containing very short poly(C) tracts, were about as pathogenic in animals as wild type (wt) EMCV and replicated at very similar rates in cells as judged by single-step growth curves. Between the poly(C) tract and the IRES element in FMDV RNA there is another stretch of sequence, about 250 nt in length, that is believed to contain multiple pseudo-knots. Different FMDV strains have been predicted to contain 2 to 4 pseudo-knots (Clarke et al., 1987; Escarmis et aI., 1995). Interestingly the cardiovirus RNAs are predicted to contain multiple pseudoknots located on the 5' side of their poly(C) tract rather than on the 3' side as in FMDV RNA (Martin and Palmenberg, 1996). It may be that the pseudoknots are involved, in some manner, in a joint function with the poly(C) tract.

Copyright © 2004 By Horizon Bioscience

\24

FMDV RNA Structure and Function

3.3. A Cis-acting Replication Element (ere) In addition to sequences at the termini of the viral RNA, recent work has established that internal RNA sequences are required for RNA replication. This feature has been termed a 'cis-acting replication element' (ere) The initial characterisation of ere structures within picornavirus RNAs resulted from the observation that a region within the PI-coding sequence was required to achieve efficient replication of replicons based on human rhinovirus-14 (HRV-14) (McKnight and Lemon, 1996). This contrasts with studies on the closely related PV RNA which have demonstrated that the entire capsid coding sequence can be deleted without affecting RNA replication (Barclay et al., 1998). Further work (McKnight and Lemon, 1998) showed that RNA replication required a specific stem-loop structure located within the coding region for the HRV-14 capsid protein ID (VPI). This structure was termed a 'ere' and subsequently, analogous elements have been identified in other picornavirus genomes (Lobert et al., 1999, Goodfellow et al., 2000; Gerber et ai., 2001). The ere elements from HRV-14, HRV-2, cardioviruses and PV are within the coding regions for ID, 2A, IB and 2C respectively. However, the ere structures can be moved without blocking function. It has been shown recently that the PV and HRV-2 ere sequences act as a template for the uridylylation of VPg (3B) by the viral RNA polymerase in vitro (Paul et al., 2000; Gerber et al., 2001). This reaction generates the products VPgpU and/or VPgpUpU that act as primers for viral RNA synthesis. Thus, it is believed that this process is an essential step in RNA replication. The development of replicons based on FMDV (McInerney et ai., 2000) showed that the entire FMDV L-coding sequence and at least 94% of the FMDV P I-coding sequence could be deleted without any adverse effect on the replication of the viral RNA. These studies indicated that if, as expected, there is a ere present within the FMDV genome then it must be present within the coding region for the non-structural proteins or within non-coding sequences. Each of the ere structures so far defined within the entero-, rhino- and cardiovirus RNAs include a conserved sequence motif, AAACA, located within a loop at the top of a stable stem-structure (Rieder et al., 2000; Gerber et al., 2001). There are multiple places in the FMDV sequence (outside of the L-PI coding region) where a copy of the AAACA motif is highly conserved between different virus isolates. These motifs occur in the 2B, 2C and 3D coding sequences of the FMDV genome and within the 5' UTR. Recently, Witwer et al. (2001) published a computer prediction for a ere structure within the 2C coding region of FMDV RNA. However, Mason et al. (2002) have now provided functional evidence for a ere, containing the AAACA motif, located within the 5' UTR of FMDV, just upstream of the IRES (see Figure I). It was demonstrated that insertion of the wt ere into the 3' UTR could restore replication ability to an RNA transcript containing a mutated (and defective) ere within its 5'UTR. When the ere structure in FMDV RNA was identified it provided an explanation for some results that had demonstrated the presence of a temperaturesensitive (ts) mutation within the 5' UTR of FMDV RNA. This mutation is located \vithin the stem-loop structure that constitutes the ere and it destabilises the structure. A non-ts revertant virus has a compensating mutation within the ere that restores the stability of the stem (Tiley et al., 2003). An important feature of this Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

251

ts mutant is that the defect in replication can be complemented in trans by other ts FMDVs (with defects elsewhere in the genome). This result seems compatible

with the earlier observations that a pool of free VPgpUpU is generated within picornavirus infected cells (Crawford and Baltimore, 1983). Thus the term 'ere' (cis-acting replication element) seems inappropriate at least for the FMDV element and a reassessment of this feature of other' ere' structures seems warranted. It was suggested by Tiley et ale (2003) that this element is called a 3B-uridylylation site (BUS). 4. The IRES Element- Structure and Function Translation of FMDV RNA is initiated internally, under the control of a sequence known as the IRES (Belsham and Brangwyn, 1990; Kuhn et al., 1990; MartinezSalas et al., 1993). The FMDV IRES consists of a highly structured region, located several hundred nucleotides away from the uncapped 5' end of the genomic RNA (see Figure 1). Initiation of translation mediated by IRES elements represents an alternative to the cap-dependent translation initiation mechanism used for most cellular mRNAs. The 5' end of all cytoplasmic eukaryotic mRNAs has a cap structure (m7GpppN ... ) which plays a crucial role during translation initiation, it is recognized by the initiation factor elF4F (Gingras et ai., 1999; Hershey and Merrick, 2000). This heterotrimer comprises the translation initiation factors eIF4E (that binds to the cap), elF4A (an RNA helicase) and eIF4G. The latter component has distinct binding sites not only for eIF4E and eIF4A but also for other proteins including elF3 (located on the small ribosomal subunit), the poly(A) binding protein (PABP) and Mnk-1 (an eIF4E kinase). It is generally believed that the small ribosomal subunit interacts with the eIF4F complex (bound at the 5' cap) and then migrates with it along the 5' UTR in a 5' to 3' direction. The presence of elF4A is believed to facilitate the unwinding of RNA secondary structure. This migration is called scanning and continues until an AUG codon in an appropriate context (as defined by Kozak, 1989) is encountered. At this point, the ribosome pauses, the large ribosomal subunit joins and polypeptide synthesis can commence. The ability of the scanning process to recognize the correct initiation codon can be inhibited by complex RNA structure and by the presence of additional AUG codons. In contrast, internal initiation of translation mediated by an IRES involves the direct recruitment of the translational machinery to an internal position in the mRNA usually with the help of cellular trans-acting factors. IRES elements were first identified within picornavirus RNAs (Pelletier and Sonenberg, 1998; lang et ai., 1998). Consistent with their role in picornavirus translation, IRES-dependent translation initiation can bypass stress conditions that are inhibitory for cap-dependent translation initiation (Hellen and Sarnow, 2001; Martinez-Salas et al., 2001). Such conditions occur following eIF4G cleavage (Gradi et al., 1998) that is induced during picornavirus infection as a consequence of the action of viral proteases Lb (FMDV) or 2A (poliovirus), or due to the dephosphorylation of the translational repressor 4E-BPl in EMCV infected cells (Gingras et al., 1996). The element in the viral RNA that constitutes the IRES is believed to adopt a tertiary structure that is essential for internal initiation of translation. However, IRES elements belonging to the picornavirus family do not show extensive primary sequence conservation. Two major groups of picornavirus IRES structures have Copyright © 2004 By Horizon Bioscience

\26

FMDV RNA Structure and Function

been described, from the entero-/rhinoviruses and from the cardio-/aphthoviruses (Belsham and Jackson, 2000). Neither type of IRES has any apparent similarity to that of hepatitis C virus (HCV) (Honda et aI., 1999). The absence of apparent structural conservation between different IRES elements may reflect the diversity of strategies used by IRES elements to interact with the translational machinery. We discuss here the features of the RNA structure of the FMDV IRES that are known to be functionally relevant and compare them to the structural features described for other IRES elements especially from other picornaviruses and the unrelated HCV IRES. 4.1. Structural Motifs in the FMDV IRES Element and Their Role in Function The initial mapping of the 5' and 3' borders of the FMDV IRES by Kuhn et ale (1990) identified a functional element of 462 nucleotides, which appear organized in 5 structural domains (see Figure 2). The strong phylogenetic conservation of the predicted secondary structure elements within a highly variable RNA virus such as FMDV (Domingo et al., 1996) argues in support of these structural features (Escarmis et al., 1992; Martinez-Salas et al., 1993). The first 85 nt contain domains 1 and 2 (see Figure 2). The contribution of domain 1 to the IRES is small since a core sequence of about 435 nt is sufficient for efficient IRES activity (Belsham and Branwgyn, 1990). Indeed, the first 21 nt of the domain 1 is now recognized as the right arm of the ere stem-loop (Mason et al., 2002). However, the presence of these sequences upstream of the IRES may, under certain conditions, help to stabilize the structure of domain 2 that is essential for IRES activity (Kuhn et al., 1990). It is apparent that RNA elements with distinct functions may use overlapping sequences. The 60 residues of domain 2 are predicted to form a highly conserved stem-loop structure that has a pyrimidine-rich sequence at its apical loop (UCUUU). This stem-loop contains the main binding site for a 57 kDa protein known as the polypyrimidine tract binding protein (PTB) (Luz and Beck, 1991). Binding of this protein to a second site on the FMDV IRES was impaired by mutations in a pyrimidine-rich sequence located near the 3' end of the IRES element (nt 438-447 in Figure 2). During early studies of the picornavirus IRES elements, it was noticed that a polypyrimidine tract was the only motif conserved between rhino-/enterovirus IRES elements and the cardio-/aphthovirus IRES elements (Meerovitch et aI., 1991; Pilipenko, et al., 1992; Meerovitch and Sonenberg, 1993). However, in spite of this conservation, it was observed later that a poly-purine tract can functionally substitute (at about 70% efficiency) for a polypyrimidine tract within the EMCV IRES (Kaminski et al., 1994), suggesting that it provides an unstructured region at the 3' end of this element. The apical part of domain 3 within the FMDV IRES contains stable stem-loop structures that are believed to form a four-way junction (Femandez-Miragall and Martinez-Salas, 2003). Two phylogenetically conserved motifs, GNRA and RAAA, that are present in the apical loops, constitute essential regions for the activity of the IRES within cells (Lopez de Quinto and Martinez-Salas, 1997; Robertson et aI., 1999). The GNRA motif appears to be responsible for the organization of the adjacent stem-loops, as deduced from accessibility studies using ribonucleases and dimethyl sulfate (Femandez-Miragall and Martinez-Salas, 2003). The GNRA tetraloop is a building block that is commonly found in structured RNA molecules (Doherty et al., 2001). Thus, a similar role for this motif within IRES elements is Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

271

3 (I)

5 (L)

1 (G) Figure 2. FMDV IRES structural organization. A representation of the secondary structure of the FMDV IRES generated by the M-Fold program of the GCG package is shown. The positions of residues cited in the text are indicated with numbers. An unused AUG triplet located 8 nt upstream of the first functional AUG is underlined. The sites on the IRES that interact with eIF4G, eIF4B, eIF3 and PTB are described in the text.

Copyright © 2004 By Horizon Bioscience

\28

FMDV RNA Structure and Function

likely. There are two pieces of evidence which suggest that tertiary interactions are promoted by the GNRA motif. Firstly, a UNCG motif (also commonly located within apical loop sequences) does not functionally substitute for the GNRA sequence. Secondly, binding of synthetic stem-loops bearing UCCG or GUAG sequences in place of the GNRA motif were significantly reduced in their binding efficiency to domain 3 compared to the wild-type GUAA hairpin (FernandezMiragall and Martinez -Salas, 2003). In contrast to the GNRA motif, mutational analysis of the conserved AAAA (nt 199-202, Figure 2) sequence did not identify a consensus motif. Although large insertions were detrimental for IRES activity, substitutions that did not modify the predicted secondary structure of this region and had a sequence close to CAAA were tolerated (Lopez de Quinto and Martinez-Salas, 1997). Another sequence in domain 3, the loop ACCC (nt 234-238, Figure 2) that is also conserved among picornavirus IRES elements, is tolerant of nucleotide substitutions. Simultaneous substitution of 3 residues of this C-rich loop to adenines reduced FMDV IRES activity only 2-fold relative to the wild-type IRES sequence (Martinez-Salas et aI., 2002). This result suggests that the conservation of this loop may be associated with a function that is unrelated to internal initiation of translation. The ACCC loop is a candidate to interact with the poly(rC) binding proteins (PCBP). A binding site for PCBP2 within the FMDV IRES has been mapped to domain 3 (Stassinopoulos and Belsham, 2001), consistent with the presence of the conserved ACCC loop. However, in contrast to the situation with the poliovirus IRES, it has been shown that PCBP2 binding to the EMCV or the FMDV 5' UTRs, is not required for IRES activity (Walter et al., 1999; Stassinopoulos and Belsham, 2001). Analysis of sequences located at the base of domain 3 showed a requirement for secondary structure involving Watson-Crick base-pairing (Martinez-Salas et aI., 1996). These results strongly suggested that formation of a helical structure around positions 88 and 297 of the FMDV IRES is needed for efficient internal initiation of translation. Computer-assisted secondary structure prediction of the whole FMDV IRES sequence represents the base of domain 3 as a long stem interrupted with bulges (Figure 2). It may be that the role of this structural element is to contribute to the stability of the entire domain. If domain 3 plays an essential role in the organization of IRES architecture as suggested (Ramos and Martinez-Salas, 1999) then small changes in the stability of this domain may induce a reorganization of the whole element with important consequences for IRES function. The functional roles of the conserved GNRA and RAAA motifs are still under study. The prominent role of GNRA motifs in organizing RNA structures (Doherty et al., 2000), prompted a search for functional GNRA receptors, resembling the 11 nt structural element described by Costa and Michel (1997), within domains 2 and 4 of the IRES but this was unsuccessful (Lopez de Quinto et aI., unpublished data). However, with this working hypothesis in mind, the possibility of RNARNA complex formation between separate domains of the FMDV IRES was analyzed. Long-range RNA-RNA interactions were shown to occur in vitro between functional domains of the FMDV IRES (Ramos and Martinez-Salas, 1999). These complexes are strand specific, and depend on RNA concentration, ionic conditions and temperature. The RNA-RNA interactions observed in vitro between separated domains of the FMDV IRES, in the absence of proteins, suggest Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

291

that the IRES fragments fold in specific forms, depending upon environmental conditions. A similar situation has been described for the HCV IRES, which adopts different structures in response to increased ionic concentration (Kieft et al., 1999) or internal mutations (Jubin et al., 2000). The central region (domain 3 or I, Figure 2) of the FMDV IRES is unique in its ability to interact with each of the other domains, including the entire IRES (Ramos and Martinez-Salas, 1999). This result has two implications: firstly, it suggests that domain 3 acts as a scaffold structure that holds together the remaining domains of the IRES, and secondly, that it is essential to determine intermolecular interactions with other IRES molecules. Such intermolecular interactions may account for the complementation data previously reported between defective IRES elements (Drew and Belsham, 1994; Roberts and Belsham, 1997). In the context of a picornavirus infection where replication occurs in a distinct region of the cell (Bolten et al., 1998), the local RNA concentration may reach the levels required to allow such intermolecular interactions. Physiological changes affecting ionic conditions, pH gradients, free radical formation, expression of specific RNAbinding proteins, etc., may induce reorganization of the IRES structure that could have important consequences for recruiting trans-acting factors. Since the intracellular concentration of cations varies with time during the course of a picornavirus infection (Carrasco, 1995), changes in tertiary structure of the IRES may occur within infected cells and, as a consequence, translation efficiency may be modulated. 4.2. Features of the RNA Structure That Mediate Interaction of Translation Initiation Factors with the FMDV IRES

Sequence comparison between the FMDV and EMCV RNAs at the base of domain 4 is of great interest since there is both primary sequence and secondary structure conservation. Two A-rich internal bulges are conserved within the secondary structure of the different field isolates of FMDV as well as in the related EMCV IRES (Hellen and Wimmer, 1995). The single point substitution of A312 to U (Figure 2) severely impaired IRES activity (Lopez de Quinto and Martinez-Salas, 2000). Moreover, deletion of A409 present in the opposite position of this internal loop, is also incompatible with IRES activity (Martinez-Salas et ai., 2002), again demonstrating the biological importance of this region. Similarly, mutations in the 5' or the 3' side of the stem (around positions 309 and 412 in Figure 2) that abrogated the predicted secondary structure were also highly detrimental for IRES activity. Restoration of the secondary structure using second-site mutations, resulted in a complete recovery of IRES activity (Lopez de Quinto and MartinezSalas, 2000). These results demonstrated that base-pair formation is required at these positions within the IRES for activity. Additionally, sequence comparison of different field isolates of FMDV showed that the AAAAA sequence in EMCV (Hellen and Wimmer, 1995; Kolupaeva et al., 1998) has as its counterpart A398AAAR402 in FMDV (Figure 2). Substitution of the A402 in this motif to U or C abrogated FMDV IRES activity in the context of dicistronic constructs (Martinez-Salas et al., 2002). Disruption of the structural motif at the base of domain 4 is associated with the modification of essential RNA-protein interactions (Lopez de Quinto and MartinezSalas, 2000). This was demonstrated using mutated FMDV transcripts that pointed Copyright © 2004 By Horizon Bioscience

\30

FMDV RNA Structure and Function

Table 1. RNA-binding proteins interacting with IRES elements a

IRES FMDV

a

eIF4G, eIF4B, eIF3

eIFs

RBPs PTB, PCBP2, ITAF45

EMCV

eIF4G, eIF4B, eIF3

PTB, PCBP2

PV

eIF4G, eIF4B, eIF3

PTB, PCBP2, La, unr

HRV

unknown

unr, PTB, PCBP2, La

Hev

eIF3

PTB, PCBP2, La

Not all of these interactions have been shown to affect activity

to the AA-internal loop (nt 312-313, Figure 2) as a critical region for binding to eIF4G. Consistent with this observation, eIF4G interacts with the related EMCV IRES in a similar position (Kolupaeva et al., 1998) and additional nucleotides within the J-domain have also been implicated in this interaction (Kolupaeva et aI., 2003; Clark et al., 2003) . A transcript corresponding to FMDV domain 4 alone binds to eIF4G as effectively as the full length IRES, strongly suggesting that no additional sites for recognition of this protein in the FMDV IRES are required. Furthermore, the carboxy-terminal end of the proteolytically processed form of eIF4G (eIF4G-Ct), present in cells transfected with the FMDV Lb protease, binds as efficiently to the FMDV IRES as the unprocessed protein (Lopez de Quinto and Martinez-Salas, 2000). Fully consistent with this result, the use of RNA transcripts corresponding to 150 nt from the 3' end of the FMDV IRES to deplete RRL extracts lead to a strong reduction of factors required for cap-dependent translation, including eIF4G (Stassinopoulos and Belsham, 2001). The second species of eIF4G, termed eIF4GII, is also cleaved by the FMDV encoded Lb protease in vitro (Lopez de Quinto et al., 2001), and its C-terminal fragment (Ct) also has a strong binding affinity for the FMDV IRES, suggesting its involvement in IRES activity. Interestingly, a fragment of the central region of eIF4GII interacts with the EMCV IRES through a non-canonical RRM region (Marcotrigiano et al., 2000). The strong correlation found between the eIF4G-IRES interaction and IRES activity in transfected cells demonstrates that eIF4G binding is an essential step in the recruitment of the translational machinery in vivo. During the last few years, it has become increasingly clear that many cellular translation initiation factors play an essential role in IRES-dependent initiation promoted by the FMDV, EMCV and HeV IRES elements (reviewed by Belsham and Jackson, 2000; Martinez-Salas et al., 2001). A summary of known protein interactions with viral IRES elements is presented in Table 1. The canonical initiation factors eIF4A, eIF4G, eIF2 and eIF3, are each required for 48S complex formation in a reconstituted 40S ribosome-binding assay with the EMCV or FMDV IRES elements (Pestova et al., 1996; Kolupaeva et al., 1998; Pilipenko et ai., 2000). The fragment of eIF4G that interacts with the EMCV and FMDV IRES (residues 630 to 1560) contains the binding sites for eIF3 and eIF4A (Gingras et ai., 1999). Accordingly, eIF4A stimulates binding of the central part of eIF4GI (amino acids 746 to 949) to the EMCV and FMDV IRES (Pilipenko et ai., 2000; Lomakin et ai., 2000). Consistent with the interaction of the eIF4G-Ct with these picornavirus IRES elements, internal initiation of translation promoted by these elements is highly efficient under conditions of eIF4G cleavage (Belsham and Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

311

Brangwyn, 1990; Martinez-Salas et al., 1993; Ohlmann et al., 1996; Roberts et al 1998; Lopez de Quinto and Martinez-Salas, 2000). The 3' region of the FMDV IRES has been shown to interact with the elF4B initiation factor (Meyer et al., 1995; Lopez de Quinto and Martinez-Salas, 2000; Stassinopoulos and Belsham, 2001). Sequence comparison of 24 isolates ofFMDV showed that domain 5 contains a conserved hairpin which is also shared with the EMCV IRES and a high affinity ligand for elF4B selected in vitro (Methot et al., 1996; Lopez de Quinto et aI., 2001). However, substitution of conserved residues within the hairpin structure did not impair FMDV IRES activity to the same extent as the mutations introduced into the A-bulge of domain 4, or the GNRA loop of domain 3. The mapping of the binding site for elF4B on the FMDV IRES to specific sequences within the apical loop of the conserved hairpin of domain 5 contrasts with earlier conclusions (Rust et al., 1999). These studies had analyzed transcripts that had large deletions or insertions in domain 4 which probably lead to large changes in the organization of the IRES structure. Moreover, Rust et ale (1999), on the basis of these results, probably mis-assigned the reduction of IRES activity to the lack of elF4B binding, since some of the mutants used must also have reorganized the binding site for eIF4G. In spite of the strong binding capacity of the FMDV IRES for eIF4B, mutants that are strongly impaired in elF4B binding were only reduced by 2-4 fold in their IRES activity (Lopez de Quinto et al., 2001). In agreement with this, the elF4B initiation factor only stimulates 48S complex formation on the EMCV IRES about two-fold (Pestova et al., 1996). Unlike the FMDV IRES, the HCV IRES element does not interact with elF4A or eIF4G, but it does bind elF3 (Buratti et al., 1998; Sizova et al., 1998; Kolupaeva et aI., 2000; Lopez de Quinto et al., 2001; Kieft et al., 2001). Results of toe-print analyses indicated that efficient 48S complex formation driven by the HCV IRES required eIF2-GTP/Met-tRNAi, elF3 and 40S ribosomal subunits (Pestova et aI., 1998). As a consequence of the use of a different mechanism to initiate translation, the HCV IRES does not seem to require elF4G to assemble a 48S initiation complex and direct contact between the HCV IRES and the 40S ribosomal subunit has been recently demonstrated (Spahn et al., 2001). Consistent with these results, it has been shown that dominant negative mutants of elF4A (which interfere with elF4F function) block picornavirus IRES activity and cap-dependent translation initiation but do not affect HCV IRES mediated translation (Pause et al., 1994; Pestova et al., 1998; Svitkin et al., 2001a). In summary, it seems that the EMCV and FMDV IRES elements use a similar strategy to interact with cellular translation initiation factors in order to promote internal initiation of translation, however this is markedly different from the strategy used by the HCV IRES. Studies on other picornavirus IRES elements are less advanced in this area but it has been suggested recently that the PV IRES also interacts with eIF4G, eIF4B and elF3 (Dchs et aI., 2003). It is also known that the activity of the PV IRES is blocked by dominant negative mutants of eIF4A (Pause et al., 1994). 4.3. RNA-Binding Proteins and FMDV IRES Organization

In the early studies on IRES elements, a number of proteins were found to associate with picornavirus IRES elements, e.g. PTB and La (reviewed in Belsham and Copyright © 2004 By Horizon Bioscience

\32

FMDV RNA Structure and Function

Sonenberg, 1996; 2000). Identification of these proteins as non-canonical initiation factors was particularly interesting since it seemed possible they could contribute to IRES tropism in different cell types, and hence determine viral spread within an infected animal. In this regard, poliovirus and rhinovirus IRES elements interact with a different set of proteins than bound by the IRES elements from EMCV or FMDV (see Table 1). Interestingly, a protein named unr (llpstream of N-ras) is unique in its capacity to interact with the rhinovirus and PV IRES elements (Hunt et al., 1999; Boussadia et al., 2003) whereas PTB appears to interact with all IRES elements tested (Luz and Beck, 1991; Hunt and Jackson, 1999). However, binding alone does not mean that interaction is necessary for activity. The requirement for PTB binding to the EMCV IRES seems to be conditional as it depends on subtle changes in sequence within the IRES or on the reporter gene (Kaminski and Jackson, 1998). In contrast to the Theiler's murine encephalitis virus (TMEV) and EMCV IRES elements, the FMDV IRES required binding to PTB and to the proliferation associated factor, ITAF45 , for 48S complex formation in vitro (Kolupaeva et al., 1996; Pilipenko et al., 2000). It is interesting to note that these proteins appear to bind to very similar sites on the RNA. Therefore, even closely related IRES elements (i.e. from EMCV and FMDV) that share secondary structure and primary sequence in essential regions behave differently in terms of functional RNA-protein association. The known IRES interacting proteins (unr, PTB, La, ITAF45 and PCBP2) contain several RNA binding motifs, display multiple sites of interaction with the IRES molecule and have the capacity to dimerize (Blyn et al., 1997; Craig et al., 1997; Hunt et al., 1999; Pilipenko et al., 2000; Conte et al., 2000; Kim et al., 2000). Hence, it has been suggested that these proteins act as RNA chaperones, directing or stabilizing the tertiary folding of the RNA. The EMCV and FMDV IRES elements seem to possess a modular organization, in which the different domains perform a precise function but none of them is active on its own. The 3' region (domains J-K or 4-5) mediates the interaction with eIFs required for 48S complex formation in vitro (Kolupaeva et al., 1998; Pilipenko et al., 2000). This region establishes RNA-eIF4G interactions which are essential for IRES activity in vivo (Lopez de Quinto and Martinez-Salas, 2000; Lopez de Quinto et al., 2001; Stassinopoulos and Belsham, 2001). The 5' and central regions (domains 1-2 and 3, or H and I) are probably involved in the organization of the IRES architecture, directing intramolecular RNA-RNA interactions (Ramos and Martinez-Salas, 1999). In agreement with this, the PTB protein that seems to act as an RNA chaperone has its main binding site near the 5' end of the IRES but it also interacts with sequences near the 3' end (Kolupaeva et ai., 1996; Luz and Beck, 1991). The tertiary structures of entire IRES elements are still unknown. Experimental evidence for tertiary structures, generated by RNA-RNA interactions, is available for portions of the IRES elements from FMDV (Ramos and Martinez-Salas, 1999), HCV (Wang et al., 1995; Kieft et al., 1999; Jubin et ai., 2000; Lafuente et al., 2002) and the intergenic region of insect picornavirus-like viruses (Kanamori et al., 2001). Moreover, the stem-loops IIId and IIIe of the HCV IRES, responsible for interaction with 40S ribosomal subunits (Spahn et al., 2001), has been shown recently to form a loop E motif, characterized by the presence of three non-canonical Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

331

Watson-Crick base-pairs (Lukavsky et ai., 2000; Klinck et al., 2000). At present, it is not known if IRES elements, other than from HCV (and the related pestiviruses), interact directly with the ribosome. The absence of conserved structural elements between unrelated IRES elements probably reflects differences in the strategies they use to interact with the translational machinery. Thus, considerable efforts are required to elucidate the structure of IRES elements before we understand how different IRES sequences accomplish the same function.

5. The Polyprotein Coding Region The major portion (ca.7000 nt) of the viral genome comprises the coding sequence for a polyprotein of about 2330 amino acids. This polyprotein is never observed within infected cells or within in vitro translation systems (e.g. rabbit reticulocyte lysate, RRL) since this protein is rapidly processed by virus-encoded proteases that are present within the polyprotein sequence (see Chapter 3, Ryan et al.). A large variety of precursors can be generated by alternative cleavage pathways (see Figure 1) and some of these precursors, as well as the mature products, may have distinct biological roles. The FMDV polyprotein yields 15 different mature proteins including the 2 forms of the Leader (L) protein and 3 different copies of VPg (3B). The function of some of these products is the subject of other chapters within this volume and we will consequently give only a brief outline of these and thus will concentrate on other issues. 5.1. Selection of the Initiation Site for Protein Synthesis

The sequencing of FMDV RNA (Forss et aI., 1984; Carroll et al., 1984) indicated that two in-frame AUG codons were present in the viral genome that could potentially act as initiation codons and these were 84 nt apart. This feature is conserved across all seven serotypes of the virus (Sangar et al., 1987). Both the initiation codons are preceded by a polypyrimidine tract. The arrangement of a polypyrimidine tract followed by an AUG codon is a general feature at the 3' end of picornavirus IRES elements (Meerovitch and Sonenberg, 1993). However, in the entero-/rhinovirus RNAs the polypyrimidine tract is upstream of an AUG codon that is not the initiation site but is believed to be the site of ribosome attachment (see Belsham and Jackson, 2000). It was demonstrated by Sangar et al. (1987) that on FMDV RNA both AUG codons are used as initiation sites for protein synthesis and thus 2 distinct forms of the Leader (L) protein, the first component of the polyprotein, are generated. These two species are termed Lab and Lb and differ in their N-termini. The production of multiple forms of the L protein was observed both within in vitro translation reactions and also within virus-infected cells although the relative usage of the two start sites varied between strains. As explained above, on capped cellular mRNAs the initiation codon is usually the first AUG encountered by the scanning ribosome as it migrates from the 5' terminus (but this can be influenced by the context of the AUG codon). In contrast IRES-directed translation initiation results in the positioning of the ribosome at an internal position within the RNA sequence and upstream elements (e.g. the presence of other AUG codons) will have no influence on the AUG codon selected. These properties need to be borne in mind when considering the experiments performed to explore the mechanism of start site selection on FMDV RNA. Copyright © 2004 By Horizon Bioscience

\34

FMDV RNA Structure and Function

Studies on EMCV RNA, that contains an IRES that is closely related to the FMDV element, demonstrated that ribosomes initiated translation at AUG-lIon the EMCV-R strain RNA without encountering AUG-I0 located just 8 nt upstream (Kaminski et ai., 1990). The Lab start site on FMDV RNA is located at the same position relative to the IRES element as AUG-II and so could be expected to be recognised as an initiation site. However, it is apparent that not all ribosomes initiate protein synthesis at this point since Lb is also made, often in excess over Lab. In an attempt to determine the mechanism of recognition of the 2 start sites on FMDV RNA, Belsham (1992) analysed the usage of these start sites under a variety of conditions. RNA transcripts including the two FMDV initiation codons, either preceded by the FMDV IRES element or not, were expressed within cells and it was observed that both the Lab and Lb start sites were used on each RNA. The Lb start site was recognised preferentially in both cases and the presence of the IRES resulted in greater usage of the Lb site than occurred when the initiation codons were recognised by scanning ribosomes (as in cap-dependent translation initiation on the transcripts lacking the IRES). When two additional AUG codons were introduced in-frame between the authentic initiation codons, it was found surprisingly that all 4 AUG codons were recognised, independently of whether translation initiation occurred by an IRES-dependent or cap-dependent mechanism. In any single initiation event, a ribosome will only start synthesis at one initiation codon and the results generated represent the accumulated collection of individual events. It seems that the 84 nt region between the two natural initiation codons has very unusual features. The most likely explanation given for the results was that the FMDV IRES directed ribosomes to bind initially to the RNA just upstream of the Lab start site, as for EMCV. The recognition of this first start site on FMDV RNA must be inefficient however, perhaps reflecting the relatively poor context of this AUG codon in the FMDV RNA, so that many ribosomes can then scan downstream until the next initiation codon is encountered (Belsham, 1992). Subsequently, Cao et ale (1995) demonstrated that the Lab initiation codon was not required for virus viability but found that modification of the Lb initiation codon was lethal. The basis for this result is not known since it has been shown that each of the known properties of the FMDV L protein (see below) are displayed by both the Lab and Lb species (Medina et al., 1993). A requirement for the Lb start site was also observed by Piccone et ale (1995a), who constructed viable mutants of FMDV that lacked the Leader protein coding region but found that transcripts lacking the Lb start site were non-infectious. These results have prompted a reexamination of the site selection mechanism. Lopez de Quinto and Martinez-Salas (1999) modified the context of the Lab start site and showed that improving the context did increase the efficiency of translation initiation at this site but surprisingly this had no apparent effect on the efficiency of initiation at the Lb start site. Hence, it was suggested that these two initiation codons could act independently. This view was supported by the observation that an antisense oligonucleotide, which annealed to the region of the Lab start site, blocked initiation at this site but had little effect on initiation at the Lb site. Thus, it may be that some ribosomes are directed by the FMDV IRES to bind to the RNA downstream of the Lab start site. However, this putative entry site must presumably be close to the Lab start site

Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

351

since the additional AUG codons introduced just 18 nt downstream of the Lab site were efficiently recognised during IRES-directed translation initiation (Belsham, 1992). It is interesting to note that different picornavirus IRES elements may direct ribosomes to the initiation codon with different degrees of precision (Ohlmann and Jackson, 1999). To analyse this process, chimeras (fused at the polypyrimidine tract) were made between the FMDV IRES and a region of the EMCV RNA around the initiation codon (AUG-II). It was shown that the FMDV IRES directed utilisation of AUG-II (the correct initiation site) and AUG-12 (12 bases downstream) with similar efficiency in vitro. In contrast, the EMCV IRES linked to its own RNA directed almost exclusive use of AUG-II. Furthermore, in the converse experiment, when the EMCV IRES was linked to the FMDV initiation codons, the Lab site was predominantly used, in contrast to the preferential use of the Lb site on natural FMDV RNA. Thus it appears that the EMCV IRES more stringently directs ribosomes to initiate translation at the position of AUG-II than does the FMDV IRES. In order to test for scanning of ribosomes on FMDV RNA, Poyry et ala (2001) introduced an iron response element (IRE) between the Lab and Lb start sites. When the iron regulatory protein (IRP) was bound to the IRE, it was found that translation initiation at the downstream Lb site was suppressed, a result consistent with the blockade of scanning. However, this suppression was more modest (only 3-4-fold) than that observed for the effect of IRP on transcripts containing a cap-proximal IRE on a capped mRNA (a 10-fold reduction was observed under these conditions). Thus these authors suggested that, although the model of lRESdirected ribosome attachment upstream of the Lab start site followed by ribosome scanning could account for the majority of initiation events, it appeared that some ribosomes access the downstream site (just for Lb production) by an alternative mechanism. This suggestion is consistent with the relative imprecision of AUG codon selection by the FMDV IRES observed by Ohlmann and Jackson (1999). It would be helpful to confirm the key observations of Ohlmann and Jackson (1999) and Poyry et ala (2001) within cells since the process of translation initiation is more stringently controlled within cells than occurs in vitro. In summary, it seems that the FMDV IRES may direct ribosome attachment to the viral RNA either just upstream or just downstream of the Lab initiation site. Ribosomes landing upstream of the Lab site can initiate protein synthesis at this point but some may fail to do so and can then scan along the RNA until the Lb site is reached. Ho\vever, ribosomes that land downstream of the Lab site can just migrate along the RNA to initiate translation at the Lb site. The reason why two initiation sites for protein synthesis are conserved across all FMDV isolates remains unknown. 5.2. The Leader Protease As indicated above, the use of two alternative initiation codons on the FMDV RNA results in the generation of two distinct forms of the Leader protein. Uniquely amongst the picornaviruses, the Leader (L) protein of aphthoviruses, including FMDV and equine rhinitis A virus, is a protease (Strebel and Beck, 1986; Hinton et al., 2002). The L protease is a member of the papain-like cysteine proteases and

Copyright © 2004 By Horizon Bioscience

\36

FMDV RNA Structure and Function

is unrelated to other picornavirus protease sequences. The active site residues have been identified (Roberts and Belsham, 1995; Piccone et al., 1995b) and the 3Dstructure of FMDV Lb has been determined (Guarne et al., 1998; see Chapter 3). Both the FMDV Lab and Lb proteins have been shown to cleave the LIP1 junction within the polyprotein (Figure 1), thus liberating the L protease from the rest of the molecule (Medina et aI., 1993). The protease can work in trans and probably also in cis (see Glaser et al., 2001). Both forms of the L protease also induce the cleavage of the translation initiation factor eIF4G (Devaney et al., 1988; Medina et al., 1993), a component of the cap-binding complex eIF4F (see above). The consequence of this cleavage is the inhibition of cap-dependent translation initiation and hence the loss of nearly all host-cell protein synthesis. The cleavage of elF4G requires a very low level of the L protease and can be observed to occur within an hour of addition of virus to cells (Belsham et al., 2000) and hence well before any virus encoded proteins can be detected directly. Indeed, replication-incompetent FMDV RNA transcripts that include the L coding sequence can still induce efficient elF4G cleavage when they are introduced into cells by electroporation, even though protein can only be produced from the input RNA (Belsham et al., 2000). A third activity of the L protease is to stimulate the activity of certain picornavirus IRES elements (Borman et al., 1997; Roberts et al., 1998; Hinton et al., 2002). There is controversy concerning whether this effect is a direct consequence of the cleavage of elF4G or whether the L protease also induces the cleavage of other proteins that modify IRES function (see Belsham and Jackson, 2000 for review). Recombinant Lb protease can cleave isolated elF4GI directly in vitro. A cleavage site has been identified on the C-terminal side of residue 674, (Kirchweger et al., 1994) which is close to that determined for the in vitro cleavage of eIF4GI by the rhinovirus 2A protease (C-terminal side of residue 681, Lamphear et ale (1993)). (Note that the locations of these cleavage sites are renumbered according to the system of Byrd et al, (2002) for the largest form of eIF4GI, termed eIF4Gla). However, in both cases, high levels of the proteases are required to achieve efficient elF4G cleavage in vitro compared to the very low levels of the proteases that are required within cells. Hence there are legitimate concerns about whether the cleavage of elF4G induced by the FMDV L and rhinovirus 2A proteases within cells is a direct effect or is mediated through cellular protease(s) (see Belsham and Jackson (2000) for review). It is noteworthy that the addition of eIF4E to eIF4G greatly enhances the efficiency of cleavage of elF4G by the 2A protease in vitro which may suggest that the conformation of the protein is important (Haghighat et aI., 1996). The identification of the cleavage site in elF4G that is generated by low levels of the Leader protease within FMDV-infected cells remains to be achieved. Such analyses should help to resolve the issue of whether the FMDV L induced cleavage of elF4G is achieved directly or not. The L protease is not essential for virus viability since the complete Lb coding sequence can be removed (Piccone et al., 1995a). This leaderless FMDV (lacking Lb) was attenuated in cattle (Brown et al., 1996), however, proof that the attenuation was reversed by re-introduction of the missing sequence was not obtained. It may be that the primary advantage to the virus of inducing the inhibition of host-cell protein synthesis lies in reducing the ability of the cell to mount an anti-viral response. It has been shown that the mRNAs encoding the alpha/beta

Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

371

interferons are induced within FMDV-infected cells, presumably this is a response to the presence of dsRNA (Chinsangaram et al., 1999; Kaufman, 2000). The shutoff of protein synthesis induced by the L protease will inhibit synthesis of the alpha/beta interferons within infected cells and hence reduce the cellular antiviral response (Chinsangaram et al., 1999). Interferons trigger a variety of responses within cells, these responses include an activation of the dsRNA-activated protein kinase (PKR, see review by Kaufman, 2000). This kinase phosphorylates the usubunit of eIF2 and hence blocks the initiation step of both host and viral protein synthesis. Thus, this process will result in a reduction in virus production within infected cells. The inability to block this host response could explain, at least in part, the apparent attenuation of the leaderless FMDV. It remains to be seen whether other mRNAs are induced in response to FMDV infection, the development of microarray technology clearly makes this a feasible approach. 5.3. The Capsid Precursor P1-2A The removal of the L protein from the N-terminus of the polyprotein reveals an N-terminal glycine residue on the Pl-2A capsid precursor (see Figure 1) that is present within the motif (GXXXS/T) required for recognition of proteins by the cellular myristoylation machinery. This modification is a common (but not universal) feature of the picornavirus capsid proteins. The addition of the C 18 lipid to the N-terminus of 1A (VP4) is important for the assembly and/or stability of the FMDV and PV capsids (Chow et al., 1987; Abrams et al., 1995). The processing of the FMDV Pl-2A precursor to lAB, 1C and 1D is achieved by the 3C protease (see below) and these products can self-assemble into empty capsid particles (Abrams et al., 1995). The cleavage of lAB to 1A and 1B normally occurs on encapsidation of the RNA genome but the precise mechanism of this cleavage is not known. The structure and properties of the FMDV capsid, including its ability to specifically interact with cellular receptors, are discussed elsewhere in this volume (see Chapter 4, Ferrer-Orta and Fita). The generation of the Pl-2A precursor clearly requires the loss of any linkage to the 2B protein and this process is mediated by the 2A sequence. However, the 2A sequence in FMDV is very short, just 19 amino acids. It seems unlikely that such a short peptide could possess protease activity in its own right. This peptide sequence is closely related to the C-terminal region of the cardiovirus 2A protein (note that the cardiovirus 2A proteins are much larger but do not contain any motifs characteristic of known proteases). It has been proposed that the FMDV 2A sequence may prevent the formation of the peptide bond at the 2A/2B junction rather than acting as a peptidase to break a bond that has been formed (see Chapter 3). 5.4. The P2 Proteins The FMDV 2BC precursor is processed to 2B and 2C by the 3C protease (Figure 1). Rather little is known about these P2 proteins. Within PV-infected cells, the P2 proteins are found within membrane-associated viral replication complexes (Bienz et ai., 1987; 1990). However, their functions are also rather poorly understood. The only apparent feature of the 2B protein is the presence of certain hydrophobic regions (van Kuppeveld et ai., 1996). The enterovirus 2B protein can increase cell permeability, as judged by the susceptibility to the translation inhibitor hygromycin

Copyright © 2004 By Horizon Bioscience

138

FMDV RNA Structure and Function

B (van Kuppeveld et al., 1997) and it blocks protein secretion (Doedens and Kirkegaard, 1995). However some mutants within 2B affect cell growth without affecting either of these functions, hence other activities for t~e 2B protein may still need to be defined (van Kuppeveld et al., 1997). The properties of the FMDV protein have not been established. The various picornavirus 2C protein sequences contain helicase and nucleotide binding motifs but no evidence for RNA helicase activity has ever been reported. The 2C protein is also the locus that determines the sensitivity of viral RNA replication to the inhibitor, guanidine (Saunders and King, 1982). Different strains of FMDV vary in their sensitivity to this agent. 5.5. The P3 Proteins The FMDV P3 precursor (Figure 1) is processed by the 3C protease to 3A, the three copies of the 3B peptide (VPg), the 3C protease and the 3D RNA polymerase plus various intermediates (e.g. 3CD). The 3A protein has hydrophobic sequences which serve to anchor it to membranes and this may be the means by which RNA replication is localised to membrane vesicles. It is thought that the 3A protein also serves to deliver the 3B peptides to the sites of RNA replication. Various strains of FMDV have been isolated that contain in-frame deletions within the 3A coding sequence and these strains are attenuated in cattle (O'Donnell et al., 2001) but remain pathogenic in pigs. This presumably reflects some differences in the interaction of 3A with cellular factors between the two species. Expression of FMDV 3A alone disrupted the Golgi apparatus of keratinocytes. It is interesting to note that some picornaviruses that disrupt Golgi function, e.g. PV, are extremely sensitive to the effect of brefeldin A (Maynell et al., 1992). However, in contrast, it has been observed that FMDV is rather insensitive to this agent (O'Donnell et ai., 2001), like EMCV (Iruzun et al., 1992). Brefeldin A modifies vesicle transport between the Golgi and the endoplasmic reticulum, and the different sensitivities of picornaviruses to this agent suggest that the different viruses recruit membranes to their replication complexes in different ways. The PV 3A protein has also been shown to block protein secretion (Doedens and Kirkegaard, 1995). Uniquely FMDV RNA encodes three similar but distinguishable copies of the 3B peptide. The coding sequences for these different peptides occur as a tandem array in the genome (see Figure 1). Priming of RNA synthesis requires 3B (VPg) and this explains the linkage of this peptide to the 5' terminus of both positive- and negative-sense RNA strands. Based on results obtained with the PV and HRV ere elements, it is now believed that the initial modification of VPg to VPgpU or VPgpUpU is achieved by the use of the 'ere' as a template (Paul et al., 2000; Gerber et ai., 2001). Each of the FMDV VPgs has been found attached to genomic RNA and thus each peptide is believed to be functionally equivalent (King et al., 1980). Even the closely related equine rhinitis A virus, the only member of the aphthovirus genus other than FMDV, only has a single copy of this peptide. By mutagenesis, deletion of one or more FMDV 3B sequences can be achieved but the mutant viruses replicate less efficiently than the wt virus (Falk et al., 1992). Indeed, deletion of 3B 3 alone destroyed virus viability but this appeared to result from a defect in polyprotein processing rather than a direct effect on RNA replication. However, overall, it is not clear why the presence of three different VPg sequences within the FMDV RNA enhances its replication efficiency (Falk et al., 1992).

Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

391

5.6. The 3C Protease

The FMDV 3C protease is responsible for most of the cleavages within the polyprotein coding sequence. It functions alone and, in contrast to the PV 3C protease (see Ypma-Wong et al., 1988), it does not require 3D sequences for any of its processing activities. The key catalytic residues of the FMDV 3C have been identified (Grubman et al., 1995) and the protease is a member of the trypsin-like family of serine proteases (except that the active serine is replaced by a cysteine, see Chapter 3, Ryan et al.). In addition to cleaving the viral polyprotein it is now established that the FMDV 3C also modifies certain cellular proteins. A histone H3 was shown to be cleaved by this protease (Falk et al., 1990) and more recently it has been shown that the 3C protease also cleaves the translation initiation factors elF4A and elF4GI within FMDV-infected cells (Belsham et aI., 2000). The precise location of the cleavage site generated by the 3C protease in elF4AI has been identified (Li et aI., 2001). The cleavage is specific for eIF4AI, since the closely related elF4AII (92% identical) is not modified. It has been suggested that the cleavage of eIF4AI, which will inactivate the protein (Li et al., 2001), may contribute to the decrease in the level of viral protein synthesis during the later phase of virus infection (Belsham et aI., 2000). The 3C protease cleaves elF4GI on the C-terminal side of the site generated by the expression of the L protease (Belsham et aI., 2000). This result is consistent with the loss of intact elF4GI within cells infected by the leaderless FMDV (Piccone et al., 1995a; Belsham et al., 2000). Sequential cleavage of elF4GI within FMDV-infected BHK cells by the L protease and then of the C-terminal cleavage product by the 3C protease has been observed, both cleavages are complete by 3h post-infection (GJB, unpublished results). Thus at the time of peak viral protein synthesis the form of elF4GI that supports IRES function is that generated by FMDV 3C. Later on during infection, further modification of eIF4GI, probably also generated by FMDV 3C, occurs and these cleavages may contribute to the decline in viral protein synthesis. There is no data on the precise locations of the FMDV 3C induced modifications of elF4GI in cells. 5.7. The 3D RNA Polymerase

The 3D protein is the RNA-dependent RNA polymerase. The replication of picornavirus RNA has two distinct aspects to it. To replicate the positive-sense genome, an antisense RNA has to be synthesised which then functions as the template for the production of new positive-sense infectious genomes (see Figure 3). Within infected cells, a large excess of positive strands accumulates over negative strands. Presumably this reflects a differential recognition of the negative-sense template over the positive strand template by the RNA polymerase and/or differences in the stability of the transcripts. The 3' terminus of the positivesense RNA (the poly(A) tail) and the negative-sense RNA template (antisense S-fragment) are very different in sequence. Thus, the nature of the recognition process by the RNA polymerase is clearly complex but is not yet defined. Interestingly the PV 3D molecule displays co-operativity in its polymerase activity in vitro (Pata et aI., 1995) and may normally function as a multimer. The crystal structure of the PV 3D polymerase has been determined and it is apparent that there is considerable interaction between adjacent molecules in the crystal (Hobson et

Copyright © 2004 By Horizon Bioscience

\40

FMDV RNA Structure and Function

Input virus

1

VPg

•

Protein synthesis Anti-host defense (L,+??)

(PItA + 3C)

Virions

~

~ or ~Viral proteins

~

Capsid assembly

+ A(n)

RNA replication

I.

Replication complex (3A, 38, 3D +??)

I

J

+

• • •

+

A(n)

:. +

A(n) A(n) A(n)

Figure 3. Distinct roles for positive-sense FMDV RNA. Following the entry of FMDV RNA into the cytoplasm of the cell, the RNA has to be translated to generate viral proteins required for viral RNA replication. Subsequently replicated RNA genomes can be used in translation, RNA replication or packaged into virions as illustrated.

al., 2001). Whether these properties are relevant to the activity of the protein within an infected cell remains to be determined. The 3CD molecule does not have RNA polymerase activity but does have RNA binding and protease activity. The PV 3CD product has been shown to bind to the cloverleaf structure at the 5' terminus of the PV genome (Gamarnik and Andino, 1998) and it is also required in the ere-dependent VPg uridylylation assay (Paul et al., 2000; Rieder et a/., 2000). PV 3CD is the protease required for the processing of the PV PI precursor whereas 3C alone can process the P2 proteins (Ypma-Wong et al., 1988). No specific requirement for the FMDV 3CD protein has been determined and as indicated above, FMDV 3C is sufficient for all FMDV polyprotein processing. 6. Structure and Function of the FMDV 3'UTR The 3' untranslated region of FMDV RNA consists of two components, a region of about 100 nt of heterogeneous sequence and the poly(A) tail. The poly(A) tail of picornavirus RNA is included within the genome of the virus, this contrasts with cellular mRNAs in which this sequence is added as a post-transcriptional modification. There is rather little information about the role of the FMDV RNA 3' UTR except that deletion of the unique (heterogeneous) sequence blocks infectivity (Saiz et aI, 2001). Recently, it has been shown that the FMDV 3' UTR sequence can stimulate the activity of the FMDV IRES (Lopez-de Quinto et al., 2002), furthermore, this effect is independent of the stimulation of IRES activity by poly(A) observed previously (Svitkin et aI., 200 1b). These results suggest further possible RNA-RNA interactions between the 5' and 3' UTRs or additional 'bridging' RNA-protein interactions. There is evidence for complex RNA-RNA interactions within the heterogeneous sequence of the PV and HRV 3' UTRs (Mirmomeni et al., 1997; Melchers et a/., Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

41\

1997) and also for protein interactions with this element (Mellits et al., 1998). However, surprisingly, it has also been shown that the unique 3' UTR sequence of PV RNA can be deleted without loss of virus viability (Todd et al., 1997) albeit that the virus replicated significantly less well than wt virus. In contrast, studies on the 3' UTR of cardiovirus RNA, have shown evidence for 3 stem-loop structures, one of which is essential for virus viability (Duque and Palmenberg, 2001). It is possible that other sequences within the coding sequence (but towards the 3' end of the genome) are required to provide the specificity of PV RNA recognition by the replication machinery that would seem to be required, especially by the mutants lacking the usual 3' UTR sequences. Studies have shown that the length of the poly(A) tail has an effect on the infectivity of PV RNA (Spector and Baltimore, 1974). This observation may reflect modification of the stability of the RNA or possibly a requirement for interactions between the 3' and 5' termini of the RNA potentially involving the poly(A) binding protein (PABP) which binds optimally to a sequence of at least 25 A residues. It should be noted that PABP interacts both with the translation initiation factor eIF4G (that binds directly to the FMDV IRES element, see above) and with the poly(rC) binding protein 2 that itself binds to the FMDV IRES (Stassinopoulos and Belsham, 2001) and presumably to the poly(C) tract. As indicated above, such interactions may serve to "circularise" the RNA and it may be that such interactions affect the efficiency of RNA translation, at least in the early stages of infection prior to cleavage of eIF4G (see Svitkin et al., 2001b). Cleavage of eIF4G by the FMDV proteases would disrupt the circularization of the RNA but also removes competition from the cellular capped mRNAs for the translation machinery. It has been suggested that circularisation of the PV RNA is required for RNA replication and is supposed to be achieved through the interaction of the poly(A) binding protein (bound to the 3' poly(A) tail) with the poly(rC) binding protein that binds to the cloverleaf structure at the 5' end of the genome (Herold and Andino, 2001). However, it is not entirely clear whether this would be achieved within the context of a cell containing many different polyadenylated mRNAs, each of which may have the poly(A) binding protein associated with it. As mentioned above, no protein interactions with the FMDV S-fragment have been identified. However, if the S-fragment does exist in a stable stem-loop structure, as predicted, then the 5' end of the RNA would be close to the poly(C) tract which is likely to be bound to the poly(rC) binding protein, potentially achieving the same effect as proposed for the PV cloverleaf. 7. Unsolved Issues There are many questions about picornavirus biology which remain to be answered and some of these areas are outlined below: 7.1. Interactions of Viral Proteins With Cellular Components

The interactions of the IRES elements with the cellular protein synthesis machinery provides a good insight into the complexity of virus -host cell interactions but we are still a long way from understanding the mechanism of IRES action. It is apparent that cellular proteins that interact with virus components can determine

Copyright © 2004 By Horizon Bioscience

\42

FMDV RNA Structure and Function

virus tropism. There are obvious examples, such as receptor molecules on the cell surface, but clearly any cellular component that is required for the virus to replicate (either for RNA replication or protein production) may vary in its availability within different cell types. Hence such factors may determine the competence of the cell to act as a host for the virus. These issues are important since they can determine the pathogenicity of the virus. 7.2. Compartmentalisation

About ten years ago, Molla et ale (1991), described a system for the in vitro synthesis of infectious PV. This system has allowed the requirements for virus assembly and RNA replication to be analysed. No equivalent system had been reported for other picornaviruses but recently Svitkin and Sonenberg (2003) have developed an in vitro replication system for EMCV and this holds out the possibility of achieving this with FMDV RNA. The ability to produce infectious virus within a cell-free system does suggest that the requirement for compartmentalisation of activities is limited. However, there are soluble and membrane-associated environments in this system and the membranes may create a protected environment. 7.3. RNA Packaging

No specific packaging signals have been identified in any picornavirus RNA. It has been suggested that PV RNA synthesis and packaging into virions are closely linked (Nugent et ai., 1999). However, there are several aspects of this process which remain to be determined. To what extent do the capsid proteins assemble prior to virion assembly? It is well known that empty capsid formation can occur in the absence of virion RNA but this mayor may not be a dead-end product. Are there discrete RNA replication complexes which are producing RNA genomes solely for packaging into virions? If so, how are these different from replication complexes that are producing RNA that will be translated and/or replicated?

8. Concluding Remarks Understanding of the structure and function of picornaviruses has grown considerably in recent years. However, the apparent simplicity of a picornavirus genome belies the great complexity of picornavirus biology. Moreover, different picornaviruses have their own individual properties that affect their interactions with the host cell. Therefore, a deep understanding of the consequences of FMDV genome organisation in co-ordinating the synthesis of viral components that ultimately leads to productive infection and virus spread demands a collaborative effort involving molecular biology, cell biology and immunology. 9. Acknowledgements We are grateful to M. Saiz and C. Gutierrez for helpful suggestions on the manuscript. Studies in the laboratory of E.M.S. were partially supported by grants PM98.0122 and BMC2002-00983, and an Institutional grant from Fundaci6n Ramon Areces. Studies in G.J.B. 's laboratory were supported by the BBSRC. Collaborative studies are supported by grant (01-0293) from INTAS.

Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

43 I

References Abrams C., KingA.M.Q. and Belsham G.J. 1995. Assembly offoot-and-mouth disease virus empty capsids synthesized by a vaccinia virus expression system. J. Gen. Virol. 76: 3089-3098. Barclay, W., Li, Q., Hutchinson, G., Moon, D., Richardson, A., Percy, N., Almond, J.W. and Evans, D.J. 1998. Encapsidation studies of poliovirus subgenomic replicons. J. Gen. Virol. 79: 1725-1734. Belsham, G.J. and Bostock, C.J. 1988. Studies on the infectivity of foot-and-mouth disease virus RNA using microinjection. J. Gen. Virol. 69: 265-274. Belsham, G.J. and Brangwyn, J.K. 1990. A region of the 5' non-coding region of foot-andmouth disease virus RNA directs efficient internal initiation of protein synthesis within cells; interaction with the role of the L protease in translational control. J. Virol. 64: 5389-5395. Belsham, G.J. 1992. Dual initiation sites of protein synthesis on foot-and-mouth disease virus RNA are selected following internal entry and scanning of ribosomes in vivo. EMBO J. 11: 1105-1110. Belsham, G.J. and Jackson, R.J. 2000. Translation initiation on picornavirus RNA. In: Translational Control of Gene Expression. Monograph 39. N. Sonenberg, J.W.B. Hershey and M.B. Mathews, eds. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. p869-900. Belsham, G.J. and Sonenberg, N. 1996. RNA-protein interactions in regulation of picornavirus RNA translation. Microbiol. Reviews 60: 499-511. Belsham, G.J. and Sonenberg, N. 2000. Picornavirus RNA translation: roles for cellular proteins. Trends in Microbiology 8: 330-335. Belsham, G.J., McInerney, G.M. and Ross-Smith, N. 2000. Foot-and-mouth disease virus 3C protease induces cleavage of translation initiation factors eIF4A and eIF4G within infected cells. J. Virol. 74: 272-280. Bienz, K., Egger, D., and Pasamontes, L. 1987. Association of polioviral proteins of the P2 genomic region with the viral replication complex and virus-induced membrane synthesis as visualized by electron microscopic immunocytochemistry and autoradiography. Virology 160: 220-226. Bienz, K., Egger, D., Troxler, M. and Pasamontes, L. 1990. Structural organization of poliovirus RNA replication is mediated by viral proteins of the P2 genomic region. J. Virol. 64: 1156-1163. Blyn, L.B., Towner, J.S., Semler, B.L., and Ehrenfeld, E. 1997. Requirement of poly(rC) binding protein 2 for translation of poliovirus RNA. J. Virol. 71: 6243-6246. Bolten, R., Egger, D., Gosert, R., Schaub, G., Landmann, L., and Bienz, K. 1998. Intracellular localization of poliovirus plus- and minus-strand RNA visualization by strand-specific fluorescent in situ hybridization. J. Virol. 72: 8578-8585. Boussadia, 0., Niepmann, M., Creancier, L., Prats, A.C., Dautry, F., and Jacquemin-Sablon, H. 2003. Unr is required in vivo for efficient initiation of translation from the internal ribosome entry sites of both rhinovirus and poliovirus. I. Virol. 77: 3353-3359. Borman, A.M., Le Mercier, P., Girard, M., and Kean, K.M. 1997. Comparison of picornaviral IRES-driven internal initiation in cultured cells of different origins. Nucl. Acids. Res. 25: 925-932. Brown, F., Newman, J., Stott, I., Porter, A., Frisby, D., Newton, C., Carey, N. and Fellner, P. 1974. Poly(C) in animal viral RNAs. Nature (London) 251: 342-344. Brown, C.C., Piccone, M.E., Mason, P.W., McKenna, T.S.C. and Grubman, M.J. 1996. Pathogenesis of wild-type and leaderless foot-and-mouth disease virus in cattle. J. Virol. 70: 5638-5641. Buratti, E., Tisminetzky, S., Zotti, M., and Baralle, F.E. 1998. Functional analysis of the interaction between HCV 5'UTR and putative subunits of eukaryotic translation initiation factor eIF3. Nuc!. Acids Res. 26: 3179-3187. Copyright © 2004 By Horizon Bioscience

\44

FMDV RNA Structure and Function

Byrd, M. P., Zamora, M., and Lloyd, R.E. 2002. Generation of multiple isoforms of eukaryotic translation initiation factor 4GI by use of alternate translation initiation codons. Mol. Cell. BioI. 22: 4499-4511. Cao, X.M., Bergmann, I.E., Fullkrug, R., and Beck, E. 1995. Functional analysis of the two alternative translation initiation sites of foot-and-mouth-disease virus. J. Virol. 69: 560-563. Carrasco, L. 1995. Modification of membrane permeability by animal viruses. Adv. Virus Res. 45: 61-112. Carroll, A.R., Rowlands, DJ., and Clarke, B.E. 1984. The complete nucleotide sequence of the RNA coding for the primary translation product of foot-and-mouth disease virus. Nucl. Acids Res. 12,2461-2472. Chinsangaram, J., Piccone, M.E., and Grubman, M.J. 1999.Ability of foot-and-mouth disease virus to form plaques in cell culture is associated with suppression of alpha! beta interferon J. Viral. 73: 9891-9898. Chow, M., Newman, J.F.E., Filman, D., Hogle, J.M., Rowlands, D.J. and Brown, F. 1987. Myristylation of picornavirus capsid protein VP4 and its structural significance. Nature 327: 482-486. Clark, A.T., Robertson, M.E.M., Conn, G.L. and Belsham, GJ. 2003. Conserved nucleotides within the J-domain of the encephalomyocarditis virus internal ribosome entry site are required for activity and for interaction with eIF4G. J. Virol. 77: 12441-12449. Clarke, B.E., Brown, A.L., Currey, K.M., Newton, S.E., Rowlands, D.J. and Carroll, A.R. 1987. Potential secondary and tertiary structure in the genomic RNA of FMDV. Nucl. Acids Res. 15: 7067-7079. Conte, M.R., Grune, T., Ghuman, J., Kelly G., Ladas, A., Matthews, S., and Curry, S. 2000. Structure of tandem RNA recognition motifs from polypyrimidine tract binding protein reveals novel features of the RRM fold. EMBO J. 19: 3132-3141. Costa, M., and Michel, F. 1997. Rules for RNA recognition of GNRA tetraloops deduced by in vitro selection: comparison with in vivo evolution. EMBO J. 16: 3289-3302. Craig, A.W.B., Svitkin, Y.V., Lee, H.S., Belsham, GJ., and Sonenberg, N. 1997 The La autoantigen contains a dimerization domain that is essential for enhancing translation. Mol. Cell BioI. 17: 163-169. Crawford, N.M., and Baltimore, D. 1983. Genome-linked protein VPg of poliovirus is present as free VPg and VPgpUpU in poliovirus-infected cells. Proc. Natl. Acad. Sci. USA. 80: 7452-7455. de la Torre, J.C., Martinez-Salas, E., Diez, J., Villaverde, A., Gebauer, F., Rocha, E., Davila, M., and Domingo, E. 1988. Coevolution of cells and viruses in a persistent infection of foot-and-mouth disease virus in cell culture. J. Virol. 62: 2050-2058. Devaney, M.A., Vakharia, V.N., Lloyd, R.E., Ehrenfeld, E., and Grubman, MJ. 1988. Leader protein of foot-and-mouth-disease virus is required for cleavage of the p220 component of the cap-binding protein complex. J. Virol. 62: 4407-4409. Diez, J., Hofner, M., Domingo, E., and Donaldson, A.I. 1990. Foot-and-mouth disease virus strains isolated from persistently infected cells cultures are attenuated for mice and cattle. Virus Res. 18: 3-8. Doedens, J.R, and Kirkegaard, K. 1995. Inhibition of cellular protein secretion by poliovirus proteins 2B and 3A. EMBO J. 14: 894-907. Doherty, E.A., Batey, R.T., Masquida, B., and Doudna, J.A. 2001. A universal mode of helix packing in RNA. Nat. Struct. BioI. 8: 339-343. Domingo, E., Mateu, M.G., Escarmis, E., Martinez-Salas, E., Andreu, D., Giralt, E., Verdaguer, N. and Fita, I. 1996. Molecular evolution of Aphthoviruses. Virus Genes 11: 197-207. Drew, J., and Belsham, G.J. 1994. Trans-complementation by RNA of defective foot-andmouth disease virus internal ribosome entry site elements. J. Virol. 68: 697-703. Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

451

Duke, G.M., and Palmenberg, A.C. 1989. Cloning and synthesis of infectious cardiovirus RNAs containing short, discrete poly(C) tracts. J. Virol. 63: 1822-1826. Duke, G.M., Osorio, J.E., and Palmenberg, A.C. 1990. Attenuation of mengo-virus through genetic-engineering of the 5' noncoding poly(C) tract. Nature 343: 474-476. Duque, H., and Palmenberg, A.C. 2001. Phenotypic characterization of three phylogenetically conserved stem-loop motifs in the mengovirus 3' untranslated region. J. Virol. 75: 3111-3120. Escarmis, C., Toja, M., Medina, M., and Domingo, E. 1992. Modifications of the 5' untranslated region of foot-and-mouth disease virus after prolonged persistence in cell culture. Virus Res. 26: 113-125. Escarrnis, C., Dopazo, J., Davila, M., Palma, E.L. and Domingo, E. 1995. Large deletions in the 5'-untranslated region of foot-and-mouth-disease virus of serotype-C. Virus Res. 35: 155-167. Falk, M.M., Grigera, P.R., Bergmann, I.E., Zibert, A., Multhaup, G., and Beck, E. 1990. Foot-and-mouth-disease virus protease-3C induces specific proteolytic cleavage of host-cell histone-H3. J. Virol. 64: 748-756. Falk, M.M., Sobrino, F., and Beck, E. 1992. VPg-gene amplification correlates with infective particle formation in foot-and-mouth-disease virus J. Virol. 66: 2251-2260. Femandez-Miragall, 0., and Martinez-Salas, E. 2003. Structural organization of a viral IRES depends on the integrity of the GNRA motif. RNA 9: 1333-1344. Follett, E.A.C., Pringle, C.R., and Pennington, T.H. 1975. Virus development in enucleate cells: echovirus, poliovirus, pseudorabies virus, reovirus, respiratory syncytial virus and Semliki Forest virus. J. Gen. Virol. 26: 183-196. Forss, S., Strebel, K., Beck, E. and Schaller, H. 1984. Nucleotide-sequence and genome organization of foot-and-mouth-disease virus. Nucl. Acids Res.12: 6587-6601. Gamarnik, A.V. and Andino, R. 1997. Two functional complexes formed by KH domain containing proteins with the 5' noncoding region of poliovirus RNA. RNA 3: 882-892. Gamarnik, A.V., and Andino, R. 1998. Switch from translation to RNA replication in a positive-stranded RNA virus. Genes Dev. 12: 2293-2304. Gerber, K., Wimmer, E. and Paul, A.V. 2001. Biochemical and genetic studies of the initiation of human rhinovirus 2 RNA replication: identification of a cis-replicating element in the coding sequence of 2Apro. J. Virol. 75: 10979-10990. Gingras, A-C., Svitkin, Y., Belsham, GJ., Pause, A., and Sonenberg N. 1996. Activation of the translational suppressor 4E-BPI following infection with encephalomyocarditis virus and poliovirus. Proc. Natl. Acad. Sci. USA 93: 5578-5583. Gingras, A.C., Raught, B. and Sonenberg, N. 1999. eIF4 initiation factors: effectors of mRNA recruitment to ribosomes and regulators of translation. Ann. Rev. Biochem. 68: 913-963. Glaser, W., Cencic, R., and Skem, T. 2001. Foot-and-mouth disease virus Leader proteinase: involvement of C-terminal residues in self-processing and cleavage of eIF4GI. J. BioI. Chern. 276: 35473-35481. Goodfellow, I.,Chaudhry, Y., Richardson, A., Meredith, J., Almond, J.W., Barclay, W. and Evans, D.J. 2000. Identification of a cis-acting replication element within the poliovirus coding region. J. Virol. 74: 4590-4600. Gradi, A., Svitkin, Y.V., Imataka, H., and Sonenberg, N. 1998. Proteolysis of human eukaryotic translation initiation factor eIF4GII, but not not eIF4GI, coincides with the shutoff of host protein synthesis after poliovirus infection. Proc. Natl. Acad. Sci. USA. 95: 11089-11094. Grubman, M.J., Zellner, M., Bablanian, G., Mason, P.W. and Piccone, M.E. 1995. Identification of the active-site residues of the 3C proteinase of foot-and-mouth-disease virus. Virology 213: 581-589.

Copyright © 2004 By Horizon Bioscience

146

FMDV RNA Structure and Function

Guarne, A., Tormo, J., Kirchweger, R., Pfistermueller, D., Fita, I., and Skern, T. 1998. Structure of the foot-and-mouth disease virus leader protease: a papain-like fold adapted for self-processing and eIF4G recognition. EMBO J. 17: 7469-7479. Haghighat, A., Svitkin, Y., Novoa, I., Kuechler, E., Skern, T. and Sonenberg, N. 1996. The eIF4G-eIF4E complex is the target for direct cleavage by the rhinovirus 2A proteinase. J. Virol. 70: 8444-8450. Hahn, H., and Palmenberg. A.C. 1995. Encephalomyocarditis viruses with short poly(C) tracts are more virulent than their mengovirus counterparts. J. Virol. 69: 2697-2699. Harris, T.J.R. and Brown, F. 1977. Biochemical analysis of a virulent and an avirulent strain offoot-and-mouth disease virus. 1. Gen. Virol. 34: 87-105. Hellen, C.V.T., and Wimmer, E. 1995. Translation of encephalomyocarditis virus RNA by internal ribosomal entry. Curro Top. Microbiol. Immunol. 203: 31-63. Hellen, C.V.T. and Sarnow, P. 2001. Internal ribosome entry sites in eukaryotic mRNA molecules. Genes Dev. 15: 1593-612. Herold, J., and Andino, R. 2001. Poliovirus RNA replication requires genome circularization through a protein-protein bridge. Mol. Cell 7: 581-591. Hershey, J.W.B. and Merrick, W.C. 2000. Pathway and mechanism of initiation of protein synthesis. In: Translational Control of Gene Expression. N. Sonenberg, J.W.B. Hershey and M.B. Mathews, eds. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. p33-88. Hinton, T., Ross-Smith, N., Warner, S., Belsham, G.J. and Crabb, B. 2002. Conservation of Land 3C proteinase activities across distantly related aphthoviruses. J. Gen. Virol. 83: 3111-3121. Hobson, S.D., Rosenblum, E.S., Richards, O.C., Richmond, K., Kirkegaard, K., and Schultz S.C. 2001. Oligomeric structures of poliovirus polymerase are important for function. EMBO J. 20: 1153-1163. Honda, M., Beard, M.R., Ping, L-H., and Lemon, S.M. 1999. A phylogenetically conserved stem-loop structure at the 5' border of the internal ribosome entry site of hepatitis C virus is required for cap-independent viral translation. J. Virol. 73: 1165-1174. Hunt, S.L., and Jackson, R.I. 1999. Polypyrimidine-tract binding protein (PTB) is necessary, but not sufficient, for efficient internal initiation of translation of human rhinovirus-2 RNA. RNA. 5: 344-359. Hunt, S.L., Hsuan, J.J., Totty, N., and Jackson, R.I. 1999. unr, a cellular cytoplasmic RNAbinding protein with five cold-shock domains, is required for internal initiation of translation of human rhinovirus RNA. Genes Dev. 13: 437-448. Imataka, H., Gradi, A. and Sonenberg N. 1998. A newly identified N-terminal amino acid sequence of human eIF4G binds poly(A)-binding protein and functions in poly(A)dependent translation. EMBO J. 17: 7480-7489. Irurzun, A., Perez, L., and Carrasco, L. 1992. Involvement of membrane traffic in the replication of poliovirus genomes: effects of brefeldin A. Virology 191: 166-175. Jang, S.K., Krausslich, H.G., Nicklin, M.J., Duke, G.M., Palmenberg, A.C., and Wimmer, E. 1998. A segment of the 5' nontranslated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation. 1. Virol. 62: 2636-2643. Jubin, R., Vantuno, N.E., Kieft, J.S., Murray, M.G., Doudna, J.A., Lau, J.Y.N., and Baroudy B.M. 2000. Hepatitis C virus internal ribosome entry site (IRES) stem loop illd contains a phylogenetically conserved GGG triplet essential for translation and IRES folding. J. Virol. 74: 10430-10437. Kaminski, A., Howell, M.T. and Jackson, R.J. 1990. Initiation of encephalomyocarditis virus RNA translation: the authentic initiation site is not selected by a scanning mechanism. EMBO J. 9: 3753-3759. Kaminski, A., Belsham, G.J., and Jackson, R.I. 1994. Translation of encephalomyocarditis virus RNA: parameters influencing the selection of the internal initiation site. EMBO J. 13: 1673-1681. Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

471

Kaminski, A, and Jackson, RJ. 1998. The polypyrimidine tract binding protein (PTB) requirement for internal initiation of translation of cardiovirus RNAs is conditional rather than absolute. RNA 4: 626-638. Kanamori, Y. and Nakashima, N. 2001. A tertiary structure model of the internal ribosome entry site (IRES) for methionine-independent initiation of translation. RNA 7: 266274. Kaufman, RJ. 2000. The double-stranded RNA-activated protein kinase PKR. In: Translational Control of Gene Expression. Monograph 39. N. Sonenberg, J.W.B. Hershey and M.B. Mathews, eds. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. p503-527. Kieft, J.S., Zhou, K., Jubin, R., Murray, M.G., Lau, J.Y. and Doudna, J.A. 1999. The hepatitis C virus internal ribosome entry site adopts an ion-dependent tertiary fold. J. Mol. BioI. 292: 513-529. Kieft, J.S., Zhou, K., Jubin, R., and Doudna, J.A. 2001. Mechanism of ribosome recruitment by hepatitis C IRES RNA. RNA 7: 194-206. Kim, lH., Hahm, B., Kim, Y.K., Choi, M. and Jang, S.K. 2000. Protein-protein interaction among hnRNPs shuttling between nucleus and cytoplasm. J. Mol. BioI. 298: 395-405. King,A.M.Q., Sangar, D.V., Harris, T.1.R. and Brown F. 1980. Heterogeneity of the genomelinked protein of FMDV. J. Virol. 34: 627-634. Kirchweger, R., Ziegler, E., Lamphear, B.1., Waters, D., Liebig, H.D., Sommergruber, W., Sobrino, F., Hohenadl, C., Blaas, D., Rhoads, R.E. and Skern, T. 1994. Foot-and-mouth disease virus leader proteinase: Purification of the Lb form and determination of its cleavage site on e1F-4 gamma. J. Virol. 61: 2711-2718. Klinck, R., Westhof, E., Walker, S., Afshar, M., Collier, A. and Aboul-Ela, F. 2000. A potential RNA drug target in the hepatitis C virus internal ribosomal entry site. RNA 6: 1423-1431. Kolupaeva, V.G., Hellen, C.V.T., and Shatsky, LN. 1996. Structural analysis of the interaction of the pyrimidine tract-binding protein with the internal ribosomal entry site of encephalomyocarditis virus and foot-and-mouth disease virus RNAs. RNA 2: 1199-1212. Kolupaeva, V.G., Pestova, T.V., Hellen, C.V.T. and Shatsky LN. 1998. Translation eukaryotic initiation factor 4G recognizes a specific structural element within the internal ribosome entry site of encephalomyocarditis virus RNA. J. BioI. Chern. 273: 18599-18604. Kolupaeva, V.G., Pestova, T.V. and Hellen, C.V.T. 2000. An enzymatic footprinting analysis of the interaction of 40S ribosomal subunits with the internal ribosomal entry site of hepatitis C virus. J. Virol 74: 6242-6250. Kolupaeva, V.G., Lomakin, LB., Pestova, T.V., and Hellen, C.V.T. 2003. Eukaryotic initiation factors 4G and 4A mediate conformational changes downstream of the initiation codon of the encephalomyocarditis virus internal ribosome entry site. Mol. Cell. BioI. 23: 687-698. Kozak, M. 1989. The scanning model for translation: an update. J. Cell BioI. 108: 229-241. Kuhn, R., Luz, N., and Beck, E. 1990. Functional analysis of the internal translation initiation site of foot-and-mouth disease virus. J. Virol. 64: 4625-4631. Lafuente, E., Ramos, R., and Martinez-Salas, E. 2002. Long-range RNA-RNA interactions between distant regions of the hepatitis C virus internal ribosome entry site element. J. Gen. Virol. 83: 1113-1121. Lamphear, B.1., Yan, R.Q., Yang, F., Waters, D., Liebig, H.D., Klump, H., Kuechler, E., Skern, T. and R.E. Rhoads. 1993. Mapping the cleavage site in protein synthesis initiation factor eIF-4y of the 2A proteases from human coxsackievirus and rhinovirus. J. BioI. Chern. 268: 19200-19203.

Copyright © 2004 By Horizon Bioscience

148

FMDV RNA Structure and Function

Li, W., Ross-Smith, N., Proud, C.G. and Belsham, G.J. 2001. Cleavage of translation initiation factor 4AI (eIF4AI) but not eIF4AII by foot-and-mouth disease virus 3C protease: determination of the eIF4AI cleavage site. FEBS Lett. 507: 1-5. Lobert, P.-E., Escriou, N., Ruelle, J., and Michiels, T. 1999. A coding RNA sequence acts as a replication signal in cardioviruses. Proc. Natl. Acad. Sci. DSA 96: 11560-11565. Lomakin, LB., Hellen, C.D.T. and Pestova, T.V. 2000. Physical association of eukaryotic initiation factor 4G (eIF4G) with eIF4A strongly enhances binding of eIF4G to the internal ribosomal entry site of encephalomyocarditis virus and is required for internal initiation of translation. Mol. Cell. BioI. 20: 6019-6029. Lopez de Quinto, S., and Martinez-Salas, E. 1997. Conserved structural motifs located in distal loops of aphthovirus internal ribosome entry site domain 3 are required for internal initiation of translation. J. Virol. 71: 4171-4175. Lopez de Quinto, S., and Martinez-Salas, E. 1999. Involvement of the aphthovirus RNA region located between the two functional ADGs in start codon selection. Virology 255: 324-336. Lopez de Quinto, S. and Martinez-Salas, E. 2000. Interaction of the eIF4G initiation factor with the aphthovirus IRES is essential for internal translation initiation in vivo. RNA 6: 1380-1392. Lopez de Quinto, S., Lafuente, E. and Martinez-Salas, E. 2001. IRES interaction with translation initiation factors: functional characterization of novel RNA contacts with eIF3, eIF4B, and eIF4GIL RNA 7: 1213-1226. Lopez de Quinto, S., Saiz, M., de la Morena, D., Sobrino, F. and Martinez-Salas, E. 2002. IRES-driven translation is stimulated separately by the FMDV 3' NCR and poly(A) sequences. Nucl. Acids Res. 30: 4398-4405. Lukavsky, P.J., Otto, G.A., Lancaster, A.M., Sarnow, P. and Puglisi, J.D. 2000. Structures of two RNA domains essential for hepatitis C virus internal ribosome entry site function. Nat. Struct. BioI. 7: 1105-1110. Luz, N., and Beck, E. 1991. Interaction of a cellular 57-kilodalton protein with the internal translation initiation site of foot-and-mouth disease virus. J. Virol. 65: 6486-6494. Marcotrigiano, J., Lomakin, LB., Sonenberg, N., Pestova, T.V., Hellen, C.D.T. and Burley S.K. 2000. A conserved HEAT domain within eIF4G directs assembly of the translation initiation machinery. Mol. Cell 7: 193-203. Martin, L.R. and Palmenberg, A.C. 1996. Tandem mengovirus 5' pseudoknots are linked to viral RNA synthesis, not poly(C)-mediated virulence. J. Virol. 70: 8182-8186. Martinez-Salas, E., Saiz, J.C., Davila, M., Belsham, GJ. and Domingo, E. 1993. A single nucleotide substitution in the internal ribosome entry site of foot-and-mouth disease virus leads to enhanced cap-independent translation in vivo. J. Virol. 67: 3748-3755. Martinez-Salas, E., Regalado, M.P. and Domingo, E. 1996. Identification of an essential domain for internal initiation of translation in the aphthovirus IRES, and implications for viral evolution. J. Virol. 70: 992-998. Martinez-Salas, E., Ramos, R., Lafuente, E., and Lopez de Quinto, S. 2001. Functional interactions in internal translation initiation directed by viral and cellular IRES elements. J. Gen. Virol. 82: 973-984. Martinez-Salas, E., Lopez de Quinto, S., Ramos, R. and Fernandez-Miragall, O. 2002. IRES elements: features of the RNA structure contributing to their activity. Biochimie. 84: 755-763. Mason, P.W., Bezborodova, S.V. and Henry, T.M. 2002. Identification and characterization of a cis-acting replication element (ere) adjacent to the IRES offoot-and-mouth disease virus. J. Virol. 76: 9686-9694. Maynell, L. A., Kirkegaard, K. and Klymkowsky, M.W. 1992. Inhibition of poliovirus RNA synthesis by brefeldin A. J. Virol. 66: 1985-1994.

Copyright © 2004 By Horizon Bioscience

Be/sham and Martinez-Sa/as

491

McInerney, G.M., King, A.M.Q., Ross-Smith, N. and Belsham GJ. 2000. Replication competent foot-and-mouth disease virus RNAs lacking capsid coding sequences. J. Gen. Viral. 81: 1699-1702. McKnight, K.L. and Lemon.S.M. 1996. Capsid coding sequence is required for efficient replication of human rhinovirus-14 RNA. J. Viral. 70: 1941-1952. McKnight, K. L. and Lemon, S.M. 1998. The rhinovirus type 14 genome contains and internally located RNA structure that is required for viral replication. RNA 4: 15691584. Medina, M., Domingo, E., Brangwyn, J.K. and Belsham, G.J. 1993. The two species of the foot-and-mouth disease virus leader protein, expressed individually, exhibit the same activities. Virology 194: 355-359. Meerovitch, K., Nicholson, R., Sonenberg, N. 1991. In vitro mutational analysis of cisacting RNA translational elements within the poliovirus type 2 5' untranslated region. J. Viral. 65: 5895-5901. Meerovitch, K. and Sonenberg, N. 1993. Internal initiation of picornavirus RNA translation. Seminars Virol. 4: 217-227. Melchers, WJ.G., Hoenderop, J.G.J., Bruins Slot, HJ., Pleij, C.W.A., Pilipenko, E.V., Agol, V.I. and Galama, J.M.D. 1997. Kissing of the two predominant hairpin loops in the coxsackie B virus 3' untranslated region is the essential structural feature of the origin of replication required for negative-strand RNA synthesis. J. Virol. 71: 686-696. Mellits, KH., Meredith, J.M., Rohll, J.B., Evans, DJ. and Almond, J.W. 1998. Binding of a cellular factor to the 3' untranslated region of the RNA genomes of entero- and rhinoviruses plays a role in virus replication. J. Gen. Virol. 79: 1715-1723. Methot, N., Pickett, G., Keene, J.D. and Sonenberg, N. 1996. In vitro RNA selection identifies RNA ligands that specifically bind to eukaryotic translation initiation factor 4B: the role of the RNA remotif. RNA 2: 38-50. Meyer, K., Petersen, A., Niepmann, M., and Beck, E. 1995. Interaction of eukaryotic initiation factor eIF-4B with a picornavirus internal translation initiation site. J. Virol. 69: 2819-2824. Mirmomeni, M.H., Hughes, P.J. and Stanway, G. 1997. An RNA tertiary structure in the 3' untranslated region of enteroviruses is necessary for efficient replication J Viral. 71: 2363-2370. MalIa, A., Paul, A.V. and Wimmer E. 1991. Cell-free, de novo synthesis of poliovirus. Science 254: 1647-1651. Murray, K.E., Roberts, A.W. and Barton, DJ. 2001. Poly(rC) binding proteins mediate poliovirus mRNA stability. RNA 7: 1126-1141. Nugent, C.I., Johnson, K.L., Samow, P. and Kirkegaard, K. 1999. Functional coupling between replication and packaging of poliovirus replicon RNA. J. Virol. 73: 427-435. Ochs, K., Zeller, A., Saleh, L., Bassili, G., Song, Y., Sonntag, A., and Niepmann, M. 2003. Impaired binding of standard initiation factors mediates poliovirus translation attenuation. J. Virol. 77: 115-122. O'Donnell, V.K., Pacheco, J.M., Henry, T.M. and Mason, P.W. 2001. Subcellular distribution of the foot-and-mouth disease virus 3A protein in cells infected with viruses encoding wild-type and bovine-attenuated forms of 3A Virology 287: 151-162. Ohlmann, T., Rau, M., Pain, V.M. and Morley, S.J. 1996. The C-terminal domain of eukaryotic protein synthesis initiation factor (elF) 4G is sufficient to support capindependent translation in the absence of eIF4E. EMBO J. 15: 1371-1382. Ohlmann T., and Jackson, R.J. 1999. The properties of chimeric picornavirus lRESes show that discrimination between internal translation initiation sites is influenced by the identity of the IRES and not just the context of the AUG codon. RNA 5: 764-778. Parsley, T.B., Towner, J.S., Blyn, L.B., Ehrenfeld, E. and Semler, BL. 1997. Poly (rC) binding protein 2 forms a ternary complex with the 5'-terminal sequences of poliovirus RNA and the viral3CD proteinase. RNA 3: 1124-1134. Copyright © 2004 By Horizon Bioscience

\50

FMDV RNA Structure and Function

Pata, J.D., Schultz, S.C. and Kirkegaard, K. 1995. Functional oligomerization of poliovirus RNA-dependent RNA-polymerase. RNA 1: 466-477. Paul, A.V., Rieder, E., Kim, D.W., van Boom, J.H. and Wimmer, E. 2000. Identification of an RNA hairpin in poliovirus RNA that serves as the primary template in the in vitro uridylylation ofVPg. J. Virol. 74: 10359-10370. Pause, A., Methot, N., Svitkin, Y.V., Merrick, W.C., and Sonenberg. 1994. Dominant negative mutants of mammalian translation initiation factor eIF-4A define a critical role for eIF-4F in cap-dependent and cap-independent initiation of translation. EMBO J. 13: 1205-1215. Pelletier, J., and Sonenberg, N. 1998. Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature 334: 320-325. Pestova, T.V., Hellen, C.V.T. and Shatsky, I.N. 1996. Canonical eukaryotic initiation factors determine initiation of translation by internal ribosomal entry. Mol. Cell. BioI. 16: 6859-6869. Pestova, T.V., Shatsky, LN., Fletcher, S.P., Jackson, R.I. and Hellen, C.V.T. 1998. A prokaryotic-like mode of cytoplasmic eukaryotic ribosome binding to the initiation codon during internal translation initiation of hepatitis C and classical swine fever virus RNAs. Genes Dev 12: 67-83. Piccone, M.E., Rieder, E., Mason, P.W. and Grubman, M.J. 1995a. The foot-and-mouthdisease virus leader proteinase gene is not required for viral replication. J. Virol. 69: 5376-5382. Piccone, M.E., Zellner, M., Kumosinski, T.F., Mason, P.W. and Grubman, M.I. 1995b. Identification of the Active-Site Residues of the L-Proteinase of Foot-and-MouthDisease Virus. J. Virol. 69: 4950-4956. Pilipenko, E.V., Gmyl, A.P., Maslova, S.V., Svitkin, Y.V., Sinyakov, A.N. and Agol, V.L 1992. Prokaryotic-like cis elements in the cap-independent internal initiation of translation on picornavirus RNA. Cell. 68: 119-131. Pilipenko, E.V., Pestova, T.V., Kolupaeva, V.G., Khitrina, E.V., Poperechnaya, A.N., Agol, V.L and Hellen, C.V.T. 2000. A cell cycle-dependent protein serves as a templatespecific translation initiation factor. Gen. Dev. 14: 2028-2045. Poyry, T.A.A., Hentze, M.W. and Jackson, R.J. 2001. Construction of regulatable picornavirus IRESes as a test of current models of the mechanism of internal translation initiation. RNA 7: 647-660. Ramos, R., and Martinez-Salas, E. 1999. Long-range RNA interactions between structural domains of the aphthovirus internal ribosome entry site (IRES). RNA 5: 1374-1383. Rieder, E., Bunch, T., Brown, F. and Mason, P.W. 1993. Genetically-engineered foot-andmouth-disease viruses with poly(C) tracts of 2 nucleotides are virulent in mice. J. Viro!. 67: 5139-5145. Rieder, E., Paul, A.V., Kim, D.W., van Boom, J.H. and Wimmer, E. 2000. Genetic and biochemical studies of poliovirus cis-acting replication element cre in relation to VPg

uridylylation. J. Virol. 74: 10371-10380. Roberts, P. and Belsham, G.I. 1995. Identification of critical amino acids within the footand-mouth disease virus Leader protein, a cysteine protease. Virology 213: 140-146. Roberts, L.O. and Belsham G.I. 1997. Complementation of defective picornavirus internal ribosome entry site (IRES) elements by the coexpression of fragments of the IRES. Virology 227: 53-62. Roberts, L.O., Seamons, R.A. and Belsham, G.J. 1998. Recognition of picornavirus internal ribosome entry sites within cells; influence of cellular and viral proteins. RNA 4: 520529. Robertson, M.E., Seamons, R.A. and Belsham, G.I. 1999. A selection system for functional internal ribosome entry site (IRES) elements: analysis of the requirement for a conserved GNRA tetraloop in the encephalomyocarditis virus IRES. RNA 5: 11671179. Copyright © 2004 By Horizon Bioscience

Copyright © 2004 By Horizon Bioscience

152

FMDV RNA Structure and Function

Witwer, C., Rauscher, S., Hofacker, I.L. and Stadler, P.F. 2001. Conserved RNA secondary structures in Picornaviridae genomes. Nucl. Acids Res. 29: 5079-5089. Ypma-Wong, M.F., Dewalt, P.G., Johnson, V.H., Lamb, J.G., and Semler, B.L. 1988. Protein 3CD is the major poliovirus proteinase responsible for the cleavage of the PI capsid precursor. Virology 166: 265-270.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 3 Foot-and-Mouth Disease Virus Proteinases

Martin D. Ryan, Michelle L.L. Donnelly, Mike Flint, Vanessa M. Cowton, Garry Luke, Lorraine E. Hughes, Caroline Knox and Pablo de Felipe

Abstract Foot-and-mouth disease virus encodes all of its proteins in the form of a polyprotein. The full-length translation product (some 2,330 amino acids) is not observed within infected cells, however, due to processing of the polyprotein. The polyprotein undergoes three co-translational, intramolecular, or 'primary', cleavages mediated by the virus-encoded proteinases Land 3C, and a short oligopeptide sequence (2A). 2A-mediated 'cleavage' is now thought to be a translational effect: an unusual ribosome 'skipping' activity. The polyprotein primary cleavage products then undergo 'secondary' proteolytic processing by a combination of inter- and intramolecular cleavages to produce the mature processing products. The aphthoviruses are unique in possessing a proteinase (Lpro) at the N-terminus of the polyprotein. The Land 3C proteinases serve not only to cleave the virus polyprotein, but to degrade certain host-cell proteins thereby greatly enhancing virus replication. The strategy of encoding proteins as polyproteins comprising virus-encoded proteinases lies, therefore, at the core of the replication strategy of these viruses. 1. Introductory Overview The first studies on what we now refer to as the 'proteome' of picornavirusinfected cells created a conundrum: estimations of the coding capacity required to account for all of the virus-specific proteins observed within infected cells far exceed even the largest estimates of the genome. This puzzle was resolved by the observation that polyprotein precursors were proteo]ytically processed to produce a multiplicity of proteins: 'intermediate' forms and 'mature' products (Summers and Maizel, 1968). We now know that some of these processing 'intermediates' are, in fact, stable end products and may have biochemical properties that are distinct from those of their component parts. The development of translation systems in vitro greatly facilitated the analysis of polyprotein processing. The use of these systems provided the first proofs that proteolytic processing of picornavirus polyproteins was mediated by virusencoded, rather than cellular, proteinases. Such systems also proved to be useful in mapping the proteolytic domains of picornavirus polyproteins and, indeed, remain a po\verful tool in the analyses of these reactions.

Copyright © 2004 By Horizon Bioscience

A pv ---§I

I

18

I

1C

10

I~I

• ~

§l

I

18

~

2C

•

Pi

...

I

1C

1D

3A3B

P2

•

I~

2C

•

P3

~

I

~

3D

[JI~E]

..

I

3D

3A38

FMDV

2C •

• ~

~ 3B l _3

2A.

•

[Pi -2A] - - - . ....- 28C ---. •

~

~

2C

~ 2A

/

2C

~

~12C

~~

2A

~~

P3

..

I~

I

3D

38 1_3

2A

B

\u

3D

I

~ 38 _ 13

.,

I

3D

~

I 3A [[13c II P

ro

3D

38 _ 13

3D

~~

~~ ~~

3D

~~~

~~

3D

~~

3D

2A

2A

2A

~~

2A

Copyright © 2004 By Horizon Bioscience

Qill[] 13C

p

ro

II

3D

I

Figure 1. Primary and secondary polyprotein cleavages. The genome structures of poliovirus (PV) and FMDV are shown with the single, long, open reading frames depicted as shaded boxed areas. The sites of the primary cleavages and the virus proteins responsible are indicated by the curved arrows. In the case of entero- and rhinoviruses, primary cleavages occur at the ID/2A junction mediated by 2Apro, and at the 2C/3A junction mediated by 3CPro. In contrast to PV, the FMDV polyprotein undergoes primary cleavages at the LilA junction and at the C-terminus of protein 2A (panel A). Secondary processing of the primary cleavage products gives rise to a series of alternative products (panel B).

Ryan et al.

551

During the 1980s the molecular cloning and determination of picornavirus genome sequences steadily increased. Sequence alignment algorithms applied to this expanded data set revealed that for any given proteinase type, only a small number of residues were completely conserved - indicating a role in the catalytic mechanism. Furthermore, similarities between the sequences of picornavirus and cellular proteinases indicated structural relationships. Predictions based upon these bioinformatic analyses stimulated (and directed) much of the subsequent experimentation into catalytic mechanisms of picornavirus proteinases. The role of different residues in catalysis or substrate binding were confirmed by sitedirected mutagenetic analyses. More recently, the early bioinformatic predictions of secondary structural similarities with cellular proteinases were vindicated by the determination of the structures of picornavirus proteinases to atomic resolution by X-ray crystallography. 2. FMDV Polyprotein Processing Genera within the picornaviridae show differences in their genome organizations, particularly in relation to the method of biogenesis of the capsid proteins precursor and the replicative protein precursors. The single, long, open reading frame (ORF) of FMDV encodes a polyprotein of some 2,330 aa. The full-length translation product is observed neither within infected cells nor translation reactions in vitro due to 'primary' proteolytic processing (Figure 1, panel A). Such processing is a feature of all picornavirus polyproteins, although FMDV shows a unique feature in this regard. In the entero- and rhinoviruses two primary cleavages are observed. The first cleavage serves to separate the encapsidation functions (the capsid proteins precursor PI) from the replicative protein domains (P2 and P3). This cleavage is mediated by the 2A proteinase (2APro) cleaving to create it's own N-terminus. The second primary cleavage is mediated by the 3C proteinase (3c pro) at a distal site, between protein 2C and 3A, serving to separate the P2 and P3 replication protein precursors (Figure 1, panel A). The aphthovirus polyprotein contains a proteolytically active protein at its Nterminus - the L proteinase (Lpro). In FMDV three primary polyprotein cleavages are observed. The first is mediated by Lpro cleaving to create it's own C-terminus. The second primary 'cleavage' occurs at the C-terminus of the very short 2A region (18 aa) of the FMDV polyprotein. It has been proposed that this 'cleavage' is, in fact, the result of a translational, rather than proteolytic, mechanism (see below). The third primary cleavage occurs between 2C and 3A, mediated by 3cpro (Figure 1, panel A) - as is the case for other picornaviruses. The four products of FMDV primary polyprotein processing therefore comprise; Lpro, the capsid protein precursor [PI-2A] and the replicative protein precursors [2BC] and P3. These primary processing reactions occur in an intramolecular manner (in cis). They are characteristically very rapid (co-translational) and, since the cleavage site and the proteinase are parts of the same molecule, the reaction is insensitive to dilution. Precursor forms spanning these primary cleavage sites are not observed. The [Pl-2A], 2BC and P3 precursor forms subsequently undergo 'secondary' processing mediated by 3cpro (Figure 1, panel B). Secondary processing reactions occur in an intermolecular manner (in trans) and are characteristically slower. In contrast to primary cleavages, secondary polyprotein processing reactions are Copyright © 2004 By Horizon Bioscience

156

FMDV Proteinases

sensitive to dilution, are more sensitive to inhibitors and generally show a greater sensitivity to sequence variations flanking the scissile amino acid pair. In addition to cleaving the virus polyprotein, picornavirus proteinases have also been shown to degrade specific cellular protein targets (discussed below). The following sections are organised by the historical order in which their mechanisms of action became understood.

3. The 3C Proteinase Virus-specific proteolytic activities were first described for the cardiovirus encephalomyocarditis virus (EMC; Lawrence and Thatch, 1975) and subsequently mapped to protein 3C (Gorbalenya et al., 1979; Palmenberg et al., 1979). Picornavirus 3C proteinases mediate a primary cleavage between 2C and 3A, although one report indicated an alternative primary cleavage between proteins 2A and 2B of poliovirus (Lawson and Semler, 1992). We have observed a similar alternative primary cleavage between FMDV proteins 2B and 2C (Flint, 1995). In comparison with other picornavirus 3C proteinases, FMDV 3cpro cleaves a wide range of amino acid pairs (Figure 2). The primary [2BC]/P3 cleavage occurs at a conserved Q/I pair, whilst cleavages within P3 all occur at E-G pairs. Similarly, the cleavage site between 2B/2C is completely conserved, although in this case at a QIL pair. Cleavages with the capsid proteins precursor [PI-2A] show a wider range of pairs; E/G, Q/G, E/T, Q/T, Q/L and Q/M. Not all such pairs present within the polyprotein are processed, however, - their position at the boundaries of polyprotein domains being the major determining factor. FMDV 3CPro is 213 aa long (predicted Mr = 23kDa), whose N- and Ctermini are defined by its own proteolytic activity. The 3Cpro-mediated secondary processing of the [PI-2A], [2BC] and P3 precursors is shown in Figure 1, panel B. Whilst the [P1-2A] precursor is processed to the end products [1 AB], 1C and 1D (the [lAB] 'maturation' cleavage occurs concomitantly with vRNA encapsidation), a multiplicity of different products are generated during processing of P3, some of which are stable products. For example, protein 3CD and those comprising the 3BCD complex (3CD with 1, 2 or 3 3B proteins still attached) are all stable products (Ryan et al., 1989). In the case of poliovirus the processing of the PI capsid protein precursor is mediated not by 3c pro, but by 3CDpro (Jore et ai., 1988; Ypma-Wong et ai., 1988). The 3cpro from other genera are, however, able to process capsid protein precursors (Vakharia et al., 1987; Clarke and Sangar, 1988; Parks et al., 1989; Jia et al., 1991; Harmon et al., 1992) and although the FMDV 3C pro efficiently cleaves all ten processing sites within the FMDV polyprotein (Bablanian and Grubman, 1993), processing of the capsid proteins precursor [PI-2A] was somewhat more efficient with 3CDPro (Ryan et al., 1989). The EMC 3C proteinase also retains proteolytic activity in 3Cpro-containing precursor forms (Jackson 1986; Parks et al., 1989), as is the case for FMDV (Bablanian and Grubman, 1993; Flint, 1995). 3.1. The Catalytic Mechanism of 3cpro

Early studies on 3cpro using proteinase inhibitors showed rather confusing inhibitor profiles. Inhibition of proteolytic activity was observed with both serine and thiol proteinase inhibitors (Summers et al., 1972; Korant, 1972; 1973; Pelham, 1978; Gorbalenya and Svitkin, 1983; Korant et al., 1985; Baum et al., 1991). A Copyright © 2004 By Horizon Bioscience

Ryan et a/. L A12 A24 AlO A22 Crp99 Crp146 Cl C3 Ol/Campos Ol/K O/China O/SKR O/PanAsia Ol/Geshur O/Akesu O/Taiwan O/Tau SAT2 SAT3

-RKLK -QKLK -KKLK -KRLK -RKLK -RKLK -RKLK -RKLK -RKLK -RKLK -KRLR -RKLR -KRLR -KRLK -KKLK -KRLR -KRLR -AYVK ---IR K R

A12 AlO A22 A24 Crp99 Crp146 Cl C3 Ol/Campos Ol/K O-China O/SKR O/PanAsia Ol/Geshur O-Akesu/58 O/Taiwan O/Tau SAT2 SAT3

-ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG -ESNPG G

2A

V

2B

T

3132 A12 AlO A22 Crp99 Crp146 Cl C3 Ol/Campos Ol/K O/China O/SKR O/PanAsia Ol/Geshur 0/Akesu/58 O/Taiwan O/Tau SAT2

1A

GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGOS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA GAGQS-/-GALLA A G

-LPQQE -LPQQE -LPQQE -LPRQE -LPRQE -LPRQE -LPQQE -LPQQE -LPQQE -LPQQE -LPQQE -LPQQE -LPQQE -LPQQE -LPQQE -LPQHE -LPQQE E

!

~

PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFFPFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFSF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/-RAEKQ PFFF-/Q P

JJ

3B3 GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-SVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE GPYAG-/-PVVKE G E

lB

lC

~

DKKTE-/-LPSKE DKKTE-/-LPSKE DKKTE-/-LPSKE DKKTE-/-LPSKE DKKTE-/-LPSKE DKKTE-/-LPSKE DKKTE-/-LPSKE OKKTE-/-LPSKE DKKTE-/-FPSKE OKKTE-/-FPSKE DKKTE-/-FPSKE DKKTE-/-FPSKE DKKTE-/-FPSKE OKKTE-/-FPSKE OKKTE-/-FPSKE DKKTE-/-LPSKE DKKTE-/-LPSKE DKKTE-/-LPGKQ DKKTE-/-LPSKE 0 E Q

2C

]A

U

lD

~

GIFPV-/-DPRSQ GIFPV-/-DPRQQ GIFPV-/-DPRTQ GIVPV-/-DPRSQ GIFPV-/-DAROQ GIFPV-/-DARQQ GIFPV-/-DARQQ GIFPV-/-DARPQ GIFPV-/-DARAE GIFPV-/-DARAE GIFPV-/-DARTQ GIFPV-/-DARTQ GIFPV-/-DARTQ GIFPV-/-OTRTE GVFPV-/-DARTQ GIFPV-/-DARTQ GIFPV-/-DARTQ GIVPV-/-DPVRQ GIFPV-/-NPATQ Q G E

~

TTATG-/-APGKQ TTTTG-/-APAKQ TTTTG-/-APAKQ TTSTG-/-APAKQ TTTTG-/-APAKQ TTTTG-/-APAKQ TTTTG-/-APAKQ TTATG-/-APAKQ TTSAG-/-APVKQ TTSAG-/-APVKQ TTSTG-/-APVKQ TTSTG-/-APVKQ TTSTG-/-APVKQ TTSAG-/-APVKQ TTSPG-/-APAKQ TTSAG-/-APAKQ TTSAG-/-APAKQ TTSAG-/-GVAKQ TTSAG-/-KPDKQ T Q

3B1

JJ

571 2A LLNFDLLNFDLLNFDLLNFDLLNFOLLNFOLLNFOLLNFDTLNFDTLNFOLLNFDLLSFDLLSFDLLNFDLLNFDLLNFDLLNFDLLNFDMCNFDL T M

3B2

JJ

LKARD-/-PIFKQ LKARD-/-PIFKQ LKARD-/-QIFKQ

ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE

GPYTG-/-LPQQE GPYSG-/-LPQQE GPYTG-/-LPQQE

GPYAGGPYAGGPYAG-

LKARD-/-PIFKQ LKARD-/-PIFKQ LKARD-/-PIFKQ LKARD-/-LIFKQ LKARD-/-PIFKQ LKARD-/-PIFKQ LKARD-/-PIFKQ LKARD-/-PIFKQ LKARD-/-PIFKQ LKARO-/-PIFKL LKARD-/-PIFKQ LKARD-/-LIFKQ LKARD-/-PIFKQ LKARD-/-PIFKQ

ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-OPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPQAE ISIPS-/-QPRAE ISIPS-/-QPRAE ISLPS-/-KPQAE

GPYAG-/-LPRQE GPYAG-/-LPRQE GPYAG-/-LPRQE GPYAG-/-LPQQE GPYAG-/-LPQQE GPYAG-/-LPQQE GPYTG-/-LPQQE GPYTG-/-LPQQE GPYTG-/-LPQQE GPYAG-/-LPQQE GPYAG-/-LPQQE GPYAG-/-LPQQE GPYAG-/-LPQHE GPYAG-/-LPQQE

GPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAGGPYAG-

L

3

Q

3C GPYEG-/-EPQHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE GPYEG-/-EPHHE G E

E

JJ

G

E

G

3D GLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDGLIVDG

Figure 2. FMDV polyprotein cleavage sites. Cleavage at the LilA site occurs at (KJR) I G pairs (V). The capsid protein IA/IB cleavage (~) occurs (by an unknown mechanism) concomitantly with vRNA encapsidation. The primary 2A/2B cleavage (~) occurs at a conserved G/P pair. 3Cpro-mediated processing (~) : the primary 2C/3A cleavage occurs at a conserved Q/I pair whilst secondary processing of P3 occurs only at E/G pairs and at the 2B/2C site at a conserved Q/L pair. A range of pairs is cleaved during processing of the capsid proteins precursor [P1-2A]. Copyright © 2004 By Horizon Bioscience

\58

FMDV Proteinases

breakthrough came when similarities between the large sub-class (trypsin-like) cellular serine proteinases and 3C proteinases were detected by sequence alignments (Gorbalenya et a!., 1986; 1989; Bazan and Fletterick, 1988; 1990). These analyses predicted both a chymotrypsin-like (serine proteinase) fold and the identity of the residues which would form a catalytic triad analogous to that of serine proteinases (Figure 3). In the case ofFMDV 3CPro the putative catalytic triad would be composed of His46 , Asp84, and, perhaps most interestingly, the active site nucleophile being Cys 163 (rather than serine, as in chymotrypsin). Subsequent site-directed mutagenesis experiments confirmed the roles of these residues in rhinovirus (Cheah et a!., 1990), polio (Hamme~le et al., 1991; Kean et a!., 1991; Lawson and Semler, 1991), hepatitis A (Jia et a!., 1991) and FMDV 3cpro catalysis (Flint, 1995; Grubman et al., 1995). 3.2. The Structure of 3C pro

The resolution of the atomic structures of the 3C proteinases from human rhinovirus (HRV) 14 (Matthews et al., 1994), hepatitis A (Allaire et a!., 1994; Bergmann et al., 1997), polio (Mosimann et a!., 1997) and HRV 2 (Mathews et al., 1999) confirmed the predicted chymotrypsin-like fold of the enzyme and catalytic mechanism. Detailed discussion of the structural and mechanistic relationships between these enzymes and cellular proteinases can be found in Dougherty and Semler (1993), Malcolm (1995), Babe and Craik (1997); Ryan and Flint (1997), Seipelt et ale (1999) and Skern et ale (2002). Alignment of all available picornavirus 3C sequences shows that the catalytic histidine and cysteine residues are both completely conserved and aligned. The alignment of the acidic residue of the triad shows that whilst this residue is a glutamate in the entero-, rhino- and kobuviruses, it is an aspartate in all other genera. The alignment shown in Figure 3 is derived from an alignment of all picornavirus 3C sequences, but only showing those sequences for which atomic structures are available together with FMDV (strain 01K). Although picornaviruses represent a significant group of human and animal pathogens, surprisingly few vaccines are available - notably polio, hepatitis A and FMDV. Sequence alignments of non-structural proteins (involved in the replication of the virus RNA) showed that for certain proteins, the high conservation made the development of anti-viral drugs feasible. The attraction with this strategy of disease control is that a drug could be effective against viruses from different serotypes, and even different genera. In this regard the virus-encoded proteinase and polymerase (3D) were obvious candidate a~ti-viral drug targets (assay systems were available for both enzymes). Resolution! of the atomic structure of the 3C proteinase has, indeed, led to structure-based drug design (Mathews et a!., 1999; Patick et ai., 1999; Wang, 1999; Lall et al., 2002; Ramtohul et al., 2001; 2002a; 2002b). 3.3. RNA Binding Properties of 3cpro

An additional and quite unexpected property of this proteinase was discovered when mutations suppressing the effect of a four base insertion within the 5' non-coding region of the RNA genome were mapped within 3cpro (Andino et al., 1990a). Subsequently it was demonstrated these mutations were affecting the binding of '3CDPro, rather than 3cpro, to +ve strand RNA (Andino et al., 1990b, 1993). This

Copyright © 2004 By Horizon Bioscience

aN ----+

4---

A 1 -----.

E-D 1-+

--+

4-C 1

4-

..---

E

B

1 ----+

I ----+

4-C 1

..-

F

1

-

V-IPTBA-------QPGDDVLVNGQKIRVKDKYKLVDPE--NINLELTVL V-VPTBA-------DPGKEIQVDGITTKVIDSYDLYNKN--GIKLEITVL I-LPTBA-------SPGESIVIDGKEVEILDAKALEDQA--GTNLEITII YLVPRBLFAEKYDKIMVDGR-AMTDSDYRVFEFEIKVKGQDML-SDAALM LLVPSBAYKFEKDYEMMEFYFNRGGTYYSI SAGNVVIQSLDVGFQDVVLM ----+ * 4I-D 1-+ 41--. * 4--+ E

HRV14 HRV2 PV FMDV OIK HAV

1 ------+

-----GPNTEFALS-LLRKNIMTITTSKGE---------FTGLGIHDRVC -----GPEEEFGMS-LIKHNSCVITTENGK---------FTGLGVYDRFV -----GPGFDYAVA-MAKRNIVTATTSKGE---------FTMLGVHDNVA SGAPPTDLQKMVMGN-TKPVEL--ILDGK------TVAICCATGVFGTAST---LEIAGLVRKNL---VQFGV--GEKNGCVRWVMN---ALGVKDDW4---

HRV14 HRV2 PV FMDV OIK HAV

591

B

+--- aN ----+ 4-A 1-+

HRV14 HRV2 PV FMDV OIK HAV

Rvan et al.

Is

-+ 4- A 2 B2TLDRNEKFRDIRGFISEDLE----G-VDATLVVHSNNFTNTILEVGPVTM KLDRNEKFRDIRRY IPNNED----DYPNCNLALLANQPEPTI INVGDVVS TLKRNEKFRDIRPHI PTQIT----ETNDGVLIVNTSKYPNMYVPVGAVTE VLHRGNRVRDITKHFRDTARM--KKGTPVVGVINNADVGRLIFSGEALTY KVPTIPKFRDITEHFIKKGDV-PRALNRLATLVTTVNGTPMLISEGPLKM

B

4-A 2 -+

2 -

D2 ----.

HRV14 HRV2 PV FMDV OIK HAV

4--+

AGLINLSSTPTN---------RMIRYDYATKTGQCGGVLCATG-----KI YGNILLSGNQTA---------RMLKYSYPTKSGYCGGVLYKIG-----QV QGYLNLGGRQTA---------RTLMYNFPTRAGQCGGVITCTG-----KV KDIVVCMD--------GDTMPGLFAYRAATKAGYCGGAVLAKDGA-DTFI EEKATYVHKKNDGTTVDLTVDQAWRGKGEGLPGMCGGALVSSNQSIQNAI -+

E

HRV14 HRV2 PV FMDV OIK HAV

C 2

41--+

2 -+

4--+

+-C 2 -+

*

+--+

+-

4-F 2-+

FGIHVGGNGRQGFSAQLKKQYFVEK----------Q LGIHVGGNGRDGFSAMLLRSYFTDV----------Q IGMHVGGNGSHGFAAALKRSYFTQS----------Q VGTHSAGGNGVGYCSCVSRSMLLKMKAHIDPEPHHE LGIHVAGGNSILVAKLVTQEMFQNID---KK-IESQ - E 2 -+ 4--+ 4--+ F2

ae

Figure 3. Sequence alignment and secondary structure of 3C proteinases. The sequences of 3C proteinases for which atomic structures are available (human rhinoviruses 2 and 14, hepatitis A and polio) is shown aligned with that of FMDV serotype 01 K. The secondary structural features of HRV14 and hepatitis A 3C proteinases are given above and below the alignment, respectively. The active site residues are indicated (bold type face, *) along with the RNA binding motif (underlined, grey shaded text).

Copyright © 2004 By Horizon Bioscience

\60

FMDV Proteinases

vRNA-binding property of 3CDpro was demonstrated for rhinoviruses and hepatitis A. Although mapping to multiple sites, when the data were combined the mutations clustered to define an RNA binding site (Andino et aI., 1990a; 1990b; 1993; Leong et al., 1993; Walker et aI., 1995; Blair et aI., 1996; 1998; Kusov and Gauss-Muller, 1997). Mutations shown to affect RNA binding were located within the atomic structure on the opposite side to the catalytic site. The majority of these mutations were observed in a domain interconnecting the two lobes of the proteinase - located between the two (anti-parallel) helices of the N- and C-termini - and also clustered on the (exterior) loop connecting ~D2 and ~E2 and centered about the conserved KFRDIR sequence of the domain linker (Figure 3). Interestingly, this interconnecting region lies in close proximity to the N- and C-termini of the proteinase. 4. The L Proteinase Early studies on FMDV showed that polyprotein processing was mediated by more than one type of proteinase (Burroughs et al., 1984). The aphthoviruses, together \vith erbovirus, possess a proteinase (LPro) at the N-terminus of the polyprotein not found in other genera. The L protein was known to exist in (at least) two forms; Labpro and Lb pro derived from initiation of translation at either of two in-frame AUG codons located 84 nucleotides apart (Clarke et aI., 1985; Sangar et al., 1987). The Lb pro form (and quite possibly the Lab pro form) undergoes a post-translational modification by a carboxypeptidase B-like activity producing LB' (and possibly Lab'; Sangar et al., 1988). Lpro cleaves co-translationally thereby generating its own C-terminus (Strebel and Beck, 1986). This cis cleavage occurs between Lys/Arg 201 and Gly202 (Figure 2). Both the Lab pro and Lb pro forms were shown to cleave at the L/P1 junction either in cis or in trans (Medina et al., 1993; Cao et al., 1995). Lpro has another function besides cleaving itself from the nascent polyprotein and this trans activity is the same as one identified previously for the 2A proteinase of the entero- and rhinoviruses - cleavage of the host-cell translation factor eIF4G. The cleavage site within eIF4G differs between the two types of proteinase - Lpro cleaving between residues Gly479 and Arg480 , whilst polio 2Apro cleaves between Arg486 and Gly487 (Kirchweger et al., 1994). The closely separated Lpro and 2Apro cleavage sites are both thought to lie in a hinge region separating the two major domains of eIF4G. The N-terminal domain comprises the binding site for eIF4E (binds the 5'-terminal m7G host-cell mRNA cap structure) and poly(A) binding protein (PABP). The C-terminal domain comprises the sites which bind eIF4A and eIF3, which act to unwind secondary RNA structure in the mRNA and recruit the 43S pre-initiation complex (reviewed by Hentze, 1997; Jackson and Wickens, 1997; Morely et al., 1997; Sachs et al., 1997). Cleavage of eIF4G by Lpro therefore uncouples these two major activities and leads to the 'shut-off' of capped mRNAhost-cell- translation. Although eIF4G cleavage activity is conserved between the aphtho- and entero-/rhinoviruses, the proteinases responsible are quite different: not only with respect to their positions in the polyprotein, but in that LPro is a thioltype proteinase whilst 2Apro has structural similarity with the sub-class of small serine proteinases (Petersen et al., 1999).

Copyright © 2004 By Horizon Bioscience

Rvan et al.

611

4.1. Lpro Catalytic Mechanism and Atomic Structure

Based upon sequence alignments Gorbalenya and co-workers (1991) suggested a structural and mechanistic relationship between FMDV Lpro and cellular thiol-type proteinases. This hypothesis was supported by site-directed mutagenesis showing that Cys 5! and His 148 were the active site residues (Piccone et ai., 1995a; Roberts and Belsham, 1995). In the former publication Lb pro was modeled onto the papain 3D structure and residues Glu 76 and ASp 164 were suggested to be involved in proteolysis by playing a role in substrate binding. Interestingly, mutation of the active site nucleophile cysteine to serine did not affect the cleavage activity in cis, but did abolish the ability of the proteinase to cleave substrates in trans (Piccone et ai., 1995a). Resolution of the atomic structure of Lb pro showed a compact globular form with a flexible C-terminal extension from D 184 to K210 (Guarne et ai., 1998,2000). The active site is located between an a-helical domain and a ~-sheet domain. The orientation of the histidine component of the catalytic dyad is optimised by its interaction with Asp 163, rather than an asparagine residue - which performs this role in thiol proteinases such as papain. Lpro cleaves at its own C-terminus: the substrate binding site must, at some stage, therefore, accommodate its own C-terminus (the 'P' side residues of the scissile bond). In the crystal structure of LbPro , the C-terminus of one molecule is bound into the substrate binding site of the adjacent molecule. This gave an additional insight into the function of the enzyme since half of the substrate was bound into this site! The branched hydrophobic P2 residue (L/V; Figure 2) forms an important interaction with a deep hydrophobic cavity, although (unlike papain) Lb pro has appreciable binding pockets for the P1 and P l' residues, which make a substantial contribution to its substrate binding specificity. 4.2. 'Leaderless' FMDV

Due to (i) its position at the N-terminus, (ii) the single activity in FMDV polyprotein processing (self-cleavage) and (iii) its role in shut-off of host-cell capped mRNA translation, it seemed both feasible and important to characterise the properties of a virus with this proteinase deleted. The Plum Island group deleted L from the infectious copy of FMDV serotype A12 and were able to rescue virus (Piccone et ai., 1995b). Initially this genetic manipulation did not, however, appear to greatly affect virus replication nor the yield of virus particles: plaques on BHK cells that were slightly smaller than those produced by wild-type (w.t.) virus, grew to slightly lower titers than w.t. virus in BHK cells, shut off host protein synthesis more slowly than w.t. virus, and were slightly attenuated in mice. The L proteinase is not, therefore, essential for FMDV replication in these commonly used tissuecultured cells. Infection of cattle \vith this leaderless virus showed, however, much more dramatic effects: the virus was less widely disseminated in the lung at 24hr postinfection (pj.), with no lesions or virus detectable in secondary sites at 72hr pj.. In a later study 2 of 3 animals inoculated with the leaderless virus did not develop lesions when challenged with W.t. virus, but showed mild signs of infection. The third inoculated animal developed some lesions, but these were less severe than in the un-inoculated control animal, which showed classical FMD. Inoculating swine

Copyright © 2004 By Horizon Bioscience

162

FMDV Proteinases

with either live or chemically inactivated leaderless virus showed an improved immune response in animals inoculated with live virus. Animals vaccinated with the live virus did not exhibit signs of FMD, did not spread virus to other animals, developed a neutralising antibody response and antibodies to the non-structural protein 3D, and were partially protected from FMD. The authors concluded that this strategy provided a good candidate live-attenuated vaccine as well as a safe source of antigen for chemically inactivated vaccines (reviewed in Mason et aI., 2003). Why did the leaderless virus replicate so well in BHK cells but show reduced virulence in animals? The answer appears to lie in the type I interferon (IFN) response. The inability of FMDV-infected cells to translate type I IFN mRNA, due to LPro-mediated shut-off, makes a substantial contribution to the virulence of the virus. The BHK cells commonly used to propagate FMDV are, however, defective in their response to IFN. 5. The 2A Oligopeptide The second primary cleavage of the FMDV polyprotein occurs between the 2A and 2B proteins - markedly different from that observed in the entero- and rhinoviruses (Figure 1). Precursor forms spanning the 2A / 2B cleavage site are not detected during native polyprotein processing. The 2A region of the FMDV polyprotein is very short (18 aa), and shares a conserved C-terminal sequence motif (-DxExNPG-) with that of the longer cardiovirus 2A protein. The N-terminal residue of the 2B proteins of both groups is also completely conserved - proline. Interestingly, in cardiovirus polyprotein processing the similar primary cleavage also occurs at the C-terminus of 2A. Cleavage at the 2A/2B site of FMDV or cardiovirus polyproteins was shown to require neither the L nor 3C proteinases (Clarke and Sangar, 1988; Roos et aI., 1989; Ryan et al., 1989). The 2A region was not, therefore, simply a substrate for a virus proteinase, and the very rapid, complete, cleavage suggested it did not function as a substrate for a host-cell proteinase. Could 2A play a role in a larger form - [1 D2A] or [2AB], for example? To answer this question, a series of in-frame deletions were made such that the upstream or downstream context of 2A was altered (Ryan et a!., 1991). These analyses suggested that this was not the case and that the cleavage was, indeed, associated with this region alone (plus the N-terminal proline of protein 2B). To test this hypothesis the 2A sequence was used in the creation of an 'artificial' reporter gene polyprotein in which sequences encoding chloramphenicol acetyltransferase (CAT) were linked via FMDV 2A (plus the N-terminal proline of 2B) to ~-glucuronidase (GUS) in a single ORF ([CAT-2A-GUS]). Analysis of this polyprotein system using in vitro translation systems showed, indeed, that the 20 amino acid FMDV sequence was able to mediate a co-translational cleavage producing the cleavage products [CAT-2A] and GUS (Ryan and Drew, 1994). In this heterologous context, therefore, the FMDV 2A sequence mediated a highly efficient cleavage at its own C-terminus - just as in FMDV polyprotein processing (-DxExNPG ~ P-). Other studies showed that deletion of the N-terminal 66% of the cardiovirus 2A did not abrogate cleavage at the 2A/2B site, and that mutations within the conserved -NPGP- sequence at the extreme C-terminus of EMC 2A abolished cleavage activity (Palmenberg et a!., 1992). Cleavage at the 2A/2B site of TME virus was highly efficient when only 2A and 2B sequences were present (Batson Copyright © 2004 By Horizon Bioscience

Ryan et al.

631

and Rundell, 1991). Later it was shown that the C-terminal 19 amino acids together with the N-terminal proline of 2B from either FMDV, EMC or TME when inserted into an artificial polyprotein are able to mediate a co-translational cleavage with high efficiency (~95%; Donnelly et al., 1997). Based upon the observed kinetics of the cleavage reaction, inability to inhibit the reaction with proteinase inhibitors plus data from site-directed mutagenesis of the 2A sequence, we have proposed that the 2A/2B cleavage is not a proteolytic cleavage (by either a virus or host-cell proteinase), but the result of a translational effect (Ryan etal., 1999; Donnelly etal., 2001a; Ryan etal., 2002). Briefly, we have proposed that the nascent 2A sequence interacts with the exit pore of the ribosome to bring about a reorientation of the tRNA-peptidyl ester linkage precluding it from nucleophilic attack by the prolyl-tRNA. Hydrolysis of this bond would lead to release of the peptide such that elongation may continue from the prolyl-tRNA, producing two, discrete, 'cleavage' products. The N-terminal proline of 2B is, therefore, a crucial part of the proposed mechanism. Site directed mutagenesis of this residue abrogates cleavage: it is completely conserved in all picornavirus 2B proteins which use this 'cleavage' mechanism and in '2A-like' sequences from other viruses (see below). An interesting test of this hypothesis came from the analysis of an artificial polyprotein system in which the first protein bore a leader sequence. A single open reading frame was constructed encoding three proteins; (i) DNaF - yeast proalpha factor with, instead of its native signal sequence, the signal recognition particle (SRP)-dependent signal sequence of dipeptidyl amino peptidase B (Dap2p) (ii) FMDV 2A and, (iii) green fluorescent protein (GFP). Here, SRP binds to the Dap2p leader sequence, arrests translation, and leads to the nascent protein being translated in a ribosome : protein conducting channel complex (formed in the process of translocating proteins across the membrane into the lumen of the endoplasmic reticulum - ER). In this situation, therefore, the nascent protein would not be accessible to cytosolic protein(ases) and the folding of the nascent protein would not occur until passing out of the protein conducting channel into the lumen of the ER. Analysis of the processing properties of this artificial polyprotein and the sub-cellular localization of the 'cleavage' products showed that; (i) 2Amediated 'cleavage' was unaffected, (ii) the first protein, DNaF, was translocated into the lumen of the ER and (iii) GFP was located in the cytoplasm (de Felipe et al., 2003). These data are consistent with our translational model of 'cleavage'. 5.1. The 'Translational' Model of 2A-Mediated Cleavage: Implications for The Biology of FMDV

Our work on the cleavage of a [GFP-2A-GUS] polyprotein system using in vitro translation systems showed that a molar excess of the [GFP-2A] 'cleavage' product accumulated above the GUS product (Donnelly et al., 2001a). If the gene order was reversed ([GUS-2A-GFP]), more [GUS-2A] accumulated than GFP. We showed this was not due to different rates of protein degradation or nonspecific dissociation of either the T7 RNA polymerase (during transcription of the template), or ribosomes (during translation). The molar excess was due to different levels of synthesis: the two different parts of the single open reading frame were being translated at different levels. This implies that in the FMDV polyprotein,

Copyright © 2004 By Horizon Bioscience

\64

FMDV Proteinases

more capsid proteins could be synthesized than replicative proteins. In a FMDVinfected cell at the latter stages of replication one might expect the depletion of cellular factors to become rate-limiting for the virus. Indeed, a decrease in the rate of ribosome elongation is typical during the course of picornavirus infections (Summers et al., 1967; Hackett et al., 1978; Ramabhadran and Thatch, 1981). It could be that under these conditions, FMDV has a mechanism for 'switching' those remaining cellular resources into the synthesis of capsid proteins - and not replication proteins. If this were true, the kinetics of vRNA encapsidation could be altered which might, in part, account for the rapidity of the growth of this amazing virus. 5.2. '2A-Like' Sequences

At the outset of our work on FMDV polyprotein processing, this type of 2A sequence was known only for FMDV and cardioviruses. We now know that 2Alike sequences are found in a range of other systems (Figure 4). Where they occur in other picornaviruses, 2A sequences are (as in FMDV) found at the junction of the capsid and replicative proteins. The short, oligopeptide type of 2A is found in; equine rhinitis A (ERAV; aphthovirus) and equine rhinitis B viruses (ERBV; erbovirus) (Li et al., 1996; Wutz et al., 1996), teschoviruses (Zell et ai., 2001) and Ljungan viruses (Johansson et ai., 2002). Furthermore, probing the databases with the -DxExNPGP- motif (conserved amongst all of these viruses) revealed the existence of '2A-like' sequences in other mammalian viruses - human (James et ai., 1999), bovine (Jiang et al., 1993) and porcine (Qian et ai., 1991) type C rotaviruses. 2A-like sequences are also found in a wide range of positive stranded RNA insect viruses; Thosea asigna virus (TaV; Pringle et al., 1999), Infectious flacherie virus (IFV; Isawa et ai., 1998), Drosophila C virus (DCV; Johnson and Christian, 1998), Acute bee paralysis virus (ABPV; Govan et ai., 2000), Cricket paralysis virus (CrPV; Wilson et al., 2000), Euprosterna elaeasa virus (EeV; Zeddam et al., 2001 accession no. AF461742 : unpublished), Perina nuda picorna-like virus (PnPV; Wu et al., 2002), Providence virus (PrV; Pringle et al., 2003). Indeed, some insect viruses contain t\VO 2A-like sequences. PnPV contains a 2Alike sequence within the capsid protein region, with the second 2A-like sequence at the junction of the capsid / replicative proteins (Figure 4). The capsid proteins precursor of Providence virus (754 aa) also contains two 2A-like sequences which are closely separated (residues 32-52 and 103-123; Figure 4). In addition we have identified insect cypoviruses encoding 2A-like sequences; Bombyx mori cypovirus 1 (BmCPV-l; Hagiwara et al., 2001) and Lynlantria dispar cypovirus 1 (LdCPV-l; AF389466: unpublished). All of the 2A-like sequences are from viruses, but perhaps the most intriguing instances are within repeated sequences of Trypanosoma spp. In the case of T. brucei, the causative agent of African trypanosomiasis (sleeping sickness), the '2A-like' sequence occurs at the junction of the trypanosome repeated sequence TRS-l (Hasan et al., 1984; Murphy et ai., 1987) and an ORF encoding a reverse transcriptase-like protein gene. In the case of T. cruzi, the causative agent of Chagas disease, the 2A-like sequence occurs, however, in a different protein ORFI of the non-LTR retrotransposon LlTc (Martin et ai., 1995), encoding an AP endonuclease-like sequence (APendo). Copyright © 2004 By Horizon Bioscience

Rvan et a/.

651

Replicative Proteins

Large CP

Small CP

Large CP

Small CP

Large CP

Small CP

TaV

PrV

PnPV

..

- - - - - - - - - - - - ORF ------------t~

o

D

Figure 4. Location of2A-like sequences within insect virus polyproteins. The 2A-like sequences (black rectangles) are found at the junction of the capsid protein (grey boxed areas) and replicative protein (unshaded, boxed, areas) precursors of picomaviruses and the insect virus infectious ftacherie virus (IFV). The locations of 2C/2C-like (circles), 3C/3C-like proteinase (rectangles) and polymerase (diamonds) sequence motifs are shown within the replicative proteins precursor. Insect viruses shown with two, non-overlapping, open reading frames (Drosophila C virus, DCV; cricket paralysis virus, CrPV; acute bee paralysis virus, ABPV) have the 2A-like sequence located near the N-terminus of the replicative proteins polyprotein (ORF 1). Other insect viruses have 2A-like sequences located within the capsid protein precursor (Thosea asigna virus, TaV; Euprosterna elaeasa virus, EeV; Providence virus, PrY). Complete genome sequences are not available for TaV nor PrY - only the capsid protein precursor region. A complete sequence is available for EeV, however, and the genome organisation of this virus is shown. In one case, Perina nuda picoma-like virus (PnPV), 2A-like sequences are present in both types of site.

We have inserted representatives of these 2A-like sequences into our [GFP-

2A-GUS] reporter polyprotein and found that all tested were active in mediating 'cleavage' (Donnelly et al., 2001b, Ryan et al., 2002). It should be noted, however, that not all proteins containing the -DxExNPGP- motif are active: we have tested two 2A-like sequences from cellular genes both of which were inactive. 6. The Cleavage of Host-Cell Proteins Whilst picornavirus proteinases play a central role in the biogenesis of virus proteins from precursor forms, they also may cleave host-cell proteins. Proteins involved in a range of cellular processes - notably gene transcription and Copyright © 2004 By Horizon Bioscience

Table 1. Host-cell proteins degraded by picornavirus proteinases Function Virus Host-ceU Virus Proteinase Protein translation polio eIF4G 2A initiation factor HRV2 2A CBV4 2A FMDV L FMDV 3C eIF4GI (isoform) polio 2A HRV? 2A eIF4GII (isoform) polio 2A HRV? 2A eIF4Al translation FMDV 3C initiation factor H3 histone (FMDV infection) FMDV FMDV 3C FMDV 3C FMDV 3ADC TFIIIC pollII polio 3C (a subunit) transcription factor cyclin B2 control of cell FMDV L (xenopus) cycle cyclin A control of cell L FMDV (human) cycle

Copyright © 2004 By Horizon Bioscience

Reference Etchison et al., 1982 Lamphear et al., 1993 Lamphear et al., 1993 Kirchweger et al., 1994 Belsham et 2000 Gradi et al., 1998 Svitkin et 1999 Gradi et al., 1998 Svitkin et al., 1999 Belsham et al., 2000 Lietal.,2001 Grigera and Tisminetzky 1984 Falk et al., 1990 Tesar and Marquardt 1990 Capozzo et al., 2002 Clark et al., 1991

Ziegler et al., 1995 Ziegler et al., 1995

TBP

TATAbox binding protein

polio

3C 2A

MAP-4

microtubuleassociated protein 4

polio

3C

Clark et al., 1993 Das & Dasgupta 1993 Yalamanchili et 1997b Joachims et al., 1995

Oct-l

transcription factor

polio

3C

Yalamanchili et al., 1997c

CREB

cAMP responsive element binding protein poly(A) binding protein

polio

3C

Yalamanchili et al., 1997a

polio CBV3 CBV4 CBV3

2A&3C 2A 2A 2A

PABP

Joachims et al., 1999 Kerekatte et al., 1999 Kerekatte et al., 1999 Badorf et al., 1999

dystrophin

structure of muscle fibres

cytokeratin 8

intermediate filament protein

HRV2

2A

~-COP

membrane traffic

FMDV

(FMDV infection)

C. Knox, unpublished

membrin

membrane traffic

FMDV

(FMDV infection)

C. Knox, unpublished

Copyright © 2004 By Horizon Bioscience

Seipelt et al., 2000

\68

FMDV Proteinases

translation - are degraded by these virus proteinases (Table 1). Recently we have observed that two proteins involved in membrane traffic (~-COP and membrin) are also degraded within FMDV-infected cells (C. Knox, unpublished). In many cases it is difficult to assess the contribution to replication efficiency or virulence of these different cleavage events. This is of great interest, however, in understanding how persistent or chronic picornavirus infections are established (see Chapter 6 and Colbere-Garapin et al., 2002). The picture that has emerged is of one of co-evolution - genetic changes occurring in both the host-cell and in the virus (de la Torre et al., 1985, 1988, 1989). Furthermore, cell-lines may be generated expressing low levels of 3cpro (Lawson et al., 1989; Martinez-Salas and Domingo, 1995). Whilst defective-interfering (DI) genomes can be established quite readily for enteroviruses by passage in tissue-culture, this is not the case with FMDV. Defective non-interfering genomes are generated during the establishment of a persistent FMDV infection - genomes with deletions located within Lpro (Charpentier et al., 1996). FMDV proteinases not only play a central role in the biogenesis of virus proteins, but are of acute interest in disease control. Firstly, they may well form drug targets - it could be argued that in certain situations (such as contingency ring vaccination) drug administration could protect susceptible animals until a protective immune response is established. Advances in antiviral development for the human rhinovirus 3C proteinase could be of particular interest here. Secondly, it is apparent that establishment of persistent infections involves changes in the activity of the virus proteinases. Thirdly, the leaderless FMDV provides an excellent basis for the development of attenuated viruses. References Allaire, M., Cheraia, M.M., Malcolm, B.A. and James, M.N.G. 1994. Picomaviral 3C cysteine proteinases have a fold similar to chymotrypsin-like serine proteinases. Nature. 369: 72-76. Andino, R., Rieckhof, G.E., Trono, D. and Baltimore, D. 1990a. Substitutions in the protease (3CPro) gene of poliovirus can suppress a mutation in the 5' noncoding region. J. Virol. 64: 607-612. Andino, R., Rieckhof, G.E. and Baltimore, D. 1990b. A functional ribonucleoprotein complex forms around the 5' end of poliovirus RNA. Cell. 63: 369-380. Andino, R., Rieckhof, G.E., Achacoso, P.L. and Baltimore, D. 1993. Poliovirus RNA synthesis utilizes an RNP complex formed around the 5'-end of viral RNA. EMBO 1. 12: 3587-3598. Babe, L.M. and Craik, C.S. 1997. Viral proteases: evolution of diverse structural motifs to optimize function. Cell. 91: 427-430. Bablanian, G.M. and Grubman, M.J. 1993. Characterization of the foot-and-mouth disease virus 3C protease expressed in Escherichia coli. Virology. 197: 320-327. Badorff, C., Lee, G.H., Lamphear, BJ., Martone, M.E., Campbell, K.P., Rhoads, R.E. and Knowlton, K.D. 1999. Enteroviral protease 2A cleaves dystrophin: evidence of cytoskeletal disruption in an acquired cardiomyopathy. Nat. Med. 5: 320-326. Batson, S. and Rundell, K. 1991. Proteolysis at the 2N2B junction in Theiler's murine encephalomyelitis virus. Virology. 181: 764-767. Baum, E.Z., Bebemitz, G.A., Palant, 0., Mueller, T. and Plotch, S.J. 1991. Purification, properties and mutagenesis of poliovirus 3C protease. Virology. 185: 140-150.

Copyright © 2004 By Horizon Bioscience

Rvan et al.

691

Bazan, J.F. and Fletterick, R.J. 1988. Viral cysteine proteases are homologous to the trypsinlike serine proteases: structural and functional implications. Proc. Natl. Acad. Sci. U.S.A. 85: 7872-7876. Bazan, J.F. and Fletterick, RJ. 1990. Structural and catalytic models of trypsin-like viral proteases. Sem. Virol. 1: 311-322. Belsham, GJ., McInerney, G.M. and Ross-Smith, N. 2000. Foot-and-mouth disease virus 3C protease induces cleavage of translation initiation factors eIF4A and eIF4G within infected cells. J. Virol. 74: 272-280. Bergmann, E.M., Mosimann, S.C., Chernaia, M.M., Malcolm, B.A. and James, M.N. 1997. The refined crystal structure of the 3C gene product from hepatitis A virus: specific proteinase activity and RNA recognition. J. Virol. 71: 2436-2448. Blair, W.S., Nguyen, J.H., Parsley, T.B. and Semler, B.L. 1996. Mutations in the poliovirus 3CD proteinase S I-specificity pocket affect substrate recognition and RNA binding. Virology. 218: 1-13. Blair, W.S., Parsley, T.B., Bogerd, H.P., Towner, J.S., Semler, B.L. and Cullen, B.R. 1998. Utilization of a mammalian cell-based RNA binding assay to characterize the RNA binding properties of picornavirus 3C proteinases. RNA. 4: 215-225. Burroughs, J.N., Sangar, D.V., Clarke, B.E., Rowlands, D.J., Billiau, A. and Collen, D. 1984. Multiple proteases in foot-and-mouth disease virus replication. 1. Virol. 50: 878-883. Cao, X., Bergman, I.E., Fullkrieg, R. and Beck, E. 1995. Functional analysis of the two alternative initiation sites of foot-and-mouth disease virus. J. Virol. 69: 560-563. Capozzo, A.V., Burke, DJ., Fox, J.W., Bergmann, I.E., La Torre, J.L. and Grigera, P.R. 2002. Expression of foot and mouth disease virus non-structural polypeptide 3ABC induces histone H3 cleavage in BHK21 cells. Virus Res. 09: 91-99. Charpentier, N., Davila, M., Domingo, E. and Escarmis, C. 1996. Long-term, largepopulation passage of aphthovirus can generate and amplify defective noninterfering particles deleted in the leader protease gene. Virology. 223: 10-18. Cheah, K-C., Leong, L.E-C. and Porter, A.G. 1990. Site-directed mutagenesis suggests close functional relationship between human rhinovirus 3C cysteine protease and cellular trypsin-like serine proteases. J. BioI. Chern. 265: 7180-7187. Clark, M.E., Hammerle, T., Wimmer, E. and Dasgupta, A. 1991. Poliovirus proteinase 3C converts an active form of transcription factor IIIC to an inactive form: a mechanism for inhibition of host cell polymerase III transcription by poliovirus. EMBO J. 10: 2941-2947. Clark, M.E., Lieberman, P.M., Berk, AJ. and Dasgupta, A. 1993. Direct cleavage of human TATA-binding protein by poliovirus 3C in vivo and in vitro. Mol. Cell. BioI. 13: 12321237. Clarke, B.E., Sangar, D.V., Burroughs, J.N., Newton, S.E., Rowlands, D.J. and Carroll, A.R. 1985. Two initiation sites for foot-and-mouth disease virus polyprotein in vivo. 1. Gen. Virol. 66: 2615-2626. Clarke, B.E. and Sangar, D. V. 1988. Processing and assembly of foot-and-mouth disease virus proteins using subgenomic RNA. J. Gen. Virol. 69: 2313-2325. Colbere-Garapin, F., Pelletier, I. and Ouzilou, L. 2002. Persistent infections by picornaviruses. In: Molecular Biology of Picornaviruses. B.L. Semler and E. Wimmer, eds. ASM Press, Washington, USA. 437-448. Das, S. and Dasgupta, A. 1993. Identification of the cleavage site and determinants required for poliovirus 3C pro- catalysed cleavage of human TATA-binding protein transcription factor TBP. J. Virol. 67: 3326-3331. de Felipe, P., Hughes, L.E., Ryan, M.D. and Brown, J.D. 2003. Co-translational, intraribosomal cleavage of polypeptides by the foot-and-mouth disease virus 2A peptide. J. BioI. Chern. 278: 11441-11448.

Copyright © 2004 By Horizon Bioscience

\70

FMDV Proteinases

de la Torre, J.C., Davila, M., Sobrino, F., Ortin, J. and Domingo, E. 1985. Establishment of cell lines persistently infected with foot-and-mouth disease virus. Virology. 145: 24-35. de la Torre, lC., Martinez-Salas, E., Diez, J., Villaverde, A., Gebauer, F., Rocha, E., Davila, M. and Domingo, E. 1988. Coevolution of cells and viruses in a persistent infection of foot-and- mouth disease virus in cell culture. J. Virol. 62: 2050-2058. de la Torre, J.C., Martinez-Salas, E., Diez, J. and Domingo, E. 1989. Extensive cell heterogeneity during persistent infection with foot-and- mouth disease virus. J. Virol. 63: 59-63. Donnelly, M., Gani, D., Flint, M., Monoghan, S. and Ryan, M.D. 1997. The cleavage activity of aphtho- and cardiovirus 2A proteins. J. Gen. Virol. 78: 13-21. Donnelly, M.L.L., Luke, G., Mehrotra, A., Li, X., Hughes, L.E., Gani, D. and Ryan, M.D. 2001a. Analysis of the aphthovirus 2A/2B polyprotein 'cleavage' mechanism indicates not a proteolytic reaction, but a novel translational effect: a putative ribosomal 'skip'. J. Gen. Virol. 82: 1013-1025. Donnelly, M.L.L., Hughes, L.E., Luke, G., Li, X., Mendoza, H., ten Dam, E., Gani, D. and Ryan, M.D. 2001b. The 'cleavage' activities of FMDV 2A site-directed mutants and naturally-occurring '2A-Like' sequences. J. Gen. Virol. 82: 1027-1041. Dougherty, W.G. and Semler, B.L. 1993. Expression of virus-encoded proteinases: functional and structural similarities with cellular enzymes. Microbiol. Rev. 57: 781-822. Etchison, D., Milburn, S.C., Edery, I., Sonenberg, N. and Hershey, J.W.B. 1982. Inhibition of HeLa cell protein synthesis following poliovirus infection correlates with the proteolysis of a 200,000-dalton polypeptide associated with eukaryotic initiation factor 3 and cap binding protein complex. 1 BioI. Chern. 257: 14806-14810. Falk, M,M., Grigera, P.R., Bergmann, I.E., Zibert, A., Multhaup, G. and Beck, E. 1990. Foot-and-mouth disease virus protease 3C induces specific proteolytic cleavage of host histone H3. J. Virol. 64: 748-756. Flint, M. 1995. A study of foot-and-mouth disease virus polyprotein processing. PhD Thesis, University of Reading, U.K. Gorbalenya,A.E., Svitkin, Y.V., Kazachkov, Y.A. and Agol, V.1. 1979. Encephalomyocarditis virus-specific polypeptide p22 is involved in the processing of the viral precursor polypeptides. FEBS Letters. 108: 1-5. Gorbalenya, A.E. and Svitkin, Y.V. 1983. Encephalomyocarditis virus protease: purification and role of the SH groups in processing of the precursor of structural proteins. Biochemistry (USSR) 48: 385-395. Gorbalenya, A.E., Blinov, V.M. and Donchenko, A.P. 1986. Poliovirus-encoded proteinase 3C: a possible evolutionary link between cellular serine and cysteine proteinase families. FEBS Letters. 194: 253-257. Gorbalenya, A.E., Donchenko, A.P., Blinov, V.M. and Koonin, E.V. 1989. Cysteine proteases of positive strand RNA viruses and chymotrypsin-like serine proteases. A distinct protein superfamily with a common structural fold. FEBS Letters. 243: 103-114. Gorbalenya, A.E., Koonin, E.V. and Lai, M. M-C. 1991. Putative papain-related thiol protease of positive strand RNA viruses: identification of rubi- and aphthovirus proteases and delineation of a novel conserved domain associated with proteases of rubi-, alpha- and coronaviruses. FEBS Letters. 288: 201-205. Govan, V.A., Leat, N., Allsopp, M. and Davison, S. 2000. Analysis of the complete genome sequence of acute bee paralysis virus sho\vs that it belongs to the novel group of insectinfecting RNA viruses. Virology. 277: 457-463. Gradi, A., Svitkin, Y.V., Imataka, H. and Sonenberg, N. 1998. Proteolysis of human eukaryotic translation initiation factor eIF4GII, but not eIF4GI, coincides with the shutoff of host protein synthesis after poliovirus infection. Proc. Natl. Acad. Sci. U.S.A. 95: 11089-11094.

Copyright © 2004 By Horizon Bioscience

Rvan at al.

711

Grigera, P.R. and Tisminetzky, S.G. 1984. Histone H3 modification in BHK cells infected with foot-and-mouth disease virus. Virology. 136: 10-19. Grubman, M.J., Zellner, M., Bablanian, G., Mason, P.W. and Piccone, M.E. 1995. Identification of the active-site residues of the 3C proteinase of foot-and-mouth disease virus. Virology. 213: 581-589. Guarne, A., Tormo, J., Kirchweger, R., Pfistermueller, D., Fita, I. and Skern, T. 1998. Structure of the foot-and-mouth disease virus leader protease: a papain- like fold adapted for self-processing and eIF4G recognition. EMBO J. 17: 7469-7479. Guarne, A., Hampoelz, B., Glaser, W., Carpena, X., Tormo, J., Fita, I. and Skern, T. 2000. Structure and biochemical features distinguish the foot-and-mouth disease virus leader proteinase from other papain-like enzymes. J. Mol. BioI. 302: 1227-1240. Hackett, P.B., Egberts, E. and Traub, P. 1978. Translation of ascites and mengovirus RNA in fractionated cell free systems from uninfected and mengovirus-infected Ehrlich-ascitestumor cells. Eur. J. Biochem. 83: 341-352. Hagiwara, K., Kobayashi, J., Tomita, M. and Yoshimura, T. 2001. Nucleotide sequence of genome segment 5 from Bombix mori cypovirus 1. Arch. Virol. 146: 181-187. Hammerle, T., Hellen, C.U.T. and Wimmer, E. 1991. Site-directed mutagenesis of the putative catalytic triad of poliovirus 3C proteinase. J. BioI. Chern. 266: 5412-5416. Harmon, S.A., Updike, W., Jia, X.Y., Summers, D.F. and Ehrenfeld, E. 1992. Polyprotein processing in cis and in trans by hepatitis A virus 3C protease cloned and expressed in Escherichia coli. J. Virol. 66: 5242-5247. Hasan, G., Turner, M. J. and Cordingley, J. S. 1984. Complete nucleotide sequence of an unusual mobile element from Trypanosoma brucei. Cell. 37: 333-341. Hentze, M.W. 1997. eIF4G: a multipurpose ribosome adapter? Science. 275: 500-501. Isa\va, H., Asano, S., Sahara, K., Iizuka, T. and Bando, H. 1998. Analysis of genetic information of an insect picorna-like virus, infectious flacherie virus of silkworm: evidence for evolutionary relationships among insect, mammalian and plant picorna(-like) viruses. Arch. Virol. 143: 127-143. Jackson, R.J. 1986. A detailed kinetic analysis of the in vitro synthesis and processing of encephalomyocarditis virus products. Virology. 149: 114-127. Jackson, R.J. and Wickens, M. 1997. Translational controls impinging on the 5'-untranslated region and initiation factor proteins. Curro Opin. Genet. Dev. 7: 233-241. James, V.L.A., Lambden, P.R., Deng, Y., Caul, E.O. and Clarke, LN. 1999. Molecular characterization of human group C rotavirus genes 6, 7 and 9. J. Gen. Virol. 80: 31813187. Jia, X-Y., Ehrenfeld,E. and Summers, D.F. 1991. Proteolytic activity of hepatitis A virus 3C protein. J. Virol. 65: 2595-2600. Jiang, B., Tsunemitsu, H., Gentsch, J.R., Saif, L.J. and Glass, R.I. 1993. Nucleotide sequences of genes 6 and 10 of a bovine group C rotavirus. Nuc. Acids Res. 21: 2250. Joachims, M., Harris, K.S. and Etchison, D. 1995. Poliovirus protease 3C mediates cleavage of microtubule-associated protein 4. Virology. 211: 451-461. Joachims, M., Van breugel, P.C. and lloyd, R. E. 1999. Cleavage of poly(A)-binding protein by enterovirus proteases concurrent with inhibition of translation in vitro. J. Virol. 73: 718-727. Johansson, S., Niklasson, N., Maizel, J., Gorbalenya, A.E. and Lindberg, A.M. 2002. Molecular analysis of three Ljungan virus isolates reveals a new, close-to-root lineage of the Picornaviridae with a cluster of two unrelated 2A proteins. J. Virol. 76: 89208930. Johnson, K.N. and Christian, P.D. 1998. The novel genome organization of the insect picorna-like virus, Drosophila C virus, suggests this virus belongs to a previously undescribed virus family. J. Gen. Virol. 79: 191-203.

Copyright © 2004 By Horizon Bioscience

\72

FMDV Proteinases

Jore, J., De Geus, B., Jackson, R.J., Pouwels, P.H. and Enger-Valk, B.E. 1988. Poliovirus protein 3CD is the active protease for processing of the precursor protein PI in vitro. J. Gen. Virol. 69: 1627-1636. Kean, K.M., Tetrina, N.L. Marc, D. and Girad, M. 1991. Analysis of putative active site residues of the poliovirus 3C protease. Virology. 163: 330-340. Kerekatte, V., Keiper, B.D., Badorff, C., Cai, A., Knowlton, K.U. and Rhoads, R.E. 1999. Cleavage of Poly(A)-binding protein by coxsackievirus 2A protease in vitro and in vivo: another mechanism for host protein synthesis shutoff? J. Virol. 73: 709-717. Kirchweger, R., Ziegler, E., Lamphear, B.J., Waters, D., Liebig, H-D., Sommergruber, W., Sobrino, F., Hohenadl, C., Blaas, D., Rhoads, R.E. and Skern, T. 1994. Foot-and-mouth disease virus leader proteinase: purification of the Lb form and determination of its cleavage site on eIF-4y. J. Virol. 68: 5677-5684. Korant, B.D. 1972. Cleavage of viral precursor proteins in vivo and in vitro. J. Virol. 10: 751-759. Korant, B.D. 1973. Cleavage of poliovirus-specific polypeptide aggregates. J. Virol. 12: 556-563. Korant, B.D., Brzin, J. and Turk, V. 1985. Cystatin, a protein inhibitor of cysteine proteases alters viral protein cleavages in infected human cells. Biochem. Biophys. Res. Comm. 127: 1072-1076. Kusov, Y.Y. and Gauss-Muller, V. 1997. In vitro RNA binding of the hepatitis A virus proteinase 3C (HAV 3CPTO) to secondary structure elements within the 5' terminus of the HAV genome. RNA. 3: 291-302. Lall, M.S., Ramtohul, Y.K., James, M.N.G. and Vederas, J.C. 2002. Serine and threonine beta-Iactones: A new class of hepatitis a virus 3C cysteine proteinase inhibitors. 1. Organic Chern. 67: 1536-1547. Lamphear, B.1., Yan, R., Yang, F., Waters, D., Liebig, H-D., Klump, H., Keuchler, E., Skern, T. and Rhoads, R.E. 1993. Mapping of the cleavage site in protein synthesis initiation factor eIF-4y of the 2A proteases from human coxsackievirus and rhinovirus. J. BioI. Chern. 268: 19200-19203. Lawrence, C. and Thatch, R.E. 1975. Identification of a viral protein involved in posttranslational maturation of the encephalomyocarditis virus capsid precursor. J. Virol. 15: 918-928. Lawson, T.G., Smith, L.L., Palmenberg, A.C., and Thach, R.E. 1989. Inducible expression of encephalomyocarditis virus 3C protease activity in stably transformed mouse cell lines. J. Virol. 63: 5013-5022. Lawson, M.A. and Semler, B.L. 1991. Poliovirus thiol proteinase 3C can utilize a serine nucleophile within the putative catalytic triad. Proc. Natl. Acad. Sci. USA. 88: 99199923. Lawson, M.A. and Semler, B.L. 1992. Alternate poliovirus nonstructural protein processing cascades generated by primary sites of 3C proteinase cleavage. Virology. 191: 309-

320. Leong, L.E.C., Walker, P.A. and Porter, A.G. 1993. Human rhinovirus-14 protease 3C (3CPTO) binds specifically to the 5'-noncoding region of the viral RNA. J. BioI. Chern. 268: 25735-25739. Li, F., Browning, G.F., Studdert, M.1. and Crabb, B.S. 1996. Equine rhinovirus 1 is more closely related to foot-and-mouth disease virus than to other picornaviruses. Proc. Nat!. Acad. Sci. USA. 93: 990-995. Li, W., Ross-Smith, N., Proud, C.G.and Belsham, G.J. 2001. Cleavage of translation initiation factor 4AI (eIF4AI) but not eIF4AII by foot-and-mouth disease virus 3C protease: identification of the eIF4AI cleavage site. FEBS Lett. 507: 1-5. Malcolm, B.A. 1995. The picornaviral 3C proteinases: cysteine nucleophiles in serine proteinase folds. Protein Sci. 4: 1439-1445.

Copyright © 2004 By Horizon Bioscience

Rvan et al.

73

Martin, F., Maranon, C., Olivares, M., Alonso, C. and Lopez, M.C. 1995. Characterization of a non-long terminal repeat retrotransposon cDNA (Ll Tc) from Trypanosoma cruzi: homology of the first ORF with the APE family of DNA repair enzymes. J. Mol. BioI. 247: 49-59. Martinez-Salas, E. and Domingo, E. 1995. Effect of expression of the aphthovirus protease 3C on viral infection and gene expression. Virology. 212: 111-120. Mason, P.W., Grubman, M.J. and Baxt, B. 2003. Molecular basis of pathogenesis ofFMDV. Virus Res. 91: 9-32. Matthews, D.A., Smith, W.W., Ferre, R.A., Condon, B., Budahazi, G., Sisson, W., Villafranca, J.E., Janson, C.A., McElroy, H.E., Gribskov, C.L. and Worland, S. 1994. Structure of human rhinovirus 3C protease reveals a trypsin-like polypeptide fold, RNA-binding site, and means for cleaving precursor polyprotein. Cell. 77: 761-771. Matthews, D., Dragovich, P.S., Webber, S.E., Fuhrman, S.A., Patick, A.K., Zalman, L.S., Hendrickson, T.F., Love, R.A., Prins, T.J., Marakovits, J.T., Zhou, R., Tikhe, J., Ford, C.E., Meador, J.W., Ferre, R.A., Brown, E.L., Binford, S.L., Brothers, M.A., Delisle, D.M. and Worland, S.T. 1999. Structure-assisted design of mechanism-based irreversible inhibitors of human rhinovirus 3C protease with potent antiviral activity against multiple rhinovirus serotypes. Proc. Nat. Acad. Sci. USA. 96: 11000- 11007. Medina. M., Domingo, E., Brangwyn, J.K. and Belsham, G.J. 1993. The two species of the foot-and-mouth disease virus leader protein, expressed individually, exhibit the same activities. Virology. 194: 355-359. Morely, S.J., Curtis, P.S. and Pain, V.M. 1997. eIF4G: translation's mystery factor begins to yield its secrets. RNA. 3: 1085-1104. Mosimann, S.C., Cherney, M.M., Sia, S., Plotch, S. and James, M.N. 1997. Refined X-ray crystallographic structure of the poliovirus 3C gene product. J. Mol. BioI. 273: 10321047. Murphy, N.B., Pays, A., Tebabi, P., Coquelet, H., Guyaux, M., Steinert, M. and Pays, E. 1987. Trypanosoma brucei repeated element with unusual structural and transcriptional properties. J. Mol. BioI. 195: 855-871. Palmenberg, A.C., Pallansch, M.A. and Rueckert, R.R. 1979. Protease required for processing picornaviral coat protein residues resides in the viral replicase gene. J. Virol. 32: 770-778. Palmenberg, A.C., Parks, G.D., Hall, D.J., Ingraham, R.H., Seng, T.W. and Pallai, P.V. 1992. Proteolytic processing of the cardioviral P2 region: primary 2A12B cleavage in clonederived precursors. Virology. 190: 754-762. Parks, G.D., Baker, J.C. and Palmenberg, A.C. 1989. Proteolytic cleavage of encephalomyocarditits virus capsid region substrates by precursors to the 3C enzyme. J. Virol. 63: 1054-1058. Patick, A.K., Binford, S.L., Brothers, M.A., Jackson, R.L., Ford, C.E., Diem, M.D., Maldonado, F., Dragovich, P.S., Zhou, R., Prins, T.J., Fuhrman, S.A., Meador, J.W., Zalman, L.S., Matthews, D.A. and Worland, S.T. 1999. In vitro antiviral activity of AG7088, a potent inhibitor of human rhinovirus 3C protease. Antimicrob. Agents Chemother. 43: 2444-2450. Pelham, H.R.B. 1978. Translation of encephalomyocarditis virus RNA in vitro yields an active proteolytic processing enzyme. Eur. J. Biochem. 85: 457-462. Petersen, J.F.W., Cherney, M.M., Liebig, H.-D., Skern, T., Kuechler, E. and James, M.N.G. 1999. The structure of the 2A proteinase from a common cold virus: a proteinase responsible for the shut-off of host cell protein synthesis. EMBO J. 18: 5463-5475. Piccone, M.E., Zellner, M., Kumosinski, T.F., Mason, P.W. and Grubman, M.J. 1995a. Identification of the active-site residues of the L proteinase of foot-and-mouth disease virus. J. Virol. 69: 4950-4956.

Copyright © 2004 By Horizon Bioscience

\74

FMDV Proteinases

Piccone, M.E., Rieder, E., Mason, P.W. and Grubman, M.J. 1995b. The foot-and-mouth disease virus leader proteinase gene is not required for viral replication. J. Virol. 69: 5376-5382. Pringle, F.M., Gordon, K.H.J., Hanzlik, T.N., Kalmakoff, J., Scotti, P.D. and Ward, V.K. 1999. Anovel capsid expression strategy for Thosea asigna virus (Tetraviridae). J. Gen. Virol. 80: 1855-1863. Pringle, F.M., Johnson, K.N., Goodman, C.L., McIntosh, A.H. and Ball, L.A. 2003. Providence virus: a new member of the tetraviridae that infects cultured insect cells. Virology. 306: 359-370. Qian, Y., Jiang, B., Saif, L.J., Kang, S.Y., Ojeh, C.K. and Green, K.Y. (1991). Molecular analysis of the gene 6 from a porcine group C rotavirus that encodes the NS34 equivalent of group A rotaviruses. Virology. 184: 752-757. Ramabhadran, T.V. and Thatch, R.E. 1981. Translational elongation rate changes in encephalomyocarditis virus-infected and interferon-treated cells. J. Virol. 39: 573583. Ramtohul, Y.K., Martin, N.!., Silkin, L., James, M.N.G. and Vederas, J.C. 2001. Pseudoxazolones, a new class of inhibitors for cysteine proteinases: inhibition of hepatitis A virus and human rhinovirus 3C proteinases. Chern. Comm. 24: 2740-2741. Ramtohul, Y.K., James, M.N.G. and Vederas lC. 2002a. Synthesis and evaluation of ketoglutamine analogues as inhibitors of hepatitis A virus 3C proteinase. J. Organic Chern. 67: 3169-3178. Ramtohul, Y.K., Martin, N.!., Silkin, L., James, M.N.G, and Vederas, J.C. 2002b. Synthesis of pseudoxazolones and their inhibition of the 3C cysteine proteinases from hepatitis A virus and human rhinovirus-14. J. Chern. Soc. Perkin Transactions 1: 1351-1359. Rao, S., Shapiro, M., Lynn, D., Hagiwara, K., Blackmon, B., Fang, G. and Carner,G.R. 2001. Identification of dsRNA electrophoretypes of two cypoviruses from a dual infection in gypsy moth, Lymantria dispar. EMBL database - accession no. AF389466. Roberts, P.1. and Belsham, G.1. 1995. Identification of critical amino acids within the footand-mouth disease virus leader protein, a cysteine protease. Virology. 213: 140-146. Roos, R.P., Kong, W. and Semler, B.L. 1989. Polyprotein processing of Theilers murine encephalomyelitis virus. J. Virol. 63: 5344-5353. Ryan, M.D., Belsham,G.J. and King, A.M.Q. 1989. Specificity of substrate-enzyme interactions in foot-and-mouth disease virus polyprotein processing. Virology. 173: 35-45. Ryan, M.D., King, A.M.Q. and Thomas, G.P. 1991. Cleavage of foot-and-mouth disease Virus polyprotein is mediated by residues located within a 19 amino acid sequence. J. Gen. Virol. 72: 2727-2732. Ryan, M.D. and Drew, J. 1994. Foot-and-mouth disease virus 2A oligopeptide mediated cleavage of an artificial polyprotein. EMBO 1. 13: 928-933. Ryan, M.D. and Flint, M. 1997. Virus-encoded proteinases of the picornavirus super-group. J. Gen. Virol. 78: 699-723. Ryan, M.D., Donnelly, M.L.L., Lewis, A., Mehrotra, A.P., Wilkie, J. and Gani, D. 1999. A model for non-stoichiometric, co-translational protein scission in eukaryotic ribosomes. Bioorganic Chern. 27: 55-79. Ryan, M.D., Luke, G., Hughes, L.E., Cowton, V.M., ten Dam, E., Li, X., Donnelly, M.L.L., Mehrotra, A. and Gani, D. 2002. The aphtho- and cardiovirus "primary" 2A/2B polyprotein "cleavage". In: Molecular Biology of Picornaviruses. B.L. Semler and E. Wimmer, eds. ASM Press, Washington, USA. 213-223. Sachs, A.B., Sarnow, P. and Hentze, M.W. 1997. Starting at the beginning, middle, and end: translation initiation in eukaryotes. Cell. 89: 831-838. Sangar, D.V., Newton, S.E., Rowlands, D.1. and Clarke, B.E. 1987. All FMDV serotypes initiate protein synthesis at two separate AUGs. Nuc. Acids Res. 15: 3305-3315.

Copyright © 2004 By Horizon Bioscience

Rvan et al.

751

Sangar, D.V., Clark, R.P., Carroll, A.R., Rowlands, DJ. and Clarke, B.E. 1988. Modification of the leader protein (Lb) of foot-and-mouth disease virus. J. Gen. Virol. 69: 23272333, Seipelt, J., Guarne, A., Bergmann, E., James, M., Sommergruber, W., Fita, 1. and Skern, T. 1999. The structures of picornaviral proteinases. Virus. Res. 62: 159-168. Seipelt, J., Liebig, H-D., Sommergruber, W., Gerner, C. and Kuechler, E. 2000. 2A proteinase of human rhinovirus cleaves cytokeratin 8 in infected HeLa cells. J. BioI. Chern. 275: 20084-20089. Skem, T., Hampolz, B., Guarne, A., Fita, I., Bergmann, E., Petersen, J. and James, M.N.G. 2002. Structure and function of picornavirus proteinases. In: Molecular Biology of Picornaviruses. B.L. Semler and E. Wimmer, eds. ASM Press, Washington, USA.. 199-212. Strebel, K. and Beck, E. 1986. A second protease of foot-and-mouth disease virus. J. Virol. 58: 893-899. Summers, D.F., Maizel, J.V. and Darnell, J.E. 1967. Decrease in size and synthetic activity of poliovirus polysomes late in the infectious cycle. Virology. 31: 427-435. Summers, D.F. and Maizel, J.V. 1968. Evidence for large precursor proteins in poliovirus synthesis. Proc. Natl. Acad. Sci. USA. 59: 966-971. Summers, D.F., Shaw, E.N., Stewart, M.L. and Maizel, J.V. 1972. Inhibition of cleavage of large poliovirus-specific precursor proteins in infected HeLa cells by inhibitors of proteolytic enzymes. J. Virol. 10: 880-884. Svitkin, Y.V., Gradi, A., Imataka, H., Morino, S. and Sonenberg, N. 1999. Eukaryotic initiation factor 4GII (eIF4GII), but not eIF4GI, cleavage correlates with inhibition of host cell protein synthesis after human rhinovirus infection. J. Virol. 73: 3467-3472. Tesar, M. and Marquardt, O. 1990. Foot-and-mouth disease virus protease 3C inhibits cellular transcription and mediates cleavage of histone H3. Virology 174: 364-374. Vakharia, V.N., DeVaney, M.A., Moore, D.M., Dunn, J.J. and Grubman, MJ. 1987. Proteolytic processing of foot-and-mouth disease virus polyproteins expressed in a cellfree system from clone derived transcripts. J. Virol. 61: 3199-3207. Walker, P.A., Leong, L.E.C. and Porter, A.G. 1995. Sequence and structural determinants of the interaction between the 5'-noncoding region of picornavirus RNA and rhinovirus protease 3C. J. BioI. Chern. 270: 14510-14516. Wang, Q.M. 1999. Protease inhibitors as potential antiviral agents for the treatment of picornaviral infections. Prog. Drug Res. 52: 197-219. Wilson, J.E., Powell, MJ., Hoover, S.E. and Sarnow, P. 2000. Naturally occurring dicistronic cricket paralysis virus RNA is regulated by two internal ribosome entry sites. Mol. Cell. BioI. 20: 4990-4999. Wu, C.-Y., Lo, C.-F., Huang, C.-J., Yu, H.-T and Wang, C.-H. 2002. The complete genome sequence of Perina nuda Picorna-like viurs, an insect-infecting RNA virus with a genome organization similar to that of the mammalian picornaviruses. Virology. 294: 312-323. Wutz, G., Auer, H., Nowotny, N., Grosse, B., Skern, T. and Kuechler, E. 1996. Equine rhinovirus serotypes 1 and 2: relationship to each other and to aphthoviruses and cardioviruses. J. Gen. Virol. 77: 1719-1730. Yalamanchili, P., Datta, U. and Dasgupta, A. 1997a. Inhibition of host cell transcription by poliovirus: cleavage of transcription factor CREB by poliovirus-encoded protease 3cpro. J. Virol. 71: 1220-1226. Yalamanchili, P., Banerjee, R., and Dasgupta, A. 1997b. Poliovirus-encoded protease 2APro cleaves the TATA-binding protein but does not inhibit host cell RNA polymerase II transcription in vitro. J. Virol. 71: 6881-6886. Yalamanchili, P., Weidman, K. and Dasgupta, A. 1997c. Cleavage of transcriptional activator Oct-l by poliovirus encoded protease 3CPro. Virology. 239: 176-185.

Copyright © 2004 By Horizon Bioscience

\76

FMDV Proteinases

Ypma-Wong, M.F., Dewalt, P.G., Johnson, V.H., Lamb, J.G. and Semler, B.L. 1988. Protein 3CD is the major poliovirus proteinase responsible for cleavage of the PI capsid precursor. Virology. 166: 265-270. Zeddam, J.-L. A., Pringle, F.M., Gordon, K.H., Ward, V.K., Luke, B.T., Gorbalenya, A.E. and Hanzlik, T.N. 2001. Genome organization of Euprosterna elaeasa virus defines it as a member of a new group of insect RNA viruses. Accession no. AF461742 : unpublished. Zell, R., Dauber, M., Krumbholz, A., Henke, A., Birch-Hirschfeld, E., Stelzner, A., Prager, D. and Wurm, R. 2001. Porcine teschoviruses comprise at least eleven distinct serotypes: molecular and evolutionary aspects. 1. Virol, 75: 1620-1631. Ziegler, E., Borman, A.M., Kirchweger, R., Skem, T. and Kean, K.M. 1995. Foot-andmouth disease virus Lb proteinase can stimulate rhinovirus and enterovirus IRESdriven translation and cleave several proteins of cellular and viral origin. J. Virol. 69: 3465-3474.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 4 Structure of Foot-and-Mouth Disease Virus Particles

Cristina Ferrer-Orta and Ignacio Fita

Abstract X-ray structures, at almost atomic resolution, of Foot-and-Mouth-Disease Virus (FMDV) particles from several serotypes and subtypes are now available mainly from the results obtained by the Oxford group during about the last 15 years. FMDVs show many of the structural features generally found in picornaviruses with virions forming icosahedral shells composed of 60 copies each of four structural proteins VPl-VP4. The disposition of the three larger proteins, VP1-VP3, follows a pseudo T=3 architecture (P=3) closely related to the one first found in small, RNA, plant viruses. The arrangement is possible because of the similar topology of VP1, VP2 and VP3, which adopt the wedge-shaped eight-stranded ~-barrel fold characteristic of most RNA viruses. The chain length and conformation of VP4 is quite variable among picornaviruses, though the protein is always internal and has an N-terminal myristoyl group that in the FMDV structures remains undefined. Despite the important features in common with picornaviruses, FMDV presents some major differences that can often be related to functional or biological peculiarities. In particular, the canyon or pit found in most picornaviruses is absent in aphthoviruses, which place the integrin cell attachment site containing the Arg-Oly-Asp (ROD) motif in the protruding, fully exposed and highly immunogenic OH loop from VPl, also called the "FMDV loop". Some flexibility seems required for the optimal biological functionality of this loop that has always been found disordered in the crystal structures of the unperturbed virions. However, the structure of the loop has been trapped, both in crystals of reduced virions from the 0 serotype, and in crystals of peptide complexes with neutralizing antibodies against the C serotype. The self-contained structure of the "FMDV loop" together with its hinge flexibility and with the recent availability of the structure of the ectodomain from integrin avf33 suggest mechanisms for cell receptor-virus recognition and specificity. 1. Introduction The first structure determination of an intact FMDV particle at almost atomic resolution was achieved with X-ray crystallography by the Oxford group (Acharya et al., 1989) without experimental phases and using the structural information that had just been obtained for other picornaviruses (Rossmann et ai, 1985; Hogle et al., 1985; Luo et al., 1987). Since that first analysis new crystallographic studies on FMDV have provided the structures of other serotypes and subtypes (Lea et al.,

Copyright © 2004 By Horizon Bioscience

\78

Structure of FMDV Particles

1994; Lea et al., 1995; Curry et al., 1996), of monoclonal-antibody-resistant mutants (MAR mutants) (Parry et al., 1990), of reduced virus (Logan et al., 1993), of empty capsids (Curry et al., 1997) and of complexes with oligosaccharide receptors (Fry et al., 1999). Cryo-electron microscopy studies have also contributed to the structural information mainly on FMDVs-neutralizing antibodies complexes (Hewat et al., 1997; Verdaguer et al., 1999). All this information, analyzed in perspective with the wealth of structural and biological data that is been accumulated for other picornaviruses (Rossmann et al., 1989; Rossmann et al., 2002), allows a deep understanding of some of the FMDV peculiarities. However, in FMDV, and in picornaviruses in general, at least two major structural issues remain elusive for high resolution analysis: i) the organization of the genomic RNA and its relationships with the viral protein shell and ii) the interactions of the virus particles with cellular receptors and the related molecular processes that determine cell tropism, receptor specificity and, subsequently, internalization and capsid dissociation. For these two issues, cryo-electron microscopy (cryo-EM) holds promising possibilities, as it is been shown for a number of picornaviruses at ever increasing resolutions (Rossmann et al., 2000) while, in general, X-ray crystallography does not appear to be very suitable. Departures from the icosahedral symmetry not affecting crystal packing, which to some extent is always the case for the picornavirus genome constituted by a single RNA molecule, would result in the structural information been averaged away during the crystallographic analysis. Suitable crystals of virus-receptor complexes appear to be intrinsically difficult to obtain mainly due to the nonstable or non-homogeneous nature of most of these complexes. At present, only two types of virus-receptor structures have been determined at high resolution by X-ray crystallography: the Human-Rhinovirus serotype 2 (HRV2) in complex with constructs from the very-Low-Density-Lipoprotein (vLDL) (Verdaguer et al., 2004) and the FMDV serotype 0 complexed with oligosaccharides (Fry et al., 1999). However structural information for the interaction of FMDV with integrins (considered the key cellular receptors in infected animals [see below]) still remains very limited. The present chapter attempts to summarize the structural information presently available on FMDV, focusing mainly on the differences to other picornaviruses. Structural aspects related to FMDV antigenic properties, including the organization of epitopes and virus-antibodies interactions, are treated extensively in Chapter 9 of this book and will be only briefly mentioned here. In turn, some of the ideas presented in the excellent recent review about FMDV with the title "Structure and receptor binding" published by the Oxford group (Jackson et al., 2003) have, for the sake of completeness, been incorporated into this chapter.

2. Structural Overview of Picornaviruses In picornaviruses structural proteins account for approximately one-third ofthe polyprotein that results from the intact translation of the single-stranded positive-sense viral genome (see also in Chapter 2). The structural proteins are encoded towards the 5' end of the open reading frame that, even prior to the specific cleavages by virus-encoded proteases, appears to fold into a structure antigenically closely related to the viral proteins VPO, VP1 and VP3 (Jackson et al., 2003). During Copyright © 2004 By Horizon Bioscience

Ferrer-Oria and Fifa

791

(a)

(b)

(c)

_a-helix

N

-t> p-burel

(d) Figure 1. (a) Disposition of the structural proteins in (left) SBMV (Abad-Zapatero et a/., 1980) and (right) FMDV with respect to the icosahedral symmetry axes (in yellow). SBMV, with a T=3 organization, contains three protein subunits, named as A Band C, in the icosahedral asymmetric unit. Relationships between these structurally independent subunits are defined by the quasi twofold and three-fold axes (in white), according to the quasi-equivalence principles (Caspar et al.., 1962). FMDV, as all the picornaviruses, presents a T=l organization of its structural proteins VPI-VP4, though the similar organization and topology of VPI-VP3 (labels correspond to the biological protomer) with respect to SBMV proteins A Band C makes it appropriated to talk about a pseudo T=3 (P=3) virus architecture. (b) Structural proteins related by the quasi-three fold axes are shown for SBMV (left) and FMDV (right). (c) Ribbon representation of the B subunit in SBMV (left) and of the structural proteins in FMDV (right): VPl (blue), VP2 and VP4 (green and dark green, respectively) and VP3 (red). (d) Topology of the wedge-shaped eight-stranded f3-barrel fold commonly found in RNA viruses (see text). See Colour Plate at the back of the book. Copyright © 2004 By Horizon Bioscience

\80

Structure of FMDV Particles

the RNA encapsidation VPO is finally cleaved into VP4 (the N-terminal region) and VP2. The mature virion, of about 8.5*106 Da in mass, consists of icosahedral capsids with an external diameter of around 300 A composed of 60 copies each of the four proteins VP1-VP4 and of a copy of the RNA genome (Figure 1). High-resolution structures of a few serotypes of many different picornaviruses have been solved by X-ray crystallography starting with the determinations of a human rhino- (HRV14) (Rossmann et at., 1985) and polio- (PVl) (Hogle et at., 1985) virus serotypes. The organization of the three major structural proteins, VPlVP3, follows a pseudo T=3 architecture (P=3) (Caspar et aI., 1962; Rossmann et aI., 1989), closely related to the one first found in the structures of the small, RNA, plant viruses TBSV (Winkler et aI., 1977) and SBMV (Abad-Zapatero et al., 1980) (Figure 1). The P=3 arrangement is possible in picornaviruses because of the broadly similar structure of the VPl, VP2 and VP3 cores. These proteins adopt the wedge-shaped eight-stranded ~-barrel fold that is often referred to as the RNA virus fold (RVF) because of its almost ubiquitous presence in RNA viruses (Abad-Zapatero et at., 1980; Rossmann et aI., 1989). The four strands of the two ~-sheets are conventionally labeled sequentially as BlDG and CHEF, respectively. Loops connecting the strands, which are highly variable and responsible of most of the distinctive features of the virus exterior surfaces, are identified by the strands they join. In picornaviruses the VP1 proteins are mostly located around the icosahedral5-fold axes while VP2 and VP3 alternate around the 2- and 3-fold axes. The structural protein VP4 differs completely from VPl-VP3. VP4 is internal and its N-terminus, which contains a covalently linked myristoyl group, approaches the 5-fold axes. Among the members of the Picomaviridae family VP4 varies considerably, both in length and conformation. With a few exceptions (Hendry et aI., 1999), only about half of the VP4 protein is visible in the electron density maps. This is likely to be due to its interactions with the genomic RNA, an interaction that remains almost totally undefined in all picornavirus structures reported. The external surfaces of rhino- and enteroviruses, the two larger genera of Picornaviridae, are noteworthy for a depression or "canyon" running around each 5-fold vertex (Figures 2 and 3). For these viruses it was suggested, in the so called "the canyon hypothesis", that the site of receptor attachment would involve the more conserved amino acid residues in the canyon, a site that appears protected from host immune surveillance by the inability of neutralizing antibodies to penetrate far into the canyon on account of their larger cross section. This hypothesis it is substantiated by cryo-EM studies of virus-receptor complexes from major-

group rhinoviruses (Bella et ai., 1998; Olson et ai., 1993; Kolatkar et al., 1999) coxsackieviruses A21 and B3 (He et aI., 2001; Xiao et ai., 2001) and polioviruses (Belnap et ai., 2000; He et aI., 2000; Xing et ai., 2000). However, the hypothesis has been questioned by the finding of some overlap between the receptor- and antibody- binding sites (Smith et aI., 1996) and by the existence of receptors that bind outside the canyon in the minor-group rhinoviruses (Hofer et ai., 1994; Hewat et ai., 2000) and in some echo- and coxsackie viruses (Bergelson et ai., 1994; He et aI., 2001; Rossmann et at., 2002). "The canyon hypothesis" is also questioned by possible alternative roles for the receptor attachment into the canyon not directly related with the virus immune response. In particular, receptor attachment into the canyon might trigger the uncoating process by a finely tuned interplay with Copyright © 2004 By Horizon Bioscience

Ferrer-Oria and Fita

811

(a)

(b)

Figure 2. (a) "The canyon hypothesis" as formulated initially for picornavirus presenting a deep canyon around the five fold axes. The internal part of the canyon could contain conserved receptor sites, for ICAM-size like molecules, that would be inaccessible to antibodies for steric reasons (Rossmann et al., 1985). The pocket, located at the canyon bottom, is shown containing a pocket compound that can act as an antiviral agent by displacing the native "pocket-factor" and interfering then with virus uncoating (Rossmann and Johnson, 1989). (b) The situation in FMDV appears radically different. FMDV lacks structural features such as the canyons or "the pockets" and its RGD receptor binding site motif situated in the VPl GH loop, or "FMDV loop", is fully exposed in the viral surface. Therefore, the FMDV loop, and in particular its RGD motif, results accessible both to the Fabs from antibodies and to the even bulkier integrins, which act as the key cellular receptors for FMDV (see in the text). The need of a conserved receptor binding site that could hide from immune surveillance seems to be achieved in FMDV by camouflaging a small constant region (mainly the RGD motif) within sequences of considerable variability (see also Capter 9). The five-stranded f)-barrel, formed by the N-termini of VP3 molecules, and the myristilated N-termini of VP4, both located in the internal part of the 5-fold axes are also shown in these figures.

Copyright © 2004 By Horizon Bioscience

\82

Structure of FMDV Particles

(a)

(b) Figure 3. (a) Solid representation of pentamers from HRV 14 (left) and FMDV (right). The canyon running around the icosahedral five-fold axes in HRV14 (the circle corresponds approximately to the deepest part of the canyon) is in part filled by the VPl C-terminal residues (dark blue) in FMDV. The hole seen around the five-fold axes in FMDV is absent in HRV14 and in most picomaviruses. (b) Disposition of the structural protein VP2 around the icosahedral three-fold axis in HRV14 (left) and FMDV (right). The tight annulus found in FMDV could be stabilized by ionic interactions mediated by Ca2+ ions situated directly on the three-fold axis (see in the text). See Colour Plate at the back of the book.

the displacement of a "pocket factor" molecule. "Pocket factors", which probably correspond to a diversity of cellularly derived fatty acids, have been found in the crystalline structures of most rhino- and enteroviruses bound into a pocket below the canyon floor overlapping with the receptor binding site (Figure 2) (Rossmann et al., 2002).

3. FMDVs Structures FMDV presents the basic traits common to picomaviruses with the structural proteins VPl, VP2 and VP3 placing their (3-barrel cores at essentially the same radius and in similar orientations to those of other picornaviruses (Figure 1). However, important differences are found in the external surfaces ofVPl, VP2 and VP3 with respect to other picornaviruses (Acharya et al., 1989). In fact, as VPl and VP2 are substantially smaller in FMDV (see below) the protein shell is generally thinner and lacks most of the features at higher radius, such as the canyon or pits, which results in the relatively smooth external surface of the virions (Figures 2 and 3). Copyright © 2004 By Horizon Bioscience

Ferrer-Dria and Fita

83 I

3.1.Conformation of the Structural Proteins

FMDVs protein VPl shows the most significant rearrangements with its B-barrel canted up from the plane of the capsid towards the 5-fold axis more pronouncedly than in other picornaviruses. VPl comprises 213 residues in the serotype 0 of FMDV, whereas in HVR14 and PVl it comprises 279 and 306 residues, respectively (Acharya et al., 1989). This truncation, which originates mainly by the trimming at the loops closer to the 5-fold axis, together with the canting of the B-barrel produces in FMDV a 5-fold axis naked of VPI loops, in marked contrast to other picornaviruses (Figures 1-3). Furthermore, the outer surface of VPl lacks most of the regions that in rhinoviruses constitute the so called north wall of the canyon, the one closer to the 5-fold axis. In turn, in FMDVs the C-terminal region of VPl forms a long "arm" that runs on the surface at a constant radius to the 5-fold axis from the VPl of one subunit over VP3 of the same protomer until it nestles onto the VPI of the 5-fold related (clockwise) protomer (Figure 3). With its unique conformation the VPl C-terminal region of FMDV fills what would be analogous to the canyon or to the pit in other picornavirus structures. In FMDV there is no room in VP 1 for a cavity such as the long, hydrophobic, pocket that accommodates the pocket factor in rhino- and enteroviruses. The totally different uncoating mechanisms in apthoviruses with respect to rhino- and enteroviruses might make the presence of pocket factors unnecessary. The two VPl protein regions showing the highest mobility correspond firstly to the GH loop at about residues 141-161 in the 0 serotype, which is the virus major antigenic site, and secondly to the C-terminal region comprising residues 200-213, which is a minor viral antigenic site. The GH loop of VP1, also called the "FMDV loop", in addition to its dominant antigenic properties also contains the integrin cell attachment site with the Arg-Gly-Asp (RGD) motif i.e. residues 145146-147 in the 0 serotype. The "FMDV loop" has been found to be disordered in all the crystalline structures of unperturbed virions: possibly adopting two extreme dispositions on the viral surface referred as "up" and "down" according to the orientation of the loop with respect to the 5-fold axis (Figure 4). Between these two extremes dispositions the "FMDV loop" appears to fluctuate as a hinged domain and this flexibility is likely to be required for the optimal functionality of the loop (Jackson et al., 2003). The "FMDV loop" in the, apparently more stable, "down" conformation was visualized by the first time in the crystalline structure of the reduced 0 serotype virion (see below) lying predominantly on the surface of VP2 (Logan et aI., 1993). The loop forms a B-strand with the RGD motif in an extended conformation, occupying a turn prior to a helix (Figure 4). This loop structure and, in particular, the disposition and conformation of the ROD motif is similar to the one found in the crystal of a synthetic peptide, which corresponded to the sequence of the loop in the FMDV serotype C, in complex with an Fab fragment from a neutralizing antibody (Verdaguer et aI., 1995). Such conservation suggests that the chemical properties of residues in the "FMDV loop" as well as the conformation they adopt must be important in integrin recognition (see below and also Chapter 9 for the immunogenic properties and interactions of this loop with antibodies). Protein VP2 comprises 218 residues in FMDV (serotype 0), whereas in HVR14 and PVI it comprises 262 and 272 residues, respectively. Truncation of VP2 in FMDV with respect to other picornaviruses corresponds, as in protein VP1,

Copyright © 2004 By Horizon Bioscience

\84

Structure of FMDV Particles

(a)

(b)

(c) Figure 4. The VPI OH loop, or "FMDV loop", (in yellow) with (a) the down conformation found in the structure of the reduced 0 serotype (Logan et ai., 1993) and (b) a fully exposed disposition obtained in the structure of complexes of the C-serotype with the neutralizing antibody SD6 (Hewat et ai., 1997). The hinge flexibility of the "FMDV loop" explains the difficulties in observing the loop in the unperturbed viruses. On the left, views are along the 5-fold axes, while a lateral view of the protomer is shown on the right. Protein subunits are colored as in Figure I. Despite the hinge movement the "FMDV loop" presents an internal structure (c) that appears well conserved among different serotypes. In this structure the RGD sequence is extended and situated occupying a tum prior to a helix with the arginine and aspartic residues pointing in opposite directions. See Colour Plate at the back of the book.

Copyright © 2004 By Horizon Bioscience

Ferrer-Dria and Fita

851

mainly to the trimming of the loops that decorate the outside surface of the virions. In particular, the reduced length of the VP2 EF loop in FMDV with respect to other picornaviruses is most remarkable. As occurs in most other picornaviruses, the first few residues of the FMDV VP2 are internal and not visible in the electron density maps, again likely because of deviations from the icosahedral symmetry when those residues interact with the genomic RNA. The VP2 N-terminal residues, when they start being visible establish a very close association with other protomers around the 3-fold axis, which results in a characteristic tight annulus (Figure 3). Within this annulus, high density directly on the symmetry axis was interpreted as corresponding to a calcium ion that, by interactions with a VP2 conserved glutamic residue, could help to stabilize the icosahedral association of the pentamers (see below). In FMDV serotype 0, the VP3 protein comprises 220 residues, all of which are well defined in the electron density map. VP3 in HVR14 and PV1 are also similar sizes Le. 238 and 236 residues, respectively. In fact, as indicated before, the structure of VP3 is the most conserved among the structural proteins of picornaviruses including FMDV. The N-termini of VP3 molecules associate to form a five-stranded ~-barrel at the 5-fold axis knitting the protomers into pentamers (Figure 2). The VP4 protein comprises 85 residues in FMDV (serotype 0), whereas it has 69 residues in both HRV14 and PV1. As mentioned previously, this internal protein varies considerably among picornaviruses. In addition, it generally contains some disordered region, which in FMDV comprises as much as just less than half of the total protein; only residues 15-39 and 65-85 could be identified in the 01 structure. The myristoyl groups covalently bound to the N-termini of VP4 are not visible in the crystalline structures of FMDVs though, by analogy with other picornaviruses, it is likely that they cluster at the base of the icosahedral 5-fold axis below the VP3 N-terminal five-stranded ~-barrel (Figure 3). In fact, in FMDV the truncation of VPlleaves this VP3 ~-barrel exposed, resulting in the virion shell having a, mostly hydrophobic, hole around the icosahedral 5-fold axes. The dimensions of the hole, with an average inner diameter of about loA and a narrowest constriction of about 6 Aat the fully conserved Cys7, are compatible with the penetration of intercalating dyes and with the rapid penetration of caesium ions that would explain the highest buoyant density of FMDV among picornaviruses. The electron density confirmed the presence of Cys7 disulphide bonds giving VP3 homodimers. This is a very unique structure, which has five cysteine residues statistically disordered between ten possible positions arranged symmetrically around the 5-fold axis. At most four out of the five residues can form disulphide bridges around anyone 5-fold axis. In

serotype 0 a disulphide bond is also found linking Cys 134 of VP1, at the entrance of the FMDV loop, and VP2 Cys130. However, this second disulphide bond is only present in FMDV serotype 0 where it seems to playa role in the flexibility of the GH loop (see below). 3.2. Capsid Stability The rhino- and enteroviruses do not readily dissociate into pentameric subunits but rather uncoat via intermediates (the A-particles) in which VP4 and the N-terminus of VP1 are externalized. In tum the integrity of FMDV capsids is very vulnerable Copyright © 2004 By Horizon Bioscience

186

Structure of FMDV Particles

to low pH, dissociating into 12S pentameric assemblies. It has been proposed that the stability of FMDV is governed by polar interactions of residues from VP2 and VP3 at the pentamer boundary. In particular, the protonation state of a cluster of conserved histidines, including VP3 His142 and His145, within 15 A. of the pentamer interface would explain the sensitivity to the low pH. Cation ligation on the 3-fold axis can also contribute, as indicated before, to maintaining capsid integrity. 3.3. Immunogenicity and Links to Receptor Binding Sites

In general, all of the accessible surface of a virus would be expected to be antigenic, with antigenic variation more likely to be apparent in residues that are not essential for capsid structure or function. Indeed multiple antigenic sites have identified for the majority of picomaviruses. For FMDVs considerations about the functional and structural aspects of the interactions with antibodies are treated in detail in Chapter 9. However, it is important to emphasize here that, in total contrast to what would be expected from the canyon hypothesis, the external GH loop from VPl in FMDV acts both as the major immunogenic site and as the integrin receptor binding site. All of the antigenic sites in this loop flank the RGD triad and the dominant mode of neutralization for anti-FMDV antibodies recognizing this site is clearly by direct blocking of receptor attachment (Verdaguer et ai., 1997). To evade immune surveillance variations around the conserved attachment site appear to be sufficient, though some distant more subtle effects have also been observed, concealing the need of a constant receptor binding site by camouflaging it within a sea of variability (Acharya et al., 1989).

4. Cellular Receptors Two major classes of receptor molecules have been recognized in FMDV; these are heparan sulfate proteoglycans (HSPGs) and integrins (Baranowski et al., 2003). HSPGs are ubiquitous proteins, located at the external surface of cells, that carry a carbohydrate component of heparan sulfates (HS). These are negatively charged polymers of disaccharide repeats with random patterns of sulfation. Heparan sulfate was originally identified as a potential enhancer (or co-receptor) of cell entry by FMDV, and was later viewed as the first point of contact between the cell and the virus, on its way towards the integrin receptor (Jackson et ai., 1996; 2003). It appears now that field straints of FMDV have an absolute dependence on integrins whereas culture-adapted viruses can use either integrins or HS as

receptors with high affinity for HS variants even dispensing their RGD integrinbinding sites (Baranowski et al., 2000). Despite clearly improving growth in cultured cells, viruses that have a high affinity for HS are attenuated for cattle. However, it has been suggested that even for field straints a low or more selective affinity for HS could confer a biological advantage for the survival of the virus, perhaps via persistent infections (Fry et ai., 1999). In the reported crystalline structure of the complex of HS with particles from 0IBFS, a FMDV tissue adapted strain, the HS-binding site is formed by a shallow depression at the junction of the three major capsid proteins (VPl, VP2 and VP3), which is equivalent to the putative receptor-binding "pit" of cardioviruses (Fry et ai., 1999) (Figure Sa). The bound sugar interacts with two pre-formed sulfateCopyright © 2004 By Horizon Bioscience

Ferrer-Dria and Fita

871

(a)

(b) Figure 5. (a) Experimentally determined structure of the FMDV-oligosaccharide receptor complex (Fry et al., 1999). Oligosaccharides are represented as balls and sticks while the virus is displayed in blue either as solid spheres (on the left) or as ribbons (on the right). (b) Modeled structure of the FMDV integrin avf33 complex obtained (Verdaguer et al., 2004) by docking the integrin structure bound to the RGD tripeptide (Xiong et aI., 2002) onto the structure of the FMDV virus serotype C with its GH-Ioop almost perpendicular to the viral surface (the one shown in Figure 4b) (Hewat et aI., 1997). The integrin is represented as ribbons (yellow) with the RGD peptide explicitly shown with balls and sticks. Views, on the left side of the figures, correspond to pentamers of the complexes seen down the 5-fold axes. On the right side of the figures the viral protomers (in blue) are seen laterally together with the receptor molecule bound to it. See Colour Plate at the back of the book.

binding sites without introducing any significant conformational changes in the proteins. Residue Arg56 of VP3 appears critical in the recognition by forming ionic interactions with the two sulphate groups found in these sulfate-binding sites. This residue is a histidine in field isolates switching to an arginine in adaptation to high affinity HS during cell-culture. Despite affinity being heavily dependent on the presence of a single amino acid it seems unlikely that a single point of attachment could explain the observed affinity of 10-9 M for FMDV bound to fixed cell HS (Jackson et al., 1996), since the central trisaccharide occludes only about 280 A2 Copyright © 2004 By Horizon Bioscience

188

Structure of FMDV Particles

of solvent accessible surface on the virus. HS chains contain discrete patches, each comprising 4-8 heavily sulfated sugars suitable for binding virus, separated by about 50 A. This spacing approximately matches the distance between neighboring HS attachment sites on the virus. Therefore, the high affinity is probably achieved by the flexible sugar chain wrapping around the virus, taking in several binding sites on the virus particle. The HS binding-site is located some 15 Afrom the RGD motif in the reduced conformation and the adaptation to HSPG receptors by 01 viruses appears not to compromise their ability to make use of integrin receptors, implying that the two attachment sites can function independently of each other. However, studies with FMDV serotype C have identified residues involved in HS adaptation at widely spaced locations on the capsid leading to suggestion that there may be more than one HS binding site in this serotype (Baranowski et al., 2000). Integrins, the second class of FMDV receptor molecules, are a large family of integral membrane receptors, with a vital role in morphogenesis and tissue remodeling and repair. Integrins were identified as FMDV receptors before HSPGs and are still seen as key determinants of cell entry (Jackson et ai., 2003). Integrins are heterodimeric molecules formed by the noncovalent association of an a and a B type I transmembrane subunits, both having large extracellular domains and, in most cases, a short cytoplasmatic tail. Integrins exist in alternative low- and highaffinity states which enables them to transmit signals both into and out of cells (Humphries et al., 2003). Several integrins recognize their ligands by binding to the Arginine-Glycine-Aspartic tripeptide, the RGD motif that, as indicated before, is also found in the GH loop of FMDV. To date, three RGD-dependent integrins, avB 1, avp3, and avp6 have been reported to serve as receptors for FMDV, while evidence has been consistently negative for the RGD-dependent integrins a5B 1 and avp5 (Jackson et al., 2003). The first integrin species reported to be a cellular receptor for FMDV was avp3 (Berinstein et al., 1995), which was later shown to use an authentic RGD-dependent interaction, though the role of this integrin during infection is less clear. Instead avp6 in addition to serving as an attachment receptor has been shown to play an active role in the events that follow receptor binding, in particular during the likely uptake of viruses into the endosomes (Miller et ai., 2001). Several lines of evidence show that residues immediately following the RGD motif of FMDV can be important for receptor recognition determining, at least in part, the specificity of the integrins ligand (Jackson et al., 2003). In particular, avp6 shows a strong preference for the RGDLXXL sequence (with X representing any amino acid) (Kraft et ai., 1999), which significantly conforms to the consensus sequence for the VPl GH loop of FMDV (Jackson et ai., 2000a; 2000b). Integrin specificity appears to be directly related with tissue tropism and host range and, though one might have thought that the virus would benefit from being able to infect any cell, in the animal FMDV appears to use just one, or at most two, species out of the whole family of integrins (Jackson et al., 2003). All the structural and biochemical data that is accumulating seems to indicate that the ability of FMDV to use integrins as receptors is, in fact, dependent on the cellular mechanisms that regulate ligand-binding affinity to integrins. The determination of the stnlctures from the ectodomain of the integrin avB3 with either Ca2+, Mn 2+ or an Arg-Gly-Asp (RGD) ligand peptide provide an initial glimpse of how recognition by integrins proceeds (Xiong et ai., 2001; Xiong et al., Copyright © 2004 By Horizon Bioscience

Ferrer-Dria and Fila

891

2002). In the av~3-RGD structure the peptide inserts into a crevice between two adjacent domains, each from a different subunit, in the integrin head. The ROD sequence makes the main contact area with the integrin and the three residues participate extensively in the interaction. The main chain conformation of the ROD sequence is closely related to that of the tripeptide as found in the diverse structures available from the "FMDV loop" with the arginine and aspartic side chains pointing in opposite directions exclusively contacting the two adjacent integrin domains, respectively (Figure 4). The peptide arginine guanidinium group is held in place by a bidentate salt bridge to two aspartic residues in the integrin. In tum, the peptide aspartate primarily involves the carboxylate group, which forms the center of an extensive network of polar interactions including coordination with a Mn 2+ ion at the metal ion-dependent adhesion site (MIDAS), a five amino acid motif conserved in integrins and required by ligand recognition. Finally, the extended conformation of the peptide glycine residue appears required to bridge the interface between the integrin a and ~ subunits. Binding of the smallest recognition peptide to the integrin, which was done by crystal soaking, results in both tertiary and quaternary changes with respect to the avB3-Mn 2+ structure. Changes in the tertiary structure reshape the adjacent environment to the MIDAS motif while quaternary changes approach the two domains at the peptide binding-site. Docking models obtained just by the superimposition of the ROD motif from both the avB3-ROD structure and the "FMDV loop" suggest integrin binding can take place when the loop adopts an exposed conformation similar to the one found when the loop is recognized by neutralizing antibodies (Verdaguer et al., 2004) (Figure 5). The peripheral location of the integrin binding site in FMDV is unlikely to impose mechanical strain on the capsid upon receptor binding, as is thought to occur in other picomaviruses. This could explain the need of the acid lability in FMDV, which would facilitate disassembly in the low pH environment of endosomes. Despite the difficulties in obtaining atomic resolution data from FMDV-integrin complexes the availability of the integrin structures might facilitate interpretation of cryo-EM reconstructions if these can be obtained from these complexes (Jackson et ai., 2003). 5. Conclusions From the several X-ray structures of FMDV particles determined is now clear that many of the structural features generally found in picomaviruses are also present in apthoviruses. These may be structurally dictated by the need to form a stable capsid and to maintain a suitable control during virus assembly and uncoating. However, FMDVs present some important structural peculiarities, which appear closely related to the biology of these viruses. In particular, the canyon or pit found in most picomaviruses is absent in aphthoviruses, which place the integrin cell attachment site containing the ROD motif in the protruding and highly immunogenic "FMDV loop". In this situation "concealment" of the constant receptor binding site seems to be achieved by camouflaging the small and constant receptor region within highly variable sequences. The self-contained structure of the "FMDV loop" together with its hinge flexibility and with the recent availability of the avB3 structures suggest mechanisms for integrin-virus recognition and specificity.

Copyright © 2004 By Horizon Bioscience

\90

Structure of FMDV Particles

6. Acknowledgements Many thanks are given to Dra. N. Verdaguer for assistance with the figures and for many valuable suggestions. This work was supported by grant BI02002-04419 to J.F. References Abad-Zapatero, c., Abdel-Meguid, S.S., Johnson, J.E., Leslie, A.G.W., Rayment, I., Rosmann, M.G., Suck, D. and Tsukihara, T. 1980. Structure of southern bean mosaic virus at 2.8 Aresolution. Nature. 286: 33 -39. Acharya, R., Fry, E., Stuart, D., Fox, G., Rowlands, D. and Brown, F. 1989 The threedimensional structure of foot-and-mouth disease virus at 2.9 A resolution. Nature. 337: 709-716. Baranowski, E., Ruiz-Jarabo, C.M., Sevilla, N., Andreu, D., Beck, E. and Domingo, E. 2000. Cell recognition by foot-and-mouth disease virus that lacks the RGD integrin-binding motif: flexibility in aphthovirus receptor usage. J. Virol. 74: 1641-1647. Baranowsky, E., Ruiz-Jarabo, C.M., Pariente, N., Verdaguer, N. and Domingo, E. 2003. Evolution of cell recognition by viruses: a source of biological novelty with medical implications. Adv. Virus Res. 62: 19-111. Bella, J., Kolatkar, P.R., Marlor, C.W., Greve, J.M. and Rossmann, M.G. 1998. The structure of the two amino-terminal domains of human ICAM-1 suggests how it functions as a rhinovirus receptor and as an LFA-l integrin ligand. Proc. Natl. Acad. Sci. USA. 95: 4140- 4145. Belnap, D.M., McDermott, B.M. Jr, Filman, D.J., Cheng, N., Trus, B.L., Zuccola, H.I., Racaniello, V.R., Hogle, J.M. and Steven, A.C. 2000. Three-dimensional structure of poliovirus receptor bound to poliovirus. Proc. Natl. Acad. Sci USA. 97: 73-78. Bergelson, J.M., Chan, M., Solomon, K.R., St John, N.F., Lin, H. and Finberg, R.W. 1994. Decay-accelerating factor (CD55), a glycosylphosphatidylinositol-anchored complement regulatory protein, is a receptor for several echoviruses. Proc. Natl. Acad. Sci. USA. 91: 6245-6249. Berinstein, A., Roivainen, M., Hovi, T., Mason, P.W. and Baxt, B. 1995. Antibodies to the vitronectin receptor (integrin av~3) inhibit binding and infection of foot-and-mouth disease virus to cultured cells. J.Virol. 69: 2664-2666. Caspar, D.L.D. and Klug, A. 1962. Physical Principles in the Construction of Regular Viruses. Cold Spring Harbor Symposia on Quantitative Biology XXVII. Cold Spring Harbor Laboratory, New York. p. 1-24. Curry, S., Fry, E., Blakemore, W., Abu-Ghazaleh, R., Jackson, T., King, A., Lea, S., Newman, J., Rowlands, D. and Stuart, D. 1996. Perturbations in the surface structure of A22 Iraq foot-and-mouth disease virus accompanying coupled changes in host cell specificity and antigenicity. Structure. 4: 135-145. Curry, S., Fry, E., Blakemore, W., Abu-Ghazaleh, R., Jackson, T., King, A., Lea, S., Newman, J. and Stuart, D. 1997. Dissecting the roles of VPO cleavage and RNA packaging in picornavirus capsidstabilization: the structure of empty capsids of foot-and-mouth disease virus. J. Virol. 71: 9743-9752. Fry, E., Lea, S.M., Jackson, T., Newman, J.W., Ellard, F.M., Blakemore, W.E., Abu-Ghazaleh, R., Samuel, A., King, A.M. and Stuart, D.I.. 1999. The structure and function of a footand-mouth disease virus-oligosaccharide receptor complex. EMBO J. 18: 543-554. He, Y., Bowman, V.D., Mueller, S., Bator, C.M., Bella, J., Peng, X., Baker, T.S., Wimmer, E., Kuhn, R.J. and Rossmann, M.G. 2000. Interaction of the poliovirus receptor with poliovirus. Proc. Nat!. Acad. Sci. USA. 97: 79-84. He, Y., Chipman, P.R., Howitt, J., Bator, C.M., Whitt, M.A., Baker, T.S., Kuhn, R.I., Anderson, C.W., Freimuth, P. and Rossmann, M.G. 2001. Interaction of coxsackievirus B3 with the full length coxsackievirus-adenovirus receptor. Nat. Struct. BioI. 8: 874878. Copyright © 2004 By Horizon Bioscience

Ferrer-Dria and Fita

91\

Hendry, E., Hatanaka, H., Fry, E., Smyth, M., Tate, J., Stanway, G., Santti, J., Maaronen, M., HyypHi, T. and Stuart, D. 1999. The crystal structure of coxackievirus A9: new insights into the uncoating mechanisms of enteroviruses Structure. 7: 1527-1538. Hewat, E.A., Verdaguer, N., Fita, I., Blakemore, W., Brookes, S., King, A., Newman, J., Domingo, E., Mateu, M.G. and Stuart, D. 1997. Structure of the complex of an Fab fragment of a neutralizing antibody with foot-and-mouth disease virus: positioning of a highly mobile antigenic loop. EMBO J. 16: 1492-1500 Hewat, E.A., Neumann, E., Conway, J.F., Moser, R., Ronacher, B., Marlovits, T.C. and Blaas, D. 2000. The cellular receptor to human rhinovirus 2 binds around the 5-fold axis and not in the canyon: a structural view. EMBO J. 19: 6317-6325. Hofer, F., Gruenberger, M., Kowalski, H., Machat, H., Huettinger, M., Kuechler, E. and Blass, D. 1994. Members of the low density lipoprotein receptor family mediate cell entry of a minor-group common cold virus. Proc. Natl. Acad. Sci. USA. 91: 18391842. Hogle, J.M., Chow, M. and Filman, DJ. 1985. Three-dimensional structure of poliovirus at 2.9 A resolution. Science. 229: 1358-1365. Humphries, M.J., Symonds, E.J.H. and Mould, A.P. 2003. Mapping functional residues onto integrin crystal structures. Current Opinion in Structural Biology. 13: 236-243. Jackson, T., Ellard, F.M., Ghazaleh, R.A., Brookes, S.M., Blakemore, W.E., Corteyn, A.H., Stuart, D., Newman, J.W. and King, A.M. 1996. Efficient infection of cells in culture by type 0 foot-and-mouth disease virus requires binding to cell surface heparan sulfate. J. Virol. 70: 5282-5287. Jackson, T., Sheppard, D., Denyer, M., Blakemore, W. and King, A.M. 2000a. The epithelial integrin avp6 is a receptor for foot-and-mouth disease virus. J. Virol. 74 : 4949-4956. Jackson, T., Blakemore, W., Newman, J.W., Knowles, N.J., Mould, A.P., Humphries, MJ. and King, A.M. 2000b. Foot-and-mouth disease virus is a ligand for the high-affinity binding conformation of integrin alpha5betal: influence of the leucine residue within the RGDL motif on selectivity of integrin binding. J. Gen. Virol. 81: 1383-1391. Jackson, T., King, A.M., Stuart, D. and Fry, E. 2003. Structure and receptor binding. Virus Res. 91: 33-46. Kraft, S., Diefenbach, B., Mehta, R., Jonczyk, A., Luckenbach, G.A. and Goodman, S.L.. 1999. Definition of an unexpected ligand recognition motif for alphav beta6 integrin J BioI Chem. 274: 1979-1985. Kolatkar, P.R., Bella, J., Olson, N.H., Bator, C.M., Baker, T.S. and Rossmann, M.G. 1999. Structural studies of two rhinovirus serotypes complexed with fragments of their cellular receptor. EMBO J. 18: 6249-6259. Lea, S., Hernandez, J., Blakemore, W., Brocchi, E., Curry, S., Domingo, E., Fry, E., AbuGhazaleh, R., King, A. and Newman, J. 1994. The structure and antigenicity of a type C foot-and-mouth disease virus. Structure. 2:123-139. Lea, S., Abu-Ghazaleh, R., Blakemore, W., Curry, S., Fry, E., Jackson, T., King, A., Logan, D., Newman, J., and Stuart, D. 1995. Structural comparison of two strains of foot-andmouth disease virus subtype 01 and a laboratory antigenic variant, 067. Structure. 3: 571-580. Logan, D., Abu-Ghazaleh, R., Blakemore, W., Curry, S., Jackson, T., King, A., Lea, S., Lewis, R., Newman, J. and Parry, N. 1993. Structure of a major immunogenic site on foot-and-mouth disease virus. Nature. 362: 566-568. Luo, M., Vriend, G., Kamer, G., Minor, I., Arnold, E., Rossmann, M.G., Boege, U., Scraba, D.G., Duke, G.M. and Palmenberg, A.C. 1987. The atomic structure of Mengo virus at 3.0 A resolution. Science. 235: 182-191. Miller, L.C., Blakemore, W.E., Sheppard, D., Atakilit, A., King, A.M.Q. and Jackson, T. 2001. Role of the cytoplasmatic domain of the b-subunit of integrin avb6 in infection by foot-and-mouth disease virus. 1. Virol. 75: 4158-4164.

Copyright © 2004 By Horizon Bioscience

\92

Structure of FMDV Particles

Olson, N.H., Kolatkar, P.R., Oliveira, M.A., Cheng, R.H., Greve, J.M., McClelland, A., Baker, T.S. and Rossmann, M.G. 1993. Structure of a human rhinovirus complexed with its receptor molecule. Proc. Natl. Acad. Sci. USA. 90: 507-511. Parry, N., Fox, G., Rowlands, D., Brown, F., Fry, E., Acharya, R., Logan, D. and Stuart, D. 1990. Structural and serological evidence for a novel mechanism of antigenic variation in foot-and-mouth disease virus. Nature. 347: 569-572. Rossmann, M.G., Arnold, E., Erickson, J.W., Frankenberger, E.A., Griffith, J.P., Hecht, H.I., Johnson, J.E., Kamer, G., Luo, M. and Mosser, A.G.. 1985. Structure of a human common cold virus and functional relationship to other picornaviruses. Nature. 317: 145-153. Rossmann, M.G.and Johnson, J.E. 1989. Icosahedral RNA virus structure. Annu. Rev. Biochem. 58: 533-573. Rossmann, M.G., Bella, J., Kolatkar, P.R., He, Y., Wimmer, E., Kuh,n R.I. and Baker, T.S. 2000. Cell recognition and entry by rhino- and enteroviruses. Virology. 269: 239-247. Rossmann, M.G., He, Y. and Kuhn, R.I. 2002. Picornavirus-receptor interactions. Trends Microbiol. 10: 324-331. Smith, T.J., Chase, E.S., Schmidt, T.J., Olson, N.H., and Baker, T.S. 1996. Neutralizing antibody to human rhinovirus 14 penetrates the receptor-binding canyon. Nature. 383: 350-354. Verdaguer, N., Mateu, M.G., Andreu, D., Giralt, E., Domingo, E. and Fita, I. 1995. Structure of the major antigenic loop of foot-and-mouth disease virus complexed with a neutralizing antibody: direct involvement of the Arg-Gly-Asp motif in the interaction. EMBO J. 14: 1690-1696. Verdaguer, N., Fita, I., Domingo, E. and Mateu, M.G. 1997. Efficient neutralization of footand-mouth disease virus by monovalent antibody binding. J. Virol. 71: 9813-9816. Verdaguer, N., Schoehn, G., Ochoa, W.F., Fita, I., Brookes, S., King, A., Domingo, E., Mateu, M.G., Stuart, D. and Hewat, E.A. 1999. Flexibility of the major antigenic loop of foot-and-mouth disease virus bound to a Fab fragment of a neutralising antibody: structure and neutralisation.Virology. 255: 260-268. Verdaguer, N., Fita, I., Reithmayer, N., Moser, R., and Blaas, D. 2004. X-ray structure of a minor group human rhinovirus bound to a fragment of its cellular receptor protein. Nat. Struct. BioI. In Press. Winkler, F.K., Schutt, C.E., Harrison, S.C. and Bricogne, G.. 1977. Tomato bushy stunt virus at 5.5 A resolution. Nature. 265: 509-513. Xiao, C., Bator, C.M., Bowman, V.D., Rieder, E., He, Y., Hebert, B., Bella, J., Baker, T.S., Wimmer, E., Kuhn, R.I. and Rossmann, M.G. 2001. Interaction of coxsackievirus A21 with its cellular receptor, ICAM-l. J. Virol. 75: 2444-2451. Xing, L., Tjarnlund, K., Lindqvist, B., Kaplan, G.G., Feigelstock, D., Cheng, R.H. and Casasnovas, J.M. 2000. Distinct cellular receptor interactions in poliovirus and rhinoviruses. EMBOl. 19: 1207-1216. Xiong, J.P., Stehle, T., Diefenbach, B., Zhang, R., Dunker, R., Scott, D.L., Joachimiak, A., Goodman, S.L. and Arnaout, M.A. 2001. Crystal structure of the extracellular segment of integrin avf33. Science. 294: 339-345. Xiong, J.P., Stehle, T., Zhang, R., Joachimiak, A., Frech, M., Goodman, S.L. and Arnaout, M.A. 2002. Crystal structure of the extracellular segment of integrin avf33 in complex with an Arg-Gly-Asp ligand. Science. 296: 151-155.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 5 Clinical Signs of Foot-and-Mouth Disease

Alex Donaldson Abstract Effective surveillance to prevent the entry of animals with foot-and-mouth disease (FMD) into countries or areas free of the disease and to rapidly identify cases during outbreaks requires that those involved with susceptible livestock species are familiar with the clinical signs of the disease. In particular this includes state and private veterinarians, livestock owners and those responsible for animals at markets, abattoirs and agricultural shows etc. In this chapter the characteristic signs of FMD in different livestock species are described and illustrated with accompanying photographs. 1. Introduction All domesticated and wild cloven-hoofed are susceptible to infection with FMD virus, however, the severity of the resulting disease varies with the level of immunity, the infectious dose, the virus strain and the species. The severity of disease may also vary between individual animals within the same species. The mortality rate in adult animals is usually low Le. below 5%, although occasionally rates as high as 30% and 50% have been recorded in cattle (Hyslop, 1970) and gazelle (Shimshony et al., 1986), respectively. By contrast a high mortality rate, i.e. above 90%, is often seen in young livestock species. The incubation period in naturally acquired FMD is generally short, ranging from two to eight days, but sometimes it can be from 1 to 14 days, depending on the infectious dose, the strain of virus and the mechanism of infection (Joubert and Mackowiak, 1968; Report, 1968; Thomson, 1994). Incubation periods will be short when susceptible and infected animals are confined in a building or a transport vehicle as under such circumstances the infectious dose will be high and infection by more than one route will be possible. Incubation periods of 24 and 33 hours, have been seen in pigs in lairages at abattoirs in Hong Kong and the United Kingdom, respectively (Ellis, 2000; Alexandersen and Donaldson, 2002). By contrast, when animals are out-of-doors longer incubation periods can be expected since exposure doses are likely to be lower. However, the periods may shorten with time as the weight of infection and challenge doses increase. In the case of farm-tofarm airborne spread the incubation period can range from 4 to 14 days, depending on the dose (Sellers and Forman, 1973).

Synonyms: fievre aphteuse (Fr.), fiebre aftosa (Sp.), Maul-und Klauenseuche (Ger.)

Copyright © 2004 By Horizon Bioscience

Figure 1. FMDV lesions in cattle. A) Cattle showing excess salivation (DEFRA). B) Foot of a steer showing an early vesicle (blanched area). C) Bovine mammary gland with lesions on the teats. D) Mouth of a steer with unruptured vesicles on the tongue and a ruptured vesicle on the upper gum. See Colour Plate at the back of the book. Copyright © 2004 By Horizon Bioscience

Figure 2. FMDV lesions in cattle and pigs. A) Mouth of a steer with lesions on the tongue and dental pad (DEFRA). B) Mouth of a steer with erosions on the dental pad and gums (DEFRA). C) Lesions on the tongue, dental pad and gum of a steer (DEFRA). D) Pig's heart with areas of necrosis (yellow) in the myocardium (PRIT). See Colour

Plate at the back of the book. Copyright © 2004 By Horizon Bioscience

\96

Clinical Manifestations of FMD

2. Cattle The early signs of FMD in cattle are likely to include the sudden loss of milk production in dairy animals accompanied by anorexia, depression and fever. After a day or so a close inspection will reveal the presence of vesicles in and around the mouth, on the feet and on the mammary glands, especially on the teats (Figure Ie). In the mouth, early lesions most commonly appear as blanched areas in the dorsal epithelium of the tongue. In 8 to 12 hours these become more prominent as fluid collects (Figure ID). The vesicles stand proud of the normal surrounding epithelium and when viewed along its surface will give the tongue a corrugated appearance. Generally, vesicles on the tongue are roughly circular in outline and reach several centimetres in diameter. Characteristically they occur on the dorsal surface of the tongue but other predilection sites are around the tip and on the lateral part of the tongue beside the dorsal prominence. Frequently the vesicles on the dorsum coalesce so that up to 50% of the surface of the tongue may be involved. The overlying epithelium becomes progressively more necrotic and friable and ruptures in one or two days leaving red, raw erosions (Figure 2A). Handling the tongue can accelerate this process resulting in the detachment of large pieces of epithelium. Vesicles followed by erosions may also develop in other parts of the mouth including along the dental pad, hard palate, gums and inside the lips (Figures 2B and 2C). Other sites where vesicles are frequently found are in the nostrils and on the muzzle. Lesions in the nostril are generally small and may coalesce to give a scab-like surface. As vesiculation develops there is evidence of acute pain in the mouth and affected cattle show trembling of the lips and lower jaw. They salivate profusely and refuse to eat. A smacking sound is often heard as the animal makes a single chewing movement. Salivation becomes more marked, and after rupture of the vesicles is one of the most obvious signs of the disease. Frequently there is drooling of frothy, ropey saliva (Figure lA). The reduced appetite results in a loss of body condition. Lesions on the feet cause acute lameness, a tucked-up stance, a reluctance to move and intermittent leg flicking. In the early stages when affected feet are palpated they will be found to be warm and painful. Vesicles usually develop first along the coronary bands near the interdigital cleft (Figure 1B) or at the bulbs of the heel. Vesicles may extend into and through the length of the interdigital space. The epithelium is white and necrotic. Generally, vesicles on the feet take a day or so longer to rupture than those in the mouth. In addition, in the acute stage of the

disease affected animals will generally have nasal and ocular discharges. Nasal discharges are usually serous at first and later become muco-purulent. The resolution of lesions, especially in the mouth, generally occurs rapidly. Soon after a vesicle ruptures serofibrinous in-filling begins. This starts around the periphery of the lesion and gradually extends across it so that by around day four the base is covered by fibrous tissue. After about seven days regeneration of the epithelium will be almost complete, except that the new epithelium lacks the typical tongue papillae. The areas of scar tissue progressively reduce with time. Foot lesions go through a similar healing process as in the mouth and resolution in the absence of severe secondary infection is quite rapid. After about 5 days the epithelium of interdigital lesions will usually have been replaced by granulation

Copyright © 2004 By Horizon Bioscience

Donaldson

971

tissue. However, damage to the coronary band may interrupt the growth of hom and as new hom grows the old one may separate off. Secondary infections may complicate the healing processes. This is more likely to occur with foot and teat lesions than with those in the mouth and can lead to chronic lameness and chronic mastitis, respectively. In tropical areas some cattle recovered from acute FMD suffer from a wasting syndrome in which they have a staring coat and dyspnoea. They have been called "hairy panters". The underlying pathology has not been determined but it has been hypothesised that it may be linked to hyperactive thyroid-adrenal function (Minett, 1949; Mullick, 1949). Mortality in adult cattle due to FMD very seldom exceeds 5% but in very young calves it can be over 90%. Sudden death in young calves results from a multifocal necrosis of the myocardium. Usually death intervenes before there are any signs of vesicular lesions. 3. Pigs The early signs of FMD are characterised by lameness, fever, depression and anorexia. The feet will be very painful and animals will be recumbent and reluctance to move. If forced to stand, affected animals are likely to adopt a tuckedup appearance or dog-sitting position. Affected feet will be warm and painful to the touch and in pale-skinned pigs initial lesions will be seen as blanched areas along the coronary bands of the digits and supernumerary digits (Figure 3B). These will quickly fill with fluid to become vesicles (Figure 3C). After a day or so they will rupture to leave red, raw, painful erosions. Vesicles may also form at the bulbs of the heels and in the interdigital space. Lame pigs will often adopt a hobbling gait and this can lead to the development of further lesions on pressure points such as the knees and hocks. Severe lesions on the feet may lead to the separation of the keratinised layers of the hoof from the corium and the sloughing of the hooves ("thimbling") (Figure 3D). This can occur in pigs of all ages but is especially likely in piglets. The severity of lesions can vary with the environmental conditions. When pigs are on hard floors the foot lesions are likely to be more severe than when they are on soft ground or bedding. Foot lesions in sows that are tethered or confined in farrowing crates may be missed. Lesions on the tongue are not as prominent as in cattle. They usually occur as one or more blanched areas on the dorsum. The necrotic tissue is quickly lost leaving an erosion. Small vesicles or circumscribed areas of necrosis are commonly seen along the gums and lips. An additional site is above the rostrum of the snout where vesicles may contain several millilitres of fluid before they rupture (Figure 3A). Vesicles may also occur on the teats and mammary glands. Pregnant sows may abort. Sucking piglets may die without developing vesicles and their deaths can occur both before and during the time when sows are beginning to develop vesicles. In some outbreaks the sudden death of piglets has been the first indication that a herd is infected (Donaldson et al., 1984). At post mortem the hearts from fatal cases may show macroscopic changes consisting of yellow coloured streaks in the myocardium; so-called "tiger hearts" (Figure 2D). The healing of vesicular lesions in pigs proceeds in the same way as has been described previously for cattle.

Copyright © 2004 By Horizon Bioscience

Figure 3. FMDV lesions in pigs. A) Pig with a large, unruptured vesicle above the rostrum (PRIT). B) Pig's feet with early, unruptured vesicles (blanched areas). C) Pig's feet with vesicles along the coronary bands of the main and supernumerary digits. D) Pig's foot with loss of horn tissue (PRIT).See Colour Plate at the back of the book. Copyright © 2004 By Horizon Bioscience

Figure 4. FMDV lesions in sheep. A) Sheep's foot with a ruptured vesicle in the interdigital area. B) Sheep's foot with a ruptured vesicle along the coronary band. C) Erosions on the tongue and dental pad of a sheep. D) Mouth of a sheep showing erosions on the dental pad and upper lip (DEFRA). See Colour Plate at the back of the

book. Copyright © 2004 By Horizon Bioscience

\100

Clinical Manifestations of FMD

4. Sheep and Goats The severity of FMD in sheep and goats varies considerably with the strain of virus, the breed of animal and the environmental conditions. Some strains cause severe lesions whereas in others the clinical signs may be mild and a careful individual examination of a high proportion of the animals in the flock or herd may be required to detect the disease. It has been reported that goats indigenous to East and South Africa generally suffer inapparent infection (Thomson, 1994). The first signs in an infected flock of sheep or herd of goats is often a rapidly increasing incidence of severe lameness accompanied by depression, anorexia and pyrexia or the sudden death of young stock if lambs or kids are present. The mortality rate among lambs and kids may be high. The cause of death, as in other cases of acute fatal infection in young stock, is heart failure due to multifocal necrosis of the myocardium. In the early stages of disease milking animals, especially goats, will show a sudden drop in production. Vesicles may be present on the teats and vulva. Rams may develop vesicles on the prepuce and be unable or unwilling to serve. Closer examination of lame animals is likely to show that their feet are hot and painful when handled. Vesicles may be found in the interdigital space, at the bulbs of the heel and along the coronary band (Figures 4A and 4B). Cleaning of the feet and careful reflection of the hair above the hoof may be necessary to see lesions along the coronary band. Vesicles on the outer coronary band are more common than in cattle. Coronary band lesions usually rupture quickly leaving shallow erosions. Early lesions in the mouth of sheep or goats are typically seen as small blanched areas of necrotic epithelium; most often on the dental pad. The superficial necrotic layer is quickly lost resulting in the formation of erosions (Figure 4D). Fluidfilled vesicles are unusual and, if they occur, are very transient as the superficial epithelium is thin and readily ruptured. Erosions may also be seen on the gums, inside the lips and occasionally on the tongue. Tongue erosions generally occur as multiple small (0.5 to 1.0 cm) areas on the dorsum (Figure 4C). Lesions in goats are usually fewer in number and less severe than in sheep. In cases uncomplicated by secondary infection the healing of lesions occurs rapidly, especially in the mouth. On the feet resolution proceeds and there is scabbing and granulation both on the coronary band and the interdigital space. At this stage it is very difficult to be certain that the lesions are those of FMD. However, if there is secondary infection lameness may continue and be severe causing affected animals to hobble on their knees or remain recumbent. In milking animals reduced production and mastitis may be sequelae. 5. Other Species FMD can also occur under natural conditions in water buffalo, deer, antelopes, hedgehogs, wild pigs and several species of African ruminants, in particular the African buffalo and impala (EI-Danaf et al." 1990; Gibbs et aI., 1974; McLaughlan and Henderson, 1947; Meeser, 1962; Shimshony et al., 1986; Hedger, 1976; Thomson, 1994; Thomson et ai., 1994).

Copyright © 2004 By Horizon Bioscience

Donaldson

1011

6. Ageing of Lesions An essential part of the history and a pre-requisite for the epidemiological investigation of an outbreak of FMD is to determine the age of lesions in affected animals, in particular to identify the oldest lesions. Readers interested in this topic can obtain details in the manual "Foot and Mouth Disease Ageing of Lesions" (MAFF, 1986) and at: www.defra.gov.uk/animalh/diseases/fmd/about/signs. 7. Acknowledgements Mr L. M. Mansley, Department for Environment, Food and Rural Affairs (DEFRA), Animal Health Divisional Office, Carlisle, UK and Drs P. C. Yang, H.-S. Hsu, J.-P. Hsu and W. Chung, Pig Research Institute Taiwan (PRIT), Chunan Miaoli 35046, Taiwan are thanked for kindly providing photographs. References Donaldson, A.I., Ferris, N.P., and Wells, G.A.H. 1984. Experimental foot-and-mouth disease in fattening pigs, sows and piglets in relation to outbreaks in the field. Vet. Rec. 115: 509-512. EI-Danaf, N.A., Nagi, A.A., Hafez, M.A.M., Shaker, M.H.M., and Tawfik, A.M. 1990. Clinical and pathological studies on foot and mouth disease in buffaloes. Assiut Vet. Med. J. 22: 44, 80-87. Ellis, T. 2000. Report of the sixth meeting of the OlE Sub-Commission for Foot-and-Mouth Disease in South-East Asia. Hanoi, Vietnam, 21-25 February 2000. OlE, Paris P 9. Garland, A.J.M., and Donaldson, A.I. 1990. Foot and mouth disease. Surveillance 17: 6-8. Gibbs, E.P.J., Herniman, K.A.J., Lawman, M.1.P., and Sellers, R.F. 1974. Foot-and-mouth disease in British deer: transmission of virus to cattle, sheep and deer. Vet. Rec. 96: 558-563. Hedger, R.S. 1976. Foot-and-mouth disease in wildlife with particular reference to the African buffalo (Syncerus caffer). In: Wildlife Diseases. L. A. Page, ed. Plenum Publishing Corporation, New York. p. 235-244. Hyslop, N.St.G. 1970. The epizootiology and epidemiology of foot and mouth disease. J. compo Path. 75: 119-126. Joubert, L., and Mackowiak, C. 1968. La fievre aphteuse. Volume 2. La fievre aphteuse spontanee. Fondation Merieux expansion scientifique, Lyon. p. 183-210. MAFF. 1986. Foot and Mouth Disease Ageing of Lesions. Reference Book 400. HMSO, London. McLaughlan, J.D., and Henderson, W.M. 1947. The occurrence offoot-and-mouth disease in the hedgehog under natural conditions. 1. Hyg. 45: 474-479. Meeser, M.J.N. 1962. Foot-and-mouth disease in game animals, with special reference to the impala (Aepyceros melampus). J. S. Afr. Vet. Med. Assoc. 33: 351-354. Minett F.C. 1949. Panting in cattle - a sequel to foot-and-mouth disease. I. Experimental observations and pathology. Am. J. vet. Res. 10: 40-48. Mullick, D.N. 1949. Panting in cattle - a sequel to foot-and-mouth disease. II. Biochemical observations. Amer. 1. Vet. Res. 10: 49-55. Report 1968. Report of the Committee of Inquiry on Foot-and-Mouth Disease 1968. Part One, London HMSO. Sellers, R.F., and Forman, A.J. 1973. The Hampshire epidemic of foot-and-mouth disease, 1967. J. Hyg., Camb. 71: 15-34. Shimshony, A., Orgad, D.. Baharav, D., Prudovsky, S., Yakobson, B., Bar Moshe, B., and Dagan, D. 1986. Malignant foot-and-mouth disease in mountain gazelles. Vet. Rec. 119: 175-176.

Copyright © 2004 By Horizon Bioscience

\102

Clinical Manifestations of FMD

Thomson, G.R. 1994. Foot-and-mouth disease. In: Infectious Diseases of Livestock with Special Reference to Southern Africa. Coetzer, J.A.W., Thomson, G.R., Tustin, R.C. , and Kriek, N.P.1., eds. Oxford University Press, Cape Town. p. 825-852. Thomson, G.R., Bengis, R.G., Esterhuysen, J.1., and Pini, A. 1984. Maintenance mechanisms for foot-and-mouth disease virus in the Kruger National Park and potential avenues for its escape into domestic animal populations. Proceedings of the XlIIth World Congress on Diseases of Cattle. Vol. I. September 1984. Durban, South Africa.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 6 Persistence of Foot-and-Mouth Disease Virus

Jeremy Salt

Abstract The clinical signs of foot-and-mouth disease (FMD) in cattle and pigs are those of an acute febrile vesicular illness. Persistent infection with FMDV (the carrier state) is considered to be a common sequel to both clinical and subclinical FMD in ruminants only, and is asymptomatic. Various factors influence the development and duration of persistent FMDV infection, including the genetics of the host and the viral challenge itself. The fear of FMDV carriers has led to restricted international trade in seropositive live ruminants. The mechanism employed by the virus to persist and evade immune elimination from the host is unknown. Despite the chronic stimulation of local IgA, the virus appears to persist in the epithelium of the soft palate and oropharynx. Evidence from in vitro work on persistently infected cell culture suggests a co-evolution of virus and cells, similar to some other picornavirus persistent infections. Thus despite an apparent antigenic stability of FMDV during persistence in vivo, it is possible that a tissue-specific variant is selected which can survive in a specific location without elimination. Whether this persisting virus represents a real source of infectious FMDV with the potential to cause patent disease in susceptible contact animals remains a mystery. 1. Introduction In 200 1 Europe was given a harsh reminder of the devastating effects that a major outbreak of foot-and-mouth disease (FMD) can have in a developed country. Our television screens were again filled with nightly images of mass burial pits and funeral pyres, and human anguish. Fortunately these images are no longer commonplace in Europe. However, FMD is still endemic in many regions of the world with whom the FMD-free countries are both neighbours and trading partners. To avoid the return of FMD to the FMD-free countries great care is taken at a national level to exclude infected animals and their products. In this context the inapparently infected foot-and-mouth disease virus (FMDV) carrier animal poses a particular threat to the susceptible livestock of all FMD-free regions. An inability to definitively diagnose FMDV carriers and the fear of their infectiousness has led to the prevention of importation of seropositive animals into non-vaccinating FMDfree countries. The nature of the FMDV carrier state, and its possible causes will be addressed in this chapter.

Copyright © 2004 By Horizon Bioscience

\104

FMDV Persistent Infection

2. The Carrier State in FMD 2.1. Historical Perspective

The primary mode of transmission of FMDV during an outbreak is direct contact with a diseased animal. Secondary modes of transmission include contact with infected animal products and fomites, wind-borne spread, and mechanical transmission by vehicles and people. However, from the beginning of this century observers predicted the occurrence and epidemiological significance of the persistent excretion of FMDV by convalescent animals. Several reports of field outbreaks of FMD in the late 19th and early part of the 20th century supported the belief that convalescent cattle may be responsible for the infection of susceptible in-contact animals (Bang, 1912). This appeared to be the case following the movement of FMD convalescent cattle onto FMD-free farms following the 192224 epidemic in the UK (Anon, 1925), and was thought to explain the introduction and perpetuation of FMD in Mexico following the importation of cattle from Brazil in the 1940's (Bureau of Animal Industry, USDA, ARS, 1947-50; van Bekkum, 1973). Retrospectively, it has been suggested that carrier animals imported from Britain may have been responsible for the last outbreaks of FMD in Australia at the end of the 19th century (Pullar, 1965). However, there is also a body of evidence that refutes the importance of the carrier animal in the epidemiology of FMD. Field observations in countries with endemic (Hedger, 1970) and epizootic FMD (Mohler, 1926) have argued against the perpetuation of FMD outbreaks by convalescent cattle. In one of the best documented field transmission trials several hundred convalescent cattle infected with FMDV at the National Dairy Show in Chicago in 1914 were mixed with susceptible cattle and subsequently pigs in a vain attempt to transmit FMDV (Mohler, 1924). Despite many reported attempts to transmit FMDV from persistently infected cattle under controlled conditions since then there have been no successes to date. Nevertheless, in 1928 Olitzky et ale isolated FMDV from the hoof of an infected steer 34 days after FMDV infection and an unsubstantiated report by Waldemann et al. in 1931 maintained that virus was present for many months in the concentrated urine of convalescent cattle. In 1959 van Bekkum et al. reported the recovery of tissue culture infectious FMDV in the oropharyngeal scrapings collected with a probang sampling cup from FMD convalescent and sub-clinically infected vaccinated cattle. This discovery irrefutably confirmed the existence of the FMDV carrier state and demonstrated that FMDV could establish a truly persistent inapparent infection in the upper respiratory tract of cattle. Subsequently, the natural occurrence of FMDV carriers was established by Sutmoller and Gaggero (1965) who described the recovery of FMDV in oropharyngeal fluid samples over prolonged periods following a field outbreak of FMD in cattle in Brazil. Since these first confirmatory reports the development of the carrier state as a common sequel to FMDV infection has become well-documented (Burrows, 1966; Hedger, 1968) and remains an area of continued investigative research and conjecture. The realisation that clinically normal convalescent or sub-clinically infected vaccinated cattle could be inapparently infected with FMDV has had a major impact on international trade in livestock. In FMD-free states strict importation control, quarantine and testing are employed to exclude potential FMDV carrier animals.

Copyright © 2004 By Horizon Bioscience

Salt 1051

2.2. Site of FMDV Persistence The accepted definition of a carrier of FMDV has become' any animal from which FMDV can be recovered in oropharyngeal scrapings using a probang sampling cup for periods of greater than 28 days post-challenge' (Sutmoller et a!., 1968). The oropharynx, and more particularly the dorsal soft palate, was suggested by Burrows (1966) and van Bekkum et ale (1966) to be the predilection site for FMDV persistence in cattle, although infectious virus has been recovered for shorter periods from a number of organs and tissues during convalescence (Cottral, 1969; Burrows et al., 1971) and from cultured tissue explants (Mohanty and Cottral, 1971). More recent analytical techniques, such as the polymerase chain reaction (peR), have confirmed these findings (Meyer et al., 1991; Hofner et al., 1993; Donn et al., 1994). In carrier sheep however, the tonsillar region was found to be the site of highest virus titre and most frequent virus recovery (Burrows, 1968). Although Hyslop (1965) recovered FMDV from atraumatically collected oral saliva samples from cattle during the prodromal and acute phases of infection, virus is not contained in saliva collected from parasympathomimetically-stimulated carrier cattle (Wittman and Eissner, 1966). This supports the widely held belief that the collection of cells and mucus in probang fluid is critical for the recovery of virus from FMDV carriers. The identity of the cells in the oropharynx within which FMDV persists is unknown. However, by sensitive in situ hybridisation techniques, FMDV RNA has been shown to persist beyond the acute stage of infection in the basal layers of the epithelium of the dorsal soft palate and pharynx (Zhang and Kitching, 2001). 2.3. Carrier Prevalence Under Field Conditions Since the development of the probang test for the detection of persistent oropharyngeal infection with FMDV many studies of the carrier state have been conducted. In FMD endemic areas carrier animals are a frequent finding. A recent survey in Anatolia (Asiatic Turkey) found 18.4% of cattle and 16.8% of sheep to be FMDV carriers (Gurhan et a!., 1994). A survey of free-living buffalo in Botswana found that between 55 and 70% of animals were virus positive at different sampling times (Hedger, 1976). As a result of the intermittent nature of FMDV recovery from persistently infected animals not all samples from carrier animals were virus positive at every sampling in this study. In areas where FMD occurs less frequently the decline in herd prevalence of FMDV carriers can be studied following an outbreak. The prevalence of persistently infected cattle after a vaccine trial breakdown in Brazil was found to be 50% after four months, and 56% after six months by Sutmoller and Gaggero (1965). In a field survey of cattle in Botswana, Hedger (1970) found an initial infection rate with a SAT 1 type FMD virus of 68% during an outbreak, which fell to 38% six months later and 5.4% by 12 months after the outbreak. In a previous follow-up survey to an FMD outbreak caused by a SAT 3 type FMD virus in cattle in Botswana, the same author found that 20% of cattle were carriers after seven months and 12% after 12 months (Hedger, 1968). However, Anderson et ale (1976) found very few carrier sheep or goats in a field survey of an endemic area of Kenya, which led them to conclude that these animals did not playa significant role in the spread of FMDV in that country. Typical data for virus recovery following an outbreak of FMD in an isolated cattle herd is presented in Table 1. The intermittent recovery of type A FMDV was Copyright © 2004 By Horizon Bioscience

Table 1. Type A FMDV isolations from probang samples collected from FMD convalescent cattle following an outbreak of FMD on a dairy unit in Saudi Arabia. All cattle were 12 month-old heifers at the time of the outbreak and all showed clinical signs of disease.

Animal Number

Months8

#1

#2

2

_b

+c

+

+

+

3

#3

#4

4

#5

#6

#7

+ +

+

+

+

6

+

+

+

+

+

8 a Months

+

#9

+

5

7

#8

+ +

#11

#12

#13

#14

+

+

+

+

+

+

+

+

+

40

+

+

+

+

40

+

+

+

47

+

+

+

47

+

+

+

40

+

20

+

after an outbreak of FMD caused by a type A virus in this dairy herd. Negative c.p.e. on BTY cells after 72 hours. C Positive c.p.e. on BTY cells, positive for type A FMDV. d Percentage of cattle positive for type A FMDV by virus isolation from probang samples collected monthly. b

Copyright © 2004 By Horizon Bioscience

#15

% +ved

#10

47

Salt

1071

found in a high proportion of the sampled cattle over a period of eight months. At any particular sample date approximately half of the cattle were virus positive and 12 of the 15 studied animals were positive at least once between two and eight months after the resolution of the outbreak (see also Chapter 14).

3. Factors That Affect the Development of FMDV Persistence 3.1. Prophylactic Vaccination In both vaccinated and non-vaccinated cattle the establishment of the carrier state

occurs frequently following sub-clinical and clinical disease, respectively (Burrows, 1966; Sutmoller et aI., 1968; Hedger, 1970). In individual animals protection from clinical FMD by vaccination does not affect the development of persistent infection. Anderson et ale (1974) reported that in an FMD endemic area in Kenya the prevalence of FMDV carriers amongst regularly vaccinated cattle was lower (0.49%) than in non-vaccinated cattle (3.34%). However, the observed reduction in prevalence of carriers in the regularly vaccinated herd was probably an indirect result of an increase in overall herd immunity reducing the number of new clinical cases in susceptible animals which could act as a source of infection for potential carriers. In studies conducted at the Institute for Animal Health (IAH), Pirbright Laboratory, 77 and 78% of vaccinated and non-vaccinated cattle respectively, developed persistent FMDV infections following a high titre challenge (Salt et al., 1996). These results are consistent with those of other workers (Auge de Mello et al., 1970). Therefore, despite evidence that vaccination of cattle can reduce post-infection virus excretion titres (McVicar and Sutmoller, 1976; Sellers et aI., 1977; Doel et al., 1993) and transmission to contact susceptible cattle (Donaldson and Kitching, 1989), there is no unequivocal experimental evidence that conventional FMD vaccination reduces the establishment or the duration of persistent infection with FMDV in individual cattle (Hedger, 1970). 3.2. Challenge Strain

Burrows (1966) studied the persistence of several different strains of FMDV in cattle under similar experimental conditions. Under these particular conditions SAT 1 and SAT 3 type FMD viruses persisted for shorter periods than two type A strains. However, few studies have made a direct strain-related comparison of the duration of persistence. Table 2 shows that in the literature there are no discemable differences between serotypes in terms of their ability to persist in cattle, a conclusion that was supported by a natural transmission study with type 0, A and C FMD viruses (McVicar and Sutmoller, 1969a). This does not of course rule out individual strain differences within serotypes. 3.3. Challenge Dose

Hedger (1970) suggested that a "severe natural field challenge" of a vaccinated herd of cattle in Botswana resulted in an increased number of FMDV carriers. Whether this finding was due to an increased chance of contact between clinically diseased and protected cattle in a herd with a high morbidity rate, or was a direct result of a higher cumulative infectious dose is impossible to say. Many other factors may have been involved and the suggestion that the establishment of persistence post-infection is challenge dose-related is not supported by

Copyright © 2004 By Horizon Bioscience

Table 2. The duration of persistent infection with FMDV in various species and breeds of cattle following challenge by various routes

a

Species

FMDV Strain

Vaccinal Status

Route of Infection

Duration (Months)

Cattle-Zebu a

A 2JC 3

Non-vaccinated

Intramuscular

8/9

Auge de Mello et al. (1970)

Cattle-European

C

Non-vaccinated

Natural

24

Straver et al. (1970)

Cattle-Baran

C

Non-vaccinated

Natural

6.5

Anderson et al. (1974)

Cattle-Zebu

C3

Non-vaccinated

Intranasal

18

Gebauer et al. (1988)

Cattle-Tswana

SAT 1

Vaccinated

Natural

>12

Hedger (1970)

Cattle-Boran

SAT 2

Vaccinated

Natural

>36

Hargreaves et al. (1994)

Cattle-Tswana

SAT 3

Vaccinated

Natural

>12

Heger (1968)

Sheep

A-4691

Non-vaccinated

Intranasal

12

Sutmoller (1970)

Goats

0

Non-vaccinated

Natural

4

Singh (1977)

Pigs

A

Vaccinated / nonvaccinated

Natural

60

Condy et al. (1985)

Breed

Copyright © 2004 By Horizon Bioscience

Reference

Salt 1091

unequivocal experimental data (Sutmoller et al., 1968). An inverse relationship has however been demonstrated between the challenge dose of FMDV and both the incubation period for detectable virus replication in the oropharynx and the development of clinical signs of disease (McVicar and Sutmoller, 1976). Low dose challenge is more likely to result in sub-clinical FMDV infection (Sutmoller et al., 1968). In a study conducted at IAH, 12 out of 16 susceptible cattle challenged with a low dose of type A 24 FMDV via the intranasal route became persistently infected, only 5/16 showed clinical signs ofFMD and 3/16 failed to be infected on the basis that FMDV was never isolated in probang samples collected from them post-infection (IIott et al., 1997). Therefore, a threshold titre of virus inoculum may be required for persistent infection to be established following sub-clinical infection in the same way that minimum infectious doses have been calculated for the development of clinical disease in different species (Donaldson et al., 1987). 3.4. Challenge Route

Very little information is available upon the effect that the route of challenge has upon the development of FMDV persistence. Once clinical disease has developed following infection by any route, persistent infection of the oropharynx is a common sequel. In goats and sheep intranasal installation of FMDV of several strains has been shown to be as efficient a method of the establishment of persistent infection as direct contact with diseased animals (McVicar and Sutmoller, 1969). Interestingly, Burrows et at. (1971) showed that the intra-mammary instillation of FMDV in cattle immune to FMD led to local virus replication only, and Sutmoller et al. (1968) showed that intramuscular inoculation did not initiate persistent oropharyngeal infection. Therefore, either a clinical disease episode following direct or indirect contact infection, or the inoculation of FMDV into the oropharyngeal cavity via the intranasal, intradermolingual or oropharyngeal route is necessary for the establishment of persistence.

4. Duration of Persistence Certain factors such as host species and breed have been shown to influence the duration of the 'carrier state', whereas the age and sex of the host is immaterial (Hedger, 1968). Different breeds of African cattle have been shown to excrete FMDV for up to two and a half (Hedger, 1968) and over three years (Hargreaves et al., 1994), sheep for up to one year (Sutmoller, 1970), goats for up to four months (Singh, 1977) and African buffalo for at least five years (Condy et al., 1985). The effect of FMDV serotype, and possibly even subtype, cannot be dismissed as the prolonged carrier periods recorded in cattle and buffalo in Africa with SAT viruses have not been repeated under controlled conditions with cattle infected with other FMDV serotypes. Pigs appear to eliminate detectable FMDV from the oropharynx within 28 days of infection (van Bekkum, 1973; Panina et al., 1988) and have, therefore, been considered as non-carriers of FMDV (see Table 2). Interestingly, Mezencio et al. (1999) have reported the prolonged detection ofFMDV RNA by RT-PCR in the serum of convalescent pigs in the presence of persistent neutralising antibody responses for over 200 days after infection. This represents a unique finding, and some heresy in the face of accepted dogma.

Copyright © 2004 By Horizon Bioscience

\110

FMDV Persistent Infection

In diseased animals a dramatic reduction in the titre of exhaled virus and the resolution of clinical signs of FMD occur coincidentally with the appearance of circulating neutralising antibody. However, after the initial decline in titre following the resolution of lesions FMDV can be detected at relatively constant titre in probang fluid samples for up to one month in cattle, and thereafter intermittently at declining titre for prolonged periods (Burrows, 1966). Intriguingly, in addition to the differences in duration of virus excretion between cattle and pigs there are marked differences in the quantity of virus excreted in the breath of pigs and ruminants in the prodromal and acute stages of FMD. Over a similar period one pig has been found to excrete an equivalent amount of infectious FMDV to 3,000 cattle (Donaldson et ai., 1970). However, in pigs the decline in aerosol excretion of FMDV following clinical recovery is dramatic and results in a discrete period of virus excretion which lasts only four to five days (Sellers et ai., 1977). Whether this represents a fundamental difference between the immune response of ruminants and monogastrics to FMDV infection, or is a function of different cellular sites of virus replication, can only be speculated upon in the absence of solid data in either area. Francis and Black (1983) showed that the neutralising antibody response in nasal secretions of pigs following FMDV infection was more rapid and effective than in cattle (Francis et ai., 1983). This observation led the authors to conclude that mucosal humoral immunity may playa more important role in the response of pigs to FMDV. Further investigation is necessary before it can be concluded that this difference between the species is the sole determinant of persistence versus elimination of FMDV. Researchers investigating the sites of FMDV replication and the early innate and cellular immune responses to FMDV infection would be well advised to exploit this apparent inter-species difference. Lessons learnt from the study of the immune response to FMDV in pigs may be invaluable for the design of future vaccines, which could induce protection from persistent infection of ruminants at mucosal sites, or favour virus elimination, rather than simply offering protection from clinical disease.

5. Evidence of Transmission of FMDV from Carrier Animals Carriers of FMDV are identified by the recovery of FMDV from oropharyngeal scrapings using a probang sampling cup. The titre of infectious virus in probang samples is generally low and after an initially stable period in the first 3-4 weeks postinfection (Burrows, 1966; Sutmoller et ai., 1968) gradually declines during the period of FMDV persistence (van Bekkum et aI., 1966), to fall below the level thought to be necessary for successful transmission to susceptible animals (Donaldson and Kitching, 1989; Terpstra and van Maanen, 1989). The decline in titre and frequency of infectious virus recovery from the oropharynx may be accompanied by the increased production of non-infectious virus particles. Using the polymerase chain reaction (peR) and dotblot hybridisation, FMDV RNA fragments have been detected in probang samples after infectious virus could no longer be isolated (Rossi et aI., 1988). How these results relate to infectivity is not currently understood. Sub-clinically infected cattle vaccinated less than seven days prior to challenge can transmit FMDV at sufficient titre to induce clinical FMD following direct contact with susceptible animals (Donaldson and Kitching, 1989). Presumably, in animals vaccinated at more than seven days pre-challenge, serum antibody limits Copyright © 2004 By Horizon Bioscience

Salt

1111

viral replication and excretion to titres below the minimum required for successful transmission (McVicar and Sutmoller, 1976). Auge de Mello et ai. (1970) showed that an attenuated live lapinised FMDV vaccine strain could be transmitted from vaccinated to susceptible cattle. However, the evidence for transmission to susceptible animals from true FMDV carriers is circumstantial. In FMD endemic areas it is often impossible to confirm suspicions that carrier animals are the cause of further outbreaks due to the problems presented by poor movement records, the number of individual field outbreaks and the lack of precise identification of FMDV outbreak strains. Field evidence suggests that transmission occurs from carrier buffalo in Africa both to buffalo calves (Condy et ai., 1985; Bengis et al., 1986) and to cattle under experimental (Hedger, 1970; Hedger and Condy, 1985; Dawe et ai., 1994a; Bastos et al., 1999) and field conditions (Dawe et al., 1994b). However, despite reports of the seroconversion (van Bekkum et ai., 1959; Sutmoller et ai., 1967) and isolation of pharyngeal FMDV (Hedger, 1968) from contact pigs and calves with carrier dams, transmission of FMD from carrier cattle to susceptible animals has not been shown under controlled conditions. Similar results have been reported for sheep in which seroconversion has occurred in an occasional susceptible animal following prolonged contact with carrier sheep (Bauer et al., 1977). Several observations have been put for\vard to explain why the demonstration of transmission from FMDV carrier cattle has been so elusive: the low virus titre in the oropharynx, a cellular association of persistent virus, the presence of neutralising antibody in secretory fluids, the dilution and swallowing of secreted virus in saliva and altered virulence/infectivity of carrier virus. However, the overall conclusion from these studies is that certain trigger factors, as yet unidentified, are involved in the successful transmission of FMDV from carrier to susceptible animals. Such factors could be responsible for occasional peaks in virus excretion from carrier animals, or alternatively cause a temporary increase in susceptibility in contacts. There have been several attempts to influence the transmission of FMDV from carriers by direct immunosuppression with corticosteroids (McVicar et al., 1976; Ilott et ai., 1997) and intercurrent infection with infectious bovine rhinotracheitis virus (IBR) (McVicar et al., 1976) and rinderpest virus (RPV) (Salt, 1993a). However, until now efforts to demonstrate FMDV transmission from carrier cattle under controlled conditions have failed. The difficulties associated with tracing cattle movements and obtaining accurate disease reports from FMDV endemic areas have hampered the epidemiological investigation of the role of carriers in field outbreaks. However, application of rapid nucleotide sequencing have allowed accurate identification and comparison of FMDV field isolates. The application of these molecular epidemiological techniques has been fruitfully applied in a collaborative project between the O.I.E/F.A.O. World Reference Laboratory for FMD and the Veterinary Services of Zimbabwe, sponsored by the Overseas Development Administration (ODA). These studies have demonstrated the long-term stability of carrier virus isolates, and have provided the most convincing evidence so far recorded for the role of FMDV carrier cattle in the initiation of field outbreaks of FMD (N.I.Knowles, unpublished data).

Copyright © 2004 By Horizon Bioscience

1112

FMDV Persistent Infection

The molecular epidemiological investigation of FMDV field isolates from Zimbabwe has demonstrated two basic patterns of FMD epidemiology. Geographically distinct subgroups of SAT 1 and 3 viruses appear to be maintained in African buffalo populations, with mostly sporadic short-lived outbreaks occurring in local cattle, whereas SAT 2 viruses appear to be maintained within the cattle population and geographically have spread more extensively. It is the study of genome sequence data of recent SAT 2 FMDV isolates that has produced the best evidence for the involvement of carrier cattle in further infection of susceptible cattle. In 1989 there \vas an isolated outbreak ofFMD in cattle due to a type SAT 2 strain of FMDV in the FMD-free zone in northern Zimbabwe. An isolate from this outbreak was sequenced and found to be closely related to the virus responsible for an outbreak of FMD south of Harare in 1987. Movement records showed that cattle had been moved from a farm involved in the 1987 outbreak via a cattle market to the farm affected in 1989. The epidemiological evidence suggests that the cattle became FMDV carriers before their movement for sale and were responsible for the outbreak of FMD in 1989 in fully susceptible cattle. Stronger circumstantial evidence for the role of FMDV carrier cattle in the epidemiology of FMD was gained from the analysis of field isolates collected from another isolated outbreak in the northern FMD-free zone of Zimbabwe in 1991. The sequence data from the SAT 2 FMD viruses isolated from this outbreak were very similar to that of the isolates collected from another previous outbreak of FMD in 1989 which occurred in the south of the country. Examination of movement records showed that cattle had moved onto the farm immediately before the FMD outbreak in 1991, via cattle sales. These animals had originated from a holding not involved in the 1989 outbreak but close enough for the cattle to have been vaccinated during the containment campaign. Zimbabwean veterinary personnel revisited the farm of origin of the moved cattle and discovered the existence of cattle persistently infected with a strain of FMDV SAT 2. The sequences of these isolates were very closely related to both the 1989 outbreak isolates and the 1991 isolates. This evidence is consistent with the theory that vaccinated cattle became persistently infected with FMDV, and subsequently transmitted the largely unchanged virus to susceptible cattle 18 months later following transportation, thereby establishing a further outbreak of FMD. In both of these histories from Zimbabwe two stressor factors were involved, transport and mixing of livestock. These factors \vere identified by Dawe et ale (1994a) as important aspects of their successful transmission experiment between carrier African buffalo and susceptible cattle.

It should be noted that in the second incident, vaccinated carrier cattle appeared to be responsible for establishing an outbreak of FMD in a non-vaccinating area two years after infection. There is currently great concern in Europe about the potential movement of cattle vaccinated following an emergency ring-vaccination campaign around a future extensive outbreak ofFMD, despite plans to identify and hold all vaccinated cattle for 12 months. At present there is no reliable method for differentiating the animals that developed a sub-clinical persistent infection from the non-infected cattle amongst such a group of vaccinated animals prior to their release.

Copyright © 2004 By Horizon Bioscience

Salt

1131

6. Diagnosis of Persistent FMDV Infection The isolation ofFMDV in oropharyngeal scrapings collected from FMD convalescent cattle using Grae and Tallgren's probang sampling cup was first described by van Bekkum et ale (1959) and subsequently by Sutmoller and Gaggero (1965) using a modification of the sampler. Since that time the 'probang test' has become the accepted method for the detection of persistent infection with FMDV (Hedger and Stubbins, 1971). Improvements upon the sensitivity of the original detection system have been made, including the pre-treatment of oropharyngeal scrapings, or probang samples, with organic solvents (Sutmoller and Cottral, 1967) and perhaps the most significant development, the use of primary calf thyroid cell monolayers (BTY) for virus isolation. Primary bovine thyroid cells have been shown to be the most sensitive in vitro assay for the detection of FMDV (Snowdon, 1966), with an equivalent sensitivity to cattle tongue inoculation (House and House, 1989). Whilst this system is reliable for the detection of FMDV in clinical samples collected during the acute stage of FMD, the low virus titre, the intermittent nature of virus recovery (Sutmoller and Gaggero, 1965; Burrows, 1966) and the possible presence of neutralising antibody in probang samples from carrier animals (Hyslop, 1965) may reduce the efficiency of virus detection. However, the isolation of FMDV from probang samples in tissue culture and subsequent typing of the virus in harvested supernate (Roeder and Le Blanc Smith, 1987) continues to be the most widely used system for the identification of FMDV carriers for international trade and epidemiological purposes. The application of modem molecular techniques to the routine diagnosis of the carrier state in FMD has so far been unsatisfactory. Procedures utilising in situ hybridisation and the polymerase chain reaction (peR) for the detection of FMDV genomic material in bovine tissues and probang samples respectively, have been reported (Brown et al., 1992; House and Meyer, 1993; Hofner et al., 1993; Donn et al., 1995; Woodbury et al., 1995; Callens et ai., 1998). However, the full potential of these techniques to detect the minute quantities of viral RNA which are undoubtedly present in FMDV carrier animals, has yet to be exploited as successfully as it has been for other picornaviruses (Kandolf et ai, 1993). Serum antibody responses to FMDV capsid proteins are an unreliable means of differentiating between persistently infected convalescent animals and those that have eliminated the infection. Similarly conventionally vaccinated, protected cattle that become FMDV carriers post-challenge are serologically indistinguishable from animals that eliminate the infection. Hedger (1970) noted that FMDV carriers were unlikely to be present in endemic regions in herds with low mean serum neutralising antibody titres. In individual carrier cattle however there is no correlation between persistent FMDV infection and the presence of serum antibody. Therefore, Hedger's observation was indicative of a boost in herd immunity by a recent FMDV infection, with the concurrent establishment of persistent infections. Indeed, the same author previously reported that 8% of FMDV carriers detected in a field survey in Botswana were seronegative for the purposes of international trade (Hedger, 1968). Auge de Mello et ale (1970) have also described the failure to seroconvert by susceptible cattle persistently infected with FMDV following contact with cattle infected with an attenuated type C 3 virus. There are two situations in which a reliable method for the diagnosis of FMDV carriers is of paramount importance. For the purposes of international trade there Copyright © 2004 By Horizon Bioscience

\114

FMDV Persistent Infection

is much interest in a serological assay which would improve the detection of FMDV carriers. More particularly a test is sought which will differentiate between post-vaccinal and convalescent animals, both of which would be serum antibody positive. In this case all convalescent animals could be considered potential FMDV carriers and as such pose too great a risk to import. For the reasons mentioned above serum antibody titres to the structural proteins of FMDV are not useful in this situation. Serological differentiation of post-vaccinal and convalescent animals must therefore concentrate upon the humeral immune response to the non-structural proteins of the FMDV. The rationale for developing a serological test based upon the response to the non-structural proteins of FMDV is that these proteins are not present in FMD inactivated viral vaccines and thus only infected animals will produce a serological response to them following virus replication. The agar gel immuno-diffusion test (AGID) (Cowan and Graves, 1966) has been widely used in epidemiological investigations (Hafez et al., 1994) to detect serum antibody responses to the non-structural protein 3D of FMDV, which is the RNA-dependent RNA polymerase. It is thought that this protein is the major component of the complex derived from FMDV-infected cell culture traditionally known as the 'Virus Infection Associated Antigen' or VIAA. Several laboratories and companies are optimising ELISA, immunoprecipitation and blotting techniques using cloned and expressed recombinant FMDV non-structural proteins for serological diagnostic purposes to replace the less sensitive AGID test (Villinger et ai., 1989; Alonso et al., 1990; Berger et ai., 1991; Rodriguez et al., 1994; Mackay, 1998). The success of such assays is unfortunately complicated by the observation that vaccinated animals can develop serum antibody responses to some of the non-structural proteins of FMDV, especially following multiple immunisations (Pinto and Garland, 1979; Lubroth and Brown, 1995). Such antibody responses are thought to be the result of the putative presence of 3D within FMDV viral particles (Newman et al., 1994), which were previously thought not to contain non-structural proteins, and/or the contamination of FMD vaccines by other FMDV non-structural proteins, present in the cell culture harvests used for antigen production following purification and concentration. However, a report by Lubroth and Brown (1995) suggested that antibody response to the non-structural protein 2C is a strong candidate for the discrimination between FMDV convalescent and post-vaccinal sera. In recent years assays based on the detection of serum antibody recognising combinations of non-structural proteins have been developed that allow discrimination of post-vaccinal from convalescent sera at the herd level (Sorensen et al., 1998). The growing confidence in such assays is reflected in the inclusion of their use in the new EC Directive for FMD control and the D.I.E. Manual of Standards for Diagnostic Tests and Vaccines. The second situation in which an improved assay for the detection of FMDV carriers would be useful is following an emergency ring-vaccination campaign in an FMD-free region. Detection of the FMDV carriers amongst the vaccinated population, all of which would be sero-positive by conventional serological assays, \vould allow the non-carriers to be mixed safely with a susceptible livestock population if so required, whilst the carriers could remain isolated. In view of the inadequacies of currently available techniques for the accurate identification Copyright © 2004 By Horizon Bioscience

Salt

1151

VACCINATES --.- CARRtERS (Na7)

2

-A- EUMtNATORS (N=2)

G)

~ ., c(

1

U)

:::J

w

~ .J

0

-1 0

30

60

90

120

150

180

210

Days post infection

NON-VACCINATES 2

--.- CARRIERS (N=10) - A - ELIMINATORS (N=3)

.,!

;;

;A

1

::J w 0

ci

0

0 .J

-1

0

30

60

90

110

150

180

210

Days post infection Figure 1. FMDV type Ol-specific IgA titres in probang samples collected following contact challenge of vaccinated/protected and non-vaccinated/susceptible cattle with FMDV type 0 1• Points represent geometric means ± standard deviation.

of FMDV carriers, several approaches are being made to improve the situation. Bergmann et ale (1993) suggested that the serological response to the non-structural proteins of FMDV was significantly prolonged in persistently infected cattle. However, the immunoblotting system used in this work is relatively cumbersome for routine diagnosis and it is hoped that an ELISA-based system will be developed for this purpose. Archetti et al. (1995) suggested that a kinetic ELISA for FMDV-specific IgA in saliva or probang samples could be used for the identification of herds of vaccinated cattle infected with FMDV on the periphery of an outbreak. Isotypespecific ELISA analysis of probang samples from FMD convalescent cattle has shown that there is significantly higher FMDV-specific IgA in carriers than noncarrier cattle (Salt et al., 1996a). This relationship is also seen in saliva and sera, which are more readily obtainable samples for analysis (see Figure 1). An assay based upon the prolonged production of FMDV-specific IgA in FMDV carriers therefore has the potential to become a useful and simple diagnostic assay. Copyright © 2004 By Horizon Bioscience

\116

FMDV Persistent Infection

Molecular techniques for the diagnosis of FMDV have been described recently which offer the potential for improved detection of a low genome copy number in clinical samples. Several assays have been described which utilise peR for the detection of FMDV in probang fluid samples either directly (Amaral-Doel et ai., 1993; Donn et al., 1994) or following inoculation of tissue culture (House and Meyer, 1993). The exquisite sensitivity and potential for cross-serotype reactivity of these systems could be a major improvement in the diagnosis of FMDV carriers. However, the interpretation of a positive result by the direct assay still has to be determined in relation to the presence of replicating virus. 7. The Carrier Threat The successful transmission of FMDV from a carrier animal to a susceptible animal may be a rare event and occurs only under a particular set of circumstances for each of the participants. However, in any FMD-free region of the world, contact between susceptible livestock and FMDV carrier animals is unacceptable. International trade in livestock and livestock products between FMD-free countries and countries in which FMD is endemic is quite rightly severely restricted. Quarantine, laboratory testing and certification of non-vaccinated status are the preconditions imposed on the entry of livestock from such countries. Following the decision of the E.U. to adopt a 'stamping out', non-vaccination policy for the control of FMD from 31 st December, 1991, prophylactic vaccination has ceased and the European herd has become fully susceptible to FMDV infection. There was ample evidence of this fact in the U.K in 2001. The most significant threat represented by FMDV carrier animals is the introduction of FMDV into FMD-free regions, and to a lesser extent the potential introduction of new strains of FMDV into regions which are not FMD-free. There are two potential sources of FMDV carrier animals for a FMD-free region: external and internal. FMDV carriers may enter a country by the legal importation of animals wrongly certificated or identified, or by the illegal entry of animals as a result of smuggling across international boundaries. The latter instance is encouraged by the Iikely price differential existing between countries free from FMD and endemic countries, because of the former's access to international markets. Secondly, carrier animals may be created following outbreaks of FMD within FMD-free countries if implementation of the 'stamping out' policy is inadequate to control an outbreak and a 'ring-vaccination' policy is resorted to for containment of the outbreak. In this situation many protected animals could be exposed to FMDV around an infection focus, and would necessarily have to be treated as potential FMDV carriers. The prolonged duration of persistent FMDV infection in cattle would require these animals to be held and monitored for at least two years. Obviously there is a risk that these cattle could become mixed with the susceptible population if not slaughtered. In addition, challenge of susceptible cattle with low doses of FMDV, as may occur around the edges of a virus plume from a future FMD outbreak, favours inapparent sub-clinical infection (Sutmoller et aI., 1968) and possibly the undetected production of carriers that would have free movement once 'emergency' movement restrictions were lifted.

Copyright © 2004 By Horizon Bioscience

Salt

8. Potential Mechanisms of FMDV Persistence Despite the intense scrutiny that animals persistently infected with FMDV have historically attracted and continue to attract, surprisingly little is known about the mechanism by which the carrier state is established or maintained. Nevertheless, because of the potential that carrier animals have for the initiation of FMD outbreaks it is important that investigation continues. Persistent viral infections of vertebrates are not uncommon. Many strategies are employed by different viruses to enable them to avoid immune clearance by the host, an area that has been extensively reviewed (Oldstone, 1991; Gooding, 1992; Oldstone and Rall, 1993). For a highly lytic virus to establish a persistent infection it must moderate its replication and escape the immune response either through evasion or direct suppression (Borrow et ai., 1991). However, viral persistence can occur in the face of an apparently functional immune response, although replication rate and expression of viral proteins is down-regulated in comparison to that seen in acute infections with the same virus (Pelletier et ai., 1991). Several reviews have described the array of decoy proteins and cytokine homologues, mostly encoded by DNA viruses, which intertere with the immune response and favour virus survival in a host (Gooding, 1992; Banks and Rouse, 1992; Alcamf, 2003; Alcamf and Koszinowski, 2000). Genomic integration as a mechanism for RNA virus persistence within a host is only possible for the retroviruses. Latency as a means of virus persistence in the host is best demonstrated by the herpesviruses, whose members are able to establish latent infections in which there is either a very slow turnover of virusinfected cells or failure to assemble virus particles due to a temporary block in structural protein synthesis (Docherty and Chopin, 1974). However, some reports have associated the persistence of apparently non-replicating RNA of two picornaviruses, coxsackieviruses Bland B2, with chronic muscle fatigue syndromes in mice (Tam et al., 1991) and in man (Cunningham et al., 1990), respectively. The role that this form of persistence plays in the carrier state of FMD is unexplored, and despite the frequency of persistent picornavirus infections in vertebrates (Colbere-Garapin et al., 2002), the mechanism of persistence is not entirely clear for any of these infections. In the remainder of this chapter the interaction of the virus and the host during persistent FMDV infection will be considered from the twin perspectives of the host immune response and the phenotypic variation of the virus. 8.1. Evasion of the Immune Response 8.1.1. Induction of an Ineffective Immune Response

One of the essential elements of a persistent virus infection is evasion of the host immune response. Viruses may achieve this goal through the induction of an ineffectual or inappropriate immune response, which allows the initial establishment of the infection, and fails to achieve its future elimination. Induction of an inappropriate or non-neutralising antibody by persistent viruses can block effective antibody either by stripping cell sunace viral antigen or by the production of infectious immune complexes \vhich allow the virus to survive in the host and even gain entry to cells not normally susceptible to infection, via Fc receptor-mediated interactions.

Copyright © 2004 By Horizon Bioscience

1171

\118

FMDV Persistent Infection

Immunity to FMD is primarily mediated by neutralising antibody to viral structural proteins. Recovery from infection and the protection afforded by inactivated virus vaccines correlate well with virus neutralising antibody titres in cattle. There is no membrane-expression of virally-encoded structural proteins during FMDV replication in infected cells, therefore a role for antibody in direct cytotoxicity mechanisms for the elimination of virus-infected cells would not be anticipated. However, the occurrence of animals apparently protected by sub-neutralising serum antibody titres (McCullough et aI., 1992; Liebermann et al., 1993) suggests that other effector mechanisms may be involved in protection. McCullough and Crowther (1986) have suggested that the major function of antibody in the clearance of FMDV is the opsonisation of virus particles for subsequent phagocytosis and degradation by macrophages. The demonstration of classical CD8+ FMDV-specific cytotoxic T cells was elusive (Collen et aI., 1984), but has ultimately been achieved (Childerstone et aI., 1999). It has to be assumed that cellular as well as humoral immunity is responsible for elimination of FMDV infection and future protection (Chapter 8). Infection of susceptible cattle with FMDV results in a rapid rise in serum neutralising antibody which can be detected from 3-4 days post-infection. This early antibody comprises mostly IgM with some IgA, is relatively serotype cross-reactive and peaks at around 10-14 days before declining to base-line titres within 30-40 days (Brown et ai., 1964; Abu Elzein and Crowther, 1981). Isotypeswitching and increased serotypic specificity occurs as the response matures into predominantly an IgG response, with measurable titres of IgA. IgG 1 is detectable at 4-7 days post-infection and is highly serotype-specific. The development of this serum response coincides with resolution of lesions, termination of viraemia and the reduction of virus excretion. Virus neutralising antibody titres peak at around 28 days and remain at protective titres for prolonged periods of up to 4.5 years in cattle (Garland, 1974). There is no correlation between post-vaccinal serum antibody titre and the development or duration of persistent infection in individual FMDV-challenged cattle (de Leeuw et ai, 1978; van Bekkum, 1973). Vaccinated, passively immunised and naive cattle appear to be equally likely to develop into FMDV carriers following either overt or sub-clinical FMD (Burrows, 1966). Bergmann et ale (1993) have suggested that the serum antibody response to non-structural FMDV proteins is prolonged in carrier animals in response to the continued replication of virus. If this is true, it would appear that in some carrier animals at

least there is continued presentation of FMDV proteins to the systemic immune system. As a general rule, once persistent infection is established, FMDV carrier animals maintain high neutralising serum antibody titres similar to or even above those of non-carrier convalescents (Garland, 1974; McVicar and Sutmoller, 1974). However, carrier cattle have been found infrequently to have low or non-detectable serum neutralising antibody titres to FMDV (Hedger, 1968). A reasonable conclusion from this data would be that serum antibody per se does not affect the establishment or maintenance of the carrier state in FMD. There is some evidence that post-vaccinal serum antibody (McVicar and Sutmoller, 1976) and secretory antibody in polyvaccinated cattle (Garland, 1974) may ameliorate the peak titre of replicating pharyngeal virus in FMDV-challenged Copyright © 2004 By Horizon Bioscience

Salt

1191

cattle, but have no influence over the outcome of the infection in terms of early oropharyngeal virus replication and persistence or elimination. Matsumoto et ale (1978) have shown that the secretory neutralising antibody response also persists for longer and at higher titres in carrier animals compared to non-carrier convalescents. Garland (1974) showed that IgA is the predominant neutralising antibody isotype in the secretions of convalescent cattle and that titres only decline following the cessation of virus isolation. Therefore, the local antibody response to FMDV in the upper respiratory tract appears to be actively stimulated during the carrier state, an observation confirmed by the author (see Figure 1) (Salt et al., 1996a). In FMD convalescent cattle, McVicar and Sutmoller (1974) showed that resistance to reinfection with homologous FMDV correlated with a history of persistent infection. These authors noted that the ability to mount systemic and local anamnestic antibody responses was not restricted to those cattle able to resist the establishment of persistent infection upon further exposure to homologous virus. The author's own results from studies upon homologous reinfection of convalescent cattle confirm the observation that resistance to local virus replication in the oropharynx following rechallenge shows a strong correlation with prior carrier status and the continued presence of IgA in secretory fluids (Salt, 1993a,b). The presence of IgA in probang fluid was a better indicator of the outcome of reinfection than serum antibody titre. Whether IgA is the mediator of protection against local virus replication or is merely an indicator of an ongoing local immune response in the pharynx which itself imparts protection against reinfection is impossible to say. Thus there seems to be no consistent failure or deficiency in the antibody response of animals that become persistently infected with FMDV. Indeed, both systemic and secretory responses are exaggerated and prolonged for the duration of the carrier period, and may render the individual immune to further acute infection with closely related virus but not to virus of another serotype (Auge de Mello et al., 1970). In FMDV carrier animals, detectable virus is effectively eliminated from all sites except the oropharynx, from where antigen presumably is released into the surrounding tissues to fuel the ongoing immune response. 8.1.2. Replication in an Immunologically Privileged Site

Persistent viral infection can occur in highly specific locations at the whole organ or cellular level. A variety of factors could be responsible for this phenomenon including the stage of differentiation of the cell, the function and site of the cell, and the distribution of the viral receptor. In certain so-called immunologically 'privileged sites' within the host, persistence is favoured because elimination of infection is less efficient. These sites include the mammary glands, salivary glands, kidneys and hepatic biliary system; all epithelium-lined secretory or excretory organs. In each case the surface of the infected cell faces into a lumen which is outside the body (Mims, 1988), and it is from the luminal surface that virus is released from epithelial cells without cytolysis. In a second group of 'privileged sites', which includes the eye, brain, testis and the foeto-maternal unit, active immunological suppression is induced by elements of the local micro-environment within the target organ (Streilein, 1993). Furthermore, a virus may 'hide' in cells Copyright © 2004 By Horizon Bioscience

\120

FMDV Persistent Infection

that naturally lack MHC class I antigen expression, as does herpes simplex virus in neurons (Banks and Rouse, 1992), or actively down-regulate the level of MHC class I antigen constitutively expressed, thereby avoiding a CTL response (Yamashita et al., 1993; Alcamf and Koszinowski, 2000). Epithelial tissues constitute an immunologically 'privileged site' because of their polarisation, their relative avascularity and the supposedly restricted accessibility of them to elements of the immune system. Possibly as a result of these factors, several viruses of vertebrates establish persistent infections in epithelial sites. Site and cell-type may play an important part in the establishment of persistent herpes and other virus infections, inasmuch as Epstein-Barr and Aujeszky's viruses are 'carried' in the epithelial lining of the oropharynx and are intermittently shed in saliva. Studies upon the early pathology of FMDV have shown that the oropharynx is a predilection site for FMDV localisation and replication in cattle (Burrows et al., 1971). The recovery of infectious virus in probang samples collected from the oropharynx prior to measurable viraemia also supports this theory (McVicar et al., 1970). There is some evidence from immunocytochemical studies on acutely infected bovine tissue that occasional infected cells occur within the epithelium of the tonsil and in the acini of mixed mucous-serous glandular cells in the lamina propria of the tonsil and dorsal soft palate of naturally infected cattle (Donn, 1993). Woodbury et ale (1995) extended this work to show that the cells of the basal and intermediate layers of the oropharyngeal epithelium are infected up to 17 days post-infection of cattle with FMDV. Using a DIG-labelled probe these authors detected viral RNA by in situ hybridisation diffusely distributed in the epithelium of the tonsillar crypt and dorsal soft palate. Further work has now extended the identification of FMDV RNA in this site into the carrier period proper (quoted in Alexandersen et al., 2002). In an excellent recent comprehensive review of the subject, Alexandersen et ale (2002) propose that the epithelium of the ruminant dorsal soft palate and roof of the adjacent oropharynx represents a unique tissue architecture. Furthermore the authors postulate that it is the unusual absence of a dead layer of epithelial cells in these sites that allows viral uptake from the lumen of the pharynx during persistence of the FMD virus, thus evading contact with serum antibody. 8.2. Interference with the Immune Response In contrast to the avoidance of elements of an effective immune response by dint of inaccessible replication sites, a far more interactive strategem employed by some viruses to enable persistence is active interference with the immune system to suppress it or induce an ineffective response (Oldstone, 1991). This may involve the infection of cellular elements of the immune system, as occurs in the persistence of poliovirus (Carp, 1981) and bovine viral diarrhoea virus (BVDV) in gut-associated lymphoid tissue in man and cattle respectively, and lymphocytic choriomeningitis virus (LCMV) in peripheral mononuclear cells and splenic lymphoid tissue in mice (Ahmed et al., 1991). Furthermore, there is ample evidence of the acquisition by some viruses of host genes that encode proteins normally involved in the immune response to viral infection. Through the selective expression of homologues of host immune mediator molecules or their receptors the virus is able to modify the Copyright © 2004 By Horizon Bioscience

Salt

1211

innate and adaptive arms of the host immune response and thereby avoid clearance (reviewed by Gooding, 1992; Alcamf, 2003; Alcamf and Koszinowski, 2000). One of the increasing number of examples of this phenomenon is Epstein-Barr virus (EBV) which possesses a homologue of the human IL-I0 gene and is able to suppress the THJ-cell mediated generation of cytotoxic T cells (CTL) through expression of the gene in persistently infected cells (Mosmann and Moore, 1991). Antigenic overload has been proposed as another means by which virus infection can subvert the immune response. In some experimental infections of mice with LCMV it has been suggested that overwhelming virus titres at the early stages of an infection may deplete the CTL response, resulting in persistent infection (Moskophidis et ai., 1993). As a small RNA virus it is extremely unlikely that FMDV could establish a persistent infection primarily through the expression of virally-encoded immunomodulatory proteins. Indeed, the genome of FMDV encodes 13 polypeptides, which along with those of other picornaviruses show no homology to known immunologically active molecules (Porter, 1993). Further elucidation of the function of the non-structural proteins of the FMD virus may yet reveal activities that could moderate infected immune cell function. Poliovirus, another picornavirus, replicates in lymphoreticular tissues of the human alimentary tract. Persistent poliovirus infection can become established in these tissues in vivo, and in vitro persistently infected tissue cultures are readily established in human lymphoid cells with the generation of defective interfering particles (DIPs) and temperature-sensitive (Ts) mutant viruses (Carp, 1981). Although as yet there is no data to support viral interference with cellular elements of the immune system in FMDV carrier animals, there is limited evidence that the virus can infect and replicate in a range of cultured mononuclear cells from susceptible species: bovine T lymphocytes (H.Takamatsu, personal communication), porcine macrophages (Baxt and Mason, 1995), porcine skinderived dendritic cells (Gregg et ai., 1995a) and bovine Langerhans cells (David et ai., 1995). Vilma (1980) described the cell-associated transport of viable FMDV from sites of inoculation to other epithelial sites in the earliest stages of FMD in cattle. He concluded that mononuclear cells were responsible for the pre-viraemic spread of FMDV to sites of primary virus replication and subsequent lesion formation prior to the ingress of neutrophils. In a later study of the pathogenesis of FMD in cattle, Burrows et ai. (1981) came to the same conclusion concerning the role of bovine macrophages in the pre-viraemic dissemination of FMDV. Studies upon experimental FMDV infection of guinea pigs (DiGirolamo et aI., 1985) and pigs (Brown et aI., 1995) have suggested a role for Langerhans cells in the early pathogenesis of FMD through the dissemination of virus. Gregg et ai. (1995a) have recently shown that a population of cultured porcine skin-derived dendritic cells normally susceptible to FMDV infection could be rendered nonpermissive by prior infection with African swine fever virus (ASFV). In parallel clinical studies a similar result showed that ASFV-infected pigs were resistant to FMDV infection (Gregg et al., 1995b). Therefore, Langerhans-type cells appear to play a role in the pathogenesis of FMD in pigs, and taken together these observations suggest that at least in the early stages of the disease FMDV can infect cells of the immune system and remain infectious. It is important to discover Copyright © 2004 By Horizon Bioscience

1122

FMDV Persistent Infection

if this is also true in ruminants in view of the pivotal role that cells of the dendritic lineage are thought to play in the initiation of the immune response to viral infections, especially those of epithelial surfaces (Yirrell et at., 1994). Impairment of the function of this important antigen-presenting cell could have implications for the induction and maintenance of the immune response, both humoral and cellular, and has the potential to influence elimination or persistence of the virus. It is interesting to speculate upon the significance of porcine Langerhans cell infection by FMDV. Bearing in mind the efficient elimination of FMDV from the oropharynx of the pig compared to the persistent infection of this site in ruminants, it is tempting to speculate that there is a qualitative difference in the immune response to FMDV between the species. The induction of a CTL response by the active infection of a particular cell type able to present FMDV-derived antigen in the context of MHC Class I following endogenous processing could constitute just such a difference. 8.3. Virus Variation 8.3.1. Antigenic Variation

One of the most effective means by which infectious organisms persist at the population and individual level is antigenic variation. Viruses may thus avoid extinction through evasion of the host's humeral or cellular immune response. At the population level the persistence of influenza virus through antigenic drift and shift is the most notable example of this phenomenon. The lentiviridae probably provide the best example of antigenic variation within the individual as a mechanism for immune evasion and subsequent persistence (Kono et aI., 1971). RNA virus genomes are highly unstable, and they show great potential for variation and mutability (Holland et aI., 1982). The mutation rate of FMDV has been shown to be similar to other RNA viruses (McCahon, 1986), and is thought to be constant throughout the length of the genome, although tolerated mutations are fixed at higher frequencies in the regions of the genome encoding the surface exposed loops of the structural proteins which are not critical for the structural integrity of the capsid (Domingo et al., 1993). The existence of seven serotypes and myriad antigenic subtypes of FMDV bears witness to the remarkable potential for antigenic diversification within the genus (Chapter 10). In vitro studies on FMDV have shown that genomic variation can occur in the absence of selective immunological pressure both as a result of point mutations caused by error-prone copying of the genome, and by recombination events between related FMDV genomes replicating in the same cell (Fellowes and Sutmoller, 1970). Rowlands et at. (1983) found that genomic variation during serial passage of FMDV in tissue culture resulted in amino acid substitutions and alteration of viral antigenicity. These authors suspected that their initial isolate may have been a mixture of virus variants, an idea expanded by Domingo et al. (1985) who have proposed that in an infected cell culture or individual animal there will be a consensus or 'master' sequence and an equilibrium distribution of variant sequences that constitute a 'quasispecies'. It has been shown that even cloned virus isolates consist of a complex distribution of variant genomes (Villaverde et aI., 1991), which Domingo et al. (1989) have suggested will drift antigenically in vivo under immune pressure to establish a new virus population with an altered 'master' sequence (Chapter 10). Copyright © 2004 By Horizon Bioscience

Salt 1231

More recently, the dual effects of random mutation and positive immune selection during the serial in vitro passage of an FMDV type C clone have been demonstrated (Borrego et al., 1993). In this study, antigenic variation was observed in passaged virus in the presence or absence of limited neutralising antibody, although the two derived variant virus populations became distinctly different. Similarly the rapid generation of antigenic diversity has been observed in FMD viruses serially passaged in both non-immune pigs (Carrillo et al., 1990) and in partly immunized cattle (Fagg and Hyslop, 1966). The genetic heterogeneity of field isolates collected from individual FMD outbreaks is well-established (Domingo et al., 1980; King et al., 1981) and the recovery of two different nucleotide sequences in a single isolate is not uncommon (N.J.Knowles, personal communication). The implication of this well-documented genomic plasticity of FMDV is that in an individual animal, immunologically resistant or tissue-tropic variants could be rapidly generated or selected from an existing virus population, thus favouring persistence over elimination. Indeed, in a recent study Escarmis et ale (1998) concluded that FMDV infection of partially permissive cells in vitro can promote the rapid selection of virus variants that show alterations in cell tropism and are highly virulent for the same cells. Co-evolution of virus and their infected host cell population is a characteristic common to several persistent picornavirus infections (Colbere-Garapin et al., 2002). However, epidemiological studies of outbreaks of poliovirus (Kinnunen et al., 1991), bovine enterovirus (BEV) (Hamada et al., 1990) and FMDV (Martinez et aI., 1992) have all shown that distinct genomic lineages are generated in which the degree of variation was related more closely to the number of outbreaks and transmission events than the temporal separation of isolations per se. This would suggest that the random sampling of a genomic variant from a population of sequences at transmission is the more likely explanation for lineage fixation than immune-driven antigenic selection (Domingo et aI., 1980). A study of FMDV type 0 1 isolates collected from persistently infected cattle showed that most of the observed genomic variation occurred early in the course of the infection when greater virus replication would be expected (Malirat et aI., 1994). Indeed, as yet antigenic variation as a means of immune evasion and the establishment of persistent infection has not been recorded for any member of the picornaviridae. Data from Zimbabwe (Bastos et aI., 2001) and Saudi Arabia (Samuel et al., 1991) suggests that the evolution of FMDV in the field is a gradual process. The implication is that major antigenic shifts do not occur frequently in carrier animals, or that if they do occur then these viruses are not transmitted to susceptible animals. Rapid antigenic diversification of FMDV has been demonstrated in pigs under controlled conditions, including variation within the major antigenic region of VPl (Carrillo et aI., 1990). However, in this species persistent infection with FMDV is not established which again questions the role of antigenic variation in the carrier state. In a study of genetic variation of FMDV during persistent infection of cattle, considerable variation was recorded over a two year period (Malirat et al., 1994). In this study sequential FMDV type 0 1 Campos isolates varied irregularly through the occurrence of point mutations. Viral populations tended to fluctuate, rather than to evolve as genomic lineages with accumulated conserved changes. Despite the

Copyright © 2004 By Horizon Bioscience

\124

FMDV Persistent Infection

authors view that escape mutants may have been generated in these carrier animals partly as a result of immune pressure, a homologous polyclonal post-vaccinal serum consistently neutralised all FMDV isolates collected throughout the period of persistence. Several authors have documented antigenic variation in the virus isolated from FMDV carrier animals in the field (Hedger, 1968; cited by Brooksby, 1968), whereas other reports have commented upon the stability of carrier virus antigenicity (Burrows, 1966) and genomic sequence (Dawe et al., 1994b). In an experimental study of the persistent infection of cattle following challenge with a cloned type C 3 FMDV, Gebauer et ale (1988) found many nucleotide substitutions in carrier virus isolates within 63 days of infection, 59% of which resulted in amino acid changes. The carrier isolates from one of three animals in the study group showed a dramatic reduction in reactivity with a panel of monoclonal antibodies (mAb's) raised against a short peptide containing the amino acid sequence of one of the major antigenic site of the challenge strain. This finding led the authors to conclude that the inapparent persistent infection of ruminants is a potential reservoir of outbreak virus and may promote the rapid antigenic drift of FMDV. A similar study on a type 0 1 persistent infection of cattle concluded that despite the continual modification of the virus population during persistence there was no predictable pattern of virus variation, and that most changes in the virus genome \vere not conserved (Malirat et al., 1994). Genetic sequencing studies conducted by the O.I.E.lF.A.O. World Reference Laboratory (WRL) for FMD, Pirbright, on FMDV carrier viruses collected from cattle in Zimbabwe over a three year period have demonstrated a remarkable genomic stability (Bastos et al., 2001). Similarly, data from a controlled study on the antigenic profile of FMDV isolated from persistently infected cattle at the Institute for Animal Health (IAH), Pirbright, supports this observation (Salt et al., 1996b). Table 3 shows the antigenic profiles of carrier isolates of a Middle Eastern type 0] FMD virus as measured by a panel of mAb's. Isolates collected over a period of 33 weeks from persistently infected cattle showed no loss of reactivity with neutralising mAb's that recognised two distinct epitopes on the antigenically dominant VPl protein of FMDV. The degree of antigenic variation required for the evasion of a polyclonal humoral response to all five of the defined neutralising antigenic sites on type 0 FMD viruses (Crowther et al., 1993) would need to be greater than that detected in this study. In addition, serum samples collected from one carrier animal at the beginning and at the end of the period of persistent infection showed equal neutralising activity against virus isolates collected between 8 and 32 weeks post-infection. Therefore, despite evidence that antigenic variation can occur in FMDV as a result of immune selection during multiple passage in tissue culture, in partially immune cattle and in persistently infected cattle, the establishment and maintenance of the FMDV carrier state is not dependent upon antigenic variation to evade the humoral immune response. Before leaving this important section on antigenic variation in persistent FMDV infection it is worth considering a final source of information. RNA viruses readily establish persistent infections in cell culture and FMDV is no exception. Baby hamster kidney (BHK) cell lines persistently infected with a plaque-purified type C 1 FMD virus have been established from cells surviving an initially lytic

Copyright © 2004 By Horizon Bioscience

Salt 1251 Table 3. Reactivities of a panel of monoclonal antibodies (mAb's), raised against FMDV type 0 1 Lausanne, with virus isolates from probang samples of cattle persistently infected with a Middle Eastern type 0 1 FMDV field isolate. Values represent percentage reactivity of each mAb with the homologous virus, \vhere H >70%, M 40-70% and L

U (Figure 8b). 2.2. Relative Importance of Monocytes and DC in Induction of Anti-FMDV Immune Responses

In the blood, monocytes by far dominate over the numbers of DC. Although this will not be maintained in other organs and tissues, particularly in the lymph nodes, the monocytes will have a role to play. Monocytes are not as suppressive as macrophages (see Figure 2) (Basta et a!., 2000; Carrasco et al., 2001), but they can regulate anti-FMDV immune responsiveness (Barnard et al., 2003). Both monocytes and DC in porcine blood carry the pan-myeloid marker SWC3, but only the monocytes are CD14+ (Summerfield et a!., 2003). Removal of all SWC3+ cells from PBMC obtained from an anti-FMDV immune animal abrogated the responsiveness to FMDV antigen (Figure 7a, "responders alone"). Only replacement of the SWC3+ cells restored the anti-FMDV response (Figure 7b). This is not surprising, considering that the DC were also removed. When the SWC3+ cells were only partially depleted from PBMC, the remaining SWC3+ cells were dominated by the CD14- cells - in the original PBMC, the SWC3+ cells are dominated by the CD14+ monocytes (Barnard et a!., 2003; Summerfield et a!., 2003). With an SWC3+ fraction dominated by the CD14- cells, a particularly efficient stimulation of the T lymphocytes was obtained (Figure 7b). Adding the removed cells back to this partially depleted SWC3- fraction again reduced the FMDV-specific lymphoproliferation (Figure 7b). These analyses demonstrated that a balance had to be reached between the stimulation of T lymphocytes via antigen presentation on DC, and the regulation of that response by monocytes. It is also likely that Treg lymphocytes would be involved in the latter process. Nevertheless, a clear interaction between monocytes and DC was evident, and the monocytes could not be defined as suppressor cells, as is the case with macrophages (see Figure 2) (Basta et al., 2000). Only when the monocytes were present at relatively high ratios with the T lymphocytes, was reduced lymphoproliferation evident (Figure 7a). Kinetic analyses also demonstrated that this "regulation" was dependent on the time post-stimulation

Copyright © 2004 By Horizon Bioscience

\186

Humoral and Cellular Immune Response Against FMDV

(a) ~ 20 o

(ij

e?

~

c::

o 0en ~

15 10

.9 t)

~ en

5

~

.c

!

2

0 1:6

ratio Sl.1Ji 1:54 C3of.:S 1ill vVC3-

SWC3only

3 day poststimulation

(b)

SWC3+ & SWC3-

SWC3-

PBMC

o

10000

20000

30000

cpm, 3H-TdR uptake Figure 7. (a). The influence of SWC3+ cells on SWC3- cells (containing CD14- SWC3+ cells on stimulation of anti-FMDV lymphocyte responses. PBMC were sorted by labelling with antibody against the pan-myeloid marker SWC3, to separate into an SWC3+ and SWC3- fractions. These fractions were tested for their response to FMDV, and the SWC3+ fraction was back-titrated into the SWC3- fraction (x-axis; ratio SWC3+ : SWC3-). The responsiveness to FMDV stimulation was measured at different times after stimulation (y-axis; day post-stimulation), in terms of the TdR incorporation) (SI; z-axis )with respect to that obtained with the SWC3relative cpm fraction alone (SWC3- only). (b) Partial removal of the SWC3+ cells from PBMC, dominated by CD14+ cells, increases the ability of the lymphocytes in the SWC3- fraction to respond to FMDV antigen presented by the remaining SWC3+ cells (dominated by CD14- cells) in this SWC3- fraction. Re-constitution of the SWC3+ fraction with the SWC3- fraction reduces the lymphoproliferation. Adapted from Barnard et al. (2003a).

eH-

Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

(a) DC uptake of FMDV

1871

(b) McIl uptake of FMDV

Figure 8. (a) Uptake of FMDV (01 Lausanne) (white) by porcine monocyte-derived dendritic cells. (b) Phagocytosis of FMDV by porcine pulmonary macrophages. The images were obtained at 4h after interaction between the ells and the virus. The FMDV was detected using antibody against a conformational neutralisable antigenic site (4C9), and confocal laser scanning microscopy with a Nomarski Interference Contrast attachment.

with FMDV antigen. Early after interaction with the antigen, immune stimulation was notable, whereas regulation became more prominent as the immune response progressed (Figure 7a). 2.3. Conclusions

The critical player in immune response development against FMDV is the DC. Certainly, monocytes can also function as APC, but the DC are the more potent at promoting the responses. The characteristics of FMDV interaction with both DC and monocytes are essential for defining how an immune response will evolve. While stimulation of T lymphocyte responses will occur, alterations in the monocytes will result in a regulation, probably of the DC capacity to stimulate the T lymphocytes. This is of course expected in terms of controlling the immune response in a positive sense, preventing over-stimulation into an immunopathological area. Such characteristics about how the immune system regulates itself are essential for understanding the mechanisms behind efficacious vaccines, and particularly effective vaccine delivery systems.

Copyright © 2004 By Horizon Bioscience

\188

Humoral and Cellular Immune Response Against FMDV

3. Induction of Immune Responses: B Lymphocyte and Antibody Responses 3.1. Epitopes Recognised by B Lymphocytes The specificity of this response lies with the recognition of antigenic sites chemically defined as epitopes - on the virus particle. B cell epitopes and antigenic sites are those recognized by anti-FMDV antibodies, because the B cell receptor for these sites is actually an immunoglobulin molecule with the same paratope (antigenbinding site) as the antibody produced by the activated B cell. Only the isotype of the secreted antibody and the B cell receptor can be different. From the point of view of immune protection against FMDV, the antibodies which neutralise FMDV infectivity are the most important. Not only do they interfere with the ability of the virus to infect susceptible cells, but they are also efficient opsonising antibodies. This is important for the uptake of FMDV by the macrophages to destroy the virus and remove the infectious threat (see sections 5 and 6). The epitopes and antigenic sites recognised by neutralising antibodies are located within exposed motifs on the surface of the FMDV particle. On at least serotype 0 of FMDV, there are five antigenic sites containing epitopes interacting with neutralising antibodies. Only one of these, contained within the site A of the G-H loop of VP1, is continuous, whereas the other epitopes and antigenic sites are discontinuous (conformational) (Francis et al., 1990; Mateu et al., 1990; McCullough et al., 1987a,1987b; Xie et al., 1987; reviewed in Brown, 1995; Mateu, 1995). This contrasts with the structure of T lymphocyte epitopes, which is, essentially, continuous (Collen, 1994; Rowlands, 1994). The epitopic B cell structure of FMDV is addressed in detail in Chapters 4 and 9, while the T cell epitopes are dealt with in section 4 of this chapter. 3.2. Anti-FMDV Antibody Protection against FMD is often associated with the induction in serum of high levels of in vitro neutralising antibodies (McCullough et al., 1992a; 1992b; Pay and Hingley, 1987; van Maanen and Terpstra, 1989). Early experiments, performed more than a century ago, indicated that passive transfer of convalescent antibodies conferred protection to cattle (Loeffler and Frosch, 1897). However, the specific antibody response per se does not ensure clinical protection (McCullough et al., 1992a; van Maanen and Terpstra, 1989). Neutralization of virus infectivity would only confer protection in vivo if the antibody interaction disrupted virion structure, but this has been reported for only one antibody against a particular conformational epitope (McCullough et al., 1987b). If the virus structure remained intact in the complexes with antibody, any elution of the antibody would permit the infectious threat of the virus to be restored. On this basis, a virus clearance mechanism was proposed as essential for the protective immune response against FMDV - the phagocytosis of virus-antibody complexes mediating that clearance in vivo (McCullough et al., 1992b). This will be discussed in sections 5 and 6 of this chapter. It is important to stress that this clearance mechanism is efficient only when opsonising antibody has interacted the virus, and antibody which neutralises virus infectivity is a particularly potent opsonising agent. 3.3. Neutralising Anti-FMDV Antibodies The neutralizing antibodies elicited upon FMDV infection or vaccination are directed against a discrete number of antigenic sites located on the surface of Copyright © 2004 By Horizon Bioscience

McCullough and Sobrino

1891

the viral particle. As addressed in detail in Chapter 9, these are referred to as immunodominant B-cell sites, due to the efficiency with which antigen-specific Blymphocytes recognise them, and are activated to produce the reactive antibodies. Of course, the immune system does not have everything its own way. The presence of antibody against one of the immunodominant B-cell sites can lead to selection of a considerable level of amino acid variation in the heterogeneous populations (quasispecies) of FMDV (Domingo et al., 1992; reviewed in Chapter 10). This is the "host immune pressure" which can result in a wide spectrum of antigenic and immunogenic variability (Domingo et aI., 2003). Induction or selection of FMDV "antibody-resistant" variants is most likely when the neutralising antibodies are of low titre or low affinity. The latter situation would arise early during primary immune responses, when the antibody "fit" for the antigenic site had not yet been refined, as occurs during the continued "maturation" of the immune response. In addition, the presence of neutralising antibody against only one antigenic site offers the virus greater potential to escape the protective role of the antibody. When antibodies against more than one antigenic site are present, the chances that virus variants can be selected to escape all antibody specificities are rapidly diminished. This is likely to be one of the driving forces in FMDV evolution, and is probably involved in the origin of the seven distinct FMDV serotypes currently in existence (reviewed in Domingo et al., 1990; Sobrino et al., 2001; see Chapter 10). It is in fact neutralising antibodies which define the FMDV serotypes (Kitching et al., 1989; Pereira, 1977). FMDV neutralising antibodies are serotype-specific, and this specificity correlates well with the lack of protection against heterologous challenge infection in animals, using FMDV of a different serotype to that employed for the immunisation (Pereira, 1981). Such antigenic and immunogenic complexity has important consequences in the fields of FMDV diagnosis (see Chapter 15), and vaccine design (see Chapters 11 and 12). 3.4. Antibody Isotypes and the Kinetics of Their Induction

A typical profile for the induction of anti-FMDV antibodies with time postvaccination is shown in Figure 5. The IgM isotype of antibody is the first detectable serum neutralising antibody appearing at 3 to 4 days following infection or vaccination. Maximum serum IgM responses are observed around 10 to 14 days post-infection, after which the response declines. Serum IgGs can be detected as early as 4 to 7 days post-infection or vaccination. These are the isotypes which become the major neutralising antibodies by two weeks after encounter with the virus antigen (Francis and Black, 1983). In both infected and vaccinated animals, titres of the IgG 1 isotype are generally higher than those of IgG2 (Salt, 1993). Following infection, or oranasal vaccination, a local antibody response is also detectable in secretions of the upper respiratory and gastrointestinal tracts (Francis and Black, 1983). The major antibody subclasses found in these secretions are initially IgM, followed by IgA and IgG as the time post-infection progresses (reviewed in Salt, 1993). The duration of the antibody response, in both infected and vaccinated animals, is dependent on the dose of virus or vaccine (Pay and Hingley, 1987). Generally speaking, the duration of protective antibodies induced by inactivated vaccine is shorter than that in convalescent animals (Salt, 1993).

Copyright © 2004 By Horizon Bioscience

\190

Humoral and Cellular Immune Response Against FMDV

3.5. Antibody Interaction with FMDV

Traditionally, analyses of serum antibody employ the dilution of sera to a point where specific and non-specific/aspecific reactions are clearly separable over a long incubation period. While these methods are useful for measuring antibody titre, they are less relevant to the mechanisms of immunological protection. Firstly, by diluting sera, one ignores the roles played by low titre antibodies, natural immunoglobulins and other opsonins in the humoral immune defences. Of particular importance is their influence on one another. Although such interactions of the humoral components of immune defences are important, it should be stressed that in the case of FMDV, the protective immune defence requires specific antibodies of minimum affinity and titre. This can be seen from the fact that nonimmunised animals are sensitive to FMDV infection. An additional problem with traditional methods for measuring antibody is that several minutes are employed for the antibody-virus interactions. Proficient immune defences must function more rapidly than this, if the infectious threat of the FMDV were to be successfully countered. Indeed, when the humoral response of immune serum itself was analysed, rather than extensively diluted material > 90% of the reaction with antigen occurred within 10-60 seconds, (Figure 5) (Scicluna and McCullough, 1999; Scicluna et al., 2001) In contrast, the nonspecific/aspecific reactions required several minutes to show interaction with the antigen. This is relevant to immunological protection. Rapid reaction during seconds would have greater capacity to inhibit the viral infectious potential - as with a protected vaccinate- whereas the slow reaction over a number of minutes found with non-immune animals would have less chance against the virus. Indeed, multiply vaccinated as well as protected animals possessed sera having short reaction times «60 seconds) with FMDV, whereas sera from non-immune and unprotected animals displayed longer reaction times (Scicluna et al., 2001). Clearly, specific antibody reactivity \vith FMDV holds the key to the protective immune defence against this virus. In order to induce anti-FMDV antibody production, it is necessary that the immunobiology of T lymphocyte function be considered. The next section will deal with this area in more detail, before proceeding to the characteristics of the effector immune defences against FMDV, in sections 5 to 7.

4. Induction of Immune Responses: Th Lymphocyte Responses 4.1. Identification of Th lymphocyte response against FMDV

Restriction of the B lymphocyte response against one of the neutralisable antigenic sites on FMDV - the continuous site at the VPl G-H loop - was first described in mice (Francis et al., 1987). These authors demonstrated that the non-responsiveness of mice to a peptide antigen covering the G-H loop of FMDV serotype 0 1 could be overcome by addition of a foreign epitope. As a peptide construction, this foreign epitope was also continuous, but it was recognised by T lymphocytes, not B-Iymphocytes. This and subsequent work clearly identified the induction of antiFMDV antibody responses to be T lymphocyte-dependent (Collen et aI., 1989; Francis et al., 1987). Although T-cell independent responses can be induced in mice (Borca et al., 1986), this requires high antigen doses, probably relating to the role of DC presenting intact antigen to B lymphocytes (Litinskiy et aI., 2002).

Copyright © 2004 By Horizon Bioscience

(A) VP1 peptide pools 35

(8) VP4 peptides

6d.p.v

~ Pig 1

30

Ii Pig 2

25 20 15

x ~

.5

10

5-

5

CP

~

35

1;)

30

~

1

2

3

4

5

6

7

8

9

10 11 12 13 14

4

5 6

7

8

9

10 11 12 13 14

10 d.p.v

25 20 15

10

5

o

1

2

3

VP1 peptide pools

o

2

3

4

5

678

9

13

14

virus

VP4 peptides Figure 9. Proliferative responses of PBMC from FMDV-vaccinated pigs against optimum stimulatory concentrations of (A) VP1 peptide pools and (B) VP4 peptides 0 to 14. The animals were vaccinated twice with oil-adjuvanted, monovalent FMDV serotype C1-0berbayern vaccine. Data are expressed as stimulation index (SI; cpm in the presence of peptide/cpm medium alone), and each bar represents the mean of triplicate cultures. Adapted from Blanco et al. (2000).

Copyright © 2004 By Horizon Bioscience

\192

Humoral and Cellular Immune Response Against FMDV

Only at high antigen doses can the DC circumvent T-cell dependency (see section 1.2). At lower doses, the T lymphocyte help is essential. Certainly, efficient B cell activation and antibody production in vaccinated cattle and pigs require T-cell help (Collen, 1994; Collen and Doel, 1990; Rodriguez et al., 1994; Saiz et al., 1992). Nevertheless, animal-to-animal variation in the recognition of the T cell epitopes has been observed (Figure 9) (Blanco et al., 2000; 2001; Collen and Doel, 1990; Rodriguez et al., 1994; Saiz et ai., 1992; van Lierop et al., 1995a). This is consistent with the restricted recognition of particular T cell epitopes, due to the polymorphism of the MHC molecules (Garcia-Briones et al., 2000; Glass et al., 1991). Only certain epitope sequences will be presented in the context of a particular MHC haplotype - the so-called haplotype-restriction. Although many T cell epitopes are restricted in terms of haplotype recognition, "promiscuous" T cell epitopes also exist. These can be recognised by a broad range of haplotypes, and between species. To date, they have been identified on the VP4 of the virion (Figure 9) (Blanco et al., 2000; van Lierop et al., 1995b). 4.2. The Nature of Th Lymphocyte Responses Against FMDV

Humoral responses against FMDV are particularly restricted by the virus serotype, especially the neutralising antibodies which are often serotype-specific - referred to as homotypic. In contrast, T lymphocyte responses are a mixture of homotypic and heterotypic responses (Blanco et at., 2001; Collen and Doel, 1990; GarciaValcarcel et al., 1996; Saiz et ai., 1992). Thus, T lymphocytes from animals immmunized with a specific serotype of FMDV can be induced to proliferate by fe-stimulation with viral particles or proteins from viruses of a different serotype. Analyses of cells from infected cattle and pigs have demonstrated that a heterotypic T cell response can be elicited by non-structural viral proteins (Blanco et ai., 2001; Collen et ai., 1998). These lymphoproliferative responses against viral particles, capsid proteins, non-structural polypeptides and synthetic peptides are clearly of a T-cell nature. A major piece of evidence is the sensitivity to inhibition by monoclonal antibodies (MAb) against CD4 molecules (Blanco et al., 2000; 2001; Collen and Doel, 1990; Saiz et al., 1992; van Lierop et al., 1995b). Of course, not only Th lymphocytes, but also the natural interferon-producing cells or plasmacytoid DC bear CD4. A requirement for monocytic cell interaction with the T lymphocytes was demonstrated by inhibition of the response with MAb against MHC class II (Figure 10) (Blanco et al., 2000; 2001). Confirmation that the proliferation was indeed due to Th lymphocytes was obtained by multiple labelling of the proliferating cells. The activated cells upregulated CD25, and were primarily CD4+, but SWC3(Figure 10). This demonstrated that they were T lymphocytes, not monocytic cells (Blanco et al., 2000). The additional presence of CD8 on these CD25+ cells was indicating that they were memory and activated Th cells - the CD4+CD8+ porcine T cell subpopulation contains the memory/activated cells. 4.3. Epitopes Recognised by Th Lymphocytes

Much of the information on the epitope specificity of anti-FMDV T lymphocyte has come from studies on peptide vaccines and from epitope mapping using overlapping synthetic peptides (reviewed by Brown, 1992; Collen, 1994; Meloen Copyright © 2004 By Horizon Bioscience

McCullough and Sobrino

1931

(A) anti-SLA MAb

7000.....-,;.....;,-------------...------...,

II a-SLA-n 0 a -SLA-I II without MAb

6000

.§ ~

5000 4000

8 3000

&

2000 1000

o pig 313

pig 315 (C ) CD4+CD8+ cells

(B) CD8+ cells 32

32

'-

Q)

.c

E ~

c

~ > ~ Q)

co 32 Q)

~

(D) CD4-CD8-

(E) CD4+ cells 32

cells 11%

M1

104

CD25 expression (fluorescence intensity)

Figure 10 (A) SLA-dependency of FMDV peptide-induced specific lymphoproliferation. AntiSLA class I (white bars) and -SLA class II (grey bars) MAb blocking of the lymphoproliferation obtained with peptide VP4-0 [20-34], compared with peptide in the absence of the MAb (black bars). The percentage of inhibition obtained with each MAb is shown. (B-E) Relative responses of cytotoxic CD4-CDS+ Tc cells (B), memory/activated helper CD4+CDS+ Th cells (C), non ThlTc CD4-CDS- cells (D), and naive CD4+CDS- Th cells (E), by measurement ofIL-2 receptor (CD25) expression (dark grey histograms) in cultures stimulated with a BT tandem peptide. Light grey histograms show the CD25 expression in control unstimulated cultures. Adapted from Blanco et al. (2000).

Copyright © 2004 By Horizon Bioscience

\194

Humoral and Cellular Immune Response Against FMDV

VP4

2A

3D

I L/L'I~I VP211 VP311~~;1ID§]0WJ[ill~1_3D_

1

~

20-34(1)

~

I

21-40(2)

~

I 21-35(3)

Figure 11. Amino acid residues on FMDV proteins corresponding to synthetic peptides reported

to induce T helper responses and to cooperate in the induction of neutralizing antibodies to FMDV G-H loop B cell site. (1) van Lierop et ale (1995); Blanco et al. (2000). (2) Collen et ale (1991). (3) Blanco et ale (2001).

et al., 2001; Rowlands, 1994; Sobrino et al., 1999; 2001). In general, the peptide sequences recognised by Th lymphocytes are found in clusters scattered along FMDV proteins. These are found on VPl, but also on other virion capsid proteins as VP4 as well as on non-structural proteins (Figures 9 and 11). Observations that distinctive Th lymphocyte clones will recognise particular peptide sequences explains the results of Collen and co-workers (Collen et aI., 1991) that 37% of vaccinated animals would recognise peptide in contrast to 89% recognising whole virus antigen. The latter would have been processed by the APC, resulting in several peptides capable of stimulating a number of Th lymphocyte-clones - those for which they were specific. Initial work studying the FMDV epitopes recognised by T lymphocytes focussed on VPl, but this has been extended to both the structural and nonstructural proteins of the virus. Several T cell epitopes on the capsid proteins VP1, VP2, VP3 (Collen et al., 1991) and VP4 (van Lierop et al., 1994; 1995b) have been identified in the context of the bovine immune response. Functional T-cell epitopes on VP2, VP3 and VP4 have also been reported using immunised mice (Perez et al., 2000). Lymphocytes from vaccinated pigs were also able to respond to T cell epitopes identifiable on all FMDV capsid proteins, particularly in VPl (Rodriguez et al., 1994) and VP4 (Blanco et al., 2000). One useful approach to identifying T-cell epitopes has employed combination of proposed T-cell epitopes with B-cell epitopes. This permits the induction of antibody as a readout, antibody being of course the important element in the protective immune response against FMDV. Much work has focussed on the G-H loop of capsid protein VP1, due to the presence of a continuous, immunodominant B -cell antigenic site in this region (Bittle et al., 1982; DiMarchi et al., 1986; Pfaff et al., 1982; Taboga et aI., 1997). Peptides synthesised on the basis of the VPl G-H loop sequence were found capable of stimulating T cells from different host species (Glass and Millar, 1995; Rodriguez et al.,1994; Taboga et al., 1997; Zamorano et al., 1994). However, not all individuals, even within the same animal species, responded in the same manner, probably due to epitope recognition by T cells being restricted by the individual MHC composition (Garcia-Briones et aI., 2000; Glass and Millar, 1994; van Lierop et aI., 1995a). Furthermore, the amino acid

Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

195\

sequence at the G-H loop varies considerably among different FMDV serotypes and even subtypes (reviewed in Domingo et al., 2003). This variability can also affect its recognition by T lymphocytes from particular individuals (Glass and Millar, 1995). The potential of FMDV Th epitopes to improve immunogenicity of peptide vaccines was shown by Collen and colleagues (Collen et al., 1991). In these experiments, cattle were immunised with a vaccine constructed from a Th epitope corresponding to VPl residues 21 to 40 (Figure 11) linked in tandem with the VP 1 G-H loop. This resulted in efficient antibody production and protection against virus challenge in cattle whose T lymphocytes did respond to the G-H loop peptide. Unfortunately, in a large-scale trial, efficient lymphocyte recognition of this Th epitope in the tandem peptide vaccine was considerably restricted (Taboga et al., 1997). It was following on such observations that the search began for T cell epitopes in conserved regions of FMDV proteins, efficiently recognised by most T lymphocytes (Blanco et aI., 2000; van Lierop et aI., 1995b). 4.4. Conserved T Cell Epitopes of Value to Vaccine Design

Th lymphocyte epitopes with conserved sequences among different FMDV isolates, and recognized by a wide spectrum of MHC Class II alleles in different host species, have potential for vaccine design. The structural protein VP4 is highly conserved among FMDV serotypes and other Picomaviruses, and possesses an interspecies MHC-restricted Th lymphocyte epitope corresponding to residues 20-34 (Figure 11) (Blanco et al., 2000; van Lierop et al., 1995b). T cell epitopes with conserved sequence, which are frequently recognized by lymphocytes from domestic pigs, have also been identified on the FMDV non-structural proteins 3D (Collen et al., 1998; Garcfa-Briones et al., 2004), 3A, 3B and 3C (Blanco et al., 2001). Such epitopes have the additional advantage of being recognized in a heterotypic manner by T-cells of different individuals. When the Th epitopes in VP4 or 3A (Figure 11) were included in peptides containing the G-H loop, the tandem peptides induced in vitro specific responses in lymphocytes from FMDV immunized animals. These responses were mediated primarily by CD4+ T lymphocytes and B lymphocytes, the stimulation of the latter being identified by an enhanced the production of anti-FMDV neutralizing antibodies (Blanco et al., 2000; 2001). The potential of such Th epitopes to improve immunogenicity of new FMDV vaccines is an ongoing topic of research investigation. Clearly, additional basic work is still required to understand the mechanisms controlling the immunomodulatory potential of FMDV T-cell epitopes. Of particular importance is the functional role of induced cytokines, and ho\v the T cell response contributes to the protective immune responses against FMDV in different host species. 4.5. Conclusions

The above analyses confirmed that CD4+ Th lymphocytes are required for the induction of protective immunity against FMDV. Their role is to promote the differentiation of antigen-activated B-Iymphocytes into antibody-producing plasma cells, and the isotype switching of the antibody response from IgM to the different IgG isotypes. This can also be seen when a peptide antigen containing both a T helper lymphocyte epitope and a B cell epitope was employed to

Copyright © 2004 By Horizon Bioscience

\196

Humoral and Cellular Immune Response Against FMDV

overcome the restriction of the immune response and induce high levels of FMDVspecific antibodies (Collen et al., 1989; 1991; Francis et al., 1987). The duration of the antibody response was less than that in animals immunized with conventional vaccines, which in tum induce duration of immunity shorter than that observed in convalescent animals. Such variations in the duration of immunity may be related to the relative efficiency with which Th memory cells are induced. These memory cells are important to ensure that the recall B-cell response is rapid. In this respect, it has been considered that FMDV non-structural proteins might contribute to the quality of the immune response following vaccination, because they are only produced in the context of a viral infection. However, it is more likely that the higher quality of the immune response following infection is due to the replication of the virus. Infection would prolong the presence of an antigenic threat, and therefore promote an immune response over a protracted period of time. These characteristics are well established for the continued maturation of an immune response towards one of higher avidity and increased efficacy. Nevertheless, the use of live FMDV vaccines is not a consideration (see Chapter 12). Consequently, current research is looking at the contribution of T cell epitopes from non-structural proteins in vaccine formulations. Unlike live virus, such non-viable entities cannot control the duration of their presence for stimulating the immune response. Nevertheless, monocytic cells, in particular the DC, can ensure retention of antigenic material for a prolonged period of time. In order to promote this role of the DC, it is important to achieve the correct interaction with and stimulation of these cells by the vaccine. For this reason, another area gaining increased interest is that focusing on vaccine delivery. While the critical element therein is the promotion of a more efficient immune response, a particular focus is on targeting vaccine delivery to the DC to enhance the rapidity and longevity of the immunity induced.

5. Effector Immune Defences As with the induction of immune responses, the effector mechanisms responsible for immune defences can be compartmentalised. The most rapidly assimilated defences, particularly in non-immune animals, are those of innate immunity. These defences will also ensure the correct and efficient stimulation of the specific or adaptive immune responses (see section 1 of this chapter). With the development of the specific immunity, the effector mechanisms therein take over in dominance from those of the innate defences. Within the specific immune defences, antibody is essential for protection against FMDV. Nevertheless, the destruction and removal of the virus infectious threat also requires involvement of the effector cells from the innate defences. The present section will discuss the current information on how immune defence systems function, as pertinent to protection against FMDV. In sections 6 and 7, the actual immunological processes of protection will be presented in terms of how immune defences tackle the threat posed by FMDV. 5.1. Overview of Innate Immune Defences

Innate defences are activated immediately after infection, more rapidly than the specific responses, but of shorter duration (Carroll and Janeway, 1999; Hoffmann et al., 1999; Medzhitov and Janeway, Jr., 2000b). In contrast to specific (adaptive) immunity, innate immune responses recognise foreign bodies in a non-specific manner, but do not develop "memory". Nevertheless, there is a certain degree Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

..-...

E

~ 0

25

~

c 20 "'C Q) .....,

~

C)

'e --I

30

(b)

15

• ••

•

10 5

n

A

" ,)/y

U

0

100

I

u 50

A \r-;'

25

•

20

• •• U

75

(c)

25

OJ

0-

1971

i'¥. 10

~ R

15 .

,-.

10 '

.&

5

p

• •I ••

I

8

Il

~

"

0

100

.&

.

.&

,

.... ,-~/

75

50

25

10

Concentration of plasma (%) Figure 12. (a) Leukocytes from vaccinated animals showing actin polarisation (arrows) and lamellipodia formation (arrow-heads), relating to FMDV vaccine induced chemotaxis. (b) Migration of peripheral blood leukocytes (PBL) from unvaccinated (PBS-treated; open symbols) and FMDV-vaccinated (filled symbols) pigs, taken 6 days post-vaccination and incubated with varying concentrations (x-axis) of autologous plasma. The PBL levels in the rectangular boxes marked "R" on the x-axis, represent the level of random migration. (c) Migration of PBL from a non-vaccinated, untreated donor animal towards plasma taken at 3 days from unvaccinated (PBStreated; open symbols) and FMDV-vaccinated (filled symbols) pigs. After a 2 h incubation, the amount of PBL that had migrated into the plasma was read. Adapted from Rigden et al. (2003).

of specificity due to the employment of what are termed pattern-associated recognition receptors (Medzhitov and Janeway, Jr., 2000a). The components of innate defences are numerous and variable in terms of their function. For example, the skin and secretions of the mucosal surfaces, as well as the acid and alkali pH in the stomach and intestine respectively, function as physical/chemical barriers in the first line of innate defences (Bals, 2000; Fellermann and Stange, 2001). If the pathogen can evade these barriers and invade the body of the host, the cellular elements of innate immune defences come into operation. The first to be encountered are the histiocytes and the monocytic cells of lymphoid organs/tissues, such as tonsils and mucosal-associated lymphoid tissue (MALT). These include monocytes/ M (Figure 12) neutrophils, dendritic cells, NK cells, mast cells, platelets, endothelial cells, and probably yo-T cells (Biron et al., 1999; Boismenu and Havran, 1997; Caux et al., 2000; Feger et ai., 2002). Copyright © 2004 By Horizon Bioscience

1198

Humoral and Cellular Immune Response Against FMD V

Soluble factors also have important roles to play. These include serum proteins, such as complement, natural antibodies, other opsonins including LPS-binding protein (LBP) and mannan-binding protein, acute phase proteins, chemokines, cytokines (particularly IL-l~, IL-6, IL-I0, IL-12, IL-15, TNF-a, TGF-~, IFNa/p/y), histamine, prostaglandins, and leukotrienes (Agah et al., 2001; Barrington et al., 2001; Biron et al., 1999; Matsushita et al., 1998; Rossi and Zlotnik, 2000; Zlotnick et al., 2000). The cellular defences of innate immune responses rely heavily on the ability of particular cell surface receptors to interact with the pathogen in question, and stimulate them into action (Gallucci and Matzinger, 2001; Wagner and Roth, 2000). Examples of these receptors are the pattern-associated recognition receptors (Beutler, 2000; Fraser et ai., 1998; Medzhitov and Janeway, Jr., 2000c; Stahl and Ezekowitz, 1998; Takeda and Akira, 2001) such as the Toll-like receptor (TLR) family, Fe receptors (FcR),complement receptors, the mannose receptor, CD14 and CD44. Integrins are also important in this respect, but in addition play critical roles in regulating cellular interactions during both innate and specific immune response development. Following vaccination or infection, there will be a degree of local tissue damage, which results in the release of exogenous immune response mediators referred to as "danger signals" (Gallucci and Matzinger, 2001). These interact with histiocytes (macrophages and neutrophils), dendritic cells and mast cells at the site of injury or vaccine deposition, resulting in the production of the endogenous mediators promoting immune defence system recruitment through the local inflammatory immune reaction. An important element therein is the local increased endothelium permeabilisation, concomitant with upregulation of integrin and other adhesion molecule expression on the endothelium. Together with the chemokines released from the endothelium and activated histiocytes, leukocytes are recruited from the blood to adhere to the endothelium possessing the modulated adhesion molecule expression. The increased binding at the endothelium permits extravasation of the recruited leukocytes between the junctions of the endothelium into the site of injury. By such means, leukocytes are recruited to the site of antigen presence to initiate the development of both innate and specific immune responses. Of particular importance therein are the DC, which will transport the antigen to the lymphoid organs and tissues wherein the specific responses will develop. 5.2. The Role of Dendritic Cells DC in situ at the local inflammatory site (dermal DC and Langerhans cells), as well as immature DC migrating from the blood into the inflammatory site, play an important role in linking the innate with the adaptive immune system. DC are also important at the mucosal surfaces, including the MALT and tonsils. They will interact with foreign material such as FMDV transported across or through the epithelial barrier at these sites. In addition, it is not necessary that the virus or vaccine antigen actually cross the mucosal epithelial barrier. Evidence is showing that local DC can open the tight junctions between epithelial cells at mucosal surface barriers, in a manner which retains the integrity of these junctions (Bozza et al., 2002; Rescigno et al., 2001a; Rescigno et al., 2001b). This permits the DC Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

199\

to send a cellular protrusion into the lumen at the mucosal surfaces, "capture" the foreign material, and transport it back into the lymphoid tissue for presentation to the immune system. From this, it can be seen that DC contact with FMDV is essential whether the virus be delivered via the respiratory route or through injection. Endocytosis of the virus by DC is known to occur (see Figure 8a). In vivo, it is probable that the DC, which had endocytosed FMDV, would migrate into lymphoid tissue, via the blood and lymph. Indeed, enhanced trafficking of leukocytes can be seen early after vaccination of pigs against FMDV (Figure 12) (Rigden et aI., 2003). Once in the lymphoid tissues or organs, the DC enter the follicles to present the peptides derived from processing the virus antigen to T lymphocytes, or deliver antigen in or more intact form to the B lymphocytes (see Figure 1) (Banchereau and Steinman, 1998; Banchereau et ai., 2000). By such means, the specific immune responses are induced. 5.3. The Role of the Phagocytes The mononuclear phagocytes - monocytes and M - are central to the acute inflammatory reaction so important in the innate effector immune defences, and are also essential in protection against FMD (McCullough et ai., 1992b). Being actively phagocytic, with a wide distribution throughout the body tissues and organs, they are important immune defence agents for attacking and killing many different types of micro-organism, as are the neutrophils. Their increased involvement following vaccination can be seen with FMDV-vaccinated pigs during the first week post-vaccination (Rigden et ai., 2003). This has been characterised in terms of the chemotactic activity induced by the vaccination. Chemotaxis is the ability of cells to migrate in response to the presence of a concentration gradient of a chemotactic factor, and is a key element in orchestrating selective recruitment of leukocytes to inflammatory sites (Mackay, 2001; Murdoch and Finn, 2000; Nelson and Krensky, 2001). Following vaccination against FMDV, the ability of the vaccine to increase the chemotactic activity of blood phagocytes was an important event early after vaccine administration (Figure 12b and c) (Rigden et aI., 2003).The rapidity and efficacy of the induction of chemotaxis, and the serum chemotactic factors, related to the rapidity with which the emergency vaccine against FMDV would induce protection. Such vaccine-induced modulation of leukocyte chemotaxis was due to the vaccine, and not just to the adjuvant employed. The migratory leukocytes were dominated by neutrophils and monocytes, which is a typical finding for innate defence responses (Murdoch and Finn, 2000). Further evidence of monocytic cell involvement early after vaccination has come from the demonstration of IL-6 activity in the serum (Barnett et ai., 2002). It was also interesting to note that these vaccinates were primed to produce IL-8 as well as the IL-6 following challenge infection. The presence of this chemokine (IL-8) would relate to the increased chemotactic activity after vaccination (see Figure 12) (Rigden et ai., 2003). The importance of phagocyte activity early post-infection or post-vaccination is the rapidity with which these innate defences respond, particularly critical in an emergency situation. M can phagocytose FMDV (see Figure 8b), leading to its destruction (Figure 13). Although the ability of M Cl

:aE u.

80···

5

60

i::

~

g-

~ .E:

~

a o

10

30

60

120

time (min.) with FMDV

0

\

---,---~,----,~~,-o, o

1

2

4

6

8

24

time (h.) with FMDV

Figure 13. The rate of uptake (a) and destruction (b) of antibody-opsonised FMDV by macrophages. In (a), the uptake of virus alone (unopsonized) is shown for comparison. Adapted from Rigden et al. (2002) and McCullough et at. (1988).

FMDV is enhanced when the virus is opsonised with specific antibody (Figure 13) (McCullough et ai., 1988; 1992b; Rigden et al., 2002), the M would be important in the early immune defence against FMDV. The host is more reliant on innate defences during the first days after vaccination. At this time, specific responses including antibody would not be at levels considered as protective. Yet, animals can be protected by emergency FMDV vaccines within 2-4 days (Cox et al., 1999; Doel et al., 1994; Salt et al., 1998). Natural antibodies and other opsonins could be playing a role due to their ability to increase the phagocytic and anti-pathogen activities of neutrophils (Konishi and Nakao, 1992), and enhance delivery of pathogens into secondary lymphoid organs (Ochsenbein et al., 1999). Analyses on natural antibody activity early post-vaccination could not identify a vaccinedependent enhancement of such factors interacting with FMDV (Rigden et al., 2003). However, natural antibodies in sera can be masked by binding to circulating antigens, which could impede their interaction with the antigen under study, such as FMDV. Nevertheless, it is unlikely that natural antibodies would be major elements in immune defences against FMDV. In the absence of specific antibodies - as induced by vaccination - animals remain fully susceptible to virus infection and disease production. It is likely that the observed early protection is multifactorial. One element is enhanced chemotaxis of phagocytes, and therefore enhanced local phagocytic activity - recruited phagocytes are more active than resident phagocytes. These immune defences would be involved in the first line of host defence against the virus, but cannot be regarded separately from the specific immune defences. If animals were not to continue with the development of the specific immunity following this enhanced innate immunity, no protection against FMDV would be seen. Certainly, the elevated innate defences could delay the onset of disease symptoms, giving the specific immune defences the necessary time to develop a more durable protection.

Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

(a) SNT

2011

(b) ELISA

640 o o

~

80-

"'C

.~

o

..c :t::::

.co coo

C

m

~

e 0-

·0

white zone

coo

~ ......,

:t::::

o o

grey zone

10

~

black zone

1.7

Figure 14. The Grey Zone of antibody tires with respect to protection against challenge infection by FMDV. Protected animals are shown with white circles, and unprotected with the black circles. The white zone identifies antibody titres more often associated with protected animals, the black zone antibody titres associated with susceptibility to infection, and the grey zone antibody titres which do not relate to protection. Adapted from McCullough et ai. (1992).

5.4. The Specific Effector Anti-FMDV Immune Response: The Role of Antibody

The immune response of major importance in providing durable protection against FMDV is that mounted by humoral immune defences, particularly the production of specific anti-FMDV antibodies. However, animals can possess similar titres of antibody titre, but only some resist challenge whereas others do not (McCullough et aI., 1992a; van Maanen and Terpstra, 1989). In reality, a "grey zone" exists for the serological estimation of antibody titres (Figure 14). Within this grey zone, it is not possible to distinguish with certainty the animals which will be protected from those remaining susceptible to FMDV challenge infection (McCullough et al., 1992a). The main reason for such inconsistencies lies with the attempts to relate antibody titres measured by in vitro tests to in vivo protection against challenge infection. In vitro tests - including the serum neutralisation test (SNT) and ELISA - need only detect antibody interacting with the virus, whereas the protective immune response must destroy virus infectivity. Although the SNT will measure the capacity of antibody to block virus infection, this is only infection of cells in culture, and does not guarantee that the virus has been destroyed. For protection in vivo, it is essential that the reticuloendothelial compartment of the immune system interacts to promote destruction of FMDV infectivity through phagocytosis of the virus/antibody complexes (McCullough et al., 1986; 1988). Copyright © 2004 By Horizon Bioscience

\202 Humoral and Cellular Immune Response Against FMDV

When antibody binds to an antigen, there are conformational alterations induced in the Fc portion of the antibody, allowing it to interact more efficiently with the low affinity FcR on the phagocyte surface. Antibody isotype can influence the outcome of interaction with FMDV, because different immunoglobulin isotypes have different capacities to interact with phagocytes. For this reason, Mulcahy and co-workers (Mulcahy et al., 1990) suggested that the induction of different isotypes of anti-FMDV antibody in cattle might be crucial to the outcome of a particular vaccination. Antibody isotypes can also influence immune responses due to their variable ability to fix complement. The complement cascade is activated by conformational alterations occurring in the antibody Fc following binding to antigen, resulting in the deposition of certain activated products of the cascade on to the antigen. Amongst these are C3b and C4b derived from the activation of C3 and C4 components of the cascade. Phagocytes possess receptors for these C3b and C4b, referred to as CRl or CD35. Similar to the role played by the FcR, CRl will be activated by interacting with the C3b or C4b, and promote phagocytosis. Certain activated components of the complement cascade will also influence leukocyte chemotaxis. Consequently, anti-FMDV antibody can promote destruction of the virus infectious threat by different means. Certainly, the removal of that threat requires specific antibody, but completion of the process has a prerequisite for phagocytic involvement. In fact, Immunological protection requires a sequence of immune reactions: antibody-virus complexing with or without complement involvement (opsonisation), enhancement of M phagocytosis of the virus via interaction of the complexes with the cell FcR or CRl, the consequential signalling of the M to intemalise the complexes, and promotion of endosomal degradation of the virus. The next section will deal in detail with this activity of the Mel> and its role in immune protection against FMDV.

6. Macrophages in Effector Immune Responses 6.1. Macrophages and the Phagocytosis of FMDV

Within innate defences, the Mel> can interact directly with FMDV in the absence of specific antibody, and present an important element of the early immune defence following emergency vaccination (Barnett et ai., 2002; Rigden et ai., 2003). M will phagocytose FMDV (see Figure 8b), with a time-dependent loss of detectable virus and virus infectivity (Figures 13 and 15) (McCullough et al., 1988; Rigden et al., 2002). Although this has a rather protracted time course, Mel> phagocytosis of FMDV in the absence of specific antibody will destroy the virus - an important early innate defensive response. This M activity can be enhanced, as has been noted early post-vaccination dependent on the vaccine formulation employed (Rigden et al., 2003). By such means, the capacity of the Mel> to interfere with the FMDV disease process will be promoted. Nevertheless, the efficiency of this process remains lower in the naIve animal compared to the immune animal. This difference in efficacy can be attributed to the influence of specific antibody opsonising the virus particles and enhancing their uptake by the phagocytes (see Figure 13). With emergency vaccination, the protection arising within 4 days postvaccination probably involves enhanced Mel> activity holding the virus advance at bay until specific antibody develops. Even though the M can destroy FMDV, Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

(a) FMDV with Mq. (o-10hr)

2031

(b) FMDV with M+ (24-48hr)

100 - - - - - - - - - - . . 0 0 Neg.

80 60

10\ ..

30min

Neg.

80

4hr¥1 I

i.

i

,

~1

140 t. :-

100000

~

1000~

Q.

.~

10

-----w-:

11 82

09

C9

1 ~

:1

1 82

C8

09

(c) in vivo protection after Fe removal

e >-

"8 -C

~ co

~

e

Q.

'0

C8

(d) silicon impairs in vivo protection

100000; ~

C9 antibody

antibody

10000~

1000~

~

100001

"8 -c

1000~i

>-

1I

100

~

co

100~

~

e

10

Q.

'0

~

100 1O~

~

1-1 82

09

C9 antibody

C8

1 82

09

C9

C8

antibody

Figure 18. The protective immune defence against FMDV requires M to remove the antibody/virus complexes.. (a) Antibodies against different epitopes important to virus infectivity protect against disease, even when the antibody has not neutralised the virus. (b) Removal of the Fc portion of antibody complexed with FMDV does not influence in vitro neutralisation of virus infectivity. (c) In contrast, rremoval of the Fc portion of antibody complexed with FMDV prevents their uptake via the FcR on M4l, and thus protection against disease. Cd) Impairment of M4l function using silicon dioxide treatment of the animals prevents antibody from protecting against disease. Adapted from McCullough et ale (1986).

Copyright © 2004 By Horizon Bioscience

1208

Humoral and Cellular Immune Response Against FMDV

Analysis of protection in vivo - using a murine model system - also demonstrated that protection could occur if the virus complexed with antibody was still infectious for susceptible cell lines in vitro (Figure 18a). That is, the antibody could protect in vivo at concentrations which did not neutralize virus infectivity in vitro (McCullough et al., 1986). Due the relationship to the in vitro observations that M could destroy antibody-complexed FMDV which was infectious for susceptible cell lines, the role of the M in vivo was examined. Removal of the Fe portion of the antibodies complexed with the FMDV did not influence in vitro neutralisation (Figure 18b), but did impart the protective defence in vivo (Figure 18c). Furthermore, when phagocytosis in vivo was blocked, the protective capacity of the antibody in the complexes with the virus was reduced or even abrogated (Figure 18d) (McCullough et al., 1986). Consequently, the direct neutralization by antibody of FMDV infectivity for susceptible cells is not of primary element in protection against disease. Certainly its ability to interfere with the spread of the virus is an important constituent in the protective defence make-up. Efficient phagocytosis and FeR-mediated uptake of antibody/FMDV complexes are the critical parameters for the protective immune response. In the absence of antibody, phagocytosis and destruction of the virus would still occur, but with reduced efficiency (Rigden et al., 2002) It is the presence of specific anti-FMDV antibody which ensures efficient McI>dependent destruction of the virus, and protection against disease. The success of M interaction with and destruction of FMDV is paramount in determining the outcome of the infection - disease or protection.

7. Cytotoxic Effector Immune Responses CTL responses are likely to play a relevant role in the early stages of infection of epithelial cells by FMDV. However, little is known on the contribution to the protection of these responses in FMDV immunized animals (Becker, 1993; Sobrino et al., 2002). Isolation of cells from NK lineage, with a non MHC-restricted FMDVspecific cytolytic activity, has been reported upon FMDV vaccination (Amadori et al., 1992); however the contribution of this NK activity to viral clearance is not yet understood. The induction of FMDV-specific effector cytotoxic T lymphocytes (CTL) has been difficult to evaluate (Childerstone et al., 1999; Rodriguez et ai., 1996; SanzParra et al., 1998a). The complexity of these studies is worsened by the problems inherent to the particular characteristics of the immune cellular responses in pigs and other animals susceptible to FMDV (Saalmuller, 1998). In addition, FMDV infection results in a rapid reduction of MHC class I expression on susceptible cells (Sanz-Parra et ai., 1998b), which could impair presentation of viral peptides by FMDV infected cells to CTLs, facilitating virus escape from this particular antiviral response of the host. However, and interestingly, monoclonal antibodies not only against CD4+ and MHC class II molecules, but also anti CD8+ or anti MHC Class I molecules, to a lesser extent, can inhibit proliferation of immune pig lymphocytes to viral particles (Childerstone et al., 1999; Saiz et al., 1992) and to synthetic peptides (Blanco et al., 2000; 2001). These observations suggest a role of CD8+ lymphocytes in the T cell response elicited by non-replicating FMDV immunogens. Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

209 I

Indirect evidence in favour of the involvement of T cells in FMDV protection come from results obtained with vaccines based on replicating eucaryotic vectors. In animals immunized with adenovirus recombinants expressing FMDV capsid proteins, the induction of protective immunity against FMDV can be observed in the absence, or associated to very low titres of anti-FMDV neutralizing antibodies (Beard et al., 1999; Sanz-Parra et ai., 1999a,b). This is of interest, because it indicates the potential protective role of the cellular immunity, in contrast to the general assumption that protection against FMDV infection is solely afforded by neutralizing antibodies (Sobrino et al., 2002). However, the precise contribution of CD8+ lymphocytes and the specific CTL response in the immunological control of FMDV infections is still unclear. Elucidation of these mechanisms would help understand further when and how the different branches of the immune response elicited by FMDV would interact towards conferring solid protection against infection. A better understanding of the cell-mediated immunity to FMDV would provide useful information to assist with the design of new vaccines. 8. Influence of FMDV on Immune Defence Compartments FMDV is not regarded as an immunomodulating virus. It is distinct from monocytotropic viruses such as classical swine fever virus, which can infect or replicate in the cells involved with the development and control of immune responses. A transient lymphopenia can be observed early after FMDV infection (Bautista et al., 2003), but this is to be expected. These authors also talked about a reduced T lymphocyte proliferation, but this is also perfectly normal under conditions of infection. FMDV can infect locally and cause cell damage due to that infection. The host immune defences will respond to both the infection and any cell damage. Firstly, there will be the response of the local histiocytes, attempting to control the infection and remove potentially harmful material. These in tum initiate a local inflammatory reaction, which promotes recruitment of more monocytic and granulocytic cells from the blood. As the inflammatory response progresses, a recruitment of lymphocytes from the blood is seen. These cells are essential to promote further the local defence against the infection. Importantly, the lymphocytes also control the inflammatory response, especially when the threat has been removed and the inflammatory reaction has to be downregulated to a resting state again. Of course, if one looks for lymphocytes in the blood at this time, there will be an apparent lymphopenia. The blood cells will also be less responsive, because there are fewer of them and they are in the process of migrating. Overall, the more prolonged the infection and/or the greater the cell damage by the local virus infection, the stronger and more prolonged will be the local inflammatory response and the migration of leukocytes from the blood. This effect should not be confused with an immunopathogenic situation as occurs with monocytotropic viruses such as classical swine fever virus. A transient migration of lymphocytes from the blood is a perfectly normal response to anything foreign, or to cell damage. It is not the same as the immunomodulatory effects of monocytotropic viruses which not only recruit leukocytes but can damage or destroy them. In addition to the virus infection promoting the above inflammatory response and recruitment of immune defences, FMDV is an antigenic substance detectable by the immune system as foreign. Thus, FMDV will interact with components of Copyright © 2004 By Horizon Bioscience

\210 Humoral and Cellular Immune Response Against FMDV

Figure 19. Confocal microscopy image of FMDV virion antigen (green) in pulmonary macrophages, detected with 4C9 antibody at 6 hr post-infection with FMDV 01 Lausanne.

Actin microfilaments were stained red using Alexa 546 phalloidin. The main picture is an x-y image through a single plane of the cell. The small pictures to the left, right and belo\v the main picture are z-stack sectional views of the left-hand and right-hand cells shown in the main picture, to demonstrate that the virus antigen has been internalised (on the cytoplasmic side of the microfilament cytoskeleton. Adapted from Rigden et al. (2002). See Colour Plate at the back of the book.

the immune system towards promoting an immune response. Such basic interaction between antigen and the immune system towards immune defence development is an immunomodulatory action. Yet, it now appears that FMDV may go beyond this basic interaction with the compartments of the immune system, and that response to FMDV infection may not be a simple case of immune defence recruitment to the site of that infeciton. 8.1. FMDV Interaction With Cells of the Innate Immune Defences

FMDV possesses an RGD sequence associated with its receptor for cells (Baxt and Becker, 1990; Fox et al., 1989; reviewed in Chapter 7) which provides the virus with the potential to interact with the integrins expressed on the M. Considering the importance of Mel> interaction with FMDV in immunological defence (McCullough et al., 1986; 1992b) (see Figures 17 and 18), this potential of the virus to bind to cell integrins is important. Both M and DC endocytose FMDV (see Figure 8). With M, this is most efficient when antibody/virus complexes are phagocytosed, leading to rapid

Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

2111

destruction of the virus (see Figure 13) (McCullough et al., 1988). When antibody is not involved, the endocytosed virus antigen remains associated with the M and DC for a prolonged period (Figure 15 a and b) (Rigden et ai., 2002). Importantly, much of this virus was still infectious, even the internalised virus (Figure 15c and d). It was not until after 10h that loss of infectivity was noted. Yet, the virus was not replicating, because neither viral RNA replication nor protein synthesis were detectable (Baxt and Mason, 1995; Rigden et ai., 2002). Although much of the infectious virus remained associated with the cells, infectious material was clearly released. An infectious centre assay determined that the cells were releasing at least some of this virus. Treatment of the cells with pronase and pH manipulations demonstrated that within the first few hours after interaction with the FMDV, most of the infectious virus was associated with the cell surface, from where some was released - "leaching" from the cell. At later times, the pronase resistance and pH-resistance of the virus demonstrated that is was internalised, an observation confirmed by confocal microscopy (Figure 19). The continued release of infectious virus demonstrated that this was from within the cells. This is possible, because following endocytosis, exocytosis of vesicles containing infectious virus could occur. This is a normal cellular intracytoplasmic re-cycling process. If the FMDV were present in a recycling structure, this would be responsible for the release of internalised infectious FMDV from the M. 8.2. FMDV Survival Within Cells of the Innate Immune Defences

This survival of FMDV after internalisation by M and DC is intriguing. If the internalisation were clathrin-dependent endocytosis, one would expect the vesicle containing the FMDV to be acidifying rapidly due to the H+-ATPase activity of the endosomes. Early endosomes acidify within minutes, and is probably the process active in susceptible cells such as BHK-21 cells and mucosal epithelial cells of the upper respiratory tract in which the virus replicates. Therein, clathrin-dependent endocytosis relates to the rapid uncoating of the virus leading to initiation of its replicative cycle. In contrast, the observed survival of FMDV infectivity in the M for hours (see Figure 15) (Rigden et al., 2002) is more akin to the virus being internalised by macropinocytosis. Considering that the virus could be interacting with integrins on the cell surface, it is interesting that integrin-mediated endocytic processes do require some time for acidification (Aderem and Underhill, 1999; Dh and Swanson, 1996). An alternative explanation is that the M and DC may employ phagocytosis involving the endoplasmic reticulum - the so-called ER-phagocytosis (Desjardins, 2003). Macropinosomes, phagosomes, and ERphagosomes are not acidic, their acidification requiring fusion with endosomal vesicles (Aderem and Underhill, 1999). Although macropinocytosis can result in relatively rapid acidification (Cardelli, 2001), such processes can take hours to acidify (Aderem and Underhill, 1999; Dh and Swanson, 1996), as seen following the uptake of collagen (Aderem and Underhill, 1999), and ovalbumin (Trombetta et al., 2003). The acidification processes can also be inhibited by certain pathogens, primarily with bacteria. Due to the absence of evidence showing an inhibition of endosomal acidification by FMDV, it would appear that the prolonged survival of the virus in M and DC reflects a process such as macropinocytosis or ER-phagocytosis. Copyright © 2004 By Horizon Bioscience

1212 Humoral and Cellular Immune Response Against FMDV

...

300

E ::» c

240~

.!

1

==8 180

!

J

120 80 0

100

10' II(

(c) FMDV inhibition of M protein synthesis

5000

t... x

cMock Mock+ACT:.D Uvevlrus • Live vlrus+ACT.D 13 InacL virus

D g

4000

E ci 3000 &

eI

InacL vtrus+ACT.D

2000

1000

O..........1IiIiIIiII1iiiiiii 2

24 TIme postInfection (hr)

Figure 20. (a, b) FMDV uptake by pulmonary macrophages treated with the 10JlM cytochalasin D Ih before (bold line), or 1 hr after (thin line) interaction with the virus, compared with the image in untreated cells (filled histogram) at 6 hr pj. Virus was detected within the cells (a) or on the cell surface (b) using the anti-FMDV monoclonal antibody 4C9. (c) De novo cell protein synthesis by pulmonary macrophages, with actinomycin D treatment (+ ACT. D) or without, following interaction with live FMDV (LIVE VIRUS), ultraviolet (UV) light-inactivated FMDV (INACT. VIRUS) or mock antigen (MOCK) for 30 min at 39°C. Protein synthesis was detected by [35S]methionine-Iabelling for 2h or 24h. Adapted from Rigden et al. (2002).

Indeed, M intemalisation of FMDV is dependent on a functional microfilament cytoskeleton (Figure 20a and b) (Rigden et al., 2002). Although this points to macropinocytosis or ER-phagocytosis, the duration of infectious FMDV survival in the M cI> and DC is still rather long. This implies an impaired vacuolar acidification. One target would be the interaction ("docking") of vesicles containing the FMDV with endosomes. It has been noted that FMDV will impair M protein synthesis (Figure 20c) (Rigden et al., 2002), and the degradation of albumin (McCullough and Gerber, unpublished data). Impaired protein synthesis would interfere with Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

2131

production proteins such as syntaxin-6 used as ligands for the EEA-1 of early endosomes to dock with vesicles containing FMDV (or anything else such as albumin present in the "infected" M or DC). 8.3. Consequences of FMDV Interaction With Cells of Innate Immune Defences

Although M and DC do not support replication of FMDV, the prolonged presence of infectious virus associated with these cells would signify that M and DC may function as infectious carriers of FMDV. Although their degradative processes would ultimately dominate, by the time this was complete the cells carrying the infectious FMDV \vould have trafficked through the body delivering their infectious cargo. This is particularly important considering that DC have a homing potential for mucosal surfaces, such as those of the upper respiratory tract where FMDV is known to replicate. Of course, this is a major problem in naive animals. With an immune animal, the endocytosis of the virus by M is enhanced by opsonisation with specific antibody, leading to activation of the M and rapid destruction of virus infectivity. Under these conditions, the virus would not pose the same infectious threat. As for the DC, the prolonged presence of virus can be an advantage. The antigen processing event would be protracted, and therefore the duration of antigen presentation to T lymphocytes (see Figure 1). A similar advantage would be accessible to the B lymphocytes, but more dependent on a lack of antigen degradation due to the capacity of DC deliver unprocessed antigen to the B lymphocytes (see Figure 1). A problem only arises if the delivered antigen is infectious virus, and the DC have homed to the mucosal surfaces of the upper respiratory tract. Then the risk emerges of the delivered virus infecting susceptible epithelial cells. This would not be a problem if the DC delivered virus to the follicular areas of lymphoid organs and tissues, where it would interact with B-Iymphocytes for the induction of specific antibody. In a vaccinated animal, specific antibody would certainly control any risk due to virus released from DC. However, this \vould be effective only in a systemic manner, and not locally at mucosal surfaces, due to the lack of mucosal immunity induced by current vaccination protocols (inactivated vaccine administered parenterally). When DC would deliver internalised infectious FMDV to mucosal surfaces, there is an increased risk that the virus would be released to infect susceptible cells in this area. Of course the problem really starts with the entry of the airborne virus into the oropharyngeal region. The lack of mucosal immunity in the vaccinates would allow infection of the mucosal surfaces by the airborne virus without recourse to transport by the DC. Indeed, infection of the soft palate has been noted in vaccinated ruminants (Alexandersen et a!., 2002; 2003). Consequently, the induction of specific immune responses in the upper respiratory tract is central for a protective immune defence excluding the infectious threat from FMDV. Without an effective immune response at the mucosal areas of the upper respiratory tract, the host may not prevent a local primary infection by FMDV, even though any systemic immunity present would prevent blood-borne viraemia and induction of disease symptoms. Due to the observation that such infections in the soft palate can be non-cytopathic without an apparent immune response induction, there would be

Copyright © 2004 By Horizon Bioscience

\214 Humoral and Cellular Immune Response Against FMDV

no immediate clearance of this primary infection in vaccinates. That is, vaccinates without mucosal immunity could permit the FMDV to establish a persistent infection, thus becoming carrier animals. The only means by which FMDV can be excluded from a host is ensuring that both effective systemic immunity and effective mucosal immunity have been induced by vaccination. This requires targeting of mucosal sites with the vaccine. Although this can be achieved by oronasal vaccination, such a method is not the most practical under field conditions. Fortunately, there is some evidence that appropriate modulation of vaccine delivery systems can permit the induction of mucosal immunity after parenteral vaccination. This theme has become a particularly important topic of research, which is being pursued not only with respect to FMDV, but also with respect to vaccine delivery in general. Once again, the critical target with respect to successful vaccination is the DC, coming back to the theme which has been running through this review - the importance of DC in the generation of all elements in immune responses against FMDV. 9. Conclusions The induction of long lasting and rapid protective immunity is the primary aim for successful vaccination against infectious diseases. With respect to FMDV, the major arm of the immune defences involved in protection is that based on specific antibody together with the increased efficiency of macrophage phagocytosis and destruction of virus/antibody complexes. For the induction of effective immune defences against FMDV, the critical player therein is the dendritic cell (DC). Des are the key controllers required for antigen presentation to T lymphocytes and antigen delivery to B-Iymphocytes, resulting in stimulation of the antigen-specific immune responses. Adjuvants play an important role in this process, particularly through induction of potent local cytokines and chemokines, which regulate the trafficking of Des. Considering vaccination, current vaccines are certainly efficient at interacting with DCs and initiating effective immune responses and defences against FMDV. Considerable detail is available on the actual epitopes on the virus involved in the antigen processing events for T lymphocytes, and the antigen stimulation events for promoting the B lymphocytes and antibody-based immunity. More recent evidence is sho\ving how FMDV interacts with the cells of the immune system, particularly the macrophages and DCs. This latter area is critical for furthering our understanding of how vaccines can be employed to modulate Des, promoting the most efficient immune defence possible. In this context, current vaccine formulations and vaccination protocols promote systemic immunity, but with a lack of mucosal immunity. This runs the risk of permitting airborne virus to establish a local infection in the dorsal soft palate, even with vaccinates. Generating mucosal immunity could circumvent such a threat. Although oranasal immunisation is the classical method for such induction, it is not always a practical solution with livestock. Recent evidence has demonstrated how to generate mucosal immunity following parenteral immunisation, and is a major area of current research in FMD immunology. The focus of this work is the DC, these cells being the central players for the successful generation of protective immune defences against FMDV. Further studies on the DC, particularly with

Copyright © 2004 By Horizon Bioscience

McCullough and Sobrino

2151

respect to their targeting through vaccination and homing to mucosal surfaces, are major themes currently under scrutiny in the furtherance of studies on the immunology of FMD. 10. Acknowledgements The authors wish to acknowledge the productive collaborations and discussions held with many colleagues, from which some of their scientific contributions are cited in this chapter. Work supported by ED grants QLK 2-CT-2002-00825 and -1304, and BI02002-04091-C03-02 from CICYT (Spain). References Aderem, A., and Underhill, D.M. 1999. Mechanisms of phagocytosis in macrophages. Annu. Rev. Immunol. 17: 593-623. Adorini, L. 2003. Tolerogenic dendritic cells induced by vitamin D receptor ligands enhance regulatory T cells inhibiting autoimmune diabetes. Ann.N.Y.Acad. Sci. 987: 258-261. Agah, A., Montalto, M.C., Young, K., and Stahl, G.L. 2001. Isolation, cloning and functional characterization of porcine mannose-binding lectin. Immunology. 102: 338-343. Alexandersen, S., Zhang, Z., and Donaldson, A.I. 2002. Aspects of the persistence of footand-mouth disease virus in animals--the carrier problem. Microbes. Infect. 4: 10991110. Alexandersen, S., Zhang, Z., Donaldson, A.I., and Garland, A.1. 2003. The Pathogenesis and Diagnosis of Foot-and-Mouth Disease. J. Compo Pathol. 129: 1-36. Amadori, M., Archetti, I.L., Verardi, R., and Berneri, C. 1992. Isolation of mononuclear cytotoxic cells from cattle vaccinated against foot-and-mouth disease. Arch.Virol. 122: 293-306. Bals, R. 2000. Epithelial antimicrobial peptides in host defense against infection. Respir. Res. 1: 141-150. Banchereau, J., Briere, F., Caux, C., Davoust, J., Lebecque, S., Liu, Y.1., Pulendran, B., and Palucka, K. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18: 767811. Banchereau, J., and Steinman, R.M. 1998. Dendritic cells and the control of immunity. Nature 392: 245-252. Barnard, A., Vincent, I., Piersma, S., Takamatsu, H., Barnett, P.V., Denyer, M., Summerfield, A., and McCullough, K.C. 2004. Myeloid cell regulation of antigen-specific lymphoproliferation. (submitted for publication). Barnard, A., Arriens, A., Cox, S. J., Barnett, P. V., Kristensen, B., Summerfield, A., and McCullough, K. C. 2004a. Immune response characteristics following emergency vaccination of pigs against foot-and-mouth disease. (submitted for publication). Barnett, P.V., Cox, S.1., Aggarwal, N., Gerber, H., and McCullough, K.C. 2002. Further studies on the early protective responses of pigs following immunisation with high potency foot and mouth disease vaccine. Vaccine. 20: 3197-3208.

Barrington, R., Zhang, M., Fischer, M., and Carroll, M.C. 2001. The role of complement in inflammation and adaptive immunity. Immunol. Rev. 180: 5-15. Basta, S., Carrasco, C.P., Knoetig, S.M., Rigden, R.C., Gerber, H., Summerfield, A., and McCullough, K.C. 2000. Porcine alveolar macrophages: poor accessory or effective suppressor cells for T-lymphocytes. Vet. Immunol. Immunopathol. 77: 177-190. Basta, S., Knoetig, S.M., Spagnuolo-Weaver, M., Allan, G., and McCullough, K.C. 1999. Modulation of monocytic cell activity and virus susceptibility during differentiation into macrophages. J. Immunol. 162: 3961-3969. Bautista, E.M., Ferman, G.S., and Golde, W.T. 2003. Induction of lymphopenia and inhibition of T cell function during acute infection of swine with foot and mouth disease virus (FMDV). Vet Immunol. Immunopathol. 92: 61-73. Copyright © 2004 By Horizon Bioscience

\216 Humoral and Cellular Immune Response Against FMDV Baxt, B., and Becker, Y. 1990. The effect of peptides containing the arginine-glycineaspartic acid sequence on the adsorption of foot-and-mouth disease virus to tissue culture cells. Virus Genes. 4: 73-83. Baxt, B., and Mason, P. W. 1995. Foot-and-mouth disease virus undergoes restricted replication in macrophage cell cultures following Fc receptor-mediated adsorption. Virology. 207: 503-509. Beard, C., Ward, G., Rieder, E., Chinsangaram, J., Grubman, M.J., and Mason, P.W. 1999. Development of DNA vaccines for foot-and-mouth disease, evaluation of vaccines encoding replicating and non-replicating nucleic acids in swine. J. Biotechnol. 73: 243-249. Becker, Y. 1993. The need for cellular and humoral immune responses in bovines to ensure protection from foot-and-mouth disease virus (FMDV)- a point of view. Virus Genes. 8: 199-214. Beutler, B. 2000. Tlr4: central component of the sole mammalian LPS sensor. Curro Opin. Immunol. 12: 20-26. Biron, C.A., Nguyen, K.B., Pien, G.C., Cousens, L.P., and Salazar-Mather, T.P. 1999. Natural killer cells in antiviral defense: function and regulation by innate cytokines. Annu. Rev. Immunol. 17:189-220. Bittle, J.L., Houghten, R.A., Alexander, H., Shinnick, T.M., Sutcliffe, J.G., Lerner, R.A., Rowlands, D.J., and Brown, F. 1982. Protection against foot-and-mouth disease by immunization with a chemically synthesized peptide predicted from the viral nucleotide sequence. Nature. 298: 30-33. Blanco, E., GarcIa-Briones, M., Sanz-Parra, A., Gomes, P., De Oliveira, E., Valero, M.L., Andreu, D., Ley, V., and Sobrino, F. 2001. Identification of T-cell epitopes in nonstructural proteins of foot-and-mouth disease virus. J. Virol. 75: 3164-3174. Blanco, E., McCullough, K., Sumnlerfield, A., Fiorini, J., Andreu, D., Chiva, C., Borras, E., Barnett, P., and Sobrino, F. 2000. Interspecies major histocompatibility complexrestricted Th cell epitope on foot-and-mouth disease virus capsid protein VP4. J.Virol. 74: 4902-4907. Boismenu, R., and Havran, W.L. 1997. An innate view of gamma delta T cells. Curr.Opin.Immunol. 9: 57-63. Borca, M.V., Fernandez, F.M., Sadir, A.M., Braun, M., and Schudel, A.A. 1986. Immune response to foot-and-mouth disease virus in a murine experimental model: effective thymus-independent primary and secondary reaction. Immunology. 59: 261-267. Bozza, S., Gaziano, R., Spreca, A., Bacci, A., Montagnoli, C., di Francesco, P., and Romani, L. 2002. Dendritic cells transport conidia and hyphae of Aspergillus fumigatus from the airways to the draining lymph nodes and initiate disparate Th responses to the fungus. 1. Immunoll68: 1362-1371. Brown, F. 1992. Ne\\' approaches to vaccination against foot-and-mouth disease [see comments]. Vaccine. 10: 1022-1026. Brown, F. 1995. Antibody recognition and neutralization of foot-and-mouth disease virus. Semin. Virol. 6: 243-248. Cardelli, J. 2001. Phagocytosis and macropinocytosis in Dictyostelium: phosphoinositidebased processes, biochemically distinct. Traffic. 2: 311-320. Carrasco, C.P., Rigden, R.C., Schaffner, R., Gerber, H., Neuhaus, V., Inumaru, S., Takamatsu, H., Bertoni, G., McCullough, K.C., and Summerfield, A. 2001. Porcine dendritic cells generated in vitro: morphological, phenotypic and functional properties. Immunology. 104: 175-184. Carroll, M.C. and Janeway, C.A. 1999. Innate immunity. Curr.Opin.Immunol. 11: 11-12. Caux, C., Ait-Yahia, S., Chemin, K., de Bouteiller, 0., Dieu-Nosjean, M.C., Homey, B., Massacrier, C., Vanbervliet, B., Zlotnik, A., and Vicari, A. 2000. Dendritic cell biology and regulation of dendritic cell trafficking by chemokines. Springer Semin. Immunopathol. 22: 345-369. Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

2171

Childerstone, A.1., Cedillo-Baron, L., Foster-Cuevas, M., and Parkhouse, M.E. 1999. Demonstration of bovine CD8+ T-cell responses to foot-and-mouth disease virus. J. Gen. Virol. 80: 663-669. Collen, T. 1994. Foot-and-mouth disease (aphthovirus): viral Tcell epitopes. In: Cell Mediated Immunity in Ruminants. B.M.L. Goddeevis and I. Morrison, eds. CRC Press Inc, Boca Raton. p. 173-197. Collen, T., Baron, J., Childerstone, A., Corteyn, A., Doel, T. R., Flint, M., Garcia-Valcarcel, M., Parkhouse, M. E., and Ryan, M.D. 1998. Heterotypic recognition of recombinant FMDV proteins by bovine T-cells: the polymerase (P3Dpol) as an immunodominant T-cell immunogen. Virus Res. 56: 125-133. Collen, T., Dimarchi, R., and Doel, T.R. 1991. A T cell epitope in VPl of foot-and-mouth disease virus is immunodominant for vaccinated cattle. J. Immunol. 146: 749-755. Collen, T. and Doel, T.R. 1990. Heterotypic recognition of foot-and-mouth disease virus by cattle lymphocytes. J. Gen. Virol. 71: 309-315. Collen, T., McCullough, K.C., and Doel, T.R. 1984. Induction of antibody to foot-andmouth disease virus in presensitized mouse spleen cell cultures. J. Virol. 52: 650-655. Collen, T., Pullen, L., and Doel, T.R. 1989. T cell-dependent induction of antibody against foot-and-mouth disease virus in a mouse model. J. Gen. Virol. 70: 395-403. Cox, S.1., Barnett, P.V., Dani, P., and Salt, J.S. 1999. Emergency vaccination of sheep against foot-and-mouth disease: protection against disease and reduction in contact transmission. Vaccine. 17: 1858-1868. Desjardins, M. 2003. ER-mediated phagocytosis: a new membrane for new functions. Nat. Rev. Immunol. 3: 280-291. Desjardins, M., Huber, L.A., Parton, R.G., and Griffiths, G. 1994. Biogenesis of phagolysosomes proceeds through a sequential series of interactions with the endocytic apparatus. J. Cell. BioI. 124: 677-688. DiMarchi, R., Brooke, G., Gale, C., Cracknell, V., Doel, T., and Mowat, N. 1986. Protection of cattle against foot-and-mouth disease by a synthetic peptide. Science. 232: 639-641. Doel, T.R., Williams, L., and Barnett, P.V. 1994. Emergency vaccination against foot-andmouth disease: rate of development of immunity and its implications for the carrier state. Vaccine. 12: 592-600. Domingo, E., Escarmis, C., Baranowski, E., Ruiz-Jarabo, C.M., Carrillo, E., Nunez, J.I., and Sobrino, F. 2003. Evolution of foot-and-mouth disease virus. Virus Res. 91: 47-63. Domingo, E., Escarmis, C., Martinez, M.A., Martinez-Salas, E. and Mateu, M.G. 1992. Foot-and-mouth disease virus populations are quasispecies. In: Current Topics in Microbiology and Immunology. J.J. Holland, ed. Springer Verlag. Vol. 176. p. 33-47. Domingo, E., Mateu, M.G., Martinez, M.A., Dopazo, J., Moya, A., and Sobrino, F. 1990. Genetic variability and antigenic diversity of foot-and-mouth disease virus. In: Applied Virology Research, Vol 2. Virus Variation and Vpidemiolgy. E. Kurstak, R. G. Marusyk, S. A. Murphy, and M. H. V. Van-Regenmortel, eds. Plenum Publishing Corp., New York, New York. p. 233-266 Feger, F., Varadaradjalou, S., Gao, Z., Abraham, S.N., and Arock, M. 2002. The role of mast cells in host defense and their subversion by bacterial pathogens. Trends Immunol. 23: 151-158. Fellermann, K., and Stange, E.F. 2001. Defensins -- innate immunity at the epithelial frontier. Eur. J. Gastroenterol.Hepatol. 13: 771-776. Fox, G., Parry, N.R., Barnett, P.V., McGinn, B., Rowlands, D.J., and Brown, F. 1989. The cell attachment site on foot-and-mouth disease virus includes the amino acid sequence RGD (arginine-glycine-aspartic acid). J. Gen.Virol. 70: 625-637. Francis, M.J. and Black, L. 1983. Antibody response in pig nasal fluid and serum following foot-and-mouth disease infection or vaccination. J. Hyg. 91: 329-334.

Copyright © 2004 By Horizon Bioscience

\218

Humoral and Cellular Immune Response Against FMDV

Francis, MJ., Hastings, G.Z., Clarke, B.E., Brown, A.L., Beddell, C.R., Rowlands, OJ., and Brown, F. 1990. Neutralizing antibodies to all seven serotypes of foot-and-mouth disease virus elicited by synthetic peptides. Immunology. 69: 171-176. Francis, MJ., Hastings, G.Z., Syred, A.D., McGinn, B., Brown, F., and Rowlands, D.J. 1987. Non-responsiveness to a foot-and-mouth disease virus peptide overcome by addition of foreign helper T-cell determinants. Nature. 330: 168-170. Fraser, LP., Koziel, H., and Ezekowitz, R.A. 1998. The serum mannose-binding protein and the macrophage mannose receptor are pattern recognition molecules that link innate and adaptive immunity. Semin. Immunol. 10: 363-372. Gallucci, S. and Matzinger, P. 2001. Danger signals: SOS to the immune system. Curro Opine Immunol. 13: 114-119. Garcia-Briones, M.M., Blanco, E., Andreu, D., Chiva, C., Ley, V., and Sobrino, F. 2004 Immunogenitity and T cell recognition in swine of foot-and-mouth disease virus polymerase 3D. Virology. In press. Garcia-Briones, M.M., Russell, G.C., Oliver, R.A., Tami, C., Taboga, 0., Carrillo, E., Palma, E.L., Sobrino, F., and Glass, EJ. 2000. Association of bovine DRB3 alleles with immune response to FMDV peptides and protection against viral challenge. Vaccine. 19: 1167-1171. Garcia-Valcarcel, M., Doel, T., Collen, T., Ryan, M., and Parkhouse, M. E. 1996. Recognition of foot-and-mouth disease virus and its capsid protein VPl by bovine peripheral T lymphocytes. J. Gen. Virol. 77: 727-735. Glass, EJ. and Millar, P. 1994. Induction of effective cross-reactive immunity by FMDV peptides is critically dependent upon specific MHC-peptide-T cell interactions. Immunology. 82: 1-8. Glass, EJ. and Millar, P. 1995. Bovine T cells preferentially recognize non-viral spacer epitopes in a putative FMDV vaccinal peptide. Vaccine. 13: 225-229. Glass, EJ., Oliver, R.A., Collen, T., Doel, T.R., DiMarchi, R., and Spooner, R.L. 1991. MHC Class II restricted recognition of FMDV peptides by bovine T cells. Immunology. 74: 594-599. Hamblin, C., Barnett, I.T., and Crowther, J.R. 1986. A new enzyme-linked immunosorbent assay (ELISA) for the detection of antibodies against foot-and-mouth disease virus. II. Application. J. Immunol.Methods. 93: 123-129. Hamblin, C., Kitching, R.P., Donaldson, A.I., Crowther, J.R., and Barnett, LT. 1987. Enzyme-linked immunosorbent assay (ELISA) for the detection of antibodies against foot-and-mouth disease virus. III. Evaluation of antibodies after infection and vaccination. Epidemiol. Infect. 99: 733-744. Hoffmann, J.A., Kafatos, F.C., Janeway, C.A., and Ezekowitz, R.A. 1999. Phylogenetic perspectives in innate immunity. Science. 284: 1313-1318. Jonuleit, H., Schmitt, E., Schuler, G., Knop, 1., and Enk, A.H. 2000. Induction of interleukin 10-producing, nonproliferating CD4(+) T cells with regulatory properties by repetitive stimulation with allogeneic immature human dendritic cells. J. Exp.Med. 192: 12131222. Kitching, R.P., Knowles, N.J., Samuel, A.R. and Donaldson, A.I. 1989. Development of footand-mouth disease virus strain characterization - a review. Trop. Anim. Hith. Prod. 21: 153-166. Konishi, E., and Nakao, M. 1992. Naturally occurring immunoglobulin M antibodies: enhancement of phagocytic and microbicidal activities of human neutrophils against Toxoplasma gondii. Parasitology. 104: 427-432. Leslie, R.G. 1985a. Critical events in the irreversible uptake of soluble immune complexes by macrophages. Immunol. Lett. 11: 153-158. Leslie, R.G. 1985b. Macrophage handling of soluble immune complexes: evaluation of mechanisms involved in the selective clearance of complexes from the circulation. Mol. Immunol. 22: 513-519. Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

2191

Leslie, R.G., and Coupland, K. 1985. The ingestion and degradation of soluble immune complexes by guinea-pig macrophages and neutrophils. Immunology. 54: 811-819. Litinskiy, M. B., Nardelli, B., Hilbert, D. M., He, B., Schaffer, A., Casali, P., and Cerutti, A. (2002). DCs induce CD40-independent immunoglobulin class switching through BLyS and APRIL. Nat. Immunol. 3: 822-829. Loeffler, F., and Frosch, P. 1897. Summarischer bericht uber die ergebnisse der untersuchungen der kommoission zur erforchung der maul-und-klamenseuche. Zentralbl. Bakterial. Parasitenkunde Infektionskranich. 22: 257-259. Mackay, C. R. 2001. Chemokines: immunology's high impact factors. Nat. Immunol. 2: 95-101. Mateu, M.G. 1995. Antibody recognition of picornaviruses and scape from neutralization: a structural view. Virus. Res. 38, 1-24. Mateu, M.G., Martinez, M.A., Capucci, L., Andreu, D., Giralt, E., Sobrino, F., Brocchi, E., and Domingo, E. 1990. A single amino acid substitution affects multiple overlapping epitopes in the major antigenic site of foot-and-mouth disease virus of serotype C. J. Gen. Virol. 71: 629-637. Matsushita, M., Endo, Y., and Fujita, T. 1998. MASPI (MBL-associated serine protease 1). Immunobiology. 199: 340-347. McCullough, K.C., Basta, S., Knotig, S., Gerber, H., Schaffner, R., Kim, Y. B., Saalmuller, A., and Summerfield, A. 1999. Intermediate stages in monocyte-macrophage differentiation modulate phenotype and susceptibility to virus infection. Immunology. 98: 203-212. McCullough, K.C., Bruckner, L., Schaffner, R., Fraefel, W., Muller, H.K., and Kihm, U. 1992a. Relationship between the anti-FMD virus antibody reaction as measured by different assays, and protection in vivo against challenge infection. Vet. Microbiol. 30: 99-112. McCullough, K.C., Crowther, J.R ., Butcher, R.N., Carpenter, W.C., Brocchi, E., Capucci, L., and De Simone, F. 1986. Immune protection against foot-and-mouth disease virus studied using virus-neutralizing and non-neutralizing concentrations of monoclonal antibodies. Immunology 58: 421-428. McCullough, K.C., Crowther, J.R., Carpenter, W.C., Brocchi, E., Capucci, L., De Simone, F., Xie, Q., and McCahon, D. 1987a. Epitopes on foot-and-mouth disease virus particles. I. Topology. Virology 157: 516-525. McCullough, K.C., De Simone, F., Brocchi, E., Capucci, L., Crowther, J.R., and Kihm, U. 1992b. Protective immune response against foot-and-mouth disease. J. Virol. 66: 18351840. McCullough, K.C., Parkinson, D., and Crowther, J.R. 1988. Opsonization-enhanced phagocytosis of foot-and-mouth disease virus. Immunology. 65: 187-191. McCullough, K.C., Pullen, L., and Parkinson, D. 1992c. The immune response against footand-mouth disease virus: influence of the T lymphocyte growth factors IL-l and IL-2 on the murine humoral response in vivo. Immunol. Lett. 31: 41-46. McCullough, K.C., Smale, C.J., Carpenter, W.C., Crowther, J.R., Brocchi, E., and De Simone, F. 1987b. Conformational alteration in foot-and-mouth disease virus virion capsid structure after complexing with monospecific antibody. Immunology. 60: 7582. Medzhitov, R. and Janeway, C., Jr. 2000a. Innate immune recognition: mechanisms and pathways. Immunol. Rev. 173: 89-97. Medzhitov, R. and Janeway, C., Jr. 2000b. Innate immunity. N. Engl. J. Med. 343: 338344. Medzhitov, R. and Janeway, C., Jr. 2000c. The Toll receptor family and microbial recognition. Trends Microbiol. 8: 452-456.

Copyright © 2004 By Horizon Bioscience

\220

Humoral and Cellular Immune Response Against FMDV

Meloen, R.H., Langeveld, J.P., Schaaper, W.M., and Slootstra, J.W. 2001. Synthetic peptide vaccines: unexpected fulfillment of discarded hope? Biologicals. 29: 233-236. Mulcahy, G., Gale, C., Robertson, P., Iyisan, S., DiMarchi, R.D., and Doel, T.R. 1990. Isotype responses of infected, virus-vaccinated and peptide-vaccinated cattle to footand-mouth disease virus. Vaccine. 8: 249-256. Mulcahy, G., Reid, E., DiMarchi, R.D., Gale, C., and Doel, T.R. 1992. Maturation of functional antibody affinity in animals immunised with synthetic foot-and-mouth disease virus. Res.Vet. Sci. 52: 133-140. Murdoch, C., and Finn, A. 2000. Chemokine receptors and their role in inflammation and infectious diseases. Blood. 95: 3032-3043. Nelson, P.1. and Krensky, A.M. 2001. Chemokines and allograft rejection: narrowing the list of suspects. Transplantation. 72: 1195-1197. Ochsenbein, A.F., Fehr, T., Lutz, C., Suter, M., Brombacher, F., Hengartner, H., and Zinkemagel, R.M. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science. 286: 2156-2159. Oh, Y. K. and Swanson, J.A. 1996. Different fates of phagocytosed particles after delivery into macrophage lysosomes. J. Cell. BioI. 132: 585-593. Pay, T.W., and Hingley, P.1. 1987. Correlation of 140S antigen dose with the serum neutralizing antibody response and the level of protection induced in cattle by foot-andmouth disease vaccines. Vaccine. 5: 60-64. Pereira, H.G. 1977. Subtyping offoot-and-mouth disease virus. Dev. BioI. Stand. 35: 167-174. Pereira, H.G. 1981. Foot-and-mouth disease. In: Virus Disease of Food Animals, Vol. 2. Gibbs, E.P.J., ed. Academic Press Inc., London. p. 333-363. Perez, F.M., Wigdorovitz, A., Romera, A., Zamorano, P., Borca, M.V., and Sadir, A.M. 2000. Detection and characterization of functional T-cell epitopes on the structural proteins VP2, VP3, and VP4 of foot and mouth disease virus 01 campos. Virology. 271: 234239. Pfaff, E., Mussgay, M., Bohm, H.O., Schulz, G.E., and Schaller, H. 1982. Antibodies against a preselected peptide recognize and neutralize foot and mouth disease virus. EMBO J. 1: 869-874. Pulendran, B., Palucka, K., and Banchereau, J. 2001. Sensing pathogens and tuning immune responses. Science. 293: 253-256. Rescigno, M., Rotta, G., Valzasina, B., and Ricciardi-Castagnoli, P. 2001a. Dendritic cells shuttle microbes across gut epithelial monolayers. Immunobiology. 204: 572-581. Rescigno, M., Urbano, M., Valzasina, B., Francolini, M., Rotta, G., Bonasio, R., Granucci, F., Kraehenbuhl, J.P., and Ricciardi-Castagnoli, P. 2001b. Dendritic cells express tight junction proteins and penetrate gut epithelial monolayers to sample bacteria. Nat. Immunol. 2: 361-367. Rigden, R.C., Carrasco, C.P., Barnett, P.V., Summerfield, A., and McCullough, K.C. 2003. Innate immune responses following emergency vaccination against foot-and-mouth disease virus in pigs. Vaccine. 21: 1466-1477. Rigden, R.C., Carrasco, C.P., Summerfield, A., and McCullough, K.C. 2002. Macrophage phagocytosis of foot-and-mouth disease virus may create infectious carriers. Immunology. 106: 537-548. Rodriguez, A., Ley, V., Ortuno, E., Ezquerra, A., Saalmuller, A., Sobrino, F., and Saiz, le. 1996. A porcine CD8+ T cell clone with heterotypic specificity for foot-and- mouth disease virus. J. Gen. Virol. 77: 2089-2096. Rodriguez, A., Saiz, J.e., Novella, I.S., Andreu, D., and Sobrino, F. 1994. Antigenic specificity of porcine T cell response against foot-and-mouth disease virus structural proteins: identification ofT helper epitopes in VP1. Virology. 205: 24-33. Rodriguez,A., Saiz, J.C., and Sobrino, F. 1995./n vitro synthesis offoot-and-mouth disease virus specific antibodies by porcine leukocytes. Arch.Virol. 140: 1645-1652.

Copyright © 2004 By Horizon Bioscience

McCullough and Sabrina

221\

Romani, N., Koide, S., Crowley, M., Witmer-Pack, M., Livingstone, A.M., Fathman, C.G., Inaba, K., and Steinman, R.M. 1989. Presentation of exogenous protein antigens by dendritic cells to T cell clones. Intact protein is presented best by immature, epidermal Langerhans cells. J. Exp. Med. 169: 1169-1178. Rossi, D. and Zlotnik, A. 2000. The biology of chemokines and their receptors. Annu. Rev Immunol. 18: 217-242. Rowlands, D.1. 1994. Experimental determination of immunogenic sites. In Synthetic Vaccines. Edited by B. H. Nicholson. Oxford: Blackwell Scientific Publications. Saalmuller, A. 1998. Antigen-specific immune response of porcine T lymphocytes to various pathogens. Rev. Sci. Tech. 17: 71-83. Saiz, J. C., Rodriguez, A., Gonzalez, M., Alonso, F., and Sobrino, F. 1992. Heterotypic lymphoproliferative response in pigs vaccinated with foot-and-mouth disease virus. Involvement of isolated capsid proteins. J. Gen. Virol. 73: 2601-2607. Salt, J.S. 1993. The carrier state in foot and mouth disease--an immunological review. Br. Vet. 1. 149: 207-223. Salt, J.S., Barnett, P.V., Dani, P., and Williams, L. 1998. Emergency vaccination of pigs against foot-and-mouth disease: protection against disease and reduction in contact transmission. Vaccine. 16: 746-754. Sanz-Parra, A., Blasco, R., Sobrino, F., and Ley, V. (1998a). Analysis of the Band T cell response in guinea pigs induced with recombinant vaccinia expressing foot-and-mouth disease virus structural proteins. Arch.Virol. 143: 389-398. Sanz- Parra, A., Jfmenez-Clavero, M.A., Garcia-Briones, M., Blanco, E. Sobrino, F. and Ley, V. 1999a. Recombinant adenovirus expressing the foot-and-mouth disease virus PI precursor polypeptide induces cellular but not humoral antiviral immunity and partial protection in pigs. Virology. 259: 129-134. Sanz-Parra, A., Sobrino, F., and Ley, V. 1998b. Infection with foot-and-mouth disease virus results in a rapid reduction of MHC class I surface expression. J. Gen. Virol. 79: 433436. Sanz-Parra, A., Vazquez, B., Sobrino, F. Cox, S.J., Ley, V. and Salt, J. 1999b. Evidence of partial protection against foot-and-mouth disease in cattle immunised with a recombinant adenovirus vector expressing the precursor polypeptide (PI) of foot-andmouth disease capsid proteins. J. Gen. Virol. 80,671-679. Schuler, G. and Steinman, R.M. 1985. Murine epidermal Langerhans cells mature into potent immunostimulatory dendritic cells in vitro. J.Exp.Med. 161: 526-546. Scicluna, L.A., Bruckner, L., and McCullough, K.C. 2001. Qualitative assessment of the humoral immune status against FMDV in post-vaccination cattle. Vaccine. 19: 29752986. Scicluna, L.A. and McCullough, K.C. 1999. Rapidity of specific antibody-antigen interactions. Scand.J. Immunol. 50: 167-176. Sobrino, F., Blanco, E., Garcia-Briones, M., and Ley, V. 1999. Synthetic peptide vaccines: foot-and-mouth disease virus as a model. Dev.Biol Stand. 101: 39-43. Sobrino, F., Blanco, E., Nunez, J.1., Jimenez-Clavero, M.A., and Ley, V. 2002. Cellular immune response to foot-and-mouth disease virus. Mod. Asp. Immunobiol. 2: 166168. Sobrino, F., Saiz, M., Jimenez-Clavero, M.A. Nunez, J.I., Rosas, M.F., Baranowski, E., and Ley, V. 2001. Foot-and-mouth disease virus: a long known virus, but a current threat. Vet. Res. 32: 1-30. Stahl, P.D. and Ezekowitz, R.A. 1998. The mannose receptor is a pattern recognition receptor involved in host defense. Curro Opine Immunol. 10: 50-55. Steinman, R.M. 1991. The dendritic cell system and its role in immunogenicity. Annu.Rev Immunol. 9: 271-296.

Copyright © 2004 By Horizon Bioscience

\222 Humoral and Cellular Immune Response Against FMDV

Steward, M.W., Stanley, C.M., Dimarchi, R., Mulcahy, G., and Doel, T.R. 1991. Highaffinity antibody induced by immunization with a synthetic peptide is associated with protection of cattle against foot-and-mouth disease. Immunology. 72: 99-103. Straub, O.C. 1995. Foot-and-mouth disease. Challenge of cattle after multiple vaccinations. Compo Immunol. Microbiol. Infect. Dis. 18: 253-257. Summerfield, A., Guzylack-Piriou, L., Schaub, A., Carrasco, C.P., Tache, v., Charley, B., and McCullough, K.C. 2003. Porcine peripheral blood dendritic cells and natural interferon producing cells. Immunology. 110: 440-449. Taboga, 0., Tami, C., Carrillo, E., Nunez, J.I., Rodriguez, A., Saiz, J.C., Blanco, E., Valero, M.L., Roig, X., Camarero, J.A., Andreu, D., Mateu, M.G., Giralt, E., Domingo, E., Sobrino, F., and Palma, E.L. 1997. A large-scale evaluation of peptide vaccines against foot-and-mouth disease: lack of solid protection in cattle and isolation of escape mutants. J. Virol. 71: 2606-2614. Takeda, K. and Akira, S. 2001. Roles of Toll-like receptors in innate immune responses. Genes Cells. 6: 733-742. Trombetta, E.S., Ebersold, M., Garrett, W., Pypaert, M., and Mellman, I. 2003. Activation of lysosomal function during dendritic cell maturation. Science. 299: 1400-1403. van Lierop, MJ., Nilsson, P.R., Wagenaar, J.P., van Noort, J.M., Campbell, J.D., Glass, E.J., Joosten, I., and Hensen, E.J. 1995a. The influence of MHC polymorphism on the selection of T-cell determinants of FMDV in cattle. Immunology. 84: 79-85. van Lierop, MJ.C., van Noort, J.M., Wagenaar, J.P.A., Rutten, V.P.M.G., Langeveld, J., Meloen, R.H., and Hensen EJ. 1994. T-cell stimulatory fragments of foot-and-mouth disease virus released by mild treatment with catepsin D. J. Gen. Virol. 75: 2937-2946. van Lierop, MJ.C., Wagenaar, J.~A., van Noort, J.M. and Hensen, EJ. 1995b. Sequences derived from the highly antigenic VPl region 140 to 160 offoot-and-mouth disease virus do not prime for a bovine T cell response against intact virus. J. Virol. 69: 4511-4514. van Maanen, C. and Terpstra, C. 1989. Comparison of a liquid-phase blocking sandwich ELISA and a serum neutralization test to evaluate immunity in potency tests of footand-mouth disease vaccines. J. Immunol. Methods. 124: 111-119. Vieira, O.V., Botelho, R.J., and Grinstein, S. 2002. Phagosome maturation: aging gracefully. Biochem. J. 366: 689-704. Summerfield, A., Guzylack-Piriou, L., Schaub, A., Carrasco, C.P., Tache, V., Charley, B., and McCullough, K.C. 2003. Porcine peripheral blood dendritic cells and natural interferon producing cells. Immunology. 110: 440-449. Wagner, J.G. and Roth, R.A. 2000. Neutrophil migration mechanisms, with an emphasis on the pulmonary vasculature. Pharmacol. Rev. 52: 349-374. Werling, D., Collins, R.A., Taylor, G., and Howard, C.J. 2002. Cytokine responses of bovine dendritic cells and T cells following exposure to live or inactivated bovine respiratory syncytial virus. J. Leukoc. BioI. 72: 297-304. Worth, R.O., Mayo-Bond, L., Kim, M.K., van de Winkel, J.O., Todd, R.F., III, Petty, H.R., and Schreiber, A.D. 2001. The cytoplasmic domain of FcgammaRIIA (CD32) participates in phagolysosome formation. Blood. 98: 3429-3434. Xie, Q.C., McCahon, D., Crowther, J. R., Belsham, G.J., and McCullough, K.C. 1987. Neutralization of foot-and-mouth disease virus can be mediated through any of at least three separate antigenic sites. J. Gen. Virol. 68: 1637-1647. Zamorano, P., Wigdorovitz, A., Chaher, M.T., Fernandez, F.M., Carrillo, C., Marcovecchio, F.E., Sadir, A.M., and Borca, M.V. 1994. Recognition ofB and T cell epitopes by cattle immunized with a synthetic peptide containing the major immunogenic site of VP 1 FMDV 01 Campos. Virology. 201: 383-387. Zlotnick, A., Aldrich, R., Johnson, J.M., Ceres, P., and Young, M.J. 2000. Mechanism of capsid assembly for an icosahedral plant virus. Virology. 277: 450-456.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 9 Functional and Structural Aspects of the Interaction of Foot-and-Mouth Disease Virus with Antibodies

Mauricio G. Mateu and Nuria Verdaguer

Abstract The capsid of FMDV presents many neutralization epitopes that cluster in several exposed regions. One of the major antigenic sites involves the G-H loop of VPl, and two other major sites are defined by discontinuous epitopes that are structurally and functionally independent of that loop. The G-H loop appears as a mobile capsid element. In some virus strains this loop participates in discontinuous epitopes which may be disrupted by mutations in other capsid elements that stabilize alternative orientations of the G-H loop. In other virus strains the G-H loop delineates true continuous epitopes that can be faithfully mimicked by synthetic peptides with native-like conformational propensities, and that have been structurally and functionally characterized in detail. Some residues within this loop have a dual function in the interaction with the cell receptor and in direct recognition by many antibodies. Some of these neutralize infectivity through monovalent binding and stenc inhibition of the interaction with the cell receptor. The cell attachment site of FMDV is not hidden from antibody attack, but it is protected from variation by the effect of negative selection. The extreme antigenic diversity of FMDV occur through multiple mutations, including a few critical ones, on the restricted subset of residues in each antigenic region that are involved in antibody binding, but that are not involved in other capsid functions. 1. Introduction Antibodies are the major effectors for protection of susceptible animals against foot-and-mouth disease (FMD) (Chapter 8). Many studies have thus focused on the structural and functional aspects of the interaction of foot-and-mouth disease virus (FMDV) with antibodies. The results have contributed to a better understanding of virus-antibody recognition and the selective routes for virus escape from neutralization, and have important implications for the design of new vaccines and disease control. Much of the work on the antigenicity of FMDV carried out during the past 20 years is reviewed in this chapter. A number of studies by different groups explore some aspects that may provide a better insight into, or that we consider of particular interest. Some of these studies and a part of our own work are presented in more detail.

Copyright © 2004 By Horizon Bioscience

\224

Interaction of FMDV With Antibodies

a Be

\\\"lI!"'"""------...r;;:::~

HI

DE FG

GH CD

b

Figure 1. (a) General fold of capsid proteins VPl, VP2 and VP3 of FMDV. The f)-strands are identified by single letters. Each loop is identified by two letters designing the strands it joins. Capsid surface is up (black arrow). (b) General schematic structure of the FMDV capsid. The approximate positions of most surface loops and C-termini of VPl (1), VP2 (2) and VP3 (3) are indicated for one biological protomer (compare Figure la). The C-terminus of a neighboring VPl is also indicated. This may facilitate the identification of the approximate positions of epitopes and antigenic sites described in the text (Section 3; see also Figure 3). The black circular area indicates the approximate relative size of an antibody footprint. (Adapted from Harrison (1989) with permission).

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2251

2. Overview of the FMDV Capsid Structure The structural characteristics of FMDV have been reviewed in Chapter 4 and will not be described here, except for a few aspects related with antigenicity. Knowledge of the X-ray structure of FMD virions of the 0, C or A serotypes, including OlBFS (Acharya et al., 1989), 01K (Lea et al., 1995), C 1Sta Pau, clone C-S8cl (Lea et al., 1994) and A22Iraq (Curry et al., 1996), has allowed a detailed comparison of the main structural features of the FMDV capsid. As in all other picornaviruses, the three major capsid proteins (VP1, VP2 and VP3, each about 210-220 residues in length), share a similar fold consisting of an eight stranded p-barrel with a jelly roll topology (Acharya et al., 1989; Figure la). In general, the highest sequence variability and major conformational differences between the capsid proteins of the three virus serotypes are found in the loops and the C-terminal segments that decorate the external surface of the virus (Figure 1b). Several of these exposed regions define much of the antigenic character of FMDV (Section 3). The FMDV capsid differs from those of other picornaviruses in having a relatively smooth surface in which any canyons, pits or marked protrusions are absent. The only exception is the long and highly exposed G-H loop of VPl (abbreviated here as the G-H loop). The G-H loop constitutes a major antigenic site (Sections 3 and 4) and also contains the conserved RGD (Arg-Gly-Asp) triplet critically involved in binding to cellular receptors of the integrin family, including avp3 (Fox et al., 1989; Mason et ai., 1994; Berinstein et al., 1995; Jackson, 1997; Sharma et al., 1997; Chapter 7). Some G-H loop residues located at a helical segment next to the RGD motif also participate in the specific attachment of FMDV to susceptible cells (Rieder et ai., 1994a; Mateu et ai., 1996; Jackson et al., 1997; 2000; Leippert et al., 1997). In the crystal structures of the FMDV particles the conformation of the functionally important G-H loop ofVPl could not be traced, which suggested this loop could be a highly mobile element. The only exception to date was found in dithiothreitol treated crystals of type 0 viruses. When the disulfide bond linking the base of the loop (Cys134 ofVPl to Cys130 ofVP2) in 0 1 viruses was reduced, the G-H loop adopted a more defined position, lying along the surface ofVP2. The RGD triplet showed an open turn conformation that was preceded by an extended region, and followed by a 3 10 helix which ended the loop (Logan et al., 1993; Figure 2). The accumulation of up to four changes within a 15-residue segment in the G-H loop of three different 0 1 viruses did not change its conformation (Lea et al., 1995). In viruses of types C and A, the G-H loop appeared disordered, even though the disulfide bond was absent (Lea et al., 1994; Curry et al., 1996). These and other lines of evidence suggested that the G-H loop may undergo hinge movements while still preserving a general intrinsic conformation (Sections 3.3, 4.4 and 4.5). 3. Antigenic Sites and Epitopes in the FMDV Capsid 3.1. Conceptual Distinctions

Some confusion has arisen in the comparison of FMDV antigenicity that may derive from the non-equivalent use of a few ambiguous terms. In this chapter we will make use of some important conceptual distinctions that have emerged from the work with many antigens, including FMDV itself.

Copyright © 2004 By Horizon Bioscience

1226 Interaction of FMDV With Antibodies

"down" G-H loop Figure 2. Ribbon diagram of a crystallographic protomer and a neighboring VPI molecule (right) in the capsid of reduced OIBFS virion. VPl, VP2 and VP3 within the protomer are depicted in dark, light and medium shades of grey, respectively. The G-H loop of VPI is located over VP2 ("down" position). The highlighted central area indicates the approximate location of the G-H loop in native OIBFS ("up" position), very close to the C-terminal segment of a neighboring VPl molecule (Sections 2 and 3.3).

3.1.1. Epitopes and Antigenic Sites

The term (B-cell) epitope will denote a defined part of the antigen molecule that is recognised by a specific antibody molecule. Antigenic site (or antigenic region) will denote a discrete area of the antigen surface where an epitope, or several structurally overlapping epitopes, have been (approximately) mapped (Wimmer and Jameson, 1984). Viral epitopes are usually grouped infunctio,!ally independent sites based on the lack of cross-neutralization activity of their defining monoclonal antibodies (MAbs) when assayed with MAb-resistant (MAR) ~irus mutants selected with other MAbs. However, epitopes assigned to functi~I)ally independent sites may sometimes belong to the same topological site (i.e., ;'they may overlap physically), based on their location on the capsid structure and on steric competition between the corresponding MAbs (or their Fab fragments) for binding the virion. In addition, some picornavirus antigenic sites previously defined as independent have been linked into a wider, single site when more complete panels of MAR mutants and MAbs have been used (Mateu, 1995; see below). Antigenic site is just an operational, not absolute, term and comparisons on the antigenic structure of FMDV should be done with caution. 3.1.2. Contact, Functional and Energetic Epitopes

In contrast to antigenic sites, epitopes can be precisely defined. We will consider three descriptions of any individual epitope: i) The contact epitope is formed by Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer

2271

those residues in the antigen molecule that establish non-covalent interactions with residues in the antibody paratope (Getzoff et al., 1988) and can be identified by X-ray crystallography of the antibody-antigen complex. ii) The functional epitope is formed by those residues that, if substituted, will affect the affinity for the antibody (Laver et al., 1990; Van Regenmortel et al., 1992), and can be analyzed in immunochemical studies using antigen mutants. iii), The energetic epitope is formed by those residues that are responsible for the major part of the interaction energy in the antigen-antibody complex (Novotny et aI., 1989), and can be defined through quantitative thermodynamic analyses, again using mutants. In general, only a part of the residues in a contact epitope participate also in the functional epitope (e.g., Air et al., 1990; Jin et aI., 1992). On the other hand, replacement of some antigen residues located outside the contact epitope may either introduce sterical restrictions or modify the local conformation, and thus affect indirectly the interaction with the antibody (Page et al., 1988; Section 3.3). Such residues may be considered a part of the functional epitope. In addition, only a few of the residues involved in the functional epitope may contribute most of the binding energy and be considered a part of the energetic epitope (Novotny et al., 1989). These are not just academic distinctions but have important practical consequences, as work with several antigens, including FMDV, clearly shows (Mateu, 1995; Sections 4 and 5). 3.1.3. Continuous and Discontinuous Epitopes The term continuous epitope will be applied to any epitope formed by residues contained within a short peptide segment (say, a single loop in the folded protein). A discontinuous epitope will be that formed by residues which are located apart in the primary structure, but that come spatially close in the folded protein. Because of the remarkable structural compactness of folded proteins, most B-cell epitopes in them are discontinuous. Some authors even proposed that no true continuous B-cell epitopes may exist in folded proteins. In fact, extensive work with FMDV and other antigens has proven that continuous B-cell protein epitopes, if infrequent, do exist and are biologically relevant (Mateu, 1995; Section 4). Of course, every epitope, be it a discontinuous ("conformational") epitope or a continuous ("linear") epitope in a protein or a peptide, can be considered conformation-dependent in that its specific recognition by the antibody does depend on its adoption of a single, folded conformation, either pre-formed or induced by the antibody itself. 3.2. The Antigenic G-H Loop of VP1 Pioneering work by Fred Brown and colleagues in Pirbright, UK showed that trypsin treatment of FMD virions of 0 and A serotypes greatly diminished their infectivity and attachment to susceptible cells, as well as their antigenicity and immunogenicity (Wild and Brown, 1967; Wild et al., 1969). Such treatment led, in strain OtK, to the excision and loss of a peptide segment located between VPl residues R138 and K154 (Strohmaier et al., 1982). This segment was later shown to form part of the exposed G-H loop in the capsid structure (Acharya et al., 1989) and to include the cell recognition site as well as immunodominant epitopes. Other early studies showed that isolated VP1, but not VP2 or VP3, induced a significant, albeit very weak, neutralizing response (Laporte et al., 1973;

Copyright © 2004 By Horizon Bioscience

\228

Interaction of FMDV With Antibodies

Bachrach et aI., 1975). Thus, VPl alone should contain those dominant epitopes not affected by capsid disruption. Use of capsid protein fragments and chemically synthesized peptides generally revealed one strongly immunogenic peptide which involved residues 138-160 (the G-H loop) of VPl and sometimes a second, less immunogenic peptide encompassing residues 200-213 (the VPl C-terminus) (Kaaden et al., 1977; Bachrach et aI., 1979; Strohmaier et al., 1982; Pfaff et aI., 1982; Bittle et al., 1982). In contrast to types 0 and A, trypsin treatment of type C virus did not lead to dramatic losses in infectivity or immunogenicity (Rowlands et al., 1971). This was later explained because of sequence differences in the hypervariable G-H loop. In a type C isolate (C-S8c 1), a single cleavage occurred, so no peptide segment was removed. Although this cleavage occurred within the ROD motif, it affected only weakly the RGD-mediated attachment to cells and impaired to different extents, but did not abolish, neutralization by MAbs that recognise the G-H loop and the RGD itself (Hernandez et al., 1996). It was proposed that the observed intraloop non-covalent interactions (Verdaguer et al., 1995; Section 4.4) could hold the cleaved ROD and the loop in partially folded, native-like conformations. The trypsin-sensitive antigenic site within the G-H loop ofVPl is now widely recognised as a major feature of all FMDV serotypes. The identification in the capsid of FMDV of an immunodominant region that could be isolated in the form of a short peptide opened the possibility to develop a synthetic peptide vaccine against FMD (Chapter 12). This has stimulated many structural and functional studies on the recognition by antibodies of this important antigenic loop of FMDV (Section 4). 3.3. Multiple Antigenic Sites in FMDV The available early evidence led some authors to consider the O-H loop of VPl as the only immunodominant region in FMDV. However, later evidence showed that aphthoviruses are, antigenically, not that different from other picornaviruses (reviewed in Minor, 1990; Mateu, 1995; Usherwood and Nash, 1995). Crossneutralization assays and the sequencing of MAR mutants have been instrumental for the identification of several independent antigenic sites in FMDV strains of different serotypes. 3.3.1. Serotype 0 The combined use of synthetic peptides as inhibitors of the binding of MAbs to virions, as well as other analyses, revealed discontinuous epitopes of FMDV OIBFS that involved the two previously identified antigenic segments (138154 and 200-213 of VP1; Parry et al., 1985; 1989). Surprisingly, amino acid substitutions in some MAR mutants were found not in the above segments, but in the B-C loop of VP1, even though these mutants exhibited a reduced reactivity with serum antibodies that should bind the G-H loop alone (because they had been raised against a peptide representing this loop). Crystallographic studies suggested a dual position for the disordered G-H loop. In the parental virus, this loop appeared to be close to the C-terminus of a neighbouring VPl molecule, and also to the B-C loop of VPl. In the mutant virus, the O-H loop appeared to be located over VP2 (Parry et aI., 1990; Figure 2). It was proposed that the discontinuous

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2291

epitopes conformed by the closely positioned G-H loop and C-terminus of VPl may be disrupted upon mutation at the neighboring B-C loop because this change would stabilize an alternative conformation of the G-H loop, a\vay from its original position (Parry et a!., 1990). According to this model, the G-H loop and the Cterminus of VPl in these mutants would behave immunogenically not as a single unit (as in the parental type 0 virus), but as independent antigenic sites. This has been experimentally observed for field strains of types C and A (see below). No 0 1BFS epitopes independent of the G-H loop were described, but several antigenic regions have been clearly detected in related type 0 viruses, including OISuisse (McCullough et aZ., 1987a), OIBrugge (Stave et a!., 1988), D1K and 0 1 Manisa (see next). Cross neutralization assays using MAR mutants derived from OIK revealed functionally independent epitopes that involved either the G-H loop of VPl, or residues in VP2 or VP3 (Pfaff et aZ., 1988). A similar approach identified four functionally independent antigenic sites of OIK, all formed by discontinuous epitopes (Xie etaZ., 1987; McCahon etaZ., 1989; Kitson etaZ., 1990). Site 1 epitopes clearly involved both the G-H loop (site la) and the C-terminus of VPl (site Ib), because mutations that allowed escape from the same MAb were found in either region. Epitopes that defined sites 3, 2 or 4 involved residues within the surface loops B-C of VP1, B-C and E-F of VP2 or the B-B "knob" of VP3, respectively (Kitson et a!., 1990). A fifth site was identified later (Crowther et a!., 1993), but the corresponding escape mutation was located within the G-H loop of VP1. Despite their functional independence (as operationally defined with the panels of MAbs and MAR mutants used), epitopes from sites 1 and 5 most probably overlap in a single topological (structural) site. Antibody competition experiments showed that sites 1,2 and 3 are topologically independent (McCullough et aZ., 1987a); MAbs against sites 4 and 5 were not used in those assays. More recently, bovine MAbs were tested for competition with the murine MAbs that defined antigenic sites 1-4. Site 3 was found independent, but the functional sites 2 and 4 were topologically linked (Barnett et aZ., 1998). MAR mutants selected with the bovine MAbs defined discontinuous epitopes which involved residues at the B-C and H-I loops of VP2 or the B-B knob of VP3 (site 2+4), or the B-C loop of VPl (site 3) (Barnett et aZ., 1998). Comparison of the structure of OIK and of a quadruple MAR mutant indicated that these discontinuous epitopes may be disrupted by the loss of single critical interactions with no other conformational changes occuring (Lea et aZ., 1995). However, disruption of G-H loop epitopes upon mutation of residues not located in the contact epitopes (as for OIBFS, see above) were also detected in OIK (Krebs et al., 1993). Recent work with yet another type 0 isolate (OlManisa) allowed the identification of several discontinuous epitopes that involved the G-H loop of VP1, the C-terminus of VPl or the B-C loop of VP2 (Aktas and Samuel, 2000). 3.3.2. Serotype A Analyses with MAbs, synthetic peptides and MAR mutants revealed several continuous epitopes in FMDV A22 Iraq that mapped in the G-H loop of VPl (residues 138-154; Bolwell et a!., 1989a). One trypsin-sensitive epitope located also in the VPl G-H loop was however disrupted by mutation(s) in VP2 (Bolwell Copyright © 2004 By Horizon Bioscience

\230

Interaction of FMDV With Antibodies

et al., 1989a; 1989b). Comparison of the structure of A22Iraq with that of a related antigenic variant suggested that residue 82, close to the B-C loop of VP2, may determine a somewhat different conformation of the G-H loop of VPl and indirectly disrupt that epitope (Curry et al., 1996). Several functionally and topologically independent sites of isolate A10Holland were identified by cross-neutralization and antibody competition analyses and the sequencing of many MAR mutants. Group 1 involved the G-H loop ofVP1, group 2 included the VPl C-terminus and group 3 was formed by epitopes defined by residues close to the B-C loop ofVP2 and the B-C, H-I and E-F loops and the B-B knob ofVP3. Again, a mutation in a different capsid element (the H-I loop ofVP1) appeared to disrupt an epitope that was located in the G-H loop of VPl (Thomas et al., 1988a). Three or four functionally independent sites were also identified in isolate A 12 119 (Baxt et ai., 1989), and the mutations located in the G-H loop, the H-I loop, the C-terminus ofVP1, or the G-H loop ofVP3, respectively. Finally, two functionally independent sites which involved the C-terminus of VPl or the B-C loop of VP2, respectively, were identified in isolate AsSpain (Saiz et al., 1991). 3.3.3. Serotype C

Studies with FMDV C-S8c 1 using large panels of neutralizing MAbs, synthetic peptides and MAR mutants revealed many overlapping, continuous epitopes located within the G-H loop of VPl (segment 136-150, termed antigenic site A; Mateu et al., 1987, 1989, 1990). A weakly neutralizing MAb defined a continuous epitope within segment 195-206 at the VPl C-terminus (site C), which was found topologically independent of the epitopes within the G-H loop (Lea et al., 1994). In addition, several discontinuous epitopes independent from sites A and C defined three mutually independent functional sites in cross-neutralization assays (sites D I, D2 and D3), but a single topological site (site D) in antibody-competition experiments (Lea et al., 1994). The escape mutations that defined site D epitopes occurred in a few solvent-exposed residues located within the beginning of the Cterminal segment ofVPl (site Dl), the B-C loop ofVP2 (site D2) or the B-B "knob" of VP3 (site D3). They all clustered in a region sligthly larger than an antibody footprint, centered on the VPI-VP2-VP3 junction in each capsid protomer. The site D area did not spatially overlap with the locations of site C or with the deduced locations of the mobile G-H loop (site A) on the capsid surface (Section 4.5). 3.3.4. Other FMDV Serotypes

Several functionally independent antigenic sites have been detected also in other serotypes, including the Asia 1 serotype (Sanyal et ai., 1997; Butchaiah and Morgan, 1997; Marquardt et ai., 2000). Comparison of the sequences of natural virus variants allowed the suggestion that three epitopes in an Asia-l isolate may respectively involve the G-H loop of VPl and B-C loop of VP2, the B-B knob of VP3 or the N-terminus of VP2 (Marquardt et al., 2000). 3.4. Immunodominance of Different Antigenic Sites in the Natural Hosts

There is compelling evidence that support the G-H loop of VPl as one of the immunodominant regions in FMDV. Some observations follo\v: i), cleavage of the G-H loop causes a generally strong reduction in immunogenicity and antigenicity (Section 3.2). ii), a large number of anti-virion Mabs recognise the G-H loop; iii), Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer

2311

G-H loop peptides may elicit high levels of virus-neutralizing antibodies, confer protection against infection by FMDV and mimic the serotype, subtype and strain-specific antigenicity of whole virions (Section 4.1). iv), immunoaffinity fractionation of polyclonal antibodies from swine --a natural host of FMDV-showed that, on average, about 60% or 30% of the virus-neutralizing activity in sera from animals that had been respectively infected or vaccinated with a type C isolate (C-S8cl) corresponded to antibodies that bound the G-H loop (Mateu et ai., 1995a). High percentages of anti G-H loop antibodies in sera from either laboratory animals or cattle have been quoted (Pfaff et ai., 1982; Brown et al., 1990). There is also equally abundant evidence that demonstrates the G-H loop is not the only immunodominant site in the FMDV capsid. A few examples follow: i) Bovines immunized with a subtype C3 vaccine strain were not fully protected against a newly emerged C 3 variant (Bergmann et al., 1988). The two viruses differed genetically and antigenically not at sites A (the G-H loop) or C, but at other sites (site D, and the region corresponding to site 3 in OlK; Feigelstock et al., 1992). Similar observations apply to other serological variants (Mateu et ai., 1994; Section 7). ii) Compared with the parental OlK virus, a quintuple mutant that exhibited a critical mutation in each of the five functionally identified antigenic sites, but not the single mutants, was substantially less neutralized by sera from laboratory animals or from post- vaccinated or infected cattle (Crowther et al., 1993). iii) Swine inoculated with a vaccine based on a chimeric type A virus whose G-H loop had been replaced by that of a type C virus were protected from challenge with the type A virus (Rieder et al., 1994b). iv) Competition between sera from susceptible animals and MAbs directed against distinct antigenic regions of A1oHolland showed that antibodies against sites different from the G-H loop were present in substantial amounts in those animals (Thomas et al., 1988b). v) Bovine MAbs recognised epitopes independent of the G-H loop that mapped in the same regions identified with murine MAbs (Barnett et ai., 1998). vi) Competition between sera from vaccinated sheep, pigs or cattle and Mabs directed against any of the independent antigenic sites (1, 2, 3) defined in OIK indicated that none of these three antigenic sites alone can be considered immunodominant in the natural hosts, and that other as yet undefined sites may be also important in the induction of a protective immune response (Aggarwal and Barnett, 2002). vii) The degree of immunodominance of the G-H loop may be dramatically different in different conditions or animals. For example, in sera from convalescent pigs the percentage of FMDV-neutralizing activity that bound the G-H loop was generally high (from 40% to 80%). However, in one animal it amounted to only 7%. In vaccinated swine the values ranged from 85% down to as low as 2% (Mateu et al., 1995a). 3.5. A Continuum of Discontinuous Epitopes?

Short peptides representing the sequence of many segments of the three major capsid proteins of FMDV were found reactive with anti-virion antibodies, which suggested the presence of epitopes scattered over the entire capsid surface (Meloen et al., 1986). Very short peptides may at best mimic a part of an epitope, and false positives are a frequent occurrence. Even so, those results provide some support to the possibility that, as observed with model proteins, the entire smooth surface of the FMDV capsid may prove antigenic. Escape mutations tend to cluster in some Copyright © 2004 By Horizon Bioscience

\232 Interaction of FMDV With Antibodies

(a)

(b)

Figure 3. Antigenic residues on the FMDV capsid as defined with antibody escape mutants. The three dimensional structure of the FMDV C-S8c I virion (a) and of a biological protomer (b) are represented in spacefilling models. The amino acid residues involved in escape mutations in type C FMDV, and the residues at positions equivalent to those found involved in escape mutations in FMDVs of types 0 and A are depicted in colour. Residues in green correspond to the mobile G-H loop of VPl, which is shown here in the orientation adopted in the complex between the virus and Fab SD6 (Figure 5). Residues in blue are those involved in the other antigenic sites identified in FMDV. Clustering of antigenic residues in the G-H loop, near the 5-fold axis (top) and near the 3-fold axis (bottom) is apparent and approximately correspond to the three general antigenic sites identified in FMDV (section 3.6). However, scattering of antigenic residues over a large part of the smooth capsid surface is also apparent (Section 3.5). See Colour Plate at the back of the book.

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2331

highly exposed positions at discrete capsid regions, but this may be partly due to a higher epitope density around the antigenic regions identified with MAR mutants and, most important, to a higher structural tolerance of such regions to genetic variation (Mateu, 1995; Sections 4 and 7; Figure 3). Many discontinuous FMDV epitopes do not depend on the integrity of the complete virion. Both natural and recombinant empty FMDV 0 1 capsids were antigenically similar to full virions when confronted \vith MAbs against different sites identified in OIK (Abrams et al., 1995). Even the unassembled and unprocessed capsid protomer precursor was able to preserve the discontinuous neutralization epitopes that had been identified with MAbs on FMDV C-S8c I (Saiz et al., 1994). However, some discontinuous epitopes are present only in the complete FMD virion. Some of these epitopes may be disrupted in capsid subassemblies because of the incidence of local conformational changes. Others could bridge neighboring pentamers. It is \vell known that the immunogenicity of a virus-inactivated FMD vaccine strongly depends on the integrity of the virions in the preparation. However, the particulate nature and immunological multivalency of complete virions may suffice to explain this observation. The relative dominance of virion-specific epitopes and of overlapping epitopes scattered over most of the FMDV capsid surface remains to be clarified. The outcome may have practical consequences regarding the potentialities and shortcomings of potential vaccines based on subcomponents of FMDV. 3.6. Conclusion: A Consensus Antigenic Structure for FMDV

The epitopes and antigenic sites identified in different isolates, subtypes and serotypes of FMDV are not completely equivalent. However, it is now clear that the FMDV capsid presents many different epitopes that can be assigned to several major antigenic sites. Three or four antigenic regions that basically appear as structurally independent have been generally identified, and each contains several overlapping epitopes that have been found frequently, though not always, functionally connected (Figure 3). We propose a general nomenclature to identify these capsid regions as (structural) antigenic sites I, II and III of FMDV. Site I would involve the mobile G-H loop of VPl (site la) and the C-terminal segment of VP1 (site Ib). Depending on the virus strain or serotype, sites la and Ib may behave either as structurally independent antigenic sites that delineate continuous epitopes, or as a single site composed of discontinuous epitopes. Some of these site I epitopes may be disrupted by altering the orientation on the capsid or the intrinsic conformation of the G-H loop, through mutations of residues located in neighboring capsid elements. Site II (site(s) 2+4 in Kitson et al., 1990 and Barnett et al., 1998; group 3 in Thomas et al., 1988a; site D in Lea et aI., 1994) would involve discontinuous epitopes that can be roughly located close to the 3-fold axis and the junction of VP1-VP2-VP3 in each biological capsid protomer. These epitopes are defined by residues within the highly exposed B-C loop of VP2, the B-B knob of VP3 and/or in other neighboring loops, including the H-I and E-F loops of VP2, the B-C and G-H loops of VP3 and a turn in VP1 not far from the C-terminus. Site III (site 3 in Kitson et aI., 1990 and Barnett et al., 1998; also a putative site in Thomas et al., 1988a) would involve discontinuous epitopes located in a region of each capsid protomer close to the 5-fold axis that includes the B-C and H-I loops ofVP1 (Figure 3). Copyright © 2004 By Horizon Bioscience

\234

Interaction of FMDV With Antibodies

The G-H loop of VPl is frequently referred to, even today, as the immunodominant site of FMDV. In fact, it is now clear that its immunodominance may vary to widely different extents (to the point of being nearly immunosilent in occasions), and that the other independent antigenic regions identified may be equally immunodominant.

4. Functional and Structural Studies on the Interaction of the G-H Loop of FMDV with Antibodies A considerable number of biochemical, immunological, spectroscopic, crystallographic and cryo-electron microscopy (cryo-EM) studies have focused on the G-H loop of FMDV as an independent antigenic element. The different approaches have proved mutually complementary, and provide a uniquely comprehensive view of the interaction between antibodies and a major antigenic loop in a virus capsid. 4.1. Functional Studies on G-H Loop Epitopes Using Synthetic Peptides and MAR Virus Mutants 4.1. 1. Early Work

Synthetic peptides that represented the VPl G-H loop (residues 141-160) ofFMDV serotypes 0, A or C (Bittle et a!., 1982), of different A subtypes (Clarke et al., 1983) or of FMDV A 12 119 variants which differed in residues 148 and 153 (Rowlands et al., 1983) were used as antigens and to immunize guinea pigs. Both the viruses and the peptides elicited antibodies that neutralized the homologous virus to a much greater extent than the heterologous viruses. Thus, the peptides mimicked the serotype, subtype and individual antigenic specificity of complete virions, and allowed the identification of the residues responsible for the specificity of clonal FMDV variants. In another early study (Geysen et al., 1984, 1985), overlapping hexapeptides that spanned G-H loop sequences of FMDV OlBFS, A10Holland or C1Detmold were reacted with sera from rabbits that had been inoculated with the homologous virions. The most antigenic hexapeptides corresponded to a segment that contained part of the (later identified) cell attachment RGD triplet, and the immediate C-terminal residues. By using replacement sets of peptides, in which each of the residues in the hexapeptides had been singly replaced by all other 19 natural amino acids, it was found that: i), not all residues were equally critical for antibody recognition. ii), the residues critically recognised by different antibody populations, even if these had been elicited against the same virus isolate, were not necessarily the same. iii), some conserved residues that formed part of the RGD(L) motif were important for antibody recognition of all three serotypes. iv), the residues critical for recognition by antibodies elicited against the peptides themselves did not match those identified with anti-virus sera. These important studies, however, used polyclonal antibodies, and thus provided a first look at the average antigenicity of the G-H loop. Also, the studies with hexapepties failed to recognise that complete G-H loop epitopes do involve longer peptide segments; this may have affected the assignment of critical residues. Later studies used MAbs and longer peptides to provide insights into individual epitopes within this loop (see next).

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaquer 2351

4.1.2. Functional Features of Individual G-H Loop Epitopes

Immunochemical analyses using MAbs and large collections of antibody escape mutants and field virus variants, as well as peptides of nested or overlapping sequences and/or containing single or multiple amino acid replacements, have allowed a precise definiton of the residues involved in specific functional epitopes within the G-H loop. For exampe, several overlapping but distinguishable, continuous epitopes were thus finely mapped in FMDV A22 Iraq (Bolwell et al., 1989a) and C-S8c1 (Mateu et al., 1989; 1990). Some epitopes spanned both sides of the RGD triplet, while a few appeared to involve mainly residues located at the C-terminus of this motif. Work with FMDV C-S8c1 also revealed that linear peptides, either coupled to a carrier protein or free in solution, quantitatively mimicked the antigenic activity and specificity of the G-H loop in the complete virion when confronted with individual MAbs. With most MAbs, 15-mer to 21-mer peptides encompassing the core sequence 138-150 of VP1 were almost as reactive, on a molar basis, as the G-H loop in the complete virion. In both cases, single (or multiple) replacements caused the same decrease in binding activity (Mateu et ai., 1989; 1992; Benito et al., 1995). Thus, variant peptides could be used to investigate the antigenic features of individual G-H loop epitopes. The neutralizing MAbs used in these studies had been elicited against FMD virions and recognised native epitopes, since their binding affected a biological activity (infectivity) of the virions. The quantitative results with peptides indicated that these native epitopes were true continuous epitopes, a finding that was confirmed in later functional and structural studies (see below and Sections 4.4 and 4.5). Studies by different groups generally revealed that specific replacements of some residues within the G-H loop had an effect on the reactivity of only one or a few anti-virus Mabs. In contrast, other replacements at specific residues (such as VP1 residue 146 in type C viruses) were found antigenically critical, in that they severely impaired or abolished the binding of all MAbs tested (Mateu et al., 1990). Residues in the RGD motif were also found critical for recognition by most type C FMDV MAbs directed against the G-H loop (Novella et ai., 1993; Verdaguer et al., 1998; Ruiz-Jarabo et ai., 1999). A study with a complete single-replacement set of 15-mer variant peptides (VP1 residues 136-150) allowed a very precise functional characterization of several epitopes within the G-H loop of FMDV C-S8c1 (Verdaguer et al., 1998). All epitopes proved functionally continuous, as many single replacements at any of 9 positions within a 10-residue stretch (138147), including the RGD triplet, affected the recognition by each MAb tested. Most MAbs recognised extensively overlapping functional epitopes. However, there were clear differences regarding the effects of some substitutions on the recognition by different MAbs, and some of these (e.g., MAb SD6 or the \veakly neutralizing MAb 4G3), clearly involved other residues as well (Mateu et al., 1990; Verdaguer et al., 1998). The subtle functional differences in the recognition of the G-H loop by different MAbs account for a part of the widely diverse antigenic specificities of field isolates of a same serotype (Mateu et ai., 1988; 1994; Martinez et ai., 1991; Section 7).

Copyright © 2004 By Horizon Bioscience

\236

Interaction of FMDV With Antibodies

4.1.3. Functional Restrictions to Variation in a Major Antigenic Site

The results obtained with peptides showed that the functional epitope in FMDV C-S8c 1 defined by the neutralizing anti-C-S8c 1 MAb SD6 involved 11 residues in a 12-amino acid stretch (residues 136-147) within the G-H loop, and that many of the 19 possible amino acid replacements of most of these residues strongly affected binding by this MAb (Verdaguer et al., 1998). In striking contrast, only 8 different single replacements in just three residues (138, 139 and 146) were detected with a collection of more than 80 MAR mutants independently selected from C-S8cl populations with that same MAb (Mateu et al., 1989; Martinez et ai., 1997). The RGD triplet (residues 141-143) and neighboring residues L144, A145 and L147 were never found mutated. Remarkably, these six residues are those specifically involved in the attachment of FMDV C-S8cl to susceptible cells (Hernandez et al., 1996; Mateu et al., 1996). Consistent, though much less complete, results have been obtained with other MAbs. To our knowledge, only one escape mutant obtained from a field isolate (A10Holland; Thomas et al., 1988a) showed a mutation at the RGD triplet. This motif is almost absolutely conserved in field isolates belonging to any of the seven serotypes. Residues 144, 145 and 147 were also rarely found mutated in type C MAR mutants, and L144 and L147 are fairly conserved in this and other serotypes (Domingo et al., 1992; Mateu et ai., 1996). Those results and the structural studies described in Section 4.5 (Verdaguer et al., 1995) provided evidence of severe functional restrictions to genetic variation in an immunodominant site of FMDV (Mateu, 1995; Verdaguer et al., 1995). Further proof came from a variant of FMDV C-S8cl obtained by multiple serial passages in cell culture. This virus (C-S8clpl00) had acquired mutations that rendered biologically dispensable the RGD motif. The more than 30 MAR mutants selected from C-S8clpl00 using MAb SD6 showed a completely altered repertoire of escape mutations, that now involved also the normally indispensable RGD triplet, and the neighboring residues L144 andA145 (Martinez etai., 1997). The functional restrictions to sequence variation in the G-H loop of natural FMDVs were relieved in C-S8clpl00 probably because this virus acquired the ability to enter cells via a different mechanism (Martinez et al., 1997; Baranowski et al., 2000). 4.1.4. Comparative Antigenic Specificities of MAbs and Polyclonal Antibody Populations from Animal Hosts

It could be argued that the MAbs used to study the antigenicity of the G-H loop may not represent the dominant antibody specificities present in the natural hosts. In fact, it was found that the consensus (average) antigenic profile of the collection of antitype C virion MAbs we used was quantitatively similar to the antigenic profiles obtained with most anti-type C virion sera from convalescent or vaccinated swine (a natural host of FMDV). For example, the same residues, including the RGD triplet, were found antigenically critical in nearly all cases (Mateu et al., 1995b). A few exceptions were however apparent. As previously observed with sera from laboratory animals (Geysen et al., 1985), a synthetic peptide that included the GH loop sequence elicited antibodies in swine that showed an average antigenic profile clearly different from those obtained with sera from pigs vaccinated with the homologous virions (Mateu et al., 1995b). The ROD triplet does not appear to be an essential component of every epitope found within the G-H loop of FMDV of every serotype (Bolwell et al., 1989b, Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer

2371

Mateu et al., 1990; 1995b; Brown et aI., 1999). As mentioned above, other residues next to the ROD are also involved in attachment to cells, so it is likely that even O-H loop epitopes that may not include the ROD still include other residues that have also a function in binding the integrin receptor(s). 4.2. Engineering Improved G-H Loop Peptide Mimics 4.2.1. The Order-Disorder Paradox and the G-H Loop of FMDV

Even if a linear peptide is able to successfully mimic a complete continuous epitope in a protein, a substantially reduced affinity for the anti-protein MAb is to be expected in most cases, because of the energetic cost of folding an essentially unstructured peptide. However, the most antigenic G-H loop linear peptides showed an affinity in solution towards most anti-virion antibodies that were very close to (though different) that of the G-H loop in the FMD virion (Mateu et al., 1989; 1995b; Benito et aI., 1995; Section 4.1). Two non-exclusive possibilities to explain this observation are: i) the peptides may exhibit strong conformational propensities towards the native G-H loop structure. The energetic cost of folding the peptide during peptide-antibody complex formation would be thus reduced, and the affinity increased; ii) the G-H loop in the virion may possess some intrinsic flexibility and sample not one, but a (restricted) range of different conformations (see below). Folding of the loop during formation of the virion-antibody complex would lower the affinity, approaching its value to that of a flexible, free peptide. Some flexibility of the G-H loop could be biologically advantageous, perhaps for binding different integrins. 4.2.2. Carrier Proteins and Virus Chimeras

Several methods have been attempted to improve even further the already good antigenicity of G-H loop linear peptides and, much more importantly, their generally unsatisfactory immunogenicity and protective capacity (Taboga et al., 1997). One such method is based on the fusion of a G-H loop segment to proteins or to heterologous virus capsids (virus chimeras; Kitson et aI., 1991; Usha et aI., 1993), that may provide a large-sized carrier, adequate antigen multivalency and/or the necessary T-cell epitopes. This approach necessarily introduces some conformational restrictions on the FMDV segment inserted. Many of the constructions described were well recognised by anti-FMDV antibodies and/or elicited high levels of antibodies that neutralized FMDV infectivity. A remarkable observation is that a high proportion of the antibodies elicited against a linear GH loop peptide (residues 141-160 of OIBFS) coupled to a carrier protein (KLH) could be effectively adsorbed to FMD virions (Parry et aI., 1988). This suggested that this peptide could sample only a restricted range of conformations including that (those) found in the virus, although it is unclear whether this was an intrinsic feature of the peptide (as discussed above) or a consequence of its coupling to the protein. Some carrier proteins clearly affected the antigenicity of the G-H loop when probed with anti-virus MAbs, either in a negative or a positive way. Detailed studies showed that the variation in affinity was dependent on the specific antivirus MAb used (Benito et aI., 1995; Feliu et aI., 1998). This is consistent with the possibility that different anti G-H loop antibodies may recognise slightly different loop conformations (see also Section 4.4). These could be differentely sampled

Copyright © 2004 By Horizon Bioscience

\238

Interaction of FMDV With Antibodies

in different protein constructs due to the different conformational restrictions imposed on the antigenic segment by the carrier protein. 4.2.3. Cyclized Peptides

An elegant approach to further improve the antigenIcIty of linear G-H loop peptides has been the introduction of additional conformational constraints through intra-peptide covalent linkages. Cyclization of long G-H loop peptides via the introduction of disulfide bonds had little effect on its reactivity with anti-virion MAbs, and it was concluded that the closure of a large cycle did not provide a strong enough conformational restriction (Camarero et al., 1993; Valero et aI., 1995). Thus, further attempts were based on a reduction of the ring size. In this way, more pronounced effects of cyclization on antigenicity were found. The imposition of a fairly rigid amide bond between the ends of a 15-mer (136-150) peptide drastically impaired recognition by anti-virus MAbs, probably because of severe structural restrictions that impeded access to the appropriate conformation. Remarkably, cyclization of the same peptide via the introduction of a less restrictive disulfide bond improved substantially the reactivity with anti-virion MAbs over that of the linear peptide, even though the latter was already a very good mimic of the G-H loop (Valero et al., 2000). 4.2.4. Retro-Inverso Peptide Analogues

Retro-inverso peptide analogues are obtained by assembling amino acid residues in the reverse order from that in the natural peptide and replacing L-amino acids by D-amino acids (Muller et aI., 1995). Retro-inverso peptides that corresponded to the G-H loop (residues 141-159) of FMDV of subtype A12 induced high levels of antibodies that neutralized virus infectivity, and were also able to protect guinea pigs against infection by FMDV (Briand et al., 1997). These peptides are specially interesting as candidate peptide vaccines because they may be degraded more slowly than L-peptides (Briand et al., 1997). 4.3. Spectroscopic Studies of G-H Loop Peptides

Some studies have made use of circular dichroism (CD) and nuclear magnetic resonance (NMR) spectroscopy to study the conformational propensities of antigenic G-H loop peptides in solution. The far-UV CD spectra of synthetic peptides that reproduced the sequence of the 141-160 antigenic region at the G-H loop of seven variants isolated from FMDV A l2 119 were obtained under different conditions (Siligardi et al., 1991; France et aI., 1994). The peptides differed at residue positions 148 and 153 of VP 1 and had been grouped on the basis of sequence comparisons and cross-reactivity in serological tests (Rowlands et al., 1983; Section 4.1). In water all peptides showed similar spectra, typical of a random-coil conformation. However, when analyzed in structure-inducing solvents they showed different conformational propensities that correlated well with the different serological behaviour of the two groups. Based on these data and on molecular modelling simulations, the authors proposed a model for the G-H loop of different variants which involved a p-tum at the cell-attachment RGD triplet and a contiguous helical segment (France et al., 1994), two structural features similar to those that had been previously observed

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2391

by X-ray crystallography of reduced type 0 virions (Logan et al., 1993; Section 2 and Figure 2). The solution structure of the peptides corresponding to two of these FMDV variants were also analyzed by a combination of NMR spectroscopy and computer calculations. Again, the two peptides exhibited different conformational features. However, the structures proposed were different from the X-ray structure of the loop in the reduced type 0 virion (Pegna et al., 1996a, b). CD and NMR analyses of a long G-H loop peptide (residues 132-162) corresponding to type 0 in the presence of a structure-inducing solvent indicated helix formation between residues 143 and 159, a conformational preference that may be not consistent with the structure of the G-H loop on the complete type 0 virion (de Prat-Gay, 1997). CD and NMR analyses of linear and cyclized (conformationally constrained) peptides which represented the sequence of the VPl G-H loop ofFMDV type C (CS8cl; Section 4.2) have been also described (Camarero et ai., 1993; Haack et ai., 1997). The NMR analyses using a ciclized G-H loop peptide of type C (residues 134-155) in a structure-inducing solvent suggested the formation of a helix Cterminal from the RGD motif and also provided some evidence for the formation of a tum in the RGD region (Haack et al., 1997). NMR analyses also showed that binding of a linear G-H loop (141-159) peptide to a solid matrix induced (in dimethylformamide) a helical conformation encompassing residues 152 to 159 (Furrer et al., 2001). In a different study, analyses by NMR of a linear antigenic peptide representing the G-H loop (residues 141-159) of FMDV A I2 and of its retro-inverso peptide analogue in aqueous solution revealed similar conformational features, particularly at the C-termini. These similarities were consistent with the similar antigenic behavior of the two peptides (Petit et al., 1999). In conclusion, the results obtained by spectroscopic techniques so far reveal that both linear and conformationally restricted, short G-H loop peptides are little structured in water but do possess clear conformational propensities that are enhanced by structure-stabilizing solvents and also by binding to a solid phase. The conformations thus induced generally include formation of a helix in the Cterminal half and appear to reflect, in some cases at least, structural propensities relatively consistent with the structure observed by X-ray crystallography of the G-H loop either as part of the capsid (Logan et al., 1993) or bound to anti-virus antibodies (Verdaguer et ai., 1995; 1998; Section 4.4). Further work is however required to provide a clearer view on this subject. 4.4. Crystallographic Studies of G-H Loop Peptide-Antibody Complexes

The ability of synthetic peptides to accurately mimic viral epitopes that involve the G-H loop of VPI provided an excellent opportunity to obtain structural information, at close to atomic resolution, on the recognition of viruses by neutralizing antiviral antibodies, and to compare the structural and functional characteristics of continuous B-cell epitopes. The results of structural studies using complexes between peptides representing the G-H loop of FMDV and anti-virus antibodies are reviewed below. Related structural studies have been carried out with peptide-antibody complexes derived from influenza virus (Rini et al., 1992; Churchill et ai., 1994), rhinovirus (Tormo et al., 1994), HIV (Ghiara et al., 1994; Saphire et al., 2001), poliovirus (Wien et ai., 1995) and hepatitis B virus (Nair et ai., 2000; 2002) Copyright © 2004 By Horizon Bioscience

1240

Interaction of FMDV With Antibodies

(b)

Figure 4. (a) Ribbon diagram of the structure of a complex between a peptide representing the G-H loop of FMDV C-S8c 1 (yellow) and the Fab fragment of anti-virus neutralizing MAb SD6. The side chains of the ROD triplet in the peptide are also depicted (b) Close-up ball-and-stick model of the O-H loop peptide in the complex (Section 4.4). See Colour Plate at the back of the book.

4.4.1. Continuous Contact Epitopes Within the G-H Loop

The three dimensional structures of complexes which involved a synthetic peptide representing antigenic site A (the G-H loop) of any of three FMDV type C isolates, and the Fab fragment of a neutralizing MAb (either SD6 or 4C4) elicited against type C virions, were analyzed by X-ray crystallography (Verdaguer et al., 1995; 1996; 1998; Ochoa et al., 2000). In the structure of Fab SD6 co-crystallized with the IS-mer peptide that represented site A ofFMDV C-S8cl (Figure 4a), all fifteen amino acid residues were solved. Fab SD6 interacted closely with ten of these residues (Verdaguer et ai., 1995). In the electron density maps of the complexes formed between Fab 4C4 and a IS-mer peptide corresponding to site A of either FMDV C1Brescia or C-S30, only the terminal residues 136 and 148-150 could not be positioned. Fab 4C4 interacted closely with eight of the eleven residues seen in the peptide model (Verdaguer et ai., 1998; Ochoa et a!., 2000). Thus, the contact epitopes defined by SD6 and 4C4 consisted of an almost continuous stretch of amino acid residues, in agreement with the continuous nature of the corresponding functional epitopes identified in immunochemical studies (Section 4.1). 4.4.2. A Genera/ Conformation for the Free and Antibody-Comp/exed G-H Loop

Each of the equivalent G-H loop residues visible in the atomic models exhibited a very similar conformation in the three peptide-antibody complexes (Verdaguer et al., 1998; Ochoa et aI., 2000). In all complexes, the peptide acquired a quasi circular conformation stabilized by several intra-peptide hydrogen bonds and van der Waals interactions (Figure 4b). A short helical segment and three hydrogen bonded turns were observed in the complexed peptides. The conserved RGD triplet adopted an open tum conformation that was almost identical to that found Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2411

for the same triplet in the uncomplexed G-H loop on the capsid of reduced type 0 FMDVs (Logan et al., 1993). A similar conformation was observed for the RGD motif of an integrin ligand complexed with the extracellular segment of integrin av~3 (Xiong et al., 2002), which is also a receptor for FMDV. The helical segment in the structures of the type C peptides involved residues Asp 143 to Leu 147 next to the RGD triplet. In the G-H loop of the reduced type 0 virus, a 310 helix is initiated at the equivalent position and its direction is essentially coincident with that in the peptides. The structural similarities between the G-H loop of reduced type 0 virus and the antibody-bound type C peptides strongly suggests that the conformation of the RGD triplet and the adjacent helical segment are structurally conserved features of this loop in aphthoviruses. Despite the high structural similarity between all G-H loop peptides, some substantial differences in the main chain conformational angles were observed, particularly around the RGD motif. In the AI5-C-S30 peptide, residues Thr140 and Arg141 were situated in the Ramachandran region corresponding to a-helices while, in the AI5-C-S8cl peptide, Ala140 and Arg141 were found in the region of the ~-strands and the left handed helices, respectively. A third situation was found for the structure of the G-H loop in the reduced type 0 1 virions, where the equivalent residues were both found in the Ramachandran region corresponding to ~-strands. The flexibility of Gly142 allowed compensatory main chain torsional angles that result on the overall structural similarity observed for the three peptides (Ochoa et al., 2000). The arrangement of the N- and C-terminal ends of the type C peptides, that include the hypervariable, antigenic residues 137 to 140 and 148 to 150, differed considerably from the equivalent segments in the G-H loop of the type 0 virus. The low level of amino acid identity between the two viruses in these segments (which include deletions/insertions of residues) provides an obvious explanation for the conformational differences and the different antigenic specificities observed. The general features of the structures of different synthetic G-H loop peptides in complex with two Fab fragments and the G-H loop in reduced FMDV 0 1, together with spectroscopic evidence of similar conformational propensities in uncomplexed peptides (Section 4.3) reinforces the notion that this loop may exist as a stable quasi-circular structure in FMDV particles. Its delocalisation in native FMD virions would probably result from some hinge movements about the anchor points of the loop on the capsid surface (Section 4.5). 4.4.3. Induced-Fit Recognition Between Antibodies and G-H Loop Peptides

The structures of the paratopes of antibodies SD6 and 4C4 complexed to the FMDV peptides showed large conformational differences when compared with the unbound structures. This indicates that the antibodies underwent considerable conformational adaptation upon binding to the peptide antigen (Verdaguer et al., 1996). The changes induced by complex formation arose mostly from the flexibility of the complementarity determining regions (CDRs), rather than from the rearrangement of the Fab quaternary structure. The largest changes were concentrated in CDR3 of the heavy chain, where most residues appeared fully reoriented upon complex formation. A similar situation has been reported for other Fab-peptide complexes (Rini et al., 1992; Tormo et al., 1994). The reorientation Copyright © 2004 By Horizon Bioscience

\242

Interaction of FMDV With Antibodies

resulted not only in an alteration of shape but also in an important redistribution of charges, providing multiple points of interaction with the antigens, and in particular with the cell attachment RGD motif (Verdaguer et al., 1996; 1998). The concave paratopes acquired similar shapes and similar distributions of charged and polar groups in the two complexes, resulting in very similar interactions with the peptide antigens. 4.4.4. Structural Description of SD6 and 4C4 Contact Epitopes: Direct Involvement of the Cell-Attachment RGD Triplet and Neighboring Residues

Asp143 in the integrin-binding RGD triplet plays a direct critical role in the recognition of the G-H loop by MAbs SD6 and 4C4 antibodies, reproducing very similar specific contacts with both. Gly 142 contacted the antigen combining site by ~80 % of its surface and the main chain of Arg141 appeared hydrogen bonded with two residues of the CDRs but remained partially exposed to the solvent (Verdaguer et al., 1995; 1998; Ochoa et al., 2000). Leu144, that forms part of the helical segment and is also involved in attachment to cells (Mateu et al., 1996; Jackson et al., 2000), participated with its entire molecular surface in a direct contact with the antibody in both complexes and dominated the hydrophobic interactions. Finally, the non-conserved N-terminal residues Thr137, Ala138, Ser139, and also His146 within the helix (none of which appear to have a role in the attachment of FMDV C-S8cl to BHK-21 cells; Mateu et al., 1996) played other important roles in the peptide-Fab interactions in all three structures determined (Verdaguer et al., 1995; 1998; Ochoa et al., 2000). The most significant difference between the SD6 and the 4C4 complexes was observed for Tyr136, a residue which was in direct contact with antibody SD6 and has not been located in the structure of the 4C4 complex. In addition, Leu147, which is partially involved in hydrophobic interactions with Fab SD6, does not make any contact with Fab 4C4 (Verdaguer et aI., 1995; 1998). Comparison of these results with those of immunochemical studies (Section 4.1) showed that, both for SD6 and for 4C4, the crystallographically defined contact epitope and the immunochemically defined functional epitope match each other with extreme accuracy. Most importantly, the structural results showed a direct participation of cell-attachment residues within the G-H loop of FMDV in strong interactions with antibody molecules and provided a clear structural basis to explain both the functional restrictions to genetic variation within a major antigenic site of FMDV (Section 4.1) and the mechanisms of neutralization by these antibodies (Section 6). 4.5. Cryo-Electron Microscopy Studies of Virion-Antibody Complexes

Cryo-EM and image reconstruction methods allow direct visualization of virusantibody complexes, which are usually too large and unstable to be analyzed at higher resolution by X-ray crystallography. In addition, \vhen high-resolution crystal structures of the virus and the Fab fragment of the antibody are available separately, a pseudoatomic model of the virus-antibody complex can be constructed by docking the atomic structures together using the cryo-EM density maps as a guide (Baker et al., 1999). Cryo-EM images of frozen hydrated complexes between FMDV C-S8cl and the Fab fragments of MAbs SD6 and 4C4 provided a direct picture on how these Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 243 I

(a)

(b)

Figure 5. (a) Isosurface representation of the FMDV C-S8cl-Fab SD6 complex reconstructed from cryo-EM images. The Fab is coloured orange and the viral capsid is depicted in grey. (b) Different orientations of the G-H loop of VPI in the FMDV virion. A lateral view of a crystallographic protomer of FMDV C-S8c 1 is presented, with VPI, VP2 and VP3 depicted in blue, green and red, respectively. The G-H loop is represented in the deduced positions it may adopt in complexes with Fab SD6 (yellow) and Fab 4C4 (light blue). For comparison, the position of the G-H loop in the reduced OlBFS virion is also indicated (deep blue). See Colour Plate at the back of the book.

antibodies interact with the complete FMD virion. The nearly spherical FMDV capsid appeared decorated with 60 Fab molecules that project almost radially from the surface, in orientations which are only compatible with a monovalent attachment of the complete antibodies (Figure 5a) The C-terminal ends of the constant domains of any two Fab molecules bound to the same virion are located far apart and in a wrong orientation to allow their bridging by a Fc fragment (Hewat et al., 1997; Verdaguer et al., 1999). This observation and the biochemical evidence (Section 6) suggests that the complete antibodies may bind monovalently to the FMDV capsid. The known structure of FMDV C-S8c 1 (Lea et aI., 1994) and that of the SD6 and 4C4 Fab fragments co-crystallized with synthetic G-H loop peptides (Verdaguer et al., 1995, 1998; Section 4.3 and Figure 4a) were placed in the reconstructed cryoEM density map of the virus-Fab complex. The best-fitted positions of Fabs onto the viral shell revealed the approximate location of the G-H loop in each complex. The position of this loop in the final docking models differed widely from each other (Figure 5b). In the SD6 complex, the loop was bent towards the viral fivefold axis and approached the B-C loop of VPl and the C-terminus of VPl from a neighbouring protomer, a position close to that proposed for native FMDV OlBFS (Parry et al., 1990; see Section 3.3). In contrast, in the 4C4 complex, the G-H loop was located in a fully exposed position, very loosely connected with the rest of the capsid, and the RGD motif was found at the maximum possible distance from the capsid surface. In both SD6 and 4C4 complexes, the Fab fragments appeared to interact almost exclusively with the G-H loop and made no other contacts with the viral capsid (Hewat et al., 1997; Verdaguer et al., 1999), which provided further proof of the true continuous nature of the epitopes defined by these anti-virus antibodies (see Sections 4.1 and 4.4). The location of the G-H loop in the two complexes differed also from the disposition of the equivalent loop in reduced FMDV OlBFS (Logan et aI., 1993; Copyright © 2004 By Horizon Bioscience

\244

Interaction of FMDV With Antibodies

Lea et al., 1995; Figure 5b). However, a single hinge rotation at the surface of the capsid would allow a transition between the three positions observed (Verdaguer et al., 1999). In the crystal structure of the uncomplexed FMDV C-S8c 1 virion the G-H loop had been found disordered. The cryo-EM results allowed the positioning of the loop in different orientations on the capsid surface, and provided clear evidence that the G-H loop of VPl may act as a mobile structural module. Recent biochemical studies related the mobility of the G-H loop with the ability of the virus to bind integrin avf33 in vitro and to infect cells via integrin receptors (Jackson et al., 2003). These observations, and the striking similarity of the integrin RGD binding motifs in all of the structures determined, suggest that the avf33 receptor may recognize the FMDV G-H loop in a way similar to that exhibited by neutralizing antibodies. 4.6. Conclusion: A Structural and Functional Model of the Antigenic G-H Loop ofFMDV

We will integrate the above results in a model on the interaction with antibodies of the G-H loop of VP 1 as an independent element of the capsid. While many aspects of this model are well founded, a few others are indeed questionable and should be taken as working hypotheses. The G-H loop exists on the FMDV capsid as a loosely connected element that may undergo hinge movements, and that may adopt preferred orientations that depends on the specific virus strain (due to specific sequence and conformational differences, either in the loop itself or in other capsid elements). The G-H loop exhibits structural features that are conserved in FMDV, including an open turn at the integrin-binding RGD triplet and a neighboring helix that contains residues also involved, to some extent, in cell attachment. The intrinsic conformation of this loop is also conserved both in the uncomplexed form and when bound to antibodies. However, the G-H loop in any FMDV isolate may exhibit some limited intrinsic flexibility, and could sample a restricted range of slightly different conformations. In type 0 viruses, the preferred orientation of the G-H loop is close to the VP1 C-terminus, and both elements participate in discontinuous epitopes. In types C and A virions the preferred orientations involve a loose connection with the rest of the capsid, so many virus-induced neutralizing antibodies recognise this loop as an independent unit and define true continuous epitopes. These antibodies make direct contact with a stretch of about ]0-12 contiguous residues. Their binding involves the immobilization of the loop in a particular orientation that depends on the specific antibody bound, perhaps due to sterical and other structural constraints. Enough affinity may be partly achieved through i), a paratope with a concave shape (as also observed with anti-peptide antibodies) to maximize contact area with the protruding, convex G-H loop, and ii), an induced-fit process in which both the antibody paratope and the G-H loop adapt to each other to maximize chemical and geometrical complementarity. Contact and functional epitopes defined in the G-H loop by different MAbs may extensively overlap. However, subtle structural differences in the recognition of the G-H loop by different antibodies may entail big differences in affinity and antigenic specificity. Such differences may include some details in the conformation of the G-H loop and the nature and relative strength of specific non-covalent interactions between any loop residue and the residues in Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2451

different paratopes. As a consequence, the effect of specific replacements in the GH loop on the binding by different antibodies may be quite different. This provides a molecular basis for the widely divergent antigenic specificities of the G-H loop found among different but highly related FMDV field variants. Severe restrictions to sequence variation in the G-H loop do occur because of the dual involvement of some residues, including the RGD triplet and some neighboring residues, in both antibody and cell receptor recognition. The receptor binding sites of other picomaviruses, including rhinovirus and some enteroviruses, are located at the base of a narrow depression or "canyon" running around each five fold axis of the icosahedral particle, and it was postulated that these sites are protected from the host immune surveillance by the inability of neutralizing antibodies to penetrate far into the canyon on account of their larger size (reviewed in Rossmann et al., 2002). The basis of this canyon hypothesis has been challenged because of the subsequent discovery that the footprints of the receptor and antibody binding sites of rhinovirus overlap on the capsid surface (Smith et ai., 1996). In FMDV, the cell attachment site is not shielded from antibody attack, but is protected from variation by the purifying effect of negative selection (Mateu, 1995; Verdaguer et al., 1995; Martinez et al., 1997). 5. Peptide Mimics of Discontinuous FMDV Epitopes Many epitopes of FMDV, including some that involve the G-H loop of VP1, are of a discontinuous nature, and this has severely impaired their detailed structural and functional characterization. However, some studies have started to investigate specific features of discontinuous FMDV epitopes (Section 3.3). A few others have attempted the immunogenic mimicking of discontinuous FMDV epitopes, mainly aimed at the design of better synthetic peptide vaccines (Chapter 12). Some of this work will be briefly considered here. Complete peptide mimics of the discontinuous epitopes formed by the G-H loop and the C-terminus of VPl (as mainly found in type 0 virions) have been attempted. Some of these constructions were based on the covalent linkage of both VPl segments through a flexible (Gly) spacer. A construction based on the OtK sequence proved to be a better immunogen than the separate peptide components, and also conferred protection in cattle (DiMarchi et ai., 1986). Direct binding and inhibition experiments indicated that the antibodies elicited in sheep by such peptide construction recognised discontinuous epitopes that were absent from the constituent VPl segments (Flynn et ai., 1990). An equivalent peptide construction based on the A24Cruzeiro sequence was also protective in cattle. In contrast to that observed with polyclonal antibodies and type 0 peptides, a MAb (C 1.1) elicited against the type A peptide construction was mapped within the G-H loop alone (Barnett et al., 1996). In our view, these and other results may not allow to settle whether the C-terminal segment is providing a part of the anchoring points needed for antibody recognition, or improving the presentation of epitopes entirely located within the G-H loop, or both. Crystallographic studies of a complex between these peptide constructions and anti-peptide, virus-neutralizing MAbs (like C 1.1) may prove of great interest. For type C, no clear improvement in the percentage of bovines protected was obtained with an equivalent peptide construction (that also included a T-cell epitope), when compared with a peptide composed only of the Copyright © 2004 By Horizon Bioscience

\246

Interaction of FMDV With Antibodies

G-H loop and the T-cell epitope (Taboga et al., 1997). Also, a recent type 0 peptide construction that contained a T-cell epitope and a cyclized G-H loop peptide, but not the VP1 C-terminal segment, conferred very good protection against infection by FMDV O}Taiwan (Wang et ai., 2002). A completely different approach resulted in the mimicking of a discontinuous G-H loop epitope by a short peptide that did not reproduce the sequence on the virus (a mimeotope). Such peptide was obtained by a combinatorial a priori approach based on serial affinity selection of the most reactive peptides among many variants (Geysen et al., 1986). Another kind of mimeotope was obtained through the rational design of a complex peptide construction aimed at mimicking site D epitopes ofFMDV C-S8c1 (Villen et al., 2002). This construction was formed by coupling synthetic peptide segments that reproduced parts of the C-terminus of VP1, the B-C loop of VP2, and the stretch around the B-B knob of VP3 of FMDV C-S8c 1, thus including the five residues that were found involved in recognition of this isolate by neutralizing anti-virion antibodies. The structural design was based on the crystal structure of FMDV C-S8c 1 with the aid of molecular dynamics. The theoretical model obtained suggested that the five antigenic residues were located in the same face of the peptide molecule and defined a set of distances and angles similar to those found on the virus capsid. Although these constructions failed to be recognised by anti-virus site D antibodies, they elicited a weak but specific neutralizing response in immunized animals, and partially protected guinea pigs against challenge with the virus (Villen et al., 2002). All of the above peptide constructions may mimic not the contact epitope(s), but just a critical part of the functional, or rather the energetic epitope(s) in the FMD virion. That is, those peptides may exhibit some steric and chemical similarity with the ensemble of residues within the true viral epitope(s) that are energetically most important for antibody binding. This consideration does not minimize the usefulness of such constructions as potential vaccines. On the contrary, it emphazises that, when dealing with the ubiquitous discontinuous epitopes of FMDV (or any other pathogen) as possible components of synthetic vaccines, being able to identify the residues critically involved in the functional epitope, or even just the energetic epitope, and not necessarily the more complex contact epitope, may constitute an important step for a successful design. 6. Neutralization of FMDV Infectivity by Antibodies Antibodies may neutralize virus infectivity by a variety of mechanisms. They may, for example, alter virus stability or prevent uncoating by bivalent binding, or they may sterically prevent attachment to cells, or cause aggregation of virus particles through cross-linking (Dimmock, 1993). There is a limited knowledge on the specific neutralization mechanisms used by anti-FMDV antibodies. An important early finding was that some anti-FMDV MAbs neutralized infectivity mainly through aggregation of viral particles, others acted by blocking virus attachment to cells, and one neutralizing MAb caused little viral aggregation but had no effect on attachment. Thus, antibodies may neutralize FMDV by at least three different mechanisms (Baxt et al., 1984). A different MAb appeared to neutralize FMDV infectivity by causing a change in the conformation of the capsid that facilitated RNA release (McCullough et ai., 1987b). In a more recent study, the wellCopyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 247\

characterized anti-virus MAbs SD6 and 4C4 that bind the G-H loop of VPl were used. The neutralization profiles showed a 90% reduction of viral infectivity over an antibody-to-virus input ratio of ~ 60. These MAbs could achieve a million-fold reduction in viral infectivity, and no substantial residual infectivity (which would be indicative of a weak neutraliser) or V-shaped neutralization curve (indicative of an aggregating antibody) were observed (Verdaguer et al., 1997; 1999). In addition, no significant differences were found when the neutralization activity of the monovalent Fab was compared with those of the bivalent Fab 2 or the intact MAb. Thus, both MAbs appear to be strong neutralizers that bind monovalently to the FMDV capsid, consistent with the cryo-EM observations (Section 4.5). These MAbs bind loop residues also involved in attachment to cells (Section 4). The major mechanism by which these antibodies neutralise FMDV infectivity seems to involve the monovalent steric inhibition of virus binding to the cell receptor. Hinge movements of the FMDV loop may help in accommodating efficient binding by different antibodies. Some MAbs that probably bind the external surface of the FMDV capsid, including the G-H loop, are non-neutralizing (Ouldridge et al., 1984). An intriguing observation that has been made \vith different viruses (Dimmock et al., 1993), including FMDV (Grubman and Morgan, 1986), is that the neutralization activity, but not the binding capacity of some antibodies may be lost upon mutation of the virus. The molecular bases for such observations remain essentially unknown. In addition to direct neutralization of infectivity, other mechanisms of viral clearance may operate in vivo (Dimmock, 1993). For example, it has been observed that anti-FMDV MAbs passively protected mice at dilutions at which they could not neutralize virus infectivity in vitro (McCullough et ai., 1986). Also, sera from guinea pigs immunized with a quintuple MAR mutant derived from OIK offered passive protection against challenge with OtK, despite having minimal neutralizing activity in vitro (Dunn et al., 1998). Much work remains to be done on the mechanisms of neutralization of FMDV infectivity by antibodies, and their role in protection.

7. Molecular Determinants of FMDV Escape From Antibody Recognition in the Field 7.1. Extreme Antigenic Variation in FMDV

Early evidence of extensive antigenic variation among FMDV isolates included the recognition of seven serotypes, more than 65 serological subtypes and many intrasubtype antigenic variants (Pereira, 1977). The advent of MAbs allowed further antigenic differentiation among field isolates (e.g., Grubman and Morgan, 1986; Pfaff et ai., 1988; Mateu et ai., 1988), even from a single outbreak (Mateu et ai., 1987, 1988), which led to the proposal that each FMDV isolate may be unique in its fine epitopic composition (Mateu et ai., 1988). Single isolates also proved antigenically heterogeneous (e.g. Rowlands et al., 1983; Mateu et ai., 1989). Direct evidence of rapid antigenic change, sometimes of a rather drastic nature, has been repeatedly found in animals (Fagg and Hyslop, 1966; Gebauer et al., 1988; Taboga et al., 1997) and in cell culture (Carrillo et al., 1989), even in the absence of antibody pressure (Bolwell et al., 1989b; Dfez et al., 1989; Borrego et ai., 1993; Meyer et al., 1994; Sevilla et ai., 1996; Holguin et al., 1997). Copyright © 2004 By Horizon Bioscience

\248

Interaction of FMDV With Antibodies

The extreme antigenic diversity of FMDV and the rapid antigenic change of the virus in the presence or absence of antibodies are important biological consequences of the quasispecies structure (extreme genetic heterogeneity) and complex evolutionary dynamics of FMDV populations (reviewed in Domingo et aI., 1985; 1990; 1992; 2001; see Chapter 10), and constitute major difficulties in controlling FMD (Pereira et al., 1977). Within an antigenic region, different individual mutations, even of the same residue, may critically affect recognition by most antibodies against that site, or be of no consequence at all (Section 4.1). The effect of multiple replacements is not necessarily additive (Mateu et al., 1992). Single mutations that disrupt some epitopes may generate different epitopes already present in genetically divergent isolates (Hernandez et al., 1992). Mutations outside the contact epitope may affect the antigenicity of the G-H loop (Section 3.3). In view of these and other considerations, one could guess that variations in sequence and antigenicity during evolution of FMDV in the field are completely unpredictable. However, comparative analyses of capsid sequence and reactivity with MAbs of many field isolates have provided evidence of some recurrent selective strategies in a FMDV serotype to escape antibody recognition in the field. A detailed study used a large panel of MAbs elicited against type C 1 or C3 viruses, and many field strains considered as representatives of each subtype and evolutionary subline of serotype C identified in Europe and South America over a period of more than six decades (Mateu et al., 1988; 1994; Martinez et al., 1991; 1992; Feigelstock et aI., 1992; 1996). Some observations derived mainly from that work follo\v. 7.2. A Case Study: Antigenic Variation of Serotype C FMDV in the Field 7.2.1. Lack of Correlation Between Genetic and Antigenic Variation Within a FMDV Serotype Most differences in capsid residues among type C field isolates were found in regions identified as antigenic in FMDV, but the total number of amino acid differences was rather limited. This is remarkable because some of the antigenically most divergent type C viruses known --some of which are taught to have escaped immunity conferred by the vaccines in use-- were included in the sample. Despite the relatively low sequence variation, immunoassays with MAbs against sites A or D revealed extremely varied reactivity patterns. In contrast to that observed between serotypes, no correlation between genetic and antigenic divergence was found within a serotype (type C), even when only antigenic regions were considered (Martinez et al., 1991; 1992; Mateu et al., 1994). 7.2.2. Two Strategies of Antigenic Variation Involving the G-H Loop of FMDV

Sequence and antigenic variation of individual major sites (sites A and D) were analyzed in detail using field viruses, MAR mutants and peptides. Two nonexcluding strategies of antigenic variation in site A (within the G-H loop of VPl) \vere detected (Martinez et aI., 1991). The first consists of a gradual antigenic differentiation ("drift") through the accumulation of amino acid replacements, most of them antigenically non-critical when considered individually (only one or a few epitopes were affected). These replacements occurred mainly in two short hypervariable segments located at either side of the stretch involved in cell

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer

249 I

recognition (the ROD triplet and the neighboring four-residue helix; Sections 4.1 and 4.4). A second strategy consists of an abrupt "shift" in antigenic specificity caused by a replacement of a single amino acid residue within the relatively conserved helical stretch, that abolished or decreased reactivity with most site A MAbs and polyclonal antibodies from natural hosts. Such drastic changes were repeatedly observed in independent field isolates (Martfnez et al., 1991; Feigelstock et al., 1996), in persistently infected cattle (Gebauer et al., 1988), in bovines vaccinated with a G-H loop peptide but not protected against challenge with FMDV (Taboga et al., 1997), in FMDVs serially passaged in cell culture in the absence of antibodies (Sevilla et aI., 1996), and in many MAR mutants selected from different type C isolates using different MAbs (Mateu et al., 1989; 1990; Hernandez et aI., 1992; Martfnez et al., 1997). Most of these changes specifically involved residue 146 of VP 1, the only residue in the helical segment for which a role in attachment to cells was not detected (Mateu et al., 1996). Interestingly, variants harboring such antigenically critical substitutions did not generally become prevalent in the field. It is tempting to speculate that these capsid mutations, located at a functional spot, may entail a reduced fitness for survival in the field. 7.2.3. Antigenic Variation and Restrictions to Genetic Variation in Sites Independent of the G-H Loop

When site D was considered, a somewhat different scenario was found (Mateu et aI., 1994). The sequence heterogeneity of site D elements among type C viruses was above average, but lower than that of site A, as observed also among serotypes. The main-chain conformation of the B-C loop of VP2, a major part of site D, is remarkably conserved between serotypes, and several site-directed mutations in the area equivalent to site D proved lethal in other picornaviruses. Type C FMDV (C-S8c 1) that had been passaged in cell culture in the absence of antibodies readily acquired mutations within the O-H loop ofVPI (Dfez et al., 1989; Borrego et al., 1993; Sevilla et al., 1996; Holgufn et aI., 1997), but no mutations were found in site D elements (Holgufn et al., 1997). Within site D, many of the relatively scarce replacements observed in field variants ocurred at anyone of the five positions found repeatedly mutated in escape mutants of serotype C, or in close positions mutated in escape mutants of other serotypes (Mateu et al., 1994). All these observations together suggested that discontinuous epitopes within site D may be subjected to restrictions to genetic variation, perhaps even more severe than those operating in site A within the G-H loop (Mateu et al., 1994; Lea et aI., 1994; see Mateu, 1995, for a more detailed discussion on the presence of restrictions to variation in antigenic sites of picornaviruses). In addition, only a few critical residues could participate in the functional (or energetic) site D epitopes, and these residues may be more frequently targeted under antibody pressure (Mateu et aI., 1998). Despite its lower sequence variability, site D was found as antigenically variable as site A (Mateu et al., 1994). Thus, a limited number of replacements does not necessarily preclude substantial antigenic divergence. Little or no variation was found in field isolates with two MAbs directed at site C, and no MAbs directed against the B-C loop ofVPl (site 3 in OIK) have been obtained \vith type C viruses. However, these two regions were also relatively hypervariable in sequence among type C field isolates (Martfnez et aI., 1991), and they may have also contributed to antigenic variation within this serotype. Copyright © 2004 By Horizon Bioscience

\250

Interaction of FMDV With Antibodies

7.2.4. Prediction of Antigenically Critical Variations During Evolution of FMDV in the Field

Individual sequence comparisons between antigenically divergent field isolates were also informative. Serological divergence and partial lack of cross protection may be associated with a drastic antigenic change in some, but not necessarily all antigenic sites identified in FMDV (e.g., C3Argentina/84 vs. C3Resende; Feigelstock et al., 1992). Conversely, drastic antigenic variation at a single site does not necessarily lead to serological divergence (e.g., C1Haute Loire vs. other subtype C 1 viruses; Mateu et ai., 1994). Remarkably, some of the most highly divergent field isolates of type C (e.g., CsArgentina/69 vs. the then current vaccine strain, C3Resende, and other type C viruses circulating in South America at the time) had accumulated one or a few substitutions in each of the antigenic sites known; at least some of the replacements found in sites A and D were antigenically critical (Mateu et ai., 1994). The above analyses led to the proposal that substantial antigenic variation of FMDV type C in the field could be repeatedly achieved by a combination of a few mutations in several or all antigenic sites identified in FMDV, and that these mutations may include: i) a number of non-critical mutations at residues 138-140 and 148-150 within site A (the G-H loop of VP1), and/or ii) a critical mutation at residue 146 in the same loop and/or iii) relatively critical mutations at one or more of the residues 74 and 79 of VP2, 58 of VP3 and 193 of VPl within site D (type C numbering). Mutations at residues 146 of VP1, 74 of VP2 and 58 of VP3 were highlighted as particularly critical (Mateu et al., 1994). Later analyses of new field variants that emerged in Argentina during 1993-1994 allowed that prediction to be confirmed. The current type C vaccine strain (C 3Argentina/85) was not effective enough to confer full protection against variants isolated during the new outbreaks. Serological analyses and immunoassays using site A and site D MAbs showed the variants analyzed to be antigenically highly divergent from the vaccine strain. All were found genetically related with the vaccine strain but differed at a number of positions, many of them located in the antigenic sites identified in FMDV. Among the mutated residues in most isolates analyzed were 146 of VPl in site A and 74 and 79 of VP2, 58 of VP3 and 193 of VPl in site D (Feigelstock et al., 1996). Antigenic variation had thus occurred through changes in many of the same critical residues that 35 years earlier led to the emergence of CsArgentina/69 and/or that were also substituted in other antigenically divergent field isolates. However, C3Argentina/85 and the new isolates belonged to a different evolutionary subline than C3Resende and CsArgentina/69. Thus, antigenically critical mutations at a few specific, predicted residues in the FMDV capsid had independently recurred in the field. 7.2.5. Conclusion: High Antigenic Divergence Through Limited Sequence Variation in the FMDV Capsid

The analyses of antigenic variation in type C viruses indicate that, within a serotype of FMDV, antigenically highly divergent viruses can arise in the field by very limited sequence variation at structurally and functionally dispensable, highly exposed key residues located in several or all of the antigenic regions identified in FMDV. Mutations of a few specific residues (including 146 of VP1, 74 of VP2 and 58 of VP3) seem to have an important role for the recurrent, independent emergence of some antigenically highly divergent serotype C FMDVs in the field. Copyright © 2004 By Horizon Bioscience

Mateu and Verdaquer

2511

8. Acknowledgements M.G.M. is indebted to Prof. E. Domingo for the guidance and support received as a former senior associate in Prof. Domingo's laboratory. M.G.M. and N.V. acknowledge Profs. E. Domingo, D. Andreu, L Fita, E. Giralt, A. King and D. Stuart, Drs. E. Baranowski, A. Benito, B. Borrego, E. Brocchi, J. Camarero, C. Escarmis, D. Feigelstock, J. Hernandez, S. Lea, M.A. Martinez, LS. Novella, E.L. Palma, C. Palomo, F. Sobrino, M.L. Valero and A. Villaverde and R. Mateo for collaboration on FMDV studies and for discussions. Work in M.G.M....s laboratory is supported by grants BI02000-0408 and 2003-04445 from Ministerio de Ciencia y Tecnologia (MCyT) and 08.2/0008/2000-2 from Comunidad Autonoma de Madrid, and by an institutional grant from Fundacion Ramon Areces. Work in N.V. 's laboratory is supported by grants BI0099-065 and BI02002-00517 from MCYT. References Abrams, C.C, King, A.M., and Belsham, G.J. 1995. Assembly of foot-and-mouth disease virus empty capsids synthesized by a vaccinia virus expression system. J. Gen. Virol. 76:3089-3098. Acharya, R., Fry, E., Stuart, D., Fox, G., Rowlands, D. and Brown, F. 1989. The threedimensional structure of foot-and-mouth disease virus at 2.9 A resolution. Nature 337: 709-716. Aggarwal, N. and Barnett, P.V. 2002. Antigenic sites of foot-and-mouth disease virus (FMDV): an analysis of the specificities of anti-FMDV antibodies after vaccination of susceptible host species. J. Gen. Virol. 83: 775-782. Air, G.M., Laver, W.G. and Webster, R.G. 1990. Mechanism of antigenic variation in an individual epitope on influenza virus N9 neuraminidase. J.Virol. 64: 5797-5803. Aktas, S. and Samuel, A.R. 2000. Identification of antigenic epitopes on the foot-and-mouth disease virus isolate 01/Manisa/Turkey/69 using monoclonal antibodies. Rev. Sci. Tech. 19:744-753. Bachrach, H.L., Moore, D.M., McKercher P.O. and Polatnick J. 1975. Immune and antibody responses to an isolated capsid protein of foot-and-mouth disease virus. J. Immunol. 115: 1636-1641. Bachrach, H.L., Morgan, D.O. and Moore, D.M. 1979. Foot-and-mouth disease virus immunogenic capsid protein VPt: N-terminal sequences and immunogenic peptides obtained by CNBr and tryptic cleavages. Intervirology 12: 65-72. Baker, T.S., Olson, N.H. and Fuller, S.D. 1999. Adding the third dimension to virus life cycles: three-dimensional reconstruction of icosahedral viruses from cryo-electron micrographs. Microbial. Mol. BioI. Rev. 63: 862-922. Baranowski, E., Ruiz-Jarabo, C.M., Sevilla, N., Andreu, D., Beck, E. and Domingo, E. 2000. Cell recognition of foot-and-mouth disease virus that lacks the RGD integrin- binding motif: flexibility in aphthovirus receptor usage. J. Virol. 74: 1641-1647. Barnett, P.V., Pullen, L., Staple, R.F., Lee, L.I., Butcher, R., Parkinson, D. and Doel, T.R. 1996. A protective anti-peptide antibody against the immunodominant site of the A24 Cruzeiro strain of foot-and-mouth disease virus and its reactivity with other subtype viruses containing the same minimum binding sequence. J. Gen. Virol. 77: 1011-1018. Barnett, P.V., Samuel, A.R., Pullen, L., Ansell, D., Butcher, R.N. and Parkhouse, R.M. 1998. Monoclonal antibodies, against 01 serotype foot-and-mouth disease virus, from a natural bovine host, recognize similar antigenic features to those defined by the mouse. J.Gen.Virol. 79: 1687-1697. Baxt, B., Morgan, D.O., Robertson, B.R. and Timpone, C.A. 1984. Epitopes on foot-andmouth disease virus outer capsid protein involved in neutralization and cell attachment. I.Virol. 51: 298-305. Copyright © 2004 By Horizon Bioscience

\252 Interaction of FMDV With Antibodies

Baxt, B., Vakharia, V., Moore, D.M., Franke, AJ. and Morgan, D.O. 1989. Analysis of neutralizing antigenic sites on the surface of type A I2 foot-and-mouth disease virus. J.Virol. 63: 2143-2151. Benito, A., Mateu, M.G. and Villaverde, A. 1995. Improved mimicry of a foot-and-mouth disease virus antigenic site by a viral peptide displayed on ~-galactosidase surface. Bio/Technology 13: 801-804. Bergmann, I.E., Tiraboschi, B., Mazzuca, G., Fernandez, E., Michailoff, C.A., Scodeller, E.A. and La Torre, J.L.1988. Serological and biochemical analysis of foot-and-mouth disease virus (serotype C3) isolated in Argentina between 1981 and 1986. Vaccine 6: 245-252. Berinstein, A., Roivainen, M., Hovi, T., Mason, P.W. and Baxt, B. 1995. Antibodies to the vitronectin receptor (integrin avp3) inhibit binding and infection of foot-and-mouth disease virus to cultured cells. J.Virol. 69: 2664-2666. Bittle, J.L., Houghten, R.A., Alexander, H., Shinnick, T.M., Sutcliffe, J.G., Lerner, R.A., Rowlands, DJ. and Brown, F. 1982. Protection against foot-and-mouth disease by immunizaton with a chemically synthesized peptide predicted from the viral nucleotide sequence. Nature 298: 30-33. Bolwell, C., Clarke B.E., Parry, N.R., Ouldridge, E.J., Brown, F. and Rowlands, DJ. 1989a. Epitope mapping of foot-and-mouth disease virus with neutralizing monoclonal antibodies. J.Gen.Virol. 70: 59-68. Bolwell, C., Brown, A.L., Barnett, P.V., Campbell, R.O., Clarke, B.E., Parry, N.R., Ouldridge, E.J., Brown, F. and Rowlands, DJ. 1989b. Host cell selection of antigenic variants of foot-and-mouth disease virus. J.Gen.Virol. 70: 45-57. Borrego, B., Novella, I.S., Andreu, D., Giralt, E. and Domingo, E. 1993. Distinct repertoire of antigenic variants in the presence or absence of immune selection. J.Virol. 67: 60716079. Briand, J.P., Benkirane, N., Guichard, G., Newman, J.F., Van Regenmortel, M.H., Brown, F. and Muller, S. 1997. A retro-inverso peptide corresponding to the G-H loop of foot-andmouth disease virus elicits high levels of long-lasting protective neutralizing antibodies. Proc. Natl. Acad. Sci. USA 94: 12545-12550. Brown, F. 1990. Picomaviruses. In: Immunochemistry of Viruses, Vol. II. The Basis for Serodiagnosis and Vaccines. M.H. V. Van Regenmortel and A.R. Neurath, eds. Elsevier, Amsterdam. p. 153-169. Bro\vn, F., Benkirane, N., Limal, D., Halimi, H., Newman, J.F., Van Regenmortel, M.H., Briand, J.P. and Muller, S. 1999. Delineation of a neutralizing subregion within the immunodominant epitope (GH loop) of foot-and-mouth disease virus VPl which does not contain the RGD motif. Vaccine 18: 50-56. Butchaiah, G. and Morgan, D.O. 1997. Neutralization antigenic sites on type Asia-l footand-mouth disease virus defined by monoclonal antibody-resistant variants. Virus Res. 52: 183-194. Camarero, J.A., Andreu, D., Cairo, J.J., Mateu, M.G., Domingo, E., and Giralt, E. 1993. Cyclic disulfide model of the major antigenic site of serotype C foot-and-mouth disease virus: synthetic, conformational and immunochemical studies. FEBS Letters. 328: 159164. Carrillo, E.C., Rieder Rojas, E., Cavallaro, L., Schiapacassi, M. and Campos, R. 1989. Modification of foot-and-mouth disease virus after serial passages in the presence of antiviral polyclonal sera. Virology 171: 599-601. Churchill, M.E., Stura, E.A., Pinilla, C., Appel, J.R., Houghten, R.A., Kono, D.H., Balderas, R.S., Fieser, G.G., Schulze-Gahmen, U. and Wilson, LA. 1994. Crystal structure of a peptide complex of anti-influenza peptide antibody Fab 26/9. Comparison of two different antibodies bound to the same peptide antigen. J. Mol. BioI. 241: 534-556.

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaquer 2531

Clarke, B.E., Carroll, A.R., Rowlands, D.J., Nicholson, B.H., Houghten, R.A., Lerner, R.A. and Brown, F. 1983. Synthetic peptides mimic subtype specificity of foot-and-mouth disease virus. FEBS Letters 157: 261-264. Crowther, J.R., Farias, S., Carpenter, W.C. and Samuel, A.R. 1993. Identification of a fifth neutralizable site on type 0 foot-and-mouth disease virus following characterization of single and quintuple monoclonal antibody escape mutants. J.Gen.Virol. 74: 15471553. Curry, S., Fry, E., Blakemore, W., Abu-Ghazaleh, R., Jackson, T., King, A., Lea, S., Newman, J., Rowlands, D. and Stuart, D. 1996. Perturbations in the surface structure of A22 Iraq foot-and-mouth disease virus accompanying coupled changes in host cell specificity and antigenicity. Structure 4: 135-145. de Prat-Gay, G. 1997. Conformational preferences of a peptide corresponding to the major antigenic determinant of foot-and-mouth disease virus: implications for peptide vaccine approaches. Arch. Biochem. Biophys. 341: 360-369. Diez, J., Mateu, M.G. and Domingo, E. 1989. Selection of antigenic variants of foot-andmouth disease virus in the absence of antibodies, as revealed by an in situ assay. J. Gen. Virol. 70: 3281-3289. DiMarchi, R., Brooke, G., Gale, C., Cracknell, V., Doel, T.R. and Mowat, N. (1986). Protection of cattle against foot-and-mouth disease by a synthetic peptide. Science 232: 639-641. Dimmock, N.J. 1993. Neutralization of animal viruses. Curro Top. Microbiol. Immunol. 183. Domingo, E., Martinez-Salas, E., Sobrino, F., de la Torre, J.C., Portela. A., Ortin, J., L6pezGalfndez, C., Perez-Breiia, P., Villanueva, N., Najera, R., VandePol, D., Steinhauer, D., DePolo, N. & Holland, J. 1985. The quasispecies (extremely heterogeneous) nature of viral RNA genome populations: biological relevance-- a review. Gene 40: 1-8. Domingo, E., Mateu, M.G., Martinez, M.A., Dopazo, J., Moya, A. and Sobrino, F. 1990. Genetic variability and antigenic diversity of foot-and-mouth disease virus. In: Applied Virology Research, Vol. 2. E. Kurstak, R.G.Marusyk, F.A.Murphy and M.H.V.Van Regenmortel, eds. Plenum Press, New York. p. 233-266. Domingo, E., Escarmis, C., Martinez, M.A., Martinez-Salas, E. and Mateu, M.G. 1992. Foot-and-mouth disease virus populations are quasispecies. Curro Top. Microbiol. Immunol. 176: 33-47. Domingo, E., Biebricher, C.K., Eigen, M. and Holland, J.J. 2001. Quasispecies and RNA virus evolution: principles and consequences. Landes Bioscience, Georgetown, Texas, USA. Dunn, C.S., Samuel, A.R., Pullen, L.A. and Anderson, J. 1998. The biological relevance of virus neutralisation sites for virulence and vaccine protection in the guinea pig model offoot-and-mouth disease. Virology 247: 51-61. Fagg, R.H. and Hyslop, N.S.G. (1966). Isolation of a variant strain of foot-and-mouth disease virus (type 0) during passage in partially immunized cattle. Hyg. 64: 397-404. Feigelstock, D., Mateu, M.G., Piccone, M.E., De Simone, F., Brocchi, E., Domingo, E. and Palma, E.L. 1992. Extensive antigenic diversification of foot-and-mouth disease virus by amino acid substitutions outside the major antigenic site. J. Gen.Virol. 73: 33073311. Feigelstock, D., Mateu, M.G., Valero, M.L., Andreu, D., Domingo, E. and Palma, E.L. 1996. Emerging foot-and-mouth disease virus variants with antigenically critical substitutions predicted by model studies using reference viruses. Vaccine 14: 97-102. Feliu, J.X., Benito, A., Oliva, B., Aviles, F.X. and Villaverde, A. 1998. Conformational flexibility in a highly mobile protein loop of foot-and-mouth disease virus: distinct structural requirements for integrin and antibody binding. J. Mol. BioI. 283: 331-338.

Copyright © 2004 By Horizon Bioscience

\254

Interaction of FMDV With Antibodies

Flynn, J.N., Harkiss, G.D., Doel, T. and DiMarchi, R. 1990. Analysis of immune responses in the sheep to synthetic peptides of foot-and-mouth disease virus using ovine polyclonal and monoclonal antibodies. Immunology 69: 1-7. Fox, G., Parry, N., Barnett, P.V., McGinn, B., Rowlands, DJ. and Brown, F. 1989. The cell attachment site on foot-and-mouth disease virus includes the amino acid sequence RGD (arginine-glycine-aspartic acid). J.Gen.Virol. 70: 625-637. France, L.L., Piatti, P.G., Newman, J.F.E., Toth, I., Gibbons, W.A. and Brown, F. 1994. Circular dichroism, molecular modeling, and serology indicate that the structural basis of antigenic variation in foot-and-mouth disease virus is a-helix formation. Proc. Natl. Acad. Sci. USA 91: 8442-8446. Furrer, J., Piotto, M., Bourdonneau, M., Limal, D., Guichard, G., Elbayed, K., Raya, J., Briand, J.P. and Bianco, A. 2001. Evidence of secondary structure by high-resolution magic angle spinning NMR spectroscopy of a bioactive peptide bound to different solid supports. J. Am. Chern. Soc. 123: 4130-4138. Gebauer, F., de la Torre, J.C., Gomes, I., Mateu, M.G., Barahona, H., Tiraboschi, B., Bergmann, I., Auge de Mello, P. and Domingo, E. 1988. Rapid selection of genetic and antigenic variants of foot-and-mouth disease virus during persistence in cattle. J. Virol. 62: 2041-2049. Getzoff, E.D., Tainer, J.A., Lerner, R.A. and Geysen, H.M. 1988. The chemistry and mechanism of antibody binding to protein antigens. Adv. Immunol. 43: 1-98. Geysen, H.M., Meloen, R.H. and Barteling S.J. 1984. Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natl. Acad. Sci. USA 81: 3998-4002. Geysen, H.M., Barteling, SJ. and Meloen, R.H. 1985. Small peptides induce antibodies with a sequence ans structural requirement for binding antigens comparable to antibodies raised against the native protein. Proc. Natl. Acad. Sci.USA 82: 178-182. Geysen, H.M., Rodda, SJ. and Mason, TJ. 1986. A priori delineation of a peptide which mimics a discontinuous antigenic determinant. Mol. Immunol. 23: 709-715. Ghiara, J.B., Stura, E.A., Stanfield, R.L., Profy, A.T. and Wilson, LA. 1994. Crystal structure of the principal neutralization site of HIV- I. Science 264: 82-85. Grubman, M.J. and Morgan, D.O. 1986. Antigenic comparison of foot-and-mouth disease virus serotypes with monoclonal antibodies. Virus Res. 6: 33-43. Haack, T., Camarero, 1.A., Roig, X., Mateu, M.G., Domingo, E., Andreu, D. and Giralt, E. 1997. A cyclic disulfide peptide reproduces in solution the main structural features of a native antigenic site of foot-and-mouth disease virus. Int. J. BioI. Macromol. 20: 209-219. Harrison, S.C. 1989. Picornaviruses. Finding the receptors. Nature. 1989 338(6212): 205206. Hernandez, J., Martfnez, M.A., Rocha, E., Domingo, E. and Mateu, M.G. 1992. Generation of a subtype-specific neutralization epitope in foot-and-mouth disease virus of a different subtype. J. Gen. Virol. 73: 213-216. Hernandez, J., Valero, M.L., Andreu, D., Domingo, E. and Mateu, M.G. 1996. Cleavage at the RGD sequence of a serotype C foot-and-mouth disease virus does not abolish antibody recognition or cell attachment. J. Gen. Virol. 77: 257-264. Hewat, E., Verdaguer, N., Fita, I., Blakemore, W., Brooks, S., King, A., Newman, 1., Domingo, E., Mateu, M.G. and Stuart, D. 1997. Structure of the complex of a Fab fragment of a neutralising antibody with foot-and-mouth disease virus. Positioning of a highly mobile loop. EMBO J. 16: 1492-1500. Holgufn, A., Hernandez, J., Martinez, M.A., Mateu, M.G. and Domingo, E. 1997. Differential restrictions to antigenic variation among antigenic sites of foot-and-mouth disease virus in the absence of antibody selection. J. Gen. Virol. 78: 601-609. Jackson, T., Sharma, A., Ghazaleh, R.A., Blakemore, W.E., Ellard, F.M., Simmons, D.L., Newman, J.W., Stuart, DJ. and King, A.M. 1997. Arginine-glycine-aspartic acidCopyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2551

specific binding by foot-and-mouth disease viruses to the purified integrin avp3 in vitro. J. Virol. 71: 8357-8361. Jackson, T., Blakemore, W., Newman, J.W., Knowles, N.J., Mould, A.P., Humphries, M.J. and King, A.M. 2000. Foot-and-mouth disease virus is a ligand for the high-affinity binding conformation of integrin av(31: influence of the leucine residue within the RGDLmotifon selectivity ofintegrin binding. J.Gen.Virol. 81: 1383-1391. Jackson, T., King, A.M., Stuart, D.I., and Fry, E. 2003. Structure and receptor binding.Virus Res. 2003 91:33-46. Jin, L., Fendly, B.M. and Wells, J.A. 1992. High resolution functional analysis of antibodyantigen interactions. J. Mol. BioI. 226: 851-865. Kaaden, O.R., Adam, K.-H. and Strohmaier, K. 1977. Induction of neutralizing antibodies and immunity in vaccinated guinea pigs by cyanogen bromide peptides of VP3 of footand-mouth disease virus. J. Gen. Virol.34: 397-400. Kitson, J.D.A., McCahon, D. and Belsham, G.l. 1990. Sequence analysis of monoclonal antibody resistant mutants of type 0 foot-and-mouth disease virus: evidence for the involvement of the three surface exposed capsid proteins in four antigenic sites. Virology 179: 26-34. Kitson, J.D., Burke, K.L., Pullen, L.A., Belsham, G.J. and Almond, l.W. 1991. Chimeric polioviruses that include sequences derived from two independent antigenic sites of foot-and-mouth disease virus (FMDV) induce neutralizing antibodies against FMDV in guinea pigs. J. Virol. 65: 3068-3075. Krebs, 0., Ahl, R., Straub, O.C. and Marquardt, O. 1993. Amino acid changes outdide the GH loop of capsid protein VPl of type 0 foot-and-mouth disease virus confer resistance to neutralization by antipeptide G-H serum. Vaccine 11: 359-362. Laporte, J., Grosclaude, J., Wantyghem, J., Bernard, S. and Rouze, P. 1973. Neutralisation en culture cellulaire du pouvoir infectieux du virus de la fievre aphteuse par des serums provenant de porcs immunises a l'aide d'une proteine virale purifiee. Comptes Rendues Hebdomadaires des Seances de I'Academie des Sciences, serie D 276: 2299-3402. Laver, W.G., Air, G.M., Webster, R.G. and Smith-Gill S.J. 1990. Epitopes on protein antigens: misconceptions and realities. Cell 61: 553-556. Lea, S., Hernandez, J., Blakemore, W., Brocchi, E., Curry, S., Domingo, E., Fry, E., AbuGhazaleh, R., King, A.M.Q., Newman, J., Stuart, D. and Mateu, M.G. 1994. The structure and antigenicity of a serotype C foot-and-mouth disease virus. Structure 2: 123-139. Lea, S., Abu-Ghazaleh, R., Blakemore, W., Curry, S., Fry, E., Jackson, T., King, A., Logan, D., Newman, J. and Stuart, D. 1995. Structural comparison of two strains of foot.andmouth disease virus subtype Oland a laboratory antigenic variant, G67. Structure 3: 571-580. Leippert, M., Beck, E., Weiland, F. and Pfaff, E. 1997. Point mutations within the (3G-(3H loop of foot-and-mouth disease virus 01K affect virus attachment to target cells. J. Virol. 71: 1046-1051. Logan, D., Abu-Ghazaleh, R., Blakemore, W., Curry, S., Jackson, T., King, A., Lea, S., Lewis, R., Newman, J., Parry, N., Rowlands, D., Stuart, D. and Fry, E. 1993. Structure of a major immunogenic site of foot-and-mouth disease virus. Nature 362: 566-568. Marquardt, 0., Rahman, M.M. and Freiberg, B. 2000. Genetic and antigenic variance of foot-and-mouth disease virus type Asia-I. Arch. Virol. 145: 149-157. Martinez, M.A., Hernandez, J., Piccone, M.E., Palma, E.L., Domingo, E., Knowles, N. and Mateu, M.G. 1991. Two mechanisms of antigenic diversification of foot-and-mouth disease virus. Virology 184: 695-706. Martinez, M.A., Dopazo, J., Hernandez, J., Mateu, M.G., Sobrino, F., Domingo, E. and Knowles, N.J. 1992. Evolution of the capsid protein genes of foot-and-mouth disease virus: antigenic variation without accumulation of amino acid substitutions over six decades. J.Virol. 66: 3557-3565. Copyright © 2004 By Horizon Bioscience

\256

Interaction of FMDV With Antibodies

Martinez, M.A., Verdaguer, N., Mateu, M.G. and Domingo, E. 1997. Evolution subverting essentiality: Dispensability of the cell attachment Arg-Gly-Asp motif in multiply passaged foot-and-mouth disease virus. Proc. Natl. Acad. Sci. USA 94: 6798-6802. Mason, P.W., Rieder, E. and Baxt, B. 1994. RGD sequence of foot-and-mouth disease virus is essential for infecting cells via the natural receptor but can be bypassed by an antibody-enhancement pathway. Proc. Natl. Acad. Sci. USA 91: 1932-1936. Mateu, M.G. 1995. Antibody recognition of picornaviruses and escape from neutralization: a structural view. Virus Res. 38: 1-24. Mateu, M.G., Rocha, E., Vicente, 0., Vayreda, F., Navalpotro, C., Andreu, D., Pedroso, E., Giralt, E. Enjuanes, L. and Domingo, E. 1987. Reactivity with monoclonal antibodies of viruses from an episode of foot-and-mouth disease. Virus Res. 8: 261-274. Mateu, M.G., Da Silva, J.L., Rocha, E., De Brum, D.L., Alonso, A., Enjuanes, L., Domingo, E. and Barahona, H. 1988. Extensive antigenic heterogeneity of foot-and-mouth disease virus of serotype C. Virology 167: 113-124. Mateu, M.G., Martinez, M.A., Rocha, E., Andreu, D., Parejo, J., Giralt, E., Sobrino, F. and Domingo, E. 1989. Implications of a quasispecies genome structure: effect of frequent, naturally occurring amino acid substitutions on the antigenicity of foot-and-mouth disease virus. Proc. Natl. Acad. Sci. USA 86: 5883-5887. Mateu, M.G., Martinez, M.A., Capucci, L., Andreu, D., Giralt, E., Sobrino, F., Brocchi, E. and Domingo, E. 1990. A single amino acid substitution affects multiple overlapping epitopes in the major antigenic site of foot-and-mouth disease virus of serotype C. J. Gen. Virol. 71: 629-637. Mateu, M.G., Andreu, D., Carreno, C., Roig, X., Cairo, J.J. Camarero, J.A., Giralt, E. and Domingo, E. 1992. Non-additive effects of multiple amino acid substitutions on antigen-antibody recognition. Eur. J. Immunol. 22: 1385-1389. Mateu, M.G., Hernandez, J., Martinez, M.A., Feigelstock, D., Lea, S., Perez, J.1., Giralt, E., Stuart, D., Palma, E.L. and Domingo, E. 1994. Antigenic heterogeneity of a foot-andmouth disease virus serotype in the field is mediated by very limited sequence variation at several antigenic sites. J. Virol. 68: 1407-1417. Mateu, M.G., Camarero, J.A., Andreu, D., Giralt, E. and Domingo, E. 1995a. Direct evaluation of the immunodominance of a major antigenic site of foot-and-mouth disease virus in a natural host. Virology 205: 298-306. Mateu, M.G., Andreu, D. and Domingo, E. 1995b. Antibodies raised in a natural host and monoclonal antibodies recognize similar antigenic features of foot-and-mouth disease virus. Virology 210: 120-127 Mateu, M.G., Valero, M.L., Andreu, D. and Domingo, E. 1996. Systematic replacement of amino acid residues within an Arg-Gly-Asp-containing loop offoot-and-mouth disease virus and effect on cell recognition. J. BioI. Chern. 271: 12814-12819. Mateu, M.G., Escarmfs, C. and Domingo, E. 1998. Mutational analysis of discontinuous epitopes of foot-and-mouth disease virus using an unprocessed capsid protomer precursor. Virus Res. 53: 27-37. McCahon, D., Crowther, J.R., Belsham, GJ., Kitson, J.D.A., Duchesne, M., Have, P., Meloen, R.H., Morgan, D.O. and de Simone, F. 1989. Evidence of at least four antigenic sites on type 0 foot-and-mouth disease virus involved in neutralization; identification by single and multiple site monoclonal antibody-resistant mutants. lGen.Viroi. 70: 639-645. McCullough, K.C., Crowther, J.R., Butcher, R.N., Carpenter, W.C., Brocchi, E., Capucci, L. and De Simone, F. 1986. Immune protection of foot-and-mouth disease virus studied using virus-neutralizing and non-neutralizing concentrations of monoclonal antibodies. Immunology 58: 421-428. McCullough, K.C., Crowther, J.R., Carpenter, W.C., Brocchi, E., Capucci, L., De Simone, F., Xie, Q. and McCahon, D. 1987a. Epitopes on foot-and-mouth disease virus particles. Virology 157: 516-525. Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer

2571

McCullough, K.C., Smale, CJ., Carpenter, W.C., Crowther, J.R., Brocchi, E. and De Simone, F. 1987b. Conformational alteration in foot-and-mouth disease virus virion capsid structure after complexing with monospecific antibody. Immunology 60: 75-82. Meloen, R.H., Puyk, W.C., Meijer, D.J.A., Lankhof, H., Posthumus, W.P.A., and Schaaper, W.M.. 1986. The use of peptides to locate epitopes on the surface of footh-and-mouth disease. Protides BioI. Fluids Proc. Collog.34: 103-106. Meyer, R.F., Pacciarini, M., Hilyard, EJ., Ferrari, S., Vakharia, V.N., Donini, G., Brocchi, E. and Molitor, T.W. 1994. Genetic variation of foot-and-mouth disease virus from field outbreaks to laboratory isolation. Virus Res. 32: 299-312. Minor, P.o. 1990. Antigenic structure of picornaviruses. Curro Top. Microbiol. Immunol. 161: 121-154. Muller, S., Guichard, G., Benkirane, N., Brown, F., Van Regenmortel, M.H. and Briand, J.P. 1995. Enhanced immunogenicity and cross-reactivity of retro-inverso peptidomimetics of the major antigenic site of foot-and-mouth disease virus. Pept. Res. 8: 138-144. Nair, D.T., Singh, K., Sahu, N., Rao, K.V. and Salunke, D.M. 2000. Crystal structure of an antibody bound to an immunodominant peptide epitope: novel features in peptideantibody recognition. J. Immunol. 165: 6949-6955. Nair, D.T., Singh, K., Siddiqui, Z., Nayak, B.P., Rao, K.V. and Salunke, D.M. 2002. Epitope recognition by diverse antibodies suggest conformational convergence in an antibody response. J.Immunol. 168: 2371-2382. Novella, I.S., Borrego, B., Mateu, M.G., Domingo, E., Giralt, E. and Andreu, D. 1993. Use of substituted and tandem-repeated peptides to probe the relevance of the highly conserved RGD tripeptide in the immune response against foot-and-mouth disease virus. FEBS Letters 330: 253-259. Novotny, J., Bruccoleri, R.E. and Saul, F. 1989. On the attribution of binding energy in antigen-antibody complexes Mcpc 603, 01.3 and Hy-hel-5. Biochemistry 28: 47354749. Ochoa, W.F., Kalko, S., Mateu, M.G., Gomes, P., Andreu, D., Domingo, E., Fita, I. and Verdaguer, N. 2000. Structure of a multiple substituted G-H loop from foot-and-mouth disease virus in complex with a neutralising antibody. J.Gen.Virol. 81: 1495-1505. Ouldridge, E.J., Barnett, P.V., Parry, N.R., Syred, A., Head, M. and Rweyemamu, M.M.1984. Demonstration of neutralizing and non-neutralizing epitopes on the trypsin-sensitive site of foot-and-mouth disease virus. J.Gen.Virol. 65: 203-207. Page, G.S., Mosser, A., Hogle, J.M., Filman, D.J., Rueckert, R.R. and Chow, M. 1988. Three-dimensional structure of poliovirus serotype 1 neutralizing determinants. J. Virol. 62: 1781-1794. Parry, N.R., Ouldridge, E.J., Barnett, P.V., Rowlands, D.J. ,Brown, F., Bittle, J.L., Houghten, R.A. and Lerner, R.A. 1985. Identification of neutralizing epitopes of foot-and-mouth disease virus. In: Vaccines 85. Cold Spring Harbor Laboratory. p. 211-216. Parry, N.R., Syred, A., Rowlands, D.J. and Brown, F. 1988. A high proportion of antipeptide antibodies recognize foot-and-mouth disease virus particles. Immunology 64: 567-572. Parry, N.R., Barnett, P.V., Ouldridge, E.J., Rowlands, D.J. and Brown, F. 1989. Neutralizing epitopes of type 0 foot-and-mouth disease virus. II. Mapping three conformational sites with synthetic peptide reagents. J.Gen.Virol. 70: 1493-1503. Parry, N., Fox, G., Rowlands, D., Brown, F., Fry, E., Acharya, R., Logan, D. and Stuart, D. 1990. Structural and serological evidence for a novel mechanism of antigenic variation in foot-and-mouth disease virus. Nature 347: 569-571. Pegna, M., Molinari, H., Zetta, L., Gibbons, W.A., Brown, F., Rowlands, D., Siligardi, G. and Mascagni, P. 1996a. The solution structure of the immunodominant aand cell receptor binding regions of foot-and-mouth disease virus serotype A, variant A. J.Pept.Sci. 2: 75-90.

Copyright © 2004 By Horizon Bioscience

\258

Interaction of FMDV With Antibodies

Pegna, M., Molinari, H., Zetta, L., Melacini, G., Gibbons, W.A., Brown, F., Rowlands, D., Chan, E. and Mascagni, P. 1996b. The solution conformational features of two highly homologous antigenic peptides of foot-and-mouth disease virus serotype A, variant A and USA, correlate with their serological properties. J. Pept. Sci. 2: 91-105. Pereira, H.G. 1977. Subtyping of foot-and-mouth disease virus. Dev. BioI. Stand. 35: 167174. Petit, M.C., Benkirane, N., Guichard, G., Du, A.P., Marraud, M., Cung, M.T., Briand, 1.P. and Muller, S. 1999. Solution structure of a retro-inverso peptide analogue mimicking the foot-and-mouth disease major antigenic site. Structural basis for its antigenic crossreactivity with the parent peptide. 1. BioI. Chern. 274: 3686-3692. Pfaff, E., Mussgay, M., Bohm, H.O., Schulz, a.E. and Schaller, H. 1982. Antibodies against a preselected peptide recognise and neutralize foot-and-mouth disease virus. EMBO J. 1: 869-874 Pfaff, E., Thiel, H.-J., Beck, E., Strohmaier, K. and Schaller, H. 1988. Analysis of neutralizing epitopes on foot-and-mouth disease virus. J.Virol. 62: 2033-2040. Rieder, E., Baxt, B. and Mason, P.W. 1994a. Animal-derived antigenic variants of footand-mouth disease virus type A12 have low affinity for cells in culture. J. Virol. 68: 5296-5299. Rieder, E., Baxt, B., Lubroth, J. and Mason, P.W. 1994b. Vaccines prepared from chimeras offoot-and-mouth disease virus (FMDV) induce neutralizing antibodies and protective immunity to multiple serotypes of FMDV. 1. Virol. 68: 7092-7098. Rini, J.M., Schulze-Gahmen, U. and Wilson, LA. 1992. Structural evidence for induced fit as a mechanism for antibody-antigen recognition. Science 255: 959-965. Rossmann, M.G., He, Y. and Kuhn, R.J. 2002. Picornavirus-receptor interactions. Trends Microbiol. 10: 324-331. Rowlands, D.J., Sangar, D.V. and Brown, F. 1971. Relationship of the antigenic structure of foot-and-mouth disease virus to the process of infection. J. Gen. Virol. 13: 85-93. Rowlands, D.J., Clarke, B.E., Carroll, A.R., Brown, F., Nicholson, B.H., Bittle, J.L., Houghten, R.A. and Lerner, R.A. 1983. Chemical basis of antigenic variation in footand-mouth disease virus. Nature 306: 694-697. Ruiz-Jarabo, C.M., Sevilla, N., Davila, M., G6mez-Mariano, G., Baranowski, E. and Domingo, E. 1999. Antigenic properties and population stability of a foot-and-mouth disease virus with an altered Arg-Gly-Asp receptor recognition motif. J. Gen. Virol. 80: 1899-1909. Saiz, J.C., Gonzalez, MJ., Borca, M.V., Sobrino, F. and Moore, D.M. 1991. Identification of neutralizing antigenic sites on VPI and VP2 of type A5 foot-and-mouth disease virus, defined by neutralization-resistant variants. J. Virol. 65: 2518-2524. Saiz, J.C., Cair6, J., Medina, M., Zuidema, D., Abrams, C., Belsham, G.J., Domingo, E. and Vlak,1.M. 1994. Unprocessed foot-and-mouth disease virus capsid precursor displays discontinuous epitopes involved in viral neutralization. J. Virol. 68: 4557-4564. Sanyal A., Venkataraman, R., Pattnaik B. 1997. Antigenic features of foot-and-mouth disease virus serotype Asia 1 as revealed by monoclonal antibodies and neutralizationescape mutants. Virus Res. 50: 107-117. Saphire, E.O., Parren, P.W., Pantophlet, R., Zwick, M.B., Morris, G.M., Rudd P.M., Dwek R.A., Stanfield R.L., Burton D.R. and Wilson LA. 2001. Crystal structure of a neutralizing human IGG against HIV-l: a template for vaccine design. Science 293: 1155-1159. Sevilla, N., Verdaguer, N. and Domingo, E. 1996. Antigenically profound amino acid substitutions occur during large population passages of foot-and-mouth disease virus. Virology 225: 400-405. Sharma, A., Rao, Z., Fry, E., Booth, T., Jones, E.Y., Rowlands, D.J., Simmons, D.L. and Stuart, D.I. 1997. Specific interactions between human integrin avf33 and chimeric

Copyright © 2004 By Horizon Bioscience

Mateu and Verdaguer 2591

hepatitis B virus core particles bearing the receptor-binding epitope of foot-and-mouth disease virus. Virology 239: 150-157. Siligardi, G., Drake, A.F., Mascagni, P., Rowlands, D., Brown, F. and Gibbons, W.A. 1991. Correlations between the conformations elucidated by CD spectroscopy and the antigenic properties of four peptides of the foot-and-mouth disease virus. Eur.J.Biochem. 199: 545-551. Smith, T.J., Chase, E.S., Schmidt, TJ., Olson, N.H. and Baker, T.S. 1996. Neutralizing antibody to human rhinovirus 14 penetrates the receptor-bonding canyon. Nature 383: 350-354. Stave, J.W., Card, J.L., Morgan, D.O. and Vakharia, V.N. 1988. Neutralization sites of type 01 foot-and-mouth disease virus defined by monoclonal antibodies and neutralizationescape virus variants. Virology 162: 21-29. Strohmaier, K., Franze, F. and K.-H. Adam, K.-H. 1982. Location and characterization of the antigenic portion of the FMDV immunizing protein. J.Gen.Virol. 59: 295-306. Taboga, 0., Tami, C., Carrillo, E., Nunez, J.I., Rodriguez, A., Saiz, lC., Blanco, E., Valero, M.L., Roig, X., Camarero, J.A. , Andreu, D., Mateu, M.G., Giralt, E., Domingo, E., Sobrino, F. and Palma, E.L. 1997. A large scale evaluation of peptide vaccines against foot-and-mouth disease: lack of solid protection in cattle. J. Virol. 71: 2606-2614. Thomas, A.A.M., Woortmeijer, RJ., Puijk, W. and Barteling, S.J. (1988a). Antigenic sites on foot-and-mouth disease virus type A10. J.Virol. 62: 2782-2789. Thomas,A.A.M., Woortmeijer, R.I., Barteling, S.I. and Meloen, R.H. (1988b). Evidence for more than one important, neutralizing site on foot-and-mouth disease virus. Arch.Virol. 99: 237-242. Tormo, J., Blaas, D., Parry, N.R., Rowlands, D., Stuart, D. and Fita, I. 1994. Crystal structure of a human rhinovirus neutralizing antibody complexed with a peptide derived from viral capsid protein VP2. EMBO J. 13: 2247-2256. Usha, R., Rohll, J.B., Spall, V.E., Shanks, M., Maule, A.J., Johnson, J.E. and Lomonosoff, G.P. 1993. Expression of an animal virus antigenic site on the surface of a plant virus particle. Virology 197: 366-374. Usherwood, E.I. and Nash, A.A. 1995. Lymphocyte recognition of picornaviruses. J.Gen.Virol. 76: 499-508. Valero, M.L., Camarero, J., Adeva, A., Verdaguer, N., Fita, I., Mateu, M.G., Domingo, E., Giralt, E. and Andreu, D. 1995. Cyclic peptides as conformationally restricted models of viral antigens: application to foot-and-mouth disease virus. Biomed. Peptides, Proteins and Nucleic Acids 1: 133-140. Valero, M.L., Camarero, J., Haack, T., Mateu, M.G., Domingo, E., Giralt, E. and Andreu, D. 2000. Native-like cyclic peptide models of a viral antigenic site. J. Mol. Recognition 13: 5-13. Van Regenmortel, M.H.V. 1992. Molecular dissection of protein antigens. In: Structure of antigens, Vol. 1. M.H.V. Van Regenmortel, ed. CRC Press, Boca Raton. pp. 1-27. Verdaguer, N., Mateu, M.G., Andreu, D., Giralt, E., Domingo, E. and Fita, I. 1995. Structure of the major antigenic loop of foot-and-mouth disease virus complexed to anti-virus neutralizing antibody. Direct involvement of Arg-Gly-Asp in the interaction. EMBO J. 14: 1690-1696. Verdaguer, N., Mateu, M.G., Bravo, J., Domingo, E. and Fita, I. 1996. Induced pocket to accommodate the cell attachment Arg-Gly-Asp motif in a neutralizing antibody against foot-and-mouth disease virus. J. Mol. BioI. 256: 364-376. Verdaguer, N., Fita, I., Domingo, E. and Mateu, M.G. 1997. Efficient neutralization of footand-mouth disease virus by monovalent antibody binding. J. Viral. 71: 9813-9816. Verdaguer, N., Sevilla, N., Valero, M.L., Andreu, D., Domingo, E., Mateu, M.G. and Fita, I. 1998. A similar pattern of interaction for different antibodies with a major antigenic site of foot-and-mouth disease virus: implications for intratypic antigenic variation. J. Virol. 72: 739-748. Copyright © 2004 By Horizon Bioscience

\260

Interaction of FMDV With Antibodies

Verdaguer, N., Schoehn, G., Ochoa, W.F., Fita, I., Brookes, S., King, A., Domingo, E., Mateu, M.G., Stuart, D. and Hewat, E.A. 1999. Flexibility of the major antigenic loop of foot-and-mouth disease virus bound to an Fab fragment of a neutralising antibody: structure and neutralisation. Virology 255: 260-268. Villen, J., Borras, E., Schaaper, W.M., Meloen, R.H., Davila, M., Domingo, E., Giralt, E. and Andreu, D. 2002. Functional mimicry of a discontinuous antigenic site by a designed synthetic peptide. Chembiochem 3: 175-182. Wang, C.Y., Chang, T.Y., Walfield, A.M., Ye, 1., Shen, M., Chen, S.P., Li, M.C., Lin, Y.L., Jong, M.H., Yang, P.C., Chyr, N., Kramer, E. and Brown, F. 2002. Effective synthetic peptide vaccine for foot-and-mouth disease in swine. Vaccine 20: 2603-2610. Wien, M.W., Filman, DJ., Stura, E.A., Guillot, S., Delpeyroux, F., Crainic, R. and Hogle, J.M. 1995. Structure of the complex between the Fab fragment of a neutralizing antibody for type 1 poliovirus and its viral epitope. Nature Struct. BioI. 2: 232-243. Wild, T.F. and Brown, F. 1967. Nature of the inactivating action of trypsin on foot-andmouth disease virus. J.Gen.Virol. 1: 247-250. Wild, T.F., Burroughs, J.N. and Brown, F. 1969. Surface structure offoot-and-mouth disease virus. J.Gen.Viroi. 4: 313-320. Wimmer, E. and Jameson, B.A. 1984. Poliovirus antigenic sites and vaccines. Nature 308: 19. Xie, Q.-C., McCahon, D., Crowther, J.R., Belsham, G.J. and McCullough, K.C. 1987. Neutralization of foot-and-mouth disease virus can be mediated through any of at least three separate antigenic sites. J.Gen.Virol. 68: 1637-1647. Xiong, J.P., Stehle, T., Zhang, R., Joachimiak, A., Frech, M., Goodman, S.L. and Arnaout, M.A. 2002. Crystal structure of the extracellular segment of integrin alphaVbeta3 in complex with an Arg-Gly-Asp ligand. Science 296: 151-155.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 10 Quasispecies Dynamics and Evolution of Foot-and-Mouth Disease Virus

Esteban Domingo, Carmen M. Ruiz-Jarabo, Armando Arias, Juan Garcia-Arriaza and Cristina Escarmis Abstract Foot-and-mouth disease virus (FMDV), as other RNA viruses characterized to date, exists and replicates as complex and dynamic mixtures of related mutants termed viral quasispecies. In this chapter we review the conceptual origins of quasispecies, and the molecular basis and biological implications of quasispecies dynamics, as viewed through FMDV. Topics covered include genetic and antigenic heterogeneity of FMDV populations, evolution of host cell recognition, molecular epidemiology of FMDV, comparative rates of evolution of FMDV and other viruses in nature, and the mechanisms operating to maintain virus identity despite high mutation rates. Considering FMDV as a complex adaptive system, we address memory in FMDV quasispecies, the consequences of Muller's ratchet, and virus entry into error catastrophe as a new antiviral strategy. Although the chapter reviews mainly experimental results, some connections with theoretical concepts are also addressed. 1. Introduction: The Impact of Genetic Variation of RNA Viruses Viruses that have RNA as genetic material (or as a replicative intermediate) share a salient feature that conditions their biology: error-prone replication. Most human and animal RNA viruses display extensive heterogeneity within infected individuals, and the genetic composition of the viral populations varies within infected hosts as the infection progresses, and also in the course of disease outbreaks and epidemics. Examples of highly variable viruses include the human and animal influenza viruses, human immunodeficiency virus, hepatitis viruses, poliovirus and several other pathogenic human picomaviruses, and viruses of impact for animal health such as aphthoviruses, the lentivirus equine infectious anemia virus, bovine viral diarrhea virus, and porcine respiratory and reproductive disease virus, to name just a few [reviews in (Domingo et aI., 1988; Kurstak et aI., 1990; Holland, 1992; Morse, 1993; 1994; Gibbs et aI., 1995; Domingo et aI., 1999; Kiss et al., 1999; Domingo et al., 2001; Domingo, 2003a; 2003b)]. Virus variation may be involved in disease progression and viral pathogenesis. Clones of SIV prepared from isolates of monkeys at different phases of immunodeficiency disease, reproduced the disease stage following inoculation of Copyright © 2004 By Horizon Bioscience

\262

Quasispecies Dynamics and Evolution of FMD V

naIve monkeys (Kimata et al., 1999). When HBV, HCV and HIV are not cleared by the host immune response, the selection of cytotoxic lymphocyte (CTL)-escape mutants may contribute to persistence of these viruses (Weiner et al., 1995; McMichael and Rowland-Jones, 2002). Genetic variation is also the basis of antigenic diversification of viruses with ensuing problems for vaccine efficacy. Mutation and recombination underlie modifications of virulence, cell tropism and host range, and the selection of mutants resistant to antiviral inhibitors or that escape from short interfering RNAs, adaptive responses that constitute major difficulties for the control of viral diseases (Domingo and Holland, 1992; Domingo et al., 2001; Gitlin et al., 2002; Menendez-Arias, 2002b; Boden et al., 2003). There are numerous examples of minimal genetic variation having major (and some times subtle) effects on the behaviour of viruses. Therefore, genetic variation is embedded in RNA virus biology. Foot-and-mouth disease virus (FMDV), the most important viral pathogen in veterinary medicine, fully participates of continuous genetic variation and of several of its biological consequences, affecting directly or indirectly viral pathogenesis, virus persistence and difficulties for disease prevention and control. Since the early studies with FMDV, phenotypic plasticity was observed in the extent of acid and thermal lability of the virus particles, plaque morphology and in the pathogenic manifestations of the virus. Indeed some FMDV variants, often depending upon their passage history in cell culture, could be associated with fulminant or degenerative forms of myocarditis, neuropathology or diabetes [(Bachrach, 1968; Domingo et al., 1990), and references therein]. Therefore, contrary to an extended belief, viral diseases need not be "invariable" manifestations of "variable" viruses. [This point was addressed by (Holland et at., 1992)]. As reviewed in this and other chapters of this book, developments in the last two decades have fully confirmed multiple biological implications of genetic variation of FMDV. Implications extend from adaptation to new host species to the surprising epidemiological dominance of new strains such as the FMDV 0 PanAsia group, at the beginning of the XXI st century (Section 3.1). This chapter deals with the origins and consequences of error-prone replication as they affect the behaviour of FMDV. First, we review the molecular basis of the generation of genetic diversity in RNA viruses, and its consequences for virus population dynamics. 1.1. Molecular Basis of Genetic Variation of RNA Viruses

The molecular mechanisms of variation of RNA viruses are the same that operate to promote genetic change in all life forms on earth, namely mutation, recombination and genome segment (chromosome) reassortment. The main difference between complex organisms and molecular parasites such as RNA viruses is the extent to which such molecular mechanisms alter their genetic material and the time frame in which alterations can be observed and consequences felt. This is due to two critical reasons: differences in mutation rates per nucleotide, and the vastly larger population sizes and numbers of replication cycles per unit time of RNA viruses as compared with differentiated organisms. For review of the different impact of genetic variation on viruses, cells and organisms see (Domingo and Holland, 1994; Wain-Hobson, 1994; Coffin, 1995; Domingo eta/., 2001). Copyright © 2004 By Horizon Bioscience

Domingo et al. 263 I

Mutation rates calculated for several RNA viruses using genetic and biochemical methods are in the range of 10-3 to 10-5 substitutions per nucleotide copied (Batschelet et a!., 1976; Drake and Holland, 1999). There is no correlation between average mutation rates, which are high for all RNA genomes examined to date, and either the evolutionary rate or the extent of antigenic variation displayed by a virus in nature. Within the Picornaviridae family, there is a single serotype of Mengo virus, three of poliovirus, seven of FMDV and around one hundred of human rhinoviruses. Yet in all these viruses the frequency of monoclonal antibody (MAb)-escape mutants is similar and in the range of 10-3 to 10-5 , and the specific frequency values are dependent on the nature of each individual epitope rather than on the degree of antigenic diversity attained by the virus in nature [reviewed in (Domingo et a!., 2002)]. Other viruses that show relative antigenic stability, such as the human measles and hepatitis A viruses, manifest genetic variation and heterogeneity that can contribute to significant biological differences among isolates of the same virus (Hsu et al., 1998; Schrag et al., 1999; Sanchez et al., 2003). Not only frequencies of MAb-escape mutants, but mutation frequencies in general (the proportion of mutated residues present in a population of genomes), are only indirectly related to mutation rates. The latter are the result of biochemical events dependent essentially on the nature of the replication machinery in interaction with the template to be copied, while the mutation frequency is a population parameter dependent not only on mutation rates at individual genomic sites, but also on relative replication capacity of the different mutants produced and that coexist at any given time (see Section 3). Introductions to terms and definitions relevant to the description of RNA virus evolution can be found in (Domingo et al., 1988; Domingo, 1996; 1999; 2003b). Mutation rates for RNA viruses are in sharp contrast with mutation rates normally operating during cellular DNA replication which are around 10- 10 substitutions per nucleotide copied (Friedberg et a!., 1995; Alberts et al., 2002). These average values do not apply to all DNA polymerization events in cells. There are low fidelity DNA polymerases involved in repair functions, and localized DNA mutagenesis is observed in some physiological processes such as hypermutagenesis in the generation of immunoglobulin gene diversity (Michael et al., 2002; Storb and Stavnezer, 2002), in some types of cancer cells (Nicolson, 1987; Strauss, 2000), and as part of defense mechanisms such as repeat-induced point mutations in some fungi, and antiretroviral cytidine deaminases [reviewed in (Domingo et al., 2001; Bushman, 2002; Arnold and Hilton, 2003; Goff, 2003)]. Functional and structural studies have provided solid evidence that most enzymes that replicate and retrotranscribe viral RNA (the RNA-dependent RNA polymerases and RNAdependent DNA polymerases, respectively) lack a 3' ~ 5' exonuclease activity which is the proofreading-repair activity associated with most cellular DNA polymerases (Steinhauer et al., 1992; Sousa, 1996; Steitz, 1999; Cameron et al., 2002). Even if some proofreading activity - a 3' to 5' exonuclease suggested by the presence of the Exo N domain in coronaviruses (Snijder et al., 2003), or other activities of the type described for human influenza virus polymerase (Ishihama et al., 1986)- operated during replication of RNA viruses with a large genome, it is unlikely that post-replicative repair pathways, which have evolved probably to ensure the necessary accuracy of copying of large chromosomal DNA, would be effective for the repair of RNA genomes. Such multiple repair pathways can Copyright © 2004 By Horizon Bioscience

\264 Quasispecies Dynamics and Evolution of FMDV

excise and correct nucleotides that have been misincoporated in DNA but not those misincorporated in RNA or in RNA-DNA hybrids (Radman et al., 1981; Friedberg et al., 1995; Kamath-Loeb et al., 1997; Wood et al., 2001). In cellular organisms, mismatch repair appears to be involved in many functions (repair of misincorporated nucleotides, prevention of some types of non-homologous recombination, etc.) that contribute to genetic stability (Harle and links-Robertson, 2000). Loss of mismatch repair functions in E. coli resulted in mutant frequencies for a lacZ allele of 1 x 10-4 (Cupples et aI., 1990), and saturation of mismatch repair by a mutagenic base analog resulted in error catastrophe of E. coli (Negishi et al., 2002) (compare with Section 4.2.2 on FMDV entry into error catastrophe). Despite very solid evidence that amino acid replacements and minimal structural alterations of enzymes can modulate polymerase copying fidelity (Menendez-Arias, 2002a), high average mutation rates have been conserved among RNA viruses and constitute a basic feature of their biology. Several forms of recombination of genetic material have been described that can be broadly divided into homologous and non-homologous recombination. The former requires high nucleotide sequence identity at the recombination site, and may result from breakage and rejoining of RNA fragments or from "copy choice" (strand switching) of the polymerase during RNA synthesis (Nagy and Simon, 1997; Gmyl et aI., 2003). Non-homologous recombination embraces a number of different processes that share the lack of a requirement for extended regions of nucleotide sequence identity at the recombination site. Therefore, it may provide a means to join divergent nucleotide sequences within the same genome or between unrelated genomes (Cooper, 2000; Alberts et al., 2002). Recombination rates are quite variable among RNA viruses. During picornavirus infections, recombination frequencies between closely related genomes have been estimated in 10% to 20% of the progeny (King, 1988; Lai, 1992). Homologous recombination appears to be infrequent for negative strand RNA viruses although it has been reported for the Tula hantavirus (Plyusnin et al., 2002). Recombination may have two disparate effects on the evolution of RNA viruses. It may serve to rescue high fitness genomes from low fitness parents. This conservative activity will often be associated with homologous recombination. In contrast, non-homologous recombination may produce new genome combinations from distant parents. This is essentially an exploratory activity, risky from the point of view of survival, but potentially innovative. There is evidence that recombination has played a key role in generating new viral pathogens. Western equine encephalitis virus was probably generated by recombination between a virus related to eastern equine encephalitis virus and some New World relative of Sindbis virus (Hahn et al., 1988; Weaver et aI., 1994). Reverse genetics of negative strand RNA virus has permitted the construction of recombinant RNA viruses that are highly altered in terms of genome composition and gene order (Rolls et aI., 1994; Ball et al., 1999; Wertz et aI., 2002). Several such constructs produce viable virus suggesting a remarkable tolerance of RNA genome to some profound genome rearrangements. These observations support a likely role of recombination in RNA virus origins and evolution. Genome segment reassortment can be a source of drastic phenotypic variation, the best characterized being the generation of antigenically divergent influenza viruses associated with influenza pandemics in humans (Webster, 1999). Copyright © 2004 By Horizon Bioscience

Domingo et a/. 265 I

Mutation stands as a universal mechanism of genetic variation in the earth biosphere. For homologous recombination to be of evolutionary value, mutations must have occurred to differentiate genetically and phenotypically the parental genomes. Therefore, mutation occupies a central role in the evolution of RNA viruses because it confers evolutionary value to homologous recombination events, and also because all evidence points to mutation as the means to modulate adaptation of new genome combinations. A key role of mutation in the evolution of FMDV is documented in Section 2. 1.2. Quasispecies Dynamics of RNA Viruses

Replication of a single, initial RNA virus genome in vivo or in cell culture rapidly generates a spectrum of related mutants. The founder genome can be either a biological clone (virus from a single viral plaque or virus rescued from an endpoint dilution series) or a molecular cDNA clone that can produce infectious transcripts. Replication then originates a dynamic mutant distribution in which the ensemble displays higher replication capacity than the individual mutants that compose the ensemble (Domingo et al., 1978; Nowak, 1992; Duarte et al., 1994; Domingo et al., 1995; Eigen, 1996). Such dynamic mutant distributions are termed quasispecies. The theoretical quasispecies concept was formulated by Eigen, Schuster and their colleagues as a model for error-prone replication and self-organization of simple replicons that perhaps constituted primitive life forms on earth (Eigen, 1971; Eigen and Schuster, 1979; Eigen and Biebricher, 1988; Eigen, 1993). Quasispecies, as other theoretical descriptions of evolutionary dynamics, is based on the fundamental principles of Darwinian evolution, namely replication, genetic variation, competition, and selection. Different formulations of evolutionary dynamics are in reality part of a single unified framework, as elegantly shown by (Page and Nowak, 2002), including the quasispecies equation of Eigen and Schuster. Among the different treatments (some very rooted in classical population genetics) quasispecies is suitable as a theoretical framework for RNA virus evolution because it emphasizes mutation as an essential component of the replication dynamics. The initial definition of quasispecies involved steadystate equilibrium conditions achieved with replicons of infinite population size, implying a static fitness landscape (environment). Recent theoretical formulations have extended quasispecies to variable fitness landscapes (Eigen, 2000; Nilsson and Snoad, 2000; Wilke et al., 2001). Since viruses encounter variable environments during their replication in vivo and in cell culture, these theoretical extensions of quasispecies provide a realistic framework for the evolutionary behaviour of finite viral populations. Another important theoretical development was the inclusion of genetic recombination as a source of genetic variation that may contribute to the composition and degree of stability of mutant spectra (Boerlijst et ai., 1996). Concordantly, virologists use an extended definition of quasispecies to mean dynamic distributions of nonidentical but closely related mutant and recombinant viral genomes subjected to a continuous process of genetic variation, competition and selection, and which act as a unit of selection (Domingo, 1999). This extended definition captures the essential features of quasispecies dynamics as they affect RNA virus biology.

Copyright © 2004 By Horizon Bioscience

\266 Quasispecies Dynamics and Evolution of FMDV

Quasispecies dynamics, as expected from a general evolutionary theory (Eigen and Schuster, 1979; Page and Nowak, 2002) is not limited to RNA viruses. Evidence of extensive genetic heterogeneity in natural populations and competitive dynamics typical of quasispecies have been documented in DNA viruses as diverse as plant geminivirus (Isnard et al., 1998), the parvovirus minute virus of mice (Lopez-Bueno et al., 2003) or in the complex DNA viruses of the thermophilic archaeon Sulfolobus (Prangishvili et ai., 2001) [review in (Domingo et al., 2001)]. Cells are no exception. The behaviour of bacterial communities has much in common with quasispecies dynamics as studied with viruses [reviewed in (Domingo, 2004)]. A model of evolution of two coupled quasispecies has been extended to the interaction of a virus with the immune receptor motif of B lymphocytes (Kamp and Bornholdt, 2002). This study defined a new "adaptation catastrophe" for viruses when viral mutation rates are insufficient to evade immune attack, and suggests experimental approaches in which cellular compartments are regarded as heterogeneous arrays of genetic entities. In summary, to end this general introduction to viral population dynamics, it must be emphasized that the great merit of quasispecies has been to regard biological entities as highly complex and dynamic. Nucleotide sequences in RNA virus isolates are "weighted averages" of multitudes of different sequences (Domingo et al., 1978). As stated by Frank, this "variable polyploidy" (the term of classical population genetics akin to mutant spectrum), which is expressed to its full extent mainly (but not exclusively) in RNA viruses, differs from "classical Mendelian genetics" (Frank, 2001). Virologists often refer to the "unique features" of RNA genetics (versus classical DNA genetics) to mean the consequences of high mutation rates and quasispecies dynamics for RNA virus behaviour. It is a spectrum of related genomes, and not a defined genetic entity, the target and the result of competition and selection. Generalizing a statement of Gomez and colleagues which referred to HCV: "An RNA virus is constantly redefining itself both genetically and phenotypically" (Gomez et al., 1999). Next we review consequences of quasispecies dynamics for FMDV.

2. Quasispecies Dynamics as Studied with FMDV 2.1. A Historical Introduction

Early experiments on nucleic acid competition-hybridization using genomic FMDV RNA, estimated nucleotide sequence identities of 40% to 70% between RNAs from FMDVs of serotypes A, 0 and C, and greater than 70% between RNAs of FMDVs of different subtypes within serotypes A and 0 (Dietzschold et al., 1971; Robson et al., 1977). Sampling of genomic sequences by T1 oligonucleotide fingerprinting of p 32-labeled FMDV RNA [a technique that represented a true revolution for the study of RNA virus variation (De Watcher and Fiers, 1972)] detected nucleotide sequence differences between FMDVs of the same serotype (Frisby et al., 1976; Robson et al., 1979). In the beginning of nucleotide sequencing (that yielded short sequences deduced from analysis of partial and complete ribonuclease digests of p 32-labeled RNA) only limited sequence identity was observed in the 40 nucleotides adjacent to the 3'-terminal polyadenylate (poly A) of FMDVs A-61, O-VI and C-997 (Porter et al., 1978). Analyses ofFMDV isolates from a single disease outbreak by T 1oligonucleotide fingerprinting revealed genetic heterogeneity within infected animals that was not Copyright © 2004 By Horizon Bioscience

Domingo et al. 2671

compatible with multiple infections of an animal, providing the first evidence of viral quasispecies for an animal virus in vivo (Domingo et ai., 1980). The rapid generation of mutants from a single genome was documented by passage of FMDV clones of the same isolates in cell culture (Sobrino et al., 1983). In this study, several features of quasispecies dynamics were revealed: (i) the generation of a broad mutant spectrum upon 30 serial passages of biological FMDV clones (single founder genomes) of serotype 0 and C in two cell lines; (ii) a dynamics of competition among ne\vly generated mutants, and fixation of different genomic sequences in parallel lineages (a stochastic versus deterministic behaviour; see Section 3.1); and (iii) an increase in the capacity to produce progeny upon passage of the virus in cell culture. These results with a pathogenic eukaryotic virus agreed with those obtained with the prokaryotic virus phage Q~ (Domingo et al., 1978), suggesting that the concept of quasispecies and its biological implications could apply to other pathogenic RNA viruses as well. Rapid generation of mutant spectra provided a molecular interpretation of the phenotypic plasticity of animal, plant and bacterial RNA viruses recorded long ago, when procedures to analyze nucleotide sequences were not available [references to early work on plasticity of RNA viruses and the first evidence of viral quasispecies can be found in (Holland et al., 1982; Domingo et al., 1985; 1988)]. Subsequent studies with FMDV, reviewed in the following Sections, have supported high mutation rates and quasispecies dynamics as influencing the behaviour of FMDV. The main consequences have been: (i) The generation of mutant spectra in infected animals as the first step in the process of genetic diversification of the virus in nature. What traditionally was termed "an FMDV isolate" is in reality "a pool of variants". (ii) The occurrence of fitness (replicative capacity) variations among components of mutant spectra and between ensembles of mutants, both in vivo and in cell culture. (iii) The demonstration that variation of several phenotypic traits of FMDV is a consequence of quasispecies dynamics. (iv) The presence of memory genomes in evolving FMDV quasispecies. (v) Experimental evidence, afforded also by studies with several other viruses, that violation of an error threshold for maintenance of genetic information (transition into error catastrophe; discussed in Section 4.2.2.) can lead to viral extinction. 2.2. The Genetic and Antigenic Heterogeneity of FMDV Genetic heterogeneity among the FMDVs circulating in the course of a single disease episode has been documented in analyses world-wide. Our own studies (Sobrino et al., 1986; Martinez et al., 1988) showed that some genetic changes resulted in amino acid substitutions at critical positions of antigenic sites (see Chapter 9) and modifications of the immunogenic properties of FMDV. Indeed, vaccines prepared with two FMDV C 1 isolates from the same geographical area (and that were indistinguishable by the classical complement fixation test) evoked complete protection against the homologous virus but only partial protection against the heterologous virus (Martinez et al., 1988). Extensive antigenic heterogeneity within serotype C was documented by reactivity with MAbs that defined different epitopes of viruses of different subtypes within serotype C (Mateu et al., 1987; 1988). Heterogeneity extended to the finding of epitopes that varied among isolates of the same subtype (Mateu et al., 1988). An epitope Copyright © 2004 By Horizon Bioscience

\268

Quasispecies Dynamics and Evolution of FMDV

involved in neutralization of FMDV subtype C 3 was generated by a single amino acid replacement at the corresponding site of FMDV subtype C 1 (Hernandez et al., 1992). Amino acid replacements at a major antigenic site, found as minority components in natural isolates, altered substantially the interaction with antibodies, implying quasispecies as an important element in viral pathogenesis (Mateu et al., 1989). Antigenic variants were isolated following experimental persistent infections of cattle (Gebauer et aI., 1988) or acute infections of swine (Carrillo et al., 1990b), initiated with biological clones of FMDV. These analyses of natural isolates and of controlled viral infections in vivo, suggest that FMDV in nature is best described as a dynamic continuum of different antigenic forms as a reflection of quasispecies dynamics at the protein level. Some times such forms may differ very slightly, indicating probably either a recent common ancestor or the survival of virus without replication. Other times, antigenic differences may be profound, and not necessarily associated with a large genetic distance (Hernandez et al., 1992). FMDVs of different serotype, represent an extreme case of antigenic diversification, and they are defined on the basis of lack of cross-protection by vaccination, or by susceptibility of animals convalescent of FMD caused by FMDV of one serotype to infection by FMDV of a different serotype (see Chapters 1, 11, 12). Coherently with this view on the antigenic properties of FMDV populations, subtyping of FMDV is no longer practiced. Despite increasing knowledge of the virus, it is still a highly empirical endeavour to characterize antigenically the FMDVs associated with new FMD outbreaks (in enzootic areas) and with disease reemergence (in areas that were free of FMD), as well as to monitor antigenic variation in the course of disease episodes (Section 3.2). The application of molecular techniques has provided new insights into the mechanisms of antigenic variation of FMDV. Amino acid sequences of a major antigenic site located within the G-H loop of capsid protein VP 1 [a protruding and mobile loop that includes also an integrin-recognition domain (Acharya et al., 1989); Chapter 4)] of viruses of different subtypes within serotype C, were related to the reactivities of the corresponding viruses with several MAbs. This comparison revealed two mechanisms of antigenic diversification in nature: (i) a gradual increase in antigenic distance brought about by accumulation of amino acid replacements at two hypervariable segments within the antigenic site, and (ii) an abrupt antigenic change, manifested by loss of several epitopes, associated with a replacement at some critical positions within the same antigenic site (Martinez et al., 1991b). These two mechanisms bear a resemblance with antigenic drift and shift, extensively characterized for human influenza viruses (Webster, 1999). FMDV has at least four different antigenic sites some of which may interact ,vith each other, depending on the viral serotype, and by mechanisms which are not well understood (Parry et al., 1990; Baranowski et al., 2001b) (see Chapter 9 and Section 2.4). This complicates even further antigenic variation of FMDV since critical amino acid replacements at several antigenic sites can contribute substantially to the antigenic heterogeneity of the virus in the field (Feigelstock et al., 1992; Mateu et al., 1994). To add further to the complexities of the effects of amino acid replacements on the genetic behaviour of FMDV, there are nonadditive effects of amino acid substitutions in antigen-antibody recognition (Mateu et al., 1992). Copyright © 2004 By Horizon Bioscience

Domingo et al. 269 I

Antigenic heterogeneity is a consequence of a broader genetic heterogeneity continuously existing and evolving in FMDV populations. Heterogeneity and modifications can occur in regulatory regions (Escarmis et aI., 1992; MartinezSalas et aI., 1993) and essential genes such as the viral polymerase (3D) (Villaverde et al., 1988), or proteinase L (Mohapatra et al., 2002), or non-structural protein 3A (Beard and Mason, 2000), to quote only a few of many studies (see also Section 3.2 and Chapter 7). In the case of the polymerase gene (3D) the genetic heterogeneity of contemporary isolates was in the range of 7.0 x 10-4 to 2.8 x 10-3 substitutions per nucleotide (albeit with 80% silent mutations) comparable to 1.6 x 10-3 to 6.4 x 10-3 for the VP1 gene (Villaverde et al., 1988). 2.3. Antigenic Variation in the Absence of Immune Selection The most accepted view is that antigenic variation is the result of immune selection. In the case of FMDV, antibodies against several antigenic sites are present in infected, convalescent and vaccinated animals, and whenever virus is not rapidly cleared, there is an opportunity to select for antigenic variants of FMDV [reviewed in different chapters of (Rowlands, 2003)]. While selection by antibodies and cytotoxic T cells probably plays a major role in antigenic variation of viruses, a number of experiments with FMDV (Bolwell et aI., 1989; Diez et aI., 1990; Martinez et aI., 1991a; Curry et aI., 1996; Sevilla and Domingo, 1996; Holguin et al., 1997), and with other viruses, have indicated that antigenic variation is not necessarily the result of immune selection (Domingo et al., 1993; Haydon and Woolhouse, 1998; Domingo et aI., 2001). The mechanism alternative to positive immune selection is the acceptance of amino acid replacements at random in antigenic sites because surface residues of viral capsids and envelopes are less subjected to structural constraints than residues that are internal in capsids or envelopes. Following the occurrence and acceptance of amino acid substitution(s), the antigenic variant may rise to dominance by the following mechanisms. The relevant antigenic site may perform functions other than the interaction with components of the immune response. This possibility is increasingly supported by evidence of multifunctionality of viral proteins and protein domains. Selection for a trait that depends on amino acid replacements that are also part of an antigenic site will result in antigenic variation in the absence of immune selection. In particular, the observation made first with FMDV (Verdaguer et al., 1995; 1998), but also with many other viruses [reviewed in (Baranowski et aI., 2001a; 2003)], that amino acid residues of antigenic sites may also be involved in recognition of cellular receptors, renders possible, and may even facilitate, a coevolution of antigenicity and host cell tropism. Depending on the tolerances of the two functions for amino acid replacements, a change in receptor specificity may result in an antigenic change. Alternatively, the use of an entirely new receptor by a virus may render the obsolete receptor binding site highly tolerant to variation (Baranowski et al., 2003). An additional mechanism for antigenic variation in the absence of immune selection relates to quasispecies dynamics and the classical concept of"hitch-hiking" of mutations. This was termed "random change" model (Domingo et al., 1993). It proposes that an antigenic variant may rise to dominance due to fluctuations in mutant distributions associated with random drift or positive selection targetted at sites other than the antigenic sites (Domingo et al., 1993; Haydon and Woolhouse,

Copyright © 2004 By Horizon Bioscience

\270

Quasispecies Dynamics and Evolution of FMDV

1998). Therefore, antigenic variation of a virus is not necessarily the result of positive selection by components of the immune response. 2.3.1. Different Repertoire of Amino Acid Replacements in the Presence or Absence of Immune Selection

Passage of FMDV in cell culture in the presence of neutralizing antibodies results in selection of antigenic variants (Carrillo et al., 1989; Rojas et al., 1992; Borrego et al., 1993). One of the studies involved selection of FMDV C-S8c 1 mutants upon passage of the cloned virus in the presence of polyclonal antibodies raised against a synthetic peptide that represented the major antigenic site A at the G-H loop ofVPl (see Chapter 9). Clone C-S8cl evolved towards complex populations of antigenic variants displaying increased resistance to polyclonal antibodies and decreased fitness. Interestingly, the repertoire of amino acid replacements at antigenic site A in mutants selected in the presence of antibodies overlapped but was not identical to the repertoire seen in the absence of immune selection (Figure 1). Extension of these results to other FMDVs and a careful statistical evaluation of the replacement repertoire could provide a basis for distinguishing the relative importance of random mutational events, tolerance to replacements, and positive selection in antigenic variation of FMDV in nature. However, a word of caution is necessary. The different subsets of amino acid replacements in the presence or the absence of antibody selection may have been affected not only by the degree of adaptation to cell culture (as shown in Figure 1) but also by the number of passages in cell culture, the m.oj. used in the infections, the virus population size (which may not have the same impact as the m.oj.) or, simply, the number of clones or populations analyzed (a statistical limitation). As an example, a double amino acid replacement at antigenic site A of FMDV C-S8cl (L-144 ~ V and A-145 ~ P) became dominant following 100 serial passages in BHK-21 cells involving 2 x 108 PFU of virus per passage, but not 2 x 105 PFU per passage (Sevilla and Domingo, 1996). Clearly, the mutant repertoire that participates in an evolutionary process (the number of mutants in the portion of the mutant spectrum involved) may affect the outcome of the process. Furthermore, in competitions between populations with the same clonal origin but with a different passage history, the winner of the competition was influenced by the m.oj. (Sevilla et ai., 1998). These facts illustrate the difficulties in interpreting comparative mutant repertoires and hint to more complex virus-host interactions than previously suspected (Woolhouse et al., 2002). 2.4. Variations of Host Cell Tropism

Passage of FMDV C-S8cl in BHK-21 cells resulted in an expansion of host cell tropism (Baranowski et ai., 1998; 2000) that made dispensable an ROD triplet involved in recognition of integrin receptors (see Chapter 7). This permitted isolation of FMDV mutants with RED, ROO and even 000 instead of ROD at the G-H loop of capsid protein VPl (Martinez et al., 1997; Ruiz-Jarabo et al., 1999). Since the ROD plays a critical role also in binding of neutralizing antibodies (Verdaguer et ai., 1995; 1999; Ochoa et al., 2000) the mutants lacking ROD were antigenically altered. In particular, FMDV with ROO showed impaired reactivity with monoclonal antibodies and with polyclonal antibodies raised in swine and

Copyright © 2004 By Horizon Bioscience

Domingo et al. 2711

N

RV

R

VP

AA KKK

Virus not subjected to multiple passages in BHK-21 cells

K

Virus subjected to multiple passages in BHK-21 cells

---IYTASARGDLAHLTTTI~-----DIV

R

S K K

PN

Q y

AAM

R

P P M

G

D I V SN

Virus not subjected to multiple passages in BHK-21 cells

EGVVRP RESDQS

NR

V

G

N

T

L

Virus subjected to multiple passages in BHK-21 cells

P

N

A

Figure 1. Patterns of variation of a major antigenic site of FMDV. The boxed sequence corresponds to positions 136 to 150 of the O-H loop of capsid protein VPl of FMDV C-S8cl. The sequence, in which the ROD is underlined, includes the minimal number of residues mapped as belonging to antigenic site A, according to reactivity of MAbs with synthetic peptides, and the analysis of MAb-escape mutants (Mateu, 1995). Above the sequence, the repertoire of amino acid replacements found in FMDV C-S8c 1 not subjected or subjected to multiple passages in BHK-21 cells, upon replication in the absence of antibodies, is given. Often a single replacement, and less frequently two or more replacements, were found in the same clone or population. [Data from (Dfez et al., 1989; Borrego et al., 1993; Sevilla and Domingo, 1996; Holguin et al., 1997)]. Below the sequence, the corresponding repertoires obtained with FMDV C-S8c 1 replicated in the presence of site A-specific polyclonal or monoclonal antibodies are represented. Again, most frequently a single replacement, and less frequently two or more replacements, were found in the same clone or population. [Data from (Borrego et al., 1993; Mateu, 1995; Rufz-Jarabo et al., 1999; 2003b)]. Note the strict conservation of Y-136 and R-141, and the tolerance of 0-142,0143 to accept mutations only in FMDV C-S8cl multiply passaged in BHK-21 cells. In addition to the variables recognized in the figure (presence or absence of antibody selection, and degree of adaptation to BHK-21 cells), the repertoires of observed substitutions may be influenced by the population size of the virus, and the total number of clones and populations analyzed, as discussed in the text.

guinea pigs using wild-type virus as immunogen (Rufz-Jarabo et aI., 1999). The results of these studies in cell culture indicate that FMDV C-S8cl passaged in BHK-21 cells could use at least three entry pathways involving either integrins, heparan sulfate or a third, unidentified component, to enter even the same cell type (Baranowski et al., 2000; 2001a). Flexibility in receptor usage is by no means restricted to FMDV in cell culture. The replication ofFMDV C3 Arg-85 in partially immune cattle resulted in selection of mutants with amino acid substitutions within Copyright © 2004 By Horizon Bioscience

1272 Quasispecies Dynamics and Evolution of FMDV

the RGD or at neighbouring positions (Taboga et ai., 1997). Studies on the stability of the mutants in cell culture suggested a coevolution of antigenicity and host cell tropism of FMDV in vivo (Tami et ai., 2003). Dispensability of the RGD was also shown with FMDV O/CHN/90, a type 0 FMDV used for vaccine production in China. Cell culture adaptation of this virus resulted in altered cell tropism, and virus lacking RGD caused mild disease in swine (Zhao et ai., 2003). The question of what are the receptors that can be potentially used by FMDV in vivo remains an intriguing one (see Chapter 7). Contrary to the traditional view that one virus fits one receptor type, increasing evidence indicates that many viruses may use multiple, unrelated receptors and that the same receptor can be used by several viruses and other pathogens (Baranowski et ai., 2003). For a minimum of 20 viruses belonging to 10 virus families there is some overlap between antibody-binding sites and receptor-recognition sites (Baranowski et ai., 2003), permitting the type of coevolutionary processes of antigenicity and tropism described for FMDV (Section 2.3). Variations in cell tropism and host-range of FMDV can also be mediated by alterations of non-structural proteins. Deletions and point mutations in 3A were associated with attenuation of FMDV for cattle (Giraudo et ai., 1990; Beard and Mason, 2000). A point mutation in 3A was critical for the adaptation of FMDV CS8cl, a virus isolate from swine, to the guinea pig (Nunez et ai., 2001). Much has to be learned of the role of non-structural proteins in determining cell permissivity to infection. The relevance of host cell tropism to FMDV pathogenesis is further discussed in Chapter 7. 3. Natural Evolution of FMDV The rapid generation of mutants during replication in infected animals is the first stage in a process of natural diversification of FMDV. Alignment of nucleotide sequences (and deduced amino acid sequences) is used to calculate nucleotide and amino acid sequence identities among sequences, and to complete matrices of genetic distances for different genomic regions. This information has been used to derive phylogenetic relationships among FMDV isolates. Nucleotide sequences are available from the EMBL/GenBank or the World Reference Laboratory, Pirbright (Knowles and Palmenberg, 1998), where a world wide sequence database is in the process of being established [quoted in the excellent review by Knowles and Samuel (2003) on the molecular epidemiology of FMDV, on which Sections 3 and 3.1 of this chapter have been inspired]. A number of statistical procedures are available and have been applied to derive phylogenetic trees relating FMDVs of the same serotype or of different serotypes. Frequently, sequence relatedness of the capsid-coding region have been used to construct trees using either the neighour-joining or the unweigthed pair group mean average (UPGMA) algorithms [program NEIGHBOR from the PHYLIP 3.5 package (Felsenstein, 1993)]. In this procedure, distances are calculated for each pair of sequences, and results are more reliable for related than for highly divergent sequences. Using this procedure, phylogenetic trees have been derived for FMDVs of different serotype, and of the same serotype for isolates of different areas of the world [review in (Knowles and Samuel, 2003)].

Copyright © 2004 By Horizon Bioscience

Domingo et al. 2731

.-----------:tt~f:

01K

I-----------~~t~: Asia 1

:~:t~: A 22

t----:::v.t#:

A 12

I------~t~f:· Alran

~IDf: A 10 r-+---:~~t~t: A22b

---~~:t~f: A5Ww ~~:t~~: A24

:~v.W:· AVen {t~f: AArg

:~i\:~: C3Arg :~I~l:' C31nd

---------------:0:t~f: SAT 3 ...--------------------~v.tf:: SAT 2 """--------------{:t~~: SAT 1

10 amino acid substitutions

I

Figure 2. A phylogenetic tree relating FMDV isolates representative of the seven FMDV serotypes. The tree is based on amino acid sequences of capsid protein VPl, and it was constructed using the neighbour-joining method, and rooted using sequences of other picomaviruses as outgroups. Clouds at the tip of each branch indicate that FMDV isolates are distributions of related variants, as discussed in the text. Adapted from Dopazo et al. (1995).

3.1. Phylogenetic Relationships Among FMDVs: Epidemiological Fitness

A phylogenetic tree based on VP1 amino acid sequences of FMDVs of the seven serotypes suggests a major diversification event, leading to the SAT serotypes and to the other FMDV serotypes, and then additional diversifications leading to the A, C, 0 and Asia serotypes (Dopazo et ai., 1995; Knowles and Samuel, 2003) (Figure 2). FMDVs of different serotype show 30% to 50% divergence in the VP1-coding region. Following early work on poliovirus, FMDVs of the same serotype have been divided in different genetic types or topotypes according to the divergence of the VP1-coding region. For FMDVs of serotype 0 nucleotide differences of up to 15% group the isolates in the same topotype. For the SAT serotypes the value was set to 20% since the VP1 gene of SAT appears as more divergent (Knowles and Samuel, 2003), perhaps as a result of a more ancestral origin of these serotypes (Figure 2). Eight topotypes of FMDV serotype 0 have been described (Figure 3). One of the most useful applications of nucleotide sequence comparisons and phylogenetic trees is the tracing of the origin of FMD outbreaks. Historically, the demonstration that several disease episodes in Europe were caused by FMDVs Copyright © 2004 By Horizon Bioscience

\274 Quasispecies Dynamics and Evolution of FMDV

O/P H117/96 1 1 - - - ::':'::i:":",

OITAW/81/97

- - - - - - - - - - - ~:::':::.. 01/KaufbeurenlFRGI66

O/lSAl1J74

O/JAV/5J72

1#".1;1"'4";'

'1m'·i"+16.' '1m'·"'i16.1

O/ALG/1/99

'W"'II;. South - East Asia

--

Middle East - South Asia

- - - - - - - - : : ; : ; : r : ' · 01IManisaITURJ69

1% ~

Figure 3. A phylogenetic tree relating FMDV isolates of serotype O. The unrooted tree is based on nucleotide sequences of the VP I-coding region, and it was constructed using the neighbourjoining method. Clouds at the tip of each branch indicate that FMDV isolates are distributions of related variants, as discussed in the text. The eight topotypes described for FMDV 0 are boxed on the right of the corresponding branches. Adapted from (Knowles and Samuel, 2003).

that were closely related to the FMDV 01 vaccine strains in use during 1970's and 1980's (Beck and Strohmaier, 1987; Carrillo et al., 1990a) influenced the implementation in 1991 of a non-vaccination policy in the ED. Difficulties for the distinction between vaccinated and convalescent animals, with consequences for trade of animal products, and the fact that current vaccines do not prevent the establishment of persistent infections in ruminants (Chapters 1, 6 and 14) also contributed to the implementation of the non-vaccination policy in the ED. However, the decision on this policy did not give sufficient weight to the fact the we live in an increasingly global \vorld in which communication between distant parts of the globe is remarkably intense. As discussed by Mahy in Chapter 17, among the first four out of thirteen principal factors which may contribute to the emergence or reemergence of infectious diseases, microbial adaptation and change, human demographics and behaviour and international travel and commerce are included (Smolinski et aI., 2003). The non-vaccination ED policy Copyright © 2004 By Horizon Bioscience

Domingo et al. 2751

left the farm animal population in Europe totally susceptible to FMDV. In a global world there is a certain probability that legal or illegal commerce may introduce infected animals or animal products into distant areas, and this is presumably what originated the UK 2001 outbreak of FMD, linked for the first time in history with a virus of Asian origin, belonging to the FMDV 0 PanAsia group (Knowles et al., 2001; Sobrino and Domingo, 2001). Rapid action by molecular biologists at the World Reference Laboratory established that the UK 2001 outbreak, that extended also to The Netherlands and France (albeit with less severe consequences than for the UK), was associated with a rapidly expanding new group of viruses, related to those that had caused FMD in Japan in 2000, a country that had been free of the disease since 1908 (Figure 3). Knowles and Samuel (2003) have given a detailed account of the evolution and world-wide spread of FMDV 0 PanAsia. These events show how the capacity of FMDV to vary genetically (as documented by model studies in other Sections of this chapter) together with environmental and sociologic influences, can give rise to highly dominant FMDV variants that may displace previously dominant FMDVs from many areas of the world (Knowles and Samuel, 2003). Epidemiological fitness describes the invasive properties of a virus type in the field. It adds to the replication capacity other characteristics that must enter a broad definition of fitness (Domingo and Holland, 1997): stability of virus particles in infected animals and in the environment, transmissibility, host range, symptomatology (including capacity of produce subclinical infections that delay virus detection), and others. FMDV 0 PanAsia stands as an epidemiologically fit virus, and its capacity to penetrate different world areas has been revealed by application of phylogenetic methods (Figure 3). The adequacy of a non-vaccination policy is a hotly debated issue (Chapters 1, 11 and 15) and brings up the question of different alternative FMD control measures in a global world. From an evolutionary standpoint, systematic vaccination may exert a selective pressure by promoting dominance of antigenic variants in partially immunized animals. Obviously, accidental introduction of vaccine strains may alter the epidemiological picture in ways that would not occur if vaccines that not require handling of infectious FMDV were available. If surveillance and slaughtering of infected animals are the basis of FMD control, then viruses that cause inapparent, subclinical infections would have a selective advantage since they have a higher probability of eluding detection. Therefore, decisions on control methods may influence the differential survival of disparate FMDV forms. Nothing eludes evolution. Not even politics! 3.2. Molecular Epidemiology, Sequence Clouds and Determinants of Phenotype

Phylogenetic studies have been extended to each of the seven FMDV serotypes, and the general conclusions of these studies are in agreement with classical serological classifications (Knowles and Samuel, 2003). High genetic diversity, with the distinction of several topotypes has been documented for FMDV of serotype A, a virus for which multiple serological subtypes had been defined by complement fixation tests (Pereira, 1977). In contrast, FMDV Asia 1 shows limited diversity and to date the available VPl-coding sequences can be grouped into a single topotype. FMDV type C viruses have been divided in eight topotypes Copyright © 2004 By Horizon Bioscience

1276

Quasispecies Dynamics and Evolution of FMDV

(Knowles and Samuel, 2003). Therefore, there is considerable variation in the number of genetic topotypes and antigenic forms [previously characterized as subtypes (Pereira, 1977)] for the seven FMDV serotypes identified to date. In the molecular epidemiology of FMDV, several important questions remain unanswered, notably the extent of variation of other genomic regions, as compared with the capsid-coding region, and the role of genetic recombination in the natural evolution ofFMDV. For FMDV type C nucleotide sequences of the 5'-UTR defined similar groupings than the capsid-coding region, and documented that, in addition to nucleotide substitutions, insertions and deletions may have played a role in the -evolution of regulatory regions of FMDV (Escarmis et al., 1995). Considerable diversity has also been recorded for non-structural proteins such as 3A, which is associated with virulence alterations (Beard and Mason, 2000; Knowles et al., 2001), and proteinases Land 3C (Mohapatra et aI., 2002; van Rensburg et al., 2002) (see also Section 2.2). Recombination has played a major role in the evolution of RNA viruses, in particular positive strand RNA viruses and retroviruses (Section 1.1). Despite high recombination rates of FMDV in cell culture (King et al., 1982; King, 1988), only a few cases of recombination in the field have been suggested from sequence comparisons of natural isolates (Krebs and Marquardt, 1992; van Rensburg et al., 2002). To study mechanisms and frequencies of FMDV recombination in vivo is an important challenge and need. Homologous recombination is more frequent in genomic regions of high nucleotide sequence identity (King, 1988). Therefore, if non-structural proteins are more conserved than structural proteins, recombination at these sites may mediate alterations of virulence which may be highly significant epidemiologically (Giraudo et al., 1990; Beard and Mason, 2000; Knowles et aI., 2001). The tips of the branches of the trees depicted in Figures 2 and 3 have been drawn as clouds to emphasize that in reality each FMDV isolate is not defined by a single nucleotide sequence but by a distribution of sequences or viral quasispecies, as discussed in other sections of this chapter. Current RT-PCR techniques allow amplification and sequencing of natural isolates of FMDV without the need to grow the virus in cell culture. It is well established that adaptation of RNA viruses to cell cultures may result in genetic changes to an extent which is not easy to predict. The studies with FMDV suggest that the number of nucleotide substitutions occurring upon extensive passage of the virus in cell culture is unlikely to alter its phylogenetic position, when comparing it with other natural isolates. A total of 16

nucleotide substitutions per genome became dominant in the consensus sequence after 200 passages of FMDV C-S8cl in cell culture, which means about 0.8% change in the VPl-coding region (compare with genetic distances in Figure 3). This brings us to a distinction: important biological traits depending on one or few nucleotide or amino acid substitutions will not be reflected in the phylogenetic position of the virus. Such traits may include the host cell tropism, host range and virulence of FMDV (Section 2.4). Think, for example, that from a biological clone of C-S8 (derived from one point of the cloud of one of the FMDV branches shown in Figure 2), mutants that can infect primate and human cells without a functional RGD, multitude of antigenic variants, mutants with increased or decreased fitness and virulence for different cell types, ts mutants, mutants fit to persist in cell culture, mutants with an internal oligoadenylate tract (with no precedent in the Copyright © 2004 By Horizon Bioscience

Domingo et al. 2771

genetics of animal viruses) and mutants adapted to cause disease in guinea pigs, among others, have been isolated and characterized (described in other sections of this chapter, and unpublished observations). And all such phenotypic diversity was derived from a "single" genome within the life-times of a few students and postdocs! Such phenotypic flexibility of viruses that originate from a single position of a phylogenetic tree is also the reason why it is not possible to relate a phylogenetic position with the antigenic properties of FMDV, a fact that impedes vaccine design based exclusively on genetic data (Section 2.2). The information on FMDV evolution contributed by molecular epidemiology must be complemented by information on antigenic behaviour and pathologic manifestations of the different isolates that have entered the comparisons. 3.3. Rates of Evolution of FMDV as Compared with Other RNA Viruses For FMDVs of type C isolated during the same disease outbreak, assumed for epidemiological reasons to derive from a recent common ancestor, rates of evolution have been estimated in < 4 x 10-4 and up to 4 x 10-2 substitutions per nucleotide and year [(Sobrino et al., 1986); reviewed in (Domingo et al., 2001)]. Based on sequence comparisons of FMDV A, 0 and C in areas of the world where outbreaks were produced by the introduction of a strain that was identified, it has been estimated that the VP1-coding region evolves at a rate of approximately 10-2 substitutions per nucleotide and year (Knowles and Samuel, 2003). These values are comparable to those calculated for other RNA viruses, and are 104- to 107fold larger than rates of evolution estimated for cellular genes (Table 1). It must be emphasized that rates for FMDV vary depending on the viral gene (or gene segment) under study, and on the time span between the isolations of the FMDVs being compared (Sobrino et al., 1986). For short time periods, a constant rate of evolution during a disease episode may be observed, and such a constancy has been attributed to random sampling of mutants from the quasispecies that are generated in each infected animal (Villaverde et al., 1991). For longer time periods, population equilibrium and disequilibrium in infected hosts, that is, perturbations that change the complexity and the consensus sequence of the mutant spectrum, render unlikely a constancy of evolutionary rates. In terms commonly used in population genetics, very often a "clock" (constancy of accumulation of mutations with time) does not operate during RNA virus evolution. This is an expected consequence of quasispecies dynamics (Section 4) and renders questionable the dating of viral ancestors based on comparisons of sequential genomic sequences [reviews in (Coffin, 1990; Domingo et ai., 2001)]. This is an added difficulty to the requirements for measuring evolutionary rates which are: (i) an unambiguous identification of the founder genome, (ii) determination of its time of appearance in the natural environment in which evolution is studied, and (iii) a reconstruction of the evolutionary lineages of its viral progeny (Hillis et al., 1996). One of the largest rates of evolution (7.4 x 10-2 substitutions per nucleotide and year!) recorded for an RNA virus was observed during persistent infections of FMDV in cattle established experimentally with biological clones of the virus. It is difficult to conceive such rates of variation for extended periods of time, given the requirements for functionality of genomes with such a compact genetic information. Comparison of amino acid sequences of antigenic sites over six Copyright © 2004 By Horizon Bioscience

Table 1. Examples of rates of evolution of viruses Virus

Genomic Region Analyzed

VPl c

Substitutions Per Nucleotide Per Year 3PD50, see below). For example, the post vaccination sera of the highly potent 01 BFS vaccine of the European Vaccine Bank(s) showed cross neutralisation levels that indicated full protection against 6 different field strains from the Middle East (Barteling and Swam, 1996). Some of these strains differed as much as 20% from the 01 BFS sequence in the routine RNA analysis. For all these reasons Veterinary services must be reluctant to ask for vaccines based on local field strains. Also, producers should not be lead solely by their commercial interests in offering so-called tailor-made vaccines with the latest field variant incorporated into the vaccine. Often, such vaccines are marketed before (in the context of GMP) all laboratory tests (e.g. shelf life of vaccines) have been carried out. In South-America the only vaccine strains that are allowed are those which are, in close co-operation with the national laboratories, carefully selected by the PANAFTOSA Laboratory in Rio de Janeiro. Some strains, e.g. the 01 Campos strain, has been successfully used since the 1960s. 4.4. Preparation of Seed Virus

Once the vaccine strains are known, they have to be adapted to grow in the cells used for virus production, in general, BHK-suspension cultures. Most FMD virus strains grow better in BHK cells grown in monolayer than in suspension (Spier and Whiteside, 1976; Bolwell et al., 1989). The fundamental background of that phenomenon is not known. Anyhow, a usual procedure is to first passage the virus in BHK-monolayer cells and when a rapid (e.g. overnight) cyto-pathogenic effect is observed, to passage the virus in suspension until the virus grows well (Radlett et ale 1985) and a "mother" seed stock and successively a "master" seed stock can be prepared. After these stocks have been tested for extraneous agents (e.g. BVD virus, mycoplasma's) the master stock can be used to produce a working stock. In general this is done at a scale of several litres that are stored away at -70°C in smaller portions. When they have been titrated they can be used at a fixed multiplicity of infection (m.oj.), e.g. 1 infectious unit per 100 cells. Another possibility is to make a seed culture in a small fermentor - always under the same conditions - and then add this culture to the large production fermentor. Virus cultures can best be carried out at temperatures from 34°C - 36°C and a pH of 7.2 (Radlett et ale 1985). Because small variations in culture conditions will cause differences in virus yields, better standardization is obtained if a large working stock is prepared that

- after concentration - is stored frozen at -70°C and that can serve for a whole series of production batches. The concentration can be carried out by ultra-filtration or by precipitation with polyethylene glycol (see below). In all cases it is - in the context of GMP - advisable to sample the used seed virus shortly before addition to the large culture tank and to verify its infectivity by titration in susceptible cells. 5. In Process Controls Good manufacturing practice requires 'in process control' of each step. The crude culture, the virus harvest (including filtration steps), inactivation, and additional antigen processing steps must all be checked for yields in live virus or inactivated Copyright © 2004 By Horizon Bioscience

1318 Inactivated FMD Vaccines

(146 S) antigen and checked for other parameters (e.g. sterility) that are crucial for the production process. Checks and controls of the functionality of equipment (pH, temperature etc.) could - under the supervision of the QAIQC management - be carried out by production operators and should be recorded on record sheets accompanying each step of the production process. Because proper inactivation of the harvested virus is such a critical step for the safety of the final vaccine, this step should be subjected to external control (e.g., that provided for in the European Pharmacopoeia). The 'in process" control of proper inactivation plots together with the inactivation of the virus will be discussed in Section 8. 5.1. Biological Tests

Since the early 1950s, the only methods for quantification of live FMD virus were titration in cattle tongue (Henderson, 1949) and in suckling mice (Skinner, 1951). In the 1960s, several in vitro methods using cell cultures became available. Primary (foetal) calf thyroid cells are the most sensitive cells for FMDV (Snowdon, 1966). However, these cells are not always available. Therefore, IBRS 2 cells (De Castro, 1964) are often preferred, or a limiting dilution assay in BHK-monolayer cells (either in tubes or, more conveniently, in micro-titer plates) (Wagner and Me Vigar, 1970). Plaque titration is performed in BHK-monolayer cells with agar overlay (Anderson et al., 1970) or in an agar cell suspension plaque assay in which cells come from a large cell stock, stored over liquid nitrogen (Barteling, 1972). Although these biological tests do not tell much about antigenic mass for vaccine formulation, they are required for titration of seed virus, to enable - during virus seeding - standardizing of the multiplicity of infection. Titration of biological activity (e.g. by a plaque test) is also required to verify that the inactivation kinetics are correct. 5.2. Immunological Tests

A quantitative complement fixation (CF) test was described in 1964 (Bradish et ai.). In this test the total antigenic mass is estimated, Le. intact 146 S virus particles together with subunit 12 S antigen. After removal of 12 S antigen by fluorocarbon treatment (Brown and Cartwright, 1960) or of 146 S antigen by ultra-centrifugation (Terpstra et al., 1976), the CF mass of the major 146 S immunogen can directly or indirectly be estimated. Monoclonal antibodies can also be used in a quantitative enzyme-linked immuno-sorbent assay (ELISA). In this test (Van Maanen and

Terpsara, 1990) only integer 146 S particles are estimated, while virus particles with reduced immunogenicity, e.g. which have been attacked by proteolytic enzymes (Wild and Brown, 1967; Barteling et al. 1979; Doel and Collen, 1984), do not react. 5.3. Quantitative Sucrose Density Gradient Analysis

Using the fact that the FMD virus particle has a sedimentation coefficient of 146 S, the concentration of FMDV antigen can also be estimated (in micro-grams per ml) by quantitative sucrose density gradient analysis (Fayet et ai., 1971; Barteling and Meloen, 1974). The method is reproducible and by using a small ultra-speed rotor, six virus culture samples can be assayed in one hour (Barteling and Meloen, Copyright © 2004 By Horizon Bioscience

Barteling

3191

1974). The virus peak, obtained with UV spectro-photometry, can be analyzed by hand but the test becomes somewhat more precise if the calculation of the peak surface is carried out by a computer (Doel et aI., 1982). The results of the 146 S test correlate well with the results of the quantitative CF test and with protection tests of some FMDV types (Black et aI., 1986). The 146 S test is now a standard, internationally accepted technique. In fact the 146 S is a minimum assay: virus particles occurring in complexes will sediment to lower positions in the gradient or to the bottom of the centrifuge tube and will not contribute to the size of the 146 S peak. Pre-treatment of the samples with trypsin (1 Jig/ml for 30 min) may disintegrate the complexes (Bolwell, et al., 1989) and thus result in higher peaks and a more realistic estimation of the 146 S contents. Another possibility is to do iso-picnic ultra-centrifugation (overnight) in caesium chloride. The 146 S test must be carried out on samples of the virus harvest and of all steps of the down-stream processing of the inactivated virus antigen (concentration and purification). 6. Cell Culture The art of cell production is to reach optimal cell densities without depletion of the medium or without building up toxic levels of break-down products that might hamper virus growth or cell growth in a following production cycle. To the author's knowledge, many vaccine producers still use the process as developed in the seventies (Telling, 1975; Radlet et aI., 1985). The cells are grown in batch cultures starting at 0.5 million cells per ml reaching cell densities in the range of 2.5 million in 2 or 3 days. Often, weekly cell growth cycles are maintained of two cultures lasting 2 days and one culture lasting 3 days, the last culture starting at Friday with a somewhat lower cell concentration (e.g. 0.3 million cells per ml). Often, cells are maintained in a volume just below the maximum level, e.g. at 500 L (to feed the 2.500 L fermentors), as long as they were "happily growing", sometimes for months. Contamination or difficulties with virus production might well finish this happiness. With the continuation of cultures there is an increasing risk of contamination. Hampered by the presence of antibiotics, a low-level of contamination might at first not cause problems and might (undetected) more or less be co-cultivated with the cells. At a later stage the bacteria might break through causing serious problems over a length of time. Also, long-term cultivation of cells might alter the properties of the cell population, which might impair the reproduction cycle of FMD virus, the source of the (enduring) problems being unclear. When, finally, it is decided to start from the liquid nitrogen cell stock, it will last another (lost) two weeks, before the thousand L scale is reached again. Thus, maintaining cells over a long period of time is not GMP. The optimum approach is to start every week with cells from the "working" stock and scaling-up according to standardised schedules. 7. Virus Production For virus production the cells must be transferred from the "clean" cell culture laboratory where activities might well be combined with e.g. medium preparation, to the bio-containment section where all the work with live virus is carried out. In the classical process, cells, having reached optimal densities (e.g. 2.5 million cells Copyright © 2004 By Horizon Bioscience

1320 Inactivated FMD Vaccines

per ml), are sedimented overnight and the supernatant, serum-containing medium is removed and replaced by serum-free medium. The final cell culture can either be carried out at the "clean" cell culture side or in the same fermentor as was used for virus culture. The latter option saves a (second) big fermentor but, since the serum containing-medium (the removed medium) originated from the bio-containment section of the laboratory, it has to be treated for virus contamination. In general the remaining cell-sediment slurry represents 5 - 10% of the original medium, which means that the virus culture still contains some serum proteins. In general, the cell slurry is supplied with fresh serum-free medium and pumped (or pressed) into the virus production fermentor where seed virus is added. In general, 1 infectious unit per 100 cells gives sufficient cyto-pathogenic effect (>90% of the cells affected) in 20 hours time. Then the virus can be harvested. An alternative way of virus production has been described in the eighties by the author as a "unit-process". Medium with PEG-treated serum is used for both, cell and virus culture, without sedimentation of cells in between. Successively, the inactivated virus antigen is concentrated and purified by two cycles of PEG-precipitation (the large precipitable serum proteins having previously been removed). Generally virus can be harvested after 20 - 24 hour. Cell debris must be removed first. Often this is done by the addition of 0.5% chloroform Uust above the saturation level) to the cooled virus harvest. The chloroform is mixed intensively for 1 or 2 hours and then the precipitates are sedimented either by gravity or by centrifugation. Additional clarification can be carried out by filtration. By both procedures approximately 5 - 1. 0% of the virus harvest is lost. If the removal of cell debris is carried out by addition of the filter-aid (e.g. purified calcified diatomaceous earth) and filtration, the complete virus harvest can be recovered by washing the filter with buffer. Successively, the virus harvest can be inactivated. Chloroform treatment has the advantage that it purifies the antigen somewhat and that it will remove or reduce contamination. However, in a GMP regime the last cannot be an argument for its use. A disadvantage of chloroform treatment is that some precipitation often continues during the following inactivation step with the risk that active virus might be captured inside the complexes, becoming inaccessible to the inactivating agent. If the inactivated product is used directly for vaccine formulation, or is used after concentration by ultra-filtration, these complexes will go into the vaccine. The difficulty of inactivating virus trapped in complexes is, for instance, seen when trying to inactivate PEG-concentrated virus (Barteling, 1983). For that reason it is strongly recommended to carry out inactivation before the concentration step.

8. Inactivation and 'In Process' Safety Testing Virus inactivation and safety tests are the most critical steps in FMD vaccine production. As outlined above, some FMD outbreaks in Europe have been clearly linked to vaccination (King et al., 1981; Beck and Strohmaier, 1987), indicating that formaldehyde did not always inactivate FMD virus properly. Later on, in the eighties most vaccine production laboratories changed to inactivation by aziridins, binary ethylene immine (BEl) in particular (Bahnemann, 1975 and 1990), giving, in principle, first order inactivation kinetics and, consequently, a safer product. Some virus particles might escape inactivation if they do not come into contact with the inactivant, for instance at the lid of the container or at "dead" spots Copyright © 2004 By Horizon Bioscience

Barte/ing

3211

inside tubes or valves. Therefore, to be absolutely sure that all virus particles are properly inactivated, the virus-inactivant mixture must be pumped into a second vessel, where the inactivation process is completed (EP). Safety tests can fail to detect low residual-infectivity in vaccine batches. The cattle test as prescribed by the European Pharmacopoeia (1997) was often considered a final proof of safety. However, in the 3 cattle that are used, only 6 ml of vaccine can be tested. Even, if a 1 litre sample of a 1000-liter batch of vaccine is checked in vitro on cell cultures for remaining infectivity (Anderson, 1970) and found negative, the whole batch may still contain a 1000 infectious particles, sufficient virus to bring down at least one cow with FMD. The safety of a vaccine batch can far better be guaranteed by verifying the inactivation kinetics. In a logarithmic plot, the whole inactivation line, or at least its last measured part, should be linear. The straight line enables extrapolation to the end of the inactivation period. Therefore, the inactivation kinetics of every vaccine batch must be checked. Test samples can be stored frozen as back-ups to enable repeated tests if necessary. Under certain conditions the inactivation plots obtained by aziridins can also show tailing-off. For instance, inactivation with acetyl-ethylenimine of (PEG) concentrated FMDV, showed tailing-off (Barteling, 1983). This example clearly shows that the estimation of inactivation kinetics, as part of "in process control", is absolutely necessary. For the labile (SAT2) vaccine strains it has been shown that stability of the vaccines increased significantly if, after inactivation with aziridin, the antigen was treated with formaldehyde (Rowlands et aI, 1972; Mowat et al., 1973). Recently, a synergistic effect has been reported (for 5 different SAT strains) if formaldehyde and BEl are used simultaneously, resulting in a more than lOO-fold increased inactivation rate and, therefore, resulting in improved safety (Barteling and Cassim, 1999). Sufficient inactivation can be attained within 8 hours - instead of over 40 hours - which makes production schedules more flexible. Also, where (cellular) proteases might, during the inactivation process, impair the quantity and quality of antigens, the shorter duration and the fixation of the antigens (and proteases) might well favourably influence the quality of the antigens. Again, at the end of the inactivation, a large sample must be screened for residual viral activity by inoculation of cultures of highly susceptible cells, e.g. BHK monolayer (Anderson et al., 1970) or suspension cells (Barteling, 1983). 9. Concentration and Purification of the Antigen In general concentration and purification of the inactivated antigen is carried out when the results of bacteriological and safety tests are not yet available. Therefore, the antigen must be considered to contain active virus and these processes must be carried out in a high-containment environment. However, recontamination with live virus must be excluded. Therefore this high-containment "quarantine" area must be absolutely separated from the rest of the facilities. After concentration and purification in the "quarantine" area the antigen can be stored at ultra-low temperatures (e.g. at -70°C) until its safety is verified. Concentrated antigen can then be stored outside at even lower temperatures e.g. over liquid nitrogen, where it is readily available for vaccine preparation. With concentrated purified antigen, smaller dose volumes (e.g. 5 ml or smaller) of trivalent vaccine can be formulated. Copyright © 2004 By Horizon Bioscience

\322

Inactivated FMD Vaccines

Finally, since FMD vaccines can only be stored for a limited number of years, vaccine banks prefer storage of highly concentrated, inactivated antigen at ultralow temperatures (Barteling and de Leeuw 1981). International or regional vaccine banks contain large quantities of such concentrated antigen, enabling the rapid generation of vaccine against several FMD serotypes, should an FMD outbreak occur. 9.1. Treatment with Chloroform Virus obtained from tongues of infected cattle, was often heavily contaminated with bacteria and other micro-organisms. Also Frenkel cultures were often not sterile but for both products contamination could be reduced considerably by vigorous mixing with chloroform (Waldmann, 1937; Frenkel, 1947). The chloroform is added to the saturation point (0.375%) or slightly above (0.5%). Cell debris and denatured proteins are removed from the virus harvest by centrifugation and/or filtration. 9.2. Sedimentation of AI(OH)3-Antigen Complex Adsorption of antigen to AI(OH)3 gel followed by (gravity) sedimentation of the complex offers a simple and efficient concentration method (Matheka, 1959). If the virus antigen was first treated with chloroform (see above) and then adsorbed to a relatively small quantity of AI(OH)3 gel, concentration factors of 10-15 times could be obtained. 9.3. Ultra-Filtration Ultra-filtration systems for concentrating FMDV were successfully pioneered by Strohmaier (1967). Later, hollow fibre systems were applied by others (Morrow et al., 1974) and both systems have been introduced by many vaccine manufacturers in South America to obtain the high antigen concentrations necessary for oiladjuvant vaccine formulation. However, even if filter materials with cut-off values of 100,000 D (molecular weight) are used, purification of the antigen by ultra-filtration is limited. In the author's experience at cut-offs of 200,000 D and higher, considerable amounts of 146 S are lost probably because antigen is caught in the network of the filter. Ultra-filtration is also used by one of the great vaccine producers (Doel, 2003). For purification of the antigen it is combined with chromatography, but what type of chromatography or any other details were not given. 9.4. Precipitation with Polyethylene Glycol (PEG) Depending on chemical and physical conditions PEG precipitates FMDV fairly specifically (Wagner et al., 1970). One precipitation step with PEG is sufficient to remove allergenic components from BHK vaccines (Kaaden et al., 1971; Panina and De Simone, 1973). Serum proteins do not co-precipitate at this stage if PEGtreated serum is used for virus culture. The precipitated antigen can be collected by sedimentation or by filtration with filter-aid (Barteling and De Leeuw, 1981) which is more convenient. The antigen collected in the filter-aid can easily be eluted in a small volume of buffer that is recycled through the filter. Highly purified products can be obtained after a second precipitation-elution cycle and UV-extinction

Copyright © 2004 By Horizon Bioscience

Bartelinq

3231

profiles of sucrose density gradients show almost exclusively 146 S antigen. No a-NSP antibodies were detected in cattle that had been repeatedly vaccinated with vaccines prepared from such antigens (Bruderer et ale in press). 10. Vaccine Formulation The inactivated (purified) antigen must be mixed with an immune stimulant, the adjuvant, to obtain a proper immune response. Although numerous substances with adjuvant activity are known, only three are used for FMD vaccines: AI(OH)3 gel, saponin, and oil emulsion. When the first two types of adjuvants are used, one speaks of aqueous vaccines. In oil emulsion vaccines, as introduced by Freund (1945), the antigen is emulsified in a light mineral oil with an emulsifying agent, mannide-mono-oleate. For further stimulation of the immune response Freund also added bacterial cell wall components. Because of adverse effects at the injection site, for veterinary practice, only the incomplete (without bacterial cell wall components) formulation, is acceptable (Mc Kercher, 1977). Some producers emulsify the primary emulsion once more in buffered saline containing Tween 80, which gives the product a low viscosity. These types of oil-based vaccines, developed by Herbert (1965) and for FMD introduced by Mowat (1972), are called double emulsion vaccines. 10.1. Aqueous Vaccines

The dose volume of the aqueous vaccine was gradually reduced from 60 ml (Waldmann, 1937) to 5 ml of concentrated antigen-AI(OH)3 complex and (purified) saponin (1 - 5 mg per dose). In Europe a similar formulation was used until 1992 vaccination was discontinued and it is still used in vaccination programmes in the Middle East and on the Asian continent. 10.2. Oil Adjuvant Vaccines

For the preparation of oil vaccines the antigen is emulsified in light mineral oil that contains 10% mannide mono-oleate (MMO). The purity of the ingredients, that of the MMO in particular, is crucial for the stability of the vaccine and, after injection of the vaccine, for the absence of adverse reactions (Mc Kercher and Graves, 1977). MMO must now pass a toxicity test before it is accepted for application in vaccines. At room temperature MMO is not stable and, therefore, long-term storage should be in the cold-room, or even better, at -20°C. Emulsions can be prepared with a colloid mill, a continuous mechanical emulsifier, or a continuous flow ultrasonic emulsifier. The water in oil emulsion must be stabile for more than one year. When the water in oil emulsion is emulsified once more in an aqueous phase containing 2% Tween 80 or a similar polysorbate, a double emulsion is obtained: water in oil in water (Herbert, 1965; Mo\vat, 1974). While the single emulsion is stable, some separation of the last aqueous phase might occur in the double emulsion vaccine. Shaking will restore the homogeneity of the vaccine. Because of their low viscosity the double emulsion vaccines are easy to inject. Good results have also been obtained with ISA 206 from Seppic, Paris (Salt et al. 1997). By a simple mixing procedure the latter emulsifies the antigen into a kind of double emulsion that is formed spontaneously. The double emulsion type vaccines prepared from the antigen that is purified and concentrated by sequential PEG precipitation can be used for both pigs and cattle (Barteling and De Leeuw 1981). Copyright © 2004 By Horizon Bioscience

\324

Inactivated FMD Vaccines

11. Approval of Vaccines Once vaccine formulation, bottling conditions, and labels fulfil the requirements of the European Pharmacopoeia (1997) or other national or international standards, safety and potency tests must be carried out under the responsibility of (independent) disease control authorities. 11.1. Safety

We already described the "in process" safety tests, which give the producer the best guarantee for a safe vaccine. Because proper inactivation of virus antigen is of crucial national (and international) importance, in the author's view national control authorities should require inactivation plots demonstrating sufficient rapid inactivation of every production batch. In addition absence of remaining virus activity can be demonstrated on (concentrated) inactivated antigen used for vaccine preparation or on the final vaccine. However the last tests will be rather "cosmetic" than being a test that really guarantees freedom of active FMD virus e.g. in batches (like in South America) contain 2 million doses. We will discuss briefly the test procedures for the final product (for a more extended discussion see Barteling and Vreeswijk, 1991). A sufficiently large sample (representing 200 doses or more) should be tested in cells of proven sensitivity. Because AI(OH)3 is toxic for cells aqueous vaccines cannot be tested in cell cultures. The (virus) antigen must be eluted first from the gel by 0.3 M sodium phosphate buffer or, more efficiently, by 1.2 M potassium phosphate (Barteling, 1983). After dialysis and concentration by ultra-filtration or by precipitation with PEG, the antigen can be tested in cells. Elution and concentration procedures may not destroy or harm FMDV because if they do, residual virus will be missed. For validation of the procedures, a standard live virus, identical to the vaccine strain(s) should be added to a paired vaccine sample. Sufficient (e.g. > 50%) recovery of the standard live virus will validate the test. This validation, although not prescribed by the European Pharmacopeia, is essential for the test procedure as a whole. The inactivated antigen should not interfere with the detection of residual live virus (Wittmann, 1964; Barteling, 1983). In the classical safety test the vaccine is inoculated in cattle tongue (intradermolingually), the part of the animal that is most susceptible to FMDV (Henderson, 1949). Because only small volumes (2 ml per animal) can be tested and only three cattle are used, this test certainly cannot guarantee that vaccine batches are free of live virus. The intramuscular injection prescribed as well by the E.P. will not contribute to a better detection because detection of FMDV by this route is several logarithms less accurate than by the intra-dermolingual route (Henderson, 1949) or by in vitro titration of virus in susceptible cells. The intradermolingual test - in the target animal - as prescribed by the E.P. is still used for psychological reasons and the test will only reveal gross contamination, which can be readily detected by in vitro tests. It is understood that, for the new version of the E.P., several of the mentioned elements - including the in vivo safety tests - will be reconsidered (De Clercq, 2001).

Copyright © 2004 By Horizon Bioscience

Barteling

3251

11.2. Potency and Duration of Immunity

The European Pharmacopeia (1997) prescribes that a potency test in cattle must be performed before a vaccine can be released on the market. The test (Terre et al., 1976) estimates the 50% protective dose (PD50) that is the volume of the vaccine that protects three weeks after vaccination 50% of the animals after challenge. Three groups of fi ve cattle are vaccinated each with fourfold dilutions of the vaccine in carbonate buffer. Cattle should not have been vaccinated earlier and must be free of antibodies against FMDV. Three weeks after vaccination, the cattle are challenged intra-dermolingually with at least 10,000 ID50 (50% bovine infectious dose) of the virus strain incorporated in the vaccine. If the feet and other areas (except the injection sites on the tongue) remain free of lesions for at least 8 days, cattle are considered protected. Control animals must develop lesions on at least three feet. In most European countries, 3 PD50 are required for a vaccine to pass the test. The prescription of the European Pharmacopoeia for the PD50 test cannot be used to oil-based vaccines because proper dilutions of the oil emulsion cannot be made with (aqueous) buffer solutions. Alternatively fourfold reduction of the dose volume can be applied. Challenge tests in cattle should be reduced to a minimum. They are expensive because the animals must be destroyed afterwards. The test also imposes ethical problems. The animals clearly suffer and quality food is destroyed. For a control authority it is only important to verify that the vaccines are better than 3 PD 50. If protection is observed in the group vaccinated with the lower (1/16) dose one may expect a PD50 value clearly above 3. Reversibly, if the real value of the vaccine is around 3 PD50 one cannot expect any cattle vaccinated with a 1/16 dose to be protected and, therefore, this group can be left from the test. Also, if in this lower group 1 protected animal lifts the vaccine to just above 3 PD50 the confidence limits of the test will become unacceptable. Therefore, the author proposed a "truncated" test to be used (Barteling, 1998) in which cattle are vaccinated with half a dose and with 1/8 of a dose. In South America it must be demonstrated for every batch of vaccine that twelve out of 16 cattle vaccinated with the vaccine are protected (Argentina requires 13) against challenge (Vianna Filho et al., 1993). If only 10 or 11 (Argentina requires lIar 12) cattle are protected the manufacturer could request a second test (paid for by the manufacturer). If the second test fails, the whole batch is placed under embargo, resulting in the destruction of 2 million or more doses of vaccine. Since there is a good correlation between serum antibody titers and protection of cattle, vaccine potency can better be assessed by serology (Black et al., 1984; Pay and Hingley, 1992). Therefore, in Brazil nowadays the oil emulsion vaccines are tested and accepted (or rejected) on the basis of induced antibody levels at 6 weeks after vaccination. Correlation between antibody levels and protection was, under the supervision of PANAFfOSA, estimated in an extended study in which a number of South American test laboratories participated (Periolo et al., 1993). For this study the sera from about a 1000 protection tests were used to establish an expected level of protection for each antibody titer. For simplicity and uniformity (standardization) an indirect ELISA became the method of choice. Nowadays the expected percentage of protection is calculated for the whole group and the vaccine will pass only if the lower confidence limit of the mean is 80% or higher. However Copyright © 2004 By Horizon Bioscience

\326

Inactivated FMD Vaccines

for the registration of the oil-emulsion vaccines produced in South America, the presence of sufficiently high antibody levels must be demonstrated 3 months after vaccination. There are no other, international, rules in existance for testing that the duration of immunity is adequate. 12. Vaccine Banks When FMD disappeared from Europe, it became even more important to be prepared for a new outbreak. In that context special attention was paid to the protection of pigs because they were not sufficiently protected by the classical aqueous vaccines. With the intensifying pig breeding industry it was clear that in an outbreak, the availability of large quantities of vaccine might be needed for this species as well. To that end inactivated antigen was purified and concentrated and stored at ultra-low temperatures. From these antigens (oil-emulsion) vaccines could be rapidly formulated, vaccines that protected both cattle and pigs (Mc Kercher and Bachrach, 1976; Doel and David, 1984). To enable a rapid response to a FMD outbreak, the first international vaccine bank was established in Pirbright (U.K.). The U.K. and a number of other countries participate in this effort. The quantities of vaccine contained in this bank (0.5 million dose per type) and the fractions that participating countries can claim are only sufficient for a first strike response against a limited outbreak. These quantities certainly cannot cover a more widespread outbreak or an outbreak in an area with a dense pig population. When Europe adopted a non-vaccination policy, European vaccine banks were set up also. Stored in these banks are a total of 5 million doses of each of 4 main serotypes and 2.5 million of some of the other less threatening strains. There is also a North American vaccine bank in which Mexico, Canada and the U.S. participate. To cover a broader variation of field strains that might occur during outbreaks, international vaccine banks require highly potent vaccines of at least 6 PD50 (3 PD50 is standard). These vaccines induced high antibody levels that not only neutralized the vaccine strain, but also a wide variety of field strains (Barteling and Swam, 1996) suggesting protection against a wide variety of strains. Unfortunately, these data are so far not supported by challenge experiments. Whether all the antigens in vaccine banks are sufficiently purified and do not induce a-NSP antibodies to allow for a-NSP testing (see below) has not yet been verified. 13. Performance of Emergency Vaccines

Several studies with potent vaccines - as contained in the international vaccine banks - showed that soon (3 to 5 days) after vaccination pigs and cattle are protected against (contact) challenge (Sellers and Herniman, 1974; Doel et al., 1994; Swam et al., 1994; Salt et aI., 1995; 1997; 1998; Cox et al., 1999). These data are in agreement with experiences in the field. However, it is feared that vaccinated cattle become carriers (Van Bekkum et al., 1959; Sutmoller and Mc Vigar, 1972) and that these carriers will cause new outbreaks. This fear (and its associated rules for international trade) hampers trade - after control of outbreaks by vaccination - to a larger extent than after control by stamping-out. Therefore, vaccination is often not used.

Copyright © 2004 By Horizon Bioscience

Barteling

327\

The fear of transmission of FMD by vaccinated carriers (or their products) seems to be unjustified. In the past,in Europe and more recently in South America, FMD has been controlled by vaccination of the cattle population only. The millions of unvaccinated pigs and sheep - in South America often grazing on the same pastures as the vaccinated cattle - did not become infected. During the past 15 years all outbreaks that have been controlled by vaccination have not resulted in recurrent disease (Leforban, 2002; Barteling and Sutmoller, 2002) even if all susceptible livestock was vaccinated. Certainly the risk that remaining (vaccinated) carriers transmit the disease must be very small and near to zero. In a previous review paper (Barteling, 2002) the author has indicated that emergency vaccination of e.g. less than 10% of the cattle population must represent an accordingly smaller risk and that consequently control of outbreaks by stampingout and by (ring-) vaccination might well deserve the same consequences for trade. In both cases a successive serological survey (e.g. by screening for a-NSP antibodies) can then demonstrate to the international community the absence of foci of active virus. Of course importing countries can ask for animals free of FMDassociated antibodies. However, if veterinary services can use vaccination for the control of outbreaks, without major constraint on export trade and with the rapid availability of quantities of vaccine from vaccine banks, FMD does not need to be the disease causing national catastrophes like in the UK. Also, deliberate spread of FMD by terrorists will become less threatening. 14. Future Vaccines The progress in (classic) FMD vaccine production was primarily directed towards safety of the vaccine, purity of the antigen, selection of a proper adjuvant, endurance of immunity, and adequate procedures for storage of inactivated FMDV antigen in vaccine banks. In addition, much research was directed towards the development of what might be called "constructed vaccines". Although antigenic proteins and epitopes have been known for two decades (Cavanagh et a!., 1977; Strohmaier et al., 1982; Rowlands et a!., 1983; Geysen et a!., 1984; for a review see Barteling, 1988), and some studies showed very promising results (e.g. Dimarchi et al., 1986), so far no vaccines have been constructed that can compete with the classical inactivated FMD vaccines. Unless a vaccine is designed that protects against all FMD strains, classical inactivated vaccines will have the advantage that they probably can be produced faster in response to new FMD variants. Since classical vaccines will continue to be necessary, research must continue to improve these vaccines as well, ideally to induce lifelong protection after one or two vaccinations. 15. World-Wide Eradication of FMD? The requirement to produce vaccines according rules for GMP represents a great dilemma for developing countries; many of which have the problem of endemic FMD. To control the disease they need quantities of vaccine which they often cannot afford to import. Therefore, in general, they want to have their own FMD vaccine production. Also, there are arguments of employment, training, education and independence that are important when it comes to the control of a critical disease like FMD. However, an often limited market and a vaccine price that must be kept low, does not allow the large investments (in the range of 5 - 10 million U.S. Copyright © 2004 By Horizon Bioscience

\328 Inactivated FMD Vaccines

dollar) that are required for a modern (GMP) vaccine plant. Even if, for example with the help of international aid, a GMP plant were put in place, it would be almost impossible to maintain at the required level, without compromising standards. An intermediary solution could be for a country (or region) to buy the (inactivated) FMD antigen in bulk on the international market and to carry out the vaccine formulation and bottling locally. To enable this, antigen/vaccine banks containing sufficient quantities of concentrated (purified) antigen should be established (to allow vaccine production in the event of an outbreak). If this were established and if sufficient political ,vill were mobilized in all countries, there may come a day in the future when outbreaks of FMD become rare events, and some FMD sero-types (e.g. type C) become history. References Anderson, E.C., Capstick, P.B., Mowat, G.N., and Lech, F.B. 1970. In vitro method for safety testing of foot-and-mouth disease vaccines. J. Hyg. Cambro 68: 159. Bahnemann, H.G. 1975. Binary ethylenimine as an inactivant for foot-and-mouth disease virus and its application for vaccine production. Arch. Virol. 47: 47. Bahnemann, H.G. 1990. Inactivation of viral antigens for vaccine preparation with particular reference to the application of binary ethylenimine. Vaccine 8: 299. Barteling, S.J. 1977. Certain aspects of Foot-and-Mouth Disease. Virus production in growing BHK suspended cell cultures. Intern. Symp. on foot-and-mouth disease, Lyon 1976. Develop. BioI. Stand. 35: 55. Barteling, S.J. 1972. The use of frozen cells in the agar-cell suspension titration technique. Arch. Ges. Virusforsch. 38: 271. Barteling, S.J. 1983. Innocuity testing of foot-and-mouth disease vaccines. ILAziridineinactivated antigen produced in baby hamster kidney cells. J. BioI. Stand. 11: 305. Barteling, S.J. 1987. Foot-and-mouth disease vaccine production and research in The Netherlands. The Veterinary Quarterly 9: 5S. Barteling, S.J. 1988. Possibilities and limitations of synthetic peptide vaccines. In: Synthetic Peptides in Biotechnology. vol. A 10. Mizrahi A., ed. New York: Alan R. Liss. p. 25. Barteling, S.J. 1998. The 'truncated' potency test. Rep. Res. Gr. Eur. Comm. Contr. FMD, Brasov, Rumenia, FAO Rome p.29. Barteling, S.J. 2002. Development and performance of inactivated vaccines against footand-mouth disease. Rev. sci. tech. Off. Int. Epiz. 21 (3): 577 Barteling, S.J. and De Leeuw, P.W. 1981. The use of stored concentrated antigens for the preparation of foot-and-mouth disease vaccines. Rep. Res. Gr. Eur. Comm. Contr. FMD, Vienna, Austria, FAD Rome p. 51. Barteling, S.J. and Ismael Cassim, N. 2000. Formaldehyde increases the BEl-inactivation rate of foot-and-mouth virus at least with a hundred-fold. Rep. Res. Gr. Eur. Comm. Contr. FMD, Borovets, Bulgaria. FAO, Rome. p.19. Barteling S.J. and Meloen R.H. 1974. A simple method for the quantification of 140S particles of foot-and-mouth disease virus. Arch Ges. Virusforsch. 45: 362. Barteling, S.J. and Swam, H. 1996. Expected cross-protection of aqueous and (double) oilemulsion type 0 foot-and-mouth disease (FMD) emergency vaccines. Rep. Res. Gr. Eur. Comm. Contr. FMD, Haaretz, Israel, FAO Rome. p.90. Barteling, S.J. and Vreeswijk, J. 1991. Developments in foot-and-mouth disease vaccines. Review. Vaccine 9: 75. Barteling, S.J. and Woortmeijer, R. 1984. Formaldehyde inactivation of foot-and-mouth disease virus. Conditions for the preparation of safe vaccine. Arch. Virol. 80: 103. Barteling, S.J., Meloen, R.H., Wagenaar, F., and Gielkens, A.L.J. 1979. Isolation and characterization of trypsin-resistant 01 variants of foot-and-mouth disease virus. J. Gen. Virol. 43: 383. Copyright © 2004 By Horizon Bioscience

Barteling

3291

Bauer, K., Kaaden, O.R. and Mussgay, M. 1970. Experimentelle Untersuchungen uber Allergien vom Spattyp nach der Schutzimpfung von Rindem mit Maul- und Klauenseuche (MKS)-Vakzinen. Berl. Munch. Tierarztl. Wschr. 83: 292. Beck, E. and Strohmaier, K. 1987. Subtyping of European foot-and-mouth disease virus strains by nucleotide sequence determination. 1. Virol. 61: 1621. Bekkum, J.G. van, Bool, P.H., and Vermeulen, CJ. 1967. Experience with the vaccination of pigs for the control of foot-and-mouth disease in the Netherlands. Tijdschr. Diergeneesk. 92: 87. Bekkum J.G. van, Frenkel H.S., Fredericks H.HJ., and Frenkel S. 1959. Observations on the carrier state of cattle exposed to foot and mouth disease virus. Tijdschr. Diergeneesk. 84: 1159. Black, L., Nichols, MJ., Rweyemamu, M.M., Ferrari, R., and Zunino, M.A. 1986. Foot-andmouth disease vaccination: a multifactorial study of the influence of antigen dose and potentially competitive immunogens on the response of cattle of different ages. Res. vet. Sci. 40: 303. Black, L., Francis, MJ., Rweyemamu, M.M.~ Umehara, 0., and Bage, A. 1984. The relationship between serum antibody titres and protection from foot-and-mouth disease in pigs after oil emulsion vaccination. J. BioI. Stand. 12: 379. Bolwell, C., Brown, A.L., Barnett, P.V., Campbell, R.O., Clarke, B.E., Parry, N.R., Ouldridge, E.J., Brown, F. and Rowlands, D.J. 1989. Host cell selection of antigenic variants of foot-and-mouth disease virus. J. Gen. Virol. 70: 45. Bradish, CJ., Jowet, R., and Kirkham, J.B. 1964. The fixation of complement by virusantibody complexes: equivalence and inhibition in the reactions of the viruses of tomato bushy shunt and foot-and-mouth disease with rabbit and guinea-pig antisera. J. Gen. Microbiol. 35: 27. Brown, F., Cartwright, B., and Stewart, DL 1963. The effect of various inactivating agents on the viral and ribonucleic acid infectivities of foot-and-mouth disease virus and on it attachment to susceptible cells. J. Gen. Microbiol. 31: 179. Brown, F. and Cartwright, B. 1960. Purification of the virus of foot-and-mouth disease by fluorcarbon treatment and its dissociation from neutralizing antibody. J. Immunol. 85: 309. Brown, F. 2001. Inactivation of viruses by aziridines. Vaccine 20: 322. Bruderer, D., S\vam, H., Haas, B., Visser, N., Brocchi, E., Grazioli, S., Esterhuysen, J., Vosloo, W., Forsyth, M., Aggarwal, N., Cox, S., Armstrong, R., and Anderson, J. Differentiating infection from vaccination in foot-and-mouth disease: evaluation of an ELISA based on recombinant 3ABC. Vet. Microbiol. In press. Capstick, P.B., Garland, A.J., Chapman, W.G., and Masters, R.C. 1965. Production of footand-mouth disease virus antigen from BHK 21 clone 13 cells grown and infected in deep suspension cultures. Nature, London 205: 1135. Capstick, P.B., Telling, R.C., Chapman, W.G., and Stewart, D.L. 1962. Growth of a cloned strain of hamster kidney cells in suspended cultures and their susceptibility to the virus of foot-and-mouth disease. Nature, London 195: 1163, Casas Olascoaga, R., Gomes, I., Rosenberg, F.J., Auge de Mello, P., Astudillo, V., and Magallanes, N. 1999. Fiebre Aftosa, Book, 1-458 Editora Atheneu, Sao Paulo, Bazil. Casas Olascoaga, R. 1978. Summary of current research of Pan American Foot-and-Mouth Disease Center on oil adjuvanted vaccines. Bull. Off. Int. Epiz. 89: 1015. Cavanagh, D., Sangar, D.V., Rowlands, D.J., and Brown, F. 1977. Immunogenic and cell attachment sites of FMDV: Further evidence for their location in a single capsid polypeptide. J. Gen. Virol. 35: 149. Cox, S.J., Barnett, P.V., Dani, P. and Salt, J.S. 1999. Emergency vaccination of sheep against foot and mouth disease: protection against disease and reduction in contact transmission. Vaccine 17: 1858.

Copyright © 2004 By Horizon Bioscience

\330

Inactivated FMD Vaccines

Cunliffe, H.R. 1973. Inactivation of foot-and-mouth disease virus with ethylenimine. Appl. Microbiol. 26: 747. Dalsgard, K. 1977. The adjuvant activity of "Quil-A" in trivalent vaccination of cattle and guinea pigs against foot-and-mouth disease. Acta Vet. Scand. 18: 367. De Castro, M.P. 1964. Behavior of foot-and-mouth disease virus in cell cultures: susceptibility of the IB-RS-2 cell line. Arq. Inst. BioI. 31: 63. De Clercq, K. 2001. Revision of the of the European Pharmacopoeia monograph for FMD vaccine and European guidelines on requirements for FMD vaccine production. Rep. Res. Gr. Eur. Comm. Contr. FMD, Island of Moen, Denmark, FAO Rome. p.174. Dekker, A. and Terpstra, C. 1996.Prevalence of foot-and-mouth disease antibodies in dairy herds in The Netherlands four years after vaccination. Res. Vet. Sci. 61: 89. Dimarchi, R., Brooke, G., Gale, C., Cracknell, V., Doel, T., and Mowat, N. 1986. Protection of cattle against foot-and-mouth disease by a synthetic peptide. Science. 232: 639. Doel, T.R. 2003. Review FMD Vaccines. Virus Research 91: 81. Doel, T.R. and Collen, T. 1984. The detection and inhibition of proteolytic enzyme activity in concentrated preparations of inactivated foot-and-mouth disease virus concentrates. J. BioI. Stand. 12: 247. Doel, T.R. and David, D.J. 1984. The stability and potency of vaccines prepared from inactivated foot-and-mouth disease virus concentrates. J. BioI. Stand. 12: 247. Doel, T.R., Fletton, B., and Staple, R.F. 1982. Further developments in the quantification of small RNA viruses by U.V. photometry of sucrose density gradients. Develop. BioI. Stand. 50: 209. Doel T.R., Williams L. and Barnett P.V. 1994. Emergency vaccination against foot-andmouth disease: The rate of development of immunity and its implications for the carrier state. Vaccine 12: 592. Dora, J.F.P., Coelho Nunes, J.C., da Silveira, J.S.G., Jorgens, H.N., Rosenberg, F.J., and Astudillo, V.M. 1984. Epidemic of foot-and-mouth disease in Bage, RS, Brazil, 1980. Evaluation of two systems of vaccination. Bol. Centro Panam. Fiebre Aftosa 50: 11. European Pharmacopoeia 3rd edition. 1997. Vaccines for veterinary use. Foot-and-mouth disease (ruminants) inactivated vaccine. Council of Europe, Strasbourg, France. Fayet, M.T., Fargeaud, D., Louisot, P., Stellmann, C., and Roumiantzeff, M. 1971. Mesure physicochimique des particles 140 S du virus de la fievre aphteuse. Ann. de l'lnstitut Pasteur 121: 107. Fontaine, J. 1985. Vaccination and prophylaxis of foot and mouth disease in France and Europe. Sci. Vet. Med. Compo 87: 47. Frenkel, H.S. 1947. La culture de virus de la fievre aphteuse sur l'(,~pithelium de la langue des bovides. Off. Int Epiz. 28: 155. Freund 1., and Thompson K.M. 1945. A simple rapid technique of preparing water-in-oil emulsions of penicillin, drugs and biologicals. Science 101: 468. Geysen, H.M., Meloen, R.H., and Barteling, S.J. 1984. Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natl. Acad. Sci. USA81: 3998. Girard, H.C., Okay, G., Kivilcim, Y., and Hogan, R. 1973. La centrifugation en continue des cellules BHK. Bull. Off. Int. Epiz. 79: 255. Graves J.H., McKercher, P.D., Farris, Jr H.E., and Cowan, K.M. 1968. Early response of cattle and swine to inactivated foot-and-mouth disease vaccine. Res. Vet. Sci. 9: 35. Henderson, W.M. 1949. The quantitative study of foot-and-mouth disease virus. ARC Rep. Sere No 8, London. Herbert W.J. 1965. Multiple emulsions. A new form of mineral-oil antigen adjuvant. Lancet 11: 771. Kaaden, O.R., Dietzschold, B., Matheka, H.D., and Tokui, T. 1971. Konzentrierung und Reinigung von Maul- und Klauenseuche (MKS)-Virus durch Polyaethylenglykol (PEG). Arch. Ges. Virusforsch. 35: 104. Copyright © 2004 By Horizon Bioscience

Barteling

331\

King, A.M.Q., Underwood, B.O., McCahon, D., Newman, J.W.I., and Brown, F. 1981. Biochemical identification of viruses causing the 1981 outbreaks of foot-and-mouth disease in the U.K. Nature. 293: 479. Leeuw, P.W. de, Tiessink, J.W.A., and Frenkel, S. 1979. Vaccination of pigs with formaldehyde inactivated aluminum hydroxide foot-and-mouth disease vaccines, potentiated with diethylaminoethyldextran (DEAE-D). Zbl. Vet. Med. B. 26: 85. Leforban, Y. 2002. Row predictable were the outbreaks of foot-and- mouth disease in Europe in 2001 and is vaccination the answer? Rev. sci. tech. Off. Int. Epiz. 21 (3): 549 MacKay DKJ, Forsyth MA, Davies PR, and Salt JS. 1998. Antibody to the non-structural proteins of foot-and-mouth disease virus in vaccinated animals exposed to infection. Vet. Quart 20 (Supplement 2): 9-1. Matheka, H.D. 1959. Ueber das Verhalten des Maul- und Klauenseuche Virus bei Adsorption an Aluminiumhydroxyd und nachfolgender Elution.Mit. 1: Untersuchung der Faktoren, die Adsorption and Elution beeinflussen. Zbl. Bakt. 1 Orig. 174: 473. McKercher, P.D. and Bachrach, H.L. 1976. A foot-and-mouth disease vaccines for swine. Can. J. Compo Med. 40: 6. McKercher, P.D. and Graves, I.H. 1977. A review of the current status of oil adjuvants in foot-and-mouth disease vaccines. Develop. BioI. Stand. 35: 107. Meloen, R.H. 1976. Localisation on foot-and-mouth disease virus (FMDV) of an antigenic deficiency induced by passage in BHK-cells. Arch. Virol. 51: 299. Michelsen, E. 1961. Experience with vaccination of pigs. Arch. Exp. Vet. Med. 15: 317. Morrow, A.W., Whittle, C.J., and Eales, W.A. 1974. A comparison of methods for the concentration of foot-and-mouth disease virus for vaccine preparation. Bull. Off. Int. Epiz. 81: 1155. Mowat, G.N. and Chapman, W.G. 1962. Growth of foot and mouth disease virus in a fibroblastic cell line derived from hamster kidneys. Nature 194: 253. Mowat, G.N., Masters, R.C., and Prince, MJ. 1973. Enhancement of immunizing potency of foot-and-mouth disease vaccine for cattle by treatment of the antigen with formaldehyde. Arch.ges.Virusforsch. 41: 365. Mowat, G.N. 1974. A comparison of the adjuvant effects of saponin and oil emulsions in foot-and-mouth disease vaccines for cattle. Bull. Off. Int. Epiz. 81: 1319. Mowat, G.N. 1972. Quantities of purified antigen required to immunze swine against footand-mouth disease. Bull. Off. Int. Epiz. 77: 887. Office International des Epizooties (OlE). 2002. International animal health code: mammals, birds and bees, 11 th Ed. OlE, Paris. (511 pp). Panina, G., and De Simone, F. 1973. Immunological activity of foot-and-mouth disease virus purified by polyethyleneglycol precipitation. Zbl. Vet. Med. B. 20: 773. Pastoret, P.P. 1996. Report on FMD Control to the European Commission, EEC Document VII7509/96. Pay, T.W.F., and Parker, M.J. 1976. Some statistical and experimental design problems in the assessment ofFMD vaccine potency. Developm. Boil. Stand. 35: 369. Pay, T.W.F. and Ringley, P.J. 1992. Foot-and -mouth disease vaccine potency test in cattle: the interrelationship of antigen dose, serum neutralizing antibody response and protection from challenge. Vaccine. 10: 699. Periolo, O.H., Seki, C. Grigera, P.R., Robiolo, B., Fernandez, G., Maradei, E., D_Aloia R., and La Torre, J.L. 1993. Large-scale use of liquid phase blocking sandwich ELISA for the evaluation of protective immunity against aphtovirus in cattle vaccinated with oiladjuvanted vaccines in Argentina. Vaccine. 11: 754. Radlett, P.J., Pay, T.W.F., and Garland, A.J.M. 1985. The use of BHK suspension cells for the commercial production of foot and mouth disease vaccines over a twenty year period. Develop. BioI. Stand. 60: 163. Radlett, P.J., Telling, R.C., Stone, C.J., and Whiteside, J.P. 1971. Improvements in the growth of BHK-21 cells in submerged culture. Appl. Microbiol. 38: 279. Copyright © 2004 By Horizon Bioscience

\332

Inactivated FMD Vaccines

Remond, M., Cruciere, C., Kaiser C., Lebreton F. and Moutou F. 1998. Preliminary results of a serological survey for residual foot and mouth disease antibodies in French cattle six years after the end of vaccination. Rep. Res. Gr. Eur. Comm. Contr. FMD, Brasov, Rumenia, FAO Rome. p. 84. Rohrer, H., and Olechnowitz, A.F. 1980. Maul- und Klauenseuche. Gustav Fisher Verlag, Jena. p. 247. Rowlands, D., Sangar, D.V., and Brown, F. 1972. Stabilizing the immunizing antigen of foot-and -mouth disease virus by fixation with formaldehyde. Arch. Ges. Virusforsch. 39: 274. Rowlands, D.J., Clarke, B.E., Carroll, A.R., Brown, F., Nicholson, B.H., Bittle, J.L., Houghten, R.A., and Lerner, R.A. 1983. Chemical basis of antigenic variation of footand-mouth disease virus. Nature. 306: 694. Rweyemamu, M.M. 1978. The selection of vaccine strains of foot-and-mouth disease. Br. Vet. J. 134: 63. Salt, J.S., Williams, L., Statham, R., and Barnett, P. V. 1995. Further Studies on the Rate of Development of Protection in Cattle Given Emergency Vaccination Against FMD. Rep. Res. Gr. Eur. Comm. Contr. FMD, Moeldling, Denmark, FAO Rome. p. 90. Salt, J. S., Dani, P., Williams, L., and Barnett, P. 1997. Efficacy Studies in Pigs with Two Novel Oil-adjuvanted Emergency' FMD Vaccines. Rep. Res. Gr. Eur. Comm. Contr. FMD, Haaretz, Israel, FAO Rome. p. 162. Salt, J. S., Barnett, P. V., Dani, P., and Williams, L. 1998. Emergency vaccination of pigs against foot-and-mouth disease: protection against disease and reduction in contact transmission. Vaccine. 16: 746. Schmidt, S. 1939. Die Bedeutung der Adsorption fur die aktive Immunisierung gegen Viruskrankheiten. Arch. Ges. Virusforsch. 1: 215. Sellers, R.F. and Herniman, K.A.1. 1974. Early protection of pigs against foot-and-mouth disease. Br. Vet. J. 130: 440. Skinner, H.H. 1951. Propagation of strains of foot-and-mouth disease virus in unweaned white mice. Proc. Royal Soc. Med. 44: 104. Snowdon, W.A. 1966. The growth of foot-and-mouth disease virus in monolayer culture of calf thyroid cells. Nature. 210: 1079. Spier, R.E. and Whiteside, J.P. 1976 The production of foot and mouth disease virus from BHK21 C13 cells grown on the surface of DEAE-Sephadex A-50 beads. Biotechnol Bioeng 19: 65. Strohmaier, K. 1967. Virus concentration by ultrafiltration. Methods in Virol III: 245. Strohmaier, K., Franze, R., and Adam, K.H. 1982. Location and characterization of the antigenic portion of FMDV immunizing protein. J. Gen. Virol. 59: 295. Sutmoller, P., BarteIing, S.1., Casas Olascuaga, R., and Sumption, K. 2003. Control of footand-mouth disease. In: Foot-and-Mouth Disease. Virus Research 91: 101 Sutmoller, P. and McVicar l.W. 1972. The epizootiological importance of foot-and-mouth carriers. III. Exposure of pigs to bovine carriers. Arch. Gesamte Virusf. 37: 78. Swam H., Davidson, F., Anemaet, D.A.1., and BarteIing, S.1. 1994. New strategies for control of foot-and-mouth disease (FMD) outbreaks in unvaccinated Europe: Use of a higly potent vaccine on infected pig farms as an alternative to "stamping-out"? Rep. Res. Gr. Eur. Comm. Contr. FMD, MoeldIing, Denmark, FAO Rome. Telling R.C. 1975. Industrial production of FMD vaccine using BHK suspension cells. Some comparative results relating in vitro assays and cattle potency. Res. Gr. Eur. Comm. Contr. FMD, Brescia, Italy, FAO, Rome. p. 95. Terpstra, C., Frenkel, S., Straver, P.1., BarteIing, S.1., and van Bekkum, J.G. 1976. Comparison of laboratory techniques for the evaluation of the antigenic potency of foot-and-mouth disease virus cultures and vaccines. Vet. Microbiol. 1: 71.

Copyright © 2004 By Horizon Bioscience

Barteling

3331

Terpstra, C., van Maanen, C., and van Bekkum, IG. 1990. Endurance of immunity against foot-and-mouth disease in cattle after three consecutive annual vaccinations. Res. in Vet. Science. 49: 236. Terre, J., Stellmann, C., Brun, A., Favre, H., and Fontaine, J. 1976. Controle d'activite du vaccin anti aphteux sur bovins: Methodes qunatitatives d'extinction antigenique. Develop. BioI. Stand. 35: 357. Ubertini, B., Nardelli, L., Dal Prato, A., Panina, G.P., and Santero, A. 1963. Large scale cultivation of foot-and-mouth disease virus on calf kidney cell monolayers in rolling bottles. ZbI. Vet. Med. B. 10: 93. Ubertini, B., Nardelli, L., Dal Prato, A., Panina, G.P., and Barei, S. 1967. BHK-21 cell cultures for the large scale production of foot-and-mouth disease. Zbl. Vet. Med. B 14: 432. Vallee, H., Carree, and Rinjard, P. 1926. Sur l'immunisation anti aphteuse par Ie virus formolee. Rev. Gen. Med. Vet. 35: 129. Van Maanen C., and Terpstra C. 1990. Quantification of intact 146S foot-and-mouth disease antigen for vaccine production by a double antibody sandwich ELISA using monoclonal antibodies. Biologicals. 18: 315-319. Van Wezel, A.L. 1985. Monolayer growth systems: homogenous unit processes. In: Anim Cell Biotechnol, Vol I. p. 265. Vianna Filho, Y.L., Astudillo, V., Gomes, L., Fernandez, G., Rozas, C.E.E., Ravison, J.A., and Alonso, A. 1993. Potency control of foot and mouth disease vaccine in cattle. Comparison of the 50% protective dose and the protection against generalization. Vaccine 11: 1424. Waldmann, O. and Trautwein, K. 1923. Die Maul- und Klauenseuche ImmuniUit nach kiinstlicher und spontaner Infektion sowie nach simultaner Impfung. Zbl Bakt I Orig. 90: 448. Waldmann, D., Kobe, K., and Pyl, G. 1937. Die aktive Immunisierung des Rindes gegen Maul- und Klausenseuche. Zbl. Bakt. I Orig. 138: 401. Wagner, G.C. and McVicar, J.W. 1970. Foot-and-mouth disease virus antibodies. Comparison of a tissue culture micro-neutralization test with the assay in suckling mice. Appl. Microbiol. 20: 995,. Wagner, G.G., Card, J.L., and Cowan, K.M. 1970. Immunochemical studies of foot-andmouth disease virus. VII. Characterization offoot-and-mouth disease virus concentrated by polyethylene glycol precipitation. Arch. ges. Virusforsch. 30: 343. Wezel, A.L. van 1967. Growth of cell-strains and primary cells on microcarriers in homogeneous culture. Nature 216: 64. Wild T.F. and Brown F. 1967. Nature of the inactivating action of trypsin on foot and mouth disease virus. 1. Gen. Virol. 1: 247. Wittmann, G. 1964. Die Interferenz zwischen inaktiviertem und aktivem Maul- und Klauenseuchevirus in Gewebskulturen. Ihre Bedeutung fUr die Unschadlichkeitsprufung von Formalinvakzinen. Zbl. Vet. Med. lIB: 135. Wittmann, G. 1972. Versuche zur Revakzinierung von Schweinen mit einer AethyHithylenimin (EEI)/DEAE-Dextran-Vakzine gegen Maul-und-Klauenseuche. Zbl. Vet. Med. B 19: 45.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 12 Foot-and-Mouth Disease Virus Peptide Vaccines David J. Rowlands

Abstract Early studies with foot-and-mouth disease virus (FMDV) showed that an important antigenic site on the particle, subsequently termed antigenic site 1, is highly sensitive to proteolytic enzymes. Only VPI is cleaved, showing that the site is present on this protein. Finally, a combination of approaches led to the location of the site within VP 1. The sequence of the site is variable both in length and composition, as might be expected for an important antigenic determinant and synthetic versions of this sequence were found to elicit reasonable levels of neutralising and protective antibodies in laboratory animals, encouraging attempts to develop synthetic vaccines. Unfortunately, for reasons that are not clear, the synthetic vaccines did not perform as well in target species. Resolution of the crystal structure of FMDV showed that antigenic site 1 lies within a highly mobile loop, the G-H loop, at the surface of the virus particle and this characteristic probably accounts for its efficacy as a synthetic immunogen. In addition, the G-H loop has been shown to be the receptor binding feature of the virus. It includes a highly conserved triplet, arginine-glycine-aspartic acid, which is key to the recognition of integrin receptors on the surface of susceptible cells. 1. History of Foot-and-Mouth Disease Vaccines Foot-and-mouth disease (FMD) has an important history in the field of virology research (Brown, 2003) The financial burden placed on the agricultural industry by FMD has long been recognised and in the late 19th century the then Prussian State offered a considerable cash prize for the discovery of the cause of the disease and followed this with funding to establish research facilities. Loeffler and Frosch (1898), demonstrated that the causative agent of the disease was not retained by filters capable of removing bacteria and so, unwittingly, founded the discipline of animal virology. FMDV was later shown to be a small RNA virus belonging to the picornavirus family. Eventually, Loeffler became director of the first' secure' FMD research institute on the island of Insel Reims in the Baltic Sea. During the early part of the 20 th century it was discovered that the serum from cattle convalescent from FMD was protective when injected into naive animals and this method was subsequently used to prevent infection of especially valuable animals. However, the availability of convalescent serum was inevitably limited and its use as a prophylactic was not a viable method for large-scale prevention of disease and could not markedly influence the incidence of infection; clearly a vaccine was necessary to affect wide-spread control of FMD. The first report of the development of a successful product was published by Vallee et ale in 1925

Copyright © 2004 By Horizon Bioscience

\336

Peptide Vaccines

and was extended by Waldman et ale in 1937. The source of the viral antigen used in this vaccine was lymph drawn from lesions produced on the tongues of deliberately infected cattle. The infectivity of the material was destroyed by treatment with formaldehyde and the immune response to the vaccine was augmented by the inclusion of aluminium hydroxide (alum) in the formulation. This vaccine proved to be effective for the protection of cattle against FMD but as with the prophylactic use of serum, the restrictions of the supply of lymph limited its value for widespread control of disease. The situation was improved markedly with the introduction of a new method for the production of antigen for vaccine manufacture (Frenkel, 1947). This involved growing the virus in culture vats on surviving epithelium stripped from the tongues of freshly slaughtered cattle. The virus grows to high titre in such tissue, which is freely available in large quantities from slaughterhouses. Again, the infectivity was inactivated with formalin and alum was used as an adjuvant. These early studies and practices established the principles of the FMD vaccines that are still in use today. Further developments which led to the modem product included the switch to the use of the baby hamster kidney cell line, BHK21, for virus culture (Mowat and Chapman, 1962). The BHK 21 cell line can be adapted to grow in suspension culture and is highly suitable for the large-scale industrial production of huge numbers of vaccine doses in a very reproducible fashion (Capstick et ai., 1962). Another advance was the change from the use of formaldehyde to aziridine compounds to inactivate the virus infectivity (Brown and Crick, 1959; Bahnemann, 1975). These compounds inactivate the virus with first order kinetics and can be guaranteed to eliminate all infectivity. This is not quite true for inactivation by formaldehyde and residual traces of live virus in vaccines have been responsible in the past for the seeding of outbreaks of the disease (King et ai., 1981). Finally, throughout the history of FMD vaccine development it became apparent that the vaccine could fail dramatically at times. It was soon realised that this is because FMDV is not a single species of virus but exists in a number of radically different forms which are antigenically unrelated (Brooksby, 1958). It is now known, of course, that there seven serotypes of the virus, each of which provides no crossimmunity against any of the other six. The situation is further complicated by the significant antigenic diversity that can occur even within serotypes. For this reason there is close monitoring of the precise cocktails of virus strains to be incorporated in vaccines tailored for use in different geographical locations (see Doel, 2003). The alternative approach to vaccine development of deriving attenuated strains of the virus to be used as live vaccines has never worked well for FMD. This is in contrast to the situation with many other viruses, including another picornavirus, poliovirus, for which very successful live attenuated vaccines have been available since the 1950s. The great antigenic variability of FMDV is one handicap to the development of live vaccines but the inherent genetic variability and high virulence properties of the virus probably also contributed to the failure of the attenuation approach. All FMD vaccines in use today are produced by growing potentially virulent virus to high titres in very large volumes prior to inactivation. Such a procedure demands highly efficient infection containment facilities and inevitably carries some risk of accidental environmental contamination. As the balance swings Copyright © 2004 By Horizon Bioscience

Rowlands

3371

towards effective control of the disease so the potential risks associated with the growth and management of large amounts of live virus take on an ever-greater significance. In such a situation the development of alternative strategies for vaccine production that do not rely on the growth of the virus itself become very attractive. Synthetic peptide vaccines provide one potential route for the development of novel and safer vaccines. However, despite early promise a number of obstacles arose which complicated their progress towards become practical vaccines and further research into their development slowed dramatically. This relative decline in interest in the continuing development of synthetic vaccines when the problems became more difficult may prove to have been premature given the revival in interest in FMD control following the disastrous outbreaks in the UK in 2001. Indeed, novel approaches to peptide vaccine development that owe their origins to the early studies on the virus and its antigenic features are now showing distinct promise. Whether they may yet be the basis for new generations of vaccines or not, synthetic peptides have been extremely useful tools in the dissection of FMDV structure and function. 2. The Antigenic Structure of FMDV - Early Studies During the 1950s biophysical studies combined with serological analyses provided evidence for the presence of different antigenic particles in crude preparations of FMDV (see Brown, 1995; Rowlands, 1993). The virus particle itself was shown to have a sedimentation coefficient of 1408 but there are also large amounts of a smaller particle which sediments at approximately 125. A similar particle was found to be the major degradation product of the 140S virus particle following heat or acid mediated disruption and is now known to be a pentameric subunit of the virus. Although the 12S subunit and the 140S virus particle share some antigenic features there are also important differences (Brown, 1995). In particular, the 12S subunit particles were found to be ineffective at inducing significant levels of virus neutralising and protective antibody responses in experimental animals. Moreover, excess amounts of 12S subunit preparations are incapable of absorbing the virus neutralising activity of sera from convalescent animals (Rowlands et al., 1971). These observations provided evidence for the importance of correct conformation for the effective presentation of antigenic features of the virus essential for the efficient induction of protective immune responses. During the 1960s and 1970s technological advances in the methods available for protein analysis permitted more detailed study of the antigenic structure of FMDV. A particularly important finding was that the antigenic properties of the virus were altered dramatically by treatment with proteolytic enzymes, especially trypsin, without obvious gross alteration of the structure of the virus particle (Brown and Smale, 1970; Wild and Brown, 1967). Although trypsin treated particles are indistinguishable from untreated virus in the electron microscope and by sedimentation analysis, their ability to induce virus neutralising antibody responses is greatly compromised. In addition, the trypsin treated particles were shown to have lost the ability to adsorb significant amounts of virus neutralising antibodies from convalescent sera (Rowlands et al., 1971). Together, these data strongly suggested that trypsin treatment results in the destruction of an antigenic site (or sites) on the virus that is of major importance for the induction of protective Copyright © 2004 By Horizon Bioscience

\338 Peptide Vaccines

Figure 1. Electron micrographs of; a) trypsin treated or, b) native FMDV particles reacted with IgM antibody and visualised following negative staining (from Brown and Smale, 1970).

immunity. In support of these conclusions, electron microscope studies showed that virus neutralising IgM antibodies attached at a limited number of sites on the surface of the virus particle and that these antibodies could no longer bind to particles that had been digested with trypsin (Brown and Smale, 1970) (Figure 1). Meanwhile, advances in protein separation techniques, such as polyacrylamide gel electrophoresis, allowed the structural proteins of the virus to be resolved easily. Analysis of trypsin treated virus compared to untreated virus showed that only one of the four virus proteins, VP1, is altered by the digestion and is cleaved into two smaller fragments (Burroughs at al., 1971). It was concluded from these observations that the most important antigenic determinant of the virus is located on VPl and that the site is susceptible to cleavage by trypsin and other proteolytic enzymes. In support of this it was demonstrated that purified VPl is capable, albeit very inefficiently, of inducing protective immune responses in experimental animals (Laporte et al., 1973; Bachrach et al., 1975; Meloen et al., 1979). With the advent of molecular cloning and recombinant protein expression techniques in the early 1980s attempts were made to produce a novel vaccine for FMD based on the VPl protein (Kleid et al., 1981). Unfortunately, despite the formal demonstration of the efficacy ofVPl as a protective antigen in experimental animals, its low potency compared to the traditional inactivated virus vaccines prevented it from being developed as a practical product. Another consequence of the cloning revolution was the increasing availability of nucleic acid sequences. As the number of FMDV sequences increased so it became apparent that, in addition to the general high degree of sequence variation seen between different virus strains and serotypes of the virus over the entire genome, there are regions of hyper variation within the structural protein coding sequences. In particular, it was noticed that there are two areas of VPl that are particularly variable (Makoff et aI., 1982; Beck et aI., 1983). As antigenic variation Copyright © 2004 By Horizon Bioscience

Rowlands

N' - - - - - - - - - - - - - - - - - C'

N'

@if!@MW¥i'iMS

M M M ~--------- - - C' KIR KIR ~ C'

R R N' DEem?iilaD~~~mmmm.l!~~~emW**i_ m _ - - - - - - C'

- -

3391

Intact Protein

Cyanogen Bromide Fragments Trypsin

Sub-maxillary Gland Protease

PREDICTED ANTIGENIC SITES _

IMMUNOGENIC FRAGMENTS

~

NON-IMMUNOGENIC FRAGMENTS

Figure 2. Diagrammatic representation of the mapping of immunogenic peptide sequences on the VPl protein of FMDV. The bars represent the entire protein or peptide fragments purified from protein cleaved with cyanogen bromide, trypsin or submaxillary gland protease. Fragments capable of eliciting neutralising antibody responses are shown in black, inactive fragmants are shown in grey (adapted from Strohmaier et ale 1982).

is a hallmark of FMDV, it was speculated that the highly variable regions of VPl equate with the sites of particular antigenic significance. The location of the antigenic determinants on VPl was resolved by elegant experiments conducted by Strohmair and colleagues (Strohmair et al., 1982). They produced a series of fragments of VPl by digestion with different proteases or by chemical cleavage with cyanogen bromide and tested the immunogenicity of these fragments after purifying them by polyacrylamide gel electrophoresis. The identity ofthe fragments was determined by limited terminal amino acid sequencing and matching the results with the nucleic acid sequence coding for the entire protein. The purified fragments were used to immunise mice, after which their sera were tested for the presence of virus neutralising activity. By comparing immunogenicities of the defined fragments it was possible to predict the location of the important antigenic sites (Figure 2). These predictions fitted well with the earlier observations of the properties of trypsin treated virus and with the emerging data on the locations of highly variable regions derived from sequence comparisons. 3. Synthetic Peptides Induce Protection Against FMD There were significant advances in the chemical synthesis of peptides in the 1970s and the resulting supply of large amounts of pure and specific peptides encouraged many immunologists and microbiologists to explore the immunogenic properties of these newly available molecules. The evidence reviewed above served to define the precise regions of the VPl protein of FMDV that most likely represent important antigenic sites on the virus and are responsible for the induction of protective, virus neutralising antibodies. Consequently, synthetic peptides corresponding

Copyright © 2004 By Horizon Bioscience

\340

Peptide Vaccines

to these regions, residues 140-160 and 200-213 at the C terminus of the protein, were coupled to 'carrier' proteins, such as keyhole limpet haemocyanin, to enhance their immunogenicity and used to immunise laboratory animals. A widely used experimental animal for FMD studies is the guinea pig, which combines the features of convenient size and susceptibility to infection by the virus, and most early investigations into synthetic peptide vaccines involved this species. Encouragingly, early results showed that the synthetic peptide preparations were able to elicit good levels of FMDV neutralising antibodies, even more excitingly, the animals were protected from severe challenge with virulent virus (Bittle et al., 1982; Pfaff et al., 1982). Although peptides representing the amino acid sequences of the 140-160 and 200-213 regions of VPl were both shown to induce virus reactive antibodies, the 140-160 peptide was markedly superior and appeared to represent the most important antigenic site. Consequently, further work to explore the potential of synthetic peptides as practical alternatives to conventional vaccines and to understand the basis for protection afforded by them concentrated on the VPl 140-160 sequence. Attempts were also made to combine the properties of the two immunogenic regions of VP1 by synthesising peptides in which the 140-160 and 200-213 sequences were linked by a spacer sequence in a continuous peptide. Experiments in cattle showed some efficacy following vaccination with this product but it was significantly less effective than conventional inactivated virus vaccine (DiMarchi et ai., 1986). Early work with conventional FMD vaccines had shown that there was a broad correlation between the virus neutralising activity induced in immunised animals and the level of immunity afforded (Martin and Chapman, 1961). Moreover, immunity was found to be relatively short lived in vaccinated animals and loss of protection could be related to waning of the virus neutralising capacity of the serum. As FMD is usually highly virulent and develops at a very rapid rate it might be rationalised that the delay involved in eliciting an anamnestic response in vaccinated animals is sufficient for the virus to establish a serious infection before it is controlled by the immune response. Thus protection is dependent upon the continued maintenance of high levels of neutralising antibodies in the serum. In view of these considerations it was of interest to compare the relationships between the neutralising antibody levels induced and the degree of protection afforded by inactivated virus vaccine or by synthetic peptide vaccine. Using the guinea pig model it was shown that the correlation between the neutralisation titres of vaccinated animal sera and protection of the animals against live virus challenge was even clearer in the peptide-vaccinated animals than for the conventionally vaccinated animals (Francis et ai., 1988). All guinea pigs that had been immunised with the synthetic peptide vaccine and had responded to produce more than a minimal virus neutralising antibody titre were found to be resistant to virus challenge. With conventionally vaccinated animals on the other hand there was a looser correlation between the neutralising antibody titre elicited and the ability to resist challenge; some animals with relatively high neutralisation titres were susceptible to infection while others with low titres of antibody proved resistant to challenge. These results suggest that; 1) protection afforded by the conventional vaccine is influenced by factors in addition to the virus neutralising activity present Copyright © 2004 By Horizon Bioscience

Rowlands

3411

in the vaccinate serum, as measured by in vitro tests and, 2) the peptide vaccine elicits a spectrum of neutralising antibodies that correlated very well with the immune status of the animals. Antigenic variation is also an important feature of FMDV and is a constant source of concern for vaccine manufacturers. The relationships between field isolate strains of FMDV and vaccine strains are usually expressed as reciprocal r values (Doel, 2003). The r value is the ratio of neutralisation titre of serum from a vaccinated animal against the homologous vaccine virus strain and the field isolate; a value of 1.0 indicates antigenic identity between the viruses and values less than unity indicate antigenic divergence. When the ratio is less than 0.4 the vaccine strain may be considered unsuitable for immunisation against the outbreak strain (Ferris and Donaldson, 1992). Surprisingly, when the neutralisation titres of antisera raised against a peptide representing the VP 1 140-160 sequence of FMDV, serotype 01 were measured against a range of field isolates of serotype 0 isolates it was found that the r values were consistently high (Parry et ai., 1989). In fact, by this in vitro test the peptide antisera performed better that antisera raised against the relevant serotype 0 conventional whole virus vaccine. In related studies neutralisation resistant viruses could not be readily selected in vitro. This contrasted strikingly with the ease with which neutralisation resistant viruses could be selected using monoclonal antibodies. Clearly the antibodies induced by the peptide antigen were sufficiently diverse to overcome the problem of rapidly evolving resistance by the selection of antigenic variants of the virus. Together, the results of these early investigations into the immunogenic properties of synthetic peptides were highly encouraging and suggested that many of their properties were precisely those required in a commercial vaccine. However, there are two important caveats to these results. Firstly, the selection experiments were done in vitro, which may not give a true reflection of the factors involved in selection in vivo. Secondly, the results were obtained using the convenient animal model system of the guinea pig and thus may not be representative of the situation in the target species for vaccination against FMD such as cattle and pigs. Most of the early studies with synthetic peptide vaccines against FMD were conducted with sequences from serotype 01 viruses. However, it was shown subsequently that the ability to induce virus neutralising antibodies using peptides representing the amino acid 140-160 region of VP 1 was not restricted to these viruses. Indeed, equivalent peptides from examples of viruses from all seven serotypes of the virus were equally effective (Francis et al., 1990). An interesting exception to this general observation was provided by experiments with serotype A12 virus. In contrast to the experience with other peptide sequences, that from the A 12 virus did not raise significant levels of neutralising antibody when used to immunise guinea pigs (Rowlands et al., 1983). The answer to this conundrum was provided when the virus used to test the neutralising activity of the sera was sequenced. At two positions within the amino acid 140-160 region of VPl the correct amino acids could not be identified due to ambiguities in the nucleotide sequence. When the virus was plaque purified and individual clones sequenced it could be seen that the virus stock was a mixture of viruses with different residues at 148 and 153, none of which corresponded to the published sequence derived from a molecular clone and from which the peptide was designed. When peptides were Copyright © 2004 By Horizon Bioscience

\342 Peptide Vaccines

made according to the individual identified sequences and used as immunogens the sera were found to neutralise homologous virus clones effectively. Moreover, the patterns of cross reactivity of the peptide antisera with the cloned viruses were qualitatively similar to those seen with vaccines made from whole inactivated viruses. These observations provided further evidence for the relative importance of the VPl 140-160 sequence in the antigenic and immunogenic structure of the whole virus and suggested the possibility of overcoming antigenic variation by using cocktails of peptides with alternative residues at defined positions in the sequence. 4. Structure of FMDV-Explanation for the Immunogenicity of the VP1 G-H Loop The early success of the synthetic approach to vaccine development with FMDV appeared to be a special feature of that system and attempts to produce peptide immunogens to a range of other pathogens were generally less successful. It was clear that knowledge of the structure of the virus at high resolution might provide important clues to understanding the relative superiority of the experimental FMDV vaccines. The structure ofFMDV serotype 01 was solved to near atomic resolution by X ray crystallographic methods in 1989 and the result \vas both illuminating and in some aspects rather disappointing (Acharya et aI., 1989). The virus particle \vas found to be constructed in a similar manner to the other picornaviruses, poliovirus and human rhinovirus, which had been solved in 1985 (Hogle, et aI. 1985; Rossmann, et al. 1985) However, there are a number of important differences between the structures of FMDV and the other viruses, the most obvious of these involving the region of the VPl protein that had shown such promise as a synthetic immunogen. In the crystal structure of FMDV serotype 01 the region between VP1 residues 137 and 156, which are located between the G and H f3 strands of VP1 and are referred to as the G-H loop, was too highly disordered to be visible on the electron density maps (Figure 3a). The interpretation of this observation is that this portion of VPl forms a flexible domain which does not occupy a defined orientation with respect to the remainder of the virus particle. Thus a plausible explanation for the effectiveness of the VP1 G-H loop as a peptide immunogen is that the sequence normally resembles an independent peptide sequence tethered flexibly to the virus surface. Figure 3. Structure of the FMDV particle and the VPl G-H loop. a) FMDV serotype O. The proteins are drawn in 'worm' representation with VPl in light grey, VP2 in mid-grey and VP3 in darker grey. The VPl G-H loop is represented by a thicker worm in black. The G-H loop appears

to behave as a hinged domain pivoting about its base (defined by a disulphide interaction in type It is not visible from crystals of the native virus and hence a transparent sphere is shown centred at its base to represent the range of flexibility of the loop. b) A crystallographic protomer of the virus with proteins shaded as above. The electron density visualised for the G-H loop in the reduced virus is shown superimposed. This corresponds to the "down" position of the loop which does not appear to be able to bind to integrin receptor. The key receptor-binding residues (RGD) are shown in black ball-and-stick (adapted from Logan et at. 1993). c) A side view of the crystallographic protomer based on cryo-EM analysis of serotype C virus binding to a monoclonal antibody FAB fragment, shaded similarly to a) and b). The loop is modelled (based on the coordinates in (b» in the putative "up" conformation, as indicated by the reconstruction (adapted from Hewat et al. 1997). Figures courtesy of E. Fry.

o viruses).

Copyright © 2004 By Horizon Bioscience

Rowlands 343 I

a

b

c

Copyright © 2004 By Horizon Bioscience

\344 Peptide Vaccines

The resolution of the structure of the virus provided an attractive rationale for the ability of synthetic versions of the G-H loop to elicit the production of antibodies that effectively cross react with the virus particle. Unfortunately, the structure of the native virus gave no indication of the conformation of the G-H loop itself. However, it could be seen from the resolved parts of the structure that a disulphide bond linking a cysteine residue at the base of the G-H loop of VPl (amino acid 134) with a cysteine in VP2 (amino acid 130) had multiple conformations and appeared to hold the loop away from the virus surface. This was confirmed when the crystal structure of virus in which the disulphide bond had been reduced with dithiothreitol was solved (Logan et ai., 1993). In this structure the VP1 G-H loop was collapsed onto the surface of the virus particle and its detailed organisation could be clearly resolved. From these studies it could be concluded that, although the VPIG-H loop is disordered with respect to the virus surface, it is itself highly ordered and has a well-defined structure (Figure 3b and 3c). This interpretation is further supported by the crystal structure of complexes between monoclonal antibodies and synthetic peptides representing the G-H loop sequence (Domingo et ai., 1999; Hewat et ai., 1997; Ochoa et ai., 2000; Verdaguer et ai., 1995). The organisation of the peptide in these complexes is similar to that seen in the reduced virus structure. The combination of the two cysteine residues that appear to tether the VP1 G-H loop of type 01 viruses in an exposed and 'disordered' configuration is not present in other serotypes of the virus. However, when the structures of more viruses became available it was found that the VPl G-H loops of these were also disordered with respect to the virus surface (Lea et ai., 1994; Curry et ai, 1996). As possession of a mobile VPl G-H loop is conserved in all the viruses that have been analysed to date it would appear that this property is of functional importance to the virus. There is some direct evidence for this since reduction of the disulphide bonds in a serotype 01 virus causes a reversible reduction in infectivity but has no such effect on a serotype C virus (King, 2003) Although not as effective as the G-H loop, synthetic peptides representing the C terminal region of the VP1 protein had also been shown to elicit virus reactive and neutralising antibodies. This sequence was also found to be located at a highly exposed position on the particle surface and to be less ordered than other parts of the virus (Acharya et ai., 1989). Both the G-H loop and the C terminus ofVPl ofFMDV have the characteristics of 'linear' epitopes. That is, they are 'self contained' antigenic entities whose

internal structures are to some extent divorced from the surrounding protein structure. This is in contrast to 'conformational' epitopes, which are regions at the surface of proteins that are recognised as antigenic entities but include portions of sequence from adjacent tertiary or quaternary features of the protein structure. Clearly, on this definition linear epitopes are more easily mimicked by synthetic peptides than conformational epitopes and this helps to explain the relative success of FMDV synthetic peptide vaccines. Unfortunately for the progress of peptide vaccines the great majority of antigenic sites on proteins fall into the 'conformational' category and are significantly more difficult to represent as synthetic constructs. However, methods for assembling peptides designed to represent the features of conformational antigenic sites are being developed (Villen et ai., 2001). Copyright © 2004 By Horizon Bioscience

Rowlands

345 I

5. Structure and Function of the VP1 G-H Loop It is clear that the VPl G-H loop of FMDV is a rather unusual structural feature at the surface of the virus and that its peculiar characteristics may explain why it is effective when presented in isolation as a synthetic immunogen. Several studies have confirmed that it is an immunodominant feature of the virus and is recognised by a high proportion of the neutralising antibodies present in immune sera (Mateu et ai., 1995). It is not, of course, the only antigenic site on the virus surface. Detailed analyses using panels of neutralising monoclonal antibodies have identified a total of 4 or 5 antigenic sites which are identified numerically or alphabetically depending on the virus serotype and laboratory in which the study was performed (Crowther et al., 1993; Kitson et al., 1990; McCahon et ai., 1989). The VPl G-H loop is referred to as site 1 or A. The other antigenic sites are conformational and are formed from the juxtaposition of residues from different parts of the primary structure and from different proteins. These sites are inherently much more difficult to mimic with synthetic peptides but, as mentioned above, new approaches in peptide design are beginning to address this problem (Villen et ai., 2001). The structure of the VPl G-H loop is shown in (Figure 3b). It forms a long hairpin with a highly conserved triplet, Arginine-Glycine-Aspartic acid (RGD) at the apex of the loop. This sequence is the hallmark motif of ligands that bind to members of a family of cell surface receptor molecules termed integrins and several studies have highlighted the importance of the motif for the attachment of FMDV to susceptible cells (Baxt and Becker, 1990; Berinstein et ai., 1995; Fox et ai., 1989; Mason et ai., 1994). Integrins are type 1 membrane proteins built from a and ~ chains and are involved in a variety of cell signalling and cell attachment functions. There are several a and ~ chains and their relative association provides a wide range of mature integrin molecules that are differentially expressed in a tissue specific fashion. The hypothesis that FMDV uses integrins as receptor molecules via binding of the VPl G-H loop is supported by several lines of evidence; 1) proteolytic cleavage within the loop destroys infectivity of the virus and prevents it from attaching to cells (Rowlands et ai., 1971), 2) the conservation of the RGD motif and, 3) the ability of RGD containing peptides to compete for virus binding to cells (Fox et al., 1989). Furthermore, the crystal structure of reduced virus sho\ved that the RGD was in an 'open' conformation, as had previously been seen in known integrin binding proteins (Logan et ai., 1993). Subsequent work has confirmed that integrins are the receptors normally used by FMDV, although it is possible to select in vitro for viruses that can utilise other receptors, including heparin sulphate (Jackson et al., 1996; Baranowski et ai., 2000; Sa-Carvalho et al., 1997). There has been considerable debate as to the specific integrin(s) that constitute the natural receptor for the virus in normal host species and susceptible tissues and the aV~6 molecule is currently the favoured contender (Jackson et al., 2000). Although there is considerable variation in the size and sequence of the VPl G-H loop between viruses of different serotypes and strains, the locations of changes are somewhat restricted and the overall structure of the loop appears to be conserved (Logan et ai., 1993). The RGD triplet was first recognised as a potential integrin binding feature but there are further conserved motifs within the VPl G-H loop, such as an LXXL sequence following the RGD, that are most likely involved in defining the integrin recognition specificity (Mateu, et ale 1996). Copyright © 2004 By Horizon Bioscience

1346 Peptide Vaccines

The VP1 G-H loop is clearly subjected to apparently conflicting evolutionary pressures. Its role as receptor binding feature is vital, not only to allow the virus simply to attach to cells but also to determine the integrin binding preferences of the virus and so its tissue tropism. On the other hand, it is a prime target for immune response and is, therefore subject to pressure for change. Detailed studies of the binding of antibodies to synthetic peptides together with structural studies of immune complexes are shedding light on the ways in which the sequence accommodates to these pressures. The crystal structure of VP1 G-H loop peptides bound to monoclonal antibody Fab fragments confirm that the structure as seen in reduced virus particles is maintained in these complexes, sometimes through distortion of the antigen binding pocket. As might be expected, many substitutions in the loop sequence abrogate binding to antibody, however, in some cases binding to antibody is maintained despite substitution in the sequence (Verdaguer et ale 1998; Ochoa et al., 2000). Attempts have been made to improve the structural mimicry of synthetic VP1 G-H loop peptides by the introduction of disulphide bonds to restrict and define tertiary structure. Although simple linking of the termini of the peptide sequence did not improve binding with monoclonal antibodies, the introduction of links within the sequence significantly improved recognition (Valero et ale 2000). In summary, there is a wealth of evidence that FMDV binds to integrins on susceptible cells to initiate the infectious process and that the VP1 G-H loop is crucial for this function. It would appear from this evidence to be a prime site on the virus surface for recognition by virus neutralising antibodies which function by inhibiting virus attachment. In fact, site All reactive monoclonal antibodies and anti VP1 G-H loop peptide polyclonal antibodies have been shown to block virus attachment to cells in culture. Whether it is necessary to block all of the VP1 G-H loops on a virus particle to abrogate attachment and infection and whether immune mechanisms in addition to the blocking of receptor binding operate to affect protection in vivo are not clear. 6. Peptide Vaccines-Improvements in Antigen Presentation From immunogenicity studies, the antigenic structures of peptides constructed from VP1 G-H loop sequences appear to resemble the native antigenic site on the virus particle since a high proportion of the antibodies induced by synthetic peptides also recognise the virus (Parry et al., 1988) and the peptides can be used to elicit protective antibody responses. Physical studies involving circular diochroism and nuclear magnetic resonance measurements have shown that the extent to which synthetic versions of the VP I G-H loop resemble the helical structures seen in the reduced virus structure depends on the nature of the solvent and the length of the peptide (Siligardi et al., 1991 a; 1991 b; de Prat-Gay, 1997). Irrespective of their ability to adopt defined conformations, short synthetic peptides are inherently poorly immunogenic. There are a number of reasons for this including their small size and rapid clearance from the body, the possible lack of strong and appropriate helper T cell epitopes, a low proportion of the peptide adopting a virus-like conformation and the monomeric presentation of the antigenic site. A number of approaches have been adopted to improve the immunogenicity of short peptide antigens which address some or all of these factors. The simplest way to increase the size, copy number per antigenic unit and to provide multiple helper Copyright © 2004 By Horizon Bioscience

Rowlands

3471

T cell epitopes is to chemically couple the synthetic product to a large carrier protein such as keyhole limpet haemocyanin. Although this method can improve immune responses to peptides, it suffers from the disadvantage of being difficult to control - the peptides are essentially randomly linked to the carrier protein. It also depends on the foreign protein providing most or all of the T cell help driving the humoral response rather than helper T cell epitopes derived from the virus itself. Thus many of the potentially attractive features of developing synthetic vaccines such as their defined and reproducible structure are lost when they are coupled to carrier proteins. More sophisticated approaches to address the immunogenicity problem have also been explored. These include the construction of larger peptides including extra T and/or B cell epitopes such as the combined VP1 G-H loop and C terminus peptides used in the first cattle immunisation experiments (DiMarchi et al., 1986). An interesting method for increasing the immunogenic mass of synthetic peptides is the multiple antigen presentation system (Tam, 1988). This involves the construction of a poly-lysine core structure using a synthetic strategy in which lysines are added in sequential rounds of synthesis to both the a and E amino groups of the growing polymer. In this way the number of amino groups available for building peptide chains is doubled at each addition. When the lysine'core' has reached a suitable size (eg. 8 or 16 available amino groups) the immunogenic peptide sequences are then built simultaneously onto each group to produce a large multivalent complex. In another approach the adjuvant effect provided by synthetic oils has been built into the peptide by derivatising it to produce a di-palmitoylated form (Volpina et ale 1996) The problem of rapid biodegradation of synthetic peptide immunogens has been addressed by constructing peptides from the unnatural D amino acids. Such peptides are resistant to hydrolysis by proteases and peptidases and so are predicted to survive for longer in the body. Of course, the conformations of peptides built from the D isomers of their component amino acids will resemble the mirror image of the native peptides comprising L amino acids. To overcome this problem the D amino peptides are assembled in the reverse order from the natural sequence. These so called 'retro-inverso' peptides adopt conformations resembling the natural antigen and are capable of inducing relevant immune responses (Van Regenmortal et al., 1998). An alternative approach to improving immunogenicity is to incorporate the peptide epitope sequence into an easily expressed protein using genetic fusion techniques. A number of 'partner' molecules have been used to produce fusion proteins linked to the FMDV VP1 G-H loop sequence and include (3 galactosidase (Broekhuijsen et al., 1987), bacterial outer membrane proteins Omp A (Ruppert et al., 1994) and Pho E (Agterberg et ai., 1990) and a viral glycoprotein, infectious bovine rhinotracheitis virus gIll (Kit et ai., 1991). Phages have also been constructed that express the sequence fused to structural proteins. Probably the most adaptable and effective fusion protein systems involve the use of viral proteins that have the inherent property of assembling spontaneously into viruslike particles (VLPs). These are usually highly immunogenic as they combine the features of appropriate molecular size and multiple epitope presentation. They often also contain multiple helper T cell epitopes, which further help to enhance Copyright © 2004 By Horizon Bioscience

\348

Peptide Vaccines

their immunogenic potential. A further benefit to be derived from the construction of VLPs expressing peptide epitopes is that the proximity of the structured fusion protein frequently induces a more 'native' conformation in the introduced peptide sequence. The core protein of hepatitis B virus has excellent properties as a fusion partner to produce recombinant VLPs and early experiments showed that the immunogenicity of the VPl G-H loop presented on such particles is comparable to that of FMDV itself (Clarke et al., 1987). 7. Comparative Efficacies of Peptide Vaccines in Different Species Much of the early work on synthetic peptides vaccines used the guinea pig model for practical reasons of convenience and economy. This species was found to respond well to the VPl G-H loop peptide and animals could be reproducibly protected from severe virus challenge by immunisation with relatively modest amounts of peptide. Results in agriculturally relevant species, cattle and pigs, have been rather less impressive. Although protection has been demonstrated in both of these species this has only been in a relatively modest proportion of the animals, especially in cattle. Pigs seem to fall between cattle and guinea pigs in the ease with which they can be protected by synthetic vaccines. A problem in the interpretation of these apparent species differences is that the parameters of ill vivo protection are not fully understood. Clearly, humoral antibody levels are of major importance as was demonstrated in the early 1900s by the passive protection of cattle using sera from convalescent animals. Indeed, protection following immunisation with conventional whole virus vaccine broadly correlates with the virus neutralising antibody levels present in the serum (McCullough et al. 1992). With peptide vaccines in the guinea pig model this correlation is very clear but becomes less so with cattle and pigs. A number of immunological factors may account for the relatively poor ability of peptide vaccines to protect cattle and pigs including the spectrum of immunoglobulin subclasses they induce, the affinity of anti-peptide antibodies for the virus and the absolute level of virus neutralising antibody induced. Virological factors are also likely to be important. For example, as we saw above, the serotype A12 strain was found to be a mixture of variant virus with minor changes in the VPl G-H loop sequence and so could readily escape neutralisation by antibodies directed against a single peptide sequence. With serotype 0 viruses, on the other hand, neutralisation resistant viruses could not be selected using anti-peptide antisera and these were found to be able to neutralise a wide range of field strains. In the largest field trial of peptide vaccine that has been conducted so far, using a serotype A virus sequence, there was found to be rapid selection and evolution of variant virus resistant to neutralisation by sera raised against the immunising peptide (Toboga et al. 1997). Factors that may account for this are the induction of inappropriate responses by the vaccine and/or too low a response (only a proportion of the animals achieved protected status). A better understanding of the in vivo mechanisms of protection against FMDV is important for the future development of alternative approaches to vaccination.

Copyright © 2004 By Horizon Bioscience

Rowlands

3491

8. Conclusion: Future Prospects for Peptide Vaccines for FMD Twenty years ago the demonstration that animals could be protected against a highly virulent infection such as FMD with a chemically synthesised vaccine was surprising and exciting. The FMD system provided the most encouraging results for the future prospects of synthetic vaccine technology and the resolution of the structure of the virus provided a rational explanation for the efficacy of the FMDV peptide immunogens. In the intervening years a great deal has been learned about the function of the VP 1 G-H loop, the principle target sequence for the development of synthetic vaccines, both in its role as the receptor binding ligand of the virus and as an antigenic determinant. However, we still do not fully understand either of these aspects of the sequence. For example, its roles as receptor binding feature and antigenic site are both subtly modulated by its interactions with other features of the virus particle surface. The VPl G-H loop sequence is highly variable both in length and sequence but there are functional constraints placed on this variation and a fuller understanding of these could be valuable in the design of peptide mixtures designed to overcome the selection of neutralisation escape variants. Many different approaches for improving the immunogenicity of peptide vaccines have been tried but the great majority of these have involved small scale laboratory experiments and few have been studied in more relevant situations. Improvements in both the rate as well as the absolute of response of the immune response are crucial to the potential successful future of synthetics vaccines as they will have to compete with an effective and relatively well understood conventional vaccine product. In summary, although synthetic vaccines against FMD have not realised their early potential, they have provided the means to study fundamental properties of FMDV. These, in tum, suggest ways in which the peptide approach may be improved and developed. The recent spread of FMD into regions of the world from which it had been eliminated has reawakened debate about the use of vaccines against the disease. This is likely to stimulate renewed interest in the development of synthetic vaccines since the production of vaccines that do not require the handling of large amounts of live virus may be an attractive possibility for many parts of the world. References Acharya, R., Fry, E., Stuart, D., Fox, G., Rowlands, D., and Brown, F. 1989. The 3dimensional structure of foot-and-mouth-disease virus at 2.9-a resolution. Nature 337: 709-716. Agterberg, M., Adriaanse, H., Lankhof, H., Meloen, R. and Tommassen, J. 1990 Outer membrane PhoE protein of Escherichia coli as a carrier for foreign antigenic determinants: immunogenicity of epitopes of foot-and-mouth disease virus. Vaccine 8: 85-91. Bachrach, H.L., Moore D.M., McKercher, P.D., and Polatnick, J. 1975. Immune and antibody responses to an isolated capsid protein, of foot and mouth disease virus. J. Immunol. 115: 1636-1641. Bahnemann, H.G. 1975. Binary ethyleneimine as an inactivant for foot-and-mouth disease virus and its application for vaccine production. Arch. Virol. 47: 47-56. Baranowski, E., Ruiz-Jarabo, C.M., and Domingo, E. 2001. Evolution of cell recognition by viruses. Science 292: 1102-1105. Copyright © 2004 By Horizon Bioscience

1350

Peptide Vaccines

Baranowski, E., Ruiz-Jarabo, C.M., Sevilla, N., Andreu, D., Beck, E., and Domingo, E. 2000. Cell recognition by foot-and-mouth disease virus that lacks the RGD integrinbinding motif: Flexibility in aphthovirus receptor usage. J. Viral. 74: 1641-1647. Baxt, B. and Becker, Y. 1990. The effect of peptides containing the arginine-glycine-aspartic acid sequence on the adsorbtion of foot-and-mouth disease virus to tissue culture cells. Virus Genes 4: 73-83. Baxt, B., Vakharia, V., Moore, D.M., Franke, A.J, and Morgan, D.O. 1989. Analysis of neutralizing antigenic sites on the surface of type A12 foot-and-mouth disease virus. J. Virol. 63: 2143-2151. Beck, E., Feil, G., and Strohmaier K. 1983. The molecular basis of the antigenic variation of foot-and-mouth disease virus. EMBO J. 2: 555-559. Berinstein, A., Roivainen, M., Hovi, T.,Mason, P. and Baxt, B.. 1995. Antibodies to the vitronectin receptor (intergrin avb3) inhibit binding and infection of foot and mouth disease virus to cultured cells. 1 Virol. 69: 2664-2666. Bittle, J. L., Houghten, R.A., Alexander, H., Shinnick, T.M., Sutcliffe, lG., Lerner, R.A., Rowlands, D.J. and Brown, F. 1982. Protection against foot and mouth disease by immunization with a chemically synthesized peptide predicted from the viral nucleotide sequence. Nature 298: 30-33. Broekhuijsen, M.P., van Rijn, J.M., Blom, A.J., Pouwels, P.H., Enger-Valk, B.E., Brown, F. and Francis, M.J. 1987 Fusion proteins with multiple copies of the major antigenic determinant offoot-and-mouth disease virus protect both the natural host and laboratory animals. J. Gen. Virol. 68: 3137-3143. Brooksby, lB. 1958 The virus of foot-and-mouth disease. Adv. Virus Res. 5: 1-37. Brown, F. 1995. Antibody recognition and neutralization of foot-and-mouth disease virus. Semin. Virol. 6: 243-248. Brown, F.2003 The history of research in foot-and-mouth disease. Virus Res. 91: 3-7. Brown F. and Crick, J. 1959 Application of gel diffusion analysis to a study of the antigenic structure of inactivated vaccines prepared from the virus of foot-and-mouth disease. J. Immunol. 82: 442-447. Brown, F. and Smale, C.J. 1970. Demonstration of three specific sites on the surface of footand-mouth disease virus by antibody complexing. J. Gen. Viral. 7: 115-127. Burroughs, J.N., Rowlands, D.J., Sangar, D.V., Talbot, P. and Brown, F. 1971 Further evidence for multiple proteins in the foot-and-mouth disease particle. J. Gen. Virol. 13: 73-84 Capstick, P.B., Telling, R.C., Chapman, W.O. and Stewart, D.L. 1962 Growth of a cloned strain of hamster kidney cells in suspended cultures and their susceptibility to the virus offoot-and-mouth disease. Nature 195: 1163-1164. Clarke, B.E., Newton, S.E., Carroll, A.R., Francis, M.J., Appleyard, G., Syred, A.D., Highfield, P.E., Rowlands, D.l and Brown, F. 1987. Improved immunogenicity of a peptide epitope after fusion to hepatitis B core protein. Nature 330: 381-384. Crowther, J.R., Farias, S., Carpenter, W.C. and Samuel, A.R. 1993. Identification of a fifth neutralizable site on type 0 foot-and-mouth disease virus following characterization of single and quintuple monoclonal antibody escape mutants. J. Gen. Virol. 74 ( Pt 8): 1547-1553. Curry, S., Fry E., Blakemore, W., Abu-Ghazaleh, R., Jackson, T., King, A., Lea, S., Newman, J., Rowlands, D., and Stuart, D. 1996. Perturbations in the surface structure of A22 Iraq foot-and-mouth disease virus accompanying coupled changes in host cell specificity and antigenicity. Structure. 4: 135-145. De Prat-Gay (1997) Conformational preferences of a peptide corresponding to the major antigenic determinant of foot-and-mouth disease virus: implications for peptidevaccine approaches. Arch. Biochem. Biophys. 341: 360-369. Di Marchi, R., Brooke, G., Gale, C., Cracknell, V., Doel, T. and Mowat, N.. 1986. Protection of cattle against foot-and-mouth disease virus by a synthetic peptide. Science 232: 639641.

Copyright © 2004 By Horizon Bioscience

Rowlands

3511

Doe], T.R. 2003. FMD vaccines. Virus Res. 91: 81-99. Domingo, E., Verdaguer, N., Ochoa, W.F., Ruiz-Jarabo, C.M., Sevilla, N., Baranowski, E., Mateu, M.G., and Fita, I. 1999. Biochemical and structural studies with neutralizing antibodies raised against foot-and-mouth disease virus. Virus Res. 62: 169-175. Dunn, C. S., Samuel, A.R., Pullen, L.A. and Anderson, J. 1998. The biological relevance of virus neutralisation sites for virulence and vaccine protection in the guinea pig model offoot-and-mouth disease. Virology 247: 51-61. Ferris, N.P. and Donaldson, A.I. 1992. The World Reference Laboratory for foot-and-mouth disease: a review of 33 years of activity (1958-1991) Rev. Sci. Tech. Off. Int. Epiz. 11: 657-684. Fox, G., Parry, N., Barnett, P.V., McGinn, B., Rowlands, D. J. and Brown F. 1989. The cell attachment site on foot-and-mouth disease virus includes the amino acid sequence RGD (arginine-glycine-aspartic acid). J. Gen. Virol. 70: 625-637. Francis, M.J. and Clarke, B.E.. 1989. Peptide vaccines based on enhanced immunogenicity of peptide epitopes presented with T cell determinants or hepatitis B core protein. Methods in Enzymology 178: 659. Francis, M.J., Fry, C.M., Rowlands, OJ. and Brown, F. 198 Qualitative and quantitative differences in the immune response to foot-and-mouth disease virus antigens and synthetic peptides. J. Gen. Virol. 69: 2483-2491. Francis, MJ., Hastings, G.Z., Clarke, B.E., Brown, A.L., Beddell, C.R., Rowlands, DJ. and Brown, F. 1990. Neutralizing antibodies to all seven serotypes of foot-and-mouth disease virus elicited by synthetic peptides. Immunol. 69: 171-176. Frenkel, H.S. 1947 La culture de la virus de la fievre aphteuse sur l'epithelium de la langue de bovides. Bull. Off. Int. des Epizoot. 28: 155-162. Haydon, D.T., Bastos, A.D.,. Knowles, NJ and Samuel, A. R. 2001. Evidence for positive selection in foot-and-mouth disease virus capsid genes from field isolates. Genetics 157: 7-15. Hewat, E.A., Verdaguer, N., Fita, I., Blakemore, W., Brookes, S., King, A., Newman, J., Domingo, E., Mateu, M.G. and Stuart, DJ. 1997. Structure of the complex of an Fab fragment of a neutralizing antibody with foot-and-mouth disease virus: positioning of a highly mobile antigenic loop. EMBO J. 16: 1492-1500. Hogle, J.M., Chow, M. and Filman, D.J. 1985. Three dimensional structure of poliovirus at 2.9 A resolution. Science, 229: 1358-1365. Jackson, T., Ellard, Abu-Ghazaleh, R., Brookes, S.M., Blakemore, W.E., Corteyn, A.H., Stuart, 0.1., Newman, J.W.I. and King A.M.Q. 1996. Efficinet infection of cells in culture by type 0 foot-and-mouth disease virus requires binding to cell surface heparan sulphate, J. Virol. 70: 5282-5287. Jackson, T., Sheppard, D., Denyer, M., Blakemore, W., and King, A.M.Q. 2000. The epithelial integrin avb6 is a receptor for foot-and-mouth disease virus. J. Virol. 74: 4949-4956. King, A M.Q. 2003 Epitopes of FMDV and their changeability. In: Foot-and-Month Disease: Control Strategies. B. Dodet and M. Vicar, eds. Editions scientifiques et medicale. Elsevier. 297-304. King, A.M.Q., Underwood, B.O., McCahon. D., Newman, J.W.I. and Brown, F. 1981. Biochemical identification of viruses causing the 1981 outbreak of foot-and-mouth disease in the UK. Nature 293: 479-480. Kit, M., Kit, S., Little, S.P., DiMarchi, R.D. and Gale, C. 1991 Bovine herpes-l (infectious bovine-rhinotracheitis virus) -based viral vector which expresses foot-and-mouth disease epitopes. Vaccine 9: 564-572. Kitson, J.D., McCahon, D. and Belsham, G.J. 1990. Sequence analysis of monoclonal antibody resistant mutants of type 0 foot and mouth disease virus: evidence for the involvement of the three surface exposed capsid proteins in four antigenic sites. Virology 179: 26-34. Copyright © 2004 By Horizon Bioscience

1352 Peptide Vaccines

Kleid, D.G., Yansura, D.G., Small, B., Dowbenko, D., Moore, D., Grubman, MJ., McKercher, P.D., Morgan, D.O., Robertson, B.H. and Bachrach, H.L. 1981. Cloned viral protein vaccine for foot and mouth disease; response in cattle and swine. Science 214: 1125-1129. Laporte, J., Grosclaude, J., Wantyghem, S. Bernard, S. and Rouze, P. 1973. Neutralization en culture cellulaire du pouvoir infectieux du virus de la fievre aphteusepar les serums provenent de porcs immunisee a l'aide d'une proteine viral purifee. C.R.SeancesHebd. Acad. Sci. Ser. D. 276: 3399. Lea, S., Hernandez, J., Blakemore, W., Brocchi, E., Curry, S., Domingo, E., Fry, E., AbuGhazaleh, R., King, A., Newman, J. and Stuart, D. 1994. The structure and antigenicity of a type C foot-and-mouth disease virus. Structure. 2: 123-139. Loeffler, F.and Frosch, P. 1998. Report of the commission for research on foot-and-mouth disease. Zentrabl Bacteriol. Parastenkunde Infektionkrankh. 23: 371-391. Logan, D., Abu-Ghazaleh, R., Blakemore, W., Curry, S., Jackson, T., King, A., Lea, S., Lewis, R., Newman, J., Parry, N., Rowlands, D. Stuart, D. and Fry, E. 1993. Structure of a major immunogenic site on foot-and-mouth disease virus. Nature. 362: 566-568. Martin, W.B. and Chapman, W.G. 1961. The tissue culture colour test for assaying the virus and neutralising antibody of foot-and-mouth disease and its application to the measuremant of immunity in cattle. Res. Vet. Sci. 2: 53-61. Mason, P., Rieder, E., and Baxt, B. 1994. RGD sequence of foot-and-mouth disease virus is essential for infecting cells via the natural receptor but can be bypassed by an antibodydependent enhancement pathway. Proc. Natl. Acad. Sci. U. S. A 91: 1932-1936. Mateu, M.G., Camarero, J.A., Giralt, E., Andreu, D. and Domingo, E. 1995. Direct evaluation of the immunodominance of a major antigenic site of foot-and-mouth disease virus in a natural host. Virology 206: 298-306. Mateu, M.G., Valero, M.L., Andreu, D. and Domingo, E. 1996. Systematic replacement of amino acid residues within an Arg-Gly-Asp- containing loop of foot-and-mouth disease virus and effect on cell recognition. 1. BioI. Chern. 271: 12814-12819. McCahon, D., Crowther, J.R., Belsham, GJ., Kitson, J.D., Duchesne, M., Have,P., Meloen, R.H.,. Morgan, D.O. and De Simone, F. 1989. Evidence for at least four antigenic sites on type 0 foot-and-mouth disease virus involved in neutralization; identification by single and multiple site monoclonal antibody-resistant mutants. J. Gen. Virol. 70 ( Pt 3): 639-645. McCullough, K.C., De Simmone, F., Brocchi, E., Capucci, L., Crowther, J.R. and Kihm, U. 1992. Protective immune response against foot-and-mouth disease. J.Virol. 66: 18351840. Meloen, R., Rowlands, OJ., and Brown, F. 1979. Comparison of the antibodies elicited by the individual structural polypeptides of foot and mouth disease virus and poliovirus. J. Gen. Virol. 45: 761-763. Mowat, G,N, and Chapman, W,G, 1962. Growth of foot-and-mouth disease virus in a fibroblastic cell line derived from hamster kidneys. Nature 194: 253-255. Ochoa, W.F., Kalko, S.G., Mateu, M.G., Gomes, P., Andreu, D., Domingo, E., Fita, I. and Verdaguer, N. 2000. A multiply substituted G-H loop from foot-and-mouth disease virus in complex with a neutralizing antibody: a role for water molecules. J. Gen. Virol. 81: 1495-1505. Parry, N., Fox, G., Rowlands, D., Brown, F., Fry, E., Acharya, R., Logan, D. and Stuart, D. 1990. Structural and serological evidence for a novel mechanism of antigenic variation in foot-and-mouth disease virus. Nature 347: 569-572. Parry, N.R., Ouldridge, EJ., Barnett, P.V., Clarke, B.E., Francis, M.J., Fox, J.D., Rowlands, D.J. and Brown, F. 1989. Serological prospects for peptide vaccines against foot-andmouth disease virus. J. Gen. Virol. 70 (Pt 11): 2919-2930.

Copyright © 2004 By Horizon Bioscience

Rowlands

3531

Parry, N.R., Syred, A.D., Rowlands, D.J. and Brown, F. 1988. A high proportion of antipeptide antibodies recognise foot and mouth disease virus particles. Immunology, 64, 567-572. Pfaff, E., Mussgay, M., Bohm, H.O., Schulz, G.E. and Schaller, H. 1982. Antibodies against a preselected peptide recognize and neutralize foot and mouth disease virus. EMBO J. 1: 869-874. Rossmann, M.G., Arnold, E., Erickson, J.W., Frankenberger, E.A., Griffith, J.P., Hecht, H.J., Johnson, J.R., Kamer, G., Luo, M., Mosser, A.G., Rueckert, R.R., Sherry, B. and Vriend, G. (1985) Structure of a human common cold virus and functional relationship to other picornaviruses. Nature, 317: 145-153. Rowlands, OJ. 1993. Progress towards peptide vaccines for foot-and-mouth disease. In: Veterinary Vaccines. Panday R., Hogland S., and Prasad G., eds. Springer-Verlag, New York. p. 54-86. Rowlands, OJ., Clarke, B.E., Carroll, A.R., Brown, F., Nicholson, B.H., Bittle, J.L., Houghten, R.A. and Lerner, R.A. 1983. Chemical basis of antigenic variation in footand-mouth disease virus. Nature 306: 694-697. Rowlands, D.J., Sangar, D.V. and Brown F. 1971. Relationship of the antigenic structure of foot-and-mouth disease virus to the process of infection. J. Gen. Virol. 13: 85-93. Ruppert, A., Arnold, N. and Hobom, G. 1994. OmpA/FMDV VPl fusion proteins: production, cell surface exposure and immune responses to the major antigenic domain of foot-and-mouth disease virus. Vaccine 12: 492-498. Sa-Carvalho, D., Rieder, E., Baxt, B., Rodarte, R., Tanuri, A. and Mason, P.W. 1997. Tissue culture adaptation of foot-and-mouth disease virus selects viruses that bind to heparin and are attenuated in cattle. J. Virol. 71: 5115-5123. Saiz, J.C., Cairo, J., Medina, M., Zuidema, D., Abrams, C., Belsham, GJ., Domingo, E. and Vlak J.M. 1994. Unprocessed foot-and-mouth disease virus capsid precursor displays discontinuous epitopes involved in viral neutralization. J. Virol. 68: 4557-4564. Siligardi, G., Drake, A.F., Mascagni, P., Rowlands, D.J., Brown, F. and Gibbons, W.A. 1991. Correlation between the conformations elucidated by CD spectroscopy and the antigenic properties of four peptides of the foot and mouth disease virus (FMDV). Euro. 1. Biochem 199: 545-551. Siligardi,G., Drake, A.F., Mascagni, P., Rowlands, OJ., Brown, F. and Gibbons,W,.A. 1991. A CD strategy for the study of polypeptide folding/unfolding: A synthetic foot and mouth disease virus immunogenic peptide. Int. J. Peptide Protein Res. 387: 519-527. Strohmaier, K., Franze, R. and Adam, K.H. 1982. Location and characterization of the antigenic portion of the FMDV immunizing protein. J. Gen. Virol. 59: 295-306. Taboga, 0., Tami, C., Carrillo, E., Nunez,J.!., Rodriguez, A., Saiz, J.C., Blanco,E., Valero, M.L., Roig, X., Camarero, J.A., Andreu, D., Mateu, M.G., Giralt, E., Domingo, E., Sobrino, F. and Palma, E. L. 1997. A large-scale evaluation of peptide vaccines against foot-and-mouth disease: lack of solid protection in cattle and isolation of escape mutants. J. Virol. 71: 2606-2614. Tam, J.P. 1988. Synthetic peptide vaccine design: Synthesis and properties of a high-density multiple antigenic peptide system. Proc. Natl. Acad. Sci. 85: 5409-5413. Thomas, A.A., Woortmeijer, R.J., Puijk, W. and Barteling, SJ. 1988. Antigenic sites on footand-mouth disease virus type A10. J. Viral. 62: 2782-2789. Valero, M-L., Camarero, J.A., Haack, T., Mateu, M.G., Domingo, E., Giralt, E. and Andreu, D. 2000. Native-like cyclic peptide models of a viral antigenic site: finding a balance between rigidity and flexibility. J. Mol. Recog. 13: 5-13. Vallee, H., Carre H. and Rinjard, P. 1925. On immunisation against foot-and-mouth disease. Rech. Med. Vet. 101: 297-299. Van Regenmortel, M.H., Guichard, G., Benkirane, N., Briand, J.P., Muller, S. and Brown, F. 1998. The potential of retro-inverso peptides as synthetic vaccines. Dev. BioI. Stand. 92: 139-143. Copyright © 2004 By Horizon Bioscience

1354 Peptide Vaccines

Verdaguer, N., Mateu, M.G., Andreu, D., Giralt, E., Domingo, E. and Fita, I. 1995. Structure of the major antigenic loop of foot-and-mouth disease virus complexed with a neutralizing antibody: direct involvement of the Arg- Gly-Asp motif in the interaction. EMBO J. 14: 1690-1696. Verdaguer, N., Mateu, M.G., Bravo, J., Domingo, E. and Fita, I.. 1996. Induced pocket to accommodate the cell attachment Arg-Gly-Asp motif in a neutralizing antibody against foot-and-mouth-disease virus. J. Mol. BioI. 256: 364-376. Verdaguer, N., Sevilla, N., Valero, M., Stuart, D., Brocchi, E., Andreu, D., Giralt, E., Domingo, E., Mateu, M.G. and Fita, I. (1998) A similar pattern of interaction for different antibodies with a major antigenic site of foot-and-mouth disease virus: implications for intratypic antigenic variation. J.Virol. 72: 739-748. Villen, J., Borras, E., Schaaper, W.M.M., Meloen, R.H., Davila, M., Domingo, E., Giralt, E. and Andreu, D. 2001. Synthetic peptides as functional mimics of a viral discontinuous antigenic site. Biologicals 29: 265-269. Volpina, O.M., Yarov, A.V., Zhmak, M.N., Kuprianova, M.A., Chepurkin, A.V., Toloknov, A.S. and Ivanov, V.T. (1996) Synthetic vaccine against foot-and-mouth disease based on a palmitoyl derivative of the VPl protein 135-159 fragment of A 22 virus strain. Vaccine 14: 1375-1380. Waldmann, 0., Kobe, K. and Pyle, G. 1937 Die aktive immunisierungdes rindes gegen maule-und klauenseuchevirus. Berl. Tierarztl Wschr. 42: 569-571. Wild, T.F. and F. Brown. 1967. Nature of the inactivating action of trypsin on foot-andmouth disease virus. J. Gen. Virol. 1: 247-250.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 13 Mathematical Models of the Epidemiology and Control of Foot-and-Mouth Disease

Mark E. J. Woolhouse

Abstract This review considers how epidemiological models are constructed, how they deal with real-life complexities such as spatial heterogeneity, how they can be applied to specific FMD outbreaks or epidemics, and how they can be used to explore the impact of control measures. A detailed description is provided of the application of a particular model, the 'Keeling' model, of the spread of FMD between farms in the UK during the 2001 epidemic. The review concludes with a brief discussion of how mathematical modelling of livestock disease is likely to develop in the future. The emphasis throughout is on 'good practice', especially how theoretical models relate to biological data and how models can sensibly be used to inform decisions about disease control strategies. 1. Introduction Mathematical models of the spread of infectious diseases were first developed early in the 20 th century and have made important contributions to improving epidemiological understanding and designing control programmes for many human diseases, including malaria, measles, tuberculosis and HIV/AIDS (see Anderson and May, 1991; Daley and Gani, 2001). In recent years there have also been applications of mathematical models to animal diseases, including BSE (Anderson et al., 1996), classical swine fever (Stegeman et al., 1999), scrapie (Matthews et al., 2001) and various wildlife diseases (Hudson et ai., 2002). Mathematical models are potentially powerful tools for infectious disease epidemiologists because infectious diseases have inherently complex dynamics. This makes it very difficult, on the basis of practical experience alone, to develop a quantitative understanding of the epidemiological process, or to generalize from one situation to others, or to predict the impact of control measures. Tackling these kinds of problems requires the kind of precise, formal, quantitative framework that mathematical models can offer, with the obvious proviso that the models must have a sound biological basis to be of practical value. The potential importance of mathematical models was underlined in the Royal Society of London's report on the 2001 epidemic of foot-and-mouth disease in the UK, which stated that: "Quantitative modelling is one of the essential tools both for developing strategies in preparation for an outbreak and for predicting and evaluating the effectiveness of control strategies during an outbreak" (Royal

Copyright © 2004 By Horizon Bioscience

\356

Mathematical Models

Table 1. Glossary Basic reproduction ratio. The average number of new cases of infection directly generated by a single case introduced into a previously unexposed host population. Also known as Ro. Bias. A systematic tendency for the value of a parameter to be over-estimated or underestimated. Bootstrapping. A method of generating confidence intervals around a parameter estimate by repeatedly sampling with replacement from the source data set to create 'new' datasets and re-estimating the parameter value each time. Case reproduction ratio. The average number of new cases generated by a single case. Also known as R or R t • It can be estimated at any stage of an epidemic and is usually expected to be less than Ro. Compartments. Discrete subsets of the host population defined according to their infection status. Commonly used compartments are susceptible, latent infected, infectious and recovered/removed. These occur in SLIR models of infection dynamics. Deterministic. The output is fully determined by the inputs. Chance is not involved. Difference equations. Equations using discrete time steps, e.g. one day. Differential equations. Equations using calculus, i.e. infinitely small time steps. Estimation. The procedure by which the value of a parameter is estimated from data. There are many different statistical approaches to parameter estimation, including least squares fitting, maximum likelihood fitting, martingale estimators and Bayesian methods. Parameters will almost always be estimated with some degree of uncertainty (giving rise to confidence intervals) and may be subject to bias, noise or non-independence. Generation time. The average interval between the time of infection of a case and the time of infection of new cases generated from it. Latin hypercube sampling. A technique for sensitivity analysis which can explore the effects of variations in many different inputs at the same time, but which is much more efficient than comparing outputs for every possible combination of inputs. Microparasites. Pathogens which cause infections which can usefully be represented in terms of compartments, e.g. susceptible or infected. The term is generally applicable to infections by viruses, bacteria or protozoa. Microsimulation. Stochastic mathematical models in which each individual in the population is represented explicity, as opposed to tracking the number of individuals in each of a set of compartments. Model. Here, a mathematical representation of a dynamic process, such as the spread of an infection through a population. Network. In this context, the numbers and distribution of links between hosts, where 'link' implies an opportunity to transmit infection. An example is the network of farms linked by movements of livestock. Noise. Inherent variability and/or imprecision in data which makes it difficult to obtain precise parameter estimates. Non-independence. Correlation between two or more parameter estimates such that there is a range of combinations of parameter values which are consistent with the data. Non-stationary. Meaning that parameter values are not constant through time. Parameter. A constant (Le. a fixed number) in a mathematical term which determines how the value of a variable in a model changes through time.

Copyright © 2004 By Horizon Bioscience

Woo/house

3571

Table 1, Continued. Sensitivity analysis. A generic term for various methods of exploring how model outputs are related to model inputs in order to determine which of the model's assumptions and/or parameter values are most important in determining its behaviour. One method for carrying out sensitivity analysis is Latin hypercube sampling. SLIR. An abbreviation for a commonly used mathematical model with four compartments: susceptible, latent infected, infectious and recovered/removed. Stochastic. The output is not fully determined by the inputs. Chance is incorporated in the process and every realization of that process can produce a different outcome. Stochastic effects are particularly important where the numbers involved are small, e.g. at the start or during the 'tail' of an epidemic when there are few infectious individuals. Transmission kernel. A mathematical function describing the relationship between transmission rate and distance. Transmission rate. The average rate at which a single infectious individual infects susceptible individuals. Net transmission rate refers to per capita transmission rates summed across all infectious individuals. Validation. The process where the outputs of a model are compared with a fully independent data set, Le. one which was not used to provide estimates of any of the model's parameters. Variable. A quantity whose value is tracked in a mathematical model. An example is the number (or fraction) of individuals in one of the compartments of a SLIR model. Well-mixed. This implies that every individual in a population is equally likely to infect every other individual. Although clearly a simplification the assumption that a population is wellmixed is often used as a starting point in developing a model of the spread of infection.

Society, 2002). The Royal Society's report should encourage the more widespread use of mathematical models in veterinary medicine. It can be anticipated that models will come to play a much more central role in epidemiological analyses, designing control programmes, contingency planning and policy making than they do at present, not just for foot-and-mouth disease (FMD) but for other infectious diseases of animals.

2. Model Structure 2.1. Types of Model There are various ways in which mathematical models can be used in epidemiological studies. One useful distinction is between retrospective and prospective models. Retrospective modelling involves fitting mathematical equations to epidemiological data and is used as a technique for the quantitative interpretation of those data (e.g. Haydon etal., 1997; Howard and Donnelly, 2000; Ferguson et al., 2001b). An extension of this approach is to take a model which has been fitted to epidemiological data and use it to examine alternative scenarios, such as the implementation of different control measures; this is sometimes referred to as 'what-if' modelling (e.g. Keeling et aI., 2001; Morris et ai., 2001). Prospective models are used in two ways. One is predictive, using current data as the basis for predicting the course of an ongoing epidemic (e.g. Ferguson et ai., 2001a). The other is exploratory, modelling a range of possible epidemiological scenarios rather than focusing on a particular event; such models are often used to aid contingency planning (e.g. Gamer and Lack, 1995; Durand and Mahul, 2000; Keeling et ai., 2003). Copyright © 2004 By Horizon Bioscience

\358

Mathematical Models

A latent

susceptible

[] B

!

ft(·)

~~

infection

infectious

h(·)

~[!]

reporting

pre-clinical

removed

13(·)

!

~~

removal

!

clinical

Figure 1. Compartments, in theory and practice. A) A diagrammatic representation of a simple SLIR model showing the flow of hosts between susceptible, latent infected, infectious and removed compartments. The numbers (or fractions or densities) of hosts in these compartments are represented by the variables S, L, I and R respectively. The rate of flow is specified by three expressions: II (.), the rate at which susceptible hosts become infected; 12(·), the rate at which latently infected hosts become infectious; and/3(e), the rate at which infectious hosts are removed. Different models use different mathematical expressions, representing different levels of detail and incorporating different numbers of parameters. B) A diagrammatic representation of the course of a FMD infection in a single host (or single farm). Note that the transition between latent and infectious does not correspond to the appearance of clinical signs: animals may be infectious before clinical signs appear. In practice, there is inevitably a further delay before clinical signs are observed and reported.

Models may be deterministic or stochastic. Deterministic models (e.g. Haydon et al., 1997; Ferguson et aI., 2001a; 2001b) generate a fixed output for a given set of inputs. Stochastic models (e.g. Gamer and Lack, 1995; Keeling et al., 2001; Morris et al., 2001) generate variable outputs for a given set of inputs; these models incorporate chance in the epidemiological process and each model run can produce a different result. Deterministic models are typically formulated as a set of coupled differential equations representing the dynamics of subsets of the host population corresponding to different infection states (see below). Stochastic models fall into two main categories: a set of coupled stochastic difference equations analogous to differential equations; and microsimulation or state transition models, where the current state of each host in the population is represented individually. Stochastic models are often (though not always) much more difficult to compute, to fit to data and to interpret than comparable deterministic models, but they do better reflect the intrinsic uncertainty in any epidemiological process. Enormous increases in computer processing power in recent years have made stochastic models, particularly microsimulations, a much more practical option, and they seem likely to be more widely used in the future.

Copyright © 2004 By Horizon Bioscience

Woo/house

359 I

Each mathematical approach has its advantages and disadvantages and it is important to recognise that no one approach is 'right' and that there can be no single model for the epidemiology of FMD or any other infectious disease. Rather, the merits of different approaches have to be judged against the purpose of the modelling exercise. Indeed, there may be advantages to modelling work being carried out in parallel: for example, four different modelling approaches were used to inform policy makers during the course of the UK 2001 FMD epidemic and the close agreement between these greatly increased confidence in the robustness of the outputs (Kao, 2002). 2.2. Compartments The standard approach for modelling the epidemiology of microparasite (virus, bacteria or protozoa) infections is to divide the host population into different compartments: susceptible, denoted S; latent, i.e. infected but not yet infectious, L; infected and infectious, I; recovered (or removed), R. The dynamics of infection are then represented by the movement of hosts from one compartment to another (Figure lA), as is described below. Such a model is usually referred to as a SLIR model or a SEIR model (the E standing for 'exposed') (Anderson and May, 1991). If vaccination is involved there may also be a compartment, denoted V, representing vaccinated individuals (Hutber and Kitching, 1996). The SLIR structure is widely used and has the advantage of simplicity, but some care is needed in its application to FMD infections. The first problem is that the compartments susceptible, latent, infectious and recovered correspond only imperfectly to the states that can be defined in the field (Figure IB). FMD infection can be demonstrated by the detection of virus, the detection of antibodies to the virus, or the appearance of clinical signs (Hughes et al., 2002a). There is, however, no simple correspondence between these assays and whether or not the host is latently infected, infectious or recovered; for example, most infected hosts are infectious before clinical signs appear (and in some hosts, notably sheep, clinical signs may not be detected at all). For this reason, considerable care must be taken when imposing the SLIR structure on field data. A second problem is that the SLIR compartments are discrete; that is, a host is either susceptible or latent or infectious or removed, and all hosts are equally susceptible when susceptible and equally infectious when infectious. This is clearly a major simplification: viraemias may vary by orders of magnitude over the time course of a single infection and between infections of different hosts (Hughes et al., 2002b); and antibody levels similarly vary greatly with time and between hosts (Woolhouse et aI., 1996b). One factor affecting the course of a single infection, and how infectious an infected host will be, is the virus dose to which it is exposed (Hughes et al., 2002c), which itself may be highly variable in practice. Mathematical models attempt to take account of this variability within compartments in various ways. One solution is to define more compartments (e.g. subclinically infected or partially immune) (Rutber and Kitching, 1996). A more sophisticated but computationally demanding alternative is to represent the 'state' of a host quantitatively rather than qualitatively (e.g. levels of viraemia rather than infected or not infected; antibody titre rather than infected or recovered ) (Woolhouse et al., 1996b; White and Medley, 1998; Stringer et ai., 1998). This Copyright © 2004 By Horizon Bioscience

\360

Mathematical Models

requires functional relationships, e.g. between antibody titre and susceptibility, to be defined. These are useful refinements \vhich merit further development, linking mathematical analyses to experimental studies of the dynamics of infections within individual hosts. A third problem is that the unit of epidemiological interest is not necessarily an individual host; many applications of mathematical models consider populations of livestock farms rather than of individual animals (e.g. Haydon etal., 1997; Durand and Mahul, 2000; Howard and Donnelly, 2000). There is no conceptual difficulty with classifying individual farms as susceptible, latent, infectious or removed, but these compartments are unlikely to be equivalent to those for individual hosts. Considerable care must be exercised in 'scaling up' from the animal level to the farm level - the timescales and nature of the epidemiological processes involved are likely to be distinctly different. As an example, consider how a farm makes the transition from latently infected to infectious. If this occurs when the first infected animal becomes infectious then the farm-level latent period will be of the same order as the individual-level latent period. But if considerably more virus is required for effective transmission between farms than for transmission between individual animals, as might be expected given that different transmission routes may be involved, then the farm-level latent period may be longer, possibly much longer. The situation is further complicated in that the dynamics of FMD outbreaks on individual farms are likely to be highly variable (Woolhouse et al., 1996b), just as are the dynamics of infections of individual animals. While its limitations must be borne in mind, it should be recognised that the SLIR approach has proved extremely useful for developing a quantitative understanding of the epidemiology and control of a wide variety of microparasite infections (Anderson and May, 1991) and is likely to remain a standard format for epidemiological models for some time to come. 2.3. Parameters and Variables A SLIR model describes infection dynamics in terms of changes in the numbers (or densities or fractions) of hosts in the compartments susceptible, latent, infectious and removed. These numbers are represented by the variables S, L, I and R respectively. Movements between these compartments occur at rates specified by a set of mathematical terms (see Figure lA). In the simplest formulation, one term specifies the rate at which susceptibles become infected, another the rate at which latently infected individuals become infectious, and another the rate at which infectious individuals recover or are removed. There may also be terms specifying, for example, the rate at which susceptibles are vaccinated and the rate at which vaccinated individuals lose their protection. Each of these terms incorporates one or more parameters (mathematical constants which determine how the values of the variables change), such as the transmission rate (see below). The exact functional form reflects the underlying assumptions of the model. To take a simple example, the term oL is often used in differential equation models to represent the rate at which latent infected individuals become infectious. Expressed thus, the implication is that latent individuals become infectious at a constant per capita rate represented by the parameter 0, that the mean latent period is 1/0, but that the modal latent period is extremely short. This last aspect is often regarded as unsatisfactory, and an alternative is to make the latent period itself a fixed constant. Copyright © 2004 By Horizon Bioscience

Figure 2. The effect of different values of the basic reproduction number, Ro. The left hand diagram illustrates the situation where Ro> 1, so that each case produces, on average, more than one new case and a major epidemic is possible. The right hand diagram illustrates the situation where Ro 1, the epidemic may fail to take off through chance alone, and that, even if Ro< 1, chains of infection constituing a minor outbreak are still possible.

Expressed thus, the implication is that all latent individuals become infectious at a fixed time after they were first infected, with no variation. The biological reality is likely to lie between these extremes. There are many potentially suitable functions to describe that situation, e.g. the gamma distribution, but these are inevitably more complex (i.e. they incorporate more parameters) and are more problematic to compute (Stringer et a!., 1998). Consequently, constant rates or constant periods are frequently assumed in practice, although it is often straightforward to compare models of both types to see if there are significant differences in their outputs (Woolhouse et ai., 1997b; Keeling and Grenfell, 2002). The number of parameters in a model is limited only by the level of biological detail which is to be represented. For example, it is possible to represent different transmission routes separately and so have one or more parameters relating to each of airborne spread, direct contact, livestock movements, vehicle movements and so on. One microsimulation model for FMD which takes this approach has over 50 different parameters (Morris et al., 2001). This allows the model to be more realistic, but creates problems with parameter estimation and model validation. 2.4. Quantifying Transmission A key parameter for any epidemiological model is the basic reproduction ratio, Ro, which is defined as the average number of secondary cases arising from the introduction of a single primary case into a previously unexposed population. If Ro>1 then each case is more than capable of replacing itself and an epidemic can take off. On the other hand, if Rol then the epidemic is growing and may be regarded as 'out of Copyright © 2004 By Horizon Bioscience

\362 Mathematical Models

control' at time t, indicating that additional control measures may be warranted. On the other hand, if Rt ;::

0.25 -

ftS

&

0 10

ciI 0

~

~

I

I

~

10

ci

~

C!

0

C\iI ~

10

C\iI

0

0

('I) I

10

C\i C\i Distance (km)

10

('I) I

0 ~

o ..0 I

10

~

Figure 3. Examples of transmission kernels, relative per capita rate of transmission as a function of distance between farms. Empirical results (symbols) derived using data tracing studies carried out during the UK 2001 epidemic after the imposition of a national ban on livestock movements (Keeling et ai., 2001) are compared with two standard functions: 1) k/cP with k=O.41 (broken line); and 2) gexp[-hd] with g=4.8 and h=2.4 (solid line). All functions show per capita transmission rates relative to that over distances of 0 to 0.5km. The constants k, g and h were fitted using the least squares method. The empirically-derived transmission kernel equates to 70% of transmission occurring over distances up to 3km. Note that function (1) overestimates transmission rates at longer distances, whereas function (2) underestimates these.

Possible representations of transmission kernels are contrasted in Figure 3. For spatially explict microsimulation models the parameter f3 is replaced by individual values for the transmission rate between a given pair of farms. These rates may be weighted by other risk factors for susceptibility and infectiousness. If epidemics in large populations of farms (the UK has over 100,000 livestock farms) are being modelled, this approach can become very computationally intensive. A refinement to modelling spatial transmission is to represent different routes of transmission explicitly, and to calculate rates of transmission for each of these (Morris et al., 2001). Such models require even more detailed inputs - e.g. road networks, milk tanker routes, footpaths, wind speed and direction - and so are often coupled to the development of geographical information system (GIS) representations of the landscape (Sanson, 1999). Many transmission parameters must be estimated and validated, and the overall transmission rate between a pair of farms is obtained by summing the rates of transmission by different routes, again weighting by known risk factors. Alternatively, there are more abstract ways in which spatial structure can be represented in mathematical models. One of these is percolation theory, a stochastic approach to modelling how an infection spreads through a geometric lattice which has been used to explore, in general terms, how the nature of the farming landscape might influence FMD dynamics (Kao, 2001). Another approach is to extend the deterministic SLIR framework using a technique known as moment closure. This approximates local spread by considering the status not just of individuals but of pairs, triplets etc. (Ferguson et ai., 2001a). A third technique, small world network modelling, is based on a SLIR framework but distinguishes between local spread Copyright © 2004 By Horizon Bioscience

Woo/house

365 I

and long-distance transmission events (Ahmed et aI., 2002). Percolation theory, moment closure and network models are, at best, relatively crude approximations to what may in practice be quite complex and variable spatial arrangements of farms. Because of this, there are many instances where it will be preferable to use spatially explict microsimulation models despite their greater computational demands.

4. Model Application 4.1. Parameter Estimation

A major challenge for developing a mathematical model for FMD (or any other infectious disease) is how to obtain estimates of the model's parameters. In general, these will come from two sources: i) prior knowledge, whether from experimental studies or from analyses of earlier epidemics; and ii) estimation from the epidemic being modelled, whether directly from inspection of the available data or indirectly from fitting the model to those data. In theory, it is desirable to obtain parameter estimates from prior knowledge - the model is then statistically independent of the data. In practice, each epidemic is likely to have its own unique features and some parameter estimates will have to be obtained using data from that same epidemic. The balance between these will depend on the availability and quality of data from independent sources and whether it is reasonable to generalize across different epidemics. For example, the incubation period can often be specified independently (mindful of differences between virus strains and livestock species and the difficulty of scaling up from individual animals to entire farms) whereas the transmission rate (whether within or between farms) may often be highly variable (depending on farm mangement practices, weather conditions, effectiveness of control measures and many other factors) and usually has to be re-estimated each time a model is applied. The mechanics of estimating parameters directly have been reviewed elsewhere (e.g. Richter and Sondgerath, 1990). In general, there are three main concerns: noise, bias and non-independence. Noise simply means that biological data tend to be inherently variable (and sometimes difficult to measure precisely): the noisier the data, the less precise the parameter estimate derived from the data. An example of noise comes from fluctuations of numbers of cases reported during the UK 2001 epidemic. Superimposed on the main trends, there was considerable day-to-day variation in numbers of cases reported per day (partly explained by day-of-the-week effects), which is especially important when numbers of cases are small, as during the epidemic's tail. This means that estimates of case reproduction ratios have limited precision which, for example, makes short-term projections difficult. Noise can sometimes be reduced by making more precise measurements. Where this is not possible, the implications of uncertain parameter values must be explored using sensitivity analysis (see below). Bias means that the available data tend to over-estimate or under-estimate the value of a parameter. A possible example of bias comes from estimated dates of FMD infection on UK sheep farms in 2001: these were often obtained by ageing lesions, but lesions were not always apparent, especially in sheep, and so early infections could have been missed and the date of infection underestimated (Gibbens and Wilesmith, 2002). Bias can be difficult to eliminate and its potential importance needs to be gauged; again this can be done using sensitivity analysis. Copyright © 2004 By Horizon Bioscience

\366

Mathematical Models

Non-independence means that uncertainties in the values of two or more parameters are related to each other, e.g. a high estimate of one parameter tends to be associated with a low value for another, and vice versa. A consequence of this is to increase the uncertainty associated with a given parameter estimate; there may be a much wider range of combinations of two or more parameter values that are consistent with the data than would be apparent by considering each one separately. Two examples come from the estimation of Ro from data on the UK 2001 epidemic: various combinations of Ro and generation time (the average time between a farm being infected and new farms being infected from it) are consistent with the initial rise in numbers of cases per day; similarly, various combinations of Ro and the number of 'at-risk' neighbouring farms are consistent with observed rates of local spread (Kao, 2001). Non-independence is most straightforwardly resolved by obtaining independent estimates of one of the parameters involved. Methods for estimating parameters indirectly by fitting a mathematical model to data are the subject of a large literature (see Becker, 1989) and are becoming increasingly sophisticated. Deterministic models can be fitted by least squares (e.g. Woolhouse et al., 1996b) or maximum likelihood methods (Anderson et al., 1996). Stochastic models can be fitted by a variety of techniques including martingale estimation, maximum likelihood and Bayesian methods - the latter two being increasingly implemented using Markov chain Monte Carlo (MCMC) methods (Gamerman, 1997). Even so, the sheer complexity of the system to be modelled and the structure of the data often mean that the available 'off-the-shelf' modelfitting procedures are inappropriate. As a result, model fitting is frequently done using ad hoc methodologies (e.g. Keeling et ai., 2001). It is desirable that modelfi tting procedures become better standardized than they are at present. An alternative approach to estimating R t values is to use a non-parametric method which does not impose a model on the data. This has been attempted for the UK 2001 epidemic using information on contact tracings and estimated dates of infection (Woolhouse et al., 2001; Haydon et ai., 2003). These data allow the entire history of the epidemic, in terms of which farm was thought to have infected which other farm, to be described. The number of cases per case, for any time period or any region, can then simply be read off directly. Uncertainty about the source of infection can be incorporated by generating epidemic histories on a probabilistic basis and generating confidence limits around R t values using bootstrapping methods (where the data are repeatedly sampled with replacement and R r re-estimated each time in order to generate confidence intervals) (Haydon et aI., 2003). This method is likely to be much more robust than methods based on model-fitting, especially for small epidemics, but does require better information on individual FMD cases than is usually available. Estimates of transmission rates can also be obtained from experimental studies of the transmission of FMD (e.g. Hughes et al., 2002c) and other infections (e.g. Bouma et al., 1995, for pseudorabies in pigs). Such studies are extremely valuable, but are normally restricted to small numbers of animals kept in a limited area under experimental conditions. The results cannot necessarily be applied directly to the field situation, and considerable care is required in extrapolating to much larger scales, e.g. transmission between farms.

Copyright © 2004 By Horizon Bioscience

Woo/house

3671

4.2. Inputs and Initial Conditions The simplest mathematical models consider only intrinsic dynamics; that is, the behaviour of the system is fully specified by the values of the model's parameters and the starting values for each of the model's variables. More complex models, however, may attempt to take into account extrinsic variables, such as weather conditions or the magnitude of the control effort, especially if these are not constant in time or space. These will appear as an additional set of inputs in the model, affecting the value of one or more parameters, e.g. transmission rates as a function of wind speed. An additional, and often crucial, set of inputs is the demographic and epidemiological characteristics of the population. The level of detail required is hugely variable, ranging from simply the initial number of susceptibles in the population (N) to comprehensive information on geographic location of, pattern of movement by and risk factors associated with every single individual in the population. An equally important, but often neglected, consideration is the starting conditions from which the model is to be run. At the simplest level, starting conditions might specify the introduction of a single primary case into a homogeneous population thus: S(O)=N-l; L(O)=I; 1(0)=0; R(O)=O. More complex models require the subpopulation or the actual individual infected to be specified. In practice, information on the early course of an epidemic is often sketchy or incomplete and, in addition, the processes important early on in the course of an epidemic may be very different from those important later on. For example, the UK 1967/68 epidemic involved early widespread wind-borne dissemination of infection from a few primary cases (Tinline, 1970) and the UK 2001 epidemic involved early dissemination via sheep markets, which was subsequently halted (Gibbens et al., 2001). As a result, it is often necessary to initialize a mathematical model based on the epidemic situation some time after the first introduction of infection (e.g. Woolhouse et aI., 1996b; Keeling et al., 2001). 4.3. Sensitivity Analysis and Validation Once a model has been constructed it is important that its behaviour is explored in detail. There are two elements to this. One is to explore the effect of various assumptions made in developing the model. Examples include comparison of the effects of assuming a constant latent period with a constant rate at which latently infected individuals become infectious, or comparison of the effects of assuming density- or frequency-dependent transmission (see above). This is referred to as the structural sensitivity of the model. Obviously, it is vital that those assumptions which are shown to have a significant effect on model behaviour are tested as rigorously as possible. The second element of exploring model behaviour is to consider the effects of different inputs: parameter values; starting conditions; and any extrinsic variables. This is done using sensitivity analysis. The simplest form of sensitivity analysis is to alter the value of each input separately (Woolhouse et aI., 1996a; Hutber and Kitching, 1996; Durand and Mahul, 2000). However, because the dynamics of infectious diseases are typically complex, a more robust approach is to explore the effects of changes in combinations of inputs. This can be done using a straightforward factorial design, but becomes unwieldy when large numbers

Copyright © 2004 By Horizon Bioscience

\368

Mathematical Models

of inputs are involved. In such circumstances a less computationally intensive alternative is to use Latin hypercube sampling (Blower and Dowlatabadi, 1994). This technique involves sampling without replacement from a set of probability distributions, one for each input. Model outputs can then be compared statistically with variations in input values, and those which have the most significant effects identified (Lord et ai., 1996). In practice, many applications of Latin hypercube sampling explore the effects of two different kinds of uncertainty in the value of a parameter or variable: uncertainty due to imperfect knowledge; and uncertainty due to inherent biological variability. In either case, if the value of an input has a significant effect then it is clearly important that that value is measured or monitored as precisely as possible. Validation of an epidemiological model is ideally achieved by comparing model output with a fully independent data set, Le. testing the model's predictions. In practice, it is often difficult to take a model parameterized for one epidemic and apply it to another: each epidemic tends to have its own unique characteristics and history. Even so, it is sometimes possible to use the same modelling framework in different settings, which constitutes partial validation even if some re-parameterization is necessary: this has been done for BSE in cattle (Anderson et ai., 1996; Ferguson et ai., 1998) and scrapie in sheep (Woolhouse et ai., 1999; Matthews et ai., 2001). There are two problems with model validation. First, it is not always clear what constitutes a good 'fit' between model and data. Second, it is often far from clear that the suggested model offers a unique solution; in other words, there may be other models, making different assumptions and using different parameter values, which fit the data as well or better. Pending objective solutions to these issues it is important that model output is evaluated critically, cautiously and, preferably, independently. 4.4. Results of Model Fitting

Several studies have attempted to fit SLIR models (or elaborations thereof) to FMD epidemic data (Haydon et ai., 1997; Howard and Donnelly, 2000; Ferguson et ai., 2001b). These studies variously used data from epidemics in the 'UK in 1967/68, Taiwan in 1997 and the UK in 2001 and an important aim was to quantify ~, the per capita transmission rate between farms. However, none of these model fitting exercises was entirely satisfactory because it proved impossible to reproduce the epidemic curve without invoking substantial changes in the value of ~, i.e. the parameter estimate was non-stationary. This occurred even when factors such as changes in the control effort (directly) and the effects of local spread (indirectly, using moment closure) were incorporated (Ferguson et ai., 2001b). It is, of course, possible that underlying transmission rates did vary throughout the course of these epidemics, but a more satisfactory resolution of the problem is suggested by using a spatially explicit microsimulation model (Keeling et ai., 2001 - see below). This model provides a good description of the UK 2001 epidemic after the imposition of the national ban on livestock movements (see below), but without having to invoke non-stationary parameter values. One interpretation of the very marked contrast between the performance of these different types of model is that the spatial distribution of farms is a key determinant of the course of a FMD epidemic and that an accurate representation of this distribution is necessary for a model to have any explanatory power. Copyright © 2004 By Horizon Bioscience

Woo/house

369 I

Estimation of FMD transmission rates within farms has been attempted by fitting a SLIR model to data from outbreaks on Saudi Arabian dairy cattle farms (Woolhouse et al., 1996b). This model gave a good fit to the early exponential phases of five different outbreaks, and provided estimates of the per capita transmission rate, (3. However, the (3 estimates for different outbreaks varied by almost two orders of magnitude. This variation may have been real, reflecting different conditions during the outbreaks, but while it remains unexplained it is difficult to use the results of this analysis to make precise predictions about the likely course of future outbreaks.

5. Modelling Control Measures 5.1. Control Options

Methods for controlling FMD epidemics are discussed in detail elsewhere in this volume. They include enhanced surveillance, epidemiological tracing, biosecurity, movement restrictions, stamping out, pre-emptive culling and vaccination. FMD control programmes often involve a combination of some or all of these measures and, even given the inevitable constraints on what is logistically feasible, the number of possible designs of control programmes greatly exceeds the number that can be tried out in practice. Mathematical models are useful tools for exploring the potential impacts of different control programmes. A problem with using models in this way is that not all interventions are readily quantifiable: for example, there are no quantitative studies of the relationship between the implementation of biosecurity measures and a decrease in transmission rates. On the other hand, the impact of stamping out, pre-emptive culling and various uses of vaccination, including prophylactic vaccination in advance of an epidemic and reactive vaccination during an epidemic, can be usefully explored using mathematical models. It should be emphasized that, from an epidemiological perspective, a key aspect of a control programme is its impact on transmission. Interventions which are clearly bad for the individual farm, e.g. pre-emptive culling, may nonetheless be good for the population by preventing onward transmission from that farm. Conversely, interventions which may be better for the individual farm, e.g. vaccination, may be less good for the population because they have a less immediate effect on transmission. Mathematical models can be used to analyse trade-offs of this kind. 5.2. Movement Restrictions

A ban on the movement of livestock between farms affects the transmission kernel (Figure 3), both in terms of its height (overall transmission rate) and shape (the fraction of long distance transmissions) (Keeling et aI., 2001). It has been calculated that, in the context of the UK 2001 epidemic, the introduction of a movement ban reduced R o from just below 3 (though with wide confidence intervals around this estimate) to approximately half that value (Woolhouse et al., 2003; Haydon et al., 2003). This single measure, therefore, would have had an enormous impact of the course of the epidemic. Indeed, a non-parametric analysis based on the results of epidemiological tracing studies suggests that if a national ban on livestock

Copyright © 2004 By Horizon Bioscience

\370

Mathematical Models

Ro

Figure 4. The relationship between the expected final size of an epidemic, l finah and the basic reproduction number, Ro. Ro is related to the net transmission rate and the infectious period. The relationship has the form lfinal=I-(I-lo)exp[-Rolfinatl (modified from Kermack and McKendrick, 1927) and is based on a SLIR model for a well-mixed, homogeneous population, with 10 as the fraction of the population initially infected. Results are shown for 10 ranging from near zero (bottom line) to 5% (top line). If R o«1 then final epidemic size is determined largely by the fraction initially infected, and if Ro» 1 by the size of the susceptible population. However, when ReF 1, small increases in Ro can lead to large increases in final epidemic size. Similarly, when Ro= 1, small increases in 10 can also lead to large increases in final epidemic size.

movements had been imposed in the UK just 2 days earlier then this would have reduced the final epidemic size by an estimated 52% (95% confidence intervals 37-61 %) (Haydon et al., 2003). 5.3. Stamping Out

Stamping out, Le. culling of livestock on farms where infection is detected, acts to increase the rate at which infected farms are removed and so decrease the infectious period. The impact of this on farm-to-farm transmission depends on how infectiousness varies through time, Le. on the dynamics of infection within the farm. Although many different dynamics are possible in principle, the average relationship, at least in the UK in 2001, seems to have been approximately linear, Le. halving the infectious period halves the number of transmissions (Royal Society, 2002). However, the relationship may well vary from farm to farm (depending on initial numbers of livestock infected, the doses they were infected with, how infection was introduced, the livestock species involved and various other factors) and may not apply in all epidemiological settings. Various modelling studies which have explored the impact of stamping out (e.g. Haydon et a/., 1997; Howard and Donnelly, 2000) concluded that more rapid removal of infected livestock has a disproportionate impact because even relatively small decreases in the infectious period can result in very large decreases in the final size of an epidemic (Figure 4). Copyright © 2004 By Horizon Bioscience

Woo/house

3711

5.4. Pre-Emptive Culling There are two main forms of pre-emptive culling: culling farms designated by epidemiological tracings as 'dangerous contacts'; and ring culling around infected farms. The culling of dangerous contacts has rarely been incorporated explicitly in epidemiological models. Its effects can be crudely approximated as a reduction in transmission, but a more realistic representation requires the removal of farms at a given rate, some of which will be infected and some not. Where the latter approach has been adopted (e.g. Keeling et aI., 2001), there has been the difficulty that there are no independent estimates of the specifity of dangerous contact identification, i.e. the fraction actually infected. Ring culling can have two effects: removal of already infected farms (much as culling dangerous contacts); and reducing the local availability of susceptible farms for infection. So, here too, it is therefore necessary to represent within the modelling framework the fraction of culled farms which were and were not infected. There is general theoretical agreement on the potential of ring culling to reduce the size of an epidemic, to reduce the total numbers of livestock lost, and to reduce economic losses, provided that there is sufficiently intense local spread (Durand and Mahul, 2000; Ferguson et al., 2001a, 2001b; Morris et al., 2001; Keeling et al., 2001; Royal Society, 2002). This conclusion stems from the nonlinear relationship between epidemic size and the net transmission rate (Figure 4), which means that the benefits of removing livestock before they transmit infection can outweigh the disadvantages of culling uninfected livestock at the same time. However, if the intensity of local spread varies, e.g. reflecting variation in livestock densities, then ring culling may not be the optimal solution in every affected area (Gamer and Lack, 1995; Durand and Mahul, 2000; Berentsen, 2001; Matthews et al., 2003). 5.5. Reactive Vaccination There are various possible designs of reactive vaccination programmes: ring vaccination around infected farms; barrier vaccination to protect at-risk areas; mass vaccination once FMD is reported in a region; or vaccination targeted at particular livestock or farm types. Modelling the effects of vaccination is less straightforward than modelling culling for the following reasons (Woolhouse and Bundy, 1997; Salt, 1997; Barnett and Carabin, 2002): (i) vaccinated livestock are not protected immediately, the delay ranging from 4 to 10 days or more; (ii) not all vaccinated animals are protected, the fraction ranging from 70% to 95% even for homologous challenge; (iii) vaccination apparently has little or no effect on already infected animals; (iv) vaccination provides protection for a limited period, which can be as little as 6 months. Realistic models of vaccination strategies must incorporate these complications, none of which apply to culling. A wide range of approaches has been used to explore the effects of ring vaccination (although, like ring culling, this is easier to represent in spatially explicit models), including network models (Ahmed et al., 2002), spatial lattices (Muller et al., 2000), discrete-time compartment models (Durand and Mahul, 2000), differential equation models with moment closure (Ferguson et al., 2001a), and microsimulation models (Keeling et al., 2001; Morris et aI., 2001). The effectiveness of ring vaccination is determined by the parameter values used for modelling the effects of the vaccine (such as the fraction protected and the delay to Copyright © 2004 By Horizon Bioscience

\372 Mathematical Models

protection) and the implementation of the vaccination programme (such as the rate at which livestock are vaccinated and any targeting of farms to be vaccinated), the degree of local spread, and the overall transmission rate. An issue of major practical importance is how far and how quickly to vaccinate. Models have been used to answer this question with respect to bringing an epidemic under control, or minimizing the costs of the epidemic (i.e. costs of vaccination versus costs of livestock losses), or achieving the loss of fewer livestock overall than could be achieved by pre-emptive culling. There is general theoretical agreement that, in these terms, ring vaccination is not necessarily preferable to preemptive culling (Garner and Lack, 1995; Durand and Mahul, 2000; Ferguson et al., 2001a; Morris et al., 2001; Keeling et al., 2001). These conclusions stem from the realization that even quite modest limitations (such as 5-10% vaccinated animals left unprotected, or a 4-day delay before vaccinated animals are protected) can allow a much bigger epidemic to develop. The limitations of vaccination can be ameliorated by more intensive or better targeted vaccination programmes (Keeling et ai., 2003). An increasingly favoured possibility (Royal Society, 2002) is to combine culling and vaccination strategies (Keeling et ai., 2003). The optimum balance between these will depend on the resources available and the precise epidemiological situation; this is a challenging problem for policy makers and one which mathematical models are well suited to addressing. 5.6. Prophylactic Vaccination There is already a substantial literature on the use of mathematical models to explore the consequences of prophylactic vaccination programmes against infectious diseases (e.g. Anderson and May, 1991; Woolhouse and Bundy, 1997; Woolhouse et al., 1997b). In various contexts many different vaccination strategies have been considered: vaccination at birth or after the waning of any maternal immunity; vaccination at other ages; vaccination of the whole population at regular intervals, so-called 'pulse' vaccination; and vaccination targeted at high risk individuals or groups. A key general result is that vaccination does not have to be 100% effective in order to provide what is known as 'herd immunity', which simply means that the number or density of susceptible individuals is reduced below a threshold value, equivalent to Ro104 TCIDso have been reported (Anderson et ai., 1979; Dawe et ai., 1994b). Individual animals may retain persistent infection for at least 5 years (Condy et al., 1985) but it is probable that a significant number of animals fail to maintain infection for a prolonged period of time because the proportion of persistently infected animals falls after reaching a peak in the 1-3 year age group (Hedger 1976; Anderson, E.C. and Knowles, N.J., personal communication, 1994). More than one type of SAT virus may be maintained by individual buffalo (Hedger, 1972; Anderson et ai., 1979). This is despite the fact that high levels of antibody to the viruses concerned occur in pharyngeal secretions (Hedger, 1976; Francis and Black, 1983). It was demonstrated that carriers are refractory to re-infection with the same strain of virus (Hedger et ai., 1972). 3. Genetic Diversity of FMDV Derived from Buffalo in Southern Africa SAT-type viruses have been shown to be constantly evolving in buffalo populations (Vosloo et ai., 1996; Bastos et al., 2001; 2003) as would be expected from what is known of the quasispecies dynamics of FMDV (Domingo et al., 2003). Thus in southern Africa, different buffalo populations can be differentiated on the basis of SAT-type viruses recovered from carrier animals representative ofthose populations (Vosloo et al., 2001). Even within the buffalo population of the KNP that numbers less than 27 000 individuals, clear intratypic differences in the genomes of SAT-I, 2 & 3 viruses from different regions of the KNP have been shown (Vosloo et al., 1995; Bastos etal., 2000; Bastos, 2001; Bastos etal., 2001; 2003). Buffalo populations in southern Africa have not been completely free ranging for at least 70 years and have been concentrated mainly in conservancies and game parks, and migratory routes have been disturbed by fences. This separation followed a drastic reduction (approximately 10,OOO-fold) in buffalo numbers following the rinderpest pandemic of 1889-1897 (Rossiter, 1994) which may be the reason why some populations within the southern-most buffalo distributional range are free of FMD (Esterhuysen et al., 1985). The resulting bottleneck for virus spread and recovery of buffalo populations in specific areas may explain the locality-specific distribution of viral topotypes apparent today. High mutation fixation rates (Vasloa Copyright © 2004 By Horizon Bioscience

1388 Natural Habitats in Which FMD Maintained KNPI7/88 (Rietpan)

' - - - - KNP/19/89 (Orpen area)

99

KNP/18195 (Orpen area)

80

KNP/31/95 (Orpen area) SARl1/01 (Mhala district) 100

r------- KNP/160/91 (Orpen area)

,------1

KNP/16/93 (Capricorn) 100

76

KNP/183191 (Crocodilebridge area) 80 " ' - - - - - - - KNP/6/96

(Crocodilebridge area)

KNP/5/91 (Satara) '---------------ZIM/Gn10/91 (Gonarezhou NP)

r - - - - - - - - - KNP/32/92 (Shingwedz; area) KNP/141/91 (Phalaborwa gate area)

100

100

KNP/143/91 (Phalaborwa gate area) ZIMlGn/3/91 (Gonarezhou NP)

100

t------------ZIM/11/02 (Manicaland province) ZIM/10/02 (Masvingo province) 99

r-----ZAM/9/93 (Kafue NP)

r-----I

99

ZAM/10/93 (Kafue NP) 100

ZAM/8/96 (Mulanga) 100

ZAMl10/96 (Mulanga) Joo ZIM/1/88 (Hwange NP)

...---------1

ZIM/4/88 (Hwange NP) ' - - - - - - - - - BOT/18/98 (Nxaraga) 100

ZIM/4/02 (Lupane area) " - - - - - - - - - NAM/286/98 (Mourna, E. Caprivi) ZIM/16/91 (Matusadona NP) 100

" - - - - - - - - - - - ZIM/267/98 (Chizarira) '------------ZIM/34/91 (Urungwe safari area) 80

BOT/13/02 (Francistown district) ZIM/11/01 (Beitbridge area)

100 86

ZIM/1/02 (Beitridge area)

Figure 1. Neighbor-joining tree of SAT-2 foot and mouth disease viruses isolated from buffalo and recent outbreaks in cattle in southern Africa. SARll/0l was isolated from the cattle outbreak in the Mhala district (South Africa) during 2001; ZIM/ll/02, ZIM/I0/02, ZIM/4/02, ZIM/l1/01 and ZIMJl/02 were isolated from the recent outbreaks in Zimbabwe between 2001 and 2002, while BOT/13/02 was isolated in Botswana during 2002. All other isolates in the tree were from buffalo. Topotypes A-C are indicated on the tree.

Copyright © 2004 By Horizon Bioscience

Vos/oo and Thomson

SOUTH AFRICA

3891

~TH

Figure 2. Map of southern Africa indicating the geographical distribution of topotypes A-C for SAT-2 viruses.

et al., 1996) and continuous, independent virus cycling within discrete buffalo populations (Condy et al., 1985), probably accounts for the extensive intratypic variation currently evident. Based on nucleotide sequence analysis of a portion of the viral genomes obtained from buffalo in southern Africa, three independently evolving viral lineages were identified for SAT-l (Bastos et aI., 2001; Vosloo et aI., 2002a). These lineages were shown to originate from three correspondingly separate geographic localities, and 3 different topotypes were proposed, viz. Topotype A: South Africa and southern Zimbabwe, Topotype B: Namibia, Botswana and western Zimbabwe and Topotype C: Zambia, Malawi and northern Zimbabwe (Bastos et al., 2001). For SAT-2 buffalo isolates three topotypes have so far been identified within the southern African region (Figures 1 and 2; Vosloo et aI., 1995; Bastos, 1998; Bastos et ai., 2003). One topotype had the same geographical distribution as SAT1 containing isolates from the KNP in South Africa and Ghonarezhou National Park in south-eastern Zimbabwe (Figure 2; Bastos et ai., 2001). Unlike the SAT-I isolates, the western and northern Zimbabwe SAT-2 isolates could not be divided into two distinct topotypes due to lack of sufficient genetic variation. Conversely, isolates obtained from the border regions of Botswana, Zimbabwe, Zambia, Namibia and Angola fell in two separate topotypes while displaying an overlap in distributional range (Figure 2). A fourth topotype containing a single cattle isolate from Angola was genetically so distinct that a different lineage was assigned (Bastos et al., 2003). Copyright © 2004 By Horizon Bioscience

1390 Natural Habitats in Which FMD Maintained

For all three SAT serotypes it has been demonstrated that the genetic differences between viruses from different topotypes is such that outbreaks should be traceable to specific countries, game parks and even to specific regions within game parks (Bastos, 2001; Bastos et aI., 2001; Vosloo et aI., 2001; 2002a, 2002b; Bastos et aI., 2003). However, some countries such as Bots\vana and Zimbabwe have more than one topotype within their borders and if uncontrolled movement of buffalo occurs, distribution of topotypes will become commingled. The high levels of genetic diversity will most likely be reflected in antigenic differences and it has been shown that for vaccination to be effective, the viruses incorporated into vaccines need to be antigenically related to viruses circulating in the field (Hunter et aI., 1996; Hunter, 1998). Therefore, the uncontrolled movement of buffalo within the southern African region could have severe implication for the control of FMD. Knowledge of the occurrence and distribution of SAT virus lineages and topotypes in buffalo populations has enabled the transmission of SAT viruses from buffalo to other species to be unequivocally proven (Bastos et al., 2000; Bruckner et al., 2002; Vosloo et al., 2002b; Thomson et al., 2003a). Further details are provided below. This confirms the early observations made by J.B. Condy and R.S. Hedger into the association between the occurrence of FMD in cattle and the distribution and behaviour of buffalo harbouring SAT-type viruses (Condy et al., 1969; Hedger et al., 1969; Condy, 1971; Hedger, 1972; Hedger et al., 1972; Condy and Hedger, 1974; Hedger, 1976; Condy, 1979; Hedger and Candy, 1985). 4. Ability of Acutely and Persistently Infected Buffalo to Transmit SAT Viruses to Cohorts Transmission of SAT type viruses between individual buffaloes appears to occur in two ways: (1) contact transmission between acutely infected and susceptible individuals that is likely to account for the majority of infections and (2) occasional transmission between carrier buffalo and susceptible individuals. Although precisely how carriers transmit the infection to susceptible cohorts is unknown (Thomson, 1996), a possibility for which the evidence remains tenuous is sexual transmission (Bastos et al., 1999; Thomson et al., 2003a). However, unless further and more convincing evidence for sexual transmission can be produced it can only be considered an interesting theory. A total of 108 sheath washes and 23 samples taken from the testes of adult male buffalo have been examined so far, and virus was isolated from only one of each type of sample. The paper of Bastos et al. (1999) was based on the isolation from the sheath wash of one animal (3 years of age) after 20 buffalo were examined and no probang was available from that animal. The viruses isolated from the probang and the testis of a later animal were identical on partial VPl nucleotide sequencing (W. Vosloo and C.1. Boshoff, unpublished results). It is not known whether these buffalo from which virus was isolated were in the acute phase of infection or in the carrier state. Due to the ages of the animal (3 and 5 years respectively) one is tempted to believe that they were carriers. No virus was isolated from the reproductive tracts of 25 adult female buffalo. These results indicate that the presence of virus in the reproductive tract of buffalo is a rare occurrence (W. Vosloo, B. Botha and C.1. Boshoff, unpublished results).

Copyright © 2004 By Horizon Bioscience

Vos/ao and Thomson

3911

The close contact that occurs between buffalo in breeding herds, especially young individuals between 6-12 months of age, is likely to ensure efficient transmission of SAT-type viruses bearing in mind the levels of excretion that have been demonstrated in naIve animals infected for the first time (Gainaru et al., 1986). Transmission between carrier buffalo and their cohorts has been shown (Condy and Hedger, 1974) but is presumed to be less efficient than that involving acutely infected animals that excrete large quantities of virus. Generally, these infections are "silent" because buffalo generally suffer few ill effects from infection (see below). Therefore, infection can only be inferred from serology or other laboratorybased investigations. To better understand the complex interactions between breeding herds of buffalo and SAT viruses, the above factors were incorporated into a stochastic model for the behaviour of a single SAT virus in a breeding herd of buffalo numbering up to 500 individuals. Included were variable transmission coefficients for spread of the infection within subgroups of the herd (such as family subgroups) as well as within the herd at large (Thomson et al., 1992). Using values for the transmission coefficients that are plausible based on current understanding, the model predicted that each year new infections in the herd would rise to a peak following loss of maternally derived immunity in the calves and then fall as the number of susceptible individuals decreased. Furthermore, periods of time where no new infections occurred within the herd were predicted resulting in the infection "dying out". The only way in which viral circulation could be re-established was through the introduction of new acutely infected individuals at a time when a new cohort of susceptible young animals had been created by the next batch of calves losing their maternal immunity or by re-activation of the virus by transmission of the virus from a carrier to the new group of susceptible calves (Thomson et al., 1992). It should be remembered, however, that, in reality, transmission of SAT viruses within buffalo herds is more complex than this because such herds usually maintain all three SAT types simultaneously. It is likely therefore that there is a period of time each year, probably covering 2-3 months, when significant numbers of young buffalo in breeding herds will be excreting SAT viruses that pose a threat of infection for other species that may come in close contact with buffalo breeding herds. 4.1 Transmission of SAT Viruses from Buffalo to Other Species

In the KNP outbreaks of FMD among impala Aepyceros melanzpus within the Park are a regular occurrence, although, strangely, other species are rarely affected. This could be due to the low minimum infectious dose (approximately 1 TCID so was shown to be capable of infecting adult impala by the respiratory route - R.G. Bengis and G.R. Thomson, unpublished data) required to infect impala in contrast to cattle and sheep that require 10-25 (Donaldson et ai, 1987; Gibson and Donaldson, 1986). Since 1938 there have been 31 recorded outbreaks of FMD in impala in the KNP, and 23 since routine surveillance was introduced in the mid 1960s. Unlike the situation in buffalo, these outbreaks have been identified by the fact that, usually, at least a proportion of impala develop clear clinical signs of FMD. Eight (26%) of the outbreaks were caused by SAT-I, 15 (48%) by SAT-2, three (10%) by SAT3, and five (16%) were untyped. However, since 1983, nine of the 10 outbreaks identified in impala were caused by SAT-2. Sequence analysis of the SAT-2 viruses Copyright © 2004 By Horizon Bioscience

~

~ en S

70

65.71 58.54

60 50

44.7

CD

> ;:;

40

I

40

38.1 40

40

35.9

I

-iii

o

Q.

.! ta E -c ca ';ft.

30

i 20

10.3 12.5 10

o

0

2.4

o m

~

12.5

12.8

l

f

2.6 2.4

n

1ft

;

0

4.6 0

0

0

g

~~~~-~~~~-~~~~-~~~~-~~~~-~~~

~~#~#~#~#~###~###~#~#~#

Date of sampling Figure 3. Graph indicating the percentage impala with antibodies to SAT-2 sampled in the central region of the Kruger National Park between September 1997 and March 2003.

Copyright © 2004 By Horizon Bioscience

Vosloo and Thomson

3931

involved has shown that these outbreaks were causally distinct (Vosloo et al., 1992; Bastos et al., 2000). Infection and attack rates have varied in outbreaks of FMD in impala that have been studied, with the latter sometimes much lower than the former, indicating that subclinical infection is common (Keet et al., 1996), as has been seen in experimentally infected impala (R. G. Bengis and G. R. Thomson, unpublished data). Serological evidence has indicated that impala in other regions of sub-Saharan Africa, \vhere impala are also abundant, have also been infected but clinical disease has not so far been reported (Anderson et al., 1993). That outbreaks of FMD in wildlife have only been recorded in the KNP could be due to the fact that surveillance in this Park is more thorough than elsewhere. Since 1997 a serological survey has been ongoing in the KNP in 3 defined locations covering the north, centre and south of the Park. Animals have been sampled on a three-monthly basis in each of the regions, involving collection of blood for serum preparation as well as examination for clinical disease from about 40 randomly selected animals for each sampling. Since the start of the survey evidence of at least three "outbreaks" of infection with SAT-2 virus(es) have been detected in the central region. During June-December 1998, clinical disease was observed in the impala sampled together with a gradual increase in the number of animals showing positive titres to SAT-2 which peaked in December of that year (Figure 3). In the other two instances no evidence of clinical disease was found at the time of sampling, but since sampling occurred only once every three months, it is possible that clinical disease was present between samplings. However, had disease occurred it is unlikely that it was severe because hoof-breaks are a common sequel to severe disease and these were not observed. In the most recent of these events that was first detected in March 2002, 59% of the animals sampled were seropositive, compared with sero-positivity values of approximately 40% experienced in past outbreaks (September to December 1998 and March to October 2000; Figure 3). In June 2002 and September 2002, the percentage of positive animals dropped to 36% and 12% respectively, but increased again to 66% in December 2002 (Figure 3). It seems therefore that between March 2002 and March 2003, two epizootics of FMD occurred in the central region. However, because the ages of animals sampled were not controlled some of this variation may have been due to age differences in the impala sampled on successive occasions. Impala in the survey area in the north of the KNP have not shown serological evidence of infection while positive animals have been continuously found in the southern region, but at lower levels than in the central region described above. The variability encountered between the proportion of sero-positive animals at successive samplings in the central region also occurred in the southern region (W. Vosloo, R.G. Bengis and B. Botha, unpublished results). Although this study has not been completed it appears to indicate that infection of impala in some regions of the KNP is more frequent than previously inferred from clinical surveillance and also that antibody responses to natural infection are ephemeral. The serological survey indicated that numbers of positive animals following each outbreak decreased progressively until no or very few sero-positive animals were detected up to the next outbreak (Figure 3; W.Vosloo, R.G. Bengis and B. Botha, unpublished results).

Copyright © 2004 By Horizon Bioscience

1394 Natural Habitats in Which FMD Maintained

FMD in impala appears to occur generally in localities where high densities of this species are present. Also, because impala depend on water, infection frequently has spread along watercourses in the KNP, i.e. it is assumed that the virus is not transmitted by means of the water itself but by contact between animals congregated along rivers and streams. During times of low rainfall both buffalo and impala congregate at watering points, giving rise to opportunities of close contact between the species. In addition to shared drinking water, buffalo and impala both graze three grass species in common (Skinner and Smithers, 1990). Obvious clinical disease in animal species other than impala in the KNP has only rarely been observed (Bengis, 1983; Keet et aI., 1996). It is thought that this is explained by the high susceptibility of impala to infection with SAT viruses and the fact that within the KNP impala are by far the most numerous large mammal. Persistent infection in impala has not been demonstrated (Hedger et al., 1972; Anderson et al., 1975; C. de W. van Vuuren, personal communication, 1997). However, FMD epidemics caused by identical viruses have recurred in impala 6-18 months after the original outbreak (Vosloo et aI., 1992; Keet et al., 1996) indicating that the virus may have been maintained within the impala population. Were that so, the mechanism whereby the viruses survived in interepidemic periods remains to be explained. The alternative explanation is that the same virus has been transmitted on more than one occasion from buffalo to impala in the same vicinity. The available evidence based on genome sequencing of appropriate viruses indicates that impala in the KNP usually, if not always, become infected with SAT viruses derived from buffalo in the vicinity (Keet et al., 1996; Bastos et al., 2000; 2003). It is presumed that this results from direct or indirect contact between resident impala and buffalo breeding herds although there is no certainty on that issue. However, there is a paradox in understanding the transmission of FMDV from buffaloes to impala because while SAT-l viruses appear to circulate more rapidly within buffalo herds in the KNP (see above) and would therefore be expected to be associated more frequently with impala outbreaks, most FMD outbreaks in impala within the KNP, certainly in the last 10 years, have been caused by SAT-2 viruses (Thomson, 1994; Bastos et aI., 2000). A possibility is that impala are, in general, more susceptible to infection with SAT-2 viruses than with SAT-lor 3 but there is no direct evidence for that. Sequence data derived from historic South African FMD outbreaks indicates that there have been occasions when related isolates of SAT viruses were obtained from cattle before they were recovered from impala. This suggests that the impala may have acquired the infections from contact with cattle rather than buffalo (W. Vosloo and A.D.S. Bastos, unpublished results). This was more likely to have occurred in the past than in recent times because only since the early 1970s has the whole KNP been game-fenced and vaccination applied to cattle immediately outside the Park on a biannual basis since the early 1980s. Conversely, as explained above, there is evidence that on occasion SAT viruses have been transmitted first from buffalo to impala and then from impala to cattle (Sutmoller et ai., 2000; Hargreaves et al., in press). In this way impala, and presumably other antelope species as well, can on occasion provide a conduit of infection between buffalo and livestock. It seems therefore that although impala and other antelope

Copyright © 2004 By Horizon Bioscience

Vos/oo and Thomson

3951

species are unable to maintain SAT viruses independently they supply an important intermediary between buffalo and livestock in the context of FMD. Experimentally, most attention has been devoted to explaining how carrier buffalo transmit the infection to cattle with which they come in contact (Hedger, 1970; Condy and Hedger, 1974; Anderson et ai., 1979; Condy et ai., 1985; Hedger and Condy, 1985; Bengis et ai., 1986; Dawe et al., 1994a, 1994b; Vosloo et al., 1996). Unlike the confusion surrounding the possibility that carrier cattle may on rare occasions transmit the infection to cohorts in close contact (Salt, 1993; Thomson, 1996), it has been shown unequivocally that carrier buffalo are able to transmit the infection not only to other buffalo (Condy and Hedger, 1974) but to cattle as well (Dawe et ai., 1994a, 1994b; Vosloo et al., 1996). This was done by showing that the viruses causing FMD in the cattle were identical to those "carried" by buffalo with which they were in contact based on nucleotide sequencing. The mechanism whereby this occurs is still uncertain although it is known that direct contact between carrier buffalo and cattle is required over a period of several weeks or months (Dawe et al., 1994a, 1994b; Vosloo et ai., 1996). Initially, it was speculated that carriers (either cattle or buffalo) may be "activated" by stress and an instance where this was thought to have occurred was reported (Hedger and Condy, 1985). Subsequent experimental work in cattle failed to substantiate this possibility; treatment of carrier cattle with cortico-steroids resulted in the disappearance of the virus from oesophageo-pharyngeal samples (Ilott et ai., 1997). Later the observation that in instances where carrier transmission occurred, male buffalo and female cattle were consistently involved, resulted in the possibility of sexual transmission being mooted (Bastos et ai., 1999). This was rendered plausible by the reported observation by farmers in Zimbabwe that buffalo bulls sometimes mount domestic cows in the field. As explained above, however, the issue of sexual transmission remains to be proven. There is no doubt that contact between buffalo and cattle regularly results in outbreaks of FMD caused by SAT viruses among cattle in southern Africa (Condy, 1979; Thomson, 1994; 1996). This conclusion has recently been reinforced by circumstances relating to outbreaks of FMD in cattle in Swaziland and South Africa. These outbreaks were traced back to cattle that became infected by contact with buffalo (Bruckner et al., 2002; Vosloo et ai., 2002b; Thomson et al., 2003a). 5. The Role of Wildlife Species Other Than Buffalo in the Epidemiology of FMD in Southern Africa Many species of wild animals have been reported as having been infected (Macaulay, 1963; Hedger, 1981) and a wide range of species in southern Africa have been shown to have antibodies to FMDV (Lees May and Condy, 1965; Condy et al., 1969). However, the bulk of evidence suggests that wildlife species other than buffalo play only a minor role in the maintenance and spread of FMDV in southern Africa. Elsewhere in sub-Saharan Africa the information available on this issue is too limited to allow an opinion to be formed. Impala seem to be the most susceptible of these species (see above) in South Africa and are considered an indicator host for the presence of SAT viruses because infection in impala in the past often presaged the occurrence of FMD in livestock (Meeser, 1962). In Zimbabwe, on the other hand, kudu Tregalaphus strepsiceros have more

Copyright © 2004 By Horizon Bioscience

\396 Natural Habitats in Which FMD Maintained

frequently been associated with clinical FMD than impala (C. Foggin, personal communication, 2003). In the Serengeti, where wildebeest are by far the most numerous large mammal species, FMD is infrequently reported although a severe outbreak caused by SAT-2 was recorded in wildebeest in 1999. At least 20% of the migratory herd of wildebeest was affected. These animals showed limping as a result of foot lesions located on the coronary band and interdigitally and in some cases the hooves sloughed off. However, typical mouth lesions were not observed. Deaths in calves occurred although no fresh carcasses for sampling were obtained and the cause of calf mortality could therefore not be proven. It seems that the disease was transmitted from cattle to the wildbeest because FMD in domestic animals was reported in February when the wildebeest were in the southern plains where there is little contact with cattle. As soon they migrated into western Serengeti in June, where they came in contact with livestock, the problem in wildebeest became evident. Surprisingly, however, other antelope species in the vicinity remained apparently unaffected (T. Mlengeya, personal communication, 2003). Essentially all cloven-hoofed animals and Camelidae (i.e. members of the order Artiodactyla) are susceptible to infection with FMDV. FMD infection is dependant on the species and even breed of animals, the strain and dose of virus causing infection and the level of immunity of the animals (Thomson, 1994). The susceptibility of most wildlife species is unknown, as are the levels of virus excretion that occur during infection. With the exception of impala no studies have been done on the minimum infectious dose of FMDV required to initiate infection (see above). The only species in which virus persistence has been unequivocally demonstrated to be important in maintenance and transmission of FMD is the African buffalo (see above). However, persistence of FMDV in the oro-pharynx for more than 4 weeks following infection - the criterion that is usually used to define carriers - probably occurs in many ruminant species (Salt, 1993). Persistent infection developed experimentally in kudu and virus was isolated for up to 140 days after infection (Hedger et al., 1972). Excretion of FMDV for a little over 4 weeks has also been shown in wildebeest (Anderson et aI., 1975) and sable antelope Hippotragus niger (Ferris et al., 1989). In impala attempts to identify carrier animals have failed (see above). FMD infection occurred in wildebeest, hartebeest Alcelaphus buselaphus, eland, gemsbok Oryx gazella and springbok Antidorcas marsupialis in the Kalahari region of Botswana after the widespread aphthisation of cattle during 1957 (Falconer, 1972). However, the infection in the Kalahari did not persist, presumably because African buffalo do not occur in that region and the other species did not maintain the infection. For other antelope species there is simply no information on this issue. It was shown in experimental studies that in kudu and buffalo, among which carriers occur (see above), antibody responses were higher and of longer duration than in impala, warthogs Phacochoerus aethiopicus and bush pigs Potamochoerus spp. in which, like domestic pigs, the carrier state has not been demonstrated (Hedger et aI., 1972). This suggests a relationship between persistent infection and the height and duration of antibody levels in susceptible species.

Copyright © 2004 By Horizon Bioscience

Vosloo and Thomson

3971

The African elephant, Loxodonta africana, (Howell et al., 1973; Hedger and Brooksby, 1976; Bengis et ai., 1984) and hippopotami, Hippopotamus amphibius, do not seem to play a role in the epidemiology of the disease on the African continent and has been described in more detail elsewhere (Thomson et ai., 2003a). Even in sub-Saharan Africa, where wildlife are clearly involved in the maintenance of FMD, livestock sometimes transmit the infection to wildlife rather than vice versa (Hedger, 1976; Thomson et al., 1984; Anderson et al., 1993). 5.1 Clinical Signs of FMD in Wildlife The clinical manifestation of FMD in wildlife has been described elsewhere (Thomson et al., 2001; 2003a). FMD in wildlife varies from evidently inapparent as seems usually to be the case in African buffalo, to acutely lethal as was seen among mountain gazelles Gazella gazella (Shimshony et al., 1986; Shimshony, 1988). Mortality due to FMD infection has also been described in impala (Hedger et al., 1972), blackbuck Antilope cervicapra (Kar et al., 1983), saiga Saiga tatarica (Kindyakov et al., 1972; Khukorov et al., 1974), white tailed-deer Odocoileus virginianus (McVicar et al., 1974), and warthogs (R. G. Bengis, personal communication). FMD infection appears to be sub-clinical in most buffalo based on observations made during culling operations in the KNP (cited by Thomson et al., 2003a). Experimentally, however, typical small lesions appeared, particularly on the feet of most animals (Anderson et al., 1979; Gainaru et ai., 1986). Two instances of clinical disease due to natural infection were reported in captive buffalo kept for unrelated experiments (Young et ai., 1972; Thomson et al., 2003a). In a recent outbreak of FMD in buffalo held in captivity for experimentation on tuberculosis, mature animals developed mouth lesions which caused them to avoid feeding and even caused them to "soak" their mouths in the water troughs but no foot lesions were evident. By the time the mouth lesions were observed a number of animals had lost weight of up to 40kg per animal (L. de Klerk, personal communication). Approximately 2 months after the clinical onset animals started to recover their loss in weight. Less mature animals did not lose weight during the outbreak; their weight gain was merely interrupted for a period of approximately a month. During the outbreak all animals suffered a significant drop in their total lymphocyte counts that lasted 3 months. However, the original lymphocyte ratio relative to the total leukocyte count was not recovered during the 5 months over which the experiment continued (L. de Klerk, personal communication). During an experiment involving buffalo persistently infected with SAT-l where SAT-2 virus was inoculated intra-dermolingually, small interdigital lesions on one or more feet and single small ruptured lesion on the dental pad of 2 out of 6 animals were seen (Hedger et al., 1972). SAT-l was isolated from a mouth lesion. Dawe et ai. (1994b) also noted a single small interdigitallesion and a small tongue lesion in one buffalo that became naturally infected. It is therefore possible that lesions do occur naturally in buffalo, but are of such a limited size and nature that the lesions have little or no obvious effect on the animals concerned as well as healing quickly. For these reasons it is accepted that FMD infection in this species is generally inapparant.

Copyright © 2004 By Horizon Bioscience

\398

Natural Habitats ;n Which FMD Maintained

5.2 Diagnosis of FMD in Wildlife

Clinical diagnosis of FMD in wildlife is more difficult than in domestic animals because most animals need to be immobilised for close examination. The tendency towards development of mild or sub-clinical infections also complicates the diagnosis. For these reasons it is essential that material be sent to suitably equipped laboratories for confirmation of a diagnosis of FMD. The issue of laboratory diagnosis has recently been addressed by Thomson and co-workers (2003a). The use of tests based on the detection of antibodies to the non-structural proteins (NSPs) of FMDV has considerable potential for diagnosis in wildlife, as these animals are not normally vaccinated and presence of antibodies will therefore be an indication of infection. However, the duration of these antibodies in wildlife has not been studied and indications are that for some strains of the SAT type viruses antibodies against the NSPs do not persist for more that 35 days in cattle (W. Vosloo and J.1. Esterhuysen, unpublished data). Furthermore, the tests have not been fully validated for use when animals have been exposed to the SAT type viruses, and in domestic stock, test results have proven difficult to interpret (W. Vosloo and J.J. Esterhuysen, unpublished results). The use of the solid phase competition ELISA (spcELISA) (Mackay et ai., 2001) has recently been accepted by the Office International des Epizooties (DIE) as the preferred test for determining antibodies to the structural proteins of FMDV (Manual of Standards for Diagnostic Tests and Vaccines, 2001). However, this test has only been validated for in cattle, pig and sheep sera (Mackay et al., 2001; Chenard et al., 2003) and there are no reports on the use of this test in wildlife. It is widely recognised that the liquid phase blocking ELISA has poor specificity, especially with sera from some species of wildlife, and therefore tends to produce false-positive results. This is especially the case with giraffe sera tested against SAT-2 (J.1. Esterhuysen, personal communication). Caution should therefore be applied when it comes to the interpretation of the results of serology conducted with unvalidated ELISAs on wildlife sera.

6. Control of FMD in Sub-Saharan Africa Where Wildlife are Involved Control of FMD in sub-Saharan Africa is more difficult than elsewhere because, as explained above, the range of viral diversity is greater than in any other region as well as being largely locality specific (Vosloo et al., 2002a). This renders the selection of viral strains for vaccine production vital for effective management of FMD, a technically involved process. Control of FMD in South America, for instance, is much less constrained by this problem (Barteling, 2002). Involvement of African wildlife, buffalo in particular, in maintaining and disseminating SAT-types in sub-Saharan Africa is a further complication and precludes, for the foreseeable future, the possibility of FMD eradication. Thirdly, "open" and largely uncontrolled borders between many countries in Africa with accompanying traditional transhumance and regional trade movement patterns means that there is little to assist the effective control of FMD. The FMD threat that livestock and livestock product imports from Africa potentially pose to other regions of the world is the most important impediment to African countries obtaining greater access for their livestock and livestock products to intercontinental markets. Consequently, better control of FMD in Africa is vital Copyright © 2004 By Horizon Bioscience

Vos/oo and Thomson

3991

in the long term for rural development in the most disadvantaged region of the world. A complication of better control of FMD in Africa is the desire to preserve Africa's rapidly dwindling wildlife heritage in natural ecosystems that, in any case, are severely threatened by human expansion. The reason for this is that the generally accepted imperative for maintaining and extending integrated wildlife and pastoral livestock ecosystems (i.e. as was traditional for centuries) is incompatible with effective FMD control using presently available technology. Presence of FMD in wildlife areas thus inhibits land-use options aimed at intensification of livestock production (because FMD has its greatest impact in intensive systems), especially those aimed at export markets. This means that FMD and its control, as is increasingly recognized in other regions of the world, have effects beyond the mere control of animal diseases (Anonymous, 2002; Perry and Randolph, 2003). Because wildlife themselves only rarely suffer severe ill effects from infection with FMDV, most control efforts involving wildlife are directed towards preventing wildlife infected with FMDV from transmitting the infection to domestic livestock. This is certainly the situation in southern Africa where control is mainly aimed at protecting livestock that may become infected by direct or indirect contact with buffalo in an effort to maintain beef exports to the European Union (Thomson, 1995). However, during the epidemics of FMD that afflicted Britain, France and the Netherlands in 2001 there was great fear that the virus would infect valuable and rare animals in zoological gardens. Fortunately, that did not occur but it focused attention on what could and should be done to prevent that eventuality (Schaftenaar, 2002). The answer to that question has not been finally resolved and much work needs to be done to determine how effective emergency vaccination would be in protecting various species of wildlife vaccinated against infection with FMDV. Until the late 1960s aphthization (deliberate infection of susceptible livestock involved in outbreaks to ensure that all became infected and develop effective immunity) together with quarantine and movement permit systems were used in southern Africa to control outbreaks of FMD in cattle. However, it was discovered that this practise exposed antelope in the region to large amounts of virus and the antelope, because they were not contained by outbreak cordons and fences, perpetuated the infection (Falconer, 1972). When it became clear that buffalo are reservoirs ofFMDV (Hedger, 1972; Hedger et al., 1972; Hedger, 1976; Condy et ai., 1985; Thomson, 1994; Thomson et al., 2001; 2003a), the approach changed. This has involved separation of domestic livestock from infected buffalo populations, usually by means of fences and vaccination of the cattle in areas adjacent to or contiguous with localities where infected buffalo occur. Movement permits and quarantining of animals that move from high risk areas to lo\ver risk areas has remained in force in most countries. This, combined with regular and frequent inspection (usually at either weekly or fortnightly intervals) of the livestock at risk for evidence of disease, has generally been successful in preventing transmission of FMD from wildlife to livestock (Thomson, 1995). 6.1 Fencing and Control of Translocation In countries such as South Africa, Botswana and Zimbabwe, extensive fencing systems have been developed to prevent contact between infected buffalo and other Copyright © 2004 By Horizon Bioscience

1400

Natural Habitats in Which FMD Maintained

Figure 4. Photograph of the electrified game fence surrounding the circumference of the Kruger National Park.

wildlife species and domestic livestock. These fences often contain steel cables for reinforcement and, because secondarily infected antelope are able to jump over fences able to contain cattle and buffalo, should be over 2m in height (Sutmoller et aI., 2000). Increasingly, double fence lines with a defoliated strip between the t\VO fences that preclude direct contact between animals on either side of the fence-line are being used. These were originally developed in south-eastern Zimbabwe to separate wildlife conservancies from surrounding cattle producing properties. In the initial design, one of the fence lines was at least 1.8 metres in height to prevent antelope from jumping over the fence (Thomson, 1999; Thomson et al., 2003a, Hargreaves et aI., in press). An analysis of the risks posed by such a system to the livestock industry of Zimbabwe was conducted soon after the establishment of three such conservancies (Sutmoller et al., 2000). This showed that there was a risk that, despite the double fence line and the height of the fence, impala and kudu would be able to get out of the conservancies in significant numbers by jumping the perimeter fences. It was therefore recommended that the height of fences needed to be increased to 2.4 metres. Conversely, for a variety of interacting reasons, the risk of air-borne spread across the perimeter fences was found to be low. In South Africa a single electrified fence of 2.4m surrounds the entire KNP, a circumference of approximately 1000 km (Figure 4). In Zimbabwe, an outbreak of SAT-2 occurred in cattle immediately adjacent to a wildlife conservancy in July 1997 where the conservancy perimeter fence was not of sufficient height to prevent antelope from jumping across (Hargreaves et aI., in press). In this instance nucleotide sequencing indicated that the virus causing the outbreak in cattle were related to viruses isolated from one of two groups of buffalo in the conservancy that had been introduced two years previously. Due to

Copyright © 2004 By Horizon Bioscience

Vos/oo and Thomson

4011

the presence of the fence, buffalo/cattle contact was precluded but antelope were previously shown to be able to jump the fence which was effectively 1.8 metres high (Sutmoller et ai., 2000). Impala sampled in the area were serologically positive to SAT-2, indicating that they were most probably the reason for transmission. However, fences have been increasingly criticised as having unacceptably high environmental, social and economic costs and it has been argued that costs outweigh the benefits (Scott Wilson Resource Consultants, 2000). Fences sometimes have blocked migration routes and access of wildlife to water, resulting in ecological disturbance and wildlife mortality (Owen and Owen, 1980; Taylor and Martin, 1987). Conversely, in South Africa it has been shown that through the use of fences in combination with vaccination, FMD can be sufficiently controlled to allow successful commercial farming and access to lucrative export markets. However, when the fence around the KNP was destroyed by flooding in 2000/2001, FMD outbreaks in cattle followed the escape of buffalo from the KNP (Briickner et al., 2002; Vosloo et al., 2002b; Thomson et al., 2003a). The recent outbreaks of FMD in Zimbabwe have similarly followed destruction of fences resulting from civil unrest. In most southern African countries movement of animals susceptible to FMD and their products from endemic areas is strictly controlled through the use of permits irrespective of whether movement is to free areas or within FMDcontrolled areas. In South Africa, for example, animals susceptible to FMD may not be moved from the buffer zone surrounding the KNP where vaccination is practised except for direct slaughter at specific non-export abattoirs. Animals may move from the surveillance zone (that surrounds the buffer zone) only after a 3-week quarantine period and negative serological testing. As indicated above all these movements are strictly controlled using a permit system. In the buffer zone animal inspection for clinical FMD by the provincial veterinary service takes place at weekly intervals while in the surveillance zone inspections occur fortnightly. This level of surveillance is greatly in excess of surveillance generally applied anywhere else in the world but experience in southern Africa has shown that it is essential to detect outbreaks at an early stage and so prevent widespread dissemination before additional control measures are applied. 6.2 Vaccination

Routine vaccination of cattle against FMD in areas adjacent or close to wildlife areas has been practiced in southern Africa since the late 1970s. However, because even the best FMD vaccines are relatively inefficient it has been shown by experience in southern Africa that reliance on vaccines exclusively is dangerous.

The reasons for this are two-fold. Firstly, immunity following primary vaccination is ephemeral (3-4 months only in cattle when the vaccine contains alhydrogel/ saponin as the adjuvant; oil-adjuvanted vaccines may be more effective in this respect). Only when individual cattle have received several inoculations does the level of immunity engendered remain high against challenge with the homologous virus (Barteling, 2002). Thereafter annual vaccination is required to keep levels of immunity at a satisfactory level. The second problem is that of antigenic variation. It is well established that animals recovered from infection with one FMDV type are susceptible to re-infection with any of the other 6 virus types (Thomson, 1994). Copyright © 2004 By Horizon Bioscience

\402

Natural Habitats in Which FMD Maintained

Due to the large number of topotypes identified for most serotypes prevalent in sub-Saharan Africa (Vosloo et al., 1995; Bastos, 1998; 2001; Bastos et al., 2001; 2003), and that different topotypes within a virus type are sometimes antigenically diverse enough to require different vaccine strains to protect against disease caused by the respective topotypes, control through vaccination alone may not be sufficient. Vaccine strains should be carefully chosen for different regions taking the topotypes occurring in that region into consideration. However, there is little incentive for vaccine producers to develop new strains as this is a costly and technically involved process and the number of topotypes is considerable. FMD vaccines are already expensive and few countries in Africa have the resources to pay for development of vaccine strains that are specific for their needs. Vaccine strains with wide antigenic cover are therefore an imperative and ways to promote this line of investigation need to be found. Vaccination of free-living wildlife as part of FMD control is not a viable option due to the logistical and financial difficulties of vaccinating large numbers of widely scattered animals in areas that are frequently inaccessible. However, the use of projectile syringes incorporating marker devices fired from helicopters could be used on a small scale. In an experiment where buffalo, impala, eland and cattle were vaccinated using an inactivated trivalent vaccine with alhydrogeV saponin as the adjuvant, it was found that all animals developed serum neutralising antibody responses of a lower order than those of cattle and there was considerable variation in the responses of individual animals (Hedger et ai., 1980). Therefore, it was recommended that a double dose of vaccine should be used for wildlife. Correlation between serum neutralisation antibody levels and resistance to challenge have not yet been established for wildlife species, but convalescent titres following experimental and contact infection with some strains of virus have been recorded (Anderson et ai., 1975; Hedger et al., 1972). Convalescent titres in buffalo and eland were of a similar order to those of cattle and it could be argued that protective antibody levels are likely to be correspondingly similar. In contrast, convalescent antibody levels of impala have been found to be lower that those of cattle (Hedger et ai., 1980). Oil adjuvanted vaccines have not been tested in wildlife except in a single pilot study that was conducted into the possibility of vaccinating buffalo calves within breeding herds in a wildlife reserve in South Africa with high antigenpayload vaccines (Hunter, P., personal communication, 1997). The results were inconclusive. For domestic animals it has been shown that such adjuvants provide

longer lasting immunity (Barteling, 2002) and the success of improved control of FMD in South America was, among other factors, attributed to the use of oiladjuvanted vaccines (cited by Allende et al., 2003.). Although vaccines have been shown to be effective in protecting domestic livestock from clinical disease caused by FMDV, protection against infection is less efficient resulting, in ruminants, in a proportion of animals that are persistently infected, i.e. so-called carriers (De Leeuw et ai., 1979; Donaldson et al., 1987; Donaldson and Kitching, 1989; Salt et al., 1996). Because of this, trade restrictions are frequently imposed on countries where there is a possibility that such animals exist. The argument that carrier animals transmit FMDV so infrequently as to be epidemiologically insignificant and that this possibility can be safely ignored Copyright © 2004 By Horizon Bioscience

Vas/aD and Thomson

4031

(Barteling, 2002; Sutmoller and Casas, 2002), is possibly dangerous in view of the enormous economic repercussions that could result if such an unlikely possibility came to pass (Muckspreader, 2001). Current technical developments in vaccine preparation (purification of viral antigens to the extent that non-structural viral proteins are excluded from the final vaccine, with appropriate certification) and companion serological tests, will soon enable animals, including wildlife, that have been vaccinated and subsequently infected, to be differentiated from animals that have been vaccinated but not infected. Once that is achieved, vaccination of wildlife to protect them against FMD should not constitute a serious problem in terms of international trade. 6.3 Establishment of Disease-Free Buffalo Herds

A solution to the dilemma posed by the widespread desire to maintain African buffalo within recognized FMD-free areas lies in the establishment of buffalo herds free of FMD and other diseases (Condy and Hedger 1978, Vosloo et al., 2001). Several protocols have been followed to ensure that calves born to infected dams are reared without becoming infected with FMDV. In South Africa, at least, a protocol has been established to ensure that breeding of these calves by private enterprises occurs under strict control of veterinary services. In one such protocol, calves are raised by their infected dams until weaning. They therefore receive buffalo colostral immunity and are removed by the time they reach 5-7 months and colostral immunity has waned significantly. The advantage is that calves are exposed to herd behaviour, but there is risk that the calves may become infected when immunity wanes. In the other protocols the calves are either removed at birth before they ingest colostrum and then fed cattle colostrum or left with their mothers for 48 hours to ingest buffalo colostrum and either hand-reared or put with lactating domestic cows for rearing. The latter 2 methods are expensive and have proven precarious as this part of rearing, described as first-stage quarantine, takes place in the endemic area of South Africa where the risk of infection is high. After a severe testing protocol, calves free of FMDV infection are allowed into the surveillance zone and only allowed into the free zone of the country after a further quarantine period and testing. These animals are subsequently used to breed herds free of FMD. However, the scarcity of so-called FMD-free buffalo has resulted in a large price difference between these animals and similar animals in FMD-infected areas. This provides an obvious incentive for smuggling infected animals and selling them as animals free ofFMD (Vosloo et ai., 2001). Such cases have been proven in the past (Vosloo et al., 2001) and the very technique that is used to ensure that buffalo herds can

be established in FMD-free zones and used for ecotourism and other purposes, now threatens the status of zones recognized as free of FMD in southern Africa. Vigilance on the part of veterinary services is therefore essential to ensure that buffalo are not moved illegally. In South Africa at least, it is now required that all buffalo movements may only occur after the animals have been subjected to serological testing. Novel fields of development, such as antiviral strategies (Domingo et al., 2003), may be utilised in future to eliminate disease from carrier buffalo. Research into the efficacy of antiviral inhibitors and mutagenic agents that target viral replication Copyright © 2004 By Horizon Bioscience

1404 Natural Habitats in Which FMD Maintained

in carrier buffalo specifically, could assist in eliminating virus from animals that could be used to establish infection-free herds. Such treatment would also be valuable should infection-free herds become infected by providing a mechanism for recovery of infection-free status and thus safe-guarding the large investments necessary. This approach, if it could be made effective and FMD in livestock better controlled, holds the potential for ridding FMD from sub-Saharan Africa in the longer term.

7. Acknowledgements Sincere thanks to Ms. Karin Boshoff for drawing the phylogenetic tree in Figure 1 and Ms. Erika Kirkbride for the graphics work done in Figure 2. We are grateful to Dr. N.J. Knowles for supplying the nucleotide sequence of BOT/13/02 used in Figure 1. Dr J.M. Musime, Director of the AU-InterAfrican Bureau for Animal Resources, is thanked for permitting G.R. Thomson to publish this chapter. References Allende, R.M., Mendes da Silva, AJ., and Comparsi Darsie, G. 2003. South American standards for foot-and-mouth disease vaccine quality. In: Foot-and-Mouth Disease: Control Strategies. B. Dodet, and M. Vicari, eds. Symposium Proceedings 2-5 June 2002, Lyon, France. Elsevier; Paris, Amsterdam, New York, Shannon, Tokyo. p. 331336. Anderson, E.C. 2003. Foot-and-mouth disease control programmes: factors that influence the strategy for wildlife. In: Foot-and-Mouth Disease: Control Strategies. B. Dodet, and M. Vicari, eds. Symposium Proceedings 2-5 June 2002, Lyon, France. Elsevier; Paris, Amsterdam, New York, Shannon, Tokyo. p. 145-151. Anderson, E.C. 1980. The role of wildlife in the epidemiology of foot-and-mouth disease in Kenya. In: Wildlife Disease Research and Economic Development. B. Karstad, B. Nestel, and M. Graham, eds. Proceedings from a workshop held at Kabete, Kenya. p. 16-18. Anderson, E.C., Anderson, J., Doughty, W.J., and Drevmo, S. 1975. The pathogenicity of bovine strains of foot-and-mouth disease virus for impala and wildebeest. J. Wildl. Dis. 11: 248-255. Anderson, E.C., Doughty, WJ., Anderson, J., and Paling, R. 1979. The pathogenesis of footand-mouth disease in the African buffalo (Syncerus caffer) and the role of this species in the epidemiology of the disease in Kenya. J. Compo Path. 89: 541-549. Anderson, E.C., Foggin, C., Atkinson, M., Sorenson, KJ., Madekurozva, R.L., and Nqindi, J. 1993. The role of wild animals other than buffalo, in the current epidemiology of foot-and-mouth disease in Zimbabwe. Epid. Infect. Ill: 559-563.

Anonymous. 2002. Dealing with an outbreak: control measures and relevant wider issues. In: Infectious Diseases in Livestock, 9. The Royal Society; London. p 111-129. Barteling, S.J. 2002. Development and perfonnance of inactivated vaccines against foot and mouth disease. Rev. Sci. Tech. DIE. 21: 577-588. Bastos, A.D.S. 1998. Detection and characterisation of foot-and-mouth disease virus in subSaharan Africa. Onderstepoort J. Vet. Res. 65: 37-47. Bastos, A.D.S. 200 I. Molecular epidemiology and diagnosis of SAT-type foot-and-mouth disease in southern Africa. PhD thesis. University of Pretoria, pp 1-148. Bastos, A.D.S., Bertschinger, H.J., Cordel, C., Van Vuuren, C, DE W., Keet, D., Bengis, R.G., Grobler, D.G., and Thomson, G.R. 1999. Possibility of sexual transmission of foot-and-mouth disease from African buffalo to cattle. Vet. Rec. 145: 77-79.

Copyright © 2004 By Horizon Bioscience

Vas/aD and Thomson 4051

Bastos, A.D.S., Haydon, D.T., Forsberg, R., Knowles, N.J., Anderson, E.C., Bengis, R.G., Nel, L.H., and Thomson, G.R. 2001. Genetic heterogeneity of SAT-l type foot-andmouth disease viruses in southern Africa. Arch. Virol. 146: 1537-1551. Bastos, A.D.S, Boshoff, C.I., Keet, D.F., Bengis, R.G., and Thomson, G.R. 2000. Natural transmission of foot-and-mouth disease virus between African buffalo (Syncerus caffer) and impala (Aepyceros melampus) in the Kruger National Park, South Africa. Epid. Infect. 124: 591-598. Bastos, A.D.S., Haydon, D.T., Sangare, 0., Boshoff, C.I., Edrich, J.L., and Thomson, G.R. 2003. The implications of viral diversity within the SAT-2 serotype for control of footand-mouth disease in sub-Saharan Africa. J. Gen. Virol. 84: 1595-1606. Bengis, R.G. 1983. Annual Report 1981-82: State Veterinarian, Skukuza (Kruger National Park). Pretoria, South Africa: Directorate of Veterinary Services, National Department of Agriculture. Bengis, R.G, Hedger, R.S, de Vos, V., and Hurter, L. 1984. The role of the African elephant (Loxodonta africana) in the epidemiology of foot and mouth disease in the Kruger National Park. In: Proceedings of the 13 th World Congress of Diseases in Cattle, World Buiatrics Association. p. 39-44. Bengis, R.G., Thomson, G.R., Hedger, R.S., De Vos, V., and Pini, A. 1986. Foot-and-mouth disease and the African buffalo (Syncerus caffer). 1. Carriers as a source of infection. Onderstepoort J. Vet. Res. 53: 69-73. Brooksby, lB. 1958. The virus offoot-and-mouth disease. Adv. in Virus Res. 5: 1-37. Brooksby, J.B. 1982. Portraits of viruses: Foot-and-mouth disease virus. Intervirology. 18: 1-23. Brooksby, J.B., and Rogers, J. 1957. In: Methods of typing and cultivation of foot-andmouth disease viruses. Project 208 of OEEC, Paris. p. 31. Brown, F. 1986. Foot-and-mouth disease--one of the remaining great plagues. Proc. R. Soc. Lond. B. BioI. Sci. 229( 1256): 215-226. Bruckner, G.K., Vosloo, W., Du Plessis, B.J.A., Kloeck, P.E.L.G., Connoway, L., Ekron, M.D., Weaver D.B., Dickason, C.J., Schreuder F.J., Marais, T., and Mogajane, M.E. 2002. Foot and mouth disease: the experience in South Africa. Rev. Sci. Tech. OlE. 21: 751-764. Bulloch, W. 1927. Foot-and-mouth disease in the 16th century. J. Compo Pathol. Therap. 40: 75-76. Casas-Olascoaga, R. 2003. The history of foot-and-mouth disease control in South America. In: Foot-and-Mouth Disease: Control Strategies. B. Dodet, and M. Vicari eds. Symposium Proceedings 2-5 June 2002, Lyon, France. Elsevier; Paris, Amsterdam, New York, Shannon, Tokyo. p. 55-72. Chenard, G., Miedema, K., Moonen, P., Schrijver, R.S., and Dekker, A. 2003. A solid-phase blocking ELISA for detection of type 0 foot-and-mouth disease virus antibodies suitable for mass serology. J. Virol. Methods. 107: 89-98. Condy, J.B. 1979. A history of foot-and-mouth disease in Rhodesia. Rhodesian Vet. J. 10: 2-10. Condy, J.B. 1971. A study of foot-and-mouth disease in Rhodesian wildlife. F.T.C.V.S. Thesis, Royal College of Veterinary Surgeons, London. Condy, J.B., and Hedger, R.S. 1978. Experiences in the establishment of a herd of foot-andmouth disease free African buffalo (Syncerus caffer). S. Afr. J. Wildl. Res. 8: 87-89. Condy, J.B., and Hedger, R.S. 1974. The survival of foot-and-mouth disease virus in African buffalo with non-transference of infection to domestic cattle. Res. Vet. Sci. 16: 182185. Condy, J.B., Hedger, R.S., Hamblin, C., and Barnett, I.T.R. 1985. The duration of the faotand-mouth disease carrier state in African buffalo (i) in the individual animal and (ii) in a free-living herd. Compo Immunol. Microbiol. Infect. Dis. 8: 259-265. Copyright © 2004 By Horizon Bioscience

1406 Natural Habitats in Which FMD Maintained

Condy, 1.B., Herniman, K.A.1., and Hedger, R.S. 1969. Foot-and-mouth disease in wildlife in Rhodesia and other African territories. A serological survey. 1. Compo Pathol. 79: 27-31. Dawe, P.S., Flanagan, F.O., Madekurozwa, R.L., Sorenson, K.1., Anderson, E.C., Foggin, C.M., Ferris, N.P., and Knowles, N.1. 1994a. Natural transmission of foot-and-mouth disease from African buffalo (Syncerus caffer) to cattle in a wildlife area of Zimbabwe. Vet. Rec. 134: 230-232. Dawe, P.S., Sorenson, K., Ferris, N.P., Barnett, I.T.R., Armstrong, R.M., and Knowles, N.1. 1994b. Experimental transmission of foot-and-mouth disease virus from carrier African buffalo (Syncerus caffer) to cattle in Zimbabwe. Vet. Rec. 134: 211-215. De Leeuw, P.W., Tiesink, 1.W.A., and Van Bekkum, 1.G. 1979. The challenge of vaccinated pigs with foot-and-mouth disease virus. Zentralbl. Veterinaermed., Riehe B 26: p.98109. Diamond, 1. 1998. Guns, germs and steel. A short history of everybody for the last 13,000 years. Vintage UK, Random House. Domingo, E., Escarmis, C., Baranowski, E., Ruiz-larabo, C.M., Carrillo, E., Nunez, 1.1., and Sobrino, F. 2003. Evolution of foot-and-mouth disease virus. Virus Res. 91: 33-46. Donaldson, A.I., Gibson, C.F., Oliver, R., Hamblin, C., and Kitching, R.P. 1987. Infection of cattle by airborne foot-and-mouth disease virus: minimal doses with Oland SAT 2 strains. Res. Vet. Sci., 43: 339-46. Donaldson, A.I., and Kitching, R.P. 1989. Transmission of foot-and-mouth disease by vaccinated cattle following natural challenge. Res. Vet. Sci. 46: 9-14. Esterhuysen, 1.1., Thomson, G.R., Flammand, lR.B., and Bengis, R.G. 1985. Buffalo in the northern Natal Game parks show no serological evidence of infection with foot-andmouth disease virus. Onderstepoort 1. Vet. Res. 52: 63-66. Fairall, N. 1968. The reproductive seasons of some mammals in the Kruger National Park. Zoologica Afr. 3: 189-210. Falconer, 1. 1972. The epizootiology and control of foot-and-mouth disease in Botswana. Vet. Rec. 91: 354-359. Ferris, N.P., Condy, lB., Barnett, I.T.R., and Armstrong, R.M. 1989. Experimental infection of eland (Taurotrages oryX), sable antelope (Ozanna grandicomis) and buffalo (Syncerus caffer) \vith foot-and-mouth disease virus. 1. Compo Path. 101: 307-316. Francis, M.1., and Black, L. 1983. Antibody responses in nasal fluids and serum following foot-and-mouth disease infection or vaccination. 1. Hyg. 91: 329-334. Funston, P.l., Skinner, J.D., and Dott, D.M. 1994. Seasonal variation in movement patterns, home range and habitat selection of buffaloes in a semi-arid habitat. Afr. 1. Ecol. 32: 100-14. Gainaru, M.D., Thomson, G.R., Bengis, R.G., Esterhuysen, 1.1., Bruce, W., and Pini, A. 1986. Foot-and-mouth disease and the African buffalo (Syncerus caffer). II. Virus excretion and transmission during acute infection. Onderstepoort J. Vet. Res. 53: 75-85. Gibson, C.F., and Donaldson, A.1. 1986. Exposure of sheep to natural aerosols of foot-andmouth disease virus. Res. Vet. Sci. 41: 45-49. Hanotte, 0., Bradley, D.G., Ochieng, 1.W., Verjee, Y., Hill, E.W., and Rege, 1.E.0. 2002. African pasturalism: Genetic imprints of origins and migrations. Science. 296: 336339. Hargreaves, S.K., Foggin, C.M., Anderson, E.C., Bastos, A.D.S., Thomson, G.R., Ferris, N., and Knowles, N. An investigation into the source and spread of foot and mouth disease virus from a wildlife conservancy in Zimbabwe. Rev. Sci. Tech. DIE. In press. Hedger, R.S. 1972. Foot-and-mouth disease and the African buffalo (Syncerus caffer). 1. Compo Path. 82: 19-28. Hedger, R.S. 1976. Foot-and-mouth disease in wildlife with particular reference to the African buffalo (Syncerus caffer). In: Wildlife Diseases. L.A. Page, ed, New York, Plenum Publishing. p. 235-244. Copyright © 2004 By Horizon Bioscience

Vas/aD and Thomson

4071

Hedger, R.S. 1981. Foot-and-mouth disease. In: Infectious Diseases of Wild Mammals, 2nd Edition. J.W. Davis, L.H. Karstad and D.O. Trainer, eds. Iowa State University Press. p.87-96. Hedger, R.S. 1970. Observations on the carrier state and related antibody titres during an outbreak of foot-and-mouth disease. J. Hyg. 68: 53-60. Hedger, R.S., and Brooksby, J.B. 1976. FMD in an Indian elephant. Vet. Rec. 99: 93. Hedger, R.S., and Condy, lB. 1985. Transmission of foot-and-mouth disease from African buffalo virus carriers to bovines. Vet. Rec. 117: 205. Hedger, R.S., Condy, J.B., and Falconer, J. 1969. The isolation of foot-and-mouth disease virus from African buffalo (Syncerus caffer). Vet. Rec. 84: 516-517. Hedger, R.S., Condy, J.B., and Golding, S.M. 1972. Infection of some species of African wildlife with foot-and-mouth disease virus. J. Compo Pathol. 82: 455-46. Hedger, R.S., Condy, J.B., and Gradwell, D.V. 1980. The response of some African wildlife species to foot-and-mouth disease vaccination. J. Wildl. Dis. 16: 431-438. Hedger, R.S., Forman, AJ., and Woodford, M.H. 1973. Foot-and-mouth disease in East African buffalo. Bull. Epizoot. Dis. Afr. 21: 99-101. Henning, M.W. 1956. Animal diseases in South Africa. 3rd edition, Pretoria: Central News Agency. Howell, P.G., Young, E., and Hedger, R.S. 1973. Foot-and-mouth disease in the African elephant (Loxodonta africana). Onderstepoort J. Vet. Res. 40: 41-52. Hunter, P. 1998. Vaccination as a means of control of foot-and-mouth disease in sub-Saharan Africa. Vacc. 16: 261-264. Hunter, P., Bastos, A.D., Esterhuysen, J.1., and van Vuuren, C. de WJ. 1996. Appropriate foot-and-mouth disease vaccines for southern Africa. All Africa Conference on Animal Agriculture, Pretoria, South Africa. 2.2.7: 1-4. 110tt, M.C., Salt, J.S., Gaskell, R.M., and Kitching, R.P. 1997. Dexamethasone inhibits virus production and the secretory IgA response in the oesophageal-pharyngeal fluid in cattle persistently infected with foot-and-mouth disease virus. Epidemiol. Infect. 118: 181-187. Kar, B.C., Hota, N., and Acharjyo, L.N. 1983. Occurrence of foot-and-mouth disease among some wild ungulates in captivity. Indian Vet. J. 60: 237-239. Keet, D.F., Hunter, P., Bengis, R.G., Bastos, A.D., and Thomson, G.R. 1996. The 1992 foot-and-mouth disease epizootic in the Kruger National Park. J. S. Afr. Vet. Assoc. 67: 83-87. Kindyakov, V.I., Nagumanov, F.M., and Tasbulatov, E.S. 1972. The epizootiological significance of contact between wild and domestic animals in relation to foot-andmouth disease (in Russian). Voprosy Prirodnoi Ochagovosti Boleznei. 5: 63-66. Khukorov, V.M., Pronina, N.A., Korsun, L.N., Karpenko, I.G., and Kruglikova, B.A. 1974. Foot and mouth disease in saiga antelopes [in Russian]. Veterinariya (Moscow) 5: 6061. Knowles, N. J., Samuel, A.R. 2003. Molecular epidemiology of foot-and-mouth disease virus. Virus Res. 91: 65-80. Lees May, T., and Condy, J. 1965. Foot-and-mouth disease in game in Rhodesia. Bull. Off. Int. Epizoot. 64: 805-811. Macaulay, J.W. 1963. Foot-and-mouth disease in non-domestic animals. Bull. Epizoot. Dis. Afr. 11: 143-146. Mackay, D.K., Bulut, A.N., Rendle, T., Davidson, F. and Ferris, N.P. 2001. A solid-phase competition ELISA for measuring antibody to foot-and-mouth disease virus. J. Virol. Methods. 97: 33-48. McVicar, J.W., Sutmoller, P., Ferris, D.H., and Campbell, C.H. 1974. Foot-and-mouth disease in white-tailed deer: clinical signs and transmission in the laboratory. Proc. Annu. Meet. US. Anim. Health. Assoc. 78: 169-180.

Copyright © 2004 By Horizon Bioscience

1408 Natural Habitats in Which FMD Maintained

Meeser, J.N. 1962. Foot-and-mouth disease in game animals with special reference to the impala (Aepyceros melampus). J. S. Afr. Vet. Med. Assoc. 33: 351-5. Mloszewski, MJ. 1983. The behaviour and ecology of the African buffalo. 15t ed. Cambridge: Cambridge University Press. p. 63-137. Muckspreader, 2001. Not The Foot and Mouth Report. Special Investigation. Private Eye November 2001. p. 3-31. Owen, M., and Owen, D. 1980. The fences of death. Afr. Wildl. 31: 25-27. Paling, R.W., Jessett, D.M., and Heath, B.R. 1979. The occurrence of infectious diseases in mixed farming of domesticated wild herbivores and domestic herbivores, including camels, in Kenya. I. Viral diseases: a serologic survey with special reference to footand-mouth disease. J. Wildl. Dis. 15: 351-8. Palmenberg,A. C. 1989. Sequence alignments of Picornaviral capsid proteins. In: Molecular Aspects of Picornavirus Infection and Detection. B.L. Semler and E. Ehrenfeld, eds. Washington D.C.: American Society for Microbiology. p. 211-241. Perry, B.D., and Randolph, T.F. 2003. The economics of foot and mouth disease, its control and its eradication. In: Foot-and-Mouth Disease: Control Strategies. Dodet, B., and Vicari, M. eds. Symposium Proceedings 2-5 June 2002, Lyon, France. Elsevier; Paris, Amsterdam, New York, Shannon, Tokyo. p. 23-41. Pienaar, U. de V. 1969. Observations on developmental biology, growth and some aspects of the population ecology of African buffalo in the Kruger National Park. Koedoe. 12: 29-52. Robson, KJ.H., Harris, TJ.R., and Brown, F. 1977. An assessment by competition hybridisation of the sequence homology between the RNAs of the seven serotypes of FMDV. J. Gen. Virol. 37: 271-276. Rossiter, PB. 1994. Rinderpest. In: Infectious Diseases of Livestock with special reference to southern Africa. J.A.W. Coetzer, G.R. Thomson and R.C. Tustin, eds. Oxford University Press, Cape Town London New York. p. 735-757. Salt, J.S. 1993. The carrier state of foot-and-mouth disease - an immunological review. Br. Vet. J. 149: 207-223. Salt, J.S., Samuel, A.R., and Kitching, R. 1996. Antigenic analysis of type 0 foot-and-mouth disease virus in persistent!y infected bovine. Arch. Virol. 141: 1407-1421. Schaftenaar, W. 2002. Use of vaccination against foot and mouth disease in zoo animals, endangered species and exceptionally valuable animals. Rev. Sci. Tech. OlE. 21: 613620. Scott Wilson Resource Consultants/Environmental Development Group, 2000. Final Report: Environmental Assessment of Veterinary Fences in Ngamiland. Vols. 1-5. Edinburgh, UK: 23 Chester Street, EH3 7ET. Shimsony, A. 1988. Foot and mouth disease in the mountain gazelle in Israel. Rev. Sci. Tech. OlE. 7: 917-923. Shimshony, A., Orgad, D., Baharav, D., Prudovsky, S., Yakobson, B., Bar Moshe, B., and Dagan, D. 1986. Malignant foot-and-mouth disease in mountain gazelles. Vet. Rec. 119: 175-176. Sinclair, A.R.E. 1977. African buffalo. A study of resource limitation of populations. The University of Chicago Press: Chicago. Skinner, J.D., and Smithers, R.H.N. 1990. The mammals of the southern African subregion. 2nd ed. Pretoria: University of Pretoria. p. 666-686. Sutmoller, P., and Casas, O.R. 2002. Unapparent foot and mouth disease infection (subclinical infections and carriers): implications for control. Rev. Sci. Tech. 21: 519-529. Sutmoller, P., Thomson, G.R., Hargreaves, S.K., Foggin, C.M., and Anderson, E.C. 2000. The foot-and-mouth disease risk posed by African buffalo within wildlife conservancies to the cattle industry of Zimbabwe. Prevo Vet. Med. 44: 43-60. Taylor, R.D., and Martin, R.B. 1987. Effects of veterinary fences on wildlife conservation in Zimbabwe. Environ. Manage. 11: 327-334. Copyright © 2004 By Horizon Bioscience

Vos/oo and Thomson

409 I

Thomson, G.R. 1999. Alternatives for controlling animal diseases resulting from interaction between livestock and wildlife in southern Africa. S. Afr. J. Sci. 95: 71-76. Thomson, G.R. 1994. Foot-and-mouth disease, In: Infectious Diseases of Livestock with Special Reference to Southern Africa. J.A.W. Coetzer, G.R. Thomson and R.C. Tustin, eds. Cape Town, London, New York: Oxford University Press. p. 825-952. Thomson, G.R. 1995. Overview of foot and mouth disease in southern Africa. Rev. Sci. Off. Int. Epiz. 14: 503-520. Thomson, G.R. 1996. The role of carrier animals in the transmission of foot and mouth disease. DIE Comprehensive Reports on Technical Items Presented to the International Committee or to Regional Commissions. p. 87-103. Thomson, G.R., Bengis, R.G., and Brown, CC. 2001. Picornaviruses. In: Infectious Diseases of Wild Mammals, 3rd Edn. E.S. Williams and I.K. Barker, eds. Iowa University Press, Ames. pp.l19-130. Thomson, G.R., Bengis, R.G., Esterhuysen, J.J and Pini, A. 1984. Maintenance mechanisms for foot-and-mouth disease virus in the Kruger National Park and potential avenues for its escape into domestic animal populations. Proceedings of the XIIIth World Congress on Diseases of Cattle. Vol. I., September 1984. Durban, South Africa. Thomson, G.R., Vosloo, W., and Bastos, A.D.S. 2003a. Foot and mouth disease in wildlife. Virus Res. 91: 145-61. Thomson, G.R, Vosloo, W., and Bastos A.D.S. 2003b. The epidemiology and control of foot-and-mouth disease in sub-Saharan Africa. In: Foot-and-Mouth Disease: Control Strategies. B. Dodet, and M. Vicari, eds. Symposium Proceedings 2-5 June 2002, Lyon, France. Elsevier; Paris, Amsterdam, New York, Shannon, Tokyo. p. 125-134. Thomson, G.R., Vosloo, W., Esterhuysen, J.1. and Bengis, R.G., 1992. Maintenance of footand-mouth disease viruses in buffalo (Syncerus caffer Sparrman, 1779) in southern Africa. Rev. Sci. Tech. DIE. 11: 1097-1107. Valee, H., and Carre, H. 1922. Sur la pluralite du virus aphteux. C. r. hebd. Seanc. Acad. Sci., Paris. 174: 1489. Vosloo, W., Bastos, A.D., Kirkbride, E., Esterhuysen, J.1., Janse van Rensburg, D., Bengis, R.G., Keet, D.F., and Thomson, G.R. 1996. Persistent infection of African buffalo (Syncerus caffer) with SAT-type foot-and-mouth disease viruses: rate of fixation of mutations, antigenic change and interspecies transmission. J. Gen. Virol. 77: 14571467. Vosloo, W., Bastos A.D.S., Michel, A., and Thomson, G.R. 2001. Tracing movement of African buffalo in southern Africa. Rev. Sci. Tech. DIE. 20: 630-639. Vosloo, W., Bastos, A.D.S., Sangare, 0., Hargreaves, S.K. and Thomson, G.R. 2002(a). Review of the status and control of foot and mouth disease in sub-Saharan Africa. Rev. Sci. Tech. DIE. 21: 437-449. Vosloo, W., Boshoff, K., Dwarka, R., and Bastos, A. 2002(b). The possible role that buffalo played in the recent outbreaks of foot-and-mouth disease in South Africa. Ann. N. Y. Acad. Sci. 969: 187-90. Vosloo, W., Kirkbride, E., Bengis, R.G., Keet, D.F., and Thomson, G.R. 1995. Genome variation in the SAT types of foot-and-mouth disease viruses prevalent in buffalo (Syncerus caffer) in the Kruger National Park and other regions of southern Africa, 1986-1993. Epidem. Infect. 114: 203-218. Vosloo, W., Knowles, N.J., and Thomson, G.R. 1992. Genetic relationships between southern African SAT-2 isolates of foot-and-mouth disease virus. Epidem. Infect. 109: 547-558. Waldmann, 0., and Trautwein, K. 1926. Die aktive Immunisiering des Rindes gegen Maulund Klauenseuche mittles Formolimpstoff. Zentbl. Bakt. Abt. I., orig. 138: 401. Young, E., Hedger, R.S., and Howell, P.G. 1972. Clinical foot-and-mouth disease in the African buffalo (Syncerus caffer). Onderstepoort J. Vet. Res. 39: 181-184.

Copyright © 2004 By Horizon Bioscience

1410

Natural Habitats in Which FMD Maintained

Zhang, Z.D., and Kitching, R.P. 2001. The localization of persistent foot and mouth disease virus in the epithelial cells of the soft palate and pharynx. J. Compo Path. 124: 89-94.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 15 Diagnosis and Control of Foot-and-Mouth Disease R.P. Kitching Abstract The recent outbreaks of foot-and-mouth disease in Europe have been a stimulus to bring to the market new diagnostic tests which can be used on the farm or in the laboratory to increase the speed with which suspect outbreaks can be confirmed. A more rapid diagnosis, particularly in species such as sheep that often show only mild clinical signs, would likely reduce the number of animals slaughtered during an outbreak, and allow a more accurate definition of the size and distribution of an outbreak. However, before these new tests are introduced, they must be rigorously tested under different conditions to clearly document their sensitivity and specificity, as, if they are to supplement or replace existing methods, they must provide advantages to a disease control program. The large scale slaughter that took place in the United Kingdom during the outbreak horrified the general public, and has prompted the political leaders to insist on a re-assessment of the use of vaccination as an alternative strategy. But political will is not sufficient to overcome all the problems associated with the use of vaccine in the face of an outbreak of FMD. 1. Introduction Initial SUspIcIon of foot-and-mouth disease (FMD) is usually following the appearance of clinical signs consistent with FMD in a susceptible species. Once the disease has become established, the clinical signs seen in totally naive cattle or pigs are almost pathognomonic, but the first animal in a herd to be affected may only have received a low dose of virus and show only mild clinical signs ofFMD, which could easily be missed or confused with another disease. FMD in adult sheep and goats is also usually mild, and the policy adopted during the 2001 UK outbreak of FMD which required veterinarians to make a clinical diagnosis of disease so that affected flocks could be immediately slaughtered, proved difficult to implement (De la Rua et al., 2001). The situation in the UK outbreak in \vhich sheep spread the disease extensively before its presence was realized is not unusual, for instance sheep took FMD into Tunisia in 1989, and although it was noticed that they were lame, the initial diagnosis had been bluetongue; not until the virus transmitted to cattle was it recognized as FMD, by which time the disease had already spread to Algeria and Morocco. The 1994 outbreak of FMD in Greece was carried by sheep from Turkey, initially onto the Island of Lesbos, and then into mainland Greece, and was not recognized for over two months (Kitching, 1998). In the very young of susceptible species, in particular in lambs and piglets, FMD is characterized by Copyright © 2004 By Horizon Bioscience

\412 Diagnosis and Control

sudden death, and although a lactating sow or ewe would likely have vesicles on the udder, the disease can be missed if only the dead young stock are examined. The 1981 outbreak of FMD in Brittany, France, was initially mis-diagnosed because the presenting signs were dead piglets, and the outbreak spread amongst the pig herds and eventually, as airborne virus to Jersey and the Isle of Wight, off the southern coast of England. Even in new outbreaks, in fully susceptible animals, FMD can sometimes fail to appear in its typical form. In 1999 the PanAsia strain of serotype 0 was isolated from probang samples collected from Chinese Yellow cattle on Kinmen Island, Taipei China; they were being routinely sampled prior to movement and were not showing any signs of clinical FMD (Huang et al., 2001). A major outbreak of FMD in Zimbabwe in 1989 was initiated when Brahman cattle carried the virus into a national agricultural show being held in the Bulawayo show ground; they were only exhibiting mild lameness, and the possibility that they had FMD was not suspected. All countries that have a control policy for FMD rely on a routine surveillance program to initially identify the first appearance of disease, and while countries that are trying to establish an internationally recognized free status to the satisfaction of potential trading partners will include in their surveillance a statistically valid random sampling regime to identify subclinical infections in their livestock population, once recognition is achieved, this will be replaced by passive surveillance, reliant on being aware of the first case of FMD. This is easier in a non-vaccinated population than in a population in which prophylactic vaccination is used, as vaccination will suppress clinical disease, but not necessarily prevent circulation of FMD virus. The declaration of an outbreak of FMD in a previously free country is likely to have major domestic and international implications, depending on the significance of the export trade in FMD susceptible animals and their products. The immediate consequences of the 1997 outbreak of FMD in Taiwan China, was an estimated cost of $400,000 for the initial control and eradication programme, and a loss of $3,600,000 in export trade (Yang et al., 1999). In order to prevent any misunderstanding as to what constitutes an outbreak, to avoid an exporting country continuing to trade and thereby put at risk its trading partners, a definition has been included in the Office International des Epizooties (OlE) Code, which provides guidelines for international trade agreements relating to live animals and animal products (DIE, 2002).

2. Definition of an Outbreak In the 200 1edition ofthe FMD Code (2001 a), a Member Country that was recognized as an FMD free country where vaccination is not practised, was required to send a declaration to the OlE that there had been no outbreak of FMD and no vaccination had been carried out for at least 12 months, together with documented evidence of an effective import and surveillance programme. An outbreak was defined as the occurrence of clinical signs of FMD, and provided an obvious loophole, such that a country could maintain its free status even when there was evidence of live FMD virus being present, in the absence of reported disease. While this was not in the spirit of what was intended, it was legally defensible. To correct this anomaly, the Copyright © 2004 By Horizon Bioscience

Kitching

4131

new Code chapter requires that Member Countries declare that there has also been no evidence of infection found during the past 12 months. This change necessarily required a definition of FMD virus infection. The following defines the occurrence of FMD virus infection: 1. 2.

3.

FMD virus has been isolated and identified as such from an animal (free-living or domestic), or a product derived from that animal, or Viral antigen or viral RNA specific to one or more of the serotypes of FMD virus has been identified in samples from one or more animals (free-living or domestic) showing clinical signs consistent with FMD, or epidemiologically linked to a confirmed or suspected outbreak of FMD, or giving cause for suspicion of previous association or contact with FMD virus, or Antibodies to structural or nonstructural proteins of FMD virus that are not a consequence of vaccination, have ben identified in one or more animals (free-living or domestic), with either epidemiological links to a confirmed or suspected outbreak of FMD, or showing clinical signs consistent with recent infection with FMD virus.

The most conclusive evidence of infection is isolation of live FMD virus, but there will be situations in which this is not possible, particularly if good laboratory support is not immediately available. The tests for FMD virus antigen (see below), are \vell validated, and when properly carried out are both sufficiently specific and sensitive to be used in a fresh outbreak in a naive population. The use of polymerase chain reaction (PCR) technology has in the past been associated with problems of cross contamination due to its extreme sensitivity, and it is currently unlikely that a declaration of an outbreak of FMD would be made on the basis of a positive PCR result alone. Thus it was considered important to link it with clinical signs of FMD or a link to a known or suspected outbreak or contact with FMD virus. The presence of antibodies to FMD virus can also lead to suspicion of disease, but no antibody test for FMD virus antigens is totally specific, as can be confirmed by anyone involved in routine import/export testing. In addition, an animal vaccinated against FMD will be positive by the virus neutralization test, liquid phase blocking ELISA, and the recently introduced solid phase competition ELISA (see below). However, vaccinated animals will not be positive to the tests for antibodies to the non-structural proteins (NSP) of FMD virus which are produced in response to live replicating virus (see below). But even a positive NSP antibody test would not in isolation be considered sufficient evidence for a declaration of infection, and there is a requirement for a link with evidence of disease or link to a confirmed or suspected outbreak. A derogation of this requirement for freedom from antibodies consequent to infection would be necessary for those countries seeking re-establishment of freedom of FMD status, three months or longer after an outbreak. It is possible that antibody positive animals would be identified in the post-outbreak surveillance required for a submission to the OlE, but, assuming that these animals are found negative for evidence of live virus, virus antigen or virus genome, and immediately slaughtered, it would not be reasonable to wait an additional three months to be declared free. Similarly, FMD free countries

Copyright © 2004 By Horizon Bioscience

1414 Diagnosis and Control

that used vaccination to control an outbreak, but did not subsequently slaughter the vaccinates could seek a restoration of status 6 months after the last case or use of vaccine; these countries would clearly have many animals positive for structural antibodies as a consequence of vaccination, and possibly some with NSP antibodies, but if there was no evidence of persistence of live virus, they could be declared free of FMD infection. 3. Laboratory Diagnosis of FMD The success of the laboratory confirmation of a presumptive diagnosis of FMD depends on the submission of adequate material, sent under suitable conditions. A minimum of 2 sq. cm. of epithelium from a ruptured vesicle in a 50/50 mixture of glycerine and 0.04 molar buffered phosphate (pH 7.4-7.6) should be sent a laboratory designated for handling live FMD virus and equipped with reagents required to type a positive sample. Whole and clotted blood samples (particularly from small ruminants from which it is often difficult to collect epithelium samples) and probang samples may also be sent. Heart muscle from dead young stock suspected of having died from FMD may also be a useful source of virus for laboratory isolation. In the laboratory the tissue samples are prepared as a 10% suspension for antigen detection ELISA (DIE, 2001 b) or used directly in a polymerase chain reaction (PCR) (Reid et al., 2000) to serotype the virus. Sensitive tissue culture, such as primary bovine thyroid or lamb kidney cells is inoculated with the tissue suspension, and/or the whole blood and serum, to grow the virus for further characterisation, and to amplify the antigen if insufficient was present in the original sample to provide an initial diagnosis by ELISA. Some strains of FMD virus such as those belonging to the Cathay topotype (Samuel and Knowles, 2001), are very pig specific, and grow only poorly if at all on cells of ruminant origin; for these strains it is essential to also inoculate pig kidney cells, such as IB-RS-2, for virus isolation. The antigenic characteristics of the strain are compared with existing vaccine strains in order to identify a suitable vaccine if one is required, or to confirm the use of one that is already helping to control the outbreak (Kitching et al., 1988). A segment of the viral genome (part of the ID gene) can also be sequenced and compared in the reference laboratory database to determine its relationship to other viruses circulating in the region, which may give an indication of its origin (Kitching et al., 1989). Laboratory confirmation of FMD in pigs is essential because of possible confusion with other diseases giving a similar clinical appearance, such as vesicular stomatitis (VS), swine vesicular disease (SVD) and vesicular exanthema (VE). Both VS and VE have only been reported in the Americas - apart from three incursions ofVS virus into France (1915) and South Africa (1884 and 1887). SVD is present in East Asia and Italy, but has in the past been more widespread, and should always be considered as a differential diagnosis for FMD in pigs (Lin and Kitching, 2000). Therefore, SVD virus reagents should always be included in the antigen detection ELISA (DIE, 2001b). The gold standard test for detecting antibodies to FMD virus antigens is the virus neutralization test (VNT), but this requires the use of live FMD virus, and can be a difficult test to carry out unless in regular use. The VNT (DIE, 2001 b) has largely been replaced for routine serology by the liquid phase blocking ELISA (LPBE) and more recently by the solid phase competition ELISA (SPCE), which Copyright © 2004 By Horizon Bioscience

Kitching

415

uses the same reagents as the LPBE (OlE, 2001 b), but is more specific and equally as sensitive as the LPBE, and thus gives fewer false positive results (MacKay et aI., 2001). Antibodies to FMD virus can be detected in milk of cattle that have recovered from FMD, using either the LPBE or a specific isotype assay (SIA) for bovine IgG 1 (Armstrong, 1997). But whereas the LPBE would not detect antibodies derived as a consequence of vaccination, the SIA was able to identify 95% of cattle vaccinated up to 12 months previously, in the study reported. There was also a strong correlation between serum antibody titres and milk antibody titres (Armstrong et al., 2000), to the extent that individual and herd immunity levels against FMD could be assessed using the SIA on individual or bulk tank milk samples respectively (Armstrong and Mathew, 2001). Once the sample is received by a competent diagnostic laboratory, and assuming the sample was properly collected and stored during transport, the available tests are extremely sensitive and specific, and for positive samples the result can usually be available in a few hours. However, before the laboratory can report that no virus was detected, it is necessary to passage the sample on tissue culture, and then blind passage it a further one or two times, each passage requiring 48 hours. Thus it may be a week after receipt of a sample that a final report is issued from the laboratory, which can and frequently does cause some frustration with the field staff, as quarantine and movement restrictions would probably have been kept in place until a result was available. However, there are many examples where preliminary negative results have been issued and restrictions prematurely removed, and then the final result has been reported positive. 4. Persistent Infection Ruminant animals that have recovered from infection with FMD virus, and vaccinated ruminants that have had contact with live virus may retain infection in their pharyngeal region for a variable period of time. The carrier is defined as an animal from which live virus can be recovered after 28 days following infection (Salt, 1993). This is not an exceptional situation, and over 50% of ruminants exposed to live FMD virus become carriers; pigs do not become carriers (Donaldson, 1987). The duration of the carrier state depends on the species and individual; the African buffalo may carry virus for over 5 years, cattle for over 3 years, sheep for up to 9 months, goats and wild ruminants for shorter periods of time, and for South American camelids no carrier state exists (David et aI., 1993). Eventually the carrier does eliminate the virus. In cattle, the virus persists in the basal layer cells of the pharyngeal epithelium, particularly of the dorsal soft palate (Zhang and Kitching, 2001), while in sheep the virus persists in the tonsil. It is not detectable by existing methods in the more superficial layers of cells, and it is not clear how the virus is excreted into the pharynx. Nor is it clear how the virus changes from a lytic agent which destroys the host cell, into one that can establish a persistent infection. It is possible that a mutation reduces the ability of the persistent virus to shut down host cell metabolism, and it may be further speculated that eventually a further back mutation restores its lytic action and the virus is ultimately eliminated; but this remains to be proven.

Copyright © 2004 By Horizon Bioscience

1416

Diagnosis and Control

The establishment of the carrier state and its duration depends on the host species, but it probably also depends on the strain and serotype of FMD virus and even on the breed of host species. All three serotypes of the SAT viruses are found in the wild African buffalo populations of Botswana and Zimbabwe, but rarely are the commercially farmed Zebu cattle of the region found to be carrying either SAT 1 or SAT 3, and outbreaks due to these serotypes are only occasionally extensive. However, outbreaks due to SAT 2 are considerably more common and one Zebu bull in particular, in Zimbabwe, remained a carrier of SAT 2 virus for over 3 years. Also during the 1991 outbreak of SAT 2 in Zimbabwe, it was notable that the European cattle, although affected, carried the virus for a shorter period than the Zebu. The SAT viruses occasionally spread out of Africa into the Middle East, most recently into Saudi Arabia during 2000. But while the 0, A and ASIA 1 serotypes persist in this region, in spite of limited attempts at control, the SAT viruses die out. This implies that the cattle, sheep and goats are unable to maintain the SAT serotypes, or conversely, these serotypes require particular host species, either the Zebu for SAT 2 or the African buffalo. Similarly, the distribution of ASIA 1 serotype would suggest that it is constrained from establishing itself outside of Asia. The definitive identification of the carrier or subclinically infected animal is recovery of live FMD virus. The predilection of the virus for the epithelium of the pharynx makes this tissue the most suitable to sample, a procedure which can be carried out using the probang sampling cup (Kitching and Donaldson, 1987). This is a hollow metal cup with a slightly sharpened edge, attached from the centre of its bowl by a long wire, approximately half a metre long, to a handle at its free end, which can be pushed into the mouth of the animal being tested, over the base of the tongue into the pharynx. The cup is then withdrawn, collecting as it is pulled out, mucous and superficial cellular material from the pharynx. The contents of the cup are usually mixed with a neutral buffer solution, and if not examined immediately, kept frozen over liquid nitrogen or on dry ice (solid carbon dioxide). Live virus can be cultured on sensitive tissue culture such as primary bovine thyroid cells or lamb kidney cells. Carrier animals, which have either recovered from clinical disease, or have been vaccinated and subsequently acquired infection following contact with live virus, will also have high levels of specific anti-FMD virus antibody present in their phayngeal mucous, and treatment of the probang sample with Freon or Arcton can help dissociate the virus/antibody complexes, and increase the possibility of recovering virus on tissue culture. Subclinically infected animals, other than those with partial vaccinal immunity, will not usually have detectable antibody levels at this stage of infection. 5. Antibody in Saliva Carrier animals have specific antibodies to FMD virus. These can be detected in the serum (DIE, 2001 b) and also in the saliva. Specific IgA is present in saliva, and this is elevated in the carrier animal, probably because of the constant low levels of virus maintaining the antigenic stimulus to the mucosal immune system. An ELISA has been developed to quantify this elevated level of specific IgA, to indicate the possibility that the animal from which the sample was collected could be a carrier (Amadori et al., 2000). Some carrier cattle, however, fail to produce a level of IgA

Copyright © 2004 By Horizon Bioscience

Kitching

4171

in their saliva significantly higher than non-carrier cattle, and while this test has potential as a herd test, further refinement and increased sensitivity is required. 6. Non Structural Proteins (NSPs) FMD virus has a positive sense, single stranded RNA genome of 8400 nucleotides that codes for 12 proteins, 4 of which are structural and make up the capsid of the virus, and 8 of which are non-structural, which together allow the virus to replicate in an infected cell. The structural genes are identified as lA, 1B, 1C and 1D, the non-structural as L, 2A, 2B, 2C, 3A, 3B, 3C and 3D. The functions of the proteins for which the non-structural genes code have not all been fully identified, and it is beyond the scope of this paper to describe the current opinions (see Rueckert, 1996). The vaccine used to help control outbreaks of FMD is a dead preparation of whole virus particles in an oil or aluminium hydroxide/saponin adjuvant. There will be no replication of the virus following vaccination, and the vaccinated animal will develop antibodies to the structural proteins of the virus present in the viral capsid, some of which will be neutralizing, and protect it from subsequent infection. Because there is no viral replication, there is no expression of the non-structural proteins (NSPs), and the animal will not develop antibodies to these proteins - although some vaccines do contain low levels of these NSPs depending on the manufacturing process, in particular 3D, and there can be a low antibody response to the NSPs, more obvious in multiply vaccinated animals. The 3D gene codes for the viral polymerase, and antibodies to this protein were in the past detected in the agar gel immuno-diffusion VIAA (virus infection associated antigen) test. Animals that have recently recovered from infection will have antibodies to the NSPs, because as the virus replicates in its tissues, these proteins will be expressed and stimulate the production by the host of specific antibodies. The detection of these antibodies can therefore be used to identify those animals that have had FMD, and which may, therefore be still carrying live virus. A variety of tests have been developed to detect these antibodies, including ELISA and EITB (OlE, 2000a), using pure NSP antigens expressed in viral (baculovirus) or plasmid (E.coli) expression systems. These tests have been predominantly designed to detect NSP antibodies in cattle, and are less useful in sheep and pigs. Sheep, in particular, probably because of the frequently subclinical nature of FMD, may fail to develop detectable levels of these antibodies. Even in cattle, there is considerable individual variation in the amount of antibody produced to each of the NSPs, and consequently the period of time after infection each can be detected. The 2C antibodies may be detectable for 12 months, while the 3ABC antibodies persist longer; it is likely that the severity of the infection is the major influence on the levels and the subsequent duration of detection of the NSP antibodies. The NSP are not FMD serotype specific proteins, and the NSP antibody tests will not distinguish between infection with any of the serotypes of FMD virus. This can have the advantage that a single screening test can be used to look for evidence of infection in a situation in which the predominant serotype(s) are not known. In South America the EITB, which uses a Western blotting technique to detect the antibodies to 5 of the NSPs, 3A, 3B, 2C, 3D and 3ABC, was very successfully used to support the local FMD control programmes, and the ultimate recognition by OlE of freedom from FMD, particularly for regions of Brazil (Bergmann et ai., Copyright © 2004 By Horizon Bioscience

1418

Diagnosis and Control

1998). A 3ABC ELISA was used to define the limits of the 1996 outbreak of FMD in the Balkans, and antibody to the 3ABC polyprotein is considered the single most reliable indicator of infection (Mackay, 1998). There is, however, a problem with the NSP tests on an individual animal level. Some cattle that have been vaccinated, particularly with a high potency vaccine as might be used in an outbreak in a previously FMD free country, will fail to develop antibodies to the NSPs should they have contact with live virus. This is because their level of immunity prevents any significant viral replication, and therefore expression of the NSPs; however, they could become carriers of live virus. On a herd basis, even potent FMD vaccine will not protect 100% of the cattle, and should the herd become exposed to live virus, some will support replicating virus, even though they do not show clinical disease, and sero-convert to some of the NSPs, in particular to 3ABC. Thus, by testing the whole herd, it would be possible to diagnose the previous encounter with live virus, and the potential for the presence of carriers, assuming, of course, that the whole herd was exposed to the same challenge. The test may fail if only a few animals contacted live virus, perhaps as an aerosol from a neighbouring infected farm, and these could all have been sufficiently immune to prevent the expression of the NSPs. The test for the antibodies to the NSPs is a significant advance in the detection of the carrier animal. However, it has its limitations, and cannot be used reliably on an individual animal to exclude its potential to be carrying live virus, and even when used on a whole herd, it would be unable to provide a guarantee. The possibility of carrier animals creating fresh outbreaks is probably extremely small, and this can be further reduced by using probang and serological tests. But, however small the risk, if importing countries have a choice, they are likely to choose to import their live animals and animal products from areas where there is no FMD vaccination or possibility of the presence of carrier animals. Until the identification of the carrier is 100% certain, FMD will probably remain a significant constraint to trade in susceptible animals and their products. 7. Pen-Side Tests A pen side diagnostic test would have been particularly valuable during the recent UK outbreak, to assist field veterinarians in their clinical diagnosis. One which had been developed for rinderpest diagnosis was in the process of validation and was used with some success towards the end of the outbreak (Ferris et at., 2001; Reid et al., 2001). The test relies on FMD viral antigen being recognized by a monoclonal antibody (Mab) attached to a coloured latex bead. The antigen/Mab/bead complex is trapped by a fixed band of additional anti-FMD virus monoclonal antibody as it migrates along a chromatographic strip, creating an easily identifiable coloured line over the fixed band of Mab, as the latex beads concentrate. The result can be read in 10 minutes. However, the test requires the amount of antigen as may be found in the epithelium of a ruptured vesicle, but would not be sufficiently sensitive to detect antigen in a blood sample; in many cases of FMD in sheep, there may not be sufficient epithelium available for the test to be used. Adaptations of the PCR could also provide a more rapid diagnosis, either to detect viral genome in nasal swabs before the development of clinical disease, or for use in portable machines taken to the suspect premise (Donaldson et at., 2001a). Pen side antibody tests for NSP and Copyright © 2004 By Horizon Bioscience

Kitching

4191

structural proteins are also being developed using similar principles to the antigen detection test which could have useful application in an outbreak situation. 8. Control Following the introduction of FMD virus into a country previously free of FMD, there is usually an initial attempt to eradicate the disease by slaughter, but most countries retain the option to vaccinate and thus maintain membership of a vaccine bank, which can provide vaccine at short notice. Control of an outbreak by slaughter of all infected and in-contact susceptible animals can be very effective in eliminating the virus and quickly re-establishing freedom for trade purposes, as was shown in the 1967/68 outbreak in the UK. The FMD virus can only spread by well defined methods, and if it is prevented from spreading, the disease will quickly die out. Most spread of FMD virus is associated with the movement of infected animals, and if these are removed as soon as possible and a ban on the movement of all susceptible species is imposed, the potential for further spread is greatly reduced. The virus can also be transported in products derived from infected animals such as meat, milk and semen, and on fomites contaminated with the excretions of infected animals, such as vehicles, the hands and clothes of people in contact with infected animals, fodder, manure, surgical instruments, and other non-susceptible species of animal. In addition, FMD virus can spread as an aerosol from infected animals, particularly pigs, sometimes over a considerable distance, depending on weather conditions and the strain of virus. All these methods of spread are well documented, and can be controlled or anticipated during an outbreak, and typically, an epidemiological team of veterinarians experienced with FMD are put in charge of identifying the likely course of an outbreak, and putting in place suitable preventative measures. The potential for wildlife species to maintain FMD is not clear in all circumstances. In southern Africa, the African buffalo population in many of the game parks are infected with the SAT serotypes of FMD virus, and there are many examples of infection spreading from the buffalo to cattle, either following direct contact (Dawe et al., 1994) or by infection passing from the buffalo to other susceptible species such as impala, and then to the cattle. In other parts of the world wild deer and feral pigs have been infected with FMD, but there is no evidence that they have spread it to domestic livestock. In the 2001 UK outbreak, although it was suspected that deer had been involved, none of the farmed or wild deer sampled were positive for virus or antibody. However, during the 2001 outbreak in the UK, policy for the control and eradication program was dictated by models developed with little understanding of the basic principles of the disease, particularly that the strain of the PanAsia topotype causing the outbreak did not spread significantly as an aerosol (Donaldson et al., 2001b). The consequence of the policy adopted was a massive overkill of healthy animals, which the models predicted were infected. In all, over 4 million animals were slaughtered, compared with 500,000 during the 1967/68 outbreak which actually affected more farms. However, the public and therefore political perception was that this scale of slaughter should never be repeated to control any future outbreak, and that vaccination should be used as an alternative. A major constraint on the use of vaccination, if the vaccinated animals are not subsequently slaughtered, is the difficulty of convincing trading partners that the virus has been Copyright © 2004 By Horizon Bioscience

\420

Diagnosis and Control

eliminated from the national herd (see above), and this was reflected in the DIE Code (DIE, 2001a). If the vaccinates are not slaughtered, a 12 month period was required before a country could re-apply for FMD freedom. This period has now been reduced to 6 months (DIE, 2002), if a serological survey using NSP tests shows that there is no longer active infection in the vaccinated population. This change was the result of both political pressure to reduce the consequences of using vaccine to help control an outbreak, and thereby reduce the number of animals slaughtered, and an optimism in the effectiveness of the NSP tests. It remains to be seen whether importing countries will be willing to accept the additional risks associated with this change in policy. If vaccination is used it is first essential to ensure that the vaccine strain selected will provide protective immunity against the stain causing the outbreak. Within each FMD virus serotype there is an infinite number of strains exhibiting a spectrum of antigenic characteristics, and while it is not expected to have available a vaccine strain identical to the outbreak strain, as close an antigenic match as possible must be achieved (Kitching et al., 1989). A decision is required whether to vaccinate outside the immediate infected area (protective vaccination) or to vaccinate close to, or even on the infected premises (suppressive vaccination). Suppressive vaccination is designed to reduce the production of virus from exposed and possibly infected animals, but has the danger of creating carriers, particularly cattle, which could maintain the virus long after the outbreak was thought to be eliminated. If the outbreak is extensive, a single round of vaccination may not be considered sufficient, both because protective immunity from a single dose of even high potency vaccine is only likely for a maximum of 4 months, or shorter if the vaccine strain is not well matched antigenically to the outbreak strain, and because there will be a growing population of young animals not included in the first campaign. Consideration must then be given to the effects of maternal immunity which will be transferred in the colostrum of the vaccinated adults to young stock born after the introduction of vaccination. Maternal immunity will protect young calves up to 4 months of age, but will then wane, leaving them susceptible to infection; however, FMD vaccine will not be effective in the presence of maternal antibodies and it is therefore necessary to adopt a policy of vaccinating the calves as they become susceptible. As this will vary according to the amount of colostral immunity received, it is inevitable that some calves will fail to respond and some will have been susceptible for some time before they are protected by vaccination. For a vaccination program to be effective in a cattle population exposed to low levels of virus challenge, at least 80% should be protected by vaccination. At higher levels of exposure, as experienced in the FMD endemic situation seen in some countries of the Middle East, even vaccinating the cattle every 10 weeks is sometimes not sufficient to prevent outbreaks of FMD. In countries usually free of FMD, control of FMD in pigs is also usually by slaughter of all clinically affected and in-contact susceptible animals, together with movement restrictions and disinfection. The speed with which FMD spread in Taiwan, Province of China, when it entered in 1997 made it impossible to control by slaughter alone, and vaccination was introduced. Because of the rapid generation time of pigs, it is usually considered uneconomic or impractical to Copyright © 2004 By Horizon Bioscience

Kitching

4211

attempt to maintain immunity in a national pig herd. However, in Taiwan, the outbreak strain was of the Cathay topotype, and only the pigs were affected (Dunn and Donaldson, 1997), and in this situation vaccination was considered a viable option. The policy has proved successful, although it has taken over 4 years to completely eradicate the disease. Similarly in the Philippines, the same topotype is present, and the control and eradication programme has concentrated on the pigs, also using vaccination. FMD vaccines for use in pigs must have an oil adjuvant, as the aluminium hydroxide/saponin adjuvant used for cattle and sheep vaccines is not effective in pigs. Protection can be provided to naive pigs by day 4 post vaccination using high potency vaccines with some of the newer oil adjuvants now available (Barnett et al., 1996). The immune response in young pigs to FMD vaccine is poor, and protection is best provided by vaccination of the pregnant sow so that immunity can be passed on in the sow's colostrum (Kitching and Salt, 1995). However, in the presence of maternally derived immunity, it is not possible to initiate an effective immune response from vaccination before 8 weeks ofage (Francis and Black, 1986). Usually vaccination is delayed until 10-12 weeks of age, and repeated 2 weeks later, which may provide sufficient immunity until their slaughter weight is reached, depending on the expected level of challenge with field virus. Sows should be vaccinated at least twice yearly, during pregnancy. In the presence of clinical disease within a pig herd it is unlikely that vaccination will provide sufficient protection, and even using vaccines of high potency (PD50 > 6) it is common that vaccinated pigs in contact with clinically affected pigs will themselves develop clinical signs. This is not evidence of poor vaccine quality but of the virulence of some strains of FMD virus in pigs. The control of FMD in sheep and goats follows the same principles as would apply with the disease in other susceptible farm livestock, movement restrictions, disinfection and either slaughter of affected and in-contact animals or vaccination. In countries that have endemic FMD, sheep may be included in the regular vaccination campaign, although they are rarely vaccinated more than once a year. Sheep and goats are usually given half or a third of the cattle vaccine dose, and either oil or aluminium hydroxide/saponin adjuvants can be used. The duration of immunity against disease will depend on the severity of challenge with live virus in the field, the antigenic relationship between the vaccine strain and the field virus and the potency of the vaccine used. The often silent nature of the disease in adult sheep and goats can also give the impression of a successful vaccination programme, when in reality the virus is freely circulating. The ability of different strains of FMD virus to maintain themselves in a small ruminant population becomes critical in these situations, but has so far received limited attention (Hughes et al., 2002a; 2002b). Attempts to vaccinate the national sheep population are usually frustrated by the numbers involved, due to the cost of the vaccine, and the resources required to administer the vaccine. The initial intention to include sheep and goats in the vaccination campaign throughout Turkey was recognized as unrealistic, and vaccination was targeted at all susceptible stock in European Turkey (Thrace), and only certain areas of Anatolia. Similarly, Morocco concentrated on vaccinating the sheep and goats on the border with Algeria, together with strategic vaccination Copyright © 2004 By Horizon Bioscience

\422

Diagnosis and Control

around affected areas during its last outbreak of FMD. Uruguay vaccinated only its cattle population during the final stages of its successful FMD eradication programme in the early 1990's, and left the approximately 20 million sheep unvaccinated during its eradication programme, and has repeated this policy following the 2002 outbreaks of serotype A. 9. Conclusion The recent UK experience has demonstrated the importance of sheep in the epidemiology of FMD caused by certain strains of virus. The difficulties encountered in identifying clinical disease resulted in large numbers of healthy animals being slaughtered, while other infected flocks went unrecognized, allowing the virus to spread. While some would argue that the policy adopted was effective, and the overkill was justified (Ferguson et al., 2001), a more traditional approach of methodical tracing and laboratory testing may have been equally effective, less expensive and morally more defensible. The change that the UK outbreak brought about in the attitude to control of the disease by slaughter will undoubtedly result in the more rapid introduction of a vaccination policy in any future outbreak, but this will also have its problems. Hopefully a consequence will be a resurgence of interest in developing rapid and reliable on farm diagnostic tests, more effective vaccines, and a better understanding of the natural history of FMD, particularly the potential role of the carrier animal. References Amadori, M., Haas, B., Moos, A. and Zerbini, I. 2000. IgA response of cattle to FMDV infection in probang and saliva samples. European Commission for the control of foot-and-mouth disease. Session of the Research Group of the Standing Technical Committee, Bulgaria. 5-8 Sept. 2000. FAD, Rome. Armstrong, R.M. 1997. Development of tests for antibodies against foot-and-mouth disease virus in cattle milk. 1. Virol. Meth. 63: 175-180. Armstrong, R.M. and Mathew, E.M. 2001. Predicting herd protection against foot-andmouth disease by testing individual and bulk milk samples. J. Virol. Meth. 97: 87-99. Armstrong, R.M., Mathew, E.S. and MacKay, D.K. 2000. Validation of the specific isotype assay to detect antibodies against foot-and-mouth disease virus in bovine milk. J. Virol. Meth. 85: 193-201. Bergmann, I.E., Astudillo, V., Malirat, V. and Neitzert, E. 1998. Serodiagnostic strategy for estimation of foot-and-mouth disease viral activity through highly sensitive immunoassays using bioengineered nonstructural proteins. J. Clin. Sci. Epid. 20/S2: 6-9. Barnett, P., Pullen, L., Statham, R. and Salt, J. 1996 - Preliminary studies on emergency foot-and-mouth disease vaccines formulated with metabolizable or semi-metabolizable oil adjuvants. Proceedings of the Research Group of the Standing Technical Committee of the European Commission for the Control of Foot-and-Mouth Disease, Israel, 2-6 Sept. p 156-169. David, M., Torres, A., Mebus, C., Carrillo, B.J., Schudel, A., Fondevilla, N., Blanco Viera, J. and Marcovecchio, F.E. 1993. Further studies on foot-and-mouth disease in the llama. Proc. Ann. Meet. U.S. animo Hlth. Assoc. 97: 280-285. Dawe, P.S., Sorenson, K., Ferris, N.P., Barnett, I.T.R., Armstrong, R.M. and Knowles, N.J. 1994. Experimental transmission of foot-and-mouth disease virus from a carrier African buffalo (Syncerus caffer) to cattle in Zimbabwe. Vet. Rec. 134: 211-215.

Copyright © 2004 By Horizon Bioscience

Kitching

4231

De la Rua, R., Watkins, G.H. and Watson, P.1. 2001. Idiopathic mouth ulcers in sheep (letter). Vet. Rec. 149: 30-31. Donaldson A.1. 1987. Foot-and-mouth disease: the principle features. Irish Vet. J. 41: 325327. Donaldson, A.I., Alexandersen, S., Sorenson, J.R. and Mikkelsen, T. 2001a. Relative risks of the uncontollable (airborne) spread of FMD by different species. Vet. Rec. 148: 602604. Donaldson, A.I., Hearps, A. and Alexandersen, S. 2001b. Evaluation of a portable, real time PCR machine for FMD diagnosis. Vet. Rec. 149: 430. Dunn, C.S. and Donaldson, A.1. 1997. Natural adaptation to pigs of a Taiwanese isolate of foot-and-mouth disease virus. Vet. Rec. 141: 174-175. Ferguson, N.M., Donnelly, C.A. and Anderson, R.M. 2001. Transmission intensity and impact of control policies on the foot and mouth epidemic in Great Britain. Nature. 413: 542-547. Ferris, N., Reid, S., Hutchings, G., Kitching, P., Danks, C., Barker, I. and Preston, S. 2001. Pen-side test for investigating FMD (letter). Vet. Rec. 148: 823-824. Francis, M.J. and Black, L. 1986 - Response of young pigs to FMD oil emulsion vaccination in the presence and absence of maternally derived neutralising antibodies. Res. Vet. Sci. 41: 33-39. Huang, C.C., Lin, Y.L., Huang, T.S., Tu, W.J., Lee, S.H., Jong, M.H. and Lin, S.Y. 2001. Molecular characterization of foot-and-mouth disease virus isolated from ruminants in Taiwan in 1999 to 2000. Vet. Microbiol. 81: 193-205. Hughes, G.1., Mioulet, V., Kitching, R.P., Woolhouse, M.E.J. Alexandersen, S. and Donaldson, A.I.. 2002a. Foot-and-mouth disease virus infection of sheep: implications for diagnosis and control. Vet. Rec. 150: 724-727. Hughes, G.1., Mioulet, V., Haydon, D.T., Kitching, R.P., Donaldson, A.I. and Woolhouse, M.E.J. 2002b. Serial passage of foot-and-mollth disease virus in sheep reveals declining levels of viraemia over time. J. Gen. Virol. 83: 1907-1914. Kitching R.P. 1998 - A recent history of foot-and-mouth disease. J. Compo Path. 118: 89108. Kitching, R.P. and Donaldson, A.I 1987. Collection and transportation of specimens for vesicular virus investigation. Rev. Sci. Tec. a.I.E. Epi. 6: 251-261. Kitching R.P. and Salt J.S. 1995 - The interference by maternally-derived antibody with active immunization of farm animals against foot-and-mouth disease. Br. Vet. J. 151: 379-389. Kitching, R.P., Knowles, N.J., Samuel, A.R. and Donaldson, A.1.1989. Development of footand-mouth disease virus strain characterisation - a review. Trop. Anim. Hlth. Prod., 21, 153-166. Kitching, R.P., Rendle, R. and Ferris, N.P. 1988. Rapid correlation between field isolates and vaccine strains of foot-and-mouth disease virus. Vaccine. 6: 403-408. Lin, F. and Kitching, R.P. 2000 - Swine vesicular disease: an overview. Vet. J. 160: 192201. MacKay, D.K.J. 1998. Differentiating infection from vaccination in foot-and-mouth disease. J. Clin. Sci. Epid. 20/52: 2-5 Mackay, D.K.J., Bulut, A.N., Rendle, T., Davidson, F. and Ferris, N.P. 2001. A solid-phase competition ELISA for measuring antibody to foot-and-mouth disease virus. J. Virol. Meth. 97: 33-48. OlE 2001 a. Foot and mouth disease, Chapter 2.1.1. In: International Animal Health Code, 10th Ed, 2001, Paris. p. 63-75. OlE 2001 b. Foot and mouth disease, Chapter 2.1.1. In: Manual of Standards for Diagnostic Tests and Vaccines, 4 th Ed., 2000, Paris. p.72-92. OlE 2002. 70th General Session, International Committee, World Organization of Animal Health, Paris, May 26-31 2002. Copyright © 2004 By Horizon Bioscience

1424 Diagnosis and Control

Reid, S.M., Ferris, N.P., Hutchings, G.H., Samuel, A.R. and Knowles, N.J. 2000. Primary diagnosis of foot-and-mouth disease by reverse transciption polymerase chain reaction. 1. Virol. Meth. 89: 167-176. Reid, S.M., Ferris, N.P., Bruning, A., Hutchings, G.H., Kowalska, Z.A. and Akerblom, L. 2001. Development of a rapid chromatographic strip test for the pen-side detection of foot-and-mouth disease virus antigen. J. VirOI. Meth. 96: 189-202. Rueckert, R.R. 1996. Picomaviridae: The viruses and their replication. In: Field's Virology. B.N.Fields, D.M.Knipe and P.H.Howley, eds. Lippincott-Raven, Philadelphia/New York. p. 609-654. Salt, J.S. 1993. The carrier state in foot-and-mouth disease, an immunological review. Br. Vet. J., 149: 207-223. Samuel, A.R. and Knowles, N.J. 2001 - Foot-and-mouth disease type 0 viruses exhibit genetically and geographically distinct evolutionary lineages (topotypes). Yang, P.C., Chu, R.M., Chung, W.B. and Sung, H.T. 1999. Epidemiological characteristics and financial costs of the 1997 foot-and-mouth disease epidemic in Taiwan. Vet. Rec. 145: 731-734. Zhang, Z.D. and Kitching, R.P. 2001. The localisation of persistent foot-and-mouth disease virus in the epithelial cells of the soft palate and pharynx. J. Compo Path. 124: 89-94.

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrino and Esteban Domingo

Chapter 16 Control of Foot-and-Mouth Disease: Role of International Organizations

J. Blancou, Y. Leforban, and J. E. Pearson

Abstract Foot and mouth disease (FMD) is one of the most contagious diseases of mammals and can cause severe economic loss. The Food and Agriculture Organization of the United Nations, Office Internatioal des Epizooties and Pan-American Health Organization have had extensive programmes for FMD surveillance and control. The efforts of these organizations focus on disease reporting, disease status evaluation, safety of world trade, diagnosis and research, standardisation of FMD vaccine production, coordinated control of outbreaks (eg: in Europe in the 90's) and international support of national and regional FMD control programmes. The coordinated activities of these organizations have contributed greatly to eradication or control of FMD in many countries, and have facilitated global trade while minimising the risk of the introduction of the virus from infected to disease free zones. 1. Introduction Foot and mouth disease (FMD) is the most contagious disease of mammals and has a great potential for causing severe economic losses in susceptible cloven-hoofed animals. Since agriculture became intensive, the disease has been a major cause for concern amongst livestock farmers, veterinarians, and those responsible for the rural economy of developing countries. In 2002, amongst the 158 countries reporting their diseases to the Office International des Epizooties (OIE:see annex), 55 were officially FMD free without vaccination, which means that one-third have been able to eradicate the disease using animal health measures while the virus still circulates in about two-thirds of the countries (Vallat, 2002). It therefore does not come as a surprise that many countries request international assistance to control this disease. The developing countries seek assistance to combat FMD with a view to gaining access to markets for their animals and animal products, and the industrialised countries try to protect their FMD free zones against the risk of importing the FMD virus. Thus, since the first International Veterinary Congress in Hamburg (Germany), in 1863, there has been a continuous effort by international organizations to develop and strengthen the FMD surveillance, prevention, and outbreak control programmes and even to attempt eradication of the disease at a regional level (Blancou, 2002). Copyright © 2004 By Horizon Bioscience

1426

Role of International Organizations

This effort was mainly directed at an improvement in the veterinary infrastructure and the quality of disease reporting, in order to facilitate the evaluation of the national disease status and to ensure the safety of world trade. This goal was mainly achieved through a better FMD diagnostic capability and the development and standardisation of FMD vaccines plus the support of research on the disease. In the meantime, international organizations have also reinforced their technical and financial support to national FMD control programmes.

2. Disease Reporting At national level, FMD surveillance is a high priority in most of the countries throughout the world. At the international level, country reports are received at the Office International des Epizooties (OIE).The DIE has designated 15 diseases, including FMD, as being List A diseases. List A diseases are transmissible diseases which have the potential for very serious and rapid spread, irrespective of national borders, and consequently are of major importance in the international trade of animals and animal products. There are also 76 list B diseases, which are considered to be less significant in the trade of animals and animal products. Countries shall make available to other countries, through the DIE, whatever information is available that will help prevent the spread of important animal diseases. The DIE Terrertrial Animal Health Code (the Code) outlines the Standards for the Member Countries, and has guidelines for disease reporting (DIE, 2003). It states that Member Countries previously considered free should report outbreaks of FMD within 24 hours to the DIE. This information is forwarded immediately to other countries. The Code also states that Member Countries shall report a provisional diagnosis of FMD to other countries, if this represents important new information of epidemiological significance. Following the initial report, monthly reports should be provided. The DIE now takes a proactive approach to disease reporting and will also report information on FMD outbreaks that is provided by the DIE FMD Reference Laboratories and from unofficial sources such as scientific publications, ProMed, and lay publications after it has been verified with the Member Country. The resulting disease reports demonstrate the spread of the disease and the threat it poses. This can assist countries to avoid the introduction of the FMD virus from infected to FMD free zones, as has been the case during the past decades. The Food and Agriculture Organization of the United Nations (FAD: see Annex) also plays a role in FMD control at the regional level. A good example of this interaction is the European Commission for the Control of Foot-and-Mouth Disease (EUFMD) established in 1954 as a Special Body of FAD, to combat FMD in Europe. The role of the Commission includes providing support to Member Countries in FMD surveillance and circulating information on their FMD situation, especially when there is a threat to Europe. The FAD Emergency Prevention System against transboundary animal and plant pests and diseases (EMPRESS) concept has also been applied by FAD to FMD surveillance. The emphasis of EMPRESS is on the early warning of the occurrence of epidemic disease (notably FMD) and the early reaction to this event in order to have effective prevention, containment and management of disease outbreaks. Early detection of a transboundary disease enables an early response, and is contingent upon having a reliable disease surveillance system in place

Copyright © 2004 By Horizon Bioscience

Blancou et al. 4271

3. Disease Status Evaluation and Safety of World Trade In 1968 the DIE approved the first Terrestrial Animal Health Code for the harmonisation of trade of animals and animal products. FMD was one of the first diseases for which Standards were established. In 1995 the Standards developed by the DIE were formally recognised by the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS Agreement) of the World Trade Organization (WTO). Chapter 2.1.1 of the Code provides requirements that must be met for a country or zone to be defined as free of FMD and the recommendations regarding the safe import of animals or animal products into a country. This chapter deals not only with the occurrence of clinical signs caused by FMD virus (FMDV) but also with the presence of infection with FMDV in the absence of clinical signs. A definition for FMDV infection is provided. The inclusion of FMDV infection as well as FMD was approved by the DIE International Committee in May 2002. The Chapter also provides the criteria for a country or zone to be FMD free with or without vaccination. It states that any country applying for FMD freedom will have a record of regular and prompt animal disease reporting. It also states that the country will provide documented evidence that an effective system of surveillance for both FMD and FMDV infection is in operation and that regulatory measures for the prevention and control of FMD have been implemented. Surveillance guidelines for FMD and FMDV infection have been developed by the DIE and sent to Member Countries for comment. The key points in the Code Chapter, with the changes that were approved by the DIE in 2002, have been recently summarized. (Pearson, 2002). This summary includes the criteria for recognition of FMD free country or zone, with or without vaccination and the criteria for regaining FMD free status after an outbreak. The requirements for demonstration of freedom from FMD infection have also been recently reviewed and updated by the DIE (Kitching, 2002). The Code and this paper are on the DIE web-site (http://www.oie.int) At the regional level, rules have also been set up by the European Commission to avoid importation of the FMDV into Europe. Those rules are included under Directive; the current Directive which applies to FMD control in European Union (EU) is the Directive 85/511/EEC. This Directive is currently under review and the preparation of a new more detailed and updated Directive is in progress in Brussels. 4. FMD Diagnosis and Research The DIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals (the Manual) provides a uniform approach to the diagnosis of FMD (DIE, 2000). It describes in detail the various FMD diagnostic tests. These include virus isolation and serological techniques. The Manual provides a list of prescribed tests; these are the tests that are required by the Code for the testing of animals in connection with international trade. The 2000 edition of the Manual specifies that the liquid phase blocking ELISA, and virus neutralization tests are the two prescribed tests. This edition also provides a general description of the non-structural protein test. In May 2002 the DIE provided a detailed protocol for the solid phase blocking ELISA and approved it as a prescribed test for international trade. It also set the criteria for the use of non-structural protein test to evaluate freedom from FMDV infection in vaccinated animals. Copyright © 2004 By Horizon Bioscience

\428

Role of International Organizations

An accurate diagnosis of FMD, including the determination of virus serotypes, is of utmost importance to ensure that proper control measures are initiated and/or to trace the origin of the outbreaks. Most of the more proficient laboratories, which have the capability to carry out this diagnosis, are included in an international network, covering all regions of the world, under the auspices of the DIE or the FAD. The list of these laboratories are presented and updated on the web-sites of these organizations. One of these laboratories, the Institute for Animal Health, Pirbright, United Kingdom, (previously Animal Virus Research Institute) was designated as the World Reference Laboratory (WRL) by the FAD and the DIE. In this laboratory, techniques for subtype differentiation and characterisation of the strains have been developed; these techniques facilitate the tracing back of outbreaks and monitoring of disease situations on a world-wide basis. Over the last 43 years, the services provided by the WRL have proven to be invaluable in that they have permitted the mapping of the virus distribution in the world and often the characterisation of specific disease situations. Molecular techniques, including the sequencing of the genome, are now applied routinely as a tool for molecular epidemiology. Some basic research on FMD virus or FMD epidemiology is supported by international organizations through their Reference Laboratories. These organizations have a number of scientific publications, some of which deal with FMD control. As an example, the DIE Scientific and Technical Review published an edition in 1999 on "Management of animal health emergencies" and published an edition in December 2002: "Foot and mouth disease: facing the new dilemmas" 5. Standardisation of FMD Vaccine Production and Control FMD vaccines have been and are still a component of disease control and eradication strategies in most parts of the world. Despite the fact that their use is now prohibited in some FMD free regions (e.g. North America and Europe), the proper control of the potency, purity, safety and efficacy of these vaccines remains of utmost importance. At the world level, guidelines for this control are recommended in the DIE Manual and technical advice or assistance can be sought from DIE or FAD Reference Laboratories. At a regional level, the Pan-American Health Organization (PAHO:see annex) plays a crucial role in FMD vaccine development and standardisation in South America. Pilot plans of vaccination with oil adjuvant vaccines were carried out by its laboratory in Brazil, the Pan American Foot-and-Mouth Disease Centre ("Panaftosa"). During the 70's, Panaftosa developed the technology for production of an oil adjuvant vaccine to be used in cattle and pigs. This technology was passed on to the vaccine producing laboratories in South America (Saraiva, 2003). Panaftosa, as a regional reference laboratory, has been testing, adapting and forwarding to the industry FMD strains that can be used for the production of vaccines. Vaccines produced by both private and official laboratories were evaluated for safety, purity and potency by Panaftosa. Panaftosa has also supported the South American countries by implementing test routines for the challenge of cattle following vaccination and for the determination of antibody titres using ELISA (Casas Olascoaga et ai., 1990)

Copyright © 2004 By Horizon Bioscience

Blancou et al. 4291

A22 - IRAQ - 1964

Figure 1. Chronology and routes of the introductions in Europe of new foot and mouth disease strains from Middle East between 1962 and 2000

6. Coordinated Control of Outbreaks During the last decade, a number of serious outbreaks of FMD threatened previously free countries or regions. International organizations played a major role in controlling the disease and preventing its spread. Three examples of international control are given below: 6.1. Buffer Zone in Turkey

The EUFMD Commission played a major role in preventing the introduction of exotic FMDV strains into Europe from the Middle East through Turkey (Figure 1). Since its establishment in 1954 the activities of the EUFMD Commission have been focused on Turkey and the Balkans with the following objectives: 1. Immediate objective: establishment of a buffer zone of vaccination in Thrace where favourable physical conditions exist for concentrating efforts in a relatively small and easily defensible area. 2. Medium-term objectives: to develop a technical infrastructure, including vaccine production units, which would allow at a later stage a shift of the buffer zone system to eastern Anatolia. 3. Long-term objectives: strengthening of the surveillance and control of the disease in Turkey. Three different exotic types (SAT1, A22, and Asial) threatened Europe between 1962 and 1984 and this threat still persists as was demonstrated by the introduction of Asia 1 type into Greek territory in July 2000. Copyright © 2004 By Horizon Bioscience

1430

Role of International Organizations

(F>"F»'//"/

~\ ('

'Sofia

*"!

Vaccinated zone

Figure 2. Example of a regional coordinated international approach to the control of Foot and Mouth Disease: the type A outbreak in the Balkans (Albania, FYRo Macedonia, FYR Yugoslavia) in 1996

Vaccination campaigns were organised by Europe in Thrace as early as 1962 to establish and maintain a buffer zone to protect Europe (Boldrini, 1975). These campaigns were organised by the EUFMD and supervised by a tripartite FAO/ EEC/OIE Committee: - SAT 1 campaigns (1962-1964) - A22 campaigns (1965-1966) -A22 and Asia 1 campaigns (1972-1975) - Asia 1 (1984) Between 1962 and 1987, thirty rounds of vaccination were carried out. During this period, 25 million doses of vaccine at a cost of US $ 12 million were supplied by Europe through the FAD. This was calculated to be the equivalent of US$ 0.12 per head of cattle owned by the 18 contributing European countries, spread over a period of 25 years. (EUFMD, 1989) Parallel to the FAO-EUFMD campaigns, action \vas taken by the Union of Soviet Socialist Republics and Romania to protect their frontier areas against invasion by exotic viruses. The latter action, which also contributed to the defence of Europe, is not dealt with here. The following two examples will demonstrate how outbreaks were controlled in recent years with the help and cooperation of international organizations.

Copyright © 2004 By Horizon Bioscience

Blancou et al. 431

I

6.2. Control of Type A Outbreak in the Balkans in 1996 A type A epidemic occurred in the Balkans (Figure 2). The disease started in Albania where FMD was confirmed in May, 1996 in the District of Korce in south-eastern Albania. Ten villages were infected (EUFMD, 1997). Clinically affected animals were destroyed at the time the disease was detected and a programme of slaughter of the remaining susceptible animals in the infected households/villages was completed later. FMD virus type A was isolated and the nucleotide sequencing of the outbreak strain showed it to be very closely related to a strain circulating in India and Saudi Arabia. A two round ring vaccination with A22 Iraq vaccine strain was carried out in a zone of approximately 50-kIn radius around Korce. A total of 266,048 animals were vaccinated twice at intervals of three to four weeks. The disease was reported in June, in the Former Yugoslav Republic of Macedonia. A total of seventeen outbreaks occurred in the Skopje District and one in Titov Veles District. The cattle population of 4,366 animals in the 18 infected villages was stamped out. Two rounds of vaccination of cattle were carried out and approximately 120,000 cattle were vaccinated around the affected villages. The Veterinary Service of the Federal Republic of Yugoslavia reported FMD in Kosovo close to the border with Macedonia on 7 July. In total, 101 villages were declared infected. The last report was on 2 August. However, this was not confirmed by the expert mission which visited the country in October and no infected animals were detected by serology. A regional Committee was established to co-ordinate the measures in the three countries. The Committee held three meetings on 03 August in Skopje, 22 August in Tirana, and 02 October in Belgrade, with the participation of the three Chief Veterinary Officers (CVOs) and of the European Commission (EC), OlE and EUFMD. Meetings were also organised by the EC and held in Brussels on 5 and 18 July. The EUFMD - FAD and EC supported the countries in the control of the disease and a regional strategy was adopted and implemented under the guidance of international experts (Figure 2).

6.3. FMD Control in the Caucasus Region in 1999 and 2000 FAG has developed a cooperative programme with the All Russian Research Institute for Animal Health (ARRIAH). A Letter ofAgreement was signed in 1999 and in 2000 between FAO and ARRIAH. ARRIAH supplied 1 million doses of bivalent vaccines (against types 0 and A) to Armenia, Azerbaijan and Georgia. With the help of the EC, FAO, and OlE, ARRIAH also organised a serosurvey in 1999 and another one in 2000. This vaccine has been useful in helping the countries to increase the protection level against FMD, and the serosurvey provided evidence of circulation of FMDV. The establishing of a buffer zone at the southern border of the Caucasus region has been only partially successful as it did not completely prevent the circulation of the virus in the region. The objectives regarding the strengthening of the surveillance of the disease in the region. are still to be improved This buffer zone project is being resumed in 2003 and a long term programme for the control of FMD in the region is under discussion between the countries concerned and the international organizations (FAD, OlE, EC) (EUFMD, 2001): see Figure 3.

Copyright © 2004 By Horizon Bioscience

\432

Role of International Organizations

Caspian Sea

TURKEY

Figure 3. Example of a coordinated FMD control programme in the Transcaucasian region in 1999-2000: establishment of vaccination buffer zones.

7. International Support to FMD Control Programmes International organizations devote great effort to help their Member Countries to initiate, develop or strengthen national FMD control or eradication programmes. This effort may vary depending on the role and missions of each organization. The following are some of the actions that the OlE has taken to accomplish its role in FMD control: 1. Sponsor of the South East Asia FMD control programme: The Southeast Asia Foot and Mouth Disease (SEA-FMD) Campaign involve the coordinated control of Foot and Mouth Disease by eight countries in the ASEAN region. The campaign is coordinated through an OlE Regional Coordination Unit in Bangkok with major sponsorship by the Australian International Development Agency (AusAID). 2. Convene emergency meetings: Following the recent epidemics of FMD in Europe, South America, Asia and Africa, FAO and OlE jointly convened an international conference at the OlE Headquarters in Paris in April 2001 to review the world FMD status and assess the needs for adjustments in the Terrestrial Animal Health Code. The OlE International Committee endorsed recommendations from this Conference in May 2001. 3. Provide emergency funds to developing countries: The OlE has an emergency fund which can be used by a developing country for disease control. The OlE has a rapid approval procedure so that these funds can be disbursed quickly. This allows a country to initiate a programme with these funds and later find other organizations to provide continuing support. 4. Make experts available to visit countries with outbreaks: The OlE can provide experts to Member Countries to assist in developing disease control methods.

Copyright © 2004 By Horizon Bioscience

Blancou et al. 4331

5.

Partner in Pan African Programme to Control Epizootics (PACE): The OlE is a partner in the PACE programme for disease control in 29 African countries. The focus of the programme is rinderpest but FMD is one of the epizootics that is addressed. The OlE representative is chairman of the PACE Advisory Committee. 6. Sponsor of international meetings and training courses: The OlE has participated in a number of training courses on disease control and improvement of veterinary services as well as in meetings organised by the WTO, all of which contribute to the control of FMD. Similarly the FAO, through its Regular and Field Programmes, plays a major role in assisting, member countries, to improve surveillance and control of transboundary animal diseases. Foot and mouth disease is considered as a priority disease under the FAO EMPRESS (Leforban et aI., 2002). FAO has developed mechanisms to help member countries in emergency situations. Sponsoring of Technical Cooperation Projects (TCP) are common means by which FAO support countries in their effort to control FMD. Between 2000 and 2002, such support for FMD control was provided by FAO upon request to more than 10 member countries in different parts of the world. The fight against FMD is also addressed at the regional level through the collaboration of neighboring countries and the exchange of information (Kitching, 1998; Leforban, 1999). FAO has a network of four FMD reference laboratories (including the WRL) and two regional commissions on Animal Health apart from the EUFMD Commission: APHCA (Animal Production and Health Commission for Asia and the Pacific) and the newly createdAHCNENA (Animal Health Commission for the Near East and North Africa). In the Near-East Region, the "Regional Animal Disease Surveillance and Control Unit (RADISCON) in North Africa, Middle East and Arab Peninsula" has been developed by FAO and IFAD (International Fund for Agricultural Development). RADISCON seeks to develop an Animal Disease Surveillance and Control Network through a process of strengthening of veterinary investigation, animal disease information and laboratory services, continued professional development and improved professional communication bet\veen neighbouring countries. The project is assisting each individual country to establish its own National Animal Disease Surveillance System. Communication among participating nations is taking place electronically, and wherever possible electronic mail facilities are being used. FAO is also collaborating with several organizations that deal essentially with FMD: The Panaftosa in South America and the OlE SEA FMD Regional Coordination Unit in Bangkok (SEA FMD-RCU). At the global level FAa, OlE

and PAHO collaborate closely not only in terms of defining norms and standards under the OlE statutes but also in elaborating regional and global strategies. Thus for examples, in 1996, FAO joined forces with WHO, through the PAHO, and the OlE to convene, in Brasilia, the first International Inter-American Conference on Perspectives for the Eradication of Foot-and-Mouth Disease in the Next Millennium and Its Impact on Food Security and Trade. On 6 November 2001, FAO convened, with the participation of the OlE, a special Ministerial Meeting during the occasion of the 31 8t Session of the FAa Conference. This meeting was attended by 80 Member Countries. In December 2001, the Director-General of FAO and the Director-General of OlE both Copyright © 2004 By Horizon Bioscience

\434

Role of International Organizations

addressed a special International Conference on FMD convened by the ED in Brussels The World Food Summit, 5 years later, held in Rome in June 2002 also addressed the global issue of animal diseases in the context of global food security, food safety, trade and poverty reduction. Following the recommendation of this submit a joint FAa/OlE "Global Framework for the progressive control of FMD and other transboundary diseases" is under preparation. At the regional level, PARa was a major player in the control of FMD in South-America. As discussed previously PARa through the Panaftosa has been instrumental in insuring the quality of the FMD vaccine used in South America. Late in the 1980's, as a result of the knowledge of FMD ecosystem's structure developed by the national programmes with the support of Panaftosa and the political commitment of the governments, the Hemispheric Plan for the Eradication of Foot-and-Mouth Disease (PHEFA) was launched. It was devised to create and maintain FMD free areas; increase availability of meat and milk; increase efficiency of livestock operations and improve the access of animals and products from originally affected countries to international markets. The strategy adapted FMD control activities to the specific regions and the administration attempted to address the social impact and need for local participation with special attention to small livestock owners, who usually resist animal health programmes. The great leap ahead was the possibility of local programming and execution of sanitary measures, especially vaccination, resulting in average levels of coverage of around 85 to 94% of the total cattle herd. The local management body was usually named Local Committees composed of representatives from the national animal health authority, the Mayor, the Veterinary School, the State Veterinary board, and the livestock associations which are politically important at that level. (Saraiva, 2003) 8. Conclusion For more than a century, a tremendous amount of effort has been devoted to promote and co-ordinate the control of FMD at the international level. Every major international organization has spent a great deal of human, technical and financial resources to reach this goal. The epidemiological situation has improved substantially in some regions, namely in Europe and the Americas. Unfortunately, the situation remains unchanged, or is even worse in much of the rest of the world. This division of the world into two zones has a great influence on world trade of susceptible animals and their products. Those countries that are not disease free face a decisive handicap in their access to world markets. Moreover, many of them are developing countries, net importers of animal proteins, whose livestock activities are based on family farms and indigenous animal breeds. The low productivity of these breeds is barely affected by any outbreak of FMD that may occur. For these and other reasons, FMD control is not yet a priority for the Veterinary Authorities in most of the developing countries. The existence of this international enzootic poses a permanent threat to the other countries that have made considerable efforts to achieve disease-free status. The threat has increased over the years due to the globalisation of movement of goods and people, a trend that has been particularly marked in the last decade and that has coincided with the decrease and abandonment, of vaccination in the developed countries. Copyright © 2004 By Horizon Bioscience

Blancou et al. 4351

The international organizations concerned have therefore developed programmes to reduce the risk of the spread of FMD though international trade. These programmes include procedures for prompt reporting of FMD outbreaks, the development of import Standards that reduce the risk of the introduction of FMD, procedures to identify FMD free countries and assistance to National Governments to develop FMD control programmes. The goal of these programmes is to facilitate trade while minimising the risk of the introduction of FMD and avoid the introduction of the FMD virus from infected to FMD free zones, as has been the case during the past decades 9. References Blancou, J. 2002. History of the control of foot and mouth disease. Compo Immunn. Microbiol. Dis. 25: 283-296. Boldrini, G.M. 1975. Thirteen year campaign against exotic foot-and-mouth disease in Southern Europe and Anatolia. World Veterinary Congress, Salonica, 6-12 July, 1975. Casas Olascoaga R., Auge de Mello P., Abarac6n D., Gomes I., Alonso A , Mesquita J., Darsie G.C., Pinkoski D.I., Deak J.G., Gubel J.G., Barbosa J.R. 1990. Producci6n y control de vacunas antiaftosa con adyuvante oleoso en el Centro Panamericano de Fiebre Aftosa y en el Laboratorio Regional de Apoyo Animal del Ministerio de Agricultura y Reforma Agraria, Brasil. Bol. Centro Panam. Fiebre Aftosa. 56: 3-51. European Commission for the Control of Foot and Mouth Disease. (EUFMD). 1989. Activities and achievements 1954-1987. FAD, Rome. European Commission for the Control of Foot and Mouth Disease. (EUFMD). 1997. Report of the 32nd Session. FAD, Rome European Commission for the Control of Foot and Mouth Disease. (EUFMD). 2001. Report of the 34th Session, FAD, Rome. http://www.fao.org/waicent/FaoInfo/Agricult/AGN AGAH/EUFMD/reports/sess34/default.htm Kitching, R.P. 1998. A recent history of foot-and-mouth disease. (Review article). J. Compo Pathol. 118 : 89-108 Kitching R.P. 2002. Foot and mouth disease diagnostics: requirements for demonstration of freedom from infection. Technical Item, DIE General Session, May 2002. DIE, Paris, www.oie.int Leforban, Y. 1999. Prevention measures against foot-and-mouth disease in Europe in recent years. Vaccine: 17: 1755-1759. Leforban Y., Gerbier G., and Rweyemamu, M. 2002. Action of FAD in the control of foot and mouth disease. Compo Immunol. Microbiol. Infect. Dis. 25(5-6):373-382 Office International Epizooties. 2000. Manual of Standards for Diagnostic Tests and Vaccines. 4th ed. Office International des Epizooties, Paris, France Office International Epizooties. 2003. Terrestrial Animal Health Code. 12th Ed. OlE, Paris, France. Pearson J.E. 2002. Office international des epizooties role in foot and mouth disease control. In: Foot and Mouth Disease: Control Strategies. Dodet, B., and Vicari, M, eds. Applewhaite, C., techno ed. Elsevier, Paris Saraiva, V. 2003. Vaccines and foot-and-mouth disease eradication in South America. In: Vaccines for DIE List A and Emerging Animal Diseases. Develop. in BioI. Stand. Karger, Basel Vallat, B. 2002. A world with Foot and Mouth Disease under control. In: Foot and Mouth Disease: Control Strategies. Dodet, B., and Vicari, M, eds. Applewhaite, C., techno ed. Elsevier, Paris

Copyright © 2004 By Horizon Bioscience

\436

Role of International Organizations

ANNEX

1.

International Organizations Currently Involved In Foot And Mouth Disease Control The Office international des epizooties (DIE), also known as the World Organization for Animal Health, is an intergovernmental organization established in Paris since 1924, with 165 Member Countries in 2004.The three principal aims of the DIE are to gather and disseminate information on animal health, to promote international coordination of research into, and control of certain animal diseases, and to harmonise import and export regulations for animals and animal products at an international level. The organization thus plays a major role in reporting the occurrence of animal diseases world-wide, and in setting international Standards in the Terrestrial Animal Health Code and the Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. The DIE has regional Representations in Africa, America, Asia and the Pacific, Europe and the Middle-East and a large network of Collaborating Centers or Reference laboratories , including five for FMD .Their role is to function as a centre of expertise and standardisation of diagnostic techniques for FMD. Each laboratory has an expert, responsible to the DIE and its Member Countries for providing scientific and technical assistance and expert advice on FMD as well as providing reference reagents. The DIE has provided direct support to many Member Countries in its role as the co-ordinator of SEAFMD campaign and as an influential partner in the EU~MD control programme. Address: 12, rue de Prony, 75 017, Paris (France). Web-site: http://www.oie.int

2. The Food and Agriculture Organization of the United Nations (FAD) is responsible for agricultural development and food production. This specialised United Nations agency, created in Rome in 1947, is concerned with livestock development through its Animal Production and Health Division . The Animal Health Service of this division plays an important role in assisting Member Countries in the control of animal disease, with the objective of improving livestock production as an integral component of general social, economic and agricultural development. In 1954, the EUFMD (see text) was established under the auspices of the FAD to combat FMD in Europe.The role of the Commission is to co-ordinate the activities of its 33 Member States in controlling and preventing FMD. It provides support in FMD surveillance, and advice on all aspects of FMD prevention and control on request. Address: Viale delle Terme di Caracalla, 00 100 Rome (Italy). Web-site: http: //wwwjao.org 3.

The Pan American Health Organization (PAHO) is a regional Office of the World Health Organization, created in 1920, to organise the control of epidemics threatening the American continent. Among these epidemics, FMD has always been considered by the PAHO as an important disease for the Americas, from the economical point of view. It has, therefore, created in 1951 the Pan-American Foot and Mouth Disease Center (Panaftosa) in Rio de Janeiro (Brazil).This Center takes an active part in developing FMD vaccine production units and in promoting vaccine quality control in the region, as well as developing epidemiological research on FMD. Address: 525 T\venty-third Street, N.W. Washington, D.C. 20037-2895 (United States of America). Web-site: http://www.paho.org

Copyright © 2004 By Horizon Bioscience

From: Foot-and-Mouth Disease: Current Perspectives. Edited by: Francisco Sobrina and Esteban Domingo

Chapter 17 Overview of Foot-and-Mouth Disease and Its Impact as a Re-emergent Viral Infection

Brian W.J. Mahy

Abstract Foot- and-mouth disease (FMD) is endemic in many countries of the world, but a number of countries have attained FMD-free status, which provides economic benefits from international trade in animals and animal products. In recent years there have been serious outbreaks of FMD in countries formerly free of the disease. This chapter summarizes the world situation and addresses the question of whether resurgences of FMD fall into a general pattern of emerging and re-emerging diseases consequent upon four interlinked domains of the determinants of the emergence of infection as defined by the 2003 Institute of Medicine Report on Microbial Threats to Health: genetic and biological factors; physical environmental factors; ecological factors; and social, economic and political factors. 1. Introduction Foot-and-mouth is a disease of great antiquity, with written records dating back at least to 1546 with a description by a monk, Hieronymus Fracastorius, of an epidemic in cattle near Verona, Italy (Fracastorius, 1546; Casas Olascoaga et aI., 1999). The concept of emerging and re-emerging infectious diseases is, on the other hand, a relatively recent one that was elegantly framed in a report of the Institute of Medicine of the National Academy of Science, USA, in 1992 (Lederberg et aI., 1992). This chapter considers the impact of foot-and-mouth disease (FMD) as a re-emerging infection, defined in the Institute of Medicine report as an infectious disease which has newly appeared, or increased in incidence, within the past two decades. FMD is primarily a highly contagious infection of clovenhoofed domestic and wild animals, and is characterized by fever and vesicles with subsequent erosions in the mouth, nares, muzzle, feet or teats. Although the average mortality from FMD is only about 1%, the morbidity is 100% in infected animals and the transmissibility is extremely high. It is the severe effect of FMD morbidity on the productivity of livestock, especially cattle, which makes FMD such a serious disease. Cattle not only develop a fall in milk production from infection, but also experience difficulty walking because of the painful foot lesions, mastitis, breeding problems, panting associated with pituitary gland damage, and diabetes mellitus. The disease tends to be milder in pigs, sheep and goats, although abortions may occur, and these species act as reservoirs from which the disease may spread to cattle.

Copyright © 2004 By Horizon Bioscience

1438

FMDVas a Re-Emergent Infection

The virus is notorious for its ability both to survive in a fully infectious state for long periods of as much as several years in soil, and is also one of the most highly transmissible viruses known. Spread of the virus is usually associated with movement of infected animals and their contact with susceptible animals. In the early stages of the infection, animals shed virus in all of their secretions and in their breath. Infected pigs, in particular, can release enormous quantities of virus as an aerosol, and so spread the virus over considerable distances to susceptible sheep or cattle. Infection by the aerosol route is enhanced in cattle because of their large respiratory tidal volume. The virus can also be carried mechanically by veterinarians, motor vehicles and other fomites between infected and susceptible animals. This must always be borne in mind when introducing measures to control an outbreak of FMD.

2. Global Impact of FMD FMD has been present throughout most of the world for many years, though New Zealand has never experienced the disease, and Australia and Japan have been free of FMD for more than 70 years. In addition, Central American countries from Guatemala to Panama have escaped the disease, despite many epidemics raging in the USA and South America over many years. The USA experienced nine outbreaks between 1870 and 1929, the most serious of which, in 1914-1916, involved 22 States and the District of Columbia. FMD was introduced into South America through the importation of breeding stock from Europe in the mid-19th century, first in Argentina (1865) then Uruguay (1870), Chile (1871) and southern Brazil (1895). The disease then spread to the rest of Brazil, to Paraguay and Peru (1910), Bolivia (1912), then much later to Mexico (1946), Venezuela (1950), Colombia (1950), and Ecuador (1956). FMD is now considered to be enzootic in South America, Africa, Asia, and parts of Europe, especially Eastern Europe. 3. Eradication of FMD There are two means of eradicating FMD from a country: stamping-out by killing all diseased and potentially infected animals, and subsequent destruction of their carcasses followed by thorough cleaning and disinfection of the affected premises; and vaccination. The stamping-out policy has been law in the UK since 1892, and is applied with compensation to the farmers for loss of livestock, resulting in freedom of the UK livestock industry from FMD over many years. Serious epidemics occurring in 1922-24, 1967-68, and 2001 were all stamped out without vaccination. This method was also used by the USA in the last outbreak in 1929, and in Canada in 1951-52. In the South American countries, control of FMD has generally employed vaccination. The difficulty with vaccination is that only inactivated vaccines have so far been developed, which do not provide long-lasting immunity and ideally need to be administered every 6 months. In addition, the serotype used in the vaccine must match the prevailing epidemic strain, and since FMD undergoes rapid antigenic changes, producing many subtypes within anyone of the seven recognized serotypes, this may further reduce the effectiveness of the vaccine. Finally, incomplete inactivation of the virus present in the vaccine has occasionally

Copyright © 2004 By Horizon Bioscience

Mahv 4391

been demonstrated in the past, leading to spread rather than control of an epidemic (Barteling, 2002). From the point of view of trade with other countries, vaccination of livestock is viewed as an admission that FMD is present in the country. Such factors led the European Union in 1992 to abolish general vaccination of cattle against FMD, so creating a large susceptible cattle population, and a situation that was ripe for the 2001 FMD epidemic which started in the UK. In 1987, the South American countries signed a Hemispheric Plan for the Eradication of Foot-and-Mouth Disease, and during the 1990's, remarkable progress was made. For example, national laboratories diagnosed 786 cases of FMD per year in the region in 1990, but by the late 1990's this had fallen to only 130 cases of FMD in the whole of South America. By 1999, the international community recognized Argentina, Chile, Guyana and Uruguay as free from FMD without vaccination. However, in 2001 FMD reappeared in the southern region of South America forcing the reintroduction of cattle vaccination in Argentina, southern Brazil (Rio Grande do SuI), Paraguay and Uruguay (Correa Melo et al., 2002). In countries where FMD is enzootic, including some African and Asian countries, there is virtually no attempt to control the disease. Countries which are free of the disease may use vaccination of cattle along a border with an infected neighbouring country to provide a "buffer zone". The World Organization for Animal Health, administered on behalf of 162 countries by the Office International des Epizooties (OlE) in Paris, maintains a code to establish guarantees that countries importing livestock are entitled to demand of exporting countries to ensure protection against the introduction of the virus without creating unjustified sanitary barriers. Countries recognized as "FMD-free without vaccination" thus have a clear commercial advantage, since the absence of vaccination and outbreaks is the most practical criterion to ensure that a country is free of FMD. For this reason countries make every effort to guard their "FMD-free" status, since a single FMD outbreak may cost the country millions of dollars in lost livestock export trade. 4. Emergence of the PanAsia Strain Serotype 0 FMDV is common throughout Asia, and exists in many distinct evolutionary lineages, that have been termed "topotypes", (Samuel and Knowles, 2001). In 1999, a type 0 FMDV was detected on Kinmen, a Taiwanese island close to China, in cattle with no clinical disease, but from a herd that had in part already been moved to mainland Taiwan and sold for slaughter there. FMD was not recognized until January 2000, when it occurred in both cattle and goats, and many young goats died of the disease. This strain of FMDV type 0 was first isolated in 1990 from outbreaks of FMD in India, and is now generally referred to as the PanAsia strain because of it's wide spread involving many Asian countries (Hemadri et aI., 2002). In March 2000 this strain spread to Japan which had been free of FMD since 1908, and Korea which had been free of FMD since 1934 (Sakamoto and Yoshida, 2002) despite controls in place to prevent the spread of the disease. It was this PanAsia type 0 strain that also moved in a westerly direction and caused FMD outbreaks in the Middle East and a devastating epidemic in the UK Copyright © 2004 By Horizon Bioscience

1440

FMDVas a Re-Emergent Infection

in 2001 (Knowles and Samuel 2003). During 1994, this virus apparently moved into Saudi Arabia due to trade of live animals with India, and from there to Turkey, Greece and Bulgaria by 1996. Outbreaks of FMD due to this strain were then reported from several other countries of the Middle East, where it replaced all other strains circulating in those countries, and has persisted in Saudi Arabia. It is interesting that over a 12 year period, the viruses causing these outbreaks of FMD appear to belong to a single lineage based on sequence analysis of the VPl gene, since this region differs by no more than 5 percent between various isolates.

5. The 2001 UK Outbreak After an absence of thirty-three years, FMDV was confirmed in pigs in mainland UK on 20 February 2001 at an abattoir in Essex, and shown to be the PanAsia type o strain shortly thereafter (Knowles et al., 2001 b). It was subsequently learned that the virus had been introduced into pigs in a farm in Northumberland, 350 km north of Essex, some two to three weeks prior to its first detection. Although not proven, it is likely that the initial introduction of the virus into pigs occurred through the illegal feeding of untreated waste food from a restaurant (swill) containing meat originating in South-East Asia. Whatever the nature of the introduction, the virus most probably spread by the airborne route from pigs to sheep on neighbouring farms, and then to various parts of the country and also to Northern Ireland, Scotland, France and the Netherlands. Although an export ban on British livestock \vas declared on February 22 nd , and all movement of animals within the UK stopped on February 24 th , more than 70 premises had already become infected by that time. This spread of the disease was undoubtedly increased by the difficulty of recognizing the disease in sheep, which show few overt clinical symptoms, as well as by the very large trade in sheep which has developed in the UK subsequent to the epidemic of bovine spongiform encephalopathy (BSE). This has caused many farmers to tum to sheep rather than cattle for their livelihood. In summary, the large size of the UK FMD epidemic was mainly due to the failure to recognize the disease and stop all livestock movements at an earlier stage. Between March 25 th and April 22 nd 2001, some 2 million animals were slaughtered because of FMD, and by October 26 th the number had risen to nearly 4 million. Of these, more than 80 percent were sheep, 15 percent cattle, and 5 percent pigs and other animals such as goats. The costs of the outbreak have been estimated as 3.1 billion pounds sterling from losses to agriculture and the food chain, and between 2.7 and 3.2 billion pounds sterling from losses to businesses and tourism (Thompson et ai., 2002). On January 22 nd 2002, the UK regained its FMD-free status, making way for resumption of meat exports. 6. Emergence of Viruses With Altered Host Range In 1997, a virulent strain of FMDV type 0 emerged in Taiwan, an island off mainland China which had been free from FMD for 68 years. This resulted in an epizootic that destroyed much of the pig export industry: more than 6000 farms were infected and 3.8 million pigs were slaughtered (Yang et al., 1999). This virus is now endemic in large parts of China, and the Taiwanese epidemic was so severe that at first it was considered to have been deliberately brought into Taiwan as a possible act of bioterrorism. No evidence has been reported to support this notion, Copyright © 2004 By Horizon Bioscience

Mahv

4411

however, and it seems highly unlikely that this virus was deliberately introduced. The World Reference Laboratory for FMD in Pirbright, UK recognizes 8 topotypes (genetically and geographically distinct evolutionary lineages) of FMDV serotype 0, and the strain resposible for the Taiwan 1997 outbreak is termed the Cathay topotype. It was subsequently shown that this topotype has an altered 3A protein with a 10-amino-acid deletion that is probably responsible for its virulence in pigs, and lack of virulence for cattle (Knowles et al., 2001). 7. Foot-and Mouth Disease in Africa As with many other virus diseases, FMD is endemic in most countries within the African continent and six of the known FMDV serotypes (all except Asia 1) can be found there. Although the disease is largely controlled in South Africa, Botswana and Namibia by strict enforcement of livestock movement control and good surveillance to detect outbreaks, most other African countries lack the necessary resources and infrastructure to deal with the problem. An additional problem is caused by the reservoir of FMDV that exists in wildlife, particularly African buffalo (Condy et aI., 1985). Importation of viruses from elsewhere in the world, including the pan-Asian topotype of FMDV type 0 (Sangare et al., 2001), further complicates the epidemiological situation. In September 2000, the Pan-Asian type 0 virus caused an outbreak of FMD on a pig farm near Pietermaritzburg, Kwazulu-Natal Province, South Africa. This was the first FMD recorded in the province since 1956, and the first occurrence of FMDV type 0 in South Africa. It is highly likely that the virus was carried in food waste fed to pigs (swill) obtained illegally from a ship visiting Durban that had originated in South-East Asia. An outbreak of FMD occurred in Zimbabwe in 2001, caused by the SAT 2 serotype of FMDV. In this case it is likely that the source of infection was wild African buffalos, and spread through illegal movement of cattle by motor transport from the Matebeleland North area to Bulawayo, where the market is more lucrative. 8. Recent Outbreaks of Other FMD Types FMDV type SAT2 had never been seen in Saudi Arabia, so this type was not included in the routine vaccination of cattle. However in April 2000 SAT2 virus was isolated from a large dairy farm in Saudi Arabia, and more than 30 percent of the cattle were affected. The source of this outbreak is unknown. FMDV type Asia 1 was detected in northeastern Iran in September 1999 from where it spread rapidly through the country, and then to eastern Anatolia and Central Turkey and on to Georgia by June 2000. This represents a considerable extension of the range of this serotype. Despite a declaration by the DIE in August 2000 that Argentina was free of FMD without vaccination, epidemiologic surveillance in the border zones detected ten bovines that had been illegally imported in the district of Clorinda, Formosa Province, on 2 August 2000. Since there was no health documentation about these animals, they were slaughtered and destroyed after blood samples were taken. The samples revealed that four of the ten animals had antibody against FMDV nonstructural proteins, indicating virus replication despite an absence of overt clinical

Copyright © 2004 By Horizon Bioscience

1442

FMDVas

a Re-Emergent Infection

symptoms of FMD. A virus was isolated by probang sampling one of the animals. The virus was type A24, a virus never before detected in Argentina, indicating that the animals had been imported from another country which harbored this strain. In 2000 there were also outbreaks of FMD type 0 in Brazil in August and in Uruguay in October.

9. FMD as a Re-Emerging Disease The 1992 Institute of Medicine Report (Lederberg et ai., 1992) listed six principal factors which contribute to the emergence or re-emergence of an infectious disease. These are: 1. Microbial adaptation and change. 2. Human demographics and behaviour. 3. Breakdown of public health measures. 4. International travel and commerce. 5. Changes in technology and industry. 6. Economic development and land use. A new Institute of Medicine Report published in 2003 (Smolinski et al., 2003) has added a further seven factors, bringing the list to thirteen. The additional factors are: 7. Human susceptibility to infection. 8. Climate and weather. 9. Changing ecosystems. 10. Poverty and social inequality. 11. War and famine. 12. Lack of political will. 13. Intent to harm. The scope of this new Report, however, is limited to infectious diseases in humans. It organizes the contributory factors into four interlinked domains of the emergence of infection:. 1. Genetic and biological factors. 2. Physical environmental factors. 3. Ecological factors. 4. Social, political and economic factors. In considering FMD as a re-emergent infection it is useful to consider to what extent these factors apply to FMD in 2003. 9.1. Genetic and Biological Factors

Under this heading it is necessary to consider the importance of structural and functional genetic changes in FMDV in influencing the reemergence of FMD. FMDV is a highly mutable virus, and within anyone of the seven recognized relatively stable serotypes there are multiple subtypes that can be distinguished genetically within the virus population (Knowles and Samuel, 2003). The FMDV population displays high mutation rates and a quasispecies dynamics (Domingo et aI., 2003). There is no doubt that within this population, mutations occur which can result in viruses with altered host range or pathogenicity, and this was an important factor in the emergence of the Cathay topotype of FMDV type 0 in Taiwan in 1997. Once established in the livestock population after its first detection on 14 March,

Copyright © 2004 By Horizon Bioscience

Mahy 4431

this strain infected susceptible hosts on the entire island, measuring 380 by 140 km, within 50 days. More than 6000 farms became infected, 3.8 million pigs were slaughtered, and 13 million doses of vaccine were administered. We have also witnessed over the past decade the emergence of the PanAsian strain of type 0, which appears to have extraordinary ability to spread, and replace other existing subtypes in a region. For example, it replaced all other strains of virus once it entered the Middle East (Knowles and Samuel, 2003). Although the precise structural changes in the virion which provide such strains with unusual success are not yet known, there is no doubt that genetic change is one important factor in the reemergence of FMD. The emergence of the PanAsian type 0 strain has clearly been a major factor in the global impact of FMD. A major biological factor in FMD emergence is the persistence of FMDV in many species of wild animals, especially in Africa, which makes global eradication of the disease virtually impossible. Even in European countries, there are also susceptible species, such as deer and hedgehogs, which may be capable of harbouring the virus and potentially acting as a source of transmission of infection. 9.2. Physical Environmental Factors

Many elements of the physical environment playa role in the survival of viruses and may affect the transmission between susceptible hosts. Humid and cold weather seems to favor the air-borne dissemination of FMDV (Donaldson, 1972). The airborne spread of FMD over many miles has been documented, as occurred in 1981 when FMD spread from France to the Isle of Wight over water. However, there are considerable species differences in such transmission (Donaldson et aI., 2001) and it seems unlikely that airborne spread played a major role in the development of the UK epidemic in 2001. It is important to realise that individual virus strains exhibit different properties which can affect transmission, and this should be taken into account when attempting to control an epidemic (Sutmoller et aI., 2003). In addition, the use of high pressure sprays for cleaning FMD -infected premises and holding facilities may generate large amounts of infectious aerosols, and could be an important factor in virus dissemination. 9.3. Ecological Factors

Although FMDV transmission is not primarily dependent upon ecological factors, except possibly in South Africa, the physical environment is constantly being modified by human activities. The human population increase in the last decade has been enormous, and soon the world population will reach 7 billion. Simply serving the nutritional needs of such a large human population has encouraged ne\v intensive livestock farming, including an increase in livestock trade. This has resulted in a great deal of movement of animals, and the spread of the UK FMD epidemic in 2001 was undoubtedly increased as a direct result of this trade. 9.4. Social, Political and Economic Factors

International trade in food and animal agriculture has increased markedly as a result of globalization. Several of the recent outbreaks of FMD can be directly attributed to such movements. For example, the original introductions of FMDV

Copyright © 2004 By Horizon Bioscience

1444

FMDVas a Re-Emergent Infection

type 0 (PanAsia strain) in South Africa in 2000 and in the UK in 2001, both appeared to have originated with pig swill that included meat from South-East Asia. Although not certain, it seems highly likely that in both these cases the importation was illegal. An important economic factor that undoubtedly played a role in the UK FMD epidemic in 2001 was the increase in sheep farming consequent upon the disastrous BSE epidemic that has made beef production unprofitable for many farmers. Sheep infected with FMDV are much more difficult to recognize as compared to cattle, and it was the spread due to movement of infected sheep over several weeks that greatly increased the magnitude of the epidemic before it was first recognized. Another important social consequence of the "stamping-out" policy during the UK epidemic in 2001 was the adverse effects of culling livestock on the nonagricultural community. Larger numbers of animals were slaughtered as a means of disease control than in any other epidemic over the last 40 years, and the funeral pyres of thousands of animals were seen on nightly television as well as by the rural community. Movement restrictions were imposed for more than ten months not only upon susceptible livestock, but also on persons walking in the countryside for recreational purposes.This undoubtedly had a damaging effect upon rural businesses as well as tourism, contributing to the enormous economic cost of the epidemic (Taylor, 2002).

10. Conclusions It is clear from a survey of FMD outbreaks during the last two decades that there is a continuing threat to the livestock industry from this disease. The adverse economic effects of FMD associated with the worldwide pandemic of the PanAsian strain of FMDV type 0 have been severe, and have provided a wake-up call to all those involved in the business of agriculture, as well as politicians, that there is no room for complacency. Effective surveillance and control measures must be maintained on a worldwide basis to respond rapidly to future outbreaks. FMD will remain a serious threat to agriculture for the foreseeable future, despite the hopes of some who have claimed that it can be eradicated using vaccination and similar strategies to those now being successfully employed in the rinderpest eradication campaign led by the Food and Agriculture Organization (FAO). It is clear from recent events, summarized above, that FMD eradication poses far more complex problems than does rinderpest. One of the reasons for success in rinderpest eradication is that the disease is caused by a single virus type, against which a live virus vaccine is extremely effective, providing lasting immunity. FMD, on the other hand, is caused by seven different serotypes and multiple subtypes of a virus with exceptional ability to persist, to mutate, and to spread amongst livestock, and only killed virus vaccines may be used which provide immunity for only a few months. Before seriously embarking on an eradication programme, better vaccines providing long-lived immunity need to be developed, and the role of wildlife in the maintenance of FMD, especially in Africa, needs to be thoroughly investigated. Such developments would require a major investment to fund much-needed research, but in the longer term the benefits to agriculture world-wide would be

Copyright © 2004 By Horizon Bioscience

Mahv 4451

considerable. If this investment is not made, FMDV will persist and continue to re-emerge to cause devastating outbreaks with serious social and political consequences in the future. References Barteling, S.J. 2002. Development and performance of inactivated vaccines against foot and mouth disease. Rev. Sci. Tech. Off. Int. Epiz. 21: 577-588. Casas Olascoaga, R., Gomes, I., Rosenberg, F.J., Auge de Mello, P., Astudillo, V. and Magallanes, N., 1999. Fiebre Aftosa. Book i-xv. Sao Paulo: Editora Atheneu. p. 1458. Domingo, E., Escarmis, C., Baranowski, E., Ruiz-Jarabo, C.M., Carrillo, E. Nunez, J.I., and Sobrino, F. 2003. Evolution of foot-and-mouth disease virus. Virus Research. 91: 47 - 63. Donaldson, A.1. 1972. The influence of relative humidity on the aerosol stability of different strains of foot-and-mouth disease virus suspended in saliva. J. Gen. Virol. 15: 25 -33. Donaldson, A.1. 1986. Aerobiology of foot-and-mouth disease (FMD): An outline of recent advances. Rev. Sci. Tech. Off. Int. Epiz. 5: 315 -321. Donaldson A.I., Alexandersen S., Sorensen J.H. and Mikkelsen T. 2001. Relative risk of the uncontrollable (airborne) spread of FMD by different species. Vet. Rec. 148: 602-604. Donaldson A.1. and Alexandersen, S. 2002. Predicting the spread of foot and mouth disease by airborne virus. Rev. sci. tech. Off. int. Epiz. 21: 569 - 575. Fracastorius, H. 1546. De contagione et contagiosis morbis et curatione. BK. 1, Chap 12 (Venencia). Hemadri, D., Tosh, C. and Venkataramanan, R. 2002. Emergence of a new strain of type 0 foot and mouth disease virus: its phylogenetic and evolutionary relationship with the PanAsia pandemic strain. Virus Genes 25: 23-34. Knowles, N.J., Davies, P.R., Henry, T., O'Donnell, V., Pacheco, J.M., Mason, P.W. 2001. Emergence in Asia of foot-and-mouth disease viruses with altered host range: characterization of alterations in the 3A protein. J. Virol. 75: 1551-1556. Knowles, N.J. and Samuel, A.R. 2003. Molecular epidemiology of foot-and-mouth disease virus. Virus Res. 91: 65-80. Knowles, N.J., Samuel, A.R., Davies, P.R., Kitching, R.P. and Donaldson, A.1. 2001. Outbreak of foot-and-mouth disease virus serotype 0 in the UK caused by a pandemic strain. Vet. Rec. 148: 258-259. Lederberg J., Shope, R.E. and Oaks, S.C. eds. 1992. Emerging Infections: Microbial Threats to Health in the United States. National Academy Press, Washington. 312 pp. Samuel, A.R. and Knowles, N.J. 2001. Foot-and-mouth disease type 0 viruses exhibit genetically and geographically distinct evolutionary lineages (topotypes). J. Gen. Virol. 82: 609 -621. Sakamoto, K and Yoshida K. 2002. Recent outbreaks of foot-and-mouth disease in countries of East Asia. Rev. Sci. Tech. Off. Int. Epiz. 21: 459-463. Smolinski, M.S. Hamburg M.A. and Lederberg J. eds. 2003. Microbial Threats to Health. Emergence, Detection, and Response. The National Academies Press, Washington. 396pp. Sutmoller, P., Barteling, 5.S., Casas Olascoaga, R., and Sumption, K.J. 2003. Control and eradication offoot-and-mouth disease. Virus Res. 91: 101 -144. Taylor, K.C. 2002. Environmental impact of the foot and mouth disease outbreak in Great Britain in 2001: The use of risk analysis to manage the risks in the countryside. Rev. Sci. Tech. Off. Int. Epiz. 21: 797-813. Thompson, D., Muriel, P., Russell, D., Osborne, P., Bromley, A., Rowland, M., Creigh-Tyte, S., and Brown, C. 2002. Economic costs of the foot and mouth disease outbreak in the United Kingdom in 2001. Rev. Sci. Tech. Off. Int. Epiz. 21: 675 -687. Copyright © 2004 By Horizon Bioscience

1446

FMDV as a Re-Emergent Infection

Yang, P.C., Chu, R.M., Chung, W.B., and Sung, H.T. 1999. Epidemiological characteristics and financial costs of the 1997 foot-and-mouth disease epidemic in Taiwan. Vet. Rec. 145: 731 - 734.

Copyright © 2004 By Horizon Bioscience

(a)

(b)

(c)

N

Uy

N

(d) Chapter 4: Figure 1. (a) Disposition of the structural proteins in (left) SBMV (Abad-Zapatero et at., 1980) and (right) FMDV with respect to the icosahedral symmetry axes (in yellow). SBMV, with a T=3 organization, contains three protein subunits, named as A Band C, in the icosahedral asymmetric unit. Relationships between these structurally independent subunits are defined by the quasi two-fold and three-fold axes (in white), according to the quasi-equivalence principles (Caspar el at .. , 1962). FMDV, as all the picornaviruses, presents a T=I organization of its structural proteins VPI-VP4, though the similar organization and topology of VPI-VP3 (labels correspond to the biological protomer) with respect to SBMV proteins A Band C makes it appropriated to talk about a pseudo T=3 (P=3) virus architecture. (b) Structural proteins related by the quasi-three fold axes are shown for SBMV (left) and FMDV (right). (c) Ribbon representation of the B subunit in SBMV (left) and of the structural proteins in FMDV (right): VPI (blue), VP2 and VP4 (green and dark green, respectively) and VP3 (red). (d) Topology of the wedge-shaped eight-stranded [3-barrel fold commonly found in RNA vimses (see text).

Copyright © 2004 Horizon Bioscience

A.2

Foot-and-Mouth Disease: Current Pers ectives

(a)

Chaptel' 4: Figure 3. (a) Solid representation of pentamers from HRV 14 (left) and FMDV (right). The canyon running around the icosahedral five-fold axes in HRV 14 (the circle cOlTesponds approximately to the deepest part of the canyon) is in part filled by the VPl Cterminal residues (dark blue) in FMDV. The hole seen around the five-fold axes in FMDV is absent in HRV14 and in most picornaviruses. (b) Disposition of the structural protein VP2 around the icosahedral three-fold axis in HRV14 (left) and FMDV (right). The tight annulus found in FMDV could be stabilized by ionic interactions mediated by Ca 2+ ions situated directly on the three-fold axis (see in the text).

(b)

(a)

(b)

Copyright © 2004 Horizon Bioscience

Chapter 4: Figure S. (a) Experimentally determined structure of the FMDVoligosaccharide receptor complex (Fry et al., 1999). Oligosaccharides are represented as balls and sticks while the virus is displayed in blue either as solid spheres (on the left) or as ribbons (on the right). (b) Modeled structure of the FMDV integrin av133 complex obtained (Verdaguer et al., 2004) by docking the integrin structure bound to the RGD tripeptide (Xiong et aI., 2002) onto the structure of the FMDV virus serotype C with its GH-Ioop almost perpendicular to the viral surface (the one shown in Figure 4b) (Hewat et aI., 1997). The integrin is represented as ribbons (yellow) with the RGD peptide explicitly shown with balls and sticks. Views, on the left side of the figures, correspond to pentamers of the complexes seen down the S-fold axes. On the right side of the figures the viral protomel's (in blue) are seen laterally together with the receptor molecule bound to it

Appendix

A.31

(a)

(b)

(c) Chapter 4: Figure 4. The VPI GH loop, or "FMDV loop", (in yellow) with (a) the down conformation found in the structure of the reduced 0 serotype (Logan et ai., 1993) and (b) a fully exposed disposition obtained in the structure of complexes of the C-serotype with the neutralizing antibody SD6 (Hewat et ai., 1997). The hinge flexibility of the "FMDV loop" explains the difficulties in observing the loop in the unperturbed viruses. On the left, views are along the 5-fold axes, while a lateral view of the protomer is shown on the right. Protein subunits are colored as in Figure I. Despite the hinge movement the "FMDV loop" presents an intemal structure (c) that appears well conserved among different serotypes. In this structure the RGD sequence is extended and situated occupying a turn prior to a helix with the arginine and aspartic residues pointing in opposite directions.

Copyright © 2004 Horizon Bioscience

):>

~

~

o

d;" ::l 9~

o SO

r::

CJ

1/;" (l)

6Jo :h

i§:'" tJ

in'

(l)

Ql

en ~

()

c:: "1

~

~

(;l "0 (l)

C1

~

en

Chapter 5: Figure 3. FMDV lesions in pigs. A) Pig with a large, unruptured vesicle above the rostrum (PRIT). B) Pig's feet with early, unruptured vesicles (blanched areas). C) Pig's feet with vesicles along the coronary bands of the main and supernumerary digits. D) Pig's foot with loss of horn tissue (PRIT).

Copyright © 2004 Horizon Bioscience

),.